Special issue

Aging Management of Concrete Structures in Nuclear Power Plants

Edited by Ippei Maruyama

Effect of radiation on concrete

Review of the current state of knowledge on the effects of radiation on concrete
Thomas M. Rosseel, Ippei Maruyama, Yann Le Pape, Osamu Kontani, Alain B. Giorla, Igor Remec, James J. Wall, Madhumita
Sircar, Carmen Andrade, Manuel Ordonez 368-383

Combined effects of temperature and irradiation on concrete damage

Yann Le Pape, Alain Giorla, Julien Sanahuja 70-86

A numerical model for concrete strength change under neutron and gamma-ray irradiation
Ippei Maruyama, Kazumoto Haba, Osam Sato, Shunsuke Ishikawa, Osamu Kontani, Masayuki Takizawa 144-162

Change in relative density of natural rock minerals due to electron irradiation

Ippei Maruyama, Shunsuke Muto 706-716

Aging management of nuclear power plant

Multi-scale review of possible Mechanisms for the natural frequency change of reinforced concrete
structures under an ordinary drying condition
Ippei Maruyama 691-705

Outline of the guidelines for maintenance and management of structures in nuclear facilities
Yoshito Umeki, Shohei Sawada, Shiro Mitsugi, Toshinobu Maenaka, Katsuki Takiguchi 643-663

Effective aging management of NPP concrete structures

Julia Tcherner, Lingam Vaithilingam, Ming Han 1-9

Maintenance management of turbine generator foundation affected by alkali-silica reaction

Ryosuke Manabe, Hiroshi Kawae, Kiyoshi Ogawa, Masayuki Matsuura 590-606

Maintenance of a prestressed concrete containment vessels in a nuclear power plant

Yoshihiro Yamaguchi, Takashi Kitagawa, Hideki Tanaka, Yasumichi Koshiro, Hideo Takahashi, Shinichi Takezaki,
Masazumi Nakao 464-474

Technologies for the aging management

The strengthening of an ASR-affected water intake tower in a hydro-electric dam by using post-tensioned
tendons and the long-term monitoring of the tower
Kazuyuki Torii, Tetsuji Kubo, Chikao Sannoh, Maki Kanitani 384-396

Numerical strategy for forecasting the leakage rate of Inner containments in double-wall nuclear reactor
buildings
Mehdi Asali, Bruno Capra, Jacky Mazars, Jean-Baptiste Colliat 408-420

Residual static strength of concrete cylinder specimen and stud anchor specimen after cyclic loadings
Osamu Kontani, Naoki Ishitobi, Junya Kawada, Noriaki Taogoshi, Makoto Koge, Yoshito Umeki 634-642

Accelerated aging of concrete dry cask storage systems for nuclear waste
Arezou Attar, Bora Gencturk, Mohammad Hanifehzadeh, Kaspar Willam 299-310

Review of research on the high temperature resistance of concrete structures in Chinese NPP
Jianzhuang Xiao, Wengang Xie, Qinghai Xie 335-343

Experimental study on hydraulic and macro-mechanical property of a mortar under heating and cooling
treatment
Zhinan Lin, Weiya Xu, Wei Wang, Jiuchang Zhang, Huanlin Wang, Rubin Wang 261-270

Effect of mineral admixture and fibers on crack of sacrificial concrete

Zuquan Jin, Xia Zhao, Tiejun Zhao, Yuanchao Cao 502-510

Click to Submit your Papers

Preface

Even after the accident of the Fukushima Daiichi Nuclear Power Plant, the Japanese government has
positioned nuclear power as an important base load power source in terms of energy policies and has
focused on safety operation of the nuclear power generation. This is because it is an indispensable power
source for the use of renewable energy in Japan with limited resources. On the other hand, the first
generation commercial reactors in the world generally have a designed life of 30 to 40 years, but most of
them have been in existence for more than 20 years. Japan is no exception, and several NPPs aged already
more than 40 years.
Analysis and research show that many commercial reactors leave enough capacity to operate beyond the
designed service period. Replacement of the concrete members used in the reactor buildings is expensive
making it difficult in many cases. Therefore, in order to continue nuclear power generation over the long
term, it is necessary to maintain the performance of the concrete member during the designed service
period. From this point of view, it is indispensable to conduct research on maintenance management,
performance evaluation, and prediction of deterioration of concrete structures.
Research on concrete in the field of civil engineering and building science has been conducted for a long
time, while reevaluation of concrete performance, physical properties and various problems in terms of
nuclear field needs, i.e. Plant Life Management and Ageing Management, and addressing issues peculiar
to the nuclear power field will be an important contribution from the field of concrete research to the field
of nuclear power engineering. As discussed at OECD / NEA, alkali silica reactions and radiation influence
are important issues for the maintenance of nuclear power plants, and many research on them has been
reported.
Also in the past, research investment on concrete in the field of nuclear power has been carried out many
times in Japan as well as in other developed countries. However, the international publication of the research
results has been limited. In this ACT special issue, especially on advanced research in recent years, I widely
urged stakeholders to make efforts to allow research trends in Japan to be forecast from other countries.
Based on these backgrounds, we applied for papers focusing on Plant Life Management at nuclear power
plant facilities and related technologies in the research related to concrete. I think that it became a State of
the Art report, which highlights the current problems. I also anticipate it could lead to a trust from the society
concerning future nuclear power generation technology and a development of the concrete research field.

December 19, 2016

Ippei Maruyama
Professor, Nagoya University
Journal of Advanced Concrete Technology Vol. 14, 368-383 July 2016 / Copyright © 2016 Japan Concrete Institute 368

Invited paper

Review of the Current State of Knowledge on the Effects of Radiation on

Concrete1)
Thomas M. Rosseel1*, Ippei Maruyama2*, Yann. Le Pape1*, Osamu Kontani3, Alain B. Giorla4,
Igor Remec4, James J. Wall5, Madhumita Sircar6, Carmen Andrade7 and Manuel Ordonez8

Received 18 May 2016, accepted 29 June 2016 doi:10.3151/jact.14.368

Abstract
A review of the current state of knowledge on the effects of radiation on concrete in nuclear power production applica-
tions is presented. Emphasis is placed on the effects of radiation damage, as reflected by changes in engineering proper-
ties of concrete, in the evaluation of the long-term operation and for plant life or aging management of nuclear power
plants (NPPs) in Japan, Spain, and the United States. National issues and concerns are described for Japan and the
United States followed by a discussion of the fundamental understanding of the effects of radiation on concrete. Spe-
cifically, the effects of temperature, moisture content, and irradiation on ordinary Portland cement paste and the role of
temperature and neutron energy spectra on radiation-induced volumetric expansion (RIVE) of aggregate-forming min-
erals are described. This is followed by a discussion of the bounding conditions for extended operation; the significance
of accelerated irradiation conditions; the role of temperature and creep; and how these issues are being incorporated into
numerical and meso-scale models. From these insights on radiation damage, analyses of these effects on concrete struc-
tures are reviewed, and the current status of work in Japan and the United States is described. Also discussed is the re-
cent formation of a new international scientific and technical organization, the International Committee on Irradiated
Concrete, to provide a forum for timely information exchanges among organizations pursuing the identification, quanti-
fication, and modeling of the effects of radiation on concrete in commercial nuclear applications. The paper concludes
with a discussion of research gaps, including (1) interpreting test-reactor data, (2) evaluating service-irradiated concrete
for aging management and to inform radiation damage models with the Zorita NPP (Spain) serving as the first compre-
hensive test case, (3) irradiated-assisted alkali-silica reactions, and (4) RIVE under constrained conditions.

1. Introduction completed. Moreover, although much of the focus has

As light water reactors (LWRs) that became operational
in the 1960s and 1970s have aged and operational effi-
ciencies have improved, national regulators, nuclear
power plant (NPP) owners, and researchers have fo-
cused on enhanced aging management (examination,
inspection, maintenance, and testing of key compo-
nents) and, for some countries, lifetime extension to
meet future national energy needs while reducing
greenhouse gas emissions. To achieve these goals, a
critical evaluation of the knowledge gaps of materials
that comprise NPP structures and components must be
1
Oak Ridge National Laboratory, Oak Ridge, USA.
*Corresponding author, E-mail: rosseeltm@ornl.gov
& lepapeym@ornl.gov
2
Nagoya University, Nagoya, Japan. *Corresponding
author, E-mail: ippei@dali.nuac.nagoya-u.ac.jp
3
Kajima Corporation, Tokyo, Japan.
4
Oak Ridge National Laboratory, Oak Ridge, USA.
5
Electric Power Research Institute, Charlotte, USA.
6 the published form of this manuscript, or allow others to
US Nuclear Regulatory Commission, Washington, DC,
USA.
7
Instituto Eduardo Torroja de Ciencias de la
Construcción, Madrid, Spain.
8
ENRESA, Madrid, Spain.
been on the performance and possible degradation
mechanisms of metals due to increased exposure time to
temperature, stress, coolant, and radiation fields, other
materials such as concrete and cables also are critical to
plant life management (PLM) and long-term operation
(LTO) (Rosseel et al. 2015b). This report focuses on
research aimed at reducing the knowledge gap of the
effects of radiation on concrete as identified by Graves
et al. (2014).
Until recently, the data on the mechanical degradation
of concrete due to radiation were thought to have been
quite limited. Moreover, in 2010, Kontani et al. (2010)
recognized that reference levels based on data reported
in Hilsdorf et al. (1978) and used to establish reference
a
1)
This manuscript has been authored by UT-Battelle,
LLC, under Contract No. DE-AC0500OR22725 with
the U.S. Department of Energy. The United States
Government retains and the publisher, by accepting the
article for publication, acknowledges that the United
States Government retains a non-exclusive, paid-up,
irrevocable, world-wide license to publish or reproduce
do so, for the United States Government purposes. The
Department of Energy will provide public access to
these results of federally sponsored research in
accordance with the DOE Public Access Plan
(http://energy.gov/downloads/doe-public-access-plan).
T. M. Rosseel, I. Maruyama, Y. Le Pape et al. / Journal of Advanced Concrete Technology Vol. 14, 368-383, 2016
levels for concrete integrity might not be suitable be-
cause data were not representative of concrete mixtures,
temperatures, and radiation fields typically seen in
LWRs. Furthermore, a comprehensive review and re-
analysis of the literature by Field et al. (2015) has
greatly expanded the database and confirmed the pre-
dominant role of radiation-induced volumetric expan-
sion (RIVE) in the degradation of irradiated concrete..
In this report, a review of the current state of knowl-
edge on the effects of radiation on concrete in nuclear
applications is presented. Emphasis is placed on the
effects of radiation damage as reflected by changes in
engineering properties of concrete for PLM and the
evaluation of the long-term operation of NPPs in Japan,
Spain, and the United States.
Following a discussion of issues and concerns in Ja-
pan and the United States, this report focuses on the
current status of the fundamental understanding of ra-
diation effects on concrete. It begins with the effects of
radiation on hardened cement paste (HCP), including
drying and temperature effects and radiolysis, and RIVE
of aggregate-forming minerals. This is followed by a
discussion of the combined effects of temperature, dry-
ing, and radiation on concrete, including the signifi-
cance of accelerated irradiation conditions, the role of
temperature and creep, as well as variations in energy
spectra and damage energy, and how these factors are
being incorporated into numerical and mesoscale mod-
els of concrete structures.
From these insights on radiation damage and the
bounding fluences at extended lifetimes, analyses of
these effects on concrete structures are reviewed, and
the current status of work in Japan and the United States
is described. Furthermore, the recent formation of a
new international scientific and technical organization
[the International Committee on Irradiated Concrete
(ICIC)] to provide a forum for timely information ex-
changes among organizations pursuing the identification,
quantification, and modeling of the effects of radiation
on concrete in commercial nuclear applications is de-
scribed.
The paper concludes with a discussion of current re-
search gaps, including (1) interpreting test-reactor data,
(2) evaluating service-irradiated concrete for aging
management and to inform radiation damage models
with the Zorita NPP (Spain) serving as the first compre-
hensive test case, (3) irradiated-assisted alkali-silica
reactions, and (4) RIVE under constrained conditions.
Japanese conditions, issues and concerns:
LWRs have been identified as an important component
of electric power generation as established by extensive
discussions following the Fukushima Daiichi nuclear
accident (METI 2014). Ensuring the safe, long-term
operation of NPPs is necessary for the public and for
electric power suppliers, particularly because construc-
tion of new NPPs is difficult in Japan. For these reasons,
aging management of concrete structures that cannot be
369
replaced is crucial for long-term operation of NPPs.
Before the Fukushima Daiichi accident, the Japanese
government’s PLM program inspected NPPs to evaluate
their integrity after an operation period of 30 years and
every 10 years thereafter. This process required data on
the integrity of structures under present and future con-
ditions (Japan Nuclear Energy Safety Organization
2013; Architectural Institute Japan 2015). To evaluate
concrete structures in PLM, two measures of deteriora-
tion were chosen: reduction in concrete strength and
reduction in shielding performance. The following fac-
tors arising from the environmental conditions and ma-
terials used in reactor buildings adversely affect the
concrete strength: elevated temperature, irradiation,
carbonation, the alkali-silica reaction (ASR), and ma-
chine vibration.
After the Fukushima Daiichi nuclear accident, new
nuclear reactor regulations came into effect on 8 July
2013, and an approval system for extending the opera-
tion period of NPPs was introduced by the Nuclear
Regulatory Authority of Japan (NRA). Generally, a 40-
year operation period is allowed for every NPP although
this period can only be extended to a maximum of 20
years after the facility passes a special inspection.
In PLM and the inspection for extending the opera-
tion period, the integrity of concrete affected by radia-
tion is evaluated with reference levels taken from data
reported in Hilsdorf et al. (1978) and interpreted by
Kontani et al. (2010): 1 × 1020 n/cm2 for fast neutrons
and 2 × 105 kGy for gamma rays. However, in a national
project organized by the Japan Nuclear Energy Safety
Organization (partly a former organization in the NRA),
the Hilsdorf compiled data were found to be unsuitable
for this purpose because they did not cover the type of
concrete used in LWRs (Kontani et al. 2010). A national
project has been launched to elucidate the effects of
radiation on concrete properties and to develop a system
for evaluating the integrity of concrete affected by ra-
diation (Maruyama et al. 2012).
US conditions, issues, and concerns:
As described in Rosseel et al. (2014), extending the
operating lifetimes of current NPPs beyond 60 years and
making additional improvements in their productivity is
essential to meeting future United States national energy
needs while reducing greenhouse gas emissions. To
meet these goals, a critical evaluation of the knowledge
gaps of the effects of exposure time to temperature,
stress, coolant, and radiation fields on materials that
comprise the structures and components of an NPP is
required. And although it is expected that the vast ma-
jority of concrete structures will continue to meet their
functional and performance requirements during future
licensing periods, there may be isolated examples where
structures may not exhibit the desired durability, primar-
ily due to environmental effects such as radiation.
To address these extended lifetimes effects, the Elec-
tric Power Research Institute (EPRI), through the Long-
T. M. Rosseel, I. Maruyama, Y. Le Pape et al. / Journal of Advanced Concrete Technology Vol. 14, 368-383, 2016
Term Operations (LTO) Program, and Oak Ridge Na-
tional Laboratory (ORNL), through the US Department
of Energy (DOE) Light Water Reactor Sustainability
(LWRS) Program, established a research plan to inves-
tigate the aging and degradation processes associated
with NPP concrete. The basis for the plan arose in part
from the Expanded Materials Degradation Analysis
(EMDA) report on the aging of concrete, an effort
jointly supported by DOE and the US Nuclear Regula-
tory Commission (NRC), and performed by expert pan-
els from US national laboratories, industry, academia,
and international organizations. A key finding of the
EMDA was the urgent need to develop a consistent
knowledge base on irradiation effects in concrete
(Graves et al. 2014). Although not part of the
EPRI/LWRS plan, the NRC has initiated confirmatory
research for the developing the technical basis for the
regulatory framework of subsequent license renewal
(see for example Section 6.2).
The LWRS program is focusing on (1) assessing the
radiation environment in concrete biological shields
(CBSs), defining the upper bound of the neutron and
gamma dose levels expected at extended operation, and
estimating adsorbed dose; (2) harvesting and testing
service-irradiated concrete to validate models; (3) irra-
diating prototypical concrete and its components under
accelerated ion, neutron and gamma dose levels to es-
tablish upper bounds to inform radiation damage mod-
els; (4) developing improved models to enhance the
understanding of the effects of radiation on concrete;
and (5) establishing an international collaborative re-
search and information exchange to leverage capabili-
ties and knowledge including developing cooperative
test programs to improve confidence in data obtained
from various concretes and from accelerated irradiation
experiments (Rosseel et al. 2015a). EPRI is now focus-
ing on the structural analyses of select US NPPs with
CBS that perform structural support and whose esti-
mated fluence at ~ 80 years of operation approaches the
bounding fluence of 6 × 1019 n/cm2 at E >0.1 MeV (Sec-
tion 4.1).
2. Fundamental understanding of radiation
effects on concrete
2.1 Introduction
Concrete is by nature and constitution, a composite ma-
terial made of HCP and aggregate. Water is present in
the HCP under various forms (i.e., chemically bound,
adsorbed and free water). Gamma and neutron irradia-
tion interact differently with the concrete compounds
[See Table 3 in Kontani et al. (2010)]. In particular, the
poorly crystallized structure of the calcium-silicate-
hydrates (C-S-H) and the HCP porosity makes them less
susceptible to neutron-induced lattice defect accumula-
tion and distortion(Kontani et al. 2010) as opposed to
well-crystallized aggregate-forming minerals. Concur-
rently, gamma-induced radiolysis occurs predominantly
370
where water is present (i.e., in the HCP). Because irra-
diation results in energy-deposition-induced heating,
temperature and moisture transport must be considered
concurrently in the study of irradiation effects. The fol-
lowing sections provide a summary of the state of
knowledge of the effects of temperature, moisture con-
tent and irradiation on ordinary Portland cement paste
and aggregate.
2.2. Hardened cement hydrates
2.2.1 Drying and temperature effects:
The shrinkage and strength properties of HCP affect the
physical properties of concrete. The aggregates restrain
the shrinkage of cement paste during drying or heating
(Carlson 1938; Hansen and Nielsen 1965; Hobbs 1974),
and the large difference in volume change, due to envi-
ronmental conditions, between the aggregates and the
cement paste matrix produces cracks in the cement paste
that act as pores. These cracks affect the strength and
Young’s modulus of the concrete (Bažant et al. 1982;
Son and Hosoda 2010; Maruyama et al. 2014b; Lin et al.
2015). Therefore, the differential shrinkage is the main
factor in the monotonic reduction of the Young’s
modulus of the concrete (Maruyama et al. 2014b).
The strength of the mortar and concrete decreases
from high relative humidity (RH) to intermediate RH
(40-50% RH at room temperature) and increases below
the intermediate RH range under drying or heating con-
ditions (Pihlajavaara 1974; Maruyama et al. 2014b).
This behavior is attributed to the change in strength of
the HCP, and the cracks due to drying are important in
reducing the compressive strength above 60% RH
through limiting the load bearing path in concrete (Ma-
ruyama et al. 2014b). This complex behavior of strength
of HCP is explained by the colloidal nature of C-S-H
(Jesser 1927; Powers and Brownyard 1946; Tomes et al.
1957; Thomas and Jennings 2006; Jennings 2008). Dur-
ing the initial drying process, the C-S-H particles form-
ing colloidal C-S-H agglomerate, producing large pores
in the cement paste. The increase in the volume of large
pores reduces the paste strength. In contrast, drying
strengthens the C-S-H by tightening the layered struc-
ture. The increase in the strength of the solid structure
overcomes the reduction of strength caused by large
pores, resulting in the strength increase of hardened
cement paste at lower RHs (Maruyama et al. 2014a).
2.2.2 Radiolysis effects:
Gamma rays affect material through electronic excita-
tions. Metastable calcium peroxide octahydrates can be
produced in HCP with high water content (Bouniol and
Aspart 1998; Vodák et al. 2005; Lowinska-Kluge and
Piszora 2008), altering the strength and pore size distri-
bution and, especially, resulting in subsequent carbona-
tion. It is also suggested that gamma radiation-induced
carbonation increases the strength of HCP (Vodák et al.
2005; Vodák et al. 2011). Underlying these phenomena
is the radiolysis of water. Radiolysis is the physico-
T. M. Rosseel, I. Maruyama, Y. Le Pape et al. / Journal of Advanced Concrete Technology Vol. 14, 368-383, 2016
chemical consequence of ionizations along the trajec-
tory of radiation. Several groups, including Bouniol and
Aspart (1998), Bouniol and Bjergbakke (2008), and
Kontani et al. (2013), have studied the radiolysis of pore
water in cement.
For example, Kontani et al. (2013) performed
gamma-ray irradiation experiments on cement paste and
collected all the gases from the samples by argon career
gas. By analyzing time-dependent gas composition
changes during gamma irradiation, they found the fol-
lowing phenomena of gas production in cement paste:
2)
• The hydrogen production G value is linear with
gamma-ray dose rate (1~10 kGy/h), and this relation
is not affected by the sample temperature (20~60ºC).
• The linearity between gamma-ray dose rate and hy-
drogen production G value implies a validation of ac-
celerated gamma-ray irradiation experiments with
gamma-ray dose rates less than10 KGy/h.
• Only a few percent of the chemically bound water
decomposed even when subjected to the reference
level of 2 × 105 kGy, since the G values were very
low at around 0.033). This means that hydration prod-
ucts can maintain their integrity.
• The hydrogen generation rate decreased as the evap-
orable water content in the sample decreased.
Further scientific research is necessary to validate the
integrity of concrete and find a possible degradation
mechanism by radiolysis in concrete.
2.2.3 Interaction between creep and irradiation:
Creep of the cement paste is a key aspect of the me-
chanical behavior of concrete, especially considering
the time scales involved in the LTO of NPPs. On the one
hand, creep may affect the response of the structure it-
self, for example, by relaxing the pre-stress in rebars
(e.g., Magura et al. 1963), and on the other hand, it
strongly affects crack propagation in concrete (Rusch
1960; Bazant and Gettu 1992). Gray (1971) measured
an increase in the creep and shrinkage of a cement grout
specimen under neutron irradiation even though the
various changes of hygro-thermal conditions during the
experiments, and the lack of control specimen, make
this specific experiment difficult to interpret.
McDowall (1971) found that under gamma irradiation,
creep of concrete was reduced, while autogeneous
shrinkage was increased. This is consistent with creep
experiments at low relative humidity (Whittmann 1970),
and therefore might be a manifestation of the radiolysis
in the cement paste. Understanding how creep is af-
fected by radiation is critical because creep controls the
long-term crack propagation by relaxation of the
2)
The G value refers to the number of specified chemical
events produced in an irradiated substance per 100 eV
of energy absorbed.
3)
This value is estimated on an assumption that only
hydroxyl groups in the cement paste sample absorb the
gamma-ray energy.
371
stresses in the cement paste [e.g., Altoubat and Lange
(2001) for shrinkage cracking, Giorla, Scrivener et al.
(2015) for ASR].
2.3. Aggregate-forming minerals
2.3.1 Radiation-induced volumetric expansion:
With about 70% of the volume fraction of concrete con-
sisting of aggregate, understanding aggregate-forming
mineral sensitivity to radiation is of the utmost impor-
tance. Alpha decay of certain radionuclides is known to
produce a metamict state in rock-forming minerals (Ew-
ing et al. 1988). Ion-beam irradiation of minerals gener-
ally results in crystalline-to-amorphization (CA) transi-
tion (Wang et al. 1991; Eby 1992; Douillard and
Duraud 1996; Harbsmeier and Boise 1998) and, in some
cases, in the precipitation of oxides (Meldrum et al.
1997) or the formation of cavities (Templier et al. 1997).
Amorphization is predominantly observed in silicate-
based minerals. For example, while quartz presents
strong long-order radiation-induced disordering, irradia-
tion of calcite causes primarily the rotation of the car-
bonate groups (Pignatelli et al. 2016). The critical CA
dose in silicates depends on several factors: (1) the
melting temperature; (2) a structural factor (e.g., degree
of SiO4 polymerization, elastic properties); and (3) the
proportion of Si-O bonding (Eby et al. 1992). Amorphi-
zation results in changes in physical, optical and me-
chanical properties (e.g., Wittels 1957; Primak 1958;
Mayer and Lecomte 1960; Wong 1974). Initial anisot-
ropic crystalline elastic properties are gradually modi-
fied with the irradiation dose to converge toward an
isotropic stiffness tensor (Mayer and Lecomte 1960;
Zubov and Ivanov 1967). Amorphization also induces
significant density change, especially in silicates. For
example, the maximum volumetric expansion of quartz
and feldspars has been shown to be as large as ~18%
(Primak 1958) and ~8% (Krivokoneva 1976), respec-
tively, while the change of density in calcite remains
rather low (~0.3% according to Wong). Depending on
the mineralogical content, considerable variations in
aggregate RIVE have been observed as described in the
comprehensive review by Field et al. (2015). Moreover,
some observed post-irradiation expansions exceed what
is considered as detrimental by ASR researchers (e.g.,
Fournier and Berube 2000; Rajabipour et al. 2015).
2.3.2 Role of temperature in radiation damage:
Increased irradiation temperature delays the onset of the
critical dose for amorphization and reduces the expan-
sion kinetics of quartz (Bykov et al. 1981) and silicates
such as feldspars (Krivokoneva 1976) and pyroxenes,
yet it does not seem to affect the total RIVE at the end
of the amorphization process. The effect of temperature
on quartz expansion is explained by the annealing of
point defects during irradiation occurring at an early and
intermediate stage of amorphization. However, no sig-
nificant annealing or anti-annealing (Primak 1958, Yano
et al. 2007) of RIVE is expected from temperature
T. M. Rosseel, I. Maruyama, Y. Le Pape et al. / Journal of Advanced Concrete Technology Vol. 14, 368-383, 2016
variations occurring in the test reactors and LWRs (Le
Pape et al. 2016).
2.3.3. Energy spectra and damage energies:
As previously noted, a review of existing data on irradi-
ated concrete reveals that neutron fluence is predomi-
nately used to characterize the radiation environment
(Hilsdorf et al. 1978; Field et al. 2015), but there is little
consistency in the neutron energy cutoff applied for the
neutron fluence (Remec et al. 2016b). However, in the
CBS of pressurized water reactors (PWRs), the neutron
fluence for a cutoff energy of E >0.1 MeV is about 8-15
times higher than the neutron fluence at E >1 MeV (and
the neutron fluence for E >0 eV can be 20-40 times
higher) (Field et al. 2015). Using consistent and appro-
priate energy cutoff is therefore crucial for the assess-
ment of concrete degradation.
RIVE is currently considered the dominant cause for
degradation of irradiated concrete (Field et al. 2015). It
is, therefore, of interest to investigate neutron-induced
atom displacements in aggregates. Remec et al. (2016b)
showed that, for the neutron spectra in the biological
shield of two-loop and three-loop (the number of reactor
cooling circuits) PWRs, neutrons with energies above
0.1 MeV contribute more than 95% of all atom dis-
placements in several widespread aggregate minerals,
while neutrons with energies above 1 MeV cause only
about 20 to 25% of the total atom displacements, as
shown in Fig. 1 (Remec et al. 2016a). The conclusions
were similar when, instead of atom displacements, neu-
tron-induced absorbed dose in minerals was considered.
These observations suggest that neutron fluence with a
cutoff of 0.1 MeV is the preferred parameter for radia-
tion field characterization in irradiation experiments and
for accessing radiation-induced concrete degradation.
Work is in progress to develop a unified irradiation
parameter that will account for neutron and gamma-ray
contributions to concrete degradation. In the typical
radiation field of the PWR biological shield, the contri-
Fig. 1 Relative contribution of neutrons with energies E > E0 to the dpa (displacements per atom) rate for the neutron
spectrum in the cavity of a three-loop PWR, for several selected minerals.
372
bution of gamma rays to the total absorbed dose is simi-
lar and may exceed the neutron-induced absorbed dose
(Remec et al. 2013). However, preliminary results indi-
cate that the gamma-ray contribution to the atom dis-
placements does not exceed a few percent of the neu-
tron-induced displacements and will therefore be insig-
nificant in most cases. In accelerated irradiation experi-
ments, the conditions may be considerably different, and
contributions of gamma rays need to be carefully evalu-
ated for each specific experiment.
3. Irradiated concrete
3.1 Test reactor data
The essential reference used to systematize the analysis
on irradiated concrete structural properties is Hilsdorf et
al. (1978). Later, Fujiwara et al. (2009) limited their
analysis to ordinary Portland structural concrete and
found no data subject to any significant effect of neutron
irradiation on concrete for fluence less than 1 × 1019
n/cm2. More recently, Field et al. (2015) revisited his-
torical test reactor irradiation data and collected 307
compression strength data, 62 tensile strength data, 138
elastic modulus data and 114 linear expansion data. The
data suggest that concrete compressive and tensile
strengths as well as the Young’s modulus decrease
above fluences around 1 × 1019 n/cm2. This threshold
fluence value, however, is subject to variations with the
concrete composition and, possibly, the exposure condi-
tions. A strong correlation with the radiation-induced
(and thermal expansion) volumetric expansion of aggre-
gate and concrete is observed, suggesting that RIVE is a
primary degradation factor (Seeberger and Hilsdorf
1982). However, Field et al. (2015) noted important
uncertainties limiting the interpretation of the gathered
data, particularly for the fluence levels of interest (i.e.,
above 1 × 1019 n/cm2): (1) The absence of energy cutoff
normalization (as described in Section 2.3.3); (2) the
absence of comparable mechanical testing procedures
T. M. Rosseel, I. Maruyama, Y. Le Pape et al. / Journal of Advanced Concrete Technology Vol. 14, 368-383, 2016 373

(age, geometry and size of specimens); (3) the effects of to the previous numerical calculation results of concrete
conditioning (moisture content) before and during irra- shrinkage (Maruyama and Sugie 2014). Based on an
diation are often not reported; and (4) most data corre- input of mortar shrinkage, damage in concrete produced
spond to testing temperatures above 100°C leading to by the restraint of coarse aggregate was evaluated ex-
enhanced damage in the HPC and at the interface with plicitly. Owing to the modeling of ITZ, shrinkage of
aggregate. Finally, although reliable data on the effects concrete under different drying conditions can be repro-
of gamma irradiation on concrete mechanical properties duced by the proposed model. After drying, specimens
are scarce (Alexander 1963; Kelly et al. 1969; Gray were loaded and strength changes calculated. The re-
1971), they seem to suggest limited impact. Further re- sults are shown as the change in the relative compres-
search is needed. sive strength (Fc/Fco) in Fig. 3. From the calculation, it
A correct interpretation of irradiation experiments re- is elucidated that compressive strength of cylinder con-
quires a physics-based mechanistic understanding and crete is determined by both damage in concrete due to
modeling of the temperature, moisture content and irra- the volume change difference between mortar and
diation effects in the concrete compounds. The next coarse aggregate and strength change of cement paste
section provides some insights regarding the recent pro- (in this calculation, it is strength mortar) due to colloidal
gress made using mesoscale models of irradiated con- alteration under heating and drying.
crete. A preliminary simulation to evaluate the impact of
aggregate expansion due to radiation was conducted. In
3.2 Numerical models the preliminary simulation, expansion of coarse aggre-
Mesoscale model: gate in addition to the presented model discussed above
The impact of the expansion of aggregates due to radia- cannot simulate the expansion of concrete and strength
tion and the resultant damage observed as changes in reduction due to damage in concrete. In the case of ag-
physical properties can be understood through numeri- gregate expansion, impact of damage in concrete is
cal calculations. This approach has two objectives. The overestimated, and further study is required of the rela-
first is to find a mechanism that describes changes in tionship of damage in concrete in large expansion to the
physical properties of concrete components due to heat physical properties of concrete.
and drying, as well as radiation. This mechanism will A 2D mesoscale model for irradiated concrete (Giorla,
contribute to the interpretation of experimental data and Vaitová et al. 2015) was developed at ORNL based on
to formulating a universal conclusion. The second is to
extrapolate the phenomena from a laboratory specimen
size to a structural size that cannot be obtained experi-
mentally.
In the mesoscale, or aggregate-size scale, it is impor-
tant to consider the role of aggregate explicitly in a nu-
merical calculation. For this objective, the finite-
element method (FEM) and rigid body spring model
(RBSM) are applied to the evaluation of concrete under
drying, heating and radiation.
Originally developed by Kawai (1978), RBSM has
been applied extensively for structural analysis. RBSM
deals with crack propagation of concrete directly since it
represents a continuum material as an assembly of rigid
particle elements interconnected by zero-size springs (a) Voronoi mesh (b) Aggregate mesh
along their boundaries. Being nonlinear, these zero-size Fig. 2. RBSM mesh (Ogawa and Maruyama 2016).
springs can simulate the cracking behavior of a contin-
uum material. The nonlinearity and discrete behavior of
the continuum material is emulated by cracks develop-
ing at the interfaces of the rigid particles. For this reason,
crack patterns and the resultant nonlinear behavior of
the target model are significantly affected when a mesh
design is employed. To solve this problem, random ge-
ometry using Voronoi diagrams was applied (Bolander
and Saito 1998).
Recently, RBSM has been applied to simulate volume
change and compressive strength of concrete concur-
rently (Ogawa and Maruyama 2016). Coarse aggregate
is explicitly modeled in the mesh (Fig. 2), and the inter-
facial transition zone (ITZ) role is considered according Fig. 3. Comparison of calculated Fc/Fco with experimental
data (Ogawa and Maruyama 2016).
T. M. Rosseel, I. Maruyama, Y. Le Pape et al. / Journal of Advanced Concrete Technology Vol. 14, 368-383, 2016
the AMIE finite element framework (Dunant and Scriv-
ener 2010, Giorla et al. 2014). The model aim is to pro-
vide the concrete expansion and internal degradation as
a function of the aggregate RIVE and the mechanical
properties of the paste and the aggregates, notably ac-
counting for the coupling between creep and damage in
the cement paste (Giorla, Le Pape and Dunant 2016).
The model was validated in an irradiation experiment
with serpentine concrete (Elleuch et al. 1972) since it is
one of the very few experiments in the literature in
which the properties of the concrete, the aggregates, and
the cement paste were independently characterized (see
Fig. 4). Under variable temperature and flux, the model
predicts a higher damage and expansion than under a
constant temperature and flux (Le Pape et al. 2016). The
model also shows an increase and orientation of internal
cracking under external restraint, which is similar to the
effect observed for ASR (Giorla, Le Pape, and Dunant
2016).
Even though these models (RBSM and FEM) are
based on different representations of the material (nota-
bly in terms of material continuity and mechanical be-
havior), they both greatly overestimate the damage in
the cement paste. This effect might be caused by the fact
that aggregates are represented as homogeneous parti-
cles that swell uniformly, and could therefore be re-
duced with a more accurate description of the aggregate
mineral microstructure. Accounting for different RIVE
within the aggregate might cause damage to initiate
within the aggregates rather than at their surface, as it is
for example the case in ASR (Ben Haha et al 2007;
Dunant and Scrivener 2010). Doing so requires an ex-
perimental investigation and characterization of the ag-
gregates themselves, as well as a well-informed data-
base of minerals and their properties.
4. Significance of irradiation for concrete
structures in NPP
4.1 Bounding fluence at extended lifetimes and
radiation transport in concrete
As previously discussed, (Remec et al. 2016b; Rosseel
et al. 2014; Rosseel et al. 2015b), the onset of radiation-
induced concrete degradation has been correlated to
neutron fluence and gamma ray dose based on data as-
sembled by Hilsdorf et al. (1978). Obtaining accurate
estimates of the bounding fluence and dose values for
the US NPP fleet at the projected extended service is
clearly important. Due to variations in plant operations
and differences in reactor designs, however, this effort is
not simple. Fortunately, because operating NPPs are
required to implement a reactor pressure vessel (RPV)
surveillance program, a wealth of fluence information is
available. In the United States, the information is avail-
able from the US NRC Agency-wide Documents Access
and Management System (ADAMS) database (US NRC
2015).
The critical first step is the neutron and gamma ray
374
transport calculations of the attenuation through the
pressure vessel and CBS (Remec et al. 2013). An ex-
ample of the attenuation for a Japanese PWR is shown
in Fig. 5.
Using the transport calculations as guidance, a multi-
step process was implemented by Esselman and Bruck
(2013) to determine the expected neutron fluence values
on the CBS for the current US boiling water reactor
(BWR) and PWR fleet for up to 80 years of operation
assuming 92% operating efficiency. The results shown
in Fig. 6 highlight the importance of plant design and
fuel loading scheme in determining the maximum ex-
pected fluence on the surface of the CBS at 80 years of
operation. For example, 2- and 3-loop PWR reactors
show significantly higher fluence values than the 4-loop
plants. It should be noted that the BWR fluence at 80
years of operation is generally lower than 4-loop PWR
fluence.
4.2 Structural analysis
Summary of concrete structural analysis litera-
ture:
The primary function of the CBS is to provide radio-
logical protection from the radiation emitted from the
nuclear steam supply system (NSSS), in particular, the
RPV. In some PWR designs, the structural function of
the CBS is to transfer the static or seismic load of the
reactor to the foundation system and to mitigate the ef-
fects of accident scenarios. The CBS is generally made
of conventional [i.e., normal weight aggregate (Essel-
man and Bruck 2013)] carbon steel reinforced concrete
(Hookham 1991). Some designs also include a steel
plate liner.
Limited irradiated concrete structural analyses have
been published in the open literature using different
models: analytical axisymmetric models (Andreev and
Kapliy 2014; Le Pape 2015), finite-element membrane
models (Mirhosseini et al. 2014), finite-element contin-
Fig. 4 Comparison of calculated concrete RIVE with ex-
perimental data from Elleuch et al (1972), and example
of simulated damage pattern at a neutron fluence of
19 2
1.3×10 n/cm ). The gray area represents the scatter in
the experimental data. Figure adapted from Giorla, Vai-
tova, et al (2015).
T. M. Rosseel, I. Maruyama, Y. Le Pape et al. / Journal of Advanced Concrete Technology Vol. 14, 368-383, 2016 375

Fig. 5 Example of Japanese PWR radiation field developed from neutron transport calculations (Fukuya et al. 2002). Dr.
O. Sato (Mitsubishi Research Institute Inc.) and Dr. T. Igari (MRI Research Associates Inc.) confirmed this calculation.

Individual US PWRs
Fig. 6 Summary of US PWR fleet fluence at the outer diameter of the RPV extrapolated to 80 years (E >0.1 MeV) and
grouped by design (heat transfer loops). Graph reinterpreted from Esselman et al. (2013).

uum models (no RIVE) (Pomaro et al. 2011), and finite-
element mesoscale models (Salomoni et al. 2014; Giorla,
Le Pape, and Huang 2016).
In the absence of RIVE (or swelling), radiation dam-
age [i.e., induced loss of mechanical properties (Hils-
dorf et al. 1978; Field et al. 2015)] has very limited ef-
fects on the CBS (Pomaro et al. 2011). Moreover, the
radiation field’s strong attenuation produces a high
RIVE gradient (Le Pape 2015) causing high biaxial
compressive elastic stresses in the vertical and hoop
directions near the reactor cavity and important tensile
hoop stresses toward the back of the CBS (Andreev and
Kapliy 2014; Le Pape 2015). Simultaneously, the pro-
longed moderate temperature exposure (<65°C by de-
sign) and strong internal moisture content gradient (Ox-
fall 2013) affect the degree of hydration of concrete, and
thus, its mechanical properties. In particular, it leads to
the development of lower strengths toward the reactor
cavity (Maruyama and Igarashi 2015). The dissipation
of the RIVE-induced elastic stresses implies a potential
relaxation in the cement paste, but, more significantly,
the development of cracking (Giorla, Le Pape, and
Dunant 2016). The extent of radiation-induced concrete
damage appears to be limited to a depth of about <20
cm (Le Pape 2015; Giorla, Le Pape, and Huang 2016).
The consequences of this damage to the structural per-
formance under seismic conditions (e.g., impact on the
RPV supports, or accident conditions; e.g., sudden in-
crease of temperature in the cavity, or seismic events)
still remain to be investigated.
Ongoing structural analysis work in the United
States:
Structural analyses currently being performed by US
industry are motivated by recent research and develop-
ment gap identification and prioritization reviews for
LWR concrete structures (Graves et al. 2014; Wall et al.
2012) and, more recently, the US NRC Draft Standard
Review Plan for Review of Subsequent License Re-
newal Applications for Nuclear Power Plants (NRC
2015). The US industry approach, as described in Sec-
tion 4.1, has been to first determine the predicted neu-
T. M. Rosseel, I. Maruyama, Y. Le Pape et al. / Journal of Advanced Concrete Technology Vol. 14, 368-383, 2016
tron fluence in the concrete biological shields for each
NPP in the US fleet for neutrons of E >0.1 MeV ex-
trapolated to 80 calendar years of operation (Esselman
and Bruck 2013). This was done using reactor-specific
surveillance data reported to the NRC, information that
is publically available in the NRC ADAMS database
(US NRC 2015). The bounding (maximum) neutron
fluence was found to be ~6 × 1019 n/cm2, which is
higher than the threshold for damage in concrete, as
described previously. To determine the effects of
chronic neutron irradiation on the structural margin in
biological shields that also perform a structural support
function, a 3D wedge FEM is being developed. This
model is based on the Westinghouse Type 1 PWR vessel
support configuration (Lapay et al. 2000). A schematic
of a typical PWR reactor vessel and biological shield is
shown in Fig. 7 for reference. The finite element mesh
was developed from a set of design drawings obtained
from a recently shut down 2-loop PWR. In this model,
the biological shield in the vicinity of a hot or cold leg
feed-through, under which the reactor vessel support sits,
and the embedded steel rebar geometry are meshed
separately and merged. The biological shield concrete
elements were segregated into annular layers to allow
variation of mechanical properties and variations in
RIVE as a function of depth into the concrete (Fig. 8).
The model will be run to failure under live and seismic
loading conditions to estimate the loss of structural
margin and changes in the internal stress morphology
due to chronic in-service neutron irradiation.
Ongoing structural analysis work in Japan:
Attempts to evaluate the impact of radiation-induced
degradation by numerical simulation are being made in
Japan. As previously discussed in Section 2.3.3, there is
a distribution of fluence in a real-scale member under
irradiation conditions because neutrons are always at-
tenuated by interaction with concrete. Consequently, the
Fig. 7 A typical PWR biological shield configuration
(shown in yellow) with the reactor vessel in the center.
376
radiation impacts show a distribution corresponding to
the fluence variation. To evaluate the time-dependent
spatial distribution of concrete properly, DEVICE
(Damage Evaluation for Irradiated Concrete) a numeri-
cal code, which takes into account the heat, moisture,
and radiation transport coupled with cement hydration,
was proposed by Maruyama et al. (2016). This code is
composed of the established computational cement-
based material model (CCBM) (Maruyama et al. 2015)
and the one-dimensional deterministic transport SN
code ANISN (Engle 1967). DEVICE predicts the prop-
erty of concrete and creates the input for further struc-
tural analysis. The rigid-body spring network, which
was originally proposed by Kawai (1978), was applied
for the structural analysis, and the impact of expansion
of the aggregate and the resultant change of seismic
performance was investigated. The preliminary result,
which is shown in Fig. 9, indicates that there is a risk of
Fig. 8 A finite element mesh of the biological shield in the
vicinity of a reactor vessel structural support showing
element layers for varying concrete properties and swell-
ing. The numbers represent annular layers of elements
used to vary properties and swelling as a function of
radial depth into the biological shield. For reference, the
total radial thickness of the shield is approximately 6 ft
(1.8 m).
Fig. 9 The expected cracks in concrete under irradiation
conditions. The fast neutron value (>0.1 MeV) at the
19
surface of the member is 6.0 × 10 n/cm .
2
T. M. Rosseel, I. Maruyama, Y. Le Pape et al. / Journal of Advanced Concrete Technology Vol. 14, 368-383, 2016
Table 1 Exposure conditions in irradiation testing facilities and in-service PWRs. (Remec et al. 2013), (Maruyama et
(c) (d) (e)
al. 2013), (Field et al. 2015), (Gray 1971), (Elleuch et al. 1972), (Dubrovskii 1967), (Kontani et al. 2010),
(i) (j)
(Seeberger and Hilsdorf 1982), (McDowall 1971), (Vodák et al. 2011), (Łowińska-Kluge and Piszora 2008), (Kelly
(m) (n)
et al. 1969), (Kitsutaka and Matsuzawa 2010), (Oxfall 2013).
PWR
Fast neutron flux (n/cm2/s) 1 to 2 × 1010 (ID) (a)
Fast neutron fluence (n/cm2) <6 × 1019 (ID) at 80 yr (c)
Gamma flux (kGy/h) 5 to 20 (a,g)
Gamma dose (MGy) 50 to 200 at 80 yr
Temperature (oC) <65 (design)
Relative Humidity (-) Strong gradient
~0.5 at 50 mm, ~0.7 at 150 mm, >0.9 at 400 mm at 30 yr (n)
spalling due to radiation-induced expansion of the ag-
gregate.
5. International forum for collaborations:
The International Committee on Irradiated
Concrete (ICIC)
Understanding the effects of radiation on concrete is
important for PLM and in determining long-term or
extended operating performance of concrete structures
in existing NPPs. Not surprisingly, this issue is being
addressed by research organizations and utilities across
the globe. In the last three years, researchers have been
actively working to build international partnerships and
collaborations in an effort to better define the issues,
develop a sound approach to resolving the major ques-
tions, and maximize resources. The result is the crea-
tion of the ICIC (Rosseel et al. 2015b).
The purpose of the ICIC is four-fold. First, it provides
a forum for timely information exchange among organi-
zations pursuing the identification, quantification, and
modeling of the effects of radiation on concrete in
commercial nuclear applications (specifically, determi-
nation of the effects of radiation damage as reflected by
changes in engineering properties). Second, it promotes
the broad application of highly specialized or unique
investigation techniques to relevant research materials
of wide interest to develop improved understanding and
predictions of the effects of radiation on concrete, ce-
ment composites, and constituents. Third, it facilitates
the conception, planning, and guidance of new coopera-
tive research programs, for example, facilitating the
collection (harvesting) and distribution of pedigreed
materials suitable for concrete irradiation damage inves-
tigations. And fourth, it promotes cross-institutional use
of resources, where possible, that reflects mutual re-
search interests or needs. Additional details may be
found at the ICIC web page:
http://web.ornl.gov/sci/psd/mst/ICICFGM/index.shtml.
6. Discussion of research gaps
6.1 Interpretation of test reactor data
Without an examination of service-irradiated LWR con-
crete, the current understanding of the effects of radia-
377
(a) (b)
(f) (g) (h)
(k) (l)
Test reactor/Gamma facility
5 × 1011 to 2 × 1014 (b)
<1020 (d,e,f)
0.02 to 200 (h,i,j,k,l,m)
0.04 to 1.3
40 to >250 (d,e)
tion on concrete is solely based on test reactor data (i.e.,
radiation under accelerated conditions). Table 1 pro-
vides a detailed comparison of the differences in radia-
tion flux and dose (E >0.1 MeV), temperature, and rela-
tive humidity. Accelerated testing raises questions of
possible rate-effect mechanisms and, consequently, the
significance of test reactor data. Interestingly, it is not
clear that accelerated testing leads to higher damage for
a given dose. For example, higher temperatures may
increase the damage in the HCP and at its interface with
an aggregate, while the RIVE rate is reduced, thereby
potentially reducing the damage in the HCP.
Another example is the effect of low/fast-RIVE on
the stress relaxation in polycrystalline minerals, aggre-
gate and HCP, which has received little attention so far.
In particular, the effects of gamma and neutron irradia-
tion on the HCP creep properties are not clearly under-
stood (McDowall 1971; Gray 1971).
Thorough fundamental understanding of both sepa-
rated and combined effects of irradiation, temperature
and internal moisture on aggregate and HCP is still
needed and should be incorporated into physics-based
modeling to derive irradiation consequences on LWRs.
6.2 Evaluating service irradiated concrete for
PLM and to inform/Confirm radiation damage
models: Harvesting zorita concrete
Harvesting of concrete cores from decommissioned
NPPs may provide not only an opportunity to generate
data from concrete that has experienced typical radia-
tion fields, but it may also provide guidance to acceler-
ated irradiation studies. Moreover, the coupling of ac-
celerated or laboratory-irradiated concrete with har-
vested cores from NPP is expected to facilitate the effort
to develop an understanding of the damage mechanisms
in irradiated concrete, including understanding potential
rate effects of accelerated irradiation studies. These re-
sults may also be used to confirm models of radiation-
induced degradation at extended lifetime such as the
ones proposed previously and for aging management
programs.
The Jose Cabrera nuclear power station or “Zorita
NPP” consists of a single-loop, Westinghouse PWR
reactor with 160-MWe output. It was commissioned in
1968 and operated until 2006, accumulating 26.5 effec-
tive full power years of operation. In February 2010,
T. M. Rosseel, I. Maruyama, Y. Le Pape et al. / Journal of Advanced Concrete Technology Vol. 14, 368-383, 2016
Empresa Nacional de Residuos Radiactivos, S.A (EN-
RESA), the Spanish national company for nuclear waste,
took responsibility for decommissioning the station. In
2009, the Consejo de Seguridad Nuclear (CSN), the
Spanish Nuclear Safety Council, convened a meeting
with national organizations to evaluate the feasibility of
obtaining concrete from Zorita to research the effects of
high radiation (neutron and gamma) and temperature on
concrete structures. Similar to the Zorita Internals Re-
search Project (ZIRP), Zorita’s concrete structures could
be of great value to developing an improved understand-
ing of the effects of radiation on aging concrete because
the neutron fluence on the Zorita CBS was approaching
the onset for possible concrete deterioration as esti-
mated by the Hilsdorf curve (Hilsdorf et al. 1978). In
January 2010, the CEIDEN (Spanish Strategic Platform
on Nuclear Research and Development) established a
working group on Zorita’s concrete cores. The current
partners of the project are ENRESA, ENDESA (utility),
IBERDROLA (utility), Gas Natural Fenosa (utility),
licensee during operation of Zorita NPP, IETcc (Instituto
Eduardo Torroja de Ciencias de la Construcción), CSN,
and recently EPRI.
The structures being considered are the CBS (radia-
tion and temperature effects), the spent fuel pool/trans-
fer channel (boric acid degradation) and the contain-
ment building (as reference). The building also will be
inspected through nondestructive testing of the steel
liner under the concrete slab. The irradiated concrete
tasks include: calculation of neutron and gamma fluence
and temperature by numerical modeling; obtaining test
specimens for chemical, mechanical, and microstruc-
tural characterization of irradiated and non-irradiated
concrete. Specifically, core samples from the CBS will
be drilled from areas of high irradiation, high tempera-
ture, and a combination of high irradiation and tempera-
ture. From each core sample, several test coupons will
be obtained: the closest to the reactor vessel, one from
the opposite side, and some intermediate coupons to
verify the attenuation effects.
On behalf of the NRC, ORNL is reviewing informa-
tion and reports provided by the Zorita Concrete Work-
ing Group and ENRESA to evaluate the concrete and
conditions to determine whether it would serve as a
suitable surrogate for US PWR NPPs. To determine the
applicability to long-term aging degradation of US
NPPs, three issues are being evaluated. The first issue
is whether the concrete, and especially the aggregate,
are the same or similar to concrete and aggregate found
in the US NPPs. A careful study of concrete cores har-
vested from any decommissioned NPP would provide
an opportunity to generate data from concrete that has
experienced typical radiation fields and, therefore,
would facilitate the effort to develop an understanding
of the damage mechanisms in irradiated concrete. If the
concrete and aggregate were not similar to concrete and
aggregate found in some US NPPs, questions would
remain as to its general applicability to extended opera-
378
tion of US NPPs. The second issue is whether the accu-
mulated fluence is sufficient to create observable dam-
age in the concrete and aggregate. Although the Zorita
CBS has accumulated the highest radiation of any reac-
tor currently scheduled for decommissioning, the flu-
ence must reach the threshold level required to produce
microstructural changes in the cement paste and the
onset of RIVE in the aggregates. The third issue is the
type of aggregate. Research has shown that siliceous
aggregates exhibit higher swelling (damage) than calcif-
erous aggregates (Field et al. 2015; Rosseel et al.
2015b). The greater the ratio of siliceous to limestone
aggregates, the larger the observable effects of radiation
on concrete. It is estimated that the evaluation will be
completed in late 2016.
6.3 Irradiation-assisted alkali-silica reaction
Ion-implantation-induced disorder (amorphization) of
the three-dimensional framework silicates, α-quartz
(Ichikawa and Koizumi 2002; Pignatelli et al. 2016) and
plagioclase (Ichikawa and Kimura 2007) results in an
increase by several orders of magnitude of the dissolu-
tion rate in contact with a highly alkaline solution close
to that of unirradiated amorphous silica (Pignatelli et al.
2016). Irradiated-amorphous silica also shows, to a
lesser extent, a greater susceptibility to dissolution
(Ichikawa and Koizumi 2002). Conversely, irradiation-
induced density change in calcite is essentially caused
by rotations and distortions of the carbonate groups with
respect to the Ca atom positions leading to minor
change of the dissolution rate (Pignatelli et al. 2016).
The increasing reactivity of irradiated aggregate is one
factor driving the formation of alkali-silica reaction gels.
Two other aspects must be considered before any defini-
tive conclusion can be made. ASR gels are highly hy-
drophilic but only internal moisture resulting from mix
excess water is available for their formation, which
could rapidly become a limiting factor. The stability of
alkali-silica hydrates under gamma irradiation (radioly-
sis) is unknown and requires further research.
6.4 RIVE in restrained conditions
Although no literature data were found on the effect of
mechanical restraint on the development of RIVE, it is
likely that restrained RIVE is subject to the same effects
as alkali-silica reaction (ASR) in concrete due to the
similarity of both mechanisms (i.e., expansion of a spe-
cific phase contained in the aggregates). ASR expansion
is reduced in the direction of the load and increased in
the lateral directions (although the total volumetric ex-
pansion is not constant with the load level), while dam-
age is oriented preferably in the direction of the load
and increases with the load level (e.g., Larive 1997;
Multon and Toulemonde 2006; Dunant and Scrivener
2012). Preliminary mesoscale simulations suggest that
RIVE is indeed subject to such anisotropy of expansion
and degradation when confined by external restraints
(Giorla, Le Pape and Dunant 2016) although experimen-
T. M. Rosseel, I. Maruyama, Y. Le Pape et al. / Journal of Advanced Concrete Technology Vol. 14, 368-383, 2016
tal verification is needed to conclusively determine this
effect.
7. Conclusions
In most countries with operating NPPs, LWRs have
been identified as an important component of electric
power generation. For that reason, ensuring the safe,
long-term operation of NPPs is critical to maintaining a
reliable energy supply. Moreover, extending the operat-
ing lifetimes of current NPPs as allowed by national
regulations, and making additional improvements in
productivity using PLM or aging management is essen-
tial to meeting future national energy needs while reduc-
ing greenhouse gas emissions. To meet these goals, a
critical evaluation of the knowledge gaps of the effects
of exposure time to temperature, stress, coolant, and
radiation fields on materials such as concrete is required.
And although it is expected that the vast majority of
concrete structures will continue to meet their functional
and performance requirements during licensing periods,
there may be isolated examples where structures may
not exhibit the desired durability, primarily due to envi-
ronmental effects such as radiation (Rosseel et al. 2014;
Rosseel et al. 2015a).
In this paper, a review of the current state of knowl-
edge on the effects of radiation on concrete in nuclear
applications is presented. Emphasis is placed on the
effects of radiation damage as reflected by changes in
engineering properties of concrete in the evaluation of
the LTO of NPPs and for PLM in Japan, Spain, and the
United States.
Following a discussion of issues and concerns in Ja-
pan (post Fukushima Daiichi) and the United States
(LTO and anticipated subsequent license renewal), this
report focuses on the current status of the fundamental
understanding of radiation effects on concrete and its
components. Specifically, it addresses drying and tem-
perature effects and radiolysis on HCP and RIVE of
aggregate-forming minerals. This is followed by a dis-
cussion of the combined effects of temperature, drying,
and radiation on concrete, including the significance of
accelerated irradiation conditions, the role of tempera-
ture and creep, as well as variations in energy spectra
and damage energy, and how these factors are being
incorporated into numerical and mesoscale models of
concrete structures. Using these insights, attempts to
evaluate the impact of radiation-induced degradation by
numerical simulations of concrete structures are re-
viewed, and the current status of work in Japan and the
United States described.
The recent formation of a new international scientific
and technical organization (the ICIC) to provide a forum
for timely information exchanges among organizations
pursuing the identification, quantification, and modeling
of the effects of radiation on concrete in commercial
nuclear applications is also described. The paper con-
cludes with a discussion of current, high-priority re-
379
search gaps including evaluating service-irradiated con-
crete from decommissioned NPPs, such as Zorita
(Spain), to inform degradation models and aging man-
agement programs, evaluation of test reactor data, irra-
diated-assisted alkali-silica reactions, and RIVE under
constrained conditions.
Acknowledgments
The authors gratefully thank the following sponsors: US
DOE, Office of Nuclear Energy, Light Water Reactor
Sustainability Program (TMR, YLP, ABG, and IR); the
Nuclear Regulation Authority, Japan (IM and OK); the
EPRI Long Term Operations Program (JJW); the US
NRC (MS); Consejo de Seguridad Nuclear (CA and
MO); IETcc-CSIC (CA); and ENRESA (MO) for sup-
port of work on the effects of radiation on concrete. The
authors also express their appreciation to Drs. Kevin
Field and Elena Tajuelo Rodriguez for their review of
the document and their helpful comments and sugges-
tions.
References
Alexander, S., (1963). “Effects of irradiation on
concrete final results.” Technical report HL.63/6438.
Harwell, Oxfordshire, UK, Atomic Energy Research
Establishment.
Altoubat, S. A. and Lange, D. A., (2001). “Creep,
shrinkage, and cracking of restrained concrete at early
age.” ACI Materials Journal, 98(4), 323-331.
Andreev, V. and Kapliy, D., (2014). “Stress-state of a
thick-walled cylindrical shell under the combined
action of radiation and temperature field.” Advanced
Materials Research, 1006-1007, 177-180
Architectural Institute Japan, (2015). “Guidelines for
maintenance and management of structures in
nuclear facilities.” Tokyo: Maruzen.
Bažant, Z. P. and Gettu, R., (1992). “Rate effects and
load relaxation in static fracture of concrete.” ACI
Materials Journal, 89(5), 456-468.
Bažant, Z. P., Chern, J. C. Abrams, M. S. and Gillen, M.
P., (1982). “Normal and refractory concretes for
LMFBR application, Vol. 1: Review of literature on
high-temperature behavior of Portland cement and
refractory concretes.” EPRI Report NP-2437.
Chicago, Illinois: Northwestern University and
Portland Cement Association.
Ben Haha, M., Gallucci, E., Guidoum, A., and Scrivener,
K. L. (2007). “Relation of expansion due to alkali
silica reaction to the degree of reaction measured by
SEM image analysis.” Cement and Concrete
Research, 37(8), 1206-1214.
Bolander, Jr., J. E. and Saito, S., (1998). “Fracture
analyses using spring networks with random geome-
try.” Engineering Fracture Mechanics, 61(5-6), 569-
91.
Bouniol, P. and Aspart, A., (1998). “Disappearance of
oxygen in concrete under irradiation: The role of
peroxides in radiolysis.” Cement and Concrete
T. M. Rosseel, I. Maruyama, Y. Le Pape et al. / Journal of Advanced Concrete Technology Vol. 14, 368-383, 2016
Research, 28, 11, 1669-1681.
Bouniol, P. and Bjergbakke, E., (2008). “A comprehen-
sive model to describe radiolytic processes in cement
medium.” Journal of Nuclear Materials, 372, 1, pp.
1-15.
Bykov, V., Denisov, A., Dubrovskii, V., Korenevskii, V.,
Krivokoneva, G. and Muzalevskii, L., (1981) “Effect
of irradiation temperature on the radiation expansion
of quartz.” Atomnaya Energiya, 51, 593-595.
Carlson, R. W., (1938). “Drying shrinkage of concrete
as affected by many factors.” American Society of
Testing and Materials Proceedings, 38(2), 419-437.
Douillard, L. and Duraud, J., (1996) “Amorphization of
alpha-quartz under irradiation.” Journal de Physique
III, 6, 1677-1687.
Dubrovskii, V., Ibragimov, S., Kulakovskii, M. Y.,
Ladygin, A. Y. and Pergamenshchik, B., (1967).
“Radiation damage in ordinary concrete.” Atomnaya
Energiya, 23(4), 310-316.
Dunant, C. F. and Scrivener, K. L., (2010). “Micro-
mechanical modelling of alkali-silica-reaction-
induced degradation using the AMIE framework.”
Cement and Concrete Research, 40(4), 517-525.
Dunant, C. F. and Scrivener, K. L., (2012). “Effects of
uniaxial stress on alkali-silica reaction induced
expansion of concrete.” Cement and Concrete
Research, 42(3), 567-576.
Eby, R., Ewing, R. and Birtcher, R., (1992). “The
amorphization of complex silicates by ion-beam
irradiation.” Journal of Materials Research, 7, 3080-
3102.
Elleuch, L., Dubois, F. and Rappeneau, J., (1972).
“Effects of neutron radiation on special concretes and
their components.” Special Publication of The
American Concrete Institute, 43,1071-1108.
Engle Jr., E. A., (1967). “A user's manual for ANISN, a
one-dimensional discrete ordinates transport code
with anisotropic scattering.” Washington DC: US
Atomic Energy Commission. USAEC Report K-1963.
Esselman, T. and Bruck, P., (2013). “Expected condition
of concrete at age 80 years of reactor operation.”
New York: Lucius Pitkin Inc. Technical Report,
A13276-R-001.
Ewing, R., Meldrum, A., Wang, L. and Wang, S., (2000).
“Radiation-induced amorphization.” Reviews in
Mineralogy and Geochemistry, 39, 319-361.
Field, K., Remec, I. and Le Pape, Y., (2015). “Radiation
effects on concrete for nuclear power plants, Part I:
Quantification of radiation exposure and radiation
effects.” Nuclear Engineering and Design, 282,126-
143.
Fujiwara, K., Ito, M., Sasanuma, M., Tanaka, H.,
Hirotani, K., Onizawa, K., Suzuki, M. and Amezawa,
H., (2009). “Experimental study of the effect of
radiation exposure to concrete.” In: Proceedings of
the 20th international conference on structural me-
chanics in reactor technology, SMiRT 20-Division I.
Fournier, B. and Bérubé, M. A., (2000). “Alkali-
380
aggregate reaction in concrete: a review of basic
concepts and engineering implications.” Canadian
Journal of Civil Engineering, 27(2), 167-191.
Harbsmeier, F. and Boise, W., (1998). “Ion beam
induced amorphization in alpha quartz.” Journal of
Applied Physics, 83, 4049-4054.
Fukuya, K., Nakata, H., Fujii, K., Kimura, I., Ohmura,
M., Kitagawa, H., Itoh, T. and Shin, K., (2002).
"Radiation field analyses in reactor vessels of
PWRs." INSS Journal, 9, 153-161.
Giorla, A. B., Scrivener, K. L. and Dunant, C. F., (2014).
“Finite elements in space and time for the analysis of
generalised visco‐elastic materials.” International
Journal for Numerical Methods in Engineering, 97(6),
454-472.
Giorla, A., Vaitová, M., Le Pape, Y. and Štemberk, P.,
(2015). “Meso-scale modeling of irradiated concrete
in test reactor.” Nuclear Engineering and Design, 295,
59-73.
Giorla, A. B., Scrivener, K. L. and Dunant, C. F., (2015).
“Influence of visco-elasticity on the stress develop-
ment induced by alkali-silica reaction.” Cement and
Concrete Research, 70, 1-8.
Giorla, A. B., Le Pape, Y. and Dunant, C. F., (2016).
“Computing creep-damage interactions in irradiated
concrete.” Journal of Micromechanics and Nanome-
chanics, under review.
Giorla, A., Le Pape, Y. and Huang, H., (2016). “Meso-
scale modeling of irradiation in pressurized water
reactor concrete biological shield.” In: Proceedings of
FraMCoS-9, Berkeley, Calif.
Graves, H., Le Pape, Y., Naus, D., Rashid, J., Saouma,
V., Sheikh, A. and Wall, J., (2014). “Expanded
material degradation assessment (EMDA), Volume 4:
Aging of concrete.” Rockville, Md.: US Nuclear
Regulatory Commission. NUREG/CR-7153,
ORNL/TM-2011/545.
Gray, B., (1971). “The effects of reactor radiation on
cement and concrete.” In: Proceedings of an
information exchange meeting on 'Results of concrete
irradiation programmes,'” Volume EUR 4751 f-e.
Brussels, Belgium, 19 April 1971. Brussels: Commis-
sion des Communautées Européennes.
Hansen, T. C. and Nielsen, K. E. C., (1965). “Influence
of aggregate properties on concrete shrinkage.” ACI
Journal, 62(7), 789-794.
Hilsdorf, H., Kropp, J. and Koch, H., (1978). “The
effects of nuclear radiation on the mechanical
properties of concrete.” Special Publication of the
American Concrete Institute, 55, 223-254.
Hobbs, D. W., (1974). “Influence of aggregate restraint
on the shrinkage of concrete.” ACI Journal, 71(9),
445-450.
Hookham, C., (1991). “Structural aging assessment
methodology for concrete structures in nuclear power
plants.” Oak Ridge National Laboratory.
Ichikawa, T. and Koizumi, H., (2002). “Possibility of
radiation-induced degradation of concrete by alkali-
T. M. Rosseel, I. Maruyama, Y. Le Pape et al. / Journal of Advanced Concrete Technology Vol. 14, 368-383, 2016
silica reaction of aggregates.” Journal of Nuclear
Science and Technology (Tokyo, Japan), 39, 880-88.
Ichikawa, T. and Kimura, T., (2007). “Effect of nuclear
radiation on alkali-silica reaction of concrete.”
Journal of Nuclear Science and Technology, 44,
1281-128.
Japan Nuclear Energy Safety Organization, (2013).
“Review manual for aging management technical
evaluation.” Tokyo: JNES-RE-2013-9012.
Jennings, H. M., (2008). “Refinements to colloid model
of C-S-H in cement: CM-II.” Cement and Concrete
Research, 38(3), 275-289.
Jesser, L., (1927). “Kolloide chemische reaktionen des
tonerdezementmörtels.” Zement, 16, 741.
Kawai, T., (1978). “New discrete models and their
application to seismic response analysis of
structures.” Nuclear Engineering and Design, 48(1),
207-229.
Kelly, B., Brocklehurst, J., Mottershead, D. and
McNearney, S., (1969). “The effects of reactor
radiation on concrete.” In: Proceedings of the second
information meeting on pre-stress concrete and
reactor pressure vessels and their thermal isolation,
volume EUR-4531, Brussels, 237-265.
Kitsutaka, Y. and Matsuzawa, K., (2010). “The effect of
gamma radiation on the fracture properties of
concrete.” In: B. Oh, Ed., Fracture mechanics of
concrete and concrete structures—Recent Advances
in Fracture Mechanics of Concrete, FramCoS-7, 23-
28 May, 2010, 61-64.
Kontani, O., Ichikawa, Y., Ishizawa, A., Takizawa, M.
and Sato, O., (2010). “Irradiation effects on concrete
structure.” In: International symposium on the ageing
management and maintenance of nuclear power
plants, 2010, 173-182.
Kontani, O., Ichikawa, Y., Ishizawa, A., Takizawa, M.
and Sato, O., (2011). “Irradiation effects on concrete
durability of nuclear power plants.” Proceedings of
ICAPP 2011, Nice, France, 2-5 May 2011.
Kontani, O., Sawada, S., Maruyama, I., Takizawa, M.
and Sato, O., (2013). “Evaluation of irradiation
effects on concrete structure—Gamma-ray irradiation
tests on cement paste.” In: Proceedings of the ASME
2013 power conference: Power2013, Boston, 29 July-
1 August, 2013. New York: American Society of
Mechanical Engineers, Vol. 2, paper no. 98099,
V002T07A002.
Krivokoneva, G., (1976). “Structural changes in
feldspars under impact of radiation,” In: Crystal
Chemistry and Structural Features of Minerals,
Leningrad: Nauka, 75-79.
Lapay, W., Wang, S. C. Y. and Kim, C., (2000). “License
renewal evaluation: aging management for reactor
coolant system supports.” Westinghouse, WCAP-
14422, Rev. 2-A.
Larive, C., (1997). “Apports combinés de
l'expérimentation et de la modélisation à la
compréhension de l'alcali-réaction et de ses effets
381
mécaniques.” PhD Thesis. Ecole Nationale des Ponts
et Chaussees.
Le Pape, Y., Giorla, A. and Sanahuja, J., (2016).
“Combined effects of temperature and irradiation on
concrete damage.” Journal of Advanced Concrete
Technology, 14, 70-86.
Le Pape, Y. , (2015). “Structural effects of radiation-
induced volumetric expansion on unreinforced
concrete biological shields.” Nuclear Engineering
and Design, 295, 534-548.
Lin, M., Itoh, M. and Maruyama, I., (2015).
“Mechanism of change in splitting tensile strength of
concrete during heating or drying up to 90°C.”
Journal of Advanced Concrete Technology, 13(2), 94-
102.
Łowińska-Kluge, A. and Piszora, P., (2008). “Effect of
gamma irradiation on cement composite observed
with XRD and SEM methods in the range of radiation
dose 0-1409 MGy.” Acta Physica Polonica, 114(2),
399-411.
Magura, D. D., Sozen, M. A. and Siess, C. P., (1963). “A
study of stress relaxation in prestressing reinforce-
ment.” University of Illinois.
Majorana, C., Salomoni, V., Gramegna, F., Xotta, G. and
Pomaro, B., (2014). “Macro- and meso-scale analysis
of concrete as a multiphase material for biological
shields against nuclear radiation.” International
Journal for Numerical and Analytical Methods in
Geomechanics, 38, 518-535
Maruyama, I., Kontani, O., Ishizawa, A., Takizawa, M.
and Sato, O., (2012). “Development of system for
evaluating concrete strength deterioration due to
radiation and resultant heat.” In: Proceedings of 3rd
international conference on NPP life management for
long term operations. IAEA-CN-194-096, Salt Lake
City, May 2012.
Maruyama, I., Kontani, O., Sawada, S., Sato, O.,
Igarashi, G. and Takizawa, M., (2013). “Evaluation of
irradiation effects on concrete structure—background
and preparation of neutron irradiation test.” In:
Proceedings of the ASME 2013 power conference
POWER2013, number 98114. Boston, USA, 29 July-
1 August 1, 2013.
Maruyama, I., Nishioka, Y., Igarashi, G., and Matsui, K.,
(2014a). “Microstructural and bulk property changes
in hardened cement paste during the first drying
process.” Cement and Concrete Research, 58(0), 20-
34.
Maruyama, I., Sasano, H., Nishioka, Y. and Igarashi, G.,
(2014b). “Strength and Young's modulus change in
concrete due to long-term drying and heating up to
90°C.” Cement and Concrete Research, 66(0), 48-63.
Maruyama, I. and Sugie, A., (2014). “Numerical study
on drying shrinkage of concrete affected by aggregate
size.” Journal of Advanced Concrete Technology,
12(8), 279-288.
Maruyama, I. and G. Igarashi, (2015). “Numerical
approach towards aging management of concrete
T. M. Rosseel, I. Maruyama, Y. Le Pape et al. / Journal of Advanced Concrete Technology Vol. 14, 368-383, 2016
structures: Material strength evaluation in a massive
concrete structure under one-sided heating.” Journal
of Advanced Concrete Technology, 13(11), 500-527.
Mayer, G., and Lecomte, M., (1960). “Effet des neutrons
rapides sur le quartz cristalline et la silice vitreuse.”
Le Journal de Physique et le Radium, 21, 846-852.
McDowall, D. C., (1971). “The effects of gamma
radiation on the creep properties of concrete.” In:
Proceedings of an information exchange meeting on
'Results of Concrete Irradiation Programmes,'”
volume EUR 4751 f-e. Brussels, Belgium, 19 April,
1971. Brussels: Commission des Communautées
Européennes.
Meldrum, A., Wang, L. and Ewing, R., (1997).
“Electron-irradiation-induced phase segregation in
crystalline and amorphous apatite: A TEM study.”
American Mineralogist, 82, 858-869.
METI (Ministry of Economy, Trade and industry of
Japan), (2016). http://www.meti.go.jp/english/press/
2014/0411_02.html. [Accessed 25 February 2016.]
Mirhosseini, S., Polak, M. A. and Pandey, M., (2014).
“Nuclear radiation effect on the behavior of
reinforced concrete elements.” Nuclear Engineering
and Design, 269, 57-65.
Multon, S. and Toutlemonde, F., (2006). “Effect of
applied stresses on alkali-silica reaction-induced
expansions.” Cement and Concrete Research, 36(5),
912-920.
Ogawa, K., and Maruyama, I., (2016). “Numerical
simulation of properties of dried concrete considering
interfacial transition zone.” In: Proceedings of JCI (to
be published.)
Oxfall, M., (2013). “Climatic conditions inside nuclear
reactor containment, monitoring campaign.” Elforsk
report 13:83. Lund, Sweden: Lund University,
Faculty of Engineering, Division of Building
Materials, Technical Report TVBM-13/3172.
www.byggnadsmaterial.lth.se
Pignatelli, I., Kumar, A., Field, K., Wang, B., Yu, Y.,
Dobbs, H., Le Pape, Y., Israelachvili, J., Bauchy, M.
and Sant, G., (2016). “Direct experimental evidence
for differing reactivity alterations of minerals
following irradiation: the case of calcite and quartz.”
Scientific Reports—Nature, 6(20155), 1-10.
Pihlajavaara, S. E., (1974). “A review of some of the
main results of a research on the ageing phenomena
of concrete: effect of moisture conditions on strength,
shrinkage and creep of mature concrete.” Cement and
Concrete Research, 4(5), 761-771.
Pomaro, B., Salomoni, V., Gramegna, F., Prete, G. and
Majorana, C., (2011). “Radiation damage evaluation
on concrete within a facility for selective production
of exotic species (SPES Project).” (Italy), Journal of
Hazardous Materials, 194, 169-177.
Powers, T. C. and Brownyard, T. L., (1946). “Studies of
the physical properties of hardened Portland cement
paste, Part 3. Theoretical interpretation of adsorption
data.” Journal of the American Concrete Institute,
382
18(4), 469-504.
Primak, W., (1958). “Fast-neutron-induced changes in
quartz and vitreous silica.” Physical Review,
American Physical Society, 110, 1240-1254.
Rajabipour, F., Giannini, E., Dunant, C., Ideker, J. H.
and Thomas, M. D., (2000). “Alkali-silica reaction:
current understanding of the reaction mechanisms and
the knowledge gaps.” Cement and Concrete Research,
76, 130-146.
Remec, I., Field, K., Naus, D., Rosseel, T. and Busby, J.,
(2013). “Concrete aging and degradation in NPPs:
LWRS program R&D progress report.” In: 2013 ANS
Winter Meeting and Nuclear Technology Expo,
Washington, DC, 10-14 November, 2013. American
Nuclear Society. 403-406.
Remec, I., Rosseel, T. M., Field, K. G. and Le Pape. Y.,
(2016a). “Characterization of radiation fields for
assessing concrete degradation in biological shields
of NPPs.” Submitted to ICRS-13 & RPSD-2016.
Remec, I., Rosseel, T. M., Field, K. G. and Le Pape. Y.,
(2016b). “Characterization of radiation fields in
biological shields of NPPs for assessing concrete
degradation.” In: European Physics Journal Web of
Conferences, 106, 02002. Available from http://dx.
doi.org/10.1051/epjconf/201610602002 (Accessed
April 2016).
Rosseel, T. M., Wall, J. J., Field, K. G., Le Pape, Y.,
Naus, D., Remec, I., Busby, J. T. and Bruck, P.,
(2014).“Radiation damage in reactor cavity concrete.
In: Fontevraud 8-contribution of materials investiga-
tions and operating experience to LWRs’ safety,
performance and reliability.” Avignon, France, 14-18
September, 2014.
Rosseel, T. M., Field, K. G., Le Pape, Y., Naus, D. J.,
Remec, I. Busby, J. T., Wall, J. J., and Bruck, P.,
(2015a). “Dommages d'irradiation dans les cavites en
beton des reacteurs aux Etats-Unis.” La Revue
Generale Nucleaire, Janvier-Fevrier, RGN number 1,
21-27.
Rosseel, T.M., Field, K. G., Le Pape, Y., Remec, I.,
Giorla, A.B. and Wall, J. J., (2015b). “Recent pro-
gress on the effects of radiation on reactor cavity
concrete.” Proceedings of the international
symposium on structural mechanics in reactor
technology, SMiRT 23, Manchester, UK, August 2015.
Paper 647. Electronic.
Rusch, H., (1960). “Researches toward a general
flexural theory for structural concrete.” Journal of the
American Concrete Institute, 57(1), 1-28.
Salomoni, V., Majorana, C., Pomaro, B., Xotta, G. and
Gramegna, F., (2014). “Macroscale and mesoscale
analysis of concrete as a multiphase material for
biological shields against nuclear radiation.” Inter-
national Journal for Numerical and Analytical
Methods in Geomechanics, 38, 418-535.
Seeberger, J. and Hilsdorf, H., (1982). “Einfluß von
Radioactiver Strahlung auf die Festogkeit and
Struktur von Beton.” Institut für Massivbau and
T. M. Rosseel, I. Maruyama, Y. Le Pape et al. / Journal of Advanced Concrete Technology Vol. 14, 368-383, 2016
Baustofftechnologie, Universität Karlsruhe. Technical
Report NR 2505.
Son, H. N. and Hosoda, A., (2010). “Detection of
microcracking in concrete subjected to elevated
temperature at very early age by acoustic emission.”
Journal of Advanced Concrete Technology, 8(2), 201-
211.
Templier, C., Desage, F., Desoyer, J., Hishmeh, G.,
Cartz, L., Donnelly, S., Vinyakov, V. and Birtcher, R.,
(1996). “Ion beam amorphization of muscovite
mica.” Journal of Materials Research, 11, 1819-1824.
Thomas, J. J. and Jennings, H. M., (2006). “A colloidal
interpretation of chemical aging of the C-S-H gel and
its effects on the properties of cement paste.” Cement
and Concrete Research 36(1), 30-38.
Tomes, L. A., Hunt, C. M. and R. L. Blaine (1957).
“Some factors affecting the surface area of hydrated
Portland cement as determined by water-vapor and
nitrogen adsorption.” Journal of Research of the
National Bureau of Standards, 59(6), 357-364.
US Nuclear Regulatory Commission, (2015). Agency-
wide Documents Access and Management System
(ADAMS). Available from http://www.nrc.gov/
reading-rm/adams.html. [Accessed April 2016].
US Nuclear Regulatory Commission, (2015). “Standard
review plan for review of subsequent license renewal
applications for nuclear power plants (DRAFT).”
Rockville, Maryland: US Nuclear Regulatory
Commission, NUREG-2192.
Vodák, F., Trtík, K., Sopko, V., Kapičková, O., and
Demo, P., (2005). ”Effect of γ-irradiation on strength
of concrete for nuclear-safety structures.” Cement and
383
Concrete Research, 35(7), 1447-1451.
Vodák, F., Vydra, V., Trtik, K. and Kapičková, O.,
(2011). “Effect of gamma irradiation on properties of
hardened cement paste.” Materials and Structures, 44
(1), 101-107.
Wall, J., (2011). “Program on technology innovation:
nuclear concrete structures aging reference manual,
concrete aging—Issues for long-term operation of
nuclear power plants.” Palo Alto, Calif.: Electric
Power Research Institute. EPRI report 1023035.
Wang, L., Eby, R., Janeczek, J. and Ewing, R., (1991).
“In situ TEM study of ion-beam-induced amorphiza-
tion of complex silicate structures.” Nuclear Instru-
ments & Methods in Physics Research Section B-
Beam Interactions with Materials and Atoms, 59/60,
395-400.
Wittmann, F. (1970). “Einfluss des
Feuchtigkeitsgehaltes auf das Kriechen des
Zementsteines.” Rheologica Acta, 9(2), 282-287.
Wittels, M., (1957). “Structural behavior of neutron
irradiated quartz.” Philosophical Magazine, 2, 1445-
1461.
Wong, C., (1974). “Neutron radiation damage in some
birefringent crystals.” Physics Letters A, 50, 346.
Yano, T., Fukuda, K., Imai, M. and Miyazaki, H., (2007).
“Physical property changes of crystalline and non-
crystalline SiO2 due to neutron irradiation and
recovery by subsequent annealing.” Journal of
Nuclear Materials, 367-370, 730-735.
Zubov, V. and Ivanov, A., (1967). “Elasticity of quartz
irradiated with fast neutrons.” Soviet Physics
Crystallography, 12, 313-314.
Journal of Advanced Concrete Technology Vol. 14, 70-86, March 2016 / Copyright © 2016 Japan Concrete Institute 70

Scientific paper

Combined Effects of Temperature and Irradiation on Concrete Damage

Yann Le Pape1*, Alain Giorla2, Julien Sanahuja3

Received 17 December 2015, accepted 1 March 2016 doi:10.3151/jact.14.70

Abstract
Aggregate radiation-induced volumetric expansion (RIVE) is a predominant mechanism in the formation of mechanical
damage in the hardened cement paste (hcp) of irradiated concrete under fast-neutron flux (Giorla et al. 2015). Among the
operating conditions difference between test reactors and light water reactors (LWRs), the difference of irradiation flux
and temperature is significant. While a temperature increase is quite generally associated with a direct, or indirect (e.g., by
dehydration) loss of mechanical properties (Maruyama et al. 2014), it causes a partial annealing of irradiation amorphi-
zation of α-quartz, hence, reducing RIVE rate. Based on data collected by Bykov et al. (1981), an incremental RIVE
model coupling neutron fluence and temperature is developed. The elastic properties and coefficient of thermal expansion
(CTE) of irradiated polycrystalline quartz are interpreted through analytical homogenization of experimental data on
irradiated α-quartz published by Mayer and Lecomte (1960). The proposed model, implemented in the meso-scale
simulation code AMIE, is compared to experimental data obtained on ordinary concrete made of quartz/quartzite ag-
gregate (Dubrovskii et al. 1967). Substantial discrepancy, in terms of damage and volumetric expansion developments, is
found when comparing irradiation scenarios assuming constant flux and temperature, as opposed to more realistic test
reactor operation conditions.

1. Introduction rise in temperature leads to decomposition of the cement

The structural significance of irradiated concrete is of
primary importance for extended long-term operation
(LTO) of LWRs (Graves et al. 2014; Le Pape et al. 2014;
Rosseel et al. 2014). Although historical data collected
from the literature (Hilsdorf et al. 1978; Field et al. 2015)
show a significant loss of mechanical properties corre-
lated with RIVE at high neutron fluence (i.e., > 1019
n.cm-2), the separation of the irradiation effects and
thermal effects remains problematic. Indeed, irradiation
experiments of concrete are typically performed in the
range of 40 °C to 250 °C. Most irradiation experiments
reaching fluences above 2×1019 n.cm-2 were conducted at
temperatures, even sporadically, above 100 °C (Field et
al. 2015). Understanding the combined effects of tem-
perature and irradiation on concrete is therefore required
to estimate the correspondence between test reactor data
and LWR conditions, i.e., with in-service temperature
limited, by design, to 65 °C.
At the concrete scale, i.e., at the macroscopic scale,
temperature affects the properties of unirradiated con-
crete, e.g., (Naus 2005, 2010), impacts the moisture
transport kinetics (Luikov 1975) and, causes dimensional
change. At the cement scale, i.e., at the microscopic scale,
1
Oak Ridge National Laboratory, Oak Ridge, TN 37831,
USA.
*Corresponding author, E-mail: lepapeym@ornl.gov
2
Oak Ridge National Laboratory, Oak Ridge, TN 37831,
USA.
3
EDF Lab, Moret-sur-Loing, 77818 Cedex, France.
hydrates (Zhang and Ye 2012) and the development of
cracking (Maruyama and Sasano 2014) due to differen-
tial deformations between the paste and the aggregates.
Most aggregates are not subject to thermal damage below
300 °C.
Irradiation causes physical changes in the cement
paste and the aggregates: (1) γ radiation induces radio-
lysis of water in the hcp, causing gas release and in-
creasing shrinkage at high temperature (Elleuch et al.
1972), and, (2) neutron radiation induces amorphization
of crystalline minerals contained in the aggregates,
leading to very high RIVE as well as a loss of mechanical
properties (Hilsdorf et al. 1978; Field et al. 2015).
However, the potential synergetic effects of tempera-
ture and irradiation on aggregate RIVE and CTE have
received limited attention, although aggregate expansion
(amplitude and rate) is a predominant effect in the for-
mation of damage in irradiated concrete (Dubrovskii et al.
1967; Elleuch et al. 1972; Seeberger and Hilsdorf 1982;
Le Pape et al. 2015; Giorla et al. 2015). RIVE is gener-
ally modeled as a function of the neutron fluence only,
neglecting a potential contribution of the temperature.
Studying combined temperature and irradiation effects
requires a significantly large and well-documented da-
tabase of irradiation test results conducted on the exact
same material. In that regard, the variations of chemical
composition within the broad classification of aggregates
(e.g., limestone, granite, sandstone, etc...) introduce an
interpretation bias. Additionally, the neutron kinetics
energy spectrum or threshold, for a given experiment, is
frequently not well-documented (Field et al. 2015),
adding another possible interpretation difficulty.
Y. Le Pape, A. Giorla, J. Sanahuja / Journal of Advanced Concrete Technology Vol. 14, 70-86, 2016
Neutron irradiation experiments of quartz at different
temperatures have been solidly reported by Bykov et al.
(1981). Dubrovskii et al. (1967) and Pedersen (1971)
also measured the irradiated properties of concrete made
with quartz aggregates. Notably, Dubrovskii et al. pro-
vide the temperature history for some of their samples,
thus allowing a detailed analysis of that particular ex-
periment. The data collected by Bykov et al. is inter-
preted in this article to develop a RIVE model for quartz
accounting for temperature and neutron flux. This model
is complemented with an analysis of the evolution of the
mechanical properties of quartz as a function of its den-
sity, using experimental measurements by Mayer and
Lecomte (1960) and analytic homogenization schemes.
This provides the sufficient information to perform a
preliminary numerical analysis of Dubrovskii et al. ex-
periment with the mesoscale model previously devel-
oped by the authors in Giorla et al. (2015). Different
scenarios in terms of neutron flux and temperature
variations are simulated to identify the potential impact
of these variables on the concrete RIVE and damage, and
determine whether coupling between temperature and
irradiation is required in further modeling efforts for
irradiated structures.
2. A model for the effects of temperature
and neutron irradiation on quartz
expansion
Figure 1 presents the experimental volumetric expansion
collected by Bykov et al. (1981) on quartz samples under
neutron irradiation at temperature ranging from 20 °C to
600 °C. It is observed that with increasing temperatures,
the expansion amplitude and rate decrease monotonically
at any given fluence. The sigmoidal nature of the ex-
pansion curves remains unchanged. The final amplitude,
i.e, ≈ 17.8% at full amorphization, is not affected by
temperature. “The relation between the value of the
volume expansion and the degree of impairment of the
lattice periodicity, according to the data of X-ray dif-
fractometry [...] is independent of the irradiation tem-
perature. This confirms that the type of defects origi-
nating and the nature of the damage caused are essen-
tially unchanged with increase of temperature” (Bykov et
al. 1981). The effects of temperature on quartz expansion
is explained by the annealing of point defect during ir-
radiation.
RIVE of quartz can be modeled by the empirical
equation developed by Zubov and Ivanov (1966) under
isothermal conditions:
eδ Φ − 1
ε * (Φ, T = cte) = κε max
ε max + κ eδ Φ
The terms ε max corresponds the maximum volumet-
ric expansion, i.e., ε max = 17.8%, for quartz. κ (di-
mensionless parameter homogeneous to a strain) and δ
71
(inverse of a fluence) govern the shape of the sigmoid. In
particular, the inflexion point of the sigmoid, i.e.,
∂ ε / ∂Φ = 0 , corresponds to a fluence of
2 * 2
ln(ε max / κ ) / δ and a swelling of (εmax − κ ) / 2 (Le Pape
2015). At the same inflexion point, the rate of swelling is
given by the expression (ε max − κ )δ / 4 . Hence, the main
characteristics of Zubov's curve, except for the final
amplitude of swelling, do not show a one-to-one de-
pendency on Zubov's model parameters. Eq. (1) can also
be interpreted in a slightly modified form, which corre-
sponds, mathematically speaking, to the form adopted by
several researchers, e.g., (Larive 1997; Saouma and
Perotti 2006), to model the expansion of concrete sub-
jected to alkali-silica reaction (ASR) as a function of
time. Substitution of time by the fluence Φ leads to:
−
Φ
1− e Φc
ε (Φ,T = cte) = ε max
*
(2)
−
Φ-Φ L
1+ e Φc
with Φ c =1/δ the characteristic fluence that primarily
governs the rate of expansion at the inflexion point of the
sigmoid, and Φ L = ln(ε max / κ ) / δ the latency fluence,
i.e., the fluence at the inflexion point, which is located at
about half of the total expansion, when κ << ε max .
As shown in Fig. 1, Zubov and Ivanov's equation
provides a good approximation of the trend lines plotted
originally by Bykov et al. although it tends to overesti-
mate the expansion rate above ≈ 15% expansion.
Fig. 1 Volumetric expansion of neutron-irradiated quartz
at different temperatures (Adapted and modified from
(Bykov et al. 1981)). (a)-(k) indexes and symbol colors
correspond to different ranges of temperature ranked by
increasing order. (a): 25-30 °C; (b): 30-45 °C; (c):
40-50 °C; (d): 60-80 °C; (e): 80-85 °C; (f): 85-90 °C; (g):
90-100 °C; (h): 105-115 °C; (i): 120-160 °C; (j): 210 °C;
(1) (k): 240 °C. Source: □ in APS-1 NPP reactor; ▽ in
BR-5 test reactor; ◇ in BR-I0 test reactor. ○ from
literature (Wittels 1957; Primak 1958; Weissmann and
Nakajima 1963). Dashed lines: Bykov et al. trend lines
(no equations given). Solid lines: fit assuming Zubov and
Ivanov's Eq. (1).
Y. Le Pape, A. Giorla, J. Sanahuja / Journal of Advanced Concrete Technology Vol. 14, 70-86, 2016 72

As noted by Bykov et al., "the very strong temperature From Eq. (2), the incremental variation of volumetric
effect in the range of 30 °C to 100 °C and the reduction of expansion at a given neutron fluence, temperature, neu-
this effect at a higher temperature indicates a nonuni- tron flux and temperature increment reads:
formity of distribution of the defects with respect to the
energy of activation, in particular, to a reduction of the d * ∂ε * dΦ ∂ε * dT
fraction of defects with an increased energy of activa- ε (Φ, T ) = ⋅ + ⋅ (4)
dt ∂Φ T dt ∂T dt
tion." Hence, instead of trying to interpret the data with Φ

the calibration of an apparent energy activation, tem-

perature effects are analyzed directly on how Zubov's The isothermal RIVE rate is obtained by direct deri-
parameters are affected. While the plots of κ and δ vation from Eq. (1):
(not shown here) exhibit a nonlinear relation with the
temperature, the plots of the fluences Φ c and Φ L show ∂ε * δ T eδ Φ (ε max + κ T )
T

a quasi linear evolutions against the temperature increase = κT ε max (5)

∂Φ T
(ε max + κT eδ Φ ) 2
T

(Fig. 2). Best fit regressions provide the empirical rela-

tions:
Φ L = aLT + bL ≈ 2.40 × 10−2 T − 6.26 ×10−2
(3)
Φ c = acT + bc ≈ 2.77 × 10−3 T + 0.128
with T in °C and Φ L,c in ×1020 n.cm-2. From Eq. (3) the
evolutions of κ = ε max / exp(δ Φ L ) and δ = 1/ Φ c with
temperature can be derived, Eq. (6). Note that the default
parameters suggested by Zubov and Ivanov (1966) for
quartz, i.e., κ = 0.27% and δ = 4.5 × 1020 n.cm-2
correspond to a temperature of 34 to 41 °C assuming Eq.
(3) holds, which is consistent with the irradiation tem-
peratures from the experiments of Wittels (1957).
Figure 3 shows the validation plot of the proposed
model against the experimental data, i.e., the data cor-
responding to marks in Fig. 1 at varied temperatures and
fluences. Although good correlation between the ex-
perimental and modeled data is found, the scatter band is
quite important because of the uncertainty on the actual
irradiation temperature, the unreliability of Zubov's
model at high expansions, i.e., above ≈ 15%, and the
strong dependency of the model paramaters to tempera-
ture.
with
1 ε max
δT = and κ T = (6)
acT + bc ⎛ a T + bL ⎞
exp ⎜ L ⎟
⎝ acT + bc ⎠
The second part of Eq. (4), i.e., in iso-neutronic con-
dition ( Φ = cte), can be derived from post-irradiation
thermal experiments (Mayer and Lecomte 1960; Primak
1958; Yano et al. 2007). Figure 4 shows the evolutions
of the postirradiation expansion of quartz during an
isochronal thermal heating up to about 1,000 °C at a rate
of about 1,000 °Ch-1 (Yano et al. 2007). For specimens
irradiated at low fluence, annealing, or “self-healing”,
i.e., the evolution toward α-quartz accompanied by a
reduction of the expansion to the initial value of quartz
density is observed. Conversely, after irradiation at high
fluence, anti-annealing is observed: quartz evolves to-
ward more amorphous forms, called cryptocrystalline by
Wittels, although different from vitreous silica unless
fluence of the order of 1020 n.cm-2 is reached (Primak

Fig. 3 Correlation between the modeled expansion and

Fig. 2 Evolution of the characteristic and latency fluences,
respectively Φc and Φ L as function of the irradiation
temperature. ● correspond to the calculation of the flu-
ences Φc and Φ L derived from the fits of Bykov et
al.'s trend lines (Fig. 1) using Zubov's Eq. (1). Linear fit
equations (solid lines) provided directly with the plots.
the measured expansion of quartz at varied fluences and
temperatures. Experimental data from (Bykov et al.,
1981). Modeled data assuming Zubov's Eq. (1) and a
linear dependence of the latency and characteristic flu-
ences on the temperature (Fig. 2). ●: T < 65 °C; ■:
T ∈ [65; 110] °C; ■: T ∈ [110; 180] °C; ◆: T >
180 °C, with T the median temperature.
Y. Le Pape, A. Giorla, J. Sanahuja / Journal of Advanced Concrete Technology Vol. 14, 70-86, 2016 73

1958). Mayer and Lecomte (1960) estimate the dose

 = 1 ln ε max (ε i + κ i +1 )
*
threshold separating the annealing/anti-annealing be- Φ (8)
i +1
haviors at ≈ 0:65 × 1019 n.cm-2 at energy above 0.3 δ i +1 κ i +1 (ε max − ε i* )
MeV (Keppens and Laermans 1996). Since limited
volumetric change (< 0.5%) is observed in the typical and, as a consequence, the volumetric expansion rates at
range of temperatures, i.e., < 400 °C, of test or commer- the beginning and the end of the time step read:
cial reactor operations, it appears legitimate to neglect
the term δε * / δ T Φ ≈ 0 in Eq. (4), for those particular δε * i

δ i eδ Φ (ε max + κ i )
i

applications. = κ i +1ε max (9)

The temperature history depends on the test reactors
∂Φ ti (ε max + κ i eδ i Φ i

)
operation, and can exhibit daily variation (e.g., JEEP-II
in Norway) and quasi-monthly cyclic evolutions (e.g., and, the volumetric expansion increment is obtained by:
BR-5 reactor in Russia). Accounting for temperature

variations implies the estimation of the volumetric ex- δε * δ i +1eδ i +1Φi −1
(ε max + κ i +1 )
= κ i +1ε max (10)
pansion ε i*+1 = ε i* + d ε * after a temperature step
dT = Ti +1 − Ti during the time increment d t = ti +1 − ti .
∂Φ ti+1 ( ε max + κ i +1e
δ i ;1Φ i +1
)
Since Bykov et al. observed that, at a given value of
volume expansion, the degree of impairment of the lat- where ξ is a numerical parameter between 0 and 1
tice periodicity is independent of the irradiation tem- controlling the integration scheme, i.e., 0 for an implicit
perature, it can be assumed that (1) The volumetric ex- scheme, 1 for an explicit scheme or in between for any
pansion is an internal state variable characterizing the semi-implicit/explicit scheme — See Fig. 5.
degree of irradiation-induced damage. (2) The expansion
rate, at a given time, appears to be a function only of the 3. Elasticity and thermal expansion of
corresponding neutron flux and temperature, and not, a irradiated quartz
function of the histories of flux an temperature. Hence,
the same volumetric expansion ε i* would have been Quartz (aggregate) is a polycrystalline assemblage of
obtained under an isothermal irradiation at temperature randomly oriented α-quartz trigonal (rhombohedral)
Ti to the equivalent fluence Φ  : monocrystals. Hence, while α-quartz possesses anisot-
i
ropic properties, polycrystalline quartz is an isotropic
materials and its properties can be derived by homog-
 = 1 ln ε max (ε i + κ i )
*
Φ (7) enization, e.g, by calculating Reuss-Voigt bounds or by
i
δ i κ i (ε max − ε i* )
using a self-consistent scheme. Under irradiation,
α-quartz and vitreous silica, e.g., fused quartz evolve
The same reasoning can be applied at temperature Ti+1 toward a similar isotropic structure caused by the disor-
 :
to derive the equivalent fluence Φ i +1 dering created by neutron irradiation (Wittels 1957;
Primak 1958). When subjected to temperature increase,
α-quartz is relatively stable until phase transformation to
β-quartz (hexagonal) at 573 °C (atm. pressure) is
reached.

Fig. 4 Post-irradiation annealing or anti-annealing of

quartz. The fluence level before thermal treatment is
20 -2
given in brackets ×10 n.cm . ●: from dimensional
change measurements (Primak 1958); □: from density
measurements (Primak 1958); ◇: from dimensional Fig. 5 Derivation of the incremental volumetric expansion
change measurements (Yano et al. 2007). accounting for temperature variations.
Y. Le Pape, A. Giorla, J. Sanahuja / Journal of Advanced Concrete Technology Vol. 14, 70-86, 2016
3.1. Elastic properties
Trigonal α-quartz is characterized by 6 independent
elastic constants, i.e., cij stiffness coefficients, including a
non-zero c14 term, while isotropic silica only requires the
determination of 2 constants. The values of cij for varied
polymorphs of crystalline silica and silica glass have
been obtained experimentally in the temperature range of
0 °C to ≈ 1, 000 °C by several authors, e.g., see review
articles (Ballato 2008; Pabst and Gregorová 2013).
Elastic constants of irradiated α-quartz and silica have
been obtained by Mayer and Gigon (1956); Mayer and
Lecomte (1960); Zubov and Ivanov (1967) in the cor-
responding range of RIVE below ≈ 6%, i.e., at about a
third of the maximun expansion. Tentative extrapolation
for RIVE beyond 6% is discussed hereafter. The elastic
properties of polycrystalline quartz can be theoretically
by derived by homogenization, i.e., Voigt-Reuss bounds
(Voigt 1889; Reuss 1929), or self-consistent scheme
(Dormieux et al. 2006). The results can either be plotted
as a function of RIVE or density. While the implemen-
tation in a finite element frame (See section 5) is done
considering properties function of the volumetric ex-
pansion (internal variable), plotting the irradiated prop-
erties of both α-quartz and silica as functions of the
density make it possible to evaluate the possibilities of
extrapolation beyond the study range, assuming with
Mayer and Lecomte, that at full amorphization, α-quartz
and silica evolve toward the same disordered state, i.e.,
of equal densities.
Young’s modulus of irradiated α-quartz.
Figure 6 shows that, in the range of studied density,
i.e., 2.55 to 2.65, irradiation affects the Young modulus
of quartz although no monotonic decreasing or increas-
ing trend can be observed. Irradiated Young modulus
remains in the range of 90 to 100 GPa. Under the same
irradiation conditions, the Young modulus of silica
Fig. 6 Young modulus of irradiated α-quartz and silica as
a function of the irradiated materials density. (s) amor-
phous silica; (r) irradiated quartz, Reuss bound; (v) irra-
diated quartz, Voigt bound; (SC) irradiated quartz,
self-consistent scheme.
74
slightly increases from 73 GPa to 76 GPa. Hence, there is
still a significant gap between the Young moduli of
α-quartz and silica at the end of Mayer and Lecomte's
experiments preventing any extrapolation. It appears that
Voigt-Reuss bounds tend to narrow with irradiation. This
suggests that the behavior of α-quartz crystals evolves,
with irradiation, towards isotropy.
Poisson’s ratio of irradiated α-quartz.
The Poisson ratio of irradiated α-quartz, v* , shows a
rapid increase from its initial value v0 = 0.08 to reach
about 0.18 at ≈ 4.5% volume expansion - See Fig. 7. In
this range of RIVE, the Poisson ratio estimated by the
selfconsistent scheme is fittted by an exponential func-
tion. In the same range of fluence, the Poisson ratio of
silica increases from 0.17 and seem to level off at about
vs* ≈ 0.21, which will be considered here, as the
maximum limit value of irradiated α-quartz at full
amorphization - See Fig. 8. The proposed evolution of
irradiated α-quartz Poisson ratio reads:
( ( b
) )
v* = min 1 − v0 + exp a ( ε * ) ; vs* (11)
with a and b two fit parameters found equal to 0.0063 and
1.938.
3.2. Coefficient of thermal expansion
The linear CTE of α-quartz at about 20 °C varies with the
crystallographic orientation: ≈ 7.4 μmm-1 °C-1 ( &c , i.e.,
optical axis) and ≈ 1 μmm-1 °C-1 ( ⊥ c ) (Amatuni and
Shevchenko, 1966). Similar values are found in natural
quartz in the range of -20 °C to 60 °C (Johnson and
Parsons 1944, nat. quartz, Minas Geraes, Brazil), i.e., α
= 12 μmm-1 °C-1. In the range of ≈ 20 °C to 350 °C,
quartz CTE can be approximated by a linear law α (T)
≈ 11.2+10.9×10-3T, with α in μmm-1 °C-1 and T in °C
Fig. 7 Poisson ratio of irradiatedα-quartz and silica as a
function of the irradiated materials density. (s) amor-
phous silica; (r) irradiated quartz, Reuss bound; (v) irra-
diated quartz, Voigt bound; (SC) irradiated quartz,
self-consistent scheme.
Y. Le Pape, A. Giorla, J. Sanahuja / Journal of Advanced Concrete Technology Vol. 14, 70-86, 2016 75

Fig. 9 Coefficient of thermal expansion of irradiated

Fig. 8 Poisson ratio of irradiated α-quartz as a function of α-quartz and silica as a function of the irradiated materi-
RIVE. Gray-colored portion: Voigt-Reuss bounds. als density. (s) amorphous silica; (r) irradiated quartz,
Dashed line: self-consistant scheme. Thick solid line: Reuss bound; (v) irradiated quartz, Voigt bound; (SC)
Suggested fit with Eq. (11). irradiated quartz, self-consistent scheme.

(Amatuni and Shevchenko, 1966) (see Kosinski et al.

(1991) for a Taylor expansion of the CTE).
However, fused-quartz has a much lower CTE, ≈
0.42 μm m-1 °C-1 at 25 °C (Beattie et al. 1941). It in-
creases nonlinearly up to 0.5-0.6 μm m-1°C-1 at ≈ 200 °C
before decreasing monotonically toward 0.42 to 0.54 μm
m-1 °C-1 at ≈ 1,000 °C. The temperature of the peak
value varies with the authors: 200 to 400 °C (Otto and
Thomas 1963). Also, α-quartz, obtained after phase
transformation at 573 °C (atm. pressure), exhibits a very
low slightly negative CTE (Welche et al. 1998). Hence, it
can be conjectured that irradiation-induced amorphiza-
tion affects the CTE of quartz toward decreasing. The
CTE of α-quartz at 20 °C shows little modification after
irradiation for fluences up to 2.2 × 1019 n.cm-2 (Mayer
and Gigon 1956). Above that fluence value, Mayer and
Lecomte (1960)'s post-irradiation CTEs (average value
Fig. 10 Coeffcient of thermal expansion of irradiated
between -190 °C to 0 °C) show a sharp decrease for
α-quartz as a function of RIVE. Gray-colored portion:
RIVE above ≈ 10% of the maximum expansion. Figure
Voigt-Reuss bounds. Thick solid line: Suggested fit with
9 shows that the CTE of α-quartz and silica appear to be
Eq. (12).
reaching similar values at the end of Mayer and Le-
comte's experiments. Interestingly, the CTE of irradiated
silica shows a decreasing trend in the negatives. Note, ⎛
* *
⎞
however, that post-irradiation measurements of CTEs 1 − e −ε / ε c
α = α + α1 ⎜ 1 −
* *
⎟⎟ (12)
(-196 °C to 20 °C) on varied limestone, dolerite, andesite
0 ⎜ 1 + e − (ε * −ε L* ) / ε c*
⎝ ⎠
and hornfels at fluence above 1019 n.cm2 (fast neutron)
show either little variation, moderate increase or large with ε c* and ε L* , two fitting coefficients found equal to
increase (dolomite) compared to the pre-irradiation 0.85% and 4.27%, respectively. The choice of the fitting
CTEs (Kelly et al. 1969; Gray 1971). Hence, generali- function is arbitrary. ε 0* corresponds to the CTE of
zation to other-than-quartz aggregate is subject to cau- fully-irradiated quartz or silica, i.e. ≈ -0.52 × 10-6 μm
tion. m-1 °C-1. α1 + α 0* is the value of the pristine α-quartz
Based on the observed trend (Fig. 10), the CTE of ir- CTE, i.e., α (T ) assuming that Eq. (12) holds beyond
radiated polycrystallin quartz (average value between the temperature range studied by Mayer and Lecomte.
-190 °C to 0 °C), α * obtained with a self-consistent After fitting, α1 is found equal to 8.63 μm m-1 °C-1.
scheme can be fitted by a sigmoidal function similar to Because the amplitude variation of fused quartz in the
Eq. (2): temperature range of 20 °C to 400 °C quite limited, i.e.,
Y. Le Pape, A. Giorla, J. Sanahuja / Journal of Advanced Concrete Technology Vol. 14, 70-86, 2016 76

Table 1 Irradiation parameters and volumetric expansion (assumed as the cube of the measured diameter expansion) for
Dubrovskii et al.'s data. Table assembled from (Dubrovskii et al., 1967, Original Tab. 1 & 3). †: estimated values.
Group No.
1 2 3 4
20 -2
integrated flux (×10 n.cm ) 3-3.3 2-2.8 1.2-1.45 0.4-0.6
13 -2 -1
maximum flux (×10 n.cm s ) 1.9-2.1 1.3-1.8 0.75-0.9 0.25-0.4
average flux (×1013 n.cm-2 s-1) 0.62†
maximum heat deposition (Wcm-3) 0.19-0.21 0.12-0.17 0.07-0.085 0.025-0.04
maximum temperature (°C) 300-350 225-275 100-150 20-50
average temperature (°C) 115 †
89 57 †
35†
concrete vol. expansion (%) 16.2 >17.4 9.3 2.4
> 19.2 >18.9 5.1 1.5
> 17.7 >15.9 5.7
> 18.0

0.4-0.6 μm m-1 °C-1, and, remains one order of magnitude n.cm-2 s-1). Assuming that the found linear correlation is
below that of pristine α-quartz CTE, the effects of tem- applicable in the range of temperatures corresponding to
perature on α 0* can be neglected ( α 0* ≈ 0.5 μm the BR-5 reactor operation, the flux history of group 2
m-1 °C-1). experiments, plotted in Fig. 11, can be derived from the
history of temperature provided in (Dubrovskii et al.
4. Interpretation of irradiation experiments 1967, original Fig. 1). Note that kinetic energies above
on quartz aggregate concrete 10 keV account for about 99.5% of the total energy (See
energy distributions in (Dubrovskii et al. 1967, Table 2.,
While quartz and quartzite sand is commonly used in p 1054)). Hence, the empirical parameters ac, bc, aL , and
concrete, coarser aggregate are usually made of other bL, derived from the interpretation of Bykov et al.'s data
rocks, e.g., limestone, granite, basalt, etc. . . Hence, ir- are assumed valid to model Dubrovskii et al.'s data. The
radiation experiments on concrete are rarely conducted average temperature and average flux calculated from
with pure quartz aggregate. We found only to sets of these plots are 89 °C and 6.2 × 1012 n.cm-2 s-1, i.e, re-
irradiation experiments on quartz aggregate concrete in spectively, 2.5-3 times, and 2-2.5 times, lower that the
the literature (Dubrovskii et al. 1967; Pedersen 1971). maximum temperature and flux. Figure 12 illustrates the
The details provided by Pedersen are unfortunately quite potential implication of different temperature and flux
limited to envision a thorough interpretation through scenarios in terms of volumetric expansion evolutions of
modeling. Dubrovskii et al. (1967) tested the effects of quartz, assuming the proposed model hold - Eqs. (2- 10):
neutron irradiation of ordinary concrete (density 2.31) (a), i.e., the “reference scenario”: variable temperature
cylinder of 40 mm in height and diameter, made of
Portland cement (w=c = 0.5). The mix contains quartz
river sand (0.15 to 0.6 mm, c : a = 1 : 1.7) and sandstone
gravel (5 to 10 mm, c : a ≈ 1 : 3). The volume fraction
of quartz sand is about 25%. The total volume fraction of
aggregate is close to 70%. Using the chemical composi-
tion of the mix provided by Dubrovskii et al., and as-
suming a typical oxide composition for the Portland
cement, we found that the molar composition of the
sandstone is about 31% of silicon and 66% of oxygen,
proving the highly-siliceous nature of the tested sand-
stone containing also traces of calcium, iron, magnesium
and aluminum, supporting Dubrovskii et al. statement
that “the total quartz content in this concrete was 70%.”
Table 1 presents the irradiation fluxes, the temperature,
and the post-irradiation volumetric expansions for the Fig. 11 Estimated time-evolution of the integrated energy
four tested groups. flux (black solid line) and fluence (thick solid red line)
Quasi-linear correlations between the maximum flux, over time - Group 2. Dashed line: fluence evolution as-

Φ max , the maximum heat deposition and maximum
12 -2 -1
suming a constant flux at 6:2×10 n.cm s . The error
temperature, Tmax can be observed ( Φ  ≈ 5.83 ×
max bar at 501 d corresponds to the intervale of fluences
-3
10 Tmax + 0.106 with Tmax in °C and Φ  13
max in ×10 provided in Table 1.
Y. Le Pape, A. Giorla, J. Sanahuja / Journal of Advanced Concrete Technology Vol. 14, 70-86, 2016
(Dubrovskii et al. 1967, original Fig. 1) and variable flux
(Fig. 11); (b): constant temperature (89 °C) and variable
flux; (c): constant temperature and constant flux 6.2×1012
n.cm-2 s-1; and, (d): variable temperature and constant
flux. At the end of the experiment, the estimated RIVE
ranges between 14.7% and 17.8%. However, important
scatter can be observed for times around 300 d: while
scenario (b) leads to almost full expansion, i.e, full
amorphization, scenario (d) leads to a RIVE under 2%.
The scenario (c) assuming constant temperature and flux
underestimates the volumetric expansion by <7%. Hence,
the total fluence received during irradiation is not a suf-
ficient parameter to describe the experiment. At high
fluence (groups 1 and 2), RIVE was large enough to fill
the space between the concrete sample and the stainless
steel capsule wall. When excessive swelling occurred,
the expansion data are presented in Table 1 with a > sign,
since the measurements were taken after the extraction of
the wedged samples. It can be observed that the volu-
metric expansions, calculated assuming a factor 3 on the
diameter change, are close to the maximum expansion of
quartz, i.e., 17.8%, for the samples from groups 1 and 2.
Dubrovskii et al. noted that “the nature of temperature
variations for samples in the first and third groups was
the same” (as the second group). Hence, the same tem-
perature and flux histories can be used for groups 1 to 3.
An approximation of the concrete expansion, ε c , can be
derived from the aggregate expansion, ε * and its vol-
ume fraction, fa, for important level of expansion leading
to important damage in the hcp (Le Pape et al. 2015)
(micromechanical theory):
Fig. 12 Time-evolution of the volumetric expansion of
quartz for group 2 specimens assuming different tem-
perature and flux scenarios. (a): variable temperature
(Dubrovskii et al. 1967, original Fig. 1) and variable flux
(Fig. 11); (b): constant temperature (89 °C) and variable
flux; (c): constant temperature and constant flux 6.2 ×
12 -2 -1
10 n.cm s ; and, (d): variable temperature and con-
stant flux. The black solid line corresponds to the flux and
temperature correlated histories.
77
2 fa
εc ≈
1+ fa
( ε * + 3αΔTmax ) (13)
The term 3αΔTmax correspond to the maximum
thermal volumetric expansion of aggregate during the
irradiation experiment. Eq. (13) assumes that the cement
paste suffers extensive micro-cracking, up to the point
that it has no elastic stiffness. The strains in the cement
paste (either due to shrinkage or thermal deformations)
have therefore no influence over the concrete homoge-
nized expansion.
In Dubrovskii et al.'s experiments on group 2, RIVE of
0.10ε max is reached approximately around 150 d (sce-
nario (a) in Fig. 12). The maximum temperature reached
during this phase is about 150 °C. Hence, the term
3αΔTmax can be estimated by 3 × 1.2 × 10-5 × 130 ≈
0.5%. As a consequence, for this particular set of ex-
periments, the thermal expansion of aggregate during
irradiation can be neglected in comparison to RIVE. For
a volume fraction of 70%, the expansion transfer coeffi-
cient to concrete, 2fa/(1+fa), equals 0.82. Hence, the
maximum theoretical concrete expansion is 0.82×17.8%
= 14.6% while the measured concrete expansion for
samples in groups 1 and 2 are around 17.8%, i.e., the
maximum expansion of quartz. Such discrepancy may be
explained by: (1) biased diameter change measurement
caused by nonuniform swelling, as suggested by (Du-
brovskii et al. 1967, Photograph of sample No. 8 in Fig.
3), resulting from the aggregate distribution in the sam-
ple, (2) displacement field discontinuities induced (mi-
cro-) cracking. This motivates the development of a
Fig. 13 Time-evolution of the volumetric expansion of
concrete for group 2 specimens assuming a variable
temperature (Dubrovskii et al. 1967, original Fig. 1) and
variable flux (Fig. 11). The boxes indicate the ranges of
maximum fluences and expansions provided by Du-
brovskii et al. (data in Table 1). Dashed lines indicate the
uncertainty of the expansion measurement because the
irradiated samples were wedged in the capsule (see
text).
Y. Le Pape, A. Giorla, J. Sanahuja / Journal of Advanced Concrete Technology Vol. 14, 70-86, 2016
numerical model, as analytical micromechanical models
have limited capabilities to capture these effects.
5. Numerical modeling
We simulate the experiments of Dubrovskii et al. (1967)
with the model previously developed by the authors for
irradiated concrete. The material is considered at the
meso-scale with a two-dimensional finite element rep-
resentation of its microstructure. The model calculates
the deformation as well as the localization and extent of
the micro-mechanical damage in the sample, accounting
for the different properties and behaviors of the cement
paste and the aggregates.
Post-irradiation deformations are the only experi-
mental result available for this set of temperature and
fluences, as Dubrovskii et al. were unable to measure
mechanical properties due to the samples being wedged
in the irradiation rig. The model provides an estimation
of the history leading to that deformation. Rigorous
validation of the simulation results cannot be achieved
due to the absence of monitoring during the experiments
and the uncertainties related to the data provided by
Dubrovskii et al.. The principal purpose of the simula-
tions is to give insights on the influence of the tempera-
ture and neutron flux variations on the expansion and
damage variations, and assess the importance of consid-
ering combined or separated effects of temperature and
irradiation in further modeling efforts.
The main components of the model are summarized
below, and described in greater details in Giorla et al.
(2015). The model is implemented in the C++ finite
element framework AMIE, which was initially devel-
oped in the context of mesoscale simulations of ASR
(Dunant and Scrivener 2010), and later applied for the
study of creep (Giorla et al. 2014) and irradiation (Giorla
et al. 2015).
5.1. Microstructure
We simulate a two-dimensional vertical slice of a con-
crete cylindrical sample. Aggregates are represented with
circles randomly placed in a square box. Dubrovskii et al.
used two different aggregate classes in their experiments:
sand (0.15-0.6 mm) and gravels (5-10 mm), without
intermediate sizes. In the simulation, only gravels are
accounted for, to avoid using a very fine finite element
mesh. While sand is not explicitely accounted for in the
simulation, the surface fraction of gravels is increased to
58% in order to partly compensate for its absence (It is
reminded that the actual aggregate volume fraction is
70% in Dubrovskii et al.'s experiment). The particle size
distribution follows a Füller curve between the given
minimum and maximum gravel diameters in absence of
further data (Fuller and Thompson 1907).
The RIVE of the missing sand fraction is neglected for
this set of simulations. The properties of the cement paste
could be adapted in order to account for that additional
expansion. However, to do so, one would need the ex-
78
pansion and degradation of the sand and cement paste
mixture not only as a function of the fluence and tem-
perature, but also as a function of the level and direction
of the current stress. Mechanical restraint is likely to
affect the relation between expansion and degradation,
notably making them both anisotropic as it is the case in
similar phenomena like ASR (Larive 1997; Multon and
Toutlemonde 2006; Dunant and Scrivener 2012). The
only experiment of neutron radiation under load in the
literature (Gray 1971) lacks the sufficient supporting data
and control measurements to be conclusive in that re-
spect. This effect is therefore neglected here due to lack
of knowledge about irradiation expansion under load.
The interfacial transition zone (ITZ) between the ag-
gregates and the cement paste is also neglected. As-
signing cohesive properties to the ITZ might control the
debonding between the cement paste and the aggregate
(e.g., Maruyama and Sugie (2014)). However, me-
chanical properties of the ITZ need to be calibrated on
experimental data, which were not readily available for
this experiment. In particular, there is no available data in
the open literature on the effects of irradiation on the ITZ
bonding properties.
The microstructure is discretized with a conforming
unstructured mesh using a Delaunay triangulation. The
calculations were run on a mesh containing ≈ 14,000
linear elements. The two-dimensional representation was
chosen as the model does not present strong anisotropy,
in terms of mechanical properties, loadings, or geometry.
However, it is recognized that this hypothesis might lead
to an earlier crack percolation compared to a
three-dimensional analysis.
The displacements of the bottom side of the sample are
fixed in the vertical direction. The central node of the
bottom side is fixed in the horizontal direction to prevent
global sliding of the sample. The microstructure and
mechanical boundary conditions are presented in Fig. 14.
The temperature, relative humidity and neutron fluence
fields are assumed homogeneous through the sample.
The four cases detailed in section 4 (See also Fig. 12) are
simulated: (a) variable temperature and flux, (b) variable
temperature and constant flux, (c) constant temperature
and variable flux, and, (d) constant temperature and flux.
5.2. Constitutive behaviors
Cement paste. The hcp is considered as a viscoelastic
material with isotropic linear damage and thermal ex-
pansion (with α p the coefficient of thermal expansion).
The creep deformation is divided into two compo-
nents: a short-term recoverable creep, modeled with a
Kelvin-Voigt unit, and a long-term irrecoverable viscous
flow, modeled with a time-dependent dashpot placed in
series (see Fig. 15). With this model, the long-term creep
behavior of the paste follows a logarithmic function of
time. The model relies on four parameters: the Young
modulus E p , the Poisson ratio v p , the creep viscosity
η p (measured as the slope of the creep curve in the
logarithmic scale), and the creep characteristic time τ p
Y. Le Pape, A. Giorla, J. Sanahuja / Journal of Advanced Concrete Technology Vol. 14, 70-86, 2016 79

Table 2: Constitutive parameters used in the simulations.

(1) Based on creep experiments for cement paste by Le
Roy (1995). (2) Assumption based on typical values for
cement paste and concrete. (3) Based on tensile strength
experiments for cement paste by Wittmann (1968). (4)
Typical value for concrete recommended by Bažant and
Steffens (2000). (5) Based on creep experiments for
cement paste by Wittmann (1970). (6) Value for
non-irradiated quartz obtained using the self-consistent
scheme on experimental data by Mayer and Lecomte
(1960). (7) Final value for the RIVE of quartz measured
by Bykov et al. (1981).
Cement paste
Ep [GPa] 12 (1)
vp [-] 0.2 (2)
αp [μm m-1 °C-1] 10 (2)
ft,p [MPa] 6 (3)
ε y ,t , p [mm m ]-1
0.75 (2)
rnl [mm] 0.3 (2)
η p [GPa d] 30 (1)
τ p [d] 20 (1)
Tact [K] 5,000 (4)
hc [-] 0.2 (5)
Aggregates
Ea [GPa] 95.7 (6)

Fig. 14 Concrete microstructure with aggregates in black,

v0 [-] 0.08 Eq. 11
cement paste in grey. vs* [-] 0.21 Eq. 11
a [-] 0.0063 Eq. 11
b [-] 1.938 Eq. 11
(which controls the instant at which the logarithmic be-
havior starts). Temperature accelerates creep with an ε max [%] 17.8 (7)
Arrhenius-type law (with Tact the activation temperature), aL [1020× ncm-1 °C-2] 0.024 Eq. 3
while relative humidity reduces the amplitude of the bL [1020× ncm-1] -0.062 Eq. 3
creep deformation (with hc a coefficient controlling the ac [1020× ncm-1 °C-2] 2.77×10-3 Eq. 3
loss of creep with relative humidity). Drying creep is 20 -1
bc [10 × ncm ] 0.128 Eq. 3
neglected in the current model, as the samples in Du-
brovskii et al. experiments have probably been pre-dried ξ [-] 0 Eq. 10
considering the reported weight loss. As the method of α 0* [μm m-1 °C-1] -0.52 Eq. 12
pre-drying (if any) was not reported by Dubrovskii et al., α1* [μm m-1 °C-1] 8.63 Eq. 12
no assumptions is made on the potential loss of me- ε c* [%] 0.85 Eq. 12
chanical properties and additional shrinkage due to dry-
ing.
ε L* [%] 4.27 Eq. 12
The damage is simulated with a non-local isotropic
linear damage adapted from the work of Dunant and neglected as the compressive stresses induced by the
Bentz (2015). The stress-strain curve of the material RIVE are well below the material strength. This as-
presents a linear softening branch after the peak (see Fig. sumption may not be valid for heterogeneous aggregates
15). This behavior is characterized by the strength at the with internal differential RIVE.
peak f t , p and the ultimate strain at the end of the frac- The mechanical and physical properties of the aggre-
ture process ε y ,t , p . Failure in compression is neglected, gates are taken as functions of the advancement of the
as most of the damage will be caused by the very high irradiation process, as measured by the current level of
tensile strains induced by the RIVE. RIVE, Eq. (11)-(12).
Aggregates. Aggregates are considered as purely The model uses the classical hypothesis of infinitesi-
elastic with thermal expansion and RIVE. The aggre- mal strain. This hypothesis may not be valid for the later
gates are furthermore considered homogeneous with no stages of the degradation, where the strains exceed one
internal defects. The failure behavior of the aggregates is percent.
Y. Le Pape, A. Giorla, J. Sanahuja / Journal of Advanced Concrete Technology Vol. 14, 70-86, 2016 80

Recoverable Non-Recoverable
Elasticity

100
creep creep Thermal expansion

η /τ
α(T-T 0 )
E η(1+t/ τ )

95
Effective properties [%]
90
η
250

continued

85
loading for
t > 100
200

80
Young's modulus
strain [arbitrary units]

Tensile strength
150

75
unloading at
100

t = 100 250 300 350 400 450 500 550 600

Temperature [K]
E η /τ
50

Fig. 16 Evolution of the Young's modulus and tensile

strength of the cement paste as a function of the maxi-
τ mum temperature reached in an element, following ex-
0

0.1 1 10 100 1000 perimental results by Fu et al. (2004).

time [d]

0.5 by Le Roy (1995). The Arrhenius coefficient for

σt,p creep is taken equal to 5,000 K as suggested in the
commonly used B3 model (Bažant and Steffens 2000).
The creep properties are reduced by a factor calibrated on
stress [arbitrary units]

Wittmann (1970) creep experiments on pre-dried sealed

cement paste samples with a water/cement ratio of 0.4.
The tensile strength of the material is adapted from
values given by Wittmann (1968) for dried cement paste
samples with a water/cement ratio of 0.45 and 0.6.
E The radius of the non-local averaging rnl is related to
ε y,t,p
the characteristic length of the heterogeneities in the
material (here, the cement paste) (Bažant and Pijaud-
0

0
ier-Cabot 1989), and is therefore taken as the radius of
strain [arbitrary units]
the largest sand particle neglected in the model. The
Fig. 15 Rheological model for the creep of cement paste elastic modulus and tensile strength of a cement paste
(top), typical creep curves for a loading at t = 0 showing element are reduced with the maximum temperature
the influence of the different components of the models reached in that element following the experimental re-
(using arbitrary parameter values) (middle) and sults of Fu et al. (2004) obtained on thin cement paste
stress-strain curve for a paste element (bottom). On the samples with a water/cement ratio of 0.5 (see Fig. 16).
creep curve E can be read from the initial deformation, The ultimate strain at rupture is dynamically changed
η from the asymptotic slope in the logarithmic scale, through the simulation to keep the fracture energy of
and τ as the intersection between the initial and final each element constant in absence of further data.
asymptotes. Aggregates. The mechanical properties of the aggre-
gates are taken as function of the irradiation, as described
in the previous sections. They are assumed to be inde-
5.3. Material properties
pendent of the temperature in the temperature range
The material properties of each phase are gathered in
recorded during the experiments. The Young's modulus
Table 2. Notations are explained in greater details in
(Giorla et al. 2015). of the quartz is taken as constant due to lack of data for
Cement paste. The material properties are taken from higher fluences. In a previous study (Giorla et al. 2015),
the authors found that the Young's modulus of the ag-
experimental results in the literature for cement pastes
gregate had no significant impact on the damage forma-
with a water/cement ratio of 0.5 as in Dubrovskii et al.
tion and propagation in the hcp during irradiation.
experiments whenever possible.
The elastic and creep properties at ambient tempera-
ture are provided from the basic creep experiments on 5.4. Results and discussion
sealed cement paste samples with a water/cement ratio of Two main outputs are derived from the simulations: (1)
Y. Le Pape, A. Giorla, J. Sanahuja / Journal of Advanced Concrete Technology Vol. 14, 70-86, 2016
the expansion of concrete, measured from the maximum
displacements variation between the opposite edges of
the sample, and, (2) the damage of the cement paste,
measured as the surface average of the damage scalar
variable over the elements representing the cement paste.
The corresponding results are plotted as a function of the
fluence on Figs. 17 and 18, respectively. The simulations
at variable temperature show an earlier increase of the
expansion and damage as a function of the fluence. When
the temperature is maintained constant, the expansion
and damage do not seem to depend on the neutron flux,
as both curves (constant and variable flux) coincide. All
simulations reach a plateau value in term of damage,
≈ 65%, which corresponds to the loss of integrity of the
sample: at that point the cement paste provides no me-
chanical restraint and the aggregates are free to swell.
Expansion. The volumetric expansion reaches values
between ≈ 15% and ≈ 18% depending on the studied
scenarios, which is consistent with the experimental
observations from Dubrovskii et al., considering that the
difference of actual aggregate volume fraction in the
experiment (0.70) and the numerical model (0.58). The
final expansion is about 30% higher than the integral of
the average strains in the sample, the later corresponding
to Eq. (13) assuming a volume fraction of aggregate of
58%.
This difference is explained by the severe mi-
cro-cracking and heterogeneities of the sample micro-
structure, notably induced by the presence of large ag-
gregates near its surface. It should be noted that gradient
effects are not considered in the present model, and may
further increase this difference. This has a direct practical
implication for the design of concrete irradiation ex-
periments, as it indicates that the maximum swelling of a
concrete sample might be higher than what would be
20
15
Concrete volumet ric expansion [%]
variable
constant T
constant Phi
constant T, Phi
10
5
0
0 5
Neutron fluence [1019 n/cm2]
Fig. 17 Volumetric expansion of the concrete as a func-
tion of fluence for the Group 2 at variable temperature
and neutron flux (thick solid line), constant temperature
(solid line), constant flux (dashed line), and constant
temperature and flux (dotted line).
81
expected from analytical models. The gap between the
sample and the irradiation capsule wall should account
for this effect in order to avoid wedging the samples into
the capsule, as what happened in Dubrovskii et al. ex-
periments.
Damage. Damage in the cement paste (Fig. 18) in-
creases rapidly in the sample after initiation. The neutron
fluence at which this rise occurs is significantly reduced
when the complete history of temperature is simulated.
Furthermore, the final stage of the damage propagation
(from 50% of damage in the cement paste to 65%) is
much sharper in the case of variable temperature than
constant temperature, as if it ranged from stable propa-
gation (constant temperature) to unstable (variable tem-
perature).
A closer analysis to the first damage increase reveals
that, when considering a variable temperature and neu-
tron flux, the initial damage increase occurs during the
first cooldown of the reactor, as shown in Fig. 19. During
that cooldown, the flux remains strictly positive due to
the hypothesis of linearity between flux and temperature
(see section 4): the small increase in neutron fluence,
coupled with the low temperature, leads to a sharp in-
crease of the RIVE. This behavior is also observed, albeit
with a smaller extent, during the second cooldown. Note
that this effect is not a modeling artifact as
quasi-identical results are obtained when assuming a
neutron zero flux for temperatures below 30°C, i.e.,
considering a complete reactor shutdown.
70
60
50
Damage in the cement paste [%]
40
30
20
variable
constant T
10
constant Phi
constant T, Phi
0
0 5 10 15 20 25
19
Neutron fluence [10 n/cm 2]
Fig. 18 Average damage in the cement paste as a func-
tion of fluence for Group 2 at: (a) variable temperature
and neutron flux (thick solid line), (b) constant tempera-
10 15 20 25
ture (solid line), (c) constant flux (dashed line), and, con-
stant temperature and flux (dotted line). The microstruc-
tures show the damage in the cement paste (in grey
level: black corresponds to a damage of 1.0; aggregates
are represented in blue) at 10%, 25% and 50% damage
in the cement paste for the case at constant temperature
and neutron flux.
Y. Le Pape, A. Giorla, J. Sanahuja / Journal of Advanced Concrete Technology Vol. 14, 70-86, 2016
temperature
100
variable Phi
constant Phi
80
Damage in cement paste [%]
60
40
20
0
0 20 40
Time [d]
Fig. 19 Average damage in the cement paste as a func-
tion of time for the Group 2 at variable temperature and
neutron flux (thick solid lines) and flux (dashed). Tem-
perature variations in red.
The difference between simulations at variable or
constant temperature has a critical impact on the analysis
of degradation of concrete structures exposed to irradia-
tion. Indeed, it indicates that the remaining service life of
such structure might be overestimated if the temperature
is averaged over time. Operation cycles, and most nota-
bly temperature cycles should be accounted for in the
analysis. Cooldown phases should be examined with care,
as a residual flux at low temperature might still be able to
initiate damage in the microstructure. The fact that the
evolution of the damage does not seem to depend on the
flux might lead to the hypothesis that creep does not play
a significant role in degradation caused by irradiation.
However, these simulations correspond to short-term
experiments compared to actual structures, with very
high deformations and without mechanical restraints, and
in which most of the damage occurs in the first months.
This may not be applicable for conditions representative
of the operation of nuclear power plants and requires
further analysis (Giorla and Le Pape 2015).
6. Conclusions
The irradiation of concrete biological shields in LWRs
occurs at relatively low temperature (< 65 °C) while the
range of irradiation temperature in test reactor varies
importantly from “low” temperature, i.e., < 40 °C) to
“high” temperature (> 150 °C). In that range of tem-
peratures, the amorphization rate, and thus, the RIVE
rate of quartz is significantly affected. Based on data
collected on quartz, RIVE rate appears to be significantly
accelerated when the irradiation temperature decreases.
The evolution of the mechanical properties (elastic
properties and CTE) of irradiated α-quartz are expressed
as a function of the current density of the material (that is,
the current level of RIVE) instead of the current fluence,
using analytic homogenization schemes. These proper-
ties have not been characterized in the literature for
82
RIVE beyond ≈ 30% of the maximum swelling, which
100
limits the predictive capabilities of the proposed model.
However, it should be noted that because of the impor-
tant rigidity contrast between the aggregate and the ce-
80
ment paste, the potential relative loss of elastic modulus
in irradiated aggregate may not lead to a significant
Temperature [oC]
60
modification of the damage level in the hcp, e.g., (Giorla
et al. 2015), as the cracking development is primarily
40
governed by the aggregate expansion amplitudes.
Based on the extensive data collected by Bykov et al.,
a model accounting for the coupled effects of tempera-
20
ture and irradiation irradiation on quartz RIVE has been
developed: RIVE rate is described as a function of tem-
0
60 80 100
perature. In the absence of neutron flux, thermal an-
nealing can be neglected in the range of temperature
observed in test reactors. After implementation in the
code AMIE, the simulation of Dubrovskii et al.'s ex-
periments on ordinary concrete made of quartz/quartzite
aggregate showed consistency with the observed expan-
sion, and pointed out that: (1) Measuring the dimensional
change of the diameter and/or height may lead to a sig-
nificant overestimation of the actual average volumetric
expansion in the specimen; (2) Assuming constant neu-
tron flux and temperature can lead to important dis-
crepancies in term of post-irradiation damage and
volumetric expansion when compared to more realistic
scenarios accounting for periodic reactor shutdowns or
other power cycles. Both effects certainly contribute to
the experimental scatter observed in data collected in the
open literature, and should be considered when analyzing
these data, as well as for future simulations of irradiated
structures.
While the coupled effects of temperature and neutron
irradiation of quartz expansion have been thoroughly
studied by Bykov et al. (1981), the same effects on other
physical and mechanical properties need to be studied to
obtain a thorough understanding of the potential effects
of differential RIVE in polyphasic, i.e., poly-minerals,
aggregates.
Nevertheless, qualitative outcomes in terms of engi-
neering applications can be made: (1) Tempera-
ture-induced point-defect annealing in irradiated miner-
als affects greatly the RIVE kinetics and occurs even at
low fluence. This effect questions the validity of using
directly radiation-induced aggregate expansion curve
obtained in test rector for LWRs conditions. In that par-
ticular perspective, it suggest that some test reactor data,
e.g., (Gray 1971), may be overly conservative because
obtained at about 45 °C, while other data should be re-
considered as relevant for LWRs operation because ob-
tained at quite high temperatures. (2) Because of the
effects of temperature on RIVE kinetics, LWR operation
cycles can potentially affect radiation-induced damage
development in the biological shield.
Acknowledgements
This research is sponsored by the U.S. Department of
Energy (DOE) Light Water Reactor Sustainability Pro-
Y. Le Pape, A. Giorla, J. Sanahuja / Journal of Advanced Concrete Technology Vol. 14, 70-86, 2016
gram. This manuscript has been authored by UT-Battelle,
LLC under Contract No. DE-AC05-00OR22725 with the
U.S. Department of Energy. The United States Gov-
ernment retains and the publisher, by accepting the arti-
cle for publication, acknowledges that the United States
Government retains a non-exclusive, paid-up, irrevoca-
ble, world-wide license to publish or reproduce the pub-
lished form of this manuscript, or allow others to do so,
for United States Government purposes. The Department
of Energy will provide public access to these results of
federally sponsored research in accordance with the DOE
Public Access Plan,
(http://energy.gov/downloads/doe-public-access-plan).
References
Amatuni, A. and Shevchenko, E., (1966). “Linear
thermal expansion on monocrystalline quartz and
aluminum oxide.” Measurement Techniques-USSR,
10, 1256-60.
Ballato, A., (2008). “Basic material quartz and related
innovations.” In: Piezoelectricity., Springer Berlin
Heidelberg. volume 114 of Springer Series in
Materials Science, 9-35.
Bažant, Z. and Steffens, A., (2000). “Mathematical
model for kinetics of alkali-silica reaction in
concrete.” Cement and Concrete Research, 30,
419-428.
Bazant, Z. P. and Pijaudier-Cabot, G., (1989).
“Measurement of characteristic length of nonlocal
continuum.” Journal of Engineering Mechanics, 115,
755-767.
Beattie, J. A., Blaisdell, B. E., Kaye, J., Gerry, N. T. and
Johnson, C. A., (1941). “An experimental study of the
absolute temperature scale. VIII. The thermal
expansion and compressibility of vitreous silica and
thermal dilation of mercury.” Proceedings of the
American Academy of Arts and Sciences, 74, 371-388.
Bykov, V., Denisov, A., Dubrovskii, V., Korenevskii, V.,
Krivokoneva, G. and Muzalevskii, L., (1981). “Effect
of irradiation temperature on the radiation expansion
of quartz.” Atomnaya Energiya, 51, 593-595.
Dormieux, L., Kondo, D. and Ulm, F. J., (2006).
“Microporomechanics.” John Wiley & Sons.
Dubrovskii, V., Ibragimov, S., Kulakovskii, M. Y.,
Ladygin, A. Y. and Pergamenshchik, B., (1967).
“Radiation damage in ordinary concrete.” Atomnaya
Énergiya, 23, 310-316.
Dunant, C. and Bentz, E., (2015). “Algorithmically
imposed thermodynamic compliance for material
models in mechanical simulations using the AIM
method.” International Journal for Numerical
Methods in Engineering, 104, 963-982.
Dunant, C. and Scrivener, K., (2010).
“Micro-mechanical modelling of alkali-silica
reaction-induced degradation using the AMIE
framework.” Cement and Concrete Research, 40,
517-525.
Dunant, C. and Scrivener, K., (2012). “Effects of
83
uniaxial stress on alkali-silica reaction induced
expansion of concrete.” Cement and Concrete
Research, 42, 567-576.
Elleuch, L., Dubois, F. and Rappeneau, J., (1972).
“Effects of neutron radiation on special concretes and
their components.” Special Publication of The
American Concrete Institute, 43, 1071-1108.
Eshelby, J., (1957). “The determination of the elastic
field of an ellipsoidal inclusion and related problems.”
Proceedings of the Royal Society, 241, 376-396.
Field, K., Remec, I. and Le Pape, Y., (2015). “Radiation
effects on concrete for nuclear power plants, Part I:
Quantification of radiation exposure and radiation
effects.” Nuclear Engineering and Design, 282,
126-143.
Fu, Y. F., Wong, Y. L., Poon, C. S., Tang, C. A. and Lin,
P., (2004). “Experimental study of micro/macro crack
development and stress-strain relations of
cement-based composite materials at elevated
temperatures.” Cement and Concrete Research, 34,
789-797.
Fuller, W. B. and Thompson, S. E., (1907). “The laws of
proportioning concrete.” Transactions of the
American Society of Civil Engineers, 59, 67-143.
Giorla, A. and Le Pape, Y., (2015). “Computing
creep-damage interactions in irradiated concrete.”
Journal of Nanomechanics and Micromechanics,
Submitted.
Giorla, A., Scrivener, K. and Dunant, C., (2014). “Finite
elements in space and time for the analysis of
generalised visco-elastic materials.” International
Journal for Numerical Methods in Engineering, 97,
454-472.
Giorla, A., Vaitová, M., Le Pape, Y. and Štemberk, P.,
(2015). “Meso-scale modeling of irradiated concrete
in test reactor.” Nuclear Engineering and Design, 295,
59-73.
Graves, H., Le Pape, Y., Naus, D., Rashid, J., Saouma, V.,
Sheikh, A., and Wall, J., (2014). “Expanded material
degradation assessment (EMDA) Volume 4: Aging of
concrete.” Technical Report NUREG/CR-7153,
ORNL/TM-2011/545. U.S. Nuclear Regulatory
Commission.
Gray, B., (1971). “The effects of reactor radiation on
cement and concrete.” In: Proceedings of an
Information Exchange Meeting on 'Results of
Concrete Irradiation Programmes', Commission des
Communautés Européennes, Brussels, Belgium.
Hilsdorf, H., Kropp, J. and Koch, H., (1978). “The
effects of nuclear radiation on the mechanical
properties of concrete.” Special Publication of The
American Concrete Institute, 55, 223-254.
Johnson, W. and Parsons, W., (1944). “Thermal
expansion of concrete aggregate materials.” Journal of
Research of the National Bureau of Standards, 32,
101-126.
Kelly, B., Brocklehurst, J., Mottershead, D. and
McNearney, S., (1969). “The effects of reactor
Y. Le Pape, A. Giorla, J. Sanahuja / Journal of Advanced Concrete Technology Vol. 14, 70-86, 2016
radiation on concrete.” In: Proceedings of the Second
Information Meeting on Pre Stress Concrete and
Reactor Pressure Vessels and their Thermal Isolation,
Brussels. 237-265.
Keppens, V. and, Laermans, C., (1996). “Influence of
heat treatment on the tunneling states in
neutron-irradiated quartz.” Physical Review, B 53,
14849-61.
Kosinski, J., Gualtereri, J. and Ballato, A., (1991).
“Thermal expansion of alpha quartz.” In: IEEE
Forty-Fifth Annual Symposium on Frequency Control,
Piscataway NJ.
Kröner, E., (1977). “Bounds for effective elastic moduli
of disordered materials.” Journal of the Mechanics
and Physics of Solids, 24, 137-155.
Larive, C., (1997). “Apports combinés de
l'expérimentation et de modélisation à compréhension
de l'alcali-réaction et de ses effets mécaniques.” Thesis
(PhD). Laboratoire Central des Ponts et Chaussées. (in
French)
Le Pape, Y., (2015). “Structural effects of
radiation-induced volumetric expansion on
unreinforced concrete biological shields.” Nuclear
Engineering and Design. 295, 534-548.
Le Pape, Y., Field, K. and Busby, J., (2014).
“Degradation of concrete for nuclear structures:
Identified mechanisms and knowledge gaps.”
Transactions of the American Nuclear Society, Reno
NV.
Le Pape, Y., Field, K. and Remec, I., (2015). “Radiation
effects in concrete for nuclear power plants - Part II:
Perspective from micromechanical modeling.”
Nuclear Engineering and Design, 282, 144-157.
Le Roy, R., (1995). “Déformations instantanées et
différées des bétons à hautes performances.” Thesis
(PhD). Ecole Nationale des Ponts et Chaussées.
Luikov, A., (1975). “Systems of differential equations of
heat and mass transfer in capilary-porous bodies
(review).” International Journal of Heat and Mass
Transfer, 18, 1-14.
Maruyama, I., Nishioka, Y., Igarashi, G. and Matsui, K.,
(2014). “Microstructural and bulk property changes in
hardened cement paste during the first drying
process.” Cement and Concrete Research, 58, 20-34.
Maruyama, I. and Sasano, H., (2014). “Strain and crack
distribution in concrete during drying.” Materials and
Structures, 47, 517-532.
Maruyama, I. and Sugie, A., (2014). “Numerical study
on drying shrinkage of concrete affected by aggregate
size.” Journal of Advanced Concrete Technology, 12,
279-288.
Mayer, G. and Gigon, J., (1956). “Effets des neutrons
rapides sur quelques constantes physiques du quartz
cristallin et de la silice vitreuse.” Le Journal de
Physique et le Radium, 18, 109-114.
Mayer, G. and Lecomte, M., (1960). “Effet des neutrons
rapides sur le quartz cristallin et la silice vitreuse.” Le
Journal de Physique et le Radium, 21, 846-852.
84
Multon, S. and Toutlemonde, F., (2006). “Effect of
applied stresses on alkali-silica reaction-induced
expansions.” Cement and Concrete Research, 36,
912-920.
Naus, D., (2005). “The effect of elevated temperature on
concrete materials and structures - A literature
review.” Technical Report ORNL/TM-2005/553. Oak
Ridge National Laboratory. Oak Ridge, TN.
Naus, D., (2010). “A compilation of elevated
temperature concrete mmaterial property data and
information for use in assessments of nuclear power
plant reinforced concrete structures.” Technical
Report NUREG/CR-7031, ORNL/TM-2009/175. Oak
Ridge National Laboratory.
Otto, J. and Thomas, W., (1963). “Die thermische
Ausdehnung von Quarzglas im Temperaturbereich
von 0 bis 1060 °C.” Zeitschrift für Physik, 175,
337-344. (in German)
Pabst, W. and Gregorová, E., (2013). “Elastic properties
of silica polymorph - a review.” Ceramics – Silikáty,
57, 167-184.
Pedersen, A., (1971). “Radiation damage in concrete -
measurements on miniature specimens of cement
mortar.” In: Benzler, H. Ed., Proceedings of an
Information Exchange Meeting on 'Results of
Concrete Irradiation Programmes, Commission of the
European Communities, Brussels, Belgium.
Primak, W., (1958). “Fast-neutron-induced changes in
quartz and vitreous silica.” Physical Review, 110,
1240-1254.
Reuss, A., (1929). “Berechnung der Fließgrenz von
Mischkristallen auf Grund der Plasticitätsbedingung
für Einkristalle.” Zeitschrift für Angewandte
Mathematik und Mechanik 9, 49-58. (in German)
Rosseel, T., Wall, J., Field, G., Le Pape, Y., Naus, D.,
Remec, I., Busby, J. and Bruck, P., (2014). “Radiation
damage in reactor cavity concrete.” In: Fontevraud 8 -
Contribution of Materials Investigations and
Operating Experience to LWRs' Safety, Performance
and Reliability, Avignon, France.
Saouma, V. and Perotti, L., (2006). “Constitutive model
for alkali-aggregate reactions.” ACI Materials Journal,
103, 194-2002.
Seeberger, J. and Hilsdorf, H., (1982). “Einuß von
radioactiver Strahlung auf die Festogkeit and Struktur
von Beton.” Technical Report NR 2505. Institut für
Massivbau and Baustofftechnologie, Universität
Karlsruhe. (in German)
Voigt, W., (1889). “Über die Beziehung zwischen den
beiden Elasticitätskonstanten isotroper Körper.”
Annalen der Physik, 274, 573-587. (in German)
Weissmann, S. and Nakajima, K., (1963). “Defect
structure and density decrease in neutron-irradiated
quartz.” Journal of Applied Physics, 34, 611-618.
Welche, P., Heine, V. and Dove, M., (1998). Negative
thermal expansion in beta-quartz.” Physics and
Chemistry of Minerals, 26, 63-77.
Wittels, M., (1957). “Structural behavior of neutron
Y. Le Pape, A. Giorla, J. Sanahuja / Journal of Advanced Concrete Technology Vol. 14, 70-86, 2016
irradiated quartz.” Philosophical Magazine, 2,
1445-1461.
Wittmann, F., (1968). “Surface tension shrinkage and
strength of hardened cement paste.” Materials and
Structure, 1, 547-552.
Wittmann, F., (1970). “Einuß des Feuchtigkeitsgehaltes
auf das Kriechen des Zementsteines.” Rheologica Acta,
9, 282-287. (in German)
Yano, T., Fukuda, K., Imai, M. and Miyazaki, H., (2007).
“Physical property changes of crystalline and
non-crystalline SiO2 due to neutron irradiation and
recovery by subsequent annealing.” Journal of
Nuclear Materials, 367-370, 730-735. Proceedings of
the Twelfth International Conference on Fusion
Reactor Materials (ICFRM-12).
Zhang, Q. and Ye, G., (2012). “Dehydration kinetics of
Portland cement paste at high temperature.” Journal of
Thermal Analysis and Calorimetry, 110, 153-158.
Zubov, V. and Ivanov, A., (1966). “Expansion of quartz
caused by irradiation with fast neutrons.” Soviet
Physics Crystallography, 11, 372-374.
Zubov, V. and Ivanov, A., (1967). “Elasticity of quartz
irradiated with fast neutrons.” Soviet Physics
Crystallography, 12, 313-314.
Appendices
A. Appendix: Micromechanical Estimates of
Thermoelastic Properties of Quartz Polycrystal
Numerical computations at the concrete scale require the
thermomechanical properties of aggregates as a function
of irradiation. These properties are well documented for
the quartz monocrystal Mayer and Lecomte (1960). The
aim of this appendix is to upscale these properties from
the quartz monocrystal to the aggregate, assumed to be
made up of a polycrystalline arrangement of quartz
crystals. The properties of interest are the stiffness and
the coefficient of thermal expansion.
Introduction to Mean-Field Homogenization of Ther-
moelastic Characteristics
The general framework of homogenization of elastic
then thermoelastic properties is briey recalled here. For
more details, see for example Dormieux et al. (2006).
A heterogeneous representative elementary volume
(rev) is considered, the behaviour at each point z being
linear elastic:
σ ( z ) = ^( z ) : ε ( z )
To formulate the effective behaviour, this rev is sub-
mitted to kinematic uniform boundary conditions, ξ ( z )
= E ⋅ z , where E is the macroscopic strain. The linearity
of the local behaviour (14) allows to introduce the strain
localization tensor A relating the microscopic ε and
macroscopic E strain tensors:
ε ( z ) = A( z ) : E
85
The macroscopic stress Σ being the spatial average
of the microscopic stress field σ , the effective behav-
iour of the REV is found to be linear elastic:
Σ = ^ eff : E with ^ eff = ^( z ) : A( z ) (16)
Ω
where Ω
denotes the spatial average of a field over
the REV Ω . The local behaviour is now considered as
linear thermoelastic:
σ ( z ) = ^( z ) : (ε ( z ) − α ( z )ΔT ) (17)
where the field of temperature increment ΔT is as-
sumed to be homogeneous across the REV. Alternatively,
this behaviour can be written using the eigenstrain ε L
σ ( z ) = ^ ( z ) : (ε ( z ) − ε L ( z )) with ε L ( z ) = α ( z )ΔT (18)
or, introducing the eigenstress σ p :
σ ( z ) = ^( z ) : ε ( z ) + σ p ( z )
(19)
with σ p ( z ) = −^( z ) : ε L ( z ) = −^( z ) : α ( z )ΔT
The effective behaviour of the REV is found to be
linear thermoelastic (Levin's theorem):
∑ = ^eff : E + ∑ effp
with ^eff = ^( z ) : A( z ) Ω
(20)
and ∑ effp = σ p ( z ) : A( z )
Ω
A( z ) still being the strain localization tensor, de-
fined in (15), where σ p = 0. The effective behaviour
can also be written in terms of the tensor of thermal
expansion:
∑ = ^ eff : ( E − α eff ΔT
(21)
with α eff = ^ eff −1 : α ( z ) : ^( z ) : A( z ) Ω
Upscalling from Crystal to Polycrystal
A specific case of rev is considered here: a polycrystal
made up of many crystals, each crystal differing from the
others only through its orientation. The following as-
(14) sumptions are considered: (1) Perfect interfaces between
crystals (continuity of displacement); (2) At the crystal
scale, anisotropy only regards the stiffness and thermal
expansion tensors, not the crystal shape; and, (3) Iso-
tropic orientation distribution of crystals, which yields an
isotropic effective behavior.
The orientation of one crystal is described by the
frame ( u1 , u2 , u3 ) or the angles θ , φ , ψ (see Fig. 20).
Each crystal stiffness tensor ^ cr (θ , φ , ψ ) thus has the
(15) same components in the frame (u1, u2, u3) associated to
the crystal. This is also the case for the tensor of thermal
Y. Le Pape, A. Giorla, J. Sanahuja / Journal of Advanced Concrete Technology Vol. 14, 70-86, 2016
expansion α cr (θ , φ , ψ ) .
Voigt and Reuss Bounds and Estimates
The classical Voigt and Reuss bounds on the poly-
crystalline stiffness ^ pc are obtained from an average
over the whole rev of the stiffness or compliance tensor:
−1
^ cr (θ , φ , ψ ) = ^ ≤ ^ pc ≤ ^ = ^ cr (θ , φ , ψ )
−1 R V
or
where the ' ≤ ' operators are considered in the sense of
quadratic forms. The averaging operator ' or . ' consid-
ers the isotropic orientation distribution, and is computed
as:
2π 2π π sin θ
T =∫ ∫ ∫ T(θ , φ ,ψ )
or . ϕ = 0 φ =0 θ = 0 8π 2
with any fourth order tensor T which depends on orien-
tation θ , φ , ψ only through the frame ( u1 , u2 , u3 ). The
tensors ^ pc , ^ R and ^V being isotropic, the inequal-
ity (22) translates into inequalities on the effective bulk
and shear moduli:
κ R ≤ κ pc ≤ κ V and g R ≤ g pc ≤ g V
The classical formulas relating the bulk and shear
moduli to the Young's modulus and Poisson ratio yield
Voigt and Reuss estimates of the latter (but not strictly
bounds).
To get an estimate of the effective tensor of thermal
expansion, an estimate of the strain localization tensor is
required, to use (20). The Voigt estimate considers that
the strain field is homogeneous in the whole REV,
ε (z) = E :
AV (θ , φ ,ψ ) = I
The Reuss estimate considers that the stress field is
homogeneous in the whole rev, σ ( z ) = ∑ :
A (θ , φ ,ψ ) = ^ cr (θ , φ ,ψ ) : ^
R −1 R
Due to isotropy of crystal orientation, the tensor of
thermal expansion is also isotropic, and its estimates
read:
α V 1 = ^V −1 : ^ cr (θ , φ ,ψ ) : α cr (θ , φ ,ψ )
and α R 1 = α cr (θ , φ ,ψ ) or .
where the averaging operator ' or .
(23) for second order tensors.
Self-Consistent Estimate
Contrary to Voigt and Reuss bounds, so-called ho-
mogenization schemes provide estimates of the effective
86
or .
(22)
Fig. 20 Characterization of the orientation of one crystal
by angles θ , φ ,ψ : the frame ( u1 , u2 , u3 ) associated to
the crystal is de_ned from the reference frame ( e1 , e2 , e3 )
dθ dφ dψ (23)
through the spherical frame ( er , eθ , eφ ), appearing as
dotted lines.
stiffness. These schemes allow estimate of the average
strain localization tensor over each phase of the rev.
Among these schemes, the self-consistent one (Kröner
1977) is particularly suited to polycrystals. Indeed, the
average strain in each crystal is estimated as the strain
arising in a sphere having the same stiffness, embedded
(24) into an infinite medium whose stiffness is the sought
effective stiffness ^ SC , and submitted at infinity to a
reference strain E 0 . The solution to this so-called aux-
iliary problem is given by Eshelby (1957):
ε cr (θ , φ ,ψ ) = [I + PSC : (^ cr (θ , φ ,ψ ) − ^ SC )]−1 : E 0 (28)
where PSC is the Hill tensor of a sphere in a medium of
stiffness ^ SC . The strain E 0 at infinity can be related
to the strain E at the boundary of the rev using the
average strain rule ε cr (θ , φ ,ψ ) or . = E . The strain lo-
calisation tensor on crystals oriented along θ , φ ,ψ is
(25) thus estimated by:
A cr (θ , φ ,ψ ) = [I + PSC : ...(^ cr (θ , φ ,ψ ) − ^ SC )]−1 : ...
−1 (29)
[I + PSC : (^ cr (θ , φ ,ψ ) − ^ SC )]−1
or .
(26)
The self consistent estimate of the effective stiffness is
then obtained from (16):
^ SC = ^ cr (θ , φ ,ψ ) : [I + PSC : (^ cr (θ , φ , ϕ ) − ^ SC )]−1 :
or .
−1
(30)
: (^ cr (θ , φ ,ψ ) − ^ )]
−1
... [I + P
SC SC
or .
or .
(27) The latter is a non linear tensorial equation, as the Hill
tensor PSC depends on the sought self-consistent stiff-
ness ^ SC . As the latter is isotropic, and as the average
' is also defined by operator yields isotropic tensors, this tensorial equation
can be reduced to two non linear equations over κ SC
and g SC , by projection over J and K . The
self-consistent estimate α SC of the coefficient of ther-
mal expansion is obtained from (21) and (29).
Journal of Advanced Concrete Technology Vol. 14, 144-162 April 2016 / Copyright © 2016 Japan Concrete Institute 144

Scientific paper

A Numerical Model for Concrete Strength Change under Neutron and

Gamma-ray Irradiation
Ippei Maruyama1*, Kazumoto Haba2, Osam Sato3, Shunsuke Ishikawa4, Osamu Kontani5 and
Masayuki Takizawa6

Received 24 January 2016, accepted 11 April 2016 doi:10.3151/jact.14.144

Abstract
For performance evaluation of existing reinforced concrete members under irradiation conditions, a numerical code called
“DEVICE” (Damage EValuation for Irradiated ConcretE), which takes into account the heat, moisture, and radiation
transport coupled with cement hydration, is proposed. This code is composed of the established computational ce-
ment-based material (CCBM) model and the one-dimensional deterministic transport Sn code “ANISN”. In the proposed
model, temperature-dependent irradiation-induced expansion of aggregate minerals and resultant strength deterioration of
concrete are introduced. Currently, the knowledge and modeling of irradiation-induced expansion of aggregate mineral is
limited only for α-quartz. DEVICE was used for evaluating the strength distribution of the decommissioned plant Japan
Power Demonstration Reactor (JPDR). Compressive strength distribution in a concrete biological shielding (CBS) wall of
the JPDR was obtained by core sampling, and the compressive loading test results were compared with the calculation
results. This comparison proved the practicality potential of DEVICE to predict the concrete strength distribution in a
CBS. In addition, concrete strength change and its distribution in a CBS of an anonymous two-loop pressurized water
reactor was simulated by DEVICE. The contributing factors for the change in the distribution of concrete strength at the
inner surface of the CBS are discussed. Furthermore, the ways of integrity evaluation other than the existing allowable
fast neutron fluence method are proposed and discussed as follows: 1) mineral composition-based allowable fast neutron
fluence; 2) strength prediction at the inner surface based on the expansion of mineral composition of aggregates and the
lower limit curve of the ratio of compressive strength of the specimen after irradiation (Fc) to that of the reference
specimen (Fco) as a function of concrete expansion; and 3) direct numerical calculation for seismic performance by
considering irradiation-induced volume expansion and degradation of concrete.

1. Introduction long-term operation of nuclear power plants (NPPs) will

Light water reactors (LWRs) are considered as an im-
portant power supplier even after the Fukushima Daiichi
accident. According to “Roadmap on R&D and Human
Resource for Light Water Reactors Safety in Japan,”
presented at the IAEA Meetings 2015, (Sekimura 2015)
and a Japanese cabinet decision passed on April 11, 2014,
nuclear power is an important base load power. Safe
1
Assoc. Professor, Graduate School of Environmental
Studies, Nagoya University, Nagoya, Japan.
*Corresponding author,
E-mail: ippei@dali.nuac.nagoya-u.ac.jp
2
Engineer, Safety Technology Policy Department, MRI
Research Associates, Inc., Tokyo, Japan.
Current position: Assistant Manager, Nuclear Facilities
Division, Taisei Corporation, Tokyo, Japan.
3
Senior Researcher, Nuclear Safety Research Division,
Mitsubishi Research Institute, Inc., Tokyo, Japan.
4
Engineer, Nuclear Power Department, Kajima Corpora-
tion, Tokyo, Japan.
5
Senior Manager, Nuclear Power Department, Kajima
Corporation, Tokyo, Japan.
6
General Manager, Nuclear Safety Research Division,
Mitsubishi Research Institute, Inc., Tokyo, Japan.
be required in the near future because construction of
new NPPs will be considerably difficult in Japan. From
this viewpoint, concrete structures in NPPs play a crucial
role because they are practically irreplaceable. Therefore,
aging management of reinforced concrete structures is
necessary. Reinforced concrete degradation can be
categorized into two types: rebar degradation due to
corrosion and concrete degradation due to freeze–thaw
cycles, alkali–silica reaction in aggregates, delayed et-
tringite formation, and chemical degradation by acids or
sulfate salts. These deterioration types, which are com-
monly observed in civil structures, have been extensively
researched. However, in NPPs, a characteristic deterio-
ration environment, i.e., irradiation, exists. Neutrons and
gamma rays are irradiated from the reactor on concrete
biological shielding (CBS) walls (in case of a pressurized
water reactor; PWR) or reactor vessel support pedestals
(in case of a boiling water reactor; BWR) during opera-
tion.
The impact of neutron and gamma-ray irradiation on
the physical properties of concrete has been summarized
by Hilsdorf et al. (1978). Their survey indicated that the
compressive strength of a concrete structure deteriorates
by fast neutrons (>0.1 MeV) above a fluence of 1 × 1019
n/cm2, and it is not affected significantly by gamma rays
I. Maruyama, K. Haba, O. Sato, S. Ishikawa, O. Kontani and M. Takizawa / Journal of Advanced Concrete Technology Vol. 14, 144-162, 2016 145

Table 1 Examples of expected fast neutron fluence and gamma-ray dose of Japanese NPPs after 60-year operation. (This
table is based on the information at 24th Jan. 2016.)
Type Name of plant Location Fast neutron (n/cm2) Gamma-ray dose (Gy)
BWR Fukushima 1st -1 Outside of reactor 3.0 × 1018 *1
< 2.0 × 108
Fukushima 1st -2 pressure vessel. 3.0 × 1018 *1 < 2.0 × 108
Shimane 1 3.0 ×1018 *1 8.0 × 107
Tsuruga 1 1.1 × 1017 *2 < 2.0 × 108
Fukushima 1st -5 Reactor pressure vessel - 2.8 × 1014 *1 2.3 × 104
Hamaoka 1 pedestal structure. 5.7 × 1013 *1 1.6 × 104
Fukushima 1st -3 1.6 × 109 *1
2.2 × 105
19 *3
PWR Mihama -2 Internal surface of the 6.2× 10 1.8 × 108
19 *3
Ikata -1 first biological shielding 6.0× 10 2.8 × 108
19 *3
Genkai -1 wall. 4.8× 10 2.8 × 108
19 *3
Mihama -3 4.7× 10 2.5 × 108
Takahama -1 4.5× 1019 *3
2.3 × 108
Takahama -2 4.5× 1019 *3
2.4 × 108
19 *3
Mihama -1 3.6× 10 1.5 × 108
: Decommisioning sites. *1 : ≥ 0.1 MeV, *2 : ≥ 1.0 MeV, *3 : ≥ 0.11 MeV.

although a decreasing trend above a dose of 5 × 105 kGy
was reported. This result is almost consistent with the
results of Dubrovskii et al. (1966, 1967, 1968, 1970),
which concluded that compressive strength might be
affected above a total fluence of ~5 × 1019 n/cm2. On the
contrary, Fujiwara et al. (2009) and Kontani et al. (2010)
pointed out that Hilsdorf et al.’s (1978) experimental
results included concrete or mortar that are not suitable
for LWRs. Therefore, Maruyama et al. (2013) summa-
rized the relationship between neutron fluence and rela-
tive compressive strength ratio (Fc/Fco, ratio of the
compressive strength of irradiated concrete to that of the
reference concrete) for concrete or mortar applicable to
concrete members in LWRs. They confirmed clear de-
creasing trend of Fc/Fco above 1–3 × 1019 n/cm2. Re-
cently, Field et al. (2015) summarized more compre-
hensive data of irradiated concrete and its components,
which also support the decreasing trend of physical prop-
erties of irradiated concrete as neutron fluence increases.
Several studies (Dubrovskii et al. 1967; Elluech et al.
1971; Pedersen 1971; Hilsdorf et al. 1978; Kontani et al.
2010; Maruyama et al. 2012; Field et al. 2015) have
suggested that the mechanism of concrete deterioration
due to irradiation is aggregate expansion that originates
from metamictization or amorphization of silicious rock
minerals due to irradiation (Primak et al. 1955; Primak
1958; Elluech et al. 1971; Eby et al. 1992; Bonnet et al.
1994; Douillard and Duraud 1996; Bolse 1999; Ewing et
al. 2000). Due to aggregate expansion, the mortar matrix
in concrete is damaged, resulting in deterioration of
concrete’s physical properties, such as compressive
strength and Young’s modulus, such as in the case of
alkali–silica reaction in concrete (Giorla et al. 2015; Le
Pape et al. 2015).
The Nuclear Regulatory Authority (NRA) of Japan has
recommended a concrete integrity evaluation process for
NPPs, called “plant life management (PLM),” which
needs to be conducted for every 10 years after 30 years of
operation of an NPP. NRA recognized 1.0 × 1020 n/cm2
for fast neutrons and 2.0 × 1010 rad (= 2.0 × 105 kGy) for
gamma-ray irradiation as reference levels. The American
Concrete Institute (2012) Committee cites Hilsdorf et
al.’s results and suggests a fluence of 1.0 × 1021 n/cm2 at
which loss of mechanical properties is observed; it also
recommended a conservative fluence limit of 1.0 × 1017
n/cm2 to prevent lifetime radiation-induced degradation of
concrete. The British Standards Institute (1990) noted a
neutron radiation level of 5.0 × 1017 n/cm2 below which its
effect on concrete structures of an NPP is not significant.
Table 1 lists the expected fast neutron fluence of
Japanese NPPs after 60 years of operation, as obtained
from the PLM document. Clearly, although fast neutron
fluence levels of all plants will not exceed NRA’s refer-
ence level, the fluence levels for some NPPs are expected
to exceed Hilsdorf et al.’s (1978) value. Consequently,
scientific knowledge is required for addressing this issue.
NRA launched a national project (Maruyama et al. 2012)
with this objective. Irradiation acceleration experiments
have been limited with respect to specimen size, tem-
perature control of the specimen, and ratio of gamma-ray
to neutron flux (Maruyama et al. 2013). Moreover, large
gaps exist between the specimens and the actual concrete
structures in a plant. In the present study, a numerical
simulation model is proposed to compensate part of such
gaps. A new numerical code called DEVICE (Damage
EValuation for Irradiated ConcretE) is developed by cou-
pling the computational cement-based material model
(CCBM) (Maruyama and Igarashi 2015), and moisture,
heat, and radiation transport codes for predicting com-
pressive strength changes and their distribution in a mas-
sive concrete member subjected to irradiation on one side.
This model succeeds the current PLM process and aims to
contribute to the material basis or material potential
evaluation for integrity of reinforced concrete structures
during PLM. However, radiation impact on the structural
performance is not addressed through this model and is a
I. Maruyama, K. Haba, O. Sato, S. Ishikawa, O. Kontani and M. Takizawa / Journal of Advanced Concrete Technology Vol. 14, 144-162, 2016
major concern (Pomaro et al. 2011; Salomoni et al. 2014;
Le Pape 2015). Therefore, this study focuses on material
strength, while similar approaches focused on the damage
in CBS walls (Pomaro et al. 2011; Salomoni et al. 2014).
The proposed model is used for evaluating the de-
commissioned plant Japan Power Demonstration Reactor
(JPDR). The strength of concrete core samples taken
from a CBS of the JPDR is evaluated and discussed by
comparing with the results obtained by DEVICE. Finally,
concrete strength and related properties of concrete in a
CBS of an anonymous PWR are calculated and the results
are discussed from the viewpoint of integrity evaluation.
2．Numerical modeling
2.1 Coupling of the CCBM and neutron and
gamma-ray transport calculation
In the present study, concrete behavior is calculated by
using the CCBM which covers the models for the rate of
hydration of cement minerals, phase composition, re-
sultant hygro-thermo-mechanical properties of cement
paste, i.e., compressive strength, Young’s modulus,
Poisson’s ratio, thermal expansion coefficient, autoge-
nous shrinkage, drying shrinkage, heat capacity, heat
transfer coefficient, water vapor sorption isotherms, and
water transfer coefficient. In addition, the model em-
pirically evaluates changes in the concrete compressive
strength under heating or drying and takes into account
cement paste strength changes due to its colloidal feature
as well as micro-defects produced around aggregates by
the volume difference between the aggregate and mortar.
For modeling the target concrete member, heat and
moisture transfer problems are calculated using one
dimensional finite differential method (1D-FDM) with
an implicit method and a time step of 0.001 hour. All the
parameters such as moisture capacity, moisture transfer
coefficient, water consumption, heat capacity, heat
transfer coefficient, and heat consumption are calculated
by the information of equilibrium and history of tem-
perature and relative humidity of the concrete at each
Fig. 1 Detailed process of introduction of ANISN-ORNL to CCBM.
146
node, and phase composition of cement paste from the
previous step.
Considering radiation conditions on concrete, a de-
terministic transport theory is used for the present study.
In the ANISN code, originally developed in Oak Ridge
National Laboratory (ORNL) (Engle Jr. 1967), the
Boltzmann transport equation, which is an inte-
gro-differential balance equation for the conservation of
radiative energy, is calculated using the discrete coordi-
nates (Sn) approximation by discretizing spatial, angular,
energy, and variables. This 1-D ANISN code is coupled
with the CCBM in the present study. The neutron and
gamma-ray flux distributions in concrete are calculated
with the ANISN-ORNL by using the spectra of radiation
energy at the inner surface of the concrete and the
cross-sections of the concrete member based on the ma-
terial compositions and temperature. The energy spectra
at the inner surface of the concrete must be calculated
separately using a nuclear reactor model. In this system,
the ANISN calculation region is divided into five zones,
in each of which the material compositions and tem-
perature are averaged. The ANISN calculation involves
P3-S16 approximation and a multi-group cross-section
library V7-200N47G in SCALE6 that has 200 neutron
groups and 47 gamma-ray groups.
The detailed procedure is described as follows:
1) Create input data “GIP” (Rhoades 1978), which is a
function to calculate the cross-section of a concrete
member on the basis of the atomic composition at each
FDM node.
2) Calculate the cross-section of a concrete member at
each FDM node from the temperature and atomic
composition data using “GIP”.
3) Calculate the radiation flux at all FDM nodes, which
correspond to the calculation nodes for the ANISN
code, based on the cross-section of the concrete
member.
4) Calculate heat production rate at each node from the
corresponding ANISN calculation results.
5) Calculate heat and moisture transfer and hydration
I. Maruyama, K. Haba, O. Sato, S. Ishikawa, O. Kontani and M. Takizawa / Journal of Advanced Concrete Technology Vol. 14, 144-162, 2016
process of cement, and then go to step 1).
Figure 1 shows the process of introduction of the
ANISN codes to the CCBM. Figure 2 shows the cou-
pling process and the calculation flow.
2.2 Modeling of concrete deterioration due to
irradiation
2.2.1 Impact of gamma rays
First, the impact of gamma rays is discussed based on
literature (Kontani et al. 2012; Maruyama et al. 2012;
Kontani et al. 2013).
Gamma rays affect material through electronic exci-
tations. The excess electronic energies of electrons in the
excited states are transformed to vibrational energies of
atoms composing the materials, resulting in breaking of
the covalent bonds and producing radicals. From this
viewpoint, the covalent nature of the concrete materials
can be affected by gamma rays. In case of the Si–O bonds
in α-quartz, the impact is significant beyond ~1.0 ｘ
1012 Gy for gamma rays (Douillard and Duraud 1996;
Ichikawa and Kimura 2007). Hydroxy groups are also
affected by gamma rays. Water is decomposed as hy-
droxyl and hydrogen radicals, forming hydrogen, oxygen,
and hydrogen peroxide (Bouniol and Aspart 1998).
As the energy of gamma rays are absorbed by the
target material during the decay process, heating always
accompanies gamma-ray irradiation. In addition to ele-
vated temperature due to gamma-ray irradiation, de-
composition of water enhances drying of the concrete.
Therefore, the combined impacts of drying and heating
alter the physical properties of concrete.
Hardened cement paste, which is a product of hydra-
tion of Portland cement and water, is the main binder
matrix in concrete. Cement is generally composed of
alite, belite, aluminate phase, ferrite phase, and gypsum.
Major cement hydrates that are responsible for the
strength of hardened cement paste are portlandite and
Fig. 2 Calculation flow of CCBM and ANISN-ORNL, their interaction, and prediction of physical properties of concrete.
147
calcium silicate hydrates (C–S–H), which are formed
from alite and belite. Portlandite and the Ca–O layer in
C–S–H are ionically bonded and may not be largely
affected by gamma rays without the interaction of water.
The Si–O covalent bonds in C–S–H may be affected
slightly, but the structure of C–S–H is sustained by the
Ca–O layer (Gartner 1997; Gartner et al. 2000). Even
though the atoms in C–S–H are dislocated by secondary
electron collision, since the Ca–O layer has a winding or
wavy shape and is not ordered because of the attachment
of silicates (Gartner 1997), such dislocations do not
largely impact the physical properties of hardened ce-
ment paste.
Metastable calcium peroxide octahydrates can be
produced in hardened cement paste with high water
content (Bouniol and Aspart 1998; Vodák et al. 2005;
Lowinska-Kluge and Piszora 2008), altering the strength
and pore size distribution and especially resulting in
subsequent carbonation.
The drying and heating associated with gamma-ray
irradiation affects the properties of hardened cement
paste (Maruyama et al. 2014a), because C–S–H formed
from Portland cement has a colloidal feature, and the
morphology of C–S–H is affected by the drying and
heating process (Pihlajavaara 1974; Litvan 1976, Parrott
et al. 1980; Parrott and Young 1981; Cong and
Kirkpatrick 1995; Thomas and Jennings 2003; Jennings
2008; Thomas et al. 2008; Maruyama et al. 2014a; Ma-
ruyama et al. 2015). Therefore, gamma heating largely
impacts the physical property of hardened cement paste.
Natural rock and chemically stable aggregates are
generally used in concrete for LWRs. Unlike hardened
cement paste, an aggregate is composed of
well-crystallized minerals, and the major bonds of rock
minerals are covalent and ionic bonds. Therefore, ag-
gregates can be considered to be not largely affected by
gamma rays through electronic excitations.
I. Maruyama, K. Haba, O. Sato, S. Ishikawa, O. Kontani and M. Takizawa / Journal of Advanced Concrete Technology Vol. 14, 144-162, 2016
Collision of gamma rays on a material surface results
in a photoelectric effect, Compton scattering, or pair
production, and thereby causes electrons, and occasion-
ally positrons, to be ejected or produced. These electrons
collide with the material and eject more electrons, the
secondary electrons; this process continues until the
energies of the secondary electrons become lower than
the ionization energy of the material. Only when the first
gamma-ray energy is very high, the secondary electrons
can collide and move atoms in target materials.
Reproducible facts indicate that α-quartz, which is one
of the major rock minerals for concrete aggregates, may
undergo expansion due to irradiation. Fully metamictized
or amorphized condition is attained at ~6 × 1027 eV/g (=
9.6 × 1012 Gy) by electron irradiation. Electron damages
are determined by the ionization cross-section; the
damage decreases with increasing electron energy.
However, if the electron energy becomes more than 1.5
MeV, the irradiation process is accelerated by a knock-on
displacement mechanism (Inui et al. 1990).
Maximum ejected electron energy of photoelectric
effect and Compton scattering is almost the same as the
energy of irradiated gamma-ray, when energy of an in-
cident electron is more than MeV, while it is 0.51 MeV in
case of pair production. The possibility of those events is
a function of atomic mass and therefore, the Compton
scattering event which will be brought by gamma-ray
energy ranging about 1.8 ~ 20 MeV should contribute to
the amorphization of α-quartz through recoiled electrons.
In case of a concrete structure and neighboring surfaces
of an LWR, the gamma-ray levels are 100 KeV to 10
MeV (Fukuya et al. 2002), therefore, gamma rays are
considered to have some effect on the concrete compo-
nents. This impact is not fully understood quantitatively
and needs further investigation for clarity, but based on
the experiment by Inui et al. (1990), the impact of
Compton recoiled electron seemed almost the same order
of that by electronic excitations, therefore, the rate of
amorphization due to gamma-ray irradiation is negligible
in case of aggregate or hardened cement paste in LWR.
In the present study, the impact of gamma rays on
hardened cement paste and concrete is assumed to be the
same as the impact of heating and drying on these mate-
rials, which has been already modeled in previous studies
(Maruyama et al. 2014a, 2014b; Maruyama and Igarashi
2015). The existing data (Vodák et al. 2005) also support
this assumption. Production of calcium peroxide octa-
hydrate must be related to the portlandite and water
content in the cement paste. However, in the case of
mature cement paste under a sealed condition, only a
small amount of capillary water exists (Muller et al.
2012), and the possibility of a reaction may be very lim-
ited due to lack of space for precipitation. In addition,
due to lack of experimental experiences, we will not
consider this phenomenon at the present research stage.
The production of gases by the decomposition of capil-
lary water may also impact the water vapor movement in
concrete, but this phenomenon is also not considered here
148
because of lack of scientific data (Kontani et al. 2013).
2.2.2 Impact of neutrons
Since neutrons are electrically neutral, they do not have
an ionizing effect. Neutrons always interact with a nu-
cleus by scattering and absorption. Two types of scat-
tering occur: inelastic and elastic scattering. Both change
the direction and speed of neutrons and the number of
protons and neutrons in the interacted nucleus remains
the same as before. In elastic scattering, a neutron col-
lides with a lighter nucleus such as that of H, C, or O and
a neutron is ejected, thus preserving the total kinetic
energy of the nucleus. On the contrary, in inelastic scat-
tering, a neutron collides with a heavier atomic nucleus,
which attains an excited state owing to internal rear-
rangement and subsequently releases radiation. When a
neutron is absorbed by a nucleus, a wide-energy-range
radiation can be released, or fission can be initiated.
Cement hydrates always coexist with evaporable wa-
ter under the environmental condition around the con-
crete in LWRs, and the dissolution and precipitation of
these hydrates are in equilibrium. Atoms in the cement
hydrates can be collided by neutrons, dislocating atoms
from the hydrates. However, the equilibrium state plays a
healing role for this neutron damage in cement hydrates.
In addition, C–S–H generally has a short-period recur-
sive structure that cannot be detected by powder X-ray
diffraction in the layer-piling direction; therefore, the
defects in the structure may not largely impact its
physical properties. The only possible damage by the
collision of neutrons is that the portlandite, with its
maximum crystal size being approximately 1 micron,
may expand because of the dislocation of atoms and
defects in its structure (Seeberger and Hilsdorf 1982).
However, the change in hardened cement strength under
irradiation conditions is not confirmed experimentally
(Gray 1971; Elleuch et al. 1972; Seeberger and Hilsdorf
1982).
As stated in the previous section, aggregates are well
crystalized, but there is a facts that α-quartz is expanded
due to irradiation. Previous neutron irradiation experi-
ments of α-quartz showed almost consistent density of
the final state (Wittels and Sherrill 1954; Primak et al.
1955; Primak 1958; Bykov et al. 1981; Bonnet et al.
1994), with its reduction value in the range 15%–18%
and a linear expansion of approximately 5%-6%. Neu-
tron irradiation damages are generally described ac-
cording to three fluence levels (Douillard and Duraud
1996). For low neutron fluence (φn ≤ 1 × 1018 n/cm2), the
irradiation induces point defects, mainly E’ centers (≡Si•)
and oxygen hole centers. The point defect concentration
increases as a power function of the fluence. For medium
neutron fluence (1 × 1018 ≤ φn ≤ 1 × 1020 n/cm2), α-quartz
undergoes an α–β transition, immediately followed by a
complete amorphization for a neutron fluence greater
than 1 × 1020 n/cm2 (Douillard and Duraud 1996).
Note that to the best of authors’ knowledge, the final
amorphized conditions attained after electron irradiation
I. Maruyama, K. Haba, O. Sato, S. Ishikawa, O. Kontani and M. Takizawa / Journal of Advanced Concrete Technology Vol. 14, 144-162, 2016
or ion bombardment have not been compared with those
after neutron irradiation in terms of α-quartz density.
Therefore, we could not conclude that gamma rays have
the same impact as neutrons.
α-Quartz is the major component of rocks and is pre-
sent in many kinds of aggregate. Therefore, aggregate is
likely expanded under the neutron irradiation. This ag-
gregate expansion in concrete leads to expansion of
concrete, accumulation of damage in concrete, and re-
sultant degradation of the physical properties of concrete
(Dubrovskii et al. 1967; Elluech et al. 1971; Hilsdorf et
al. 1978; Field et al. 2015; Le Pape et al. 2015). There-
fore, the modeling of aggregate expansion is necessary.
Damage accumulation and volume expansion of α-quartz
are well described by the nucleation and growth model
proposed by Avrami (1941) (Weber 1990; Harbsmeier
and Bolse 1998; Bolse 1999; Field et al. 2015). Our
proposed model is also based on this model. In terms of
the temperature condition, the result of a distinguished
experiment reported by Bykov et al. (1981) is shown in
Fig. 3. Based on this figure, the nucleation and growth
model, which is shown in the following equation, is used
to evaluate the temperature condition and expansion
strain εn(n).
⎛ ⎛ −n ⎞
d
ε n , quartz ( n ) = ε n , quartz ,∞ ⋅ ⎜1 − exp ⎜
⎜ ⎟
⎝ K (T ) ⎠
⎝ ⎠
where εn,quartz, ∞ : maximum expansion strain of α-quartz
(18% in case of volume change and 6% in case of linear
strain); n: neutron fluence (n/cm2) with given energy
criteria, indicating the minimum energy level for amor-
phization; K(T): temperature-dependent factor (n/cm2)
for nucleation and growth of the amorphized phase; T:
absolute temperature (K); and d: dimensionality factor,
normally ranging from 2 to 5 and is 2.38 in this study
(Field et al. 2015).
The neutron fluence at which half the volume expan-
sion occurs at the fully amorphized condition is evalu-
ated by an Arrhenius plot as shown in Fig. 4. As the
figure shows, the overall trend conforms with the Ar-
rhenius law; however, in the realistic temperature range
Fig. 3 Temperature-dependent behavior of the expansion
of α-quartz due to neutron irradiation, described by
Bykov et al. (1981).
149
for NPPs, a smaller tangent than that for the overall trend
is confirmed. Therefore, based on this, the following
equation is formulated:
⎛ − Ea , quartz ⎞
K (T ) = 3.3 ×1023 ⋅ exp ⎜ ⎟ (2)
⎝ RT ⎠
where Ea,quartz: activation energy for the nucleation and
growth of the amorphized region in α-quartz, 22500
(J/mol), and R: gas constant (J mol-1 K-1). During this
formulation, minimum energy of the neutron flux that
affects damage accumulation of α-quartz is 0.1 MeV, as
obtained from the flux information of reactors used in the
experiment by Bykov et al (1981).
Assuming simplified homogenization, aggregate ex-
pansion due to irradiation is expressed by the following
equation:
ε n , agg = ∑ ε n ,i vi (3)
i
where εn,i : expansion (or shrinkage) of mineral i caused
by neutron irradiation, and vi: volumetric ratio of mineral
i in the aggregates. Rock minerals other than α-quartz
may possibly expand by the irradiation (Ichikawa and
Kimura 2007); however, in the present study, due to the
⎞
⎟ (1) limitation of reliable data, only α-quartz is considered.
⎟
The role of the matrix around α-quartz grains is also
ignored. For example, a chlorite matrix produced in
sandstone by diagenesis shows large shrinkage under
drying and heating (Igarashi et al. 2015), but such prop-
erties of the cement mineral are not considered directly
here. In addition, stress re-distribution between the ma-
trix and many types of mineral grains due to α-quartz
grain expansion is not considered. This phenomenon
should play an important role in predicting the aggregate
expansion at the stage of point defect concentration.
These are the limitations of the present model and need to
be improved in future work for further precise prediction.
Based on aggregate expansion, concrete expansion is
Fig. 4 Arrhenius plot for α-quartz expansion due to neu-
tron irradiation as evaluated by the nucleation and growth
model.
I. Maruyama, K. Haba, O. Sato, S. Ishikawa, O. Kontani and M. Takizawa / Journal of Advanced Concrete Technology Vol. 14, 144-162, 2016
estimated for predicting the physical properties of the
concrete. As the concrete properties are dominated by
cracks due to mismatched volume change between the
aggregate and matrix (paste or mortar) and due to the
colloidal nature of hardened cement paste (Maruyama et
al. 2014b), concrete expansion strain must be a key fac-
tor that reflects the damage and cracks of concrete caused
by aggregate expansion (Dubrovskii et al. 1967; Le Pape
et al. 2015).
There exist models regarding the concrete volume
change, especially under drying, based on the informa-
tion of mixture proportion and properties and volume
change of the components. This topic has been discussed
since the early 20th century (Carlson 1938 ； Pickett
1956). Composite laws based on the continuous theory
are developed by many researchers (Hansen and Nielsen
1965; Hobbs 1974; Kishitani and Baba 1975; Hansen
1987). However, for the prediction of concrete volume
change, cracking around the aggregate as a result of
mismatched volume change between the aggregate and
matrix should be taken into account. This impact is
generally implicitly considered as a result of a lower
Young’s modulus of matrices than the measured value
under an instantaneous loading experiment, which is
attributed to creep or invisible cracking (Hansen 1987).
Recently, the impact of the interfacial transition zone
(ITZ) is also emphasized for evaluating concrete
shrinkage, because such continuous theory-based on
prediction cannot precisely explain the concrete shrink-
age affected by aggregate size (Maruyama and Sugie
2014). Therefore, in the present study, a composite law
proposed by Hobbs (1971) is applied with slight modi-
fication regarding the ITZ impact, as given below.
ε c n + 1 + ( 2mn − n − 1) VAgg ,c
= (4)
εm n + 1 + ( n − 1) VAgg ,c
n = E Agg / Em
m = ε Agg ε m
ε Agg = ε Agg , sh + ε Agg ,T + ε agg , Irr + ε ITZ
ε ITZ = d ITZ RAgg .av
where ε c: linear strain of concrete; ε m: linear strain of
mortar matrix; n: ratio of Young’s moduli of aggregate
(EAgg) to mortar (Em); m: ratio of linear strains of aggre-
gate (εAgg) to mortar (εm); ε ITZ: representative strain of the
ITZ effect compensating the damage around the aggre-
gates caused by drying (e.g., if the average radius of a
coarse aggregate (RAgg,av) is 10 mm and the ITZ apparent
thickness (dITZ) is 5 μm, then ε ITZ is calculated to be 500
μm; εAgg,sh : drying shrinkage strain of the aggregates;
εA gg,T: thermal strain of the aggregates; and VAgg,c: volu-
metric ratio of the coarse aggregates in concrete.
For the Young’s modulus of mortar, 33% of the
Young’s modulus of mortar (Hansen 1987) under in-
150
Table 2 Operation summary of JPDR-I and JPDR-II (Idei
et al. 1990).
JPDR-I JPDR-II
Total operation time 14096 hour 3014 hour
Reactor thermal power 18501 MWD 2951 MWD
output
Total power generation 12236 hour 1992 hour
time
Total power generation 1.22 × 108 kWh 1.47 × 104 kWh
Equivalent rated power 1.13 EFPY 0.09 EFPY
operation time
stantaneous loading is applied in order to consider
cracking in the mortar and long-term compressive creep
due to aggregate expansion.
In the present study, the effect of fine aggregate ex-
pansion in mortar is also considered in the same way as
for coarse aggregate and mortar expansion.
The expansion strain is a key parameter for deter-
mining the deterioration of physical properties of con-
crete, and the compressive strength is a representative
property for concrete structures. The relationship be-
tween the ratio of compressive strength after irradiation
(Fc) to that of the reference specimen (Fco) and the linear
expansion strain of the specimen due to irradiation is
obtained as shown in Fig. 5. As shown in Fig. 5, Fc/Fco
decreases with the expansion of the specimen. The ex-
perimental results by Dubrovskii et al. (1967), Du-
brovskii et al. (1970), Pedersen (1971), and Elleuch et al.
(1972) showed relatively the same trend, while the ex-
perimental results by Dubrovskii et al. (1967) showed an
upper trend compared with those by others. To evaluate
these trends, Fc/Fco of concrete specimens affected by
alkali–silica reaction (ASR) and the linear expansion
strain of the samples are plotted as shown in Fig. 5 (Na-
(5)
(6)
(7)
(8)
Lower boundary curve
Fig. 5 Relationship between the expansion strain of con-
crete and the ratio of compressive strengths of expanded
concrete specimen and the original concrete (Fc/Fco).
Experimental results of the samples expanded due to
neutron irradiation or alkali–silica reaction are plotted.
Since the lower boundary is almost consistent in these
data, a lower boundary curve is derived from these plots.
I. Maruyama, K. Haba, O. Sato, S. Ishikawa, O. Kontani and M. Takizawa / Journal of Advanced Concrete Technology Vol. 14, 144-162, 2016 151

Table 3 Chemical composition of cement (Idei et al. 1990).

Date LOI (%) Insoluble residue (%) SiO2 (%) Al2O3 Fe2O3 CaO MgO SO3 Total
Oct 1962 0.5 0.8 21.6 5.2 3.2 65.0 1.3 1.7 99.3
Nov 1962 0.5 0.7 22.0 5.3 3.0 64.7 1.3 1.8 99.3
Dec 1962 0.7 0.6 21.8 5.3 3.1 64.6 1.5 1.9 99.5

Table 4 Physical properties of cement (Idei et al. 1990).

Setting JIS-mortar Compressive
Date Density (g/cm3) Blaine value (cm2/g) Plus 88-m sieve (%)
Initial Final strength (MPa)
Oct 1962 3.14 3270 2.0 2–35 3–33 40.1
Nov 1962 3.15 3270 2.0 2–45 3–44 40.4
Dec 1962 3.15 3170 2.1 2–39 3–37 41.0

Table 5 Mixture proportion of concrete (Idei et al. 1990).

Air-entraining and water
W/C Water (kg/m3) Cement (kg/m3) Coarse aggregate (kg/m3) Fine aggregate (kg/m3)
reducing agent (xC)
0.47 139 290 1290 653 0.3 mass%

kano et al. 1987; Swamy and Al-Asali 1988; Tomita et al.
1988; Kubo et al. 2006; Ueda et al. 2011; Itoh and Ma-
ruyama 2014). This is because the deterioration mecha-
nism due to aggregate expansion is common to those
concrete degradations (Graves et al. 2011). As a result,
the data by Dubrovskii et al. (1970), Pedersen (1971),
Elleuch et al. (1972) almost correspond to the minimum
values of ASR data and the data by Dubrovskii et al.
(1967) almost correspond to the maximum values of
ASR data. The trends are consistent. A lower boundary
curve is represented by the following equation:
Fc / Fco = 0.3 − 0.7exp ( −1.4 ⋅104 ε n,c )
Note that some differences exist in the experimental
conditions of the abovementioned previous studies. For a
specimen of young concrete at the commencement of
exposure to irradiation or ASR environment, an increase
in the strength occurs because continuous hydration
increases the compressive strength of the specimen. High
temperature (~100 ºC) also causes an increase in the
C–S–H strength and additional hydration of residual
cement minerals. These impacts yield variation in the
experimental results. Using the lower boundary curve as
the safety values, such unexplained phenomena are
evaluated.
JPDR-I attained self-sustaining nuclear chain reaction
in August 1963, and continued intermittent operation
until September 1969. It was then upgraded to JPDR-II,
which operated from May 1972 to March 1976. The
operation summary is given in Table 2 and the operation
history of the reactors is presented in Fig. 6 (Idei et al.
1990).
The concrete strength of cores extracted from a CBS
of the JPDR was evaluated by the proposed numerical
method. During the construction of the JPDR, concrete
for the CBS walls was casted from March to November
1962. The concrete was composed of ordinary Portland
(9) cement from Hitachi Cement Co. Ltd., Naka River gravel,
and sea sand taken from a place near the site. Chemical
composition and physical properties of the cement used
are shown in Tables 3 and 4. The timing of these records
does not correspond to the date of construction of the
CBS, but the variation in cement quality was small, and
hence, these data seem applicable to the cement of the
CBS. The mixture proportion of concrete for the CBS is
shown in Table 5. The strength for mixture proportion
was 35 MPa and the slump was 7.5 cm (Idei et al. 1990;

3．Lessons from JPDR

3.1 Overview of JPDR

The JPDR, the first nuclear reactor that generated elec-
trical power in Japan in 1963, was built in 1960 and the
total operation time was 957 days. Even though the op-
eration time is limited, the total dose or fluence on its
CBS walls was considered comparable to that of com-
mercial light water NPPs after 40-year operation. Be-
cause of this, the JPDR has been experimented widely
from the viewpoint of decommissioning and aging Fig. 6 History of operations and relative power output of
management of concrete structures. JPDR (Idei et al. 1990).
I. Maruyama, K. Haba, O. Sato, S. Ishikawa, O. Kontani and M. Takizawa / Journal of Advanced Concrete Technology Vol. 14, 144-162, 2016 152

Kakizaki et al. 1996).

The planar and elevational sections of the target CBS
are shown in Fig. 7 (Kakizaki et al. 1996). As can be seen
from the figure, the thickness of CBS is 1500–3000 mm.
The internal surface of the member was covered by a
steel liner plate of thickness 13 mm, and 1-inch-diameter
(25.4 mm) cooling pipes were placed at 76 mm and 300
mm from the internal surface and at every 5°–11° in the
radial direction. The main rebars, which were deformed
bars of 29-mm diameter, were located horizontally and
vertically at every 150 mm. Shear reinforcements of a
25-mm-diameter deformed bar were located or-
thogonally to the main rebars at every 300 mm vertically
and 450 mm horizontally.
The outer side of the CBS was coated by resin; the
detail of this material was, however, not recorded.
The radiation condition has been reported by Idei et al.
(1990). Regarding the neutron flux, the calculated result
by ANISN is shown in Fig. 8. The calculation conditions
have been validated by the 152Eu activity that was ex-
perimentally obtained at the decommissioning process
(Kakizaki et al. 1996). The typical calculation results of
neutron and gamma-ray spectra at the full-power reactor Fig. 7 Elevation and planar section of the concrete bio-
condition at the surface of the CBS are shown in Fig. 9. logical shielding wall. The gray areas indicate the place
The reactor containment vessel was ventilated with of core sampling whose data are used in this study. (Ka-
atmospheric air at 2 × 105 m3/h. Therefore, the outer side kizaki et al. 1996).
of the CBS had almost the same temperature as the
temperature of the room where humans work. The inside
of the CBS was ventilated with atmospheric air at 9.1 ×
103 m3/h and a speed of 1.5–2.1 m/s. The temperature of
atmosphere on the inner surface of the CBS was reported
to be approximately 46 °C and the temperature around
the cooling pipe was 25 °C (Idei et al. 1990).
The core sample specimens were taken from the CBS.
The locations were determined so that depth, fluence,
and temperature conditions of the samples can be evalu-
ated. The core drilling location is indicated in Fig. 7. The
diameter of the specimen was 80 mm and a wet drilling
method was applied. The compressive strength test
specimen had a height h = 2d, where d is diameter of the
specimen. Specimens were formed by a diamond saw
and were sulphur-capped at both ends for the loading. Fig. 8 Neutron flux distribution in the radial direction from
The test method followed JIS A 1108. the core center of the JPDR.
The experiments were carried out between 1986 and
1988.

3.2 Conditions for simulation

In the present study, DEVICE was used for the evalua-
tion of concrete property changes in the CBS of the JPDR.
From the previous section, the details of the calculation
input were determined as follows. The averaged values
of Tables 3 and 4 were considered for the calculation of
the properties of the cement. Based on the loss on igni-
tion (LOI) data, the cement was considered pure Portland
cement. The mineral composition of cement was as-
sumed by using the Bogue equation (Bogue 1929). The
mixture proportion of concrete was used as it is. The
average density of river gravel is around 2.66 g/cm3 in Fig. 9 Neutron and gamma-ray spectra at the internal
surface of the target CBS.
I. Maruyama, K. Haba, O. Sato, S. Ishikawa, O. Kontani and M. Takizawa / Journal of Advanced Concrete Technology Vol. 14, 144-162, 2016
Japan (Iwata 1981) and using this value, the air content in
concrete was calculated as 5.5%, which is relatively large
for the air content required for resistivity to freezing and
thawing reaction. Therefore, the air content and density
of river gravel and sea sand were assumed to be 4.5% and
2.61 g/cm3, respectively.
No information regarding the quartz content in ag-
gregates is available in literature. However, the estimated
density is similar to that of α-quartz (2.65 g/cm3), and to
take into account the risk of neutron irradiation in the
calculation as much as possible, the aggregates were
considered as composed of α-quartz only. Therefore, the
shrinkage of aggregates is zero and thermal expansion
coefficient is 10.0 × 10-6 /K (Igarashi et al. 2015).
For the construction process, we assumed 1 hour for
transport from a ready-mixed concrete plant to the CBS
location. The mold was assumed to be made of steel and
the heat transfer coefficient to be 14 W/m2/K. The inside
steel plate liner was considered to be used for the mold.
Therefore, the inner part of the CBS was not exposed to
the dry atmospheric air. Environmental conditions were
293 K and 60% RH, which are the average values of the
Japanese climate. The timing of demolding was not re-
corded. Based on the history of concrete casting, the age
of demolding of concrete was assumed to be 28 days
after casting. Then, immediately before the operation and
11 months after concrete casting, the resin coating was
finished at the outer side of the CBS. The resin coating
was assumed to have 1/100 of the permeability of con-
crete.
A 1500-mm-thick member was used for this calcula-
tion because this was the minimum thickness of the CBS.
During the operation, the temperature of the air on the
surface of the steel plate liner was assumed to be 43 °C,
and the concrete temperature around the pipe was as-
sumed to be 25 °C. This is because the data of tempera-
ture of cooling water and ventilating air are not available
and existing experimental data and former calculations
are reliable. In the off-operation state, environmental
condition of both sides of the CBS was assumed to be
298 K and 50% RH.
Regarding the radiation condition, neutron and
gamma-ray flux was assumed to have a linear relation-
ship with the output power, as shown in Fig. 6. In the
present calculation, the impact of the water cooling pipe
has been studied beforehand, because water content and
iron within the cooling pipe may affect the shielding and
resultant secondary gamma rays and heat production.
Our prior calculation, which is shown in Fig. 10, exhibits
little impact on secondary gamma rays. Therefore, the
calculation mesh did not consider these water cooling
pipes.
As for meshing for the calculation, a 7.5 mm discrete
mesh was applied.
3.3 Results and discussion
Concrete in the CBS of the JPDR was evaluated up to a
concrete age of 25 years. The results of typical tem-
153
Table 6 Mineral composition of cement for calculation.
Alite Aluminate Ferrite Gypsum
Belite (%)
(%) phase (%) phase (%) (%)
53.9 23.1 9.0 9.6 4.4
perature, water content per cement paste in concrete,
degree of hydration of cement, and compressive strength
distribution in CBS at arbitrary concrete ages are shown
in Fig. 11. Due to its size, the heat of hydration accu-
mulated up to ~60 °C. In case of some aggregates, con-
crete strength development was stagnated or deteriorated
because of increase in number of large pores by forma-
tion of densified outer C-S-H due to elevated temperature
(Sugiyama and Masuda 1999; Gallucci et al. 2013) and
the difference between the volume changes in the ag-
gregate and matrix and resultant damage accumulation
around aggregates (Son and Hosoda 2010; Maruyama et
al. 2014b). However, in this calculation, silicate content
was high and the thermal expansion of the aggregate was
10 × 10-6 /K. Therefore, the damage does not have a large
impact on this calculation and the strength distribution is
similar to that of the hydration degree. After 1 year, the
degree of hydration was about 0.87 and the distribution
of hydration was almost homogenous except for the outer
surface where the cement hydration was stagnant owing
to water loss after demolding and resultant moisture
evaopration to the surrounding air.
After about 16 years, operation of the JPDR was
stopped. During the operation period of 6.3 years, the
maximum temperature was observed. In this case, the
maximum temperature was recorded at a 400 mm depth
from the inner surface. This temperature was attained
mainly by gamma heating accompanied by water cooling
from the embedded pipes. Therefore, we can estimate
that the water cooling pipe largely impacts the reduction
of temperature in the CBS caused by gamma heating. In
general, water movement is derived by the temperature
gradient in the thick concrete wall (England and Ross
1972), but the CBS of the JPDR does not have a large
temperature difference between its inner and outer sur-
face parts, because the water cooling pipe was effective.
Resultantly, although a small variation in the water con-
Fig. 10 Impact of water cooling pipes in concrete on the
flux distribution in the CBS.
I. Maruyama, K. Haba, O. Sato, S. Ishikawa, O. Kontani and M. Takizawa / Journal of Advanced Concrete Technology Vol. 14, 144-162, 2016
tent was observed around the cooling pipe, it is generally
almost constant. This is the main reason that the concrete
strength is almost homogenous in the CBS. In addition,
the fast neutron fluence does not exceed 1.0 × 1018 n/cm2
after the operation. Within this value, the α-quartz is
almost inert and we cannot see any impact on the con-
crete strength. The experimental results of the concrete
strength in the BSW were compared with the numerical
results as shown in Fig. 12. The comparison shows that
the calculation results are comparable to the experimen-
tal data. From the experimental data and supporting
numerical simulation, the following points can be drawn,
which can contribute to the understanding of aging
management.
1) A fast neutron fluence of less than 1.0 × 1018 n/cm2 at
room temperature (<40 °C) does not significantly
impact the concrete strength.
2) The control of temperature at the inner part of the CBS
has an impact to homogenize the concrete strength in
the CBS. This tendency is favorable to evaluate the
representative concrete strength in thick concrete walls.
3) The sealing conditions at the inner surface by a steel
Fig. 11 Calculation results of the distribution and devel-
opment of temperature, fast neutron fluence, water con-
tent, degree of hydration of cement, and compressive
strength of concrete in the BSW of the JPDR.
154
liner plate as well as at the outer surface by resin
coating also have a homogenizing effect on the con-
crete strength of the CBS. In the present calculation,
the depth affected by drying is limited to 100 mm from
the surface.
4. Calculation of an anonymous PWR
reactor
In the present study, a two-loop pressurized water reactor
(PWR) is studied for the discussion of the integrity
evaluation of a CBS. This is because a two-loop PWR
imparts a high risk of neutron fluence on the CBS. The
target CBS is assumed to be 2.0 m thick and the concrete
component is assumed to be the same as that of the JPDR.
The radiation boundary condition is calculated from
Fukuya et al. (2002), and the results of neutron and
gamma-ray flux at the inner surface of the CBS are
shown in Fig. 13. With regards to the boundary condition
of moisture transfer at the inner and outer surfaces, the
conditions of sealed from and open to the ventilated air
are assumed. For temperature conditions at operation,
64 °C and 53 °C are considered at the inner surface,
based on the code regulation and our personal interview
for the two-loop PWR condition, respectively, while
Fig. 12 Comparison of predicted concrete strength after
25 years (line) and experimental data (circles).
Fig. 13 Neutron and gamma-ray flux at the surface of the
CBS.
I. Maruyama, K. Haba, O. Sato, S. Ishikawa, O. Kontani and M. Takizawa / Journal of Advanced Concrete Technology Vol. 14, 144-162, 2016 155

Table 7 Calculation parameters and conditions.

Temperature condition Moisture condition
Operation Inspection Operation Inspection
Notation
(274 days per year) (91 days per year) (274 days per year) (91 days per year)
Inner surface Outer surface Inner surface Outer surface Inner surface Outer surface Inner surface Outer surface
Sealed by Sealed by Sealed by Sealed by
337s 64 °C 35 °C 20 °C 20 °C
Liner *1 resin *2 Liner *1 resin *2
Concrete Concrete Concrete Concrete
337o 64 °C 35 °C 20 °C 20 °C
surface *3 surface *3 surface *3 surface *3
Sealed by Sealed by Sealed by Sealed by
326s 53 °C *4 35 °C 20 °C 20 °C
Liner *1 resin *2 Liner *1 resin *2
*1 The moisture movement at the liner is assumed to be zero.
*2 The value is the same as that of JPDR calculation.
*3 Ambient air is assumed to be ventilated air whose condition is 60% RH at 20 °C.
*4 This value is based on our personal interview with the power company.

35 °C is assumed for the outer surface. The temperatures
at both sides of the CBS during the periodical inspection,
corresponding to the absence of heat from the reactor and
continuous 91 days (25% of a year) allocated for this
period, is 20 °C. The conditions and calculation pa-
rameters are summarized in Table 7. In the construction
process, demolding is after 28 days of casting, and 1 year
and 3 months is assumed as the construction period. The
construction period is relatively short compared with the
real construction process. Moreover, after the construc-
tion process, the first operation was started.
The calculation results of 337s (Notation and pa-
rameters are shown in Table 7) in which the CBS was
sealed on both surfaces and irradiated from the inner
surface are summarized in Fig. 14. The fast neutron
fluence at the surface position was about 6.2 × 1019 n/cm2
after 60 years of operation (Fig. 14 (b).) During the op-
eration, temperature at the surface of the concrete was
about 64 °C, but due to neutron scattering and
gamma-ray heating, maximum temperature was ob-
served at 40 mm from the inner surface. In addition, a
negative gradient of temperature was confirmed from the
position of peak temperature to the outer surface of the
CBS (2000 mm from the inner surface) (Fig. 14 (d)).
Due to this temperature gradient, the resulting accumu-
lation of vapor pressure and a little, but not negligible,
water transport at the outer surface of the CBS, water
slightly moved from the inner part to the outer part (Fig.

Fig. 14 Calculation results for 338s until 60-year operation. (a) Compressive strength, (b) fast neutron (>0.1 MeV) fluence,
(c) evaporable water per cement paste, and (d) temperature.
I. Maruyama, K. Haba, O. Sato, S. Ishikawa, O. Kontani and M. Takizawa / Journal of Advanced Concrete Technology Vol. 14, 144-162, 2016
14 (c)). In particular, from 0 to about 500 mm from the
inner surface, the water was slightly reduced. Although
the evaporable water tended to move to the outer surface,
a sufficient amount of water remained in the inner part of
the CBS and cement hydration gradually continued (Fig.
14 (a)).
Theoretically, water loss should relate to the reduction
of both heat capacity and shielding performance; how-
ever, in this case, significant temperature difference
could be found. The temperature difference was within
0.02 °C during the operation period.
After 20 years, a substantial drop in the concrete
strength was observed at the inner surface. The fast neu-
tron fluence was 1.61 × 1019 n/cm2 at the surface, and the
concrete strength was deteriorated by neutron irradiation
at the region from 0 mm to 40 mm from the inner surface.
At this point, the concrete expansion in a hypothetical
free restraint condition at the surface was 0.32% and at
the region 40 mm from the surface was 0.066%.
After 60 years operation, the depth at which concrete
strength was deteriorated by neutron irradiation was
about 100 mm. The expansion strain of concrete at the
surface was 3.2% and at 100 mm from the inner surface,
where the main rebars are generally embedded, was
0.12%. Therefore, in this case, the structural integrity is
questionable, because we do not have enough knowledge
of the bond behavior with such expanded concrete con-
ditions.
Next, the case of the CBS with open surfaces is dis-
cussed. The comparison of calculation results of 337s
and 337o is shown in Fig. 15. A large difference was
confirmed in the evaporable water content, because wa-
ter was evaporated from the surfaces of the CBS. Espe-
cially, near the inner surface, which was heated by irra-
diation and attained about 64 °C, a large amount of water
was evaporated from the surface after operation began.
The depth of the dried zone was approximately 500 mm.
This dried zone reflected the water vapor desorption
isotherm model considered in the CCBM. In the model,
water vapor BET surface area is a function of tempera-
ture and humidity, and water amount at 98% RH is de-
termined by the temperature. This model was based on
our experimental results of the colloidal behavior of
calcium silicate hydrates (Maruyama et al. 2014a). Un-
der the heating and drying conditions, the concrete sorp-
tion amount at the inner surface decreased with increas-
ing temperature, and water vapor was accumulated in the
pore space, which then tended to attain the equilibrium
condition. This process is strongly affected by the ele-
vated temperature rather than the relative humidity. After
the drying process, concrete strength development was
stagnant immediately after the operation began. In addi-
tion to this stagnation of compressive strength devel-
opment, deterioration due to neutron irradiation was
confirmed after several years of operation.
The sealing condition at the inner surface of the CBS
will contribute to attain the strength margin because of
continuous cement hydration through conservation of
156
evaporable water at the inner surface.
Furthermore, the temperature impact on strength
change at the inner surface of the CBS of 337s and 326s
is discussed. Calculation results are shown in Fig. 16.
The distribution of evaporable water content is not dif-
ferent from the others. The largest impact was deteriora-
tion of concrete at the inner part. In the case of 337s, the
strength of concrete at the inner surface was 16 MPa,
while it was 14 MPa in the case of 326s. The depth of
concrete affected by irradiation was 12 cm in the case of
326s, while it was 10 cm in the case of 337s. This dif-
ference was caused by the temperature difference of
11 °C at the inner surface, which will result in the dif
ference in neutron-induced expansion of α-quartz (see
Fig. 3). The changes in compressive strength at the inner
surfaces for 337s and 326s are almost the same. However,
the difference of depth of concrete strength deterioration
might considerably affect the structural performance of
reinforced concrete members.
Water, which is the main source of hydrogen atom in
concrete, is considered to have a large impact on
shielding performance of concrete. Therefore, the evap-
orable water content distribution and the ratio of fast
neutron fluence after 60 years of operation are shown in
Fig. 17. In this figure, the ratios of fast neutron fluence
for 326s and 337o to that of 337s as well as the evapor-
able water content distributions for 337s, 337o, and 326s
are plotted. The figure shows that increase (or preserv-
ing) of water content in the CBS reduces the neutron
fluence in the CBS, while decrease of water content at
the inner surface increases the neutron fluence in the
CBS. However, concrete deterioration caused by irra-
diation-induced aggregate expansion was confirmed
within 100 mm from the inner surface. Considering the
range up to which neutrons have the impact, the sealing
condition at the inner surface has a large effect. At a
depth of 100 mm, the fluence varied about 10%. This
value is linearly related to the time span of concrete
deterioration; therefore, the steel liner is very effective
for neutron-induced concrete deterioration in the CBS.
5. Discussion
5.1 Limits of DEVICE
While the DEVICE takes into account many aspects of
concrete phenomena, there are limitations. Moisture
transport in DEVICE is validated by the experiments
(Maruyama and Igarashi 2015), but long-term drying
process in massive concrete under the relatively high
temperature condition needs more validation. Because
data are very scarce, we need more scientific data of the
moisture transport in massive concrete under the high
temperature conditions as well as the temperature gra-
dient conditions.
Reduction of concrete due to neutron irradiation is
modeled by a statistical way, but the background data
have a large variation in the relationship between ex-
pansion strain of concrete and Fc/Fco. This variation is
I. Maruyama, K. Haba, O. Sato, S. Ishikawa, O. Kontani and M. Takizawa / Journal of Advanced Concrete Technology Vol. 14, 144-162, 2016 157

Fig. 15 Comparison of calculation results for 337s and 337o for 60-year operation.

Fig. 16 Comparison of calculation results for 337s and 326s. (a) (top) strength distribution and (bottom) temperature dis-
tribution of 337s. (b) (top) strength distribution and (bottom) temperature distribution of 326s. The elevated temperature at
the inner surface strongly affects the distribution of evaporable water and resultant strength development.
I. Maruyama, K. Haba, O. Sato, S. Ishikawa, O. Kontani and M. Takizawa / Journal of Advanced Concrete Technology Vol. 14, 144-162, 2016
larger than that of loading experiment for compressive
strength, therefore, there should be a mechanism to ex-
plain this variation. Concrete’s physical property changes
affected by aggregate expansion depend on the crack
propagation in it. This implies that the deterioration
might be anisotropic due to concrete stress caused by
dead load. This is not considered in DEVICE. In addition,
a combination of dead load and aggregate expansion may
cause compaction of mortar which might increase the
strength by closing the crack width and altering C–S–H
by the compaction effect (Sereda et al. 1966). The dif-
ference in strength from the lower boundary curve shown
in Fig. 5 can possibly be explained via compaction of
mortar or cement paste in a capsule that is used in irra-
diation experiments along with the irradiation-induced
aggregate expansion.
The temperature and duration of irradiation highly
affect α-quartz expansion, as shown in Fig. 3. A ther-
modynamic healing impact of silicate crystal structures
exists, which is a function of temperature and time. If the
temperature is elevated, the healing process is enhanced,
and consequently, the expansion rate under irradiation
conditions is low. This implies that the accelerated ex-
periment result might underestimate the healing process
of silicate structure, because the realistic time might be
more than 10 times the experimental time span. There-
fore, defect accumulation and consequent macroscopic
volume change must be scientifically studied and the
healing process must be independently understood for
achieving a comprehensive understanding of irradia-
tion-induced expansion of α-quartz. In addition, a more
accurate estimation may be obtained by considering this
healing impact. In addition, impacts of other aggregate
minerals also should be considered, but the current code
can not take into account due to lack of data.
5.2 Integrity evaluation
The CBS that is exposed to radiation develops a risk of
irradiation-induced concrete deterioration. The expected
degradation can be categorized into two phenomena:
large expansion (Le Pape 2015) and mechanical property
deterioration. Undoubtedly, these two phenomena are
strongly interconnected through stress and resultant
cracks in the concrete of the CBS; however, to the best of
author’s knowledge, there is no experimental report
about reinforced concrete member affected by irradiation
induced concrete deterioration. As shown in the intro-
duction, the integrity of a CBS is normally evaluated by
determining the limit of fast neutron fluence and/or
gamma-ray dose, which depend on the authorities or
countries. Previous studies have suggested that the con-
crete strength deteriorates above a fast neutron (>0.1
MeV) fluence of 1019 n/cm2 under the temperature con-
dition of light water reactors (Hilsdorf et al. 1978; Field
et al. 2015), while the minimum neutron energy that
affects α-quartz is 0.01 MeV (Primak 1958; Weissmann
and Nakajima 1963; Bykov et al. 1981). Further scien-
tific research is needed to confirm this criterion.
158
Based on previous accelerated irradiation experiments
of concrete and mortar, the results were numerically
extrapolated to the problem in a CBS by using DEVICE.
The calculation results indicate that the depth of concrete
deterioration is approximately 100 mm from the inner
surface of the CBS after 60 years of operation of the
reactor. We discuss how to evaluate this result consider-
ing the limitations of our knowledge.
An argument about the aggregate type or the mineral
composition in the aggregate is addressed here. In the
present calculation, the aggregate is considered com-
posed of 100% α-quartz. If this amount is half, the
volume of the elements that can expand in the concrete
clearly becomes half and thus the damage in concrete and
the degree of resultant deterioration should be decreased.
The ionic minerals, such as calcite, probably have high
resistivity to irradiation, and concrete with pure lime-
stone may show favorable performance under irradiated
conditions. Even though we do not have complete data-
base for the volume stability of aggregate minerals under
irradiated conditions, we can conclude that the aggregate
type and its mineral composition play an important role
in irradiation-induced concrete degradation. Therefore, it
is possible to rebuttal with a viewpoint of mineral com-
position in the used aggregate.
Secondary integrity evaluation is possible after the
fluence limit is evaluated. First, the aggregate type and
the mineral composition must be specified, and then, the
aggregate expansion strain needs to be evaluated by
considering the temperature at irradiation condition,
which requires data on the resistivity of rock minerals.
The allowable fast neutron fluence can be then proposed
according to the mineral composition of the aggregate.
Alternatively, the strength can be evaluated, and an al-
lowed reduction of concrete strength can be determined
such that it satisfies the required design strength. The
lower boundary curve shown in Fig. 5 and Eq (9) give
Fig. 17 Impact of water content on the fast neutron flu-
ence after 60-year operation. Water content distribution
and the ratio of fast neutron fluence after 60-year opera-
tion are considered. Water content distribution changes
during operation and the neutron fluence are affected by
this water content distribution. The data are accumulated
results from the onset of operation to the final state of the
calculation.
I. Maruyama, K. Haba, O. Sato, S. Ishikawa, O. Kontani and M. Takizawa / Journal of Advanced Concrete Technology Vol. 14, 144-162, 2016
the worst-case value of Fc/Fco. Therefore, based on the
DEVICE calculation with a calibration of the obtained
data, we can evaluate the strength at the inner surface as a
safe-side value, which can be compared with the required
compressive strength, and thus we can judge the integrity
of a CBS through material-basis evaluation.
The last option is a more comprehensive approach.
The integrity of a CBS with regards to the seismic per-
formance can be evaluated by omitting a part of a section
in the CBS by estimating it to be deteriorated by the
gamma-ray dose. In Japan, although a gamma-ray limit
of 2.0 × 105 kGy exists, some PWRs have experienced
more than this limit. An electrical power company has
supported this condition using a calculation that consid-
ered the deteriorated part (The Nuclear Regulation Au-
thority, Japan 2015). This type of approach can be used
for evaluating the integrity of a CBS. The impact of
concrete expansion, which may cause spalling on the
surface and deterioration of bond in the rebars, should be
considered for seismic performance evaluation.
The validation relating to important phenomena such
as healing process of α-quartz, stress-related physical
property change, bond performance affected by volume
expansion of concrete, and irradiation-induced deterio-
ration of concrete can be carried out only on decommis-
sioned plants. Therefore, such research projects using a
decommissioned plant are necessary for long-term op-
eration of aging plants.
6. Conclusion
A numerical code called “DEVICE” (Damage EValua-
tion for Irradiated ConcretE), which takes into account
heat, moisture, and radiation transport coupled with
cement hydration, is proposed. This code is based on the
computational cement-based material model (CCBM)
and the one-dimensional deterministic transport Sn code
ANISN. First, DEVICE was used for evaluating the
decommissioned plant JPDR. The compressive strength
distribution in a biological shielding wall of the JPDR is
obtained by core sampling. The compressive loading test
results were compared with the calculation results, which
presented the potential of DEVICE to predict concrete
strength distribution in a concrete biological shielding
(CBS) wall. The fast neutron fluence level of the CBS
under which significant impact on concrete strength is
not observed is 1.0 × 1018 n/cm2 at room temperature
(<40 °C). The comparison results also showed that the
sealing conditions at the inner and outer surfaces of the
CBS by a steel liner plate and resin coating, respectively,
have a homogenizing effect on the concrete strength of
the CBS.
In addition, simulation results for the concrete strength
change and its distribution in a biological shielding wall of
an anonymous two-loop pressurized water reactor up to 60
years’ operation indicated the following conclusions:
1) A steel liner plate and resin coating, which restrain
water vapor evaporation from the surface of a CBS,
159
can produce the strength margin owing to continuous
cement hydration.
2) The temperature difference at the inner surface of the
CBS has large impact on irradiation-induced concrete
degradation through temperature dependence of irra-
diation-induced expansion of α-quartz.
3) However, the difference in compressive strength due
to temperature-dependent irradiation-induced con-
crete degradation is almost the same order as that by
the strength margin built by the sealing condition.
4) While the volume stability of aggregate minerals,
minimum neutron flux energy for the damage in
α-quartz, healing effect of the silicate structure as a
function of temperature, and degradation process un-
der stress in concrete are limitations of our current
evaluation method, and they need more scientific
background.
In addition, possible integrity evaluation methods
other than allowable fast neutron fluence evaluation for a
biological shielding wall are proposed as follows.
1) Mineral composition-based allowable fast neutron
fluence.
2) Strength prediction at the inner surface on the basis of
mineral composition of aggregate and the lower limit
curve of Fc/Fco as a function of concrete expansion.
3) Direct numerical calculations for the seismic per-
formance by considering irradiation-induced volume
expansion and concrete degradation.
Acknowledgement
This research was supported by the Nuclear Regula-
tion Authority (NRA, Japan), and we thank the NRA for
their sponsorship. The information presented in this
paper is the sole opinion of the authors and does not
necessarily reflect the views of the sponsoring agencies.
References
American Concrete Institute, (2012). “Code
requirements for nuclear safety related concrete
structures.” ACI 349 - 12.
Avrami, M., (1941). “Granulation, phase change, and
microstructure kinetics of phase change. III.” The
Journal of Chemical Physics, 92, 177-184.
Bogue, R. H., (1929). “Calculation of the compounds in
Portland Cement.” Industrial & Engineering Chemis-
try Analytical Edition, 14, 192-197.
Bolse, W., (1999). “Amorphization and recrystallization
of covalent tetrahedral networks.” Nuclear Instru-
ments and Methods in Physics Research, Section B:
Beam Interactions with Materials and Atoms 148,
83-92.
Bonnet, J. P., Boissier, M. and Ait Gherbi, A., (1994).
“The amorphization process of neutron-irradiated
crystalline quartz studied by Brillouin scattering.”
Journal of Non-Crystalline Solids, 167, 199-204.
Bouniol, P. and Aspart, A., (1998). “Disappearance of
oxygen in concrete under irradiation: the role of
peroxides in radiolysis.” Cement and Concrete Re-
I. Maruyama, K. Haba, O. Sato, S. Ishikawa, O. Kontani and M. Takizawa / Journal of Advanced Concrete Technology Vol. 14, 144-162, 2016
search, 28, 1669-1681.
British Standards Institute, (1990). “Specification for
prestressed concrete pressure vessels for nuclear
reactors.” Technical Report BS 4975:1990.
Bykov, V. N., Denisov, A. V. Dubrovskii, V. B.,
Korenevskii, V. V., Krivokoneva, G. K. and
Muzalevskii, L. P., (1981). “Effect of irradiation
temperature on the radiation expansion of quartz.”
Soviet Atomic Energy, 51, 593-595.
Carlson, R. W. (1938). “Drying shrinkage of concrete as
affected by many factors.” American Society Testing &
Materials Proceedings, 38, 419-437.
Cong, X. and Kirkpatrick, R. J., (1995). “Effects of the
temperature and relative humidity on the structure of
C-S-H gel.” Cement and Concrete Research, 25,
1237-1245.
Douillard, L. and Duraud, J. P., (1996). “Amorphization
of α-quartz under irradiation.” Journal de Physique III
6, 1677-1687.
Dubrovskii, V. B., Ibragimo, S., Korenevs, V., Ladygin,
A. Y., Pergamen, V. and Perevalo, V., (1970).
“Hematite concrete for shielding against high neutron
fluxes.” Soviet Atomic Energy-Ussr 28(3), 336-338.
Dubrovskii, V. B., Ibragimo, S., Ladygin, A. Y.,
Kulakovs.M. Y. and Pergamen. B, k., (1968).
“Radiation stability of serpentine concrete.” Soviet
Atomic Energy-Ussr 25(6), 1345-1346.
Dubrovskii, V. B., Ibragimov, A., Ladygin S. and
Pergamenshckik, B. K., (1966). “The effect of neutron
irradiation on certain properties of refractory con-
cretes.” Atomnaya Energiya 21, 108-112.
Dubrovskii, V. B., Ibragimov, S. S., Kulakovskii, M. Y.,
Ladygin, A. Y. and Pergamenshchik, B. K., (1967).
“Radiation damage in ordinary concrete.” Soviet
Atomic Energy 23(4), 1053-1058.
Eby, R. K., Ewing, R. C. and Birtcher, R. C., (1992).
“Amorphization of complex silicates by ion-beam
irradiation.” Journal of Materials Research 7(11),
3080-3102.
Elleuch, L. F., Dubois F. and Rappeneau, J., (1972).
“Effects of Neutron Radiation on Special Concretes
and Their Components.” ACI SP-34, 1071-1108.
Elluech, M. R., Dubois, F. and Rappenau, J., (1971).
“Behavior of special shielding concretes and of their
constituents under neutron irradiation.” Fourth United
Nations International Conference on the Peaceful
Uses of Atomic Energy 7.
England, G. L. and Ross, A. D., (1972). “Shrinkage,
moisture, and pore pressures in heated concrete.” ACI
Special Publication, 34, 883-908.
Engle Jr., E. A., (1967). “A user's manual for ANISN, A
one-dimensional discrete ordinates transport code with
anisotropic scattering.” USAEC Report K-1963.
Ewing, R. C., Meldrum, A., Wang, L. and Wang S.,
(2000). “Radiation-induced amorphization.” Reviews
in Mineralogy and Geochemistry. 39, 319-361.
Field, K. G., Remec I. and Le Pape, Y., (2015).
“Radiation effects in concrete for nuclear power plants
160
– Part I: Quantification of radiation exposure and
radiation effects.” Nuclear Engineering and Design,
282, 126-143.
Fujiwara, K., Ito, M., Sasanuma, M., Tanaka, H.,
Hirotani, K., Onizawa, K., Suzuki, M. and Amezawaf,
H., (2009). “Experimental study of the effect of
radiation exposure to concrete.” In: Proceedings of the
20th International Conference on Structural Mechan-
ics in Reactor Technology.
Fukuya, K., Nakata, H., Fujii, K., kimura, I., Ohmura, M.,
Kitagawa, H., Itoh T. and Shin, K., (2002). “Radiation
field analyses in reactor vessels of PWRs.” INSS
Journal, 9, 153-161.
Gallucci, E., Zhang, X. and Scrivener, K. L., (2013).
“Effect of temperature on the microstructure of
calcium silicate hydrate (C-S-H).” Cement and Con-
crete Research, 53, 185-195.
Gartner, E. M., (1997). “A proposed mechanism for the
growth of C-S-H during the hydration of tricalcium
silicate.” Cement and Concrete Research, 27(5),
665-672.
Gartner, E. M., Kurtis, K. E. and Monteiro, P. J. M.,
(2000). “Proposed mechanism of C-S-H growth tested
by soft X-ray microscopy.” Cement and Concrete
Research 30(5), 817-822.
Giorla, A., Vaitová, M., Le Pape, Y. and Štemberk, P.,
(2015). “Meso-scale modeling of irradiated concrete
in test reactor.” Nuclear Engineering and Design, 295,
59-73.
Graves, H., Le Pape, Y., Naus, D., Rashid, J., Saouma, V.,
Sheikh A. and Wall, J., (2011). “Expanded Material
Degradation Assessment (EMDA).” Volume 4: Aging
of Concrete. Technical Report NUREG/CR-7153,
ORNL/TM-2011/545.
Gray, B. S., (1971). “The effects of reactor radiation on
cement and concrete.” In: Proceedings of an Infor-
mation Exchange Meeting on 'Results of Concrete
Irradiation Programmes'.
Hansen, T. C. and Nielsen, K. E. C., (1965). “Influence of
aggregate properties on concrete shrinkage.” ACI
Journal, 62(7), 789-794.
Hansen, W.. (1987). “Constitutive model for predicting
ultimate drying shrinkage of concrete.” Journal of the
American Ceramic Society, 70(5), 329-332.
Harbsmeier, F. and Bolse W., (1998). “Ion beam induced
amorphization in α quartz.” Journal of Applied
Physics, 83(8), 4049-4054.
Hilsdorf, H., Kropp J. and Koch, H., (1978). “The effects
of nuclear radiation on the mechanical properties of
concrete.” ACI SP-55, 223-254.
Hobbs, D. W., (1971). “The dependence of the bulk
modulus, Young's modulus, creep, shrinkage and
thermal expansion of concrete upon aggregate volume
concentration.” Matériaux et Construction, 4(2),
107-114.
Hobbs, D. W., (1974). “Influence of aggregate restraint
on the shrinkage of concrete.” ACI Journal, 71(9),
445-450.
I. Maruyama, K. Haba, O. Sato, S. Ishikawa, O. Kontani and M. Takizawa / Journal of Advanced Concrete Technology Vol. 14, 144-162, 2016
Ichikawa, T. and Kimura T., (2007). “Effect of nuclear
radiation on alkali-silica reaction of concrete.” Journal
of Nuclear Science and Technology, 44(10),
1281-1284.
Idei, Y., Kamata, H., Akutsu, Y., Onizawa, M., Nakajima,
N., Sukegawa T. and Kakizaki, M., (1990). “Strength
of Biological Shield Concrete of JPDR.” Technical
Report JAERI-M 90-205.
Igarashi, G., Maruyama, I., Nishioka Y. and Yoshida, H.,
(2015). “Influence of mineral composition of siliceous
rock on its volume change.” Construction and
Building Materials, 94, 701-709.
Inui, H., Mori, H., Sakata T. and Fujita, H., (1990).
“Electron irradiation induced crystalline-to-amor-
phous transition in quartz single crystals.” Journal of
Non-Crystalline Solids, 116(1), 1-15.
Itoh, M. and Maruyama, I., (2014). “Evaluation of
compressive strength of concrete expanded by
alkali-silica reaction.” Summaries of the annual CAJ
meeting of cement and concrete engineering, 68,
250-251. (in Japanese)
Iwata, S., (1981). “Transition of aggregate demand and
supply and current status of survey of natural
resources.” Cement & Concrete, 415(9), 6-13. (in
Japanese)
Jennings, H. M., (2008). “Refinements to colloid model
of C-S-H in cement: CM-II.” Cement and Concrete
Research, 38(3), 275-289.
Kakizaki, M., Idei, Y., Sukegawa, T., Akutsu, Y., Hatano
H. and Kurioka, H., (1996). “Study on environmental
and mechanical properties of irradiated concrete.”
Journal of structural and construction engineering,
488, 1-10. (in Japanese)
Kishitani, K. and Baba, A., (1975). “An essay on the
composite mechanisms of drying shrinkage of
concrete.” Transactions of the Architectural Institute
of Japan, 234, 1-11. (in Japanese)
Kontani, O., Y. Ichikawa, A. Ishizawa, M. Takizawa and
O. Sato (2010). “Irradiation effects on concrete
structure.” In: Int. Symp. On the Ageing Manag. of
Nucl. P. P., 173-182.
Kontani, O., Ichikawa, Y., Ishizawa, A., Takizawa M.
and Sato, O., (2013). “Irradiation effects on concrete
structures.” Infrastructure Systems for Nuclear Energy,
459-473.
Kontani, O., Ishizawa, A., Maruyama, I., Takizawa M.
and Sato, O., (2012). “Evaluation of irradiation effects
on concrete structure.” In: Proceedings of the 2012
International Congress on Advances in Nuclear Power
Plants, Chicago,
Kontani, O., Sawada, S., Maruyama, I., Takizawa, M.
and Sato, O., (2013). “Evaluation of irradiation effects
on concrete structure - Gamma ray irradiation tests on
cement paste.” In: Proceedings of the ASME 2013
Power Conference POWER2013.
Kubo, Y., Ueda, T., Kuroda, T. and Nomura, M., (2006).
“Influence of ASR expansion on mechanical proper-
ties of concrete.” Proceedings of the Japan Concrete
161
Institute 28(1), 1691-1696.
Le Pape, Y., (2015). “Structural effects of
radiation-induced volumetric expansion on unrein-
forced concrete biological shields.” Nuclear
Engineering and Design, 295, 534-548.
Le Pape, Y., Field, K. G., and Remec, I., (2015).
“Radiation effects in concrete for nuclear power plants,
Part II: Perspective from micromechanical modeling.”
Nuclear Engineering and Design, 282, 144-157.
Litvan, G. G., (1976). “Variability of the nitrogen surface
area of hydrated cement paste.” Cement and Concrete
Research, 6(1), 139-143.
Lowinska-Kluge, A. and Piszora, P., (2008). “Effect of
gamma irradiation on cement composites observed
with XRD and SEM methods in the range of radiation
dose 0-1409 MGy.” Acta Physica Polonica-Series A
General Physics, 114(2), 399.
Maruyama, I. and Igarashi, G., (2015). “Numerical
approach towards aging management of concrete
structures: Material strength evaluation in a massive
concrete structure under one-sided heating.” Journal
of Advanced Concrete Technology 13(11), 500-527.
Maruyama, I., Igarashi, G. and Nishioka, Y., (2015).
“Bimodal behavior of C-S-H interpreted from
short-term length change and water vapor sorption
isotherms of hardened cement paste.” Cement and
Concrete Research, 73(7), 158-168.
Maruyama, I., Kontani, O., Ishizawa, A., Takizawa, M.
and Sato O., (2012). “Development of system for
evaluating concrete strength deterioration due to
radiation and resultant heat.” In: Proceedings of 3rd
International Conference on NPP Life Management
for Long Term Operations, IAEA-CN-194-096, Salt
Lake City, May.
Maruyama, I., Kontani, O., Sawada, S., Sato, O., Igarashi
G. and Takizawa M., (2013). “Evaluation of irradiation
effects on concrete structure-background and prepara-
tion of neutron irradiation test.” In: POWER2013,
American Society of Mechanical Engineers, Power
Division.
Maruyama, I., Nishioka, Y., Igarashi, G. and Matsui, K.,
(2014a). “Microstructural and bulk property changes
in hardened cement paste during the first drying
process.” Cement and Concrete Research, 58(4),
20-34.
Maruyama, I., Sasano, H., Nishioka Y. and Igarashi, G.,
(2014b). “Strength and Young's modulus change in
concrete due to long-term drying and heating up to
90 °C.”Cement and Concrete Research, 66(4), 48-63.
Maruyama, I. and Sugie, A., (2014). “Numerical study
on drying dhrinkage of concrete affected by aggregate
size.” Journal of Advanced Concrete Technology,
12(8), 279-288.
Muller, A. C. A., Scrivener, K. L., Gajewicz, A. M. and
McDonald, P. J., (2012). “Densification of C–S–H
measured by 1H NMR relaxometry.” The Journal of
Physical Chemistry C, 117(1), 403-412.
Nakano, K., Kobayashi, S. and Nakagami, A., (1987).
I. Maruyama, K. Haba, O. Sato, S. Ishikawa, O. Kontani and M. Takizawa / Journal of Advanced Concrete Technology Vol. 14, 144-162, 2016
“Degree of damage due to alkari-silica reaction and
physical properties of concrete.” Cement Association
Japan (CAJ) Review of the 41st General Meeting /
Technical Session 41, 419-422.
Parrott, L. J., Hansen, W. and Berger, R. L., (1980).
“Effect of first drying upon the pore structure of
hydrated alite paste.” Cement and Concrete Research,
10(5), 647-655.
Parrott, L. J. and Young, J. F., (1981). “Effect of
prolonged drying upon the silicate structure of
hydrated alite pastes.” Cement and Concrete Research,
11(1), 11-17.
Pedersen, A., (1971). “Radiation damage in concrete -
Measurements on miniature specimens of cement
mortar.” In: Proceedings of an Information Exchange
Meeting on 'Results of Concrete Irradiation
Programmes'.
Pickett, G., (1956). “Effect of aggregate on shrinkage of
concrete and a hypothesis concerning shrinkage.” ACI
Journal proceedings, 52(1), 581-590.
Pihlajavaara, S. E., (1974). “A review of some of the
main results of a research on the ageing phenomena of
concrete: Effect of moisture conditions on strength,
shrinkage and creep of mature concrete.” Cement and
Concrete Research, 4 (5), 761-771.
Pomaro, B., Salomoni, V. A., Gramegna, F., Prete, G. and
Majorana, C. E., (2011). “Radiation damage evalua-
tion on concrete within a facility for Selective
Production of Exotic Species (SPES Project), Italy.”
Journal of Hazardous Materials, 194, 169-177.
Primak, W., (1958). “Fast-neutron-induced changes in
quartz and vitreous silica.” Physical Review, 110(6),
1240-1254.
Primak, W., Fuchs, L. and Day, P., (1955). “Effects of
nuclear reactor exposure on some properties of
vitreous silica and quartz.” Journal of the American
Ceramic Society, 38(4), 135-139.
Rhoades, W. A., (1978). “The GIP program for
preparation of group-organized cross-section librar-
ies.” Informal notes (RSIC CODE PACKAGE
PSR-229 MICRO).
Salomoni, V. A., Majorana, C. E., Pomaro, B., Xotta G.
and Gramegna, F., (2014). “Macroscale and mesoscale
analysis of concrete as a multiphase material for
biological shields against nuclear radiation.” Interna-
tional Journal for Numerical and Analytical Methods
in Geomechanics, 38(5), 518-535.
Seeberger, J. and Hilsdorf, H., (1982). “Einfluß von
Radioactiver Strahlung auf die Festogkeit and
Struktur von Beton.” Technical Report NR 2505,
Institut für Massivbau and Baustofftechnologie,
Universität Karlsruhe.
Sekimura, N., (2015). “Roadmap on R&D and Human
Resource for Light Water Reactors Safety in Japan,
-Nuclear Safety Visions and Technical basis
Reconstruction after the Fukushima Accident -.” In:
International Experts Meeting on Strengthening
162
Research and Development Effectiveness in the Light
of the Accident at the Fukushima daiichi Nuclear
Power Plant, Vienna, Austria, International Atomic
Energy Agency.
Sereda, P. J., Feldman, R. F. and Swenson, E. G., (1966).
“Effect of sorbed water on some mechanical properties
of hydrated portland cement pastes and compacts.”
Highway Research Board Special Report 90, 58-73.
Son, H. N. and Hosoda A., (2010). “Detection of
microcracking in concrete subjected to elevated
temperature at very early age by acoustic emission.”
Journal of Advanced Concrete Technology, 8(2)
201-211.
Sugiyama, H. and Masuda, Y., (1999). “Strength
development of concrete cured under high temperature
conditions in an early age.” Journal of structural and
construction engineering, 515, 23-30. (in Japanese)
Swamy, R. N. and Al-Asali, M. M., (1988). “Engineering
properties of concrete affected by alkali-silica reac-
tion.” Materials Journal, 85(5), 367-374.
The Nuclear Regulation Authority Japan, (2015).
“Approval of operational safety programs for Kyushu
electric power company's Sendai nuclear power
station units 1 and 2.” Tokyo, NRA.
Thomas, J. J., Allen, A. J. and Jennings, H. M., (2008).
“Structural changes to the calcium–silicate–hydrate
gel phase of hydrated cement with age, drying, and
resaturation.” Journal of the American Ceramic
Society, 91(10), 3362-3369.
Thomas, J. J. and Jennings, H. M., (2003). “Changes in
the size of pores during shrinkage (or expansion) of
cement paste and concrete.” Cement and Concrete
Research, 33(11), 1897-1900.
Tomita, Y., Kousa, K., Nakano K. and Nakagami, A.,
(1988). “Diagnosis method using core sampling for
concrete structure suffering from ASR.” Review of the
42st General Meeting / Technical Session, Cement
Association Japan (CAJ), 42, 335-338. (in Japanese)
Ueda, N., Nakamura, H., Kunieda, M., Maeno, H.,
Morishit, N. and Asai, H., (2011). “Evaluations on
ASR damage of concrete structure and its structural
performance.” Doboku Gakkai Ronbunshuu E 67 1
28-47. (in Japanese)
Vodák, F., Trtík, K., Sopko, V., Kapičková, O. and Demo,
P., (2005). “Effect of γ-irradiation on strength of
concrete for nuclear-safety structures.” Cement and
Concrete Research, 35(7), 1447-1451.
Weber, W. J., (1990). “Radiation-induced defects and
amorphization in zircon.” Journal of Materials
Research, 5(11), 2687-2697.
Weissmann, S. and Nakajima, K., (1963). “Defect
Structure and density decrease in neutron‐irradiated
quartz.” Journal of Applied Physics ,34(3), 611-618.
Wittels, M. and Sherrill, F. A., (1954). “Radiation
damage in SiO2 structures.” Physical Review, 93(5),
1117-1118.
Journal of Advanced Concrete Technology Vol. 14, 706-716, November 2016 / Copyright © 2016 Japan Concrete Institute 706

Scientific paper

Change in Relative Density of Natural Rock Minerals Due to Electron

Irradiation
Ippei Maruyama1* and Shunsuke Muto2

Received 18 July 2016, accepted 3 November 2016 doi:10.3151/jact.14.706

Abstract
We investigated changes in the density of natural rock minerals following high-energy electron irradiation, using the
plasmon peak shift of electron energy-loss spectra and transmission electron microscopy. The target materials were the
natural rock minerals α-quartz, orthoclase, anorthite, albite, biotite, muscovite, and chlorite. These crystalline minerals
can be classified into three groups based on their Si-network geometries: 3-dimensional 6-member ring; 4-member ring +
6-member ring; and planar 6-member ring. The metamictization rates and changes in relative density are discussed using
a phenomenological model, which we used to identify the physical parameters that describe the metamictization process
as a function of the volume density of Si and Al atoms, or Si atoms alone, in the crystal structures. The relative densities
following metamictization all decreased by more than a few percent, except for albite, which became denser. These results
suggest that radiolysis damage causes initial compaction, then metamictization, characterized by the expansion of the Si-
and Al-polyhedra in the aggregate. The stability of concrete containing α-quartz, orthoclase, and anorthite should be
further investigated in the light of the present results.

1. Introduction concrete. Maruyama et al. (2013) then summarized the

Light water reactors (LWRs) are considered an important
power supplier even after the Fukushima Daiichi acci-
dent. A Japanese cabinet decision passed on April 11,
2014, nuclear electric power is considered as an impor-
tant base load power. Safe long-term operation of nuclear
power plants (NPPs) will be required in the near future in
Japan, because construction of new NPPs will be con-
siderably difficult.
From a viewpoint of aging management and plant life
management for concrete structures of LWR, deteriora-
tion of concrete is a key research topic, because concrete
structures are hardly replaced. While deteriorations of
concrete structures have been extensively researched in
civil engineering field, there still remains the character-
istic deterioration environment, i.e., irradiation, fully
unexploited in NPPs (Rosseel et al. 2016). Biological
shielding concrete walls (in case of a pressurized water
reactor; PWR) or reactor vessel support pedestals (in
case of a boiling water reactor; BWR) are irradiated by
neutrons and gamma rays from the reactor during op-
eration.
The neutron and gamma-ray irradiation effects on the
physical properties of concrete has been firstly summa-
rized by Hilsdorf et al. (1978), whose results, however,
included concrete and mortar not representative of LWRs
1 natural minerals.
Professor, Graduate School of Environmental Studies,
Nagoya University, Nagoya, Japan. *Corresponding
author, E-mail: ippei@dali.nuac.nagoya-u.ac.jp
2
Professor, Advanced Measurement Technology Center,
Institute of Materials and Systems for Sustainability,
Nagoya University, Nagoya, Japan.
relationship between the fast neutron fluence and relative
compressive strength for concrete or mortar applicable to
concrete members in LWRs. They confirmed a clear
trend of the decrease in the relative strength with the
fluence for more than 1–3 × 1019 n/cm2. This trend is
almost consistent to the integrated data of irradiated
concrete recently reported by Field et al. (2015).
It has been reported that deteriorated concrete showed
expansion after irradiation (Dubrovskii et al. 1970;
Elleuch et al. 1971; Elleuch et al. 1972; Hilsdorf et al.
1978; Seeberger and Hilsdorf 1982). The main reason is
considered as an expansion of aggregate, because
α-quartz is well-known to show an expansion and den-
sity reduction under neutron and electron irradiation
(Primak et al. 1955; Primak 1958; Bonnet et al. 1994;
Douillard and Duraud 1996; Bolse 1999; Ewing et al.
2000; Field et al. 2015). It is hence considered that the
natural rock should be sensitive to irradiation, which
poses very important issues on the durability of concrete
under irradiation as 1) expansion of aggregate, and 2)
degradation of concrete performance as a result of 1)
(Maruyama et al. 2012; Field et al. 2015; Giorla et al.
2015; Le Pape et al. 2015, 2016; Maruyama et al. 2016).
In the present contribution, we conducted electron ir-
radiation experiments to examine how the sili-
cate/aluminate unit polyhedral aggregates respond to
irradiation, especially of which radiolysis impact, with
respect to the density change in several fundamental
2. Experimental procedure
2.1 Materials
We examined the following materials in the experiment:
I. Maruyama and S. Muto / Journal of Advanced Concrete Technology Vol. 14, 706-716, 2016 707

Table 1 Chemical composition, origin, density, molar mass, and silicate structure of each natural rock mineral analyzed.
Chemical composition Density Molar mass Silicate (includ-
Mineral Origin
(General form) (g/cm2) (g/mol) ing Al) structure
3-dimensional
α-quartz SiO2 Nellore, Andhra Pradesh, India 2.65 60.08
6-ring
Orthoclase KAlSi3O8 Nellore, Andhra Pradesh, India
2.56 278.33 4 ring + 6 ring
2.73
Anorthite CaAl2Si2O8 Yoichi, Hokkaido, Japan 277.41 4 ring + 6 ring
(2.72 - 2.75)
Governador Valadares, Minas, Gerais, 2.62
Albite NaAlSi3O8 263.02 4 ring + 6 ring
Brazil (2.61 - 2.63)
3.09
Biotite K(Mg,Fe)3(AlSi3O10)(F,OH)2 Nellore, Andhra Pradesh, India 433.53 Planate 6 ring
(2.8 - 3.4)
2.82
Muscovite KAl2(AlSi3O10)(F,OH)2 Ishikawayama, Ishikawa, Ishiwaka, Japan 398.71 Planate 6 ring
(2.77 - 2.88)
2.65
Chlorite (Mg,Fe)5Al(Si3Al)O10(OH)8 Miwa, Nagasaki, Japan 595.22 Planate 6 ring
(2.55 - 2.75)

α-quartz; orthoclase; anorthite; albite; biotite; musco-
vite; and chlorite (Table 1). All samples were supplied
by Nichika Co., Japan. The powder X-ray diffraction
patterns of the samples confirmed that each sample con-
sisted of a single phase (Fig. 1).
tional to its atomic density, and the relative volume
change can be estimated from the displacement of the
plasmon peak. Drude model was originally developed
by assuming the free-electron gas model for the valence

2.2 Method
Pristine samples were ground into powders with parti-
cles < 25μm in diameter, using a planetary ball mill. The
ground powder was sprinkled onto a carbon microgrid
supported by a copper mesh disc, which was mounted to
a transmission electron microscope (TEM) sample
holder. The holder was inserted into a Jeol JEM2100
TEM equipped with a Gatan ENFINA 1000 electron
energy-loss spectrometer (EELS). The TEM was oper-
ated at 100 kV, mostly below the threshold for knock-on
damage to these minerals. Electron flux was measured
using an electron current meter placed in the TEM fluo-
rescent screen.
Changes in the low-loss spectrum of α-quartz with Fig. 1 X-ray diffraction data for the rock minerals ana-
irradiation time are shown in Fig. 2. The strongest peaks lyzed.
at zero energy loss are the zero-loss peak, and the
broader peaks at ~22-25 eV correspond to the plas-
mon-loss peaks. During electron irradiation, amorphiza-
tion was confirmed by the occurrence of halo rings in
the electron diffraction pattern, as shown for α-quartz in
Fig. 3.
Plasmon energy Ep is expressed by Eq. (1), based on
the Drude model (Raether 1982, Egerton 2009):

ne 2 N
Ep = = ∝ , (1)
ε0m V

where n is the density of the valence electrons, e is elec-

tron charge, m is effective electron mass, = is the re-
duced Planck’s constant, and ε0 is the vacuum permittiv-
ity. Equation (1) indicates that plasmon energy is pro-
portional to the square root of valence electron density.
Assuming the composition and the chemical bonding
character of the sample remain more or less constant,
the plasmon peak position is supposed to be propor- Fig. 2 Changes in the low-loss spectrum of -quartz with
irradiation time.
I. Maruyama and S. Muto / Journal of Advanced Concrete Technology Vol. 14, 706-716, 2016 708

electrons in metals, though it has been well established Table 2 Change in density of rock minerals and the
that the model is applicable to materials having covalent electron dose required for metamictization.
and/or ionic bonds when the plasmon peak energy is Minerals Change in relative density Electron dose
significantly larger than the band gap energy (Egerton
α-quartz −4% 6 × 1020 e/cm2
2011) and Eq. (1) provides reasonable agreement with
experimental values. It is thus considered that the pre- Orthoclase −2% 2 × 1021 e/cm2
sent analyses can be effective unless a significant com- Anorthite < −3% >2 × 1021 e/cm2
positional change or appreciable phase separation is Albite +4% 8 × 1021 e/cm2
observed during the irradiation. Biotite −8%. 1 × 1022 e/cm2
Muscovite −4% 1 × 1023 e/cm2
3. Results Chlorite < −6% >7 × 1023 e/cm2

The changes in the relative density of the rock minerals

tested, as a function of electron dose, are summarized in
Fig. 4 and Table 2. Each mineral demonstrated different
degrees of tolerance to electron irradiation. The relative
density of α-quartz and orthoclase decreased monotoni-
cally, then asymptotically approached its final value. For
α-quartz this was −4%, at a dose of ~6 × 1020 e/cm2, and
dose at metamictization was similar to the previous re-
sults (Inui et al. 1990). Orthoclase exhibited a greater
tolerance and reached its final value of about −2% at ~2
× 1021 e/cm2. The metamictization of α-quartz pro-
ceeded very rapidly.
Anorthite showed a different pattern. The relative
density initially increased to 3% at ~3 × 1022 e/cm2, then
decreased in a manner similar to orthoclase and α-quartz.
Although the electron irradiation was terminated when
complete metamictization was confirmed (based on the
appearance of a halo pattern in the electron diffraction,
as shown in Fig. 4(c)), the change in relative density
was not yet complete. The final change in density fol-
lowing metamictization was estimated by extrapolating
the curve in Fig. 4(c) until it reached a decrease of more
than 3%. Electron diffraction patterns (see inset) sug-
gests that the irradiation destroyed the long-range struc-
ture first, followed by the shorter-range structures.
Albite behaved in a completely different manner from
the above minerals. Its density first decreased to −8%, at
~2 × 1021 e/cm2, then suddenly reversed the direction of
change and increased. It finally reached metamictization
at more than 8 × 1021 e/cm2 and a corresponding density
change of +4%.
Biotite and Muscovite had a two-stage change in
density: a gradual reduction followed by an exponential
decrease. The first stage ended at ~1 × 1022 e/cm2 and 1
× 1021 e/cm2, respectively, at −2.5% for both minerals,
and the second stage at `1 × 1023 e/cm2 and 1 × 1022
e/cm2, respectively, with final changes of −8% and
−4%.
Of the seven minerals tested, chlorite exhibited the
highest tolerance for electron irradiation. Its density
initially decreased to −1%, at up to 1 × 1021 e/cm2, then
increased until it reached a maximum of +3% at a dose
of 2 × 1022 e/cm2, then finally decreased again to its
eventual stable state, a 6% density reduction, at ~7 ×
1023 e/cm2. The electron diffraction pattern implies that
metamictization was incomplete even at a dose of ~1 ×
1022 e/cm2 (Fig. 4(g)).

Fig. 3 TEM images and electron diffraction images of α-quartz before irradiation and after amorphization, under 100 KeV
electron irradiation.
I. Maruyama and S. Muto / Journal of Advanced Concrete Technology Vol. 14, 706-716, 2016 709

(a) α-quartz (b) Orthoclase

(c) Anorthite (d) Albite

(e) Biotite (f) Muscovite

(g) Chlorite
Fig. 4 Change in relative density of rock minerals as a function of electron dose at 100 KeV. Inset images show the electron
differentiation patterns used to identify the completion of metamictization.
I. Maruyama and S. Muto / Journal of Advanced Concrete Technology Vol. 14, 706-716, 2016
4. Discussion
Minerals that are largely based on silicates consist of a
flexibly joined network of Si-Ox polyhedra, with heavier
cations located in the spacious channels between the
polyhedra. In the case of α-quartz, the metamictization
process was proposed previously (Hobbs and Pascucci
1980; Hobbs 1995). These authors suggested that the
tetrahedral Si-O4 network is metamictized by electron
irradiation via the excitonic excitation of the Si-O bond,
which results in the configurational instability of the
oxygen atom. This atom moves to an off-center position,
immediately connecting with a neighboring oxygen
atom to form a peroxy geometry. This generation of a
Frenkel pair (an oxygen vacancy and a molecular oxy-
gen crowdion interstitial) effectively breaks the silica
network, and the adjacent SiO4 units are relaxed to form
flexibly linked triangular structures (SiO3). When the
density of such affected structures reaches a critical
level, associated local bond-breaking and stochastic
local reconstruction will bring about topological disor-
dering (Hobbs 1995). This final state is called a
metamict state.
In this study, we confirmed metamictization by the
appearance of a halo ring pattern in the electron diffrac-
tion insets in Fig. 4, particularly those in (e), (f), and (g),
in which the halo ring was apparently exhibited in the
lower-order superlattice reflections, while the funda-
mental reflections remained spotty. This suggests that
the metamictization began with the deformation of the
long-range superstructures, followed by the collapse of
the shorter-range structures, possibly the unit polyhedral
structures. During the initial process of long-range
structural deformation, a slight increase in density was
common (e.g. Figs. 4(b), (c), and (g)). This confirms the
findings in a previous report that α-quartz showed a
slight increase in density under a low-dose regime
(those experimental data showed slight hardening;
(Nakano et al. 2005)). The metamictization dose for
SiO4 tetrahedra is reported to be ~7 eV/atom (Eby et al.
1992), and the displacement threshold energies for Si
and O are 11-22 eV and 9 eV respectively (Inui et al.
1990; Hobbs et al. 1994). Al2O3 and MgO also exhibit
metamictization as a result of electron irradiation. The
displacement threshold energies for Al and O in Al2O3
are ~20 eV and 50~75 eV respectively, and the
metamictization dose is 3500 eV/atom, while for Mg
and O they are 60 and 65 eV/atom respectively, and the
metamictization dose for MgO is 6000 eV/atom (Hobbs
1994). It is hence thought that ionic bonded materials
are more tolerant of irradiation (Hobbs et al. 1994), and
that the metamictization and resultant relative density
changes observed in this study are attributable to radio-
lysis by way of the deformation of the silica (or alumi-
num) bonding network.
The change in relative density following electron ir-
radiation that we observed for α-quartz was about −4%,
much smaller than the effect of neutron irradiation,
710
which is reported as 14−15% (Primak 1958; Lell et al.
1968; Bykov et al. 1981). Other research has shown that
the metamictized structure resulting from neutron irra-
diation is not the same as that caused by electron irra-
diation (Hobbs 1995). This suggests that electron irra-
diation experiments do not simulate the metamictization
process of rock minerals caused by neutron irradiation.
A previous electron irradiation experiment, using an
electron energy of > 1500 KeV, revealed a ballistic ef-
fect (or knock-on impact) on metamictization, and the
resultant EELS structure evaluation showed a new and
unexpected structure.
Each of the minerals we tested had a different toler-
ance for 100 KeV-electron irradiation, with α-quartz
having the lowest and chlorite the highest. We found
that the ratio of the number of Si and Al atoms to the
total number of atoms in each mineral was roughly
proportional to the metamictization dose (Fig. 5), sug-
gesting that the crystal structure, especially the density
of tetrahedral units containing Si or Al, is a key factor
for metamictization tolerance.
In the following we examine metamictization in
terms of more fundamental processes. To exploit the
common physical properties of these materials, we in-
troduce a phenomenological model used to describe
mechanical property changes in ion-irradiated ceramic
materials (Nakano et al. 2005). We assume that the rela-
tive density as a function of irradiation dose, R, is ex-
pressed by the sum of the densities of initial crystalline,
intermediate, and final metamict phases:
R = R1C1 + R2 C2 + R3C3 (2)
C1 + C2 + C3 = 1 (3)
where C1, C2, and C3 are the relative fractions of the
three phases respectively, and R1, R2, and R3 are the
relative densities of the three phases (R1 should thus be
unity). The nominal three phases were introduced to
model the changes in density of the materials using a set
of simple, common parameters, which should be func-
tions of the numbers of defects involved and of local
Fig. 5 Relationship between metamictization dose and
the ratio of number of Si and Al atoms to total atoms.
I. Maruyama and S. Muto / Journal of Advanced Concrete Technology Vol. 14, 706-716, 2016
Table 3 Derived parameters for equations (2), (3), and (4), based on the metamictization process.
Minerals b σ1
α-quartz 1.4E−02 5.30E−21
Orthoclase 1.5E−04 5.00E−20
Anorthite 5.0E−04 1.80E−21
Albite 1.0E−02 7.00E−20
Biotite 2.3E−07 1.50E−23
Muscovite 1.3E−04 1.50E−21
Chlorite 7.5E−06 1.00E−22
structures. The metamictization process can be ex-
pressed using the following reaction rate equations:
⎧ dC1
⎪ dt = −φσ 1C1 + bC2
⎪
⎪ dC2
⎨ = φσ 1C1 − φσ 2 C2 − bC2
⎪ dt
⎪ dC3
⎪ dt = φσ 2 C2
⎩ the other hand, has a 6-member Si ring wedge between
where φ is electron flux (e/cm2), and σ 1 and σ 2 are
the effective cross sections for damage production from
the initial crystalline phase to the intermediate phase,
and from the intermediate phase to the metamict phase,
respectively. We assume that the intermediate phase can
be thermally reversed to the initial phase, whereas the
metamict phase is irreversible once reached. This as-
sumption is drawn from the facts that (neutron-)
metamicted structure showed consistent volume change
regardless temperature during irradiation (Bykov et al.
1981). It should be noted that this model is quite simpli-
fied but this has been successfully applied to predict the
physical property of electron-irradiated materials
(Nakano et al. 2005). The set of equations (4) can be
solved analytically, and the experimental data were fit to
the solutions using least-squares (solid lines in Fig. 4) to
yield the parameters σ 1 , σ 2 , b, R2, and R3. (Table 3).
The thermal recovery parameter b can take arbitrary
small values for orthoclase and biotite, where the cross
sections for damage formation, σ 1 and σ 2 , are not
significantly different from each other. This suggests
that metamictization proceeds according to a single ra-
diolysis process in those minerals. On the other hand,
the other minerals exhibited considerable thermal re-
covery and different values for σ 1 and σ 2 , which
suggests that metamictization follows a two-step proc-
ess in those minerals. The derived cross section values
are also reasonable for radiolysis (Ugurlu et al. 2011).
These results are obviously supporting that the applied
eq. (4) is good enough as the first approximation of the
observed phenomenon and not far from the real phe-
nomenon.
The σ 1 s values for the seven minerals are plotted as
a function of the density of Si and Al atoms in Fig. 7(a),
since the bonding of the oxygen atom in Si-O-Si,
Si-O-Al, or Al-O-Al is likely the most susceptible point
for irradiation damage, and the density of Si and Al can
711
σ2 R2 R3
1.20E−20 1.01E+00 9.55E−01
6.00E−22 1.01E+00 9.60E−01
4.00E−23 1.04E+00 9.60E−01
7.00E−21 9.00E−01 1.05E+00
1.50E−22 1.01E+00 9.10E−01
5.00E−22 9.70E−01 9.55E−01
6.00E−24 1.04E+00 9.40E−01
thus be used as a measure of damage tolerance. All the
data points except that for α-quartz exhibit a linear rela-
tionship. Muscovite, chlorite, and biotite consist of a
planar 6-ring Si structure (as does Al to some extent and
in some cases), which is thought to be more tolerant of
irradiation than a 6-member ring and 4-member ring
(4) structure or a 3-dimensional 6-ring structure. Biotite is
characterized by 6-member Si ring wedges and K+ and
Mg/Al-O layers interpenetrating each other. Chlorite, on
a Mg/Al-O layer and interlayer Mg2+/Al2+ cations, and
muscovite includes a 6-member Si ring wedge between
interlayer K+ cations and Al-O layers. Of these three
minerals, muscovite shows the largest cross section σ 1 ,
probably because of the larger number of Al-O-Al,
Al-O-Si, and Si-O-Si networks it contains.
Albite, orthoclase, and anorthite have 4-member Si
ring and 6-member Si ring structures, with larger cations
in the channels between the networks, and have larger
σ 1 values than the minerals with planar 6-member ring
structures. Among these three minerals, albite has the
largest σ 1 , which can be explained by the Si-O-Si an-
gles linking the Si-polyhedra being more flexible due to
the small cations (Na+) in the spacious channels of the
Si-networks. In this context, α-quartz is an outlier (Fig.
7(a)), probably because it has a greater capacity for the
Si-O-Si angles to change, since it has no other cations in
the channels to resist local deformation.
The relationship of the thermal recovery parameter, b,
relative to the density of Si and Al atoms is similarly
linear, and α-quartz is not as much of an outlier (Fig.
7(b)). We can thus conclude that both σ 1 and b are
governed by the flexibility inherent in their Si-O-Si,
Si-O-Al, and Al-O-Al networks and the location of
other cations in these networks. Larger cations probably
play a role in restricting the change in angle possible in
these networks.
The value of σ 2 , the effective cross section of trans-
formation from the intermediate phase to the metamict
phase, is thought to be related to the collapse of Si-O-Si
bonds. Plotting σ 2 as a function of the density of Si
atoms (Fig. 7(c)) yielded a strong linear relationship,
although this parameter did not show a good correlation
with the density of Si and Al atoms. We therefore con-
clude that the final metamictization process is mainly
attributable to the collapse of Si-O-Si bonds. However,
α-quartz was an outlier in this respect as well, and ex-
hibited different tendencies than the other minerals.
I. Maruyama and S. Muto / Journal of Advanced Concrete Technology Vol. 14, 706-716, 2016
The relative density of the final metamictized phase
was closely related to the extent of volume expansion
after metamictization, which is crucial for concrete per-
formance under radiolysis impact (such as gamma-ray
irradiation). When plotting the corresponding parameter,
R2, as a function of the density of Al and Si atoms, all
tested minerals except albite appear to lie on a curve
(Fig. 8). The structures containing planar 6-member
(a) σ 1 (cm2/e)
(b) b (1/s)
(c) σ 2 (cm2/e)
Fig. 7 Relationships between parameters for rate of
metamictization by electron irradiation and the index
obtained for the crystal structure.
712
rings tended to show significant expansion despite hav-
ing a higher tolerance for electron irradiation. We note
that among the minerals that have similar Si and Al
atom densities, anorthite and orthoclase exhibited an
almost identical reduction in density, whereas albite
underwent a substantial increase in density. This distinct
difference is probably explained by the size of the
cations in the spacious channels of the Si (or Al) net-
works. In albite, Na+ ions are small enough to allow the
Si-O-Si units to shrink, and stabilize the structure fol-
lowing shrinkage. If there were no cations within the
Si-network, the SiOx polyhedra would expand at
metamictization and then be stable, as in the case of
α-quartz.
It has been already suggested that understanding the
metamictization process due to neutron irradiation of
α-quartz should take into account both the radiolysis
impact and the ballistic impact. And both of them is not
clarified for other rock minerals. Even in the case of
neutron irradiation experiment, we always have an im-
pact of gamma-rays (radiolysis) from reactor or secon-
dary effect of neutron. Therefore, better understanding
of radiolysis impact on rock forming minerals is impor-
tant for volume expansion under neutron and
gamma-ray irradiation. From a such view point, the
present study is significant. And our findings are that
nominal sections for 2 different phases under metamic-
tization and thermal recovery/annealing parameter are
determined based on simplified phenomenalogical
equations and the obtained nominal section and thermal
recovery parameters are discussed through the indices
obtained by crystal structures. As a results, the density
of Si and Al atoms are key parameters for those.
The final density change of radiolysis impact (4% for
α-quartz) is far from the value of neutron-irradiated
samples (15% for α-quartz). This supports that ballistic
impact has a large impact on the final density change of
metamictized condition and radiolysis impact is less,
while the incubation process, in which thermal anneal-
ing has a large impact, should take into account the ra-
diolysis impact.
Fig. 8 Relative density of the metamictization phase of the
seven minerals tested.
I. Maruyama and S. Muto / Journal of Advanced Concrete Technology Vol. 14, 706-716, 2016
5. Conclusion
We investigated the effects of electron irradiation on the
relative density of the natural rock minerals α-quartz,
orthoclase, anorthite, albite, biotite, muscovite, and
chlorite, in terms of plasmon peak shift, using
TEM-EELS. Each mineral showed a substantially dif-
ferent tolerance for electron fluence, in terms of the ex-
tent of its amorphization, as confirmed by electron dif-
fraction and the change in relative density. The relative
density increased only for albite, and decreased for the
other minerals. The changes in density amounted to
several percent, which would likely be non-negligible for
components of concrete. The density change due to
electron-irradiation was 4% in case of α-quartz and this
value is far from the density change due to neu-
tron-irradiation. Consequently, the present elec-
tron-irradiation experiments are limited to understand the
radiolysis impact. Applying phenomenological kinetic
reaction rate equations to this process, we found that the
metamictization rate was linearly correlated with the
density of Si and Al atoms or of Si atoms alone. Crystal
structures containing Si networks of 4- and 6-member
rings, such as albite, orthoclase, and anorthite, had al-
most identical tolerance for electron irradiation to that
of α-quartz. We also found that the relative density of the
metamict phase could be expressed as a smooth function
of the density of Si and Al atoms.
The fact that irradiation induced swelling in these
mineral aggregates is an important issue for the concrete
used in LWRs. These results highlight the importance of
neutron and gamma-ray irradiation experiments on the
natural rock minerals that are potentially used in con-
structing LWRs. Aggregates in concrete members that
are subject to neutron and gamma-ray irradiation are
thought to be degraded by radiolysis impact as well as
ballistic impact, and would expand in volume if they
contain the minerals we have studied here. Several as-
pects of this process, highlighted in the present study,
remain to be clarified. These include the energy and
temperature dependencies of the volume expansion, the
metamictizaion and recrystallization processes of each
mineral under neutron and gamma-ray irradiation (Bolse
1999), and how all of these aspects affect composites,
including cracking at mineral grain boundaries due to
heterogeneous volume changes of the constituent min-
erals in an aggregate.
Acknowledgements
This work was partially supported by a Grant-in-Aid on
Innovative Areas “Nano Informatics” (Grant No.
25106004) from the Japan Society for the Promotion of
Science and crystal figures in appendix were drawn with
VESTA.
References
Bolse, W., (1999). “Amorphization and recrystallization
of covalent tetrahedral networks.” Nuclear Instru-
713
ments and Methods in Physics Research, Section B:
Beam Interactions with Materials and Atoms, 148(1-4),
83-92.
Bonnet, J. P., Boissier, M. and Ait Gherbi, A., (1994).
“The amorphization process of neutron-irradiated
crystalline quartz studied by Brillouin scattering.”
Journal of Non-Crystalline Solids, 167(1–2), 199-204.
Bykov, V. N., Denisov, A. V., Dubrovskii, V. B.,
Korenevskii, V. V., Krivokoneva, G. K. and
Muzalevskii, L. P., (1981). “Effect of irradiation
temperature on the radiation expansion of quartz.”
Soviet Atomic Energy, 51(3), 593-595.
Douillard, L. and Duraud, J. P., (1996). “Amorphization
of α-quartz under irradiation.” Journal de Physique III,
6 (12), 1677-1687.
Dubrovskii, V. B., Ibragimov, S. S, Korenevskii, V. V.,
Ladygin, A. Y., Pergamenshchik, V. K. and Perevalov,
V. S., (1970). “Hematite concrete for shielding against
high neutron fluxes.” Soviet Atomic Energy-Ussr,
28(3), 336-338.
Eby, R. K., Ewing, R. C. and Birtcher, R. C., (1992).
“Amorphization of complex silicates by ion-beam
irradiation.” Journal of Materials Research, 7(11),
3080-3102.
Egerton, R. F., (2009). “Electron energy-loss spectros-
copy in the TEM.” Reports on Progress in Physics,
72(1), 016502.
Egerton, R. F., (2011). “Electron energy-loss spectros-
copy in the electron microscope.” Third. Springer
Science & Business Media.
Elleuch, L. F., Dubois, F. and Rappeneau, J., (1972).
“Effects of neutron radiation on special concretes and
their components.” Special Publication, 34.
Elleuch, M. R., Dubois, F. and Rappenau, J., (1971).
“Behavior of special shielding concretes and of their
constituents under neutron irradiation.” Fourth United
Nations International Conference on the Peaceful
Uses of Atomic Energy, 7.
Ewing, R. C., Meldrum, A., Wang, L. and Wang, S.,
(2000). “Radiation-induced amorphization.” Reviews
in Mineralogy and Geochemistry, 39, 319-361.
Field, K. G., Remec, I. and Le Pape, Y., (2015).
“Radiation effects in concrete for nuclear power plants
– Part I: Quantification of radiation exposure and
radiation effects.” Nuclear Engineering and Design,
282 0 126-143.
Giorla, A., Vaitová, M., Le Pape, Y. and Štemberk, P.,
(2015). “Meso-scale modeling of irradiated concrete
in test reactor.” Nuclear Engineering and Design ,295,
59-73.
Harlow, G. E., (1982). “The anorthoclase structures; the
effects of temperature and composition.” American
Mineralogist , 67(9-10), 975-996.
Hilsdorf, H., Kropp, J. and Koch, H., (1978). “The
effects of nuclear radiation on the mechanical
properties of concrete.” ACI SP-55, 223-251.
Hobbs, L. and Pascucci, M., (1980). “Radiolysis and
defect structure in electron-irradiated α-quartz.” Le
I. Maruyama and S. Muto / Journal of Advanced Concrete Technology Vol. 14, 706-716, 2016
Journal de Physique Colloques, 41, C6
C6-237-C236-242.
Hobbs, L. W., (1995). “The role of topology and
geometry in the irradiation-induced amorphization of
network structures.” Journal of Non-Crystalline Solids,
182(1), 27-39.
Hobbs, L. W., Clinard Jr, F. W., Zinkle, S. J. and Ewing,
R. C., (1994). “Radiation effects in ceramics.” Journal
of Nuclear Materials, 216(C), 291-321.
Inui, H., Mori, H., Sakata, T. and Fujita, H., (1990).
“Electron irradiation induced
crystalline-to-amorphous transition in quartz single
crystals.” Journal of Non-Crystalline Solids, 116(1),
1-15.
Le Pape, Y., Field, K. G. and Remec, I., (2015).
“Radiation effects in concrete for nuclear power plants,
Part II: Perspective from micromechanical modeling.”
Nuclear Engineering and Design, 282, 144-157.
Le Pape, Y., Giorla, A. and Sanahuja, J., (2016).
“Combined Effects of Temperature and Irradiation on
Concrete Damage.” Journal of Advanced Concrete
Technology, 14(3), 70-86.
Lell, E., Kreidl, N. J. and Hensler, J. R., (1968).
“Radiation effects in quartz, silica, and glasses.” In:
Progress in Ceramic Science. Volume 4. Burke, JE
(ed.). New York, Pergamon Press, Inc., 1-93.
Levien, L., Prewit, C. T. and Weidner, D. J., (1980).
“Structure and elastic properties of quartz at pressure.”
American Mineralogist, 65(9-10), 920-930.
Maruyama, I., Haba, K., Sato, O., Ishikawa, S., Kontani,
O. and Takizawa, M., (2016). “A numerical model for
concrete strength change under neutron and
gamma-ray irradiation.” Journal of Advanced
Concrete Technology 14(4), 144-162.
Maruyama, I., Kontani, O., Ishizawa, A., Takizawa, M.
and Sato, O., (2012). “Development of system for
evaluating concrete strength deterioration due to
radiation and resultant heat.” Proceedings of 3rd
International Conference on NPP Life Management
for Long Term Operations, IAEA-CN-194-096, Salt
Lake City, May.
Maruyama, I., Kontani, O., Sawada, S., Sato, O., Igarashi,
G. and Takizawa, M., (2013). “Evaluation of
irradiation effects on concrete structure-background
and preparation of neutron irradiation test.” American
Society of Mechanical Engineers, Power Division
(Publication) POWER.
Momma, K. and Izumi, F., (2011). “VESTA 3 for
three-dimensional visualization of crystal, volumetric
and morphology data.” J. Appl. Crystallogr., 44,
1272-1276.
Nakano, S., Muto, S. and Tanabe, T., (2005). “Change in
714
mechanical properties of ion-irradiated ceramics
studied by nanoindentation method.” Journal of the
Japan Institute of Metals and Materials 69(9),
815-824.
Primak, W., (1958). “Fast-neutron-induced changes in
quartz and vitreous silica.” Physical Review, 110(6),
1240-1254.
Primak, W., Fuchs, L. and Day P., (1955). “Effects of
nuclear reactor exposure on some properties of
vitreous silica and quartz.” Journal of the American
Ceramic Society, 38(4), 135-139.
Raether, H., (1982). “Excitation of plasmons and
interband transitions by electrons.” Springer Tracts in
Modern Physics, 88, Springer-Verlag. Berlin.
Richardson, S. M. and Richardson, J. W., (1982).
“Crystal structure of a pink muscovite from Archer's
Post, Kenya; implications for reverse pleochroism in
dioctahedral micas.” American Mineralogist, 67(1-2),
69-75.
Rosseel, T. M., Maruyama, I., Le Pape, Y., Kontani, O.,
Giorla, A. B., Remec, I. Wall, J. J., Sircar, M., Andrade,
C. and Ordonez, M., (2016). “Review of the current
state of knowledge on the effects of radiation on
concrete.” Journal of Advanced Concrete Technology,
14(7), 368-383.
Seeberger, J. and Hilsdorf, H., (1982). “Einfluß von
radioactiver Strahlung auf die Festogkeit and Struktur
von Beton.” Technical Report NR 2505, Institut für
Massivbau and Baustofftechnologie, Universität
Karlsruhe.
Takeda, H. and Ross, M., (1975). “Mica polytypism -
Dissimilarities in crystal-structures of coexisting 1M
and 2M Biotite.” American Mineralogist, 60(11-1),
1030-1040.
Ugurlu, O., Haus, J., Gunawan, A. A., Thomas, M. G.,
Maheshwari, S., Tsapatsis,M. and Mkhoyan, K. A.,
(2011). “Radiolysis to knock-on damage transition in
zeolites under electron beam irradiation.” Physical
Review, B 83(11), 113408.
Viswanathan, K. and Kielhorn, H.-M., (1983). “Al, Si
distribution in a ternary (Ba, K, Na)-feldspar as
determined by crystal structure refinement.” American
Mineralogist, 68(1-2), 122-124.
Wenk, H. R., Joswig, W., Tagai, T., Korekawa, M. and
Smith, B. K., (1980). “The average structure of An
62-66 labradorite.” American Mineralogist 65(1-2),
81-95.
Zanazzi, P. F., Montagnoli, M. Nazzareni S. and Comodi
P., (2007). “Structural effects of pressure on
monoclinic chlorite: A single-crystal study.” American
Mineralogist, 92(4), 655-661.
I. Maruyama and S. Muto / Journal of Advanced Concrete Technology Vol. 14, 706-716, 2016 715

Appendix
The crystal structure of used minerals are summarized
here. The crystal structure of used minerals are summa-
rized here. The figures were drawn with VISTA
(Momma and Izumi 2011).

(a) α-quartz (6-ring 3-dimensional) (Levien et al. 1980)

(b) Orthoclase (4-ring + 6-ring) (Viswanathan and Kielhorn (d) Albite (4 ring + 6 ring) (Harlow 1982)
1983)

(c) Anorthite (4 ring + 6 ring) (Wenk et al. 1980) (e) Biotite (Planate 6-ring) (Takeda and Ross 1975)
I. Maruyama and S. Muto / Journal of Advanced Concrete Technology Vol. 14, 706-716, 2016 716

(f) Muscovite (Planate 6-ring) (Richardson and Richard- (g) Chlorite (Planate 6-ring) (Zanazzi et al. 2007)
son 1982)
Fig. A. 1 Structure of each mineral. Blue tetrahedral represent the Si or Al atom.
Journal of Advanced Concrete Technology Vol. 14, 691-705, November 2016 / Copyright © 2016 Japan Concrete Institute 691

Invited paper

Multi-Scale Review for Possible Mechanisms of Natural Frequency

Change of Reinforced Concrete Structures under an Ordinary Drying
Condition
Ippei Maruyama1*

Received 14 March 2016, accepted 1 November 2016 doi:10.3151/jact.14.691

Abstract
Based on a review from a view point of multi-scale perspective, it is concluded that possible mechanisms of natural
frequency change of concrete structures under an ordinary drying condition are reduction of Young’s modulus of concrete
and stiffness reduction of reinforced concrete member due to shrinkage-induced cracking. Both two mechanisms can be
provoked by drying, which is natural environmental condition for all the building structures. This natural frequency
change is important for the integrity analysis of nuclear power plants because seismic response change and sympathetic
vibration between structure and installed implementations are expected. In addition, to understand the mechanism this
mechanism should contributed to post-earthquake integrity evaluation which is necessary to efficient and continuous
operation of nuclear power plants. The lost parts of scientific knowledge and needs of experimental data for whole picture
of this natural frequency change have been pointed out.

1. Introduction the natural frequency change is discussed through a view

Aging management of concrete structures becomes im-
portant issue in contexts of economy and environment. It
has been reported that the life of architectural reinforced
concrete building in Japan is about 50 years (Komatsu
2008), and this value is significantly smaller than that in
Europe or US (Komatsu 2000). This short life of dwell-
ings in Japan has large impact on environment through
material waste, natural resource consumption, and CO2
emission, and design and maintenance of long-life
dwellings are desired in Japan.
With regards to civil infrastructures in Japan, they
were originally designed without any expected life-span,
and they were mainly built from 60’s to 80’s, while cur-
rently many deteriorations have been reported
(Kobayashi 2000). About 30% of civil structures have
passed more than 40 years, for this reason, significance
of monitoring, maintenance, prediction, and integrity
evaluation is emphasized recently and based on these
background as well as the impacts of Great East Japan
Earthquake on Japanese society, new act was enacted
(Japanese government 2013).
Industry plants are also in the same way. They are
required the proper aging management. As for nuclear
power plants in Japan, in addition to the economical
benefit, they are also required to keep the integrity and
safety.
Recently, a new phenomenon of structural perform-
ance change of concrete structures has been spotlighted,
that is natural frequency. Here, possible mechanism of
1
Professor, Graduate School of Environmental Studies,
Nagoya University, Nagoya, Japan. *Corresponding
author, E-mail: ippei@dali.nuac.nagoya-u.ac.jp
of multi-scale characteristic of concrete. In addition to
the natural frequency change, impact of drying on con-
crete structure through concrete properties change, which
has not been cleared, is discussed for aging management
of concrete structure for nuclear power plants.
2. Facts of natural frequency of the
concrete structures.
Kashima and Kitagawa (2006) have reported a natural
frequency change of 8-story steel reinforced concrete
(SRC) building, by collecting the data by densely in-
strumented with 22 accelerometers during small and
medium-sized earthquakes. The data is analyzed by us-
ing Evolution Strategy to optimize analysis parameters
for a multi-story sway-rocking model. And 18% decrease
in natural frequency of building after 7 years after con-
struction was discovered. Li et al. (2014) also use the
data from the same system but the analysis is expanded to
14 years after construction, and 45% reduction of natural
frequency was found including the damage during the
Great East Japan Earthquake. It should be noted that all
the earthquakes except for the Great East Japan Earth-
quake shows no change of stiffness during the earth-
quakes.
Toyobe et al. (2013) have collected the data of accel-
erometers implemented in the buildings and they found
time-dependent behavior of natural frequencies of
buildings. After 8 years of construction, 20% decrease in
natural frequency was observed in case of 7-story SRC
building, while the reduced value was 33% in case of
reinforced concrete (RC) building. It has been also re-
ported that the 4-story RC structure after 40 years of
construction did not show any reduction of natural fre-
quency change. Therefore, this tendency can not be at-
I. Maruyama / Journal of Advanced Concrete Technology Vol. 14, 691-705, 2016
Table 1 Structure of the paper.
Chapter Section Theme
1 Introduction
2 - Facts of Natural frequency of building
3 3.1 - Young’s modulus and shrinkage of Cement paste
- Nature of C-S-H and shrinkage mechanism
3.2 - Aggregate shrinkage
3.3 - Young’s modulus, shrinkage, and creep of concrete.
- Shrinkage of mortar and fine aggregate roles.
- Role of ITZ on shrinkage of concrete
4 - Relationship the alteration of concrete under drying and structural performance of RC members.
5 Discussion for aging management
6 Summary
tributable to the soil-structure interaction problem solely
unlike the cases found in European buildings.
In case of nuclear power plant, Unit No.1 of Onagawa
Nuclear Power Plant has been analyzed since 1985, and
they found the continuous reduction of natural frequency
of the plant and it was 22% after 25 years of construction.
This means that even the RC members are thick enough,
it has a natural trend to decrease the stiffness of the
members (Nuclear and Industrial Safety Agency 2011;
Ogata et al. 2011). In addition to this phenomenon, there
were some cases that the natural frequency showed
sudden drop during or after the earthquakes but there is
some recovering mechanism and it turned back to the
natural decreasing trend. Similar phenomenon have been
reported by Saito (2012). In case of very large size of
earthquake, it is possible that this recovering mechanism
can not bring the frequency back to the natural trends
(Takahashi et al. 2012; Uebayashi and Nagano 2012).
Summary of natural frequency of concrete related
buildings are shown in Fig. 1.
Based on these previous researches, it is concluded as
follows; almost all the reinforced concrete structures
show decrease of natural frequency and this attributes to
the reduction of stiffness of concrete structure, because
there is little difference of mass of structure during the
monitoring in some cases. Consequently, there is a
common environmental condition which cause reduction
Fig. 1 Natural frequency change as a function of time, the relative natural frequency is the ratio of natural frequency to the
natural frequency of buildings just after construction.
692
of stiffness of reinforced concrete member, and two
mechanisms for reduction of stiffness of reinforced
concrete member are derived; reduction of Young’s
modulus of concrete and cracking in reinforced concrete
member. Now, the common environmental condition,
drying, is discussed to emphasize the importance of
scientific research for prediction of performance of
concrete structures to contribute to aging management of
concrete structures. The whole structure of the manu-
script is shown in Table 1.
3. Concrete components and concrete
3.1 Cement paste properties under the first
drying
Hardened cement paste (hcp) made from cement paste
with relatively high water cement ratio (W/C = ~0.55)
has notable colloidal nature. This nature is attributed to
calcium silicate hydrate (C-S-H) with low crystallinity
which is hydration product of alite and belite. There are
many alibi of the colloidal nature of C-S-H. For example,
BET surface from nitrogen sorption isotherm (Tomes et
al. 1957; Hunt et al. 1960; Parrott et al. 1980; Litvan and
Myers 1983), water vapor sorption isotherm (Tomes et al.
1957; Maruyama et al. 2014a), and surface area obtained
by SAXS (Winslow and Diamond 1974; Kropp et al.
1985; Völkl et al. 1987) are affected by long-term drying
I. Maruyama / Journal of Advanced Concrete Technology Vol. 14, 691-705, 2016 693

condition. A shape of water vapor sorption isotherms is modulus change of hardened cement paste under the first
also dramatically affected by the drying (Maruyama et al. desorption process and it showed concave curve in the
2014a). It has been reported that irreversible shrinkage relationship between relative humidity and Young’s
after the first drying is large when W/C of paste is large modulus, and the minimum value was observed around
(Helmuth and Turk 1967). A part of these phenomena is 30 ~ 40% RH. The variation was within 18% of the
considered to attribute to the low density (LD) C-S-H, Young’s modulus at the saturated state. This Young’s
which is quantitatively proposed by Tennis and Jennings modulus is obtained by compressive loading test.
(2000). A picture obtained by field emission (FE) scan- In our previous experiments (Horiguchi et al. 2011;
ning electron microscopy (SEM) of hydrated cement Maruyama et al. 2014a), similar trends are confirmed by
paste is shown in Fig. 2. In this picture, the low density using the cement paste made from white cement by
C-S-H is clearly confirmed in white Portland cement bending test and cement paste made from white cement
paste with W/C = 0.55 and it is produced outer area of and ordinary Portland cement by ultrasonic pulse veloc-
original boundary of cement particle. This LD C-S-H is ity measurement. The summary is shown in Fig. 4. In
almost similar definition of that of outer product firstly case of Young’s modulus determined by bending test and
addressed by Goto et al. (1976) ultrasonic pulse velocity measurement, they show 50%
This LD C-S-H is believed to shrink and be compacted and 25% variation under the first drying, and around 40 ~
by drying and the nitrogen accessible area is reduced. A 50% RH the minimum values were observed. It should
morphology change of LD C-S-H is observed by envi- be mentioned that there is a discrepancy in the range
ronmental SEM (Fonseca and Jennings 2010) and this is from 100% to 80% RH between bending test data shown
also reproduced by FE-SEM as they were shown in Fig. in Fig. 4 and ultra-sonic measurement results or Witt-
3. In wet condition, LD C-S-H has indeterminate mor-
phology while it has fibrillar shape in the sample slowly
dried under 40% RH for 1.5 years. These observations
clearly show the dynamic microstructural change in
hardened cement paste under the first drying process.
And thus, cement paste has a colloidal nature (Jesser
1927; Tomes et al. 1957; Jennings 2000, 2008).
Based on these facts, modulus of elasticity and
shrinkage strain of hardened cement paste should cause
the change of Young’s modulus and volume of concrete.
Sereda et al. (1966) substantially studied the relation
between relative humidity and Young’s modulus of hcp,
but they focused mainly on the Young’s modulus change
after strongly dried condition and avoided to discuss
about the unstable calcium silicate hydrate under the first
desorption process because their objective is to clarify
the role of well-stabilized calcium silicate hydrate in Fig. 2 FE-SEM observation of LD C-S-H and HD C-S-H
cement paste. Wittmann (1973) showed the Young’s in white Portland cement paste with W/C=0.55.

Fig. 3 FE-SEM observation of morphology change of LD C-S-H in white Portland cement paste with W/C=0.55. In case of
lime-water cured sample with rapid drying for pre-drying showed indeterminate morphology of C-S-H on the crushed
surface (left). Fibrillar LD C-S-H shown on the surface of cracked surface of matured white Portland cement paste dried
after 40% RH for 1.5 year (right). The sample detail is shown in references (Maruyama et al. 2014a, Maruyama et al.
2015), and they are SDS100 and SDS40 respectively. These SEM data are provided by Drs. Watanabe and Matsui (Asahi
Kasei Co.).
I. Maruyama / Journal of Advanced Concrete Technology Vol. 14, 691-705, 2016 694

(a) Young’s modulus by bending test (b) Young’s modulus by ultrasonic pulse velocity
Fig. 4 Young’s modulus of hardened cement paste as a function of relative humidity for long-term drying affected by ce-
ment type, water to cement ratio, and loading type. W: white cement, N: ordinary Portland cement, XX: water to cement
ratio of cement paste. All the cements are categorized as Type-I. Data are from Horiguchi et al. (2011) and Maruyama et
al. (2014a).

mann’s results. The results of compressive test or ul-

tra-sonic measurement test might be reflected by the load
bearing water molecules in the cement paste while they
will not bear the stress in tension case. This is a possible
explanation for the discrepancy.
With regards to the Poisson’s ratio, it has decreasing
trend as a function of relative humidity. ~ 0.23 is the
value under dried state, while it is about 0.26 ~0.28 in
saturated state as shown in Fig. 5.
These trend might be explained by the change in ag-
glomeration of C-S-H and atomic crystal structure of
C-S-H, which are still under discussing. Small angle
scattering profiles (Chiang et al. 2012, 2013; Nicoleau et
al. 2013; Trapote-Barreira et al. 2015) and 1H-NMR
relaxometry (Korb 2009; McDonald et al. 2010; Muller
et al. 2012, 2013, 2015; Valori et al. 2013) give insights
on this topic. An example of elaboration to explain the
relationships between C-S-H microstructure change
under the first drying and physical properties is shown in
the references (Maruyama et al. 2014a, 2015, 2016a).
Drying shrinkage of cement paste might have large
impact on concrete property, because large shrinkage
difference is expected in concrete and cracks in cement
paste around aggregates are going to be produced under
drying (Pickett 1956; Hansen and Nielsen 1965; Hobbs
1974; Grassl et al. 2010; Idiart et al. 2011; Lagier et al.
2011; Zhang et al. 2013; Maruyama and Sasano 2014),
while the mechanism of shrinkage of hcp is not clarified
(Powers 1965; Helmuth and Turk 1967; Feldman and
Sereda 1968; Wittmann 1968; Bažant 1972; Beltzung
and Wittmann 2005; Setzer 2007; Wittmann 2008;
Beaudoin et al. 2010; Maruyama 2010). The most lost
part is irreversible shrinkage of hcp under the first de-
sorption process. In author's previous study (Maruyama
2010), shrinkage of hcp is well-predicted by the statis-
tical thickness of water vapor adsorption based on a
disjoining (or hydration) pressure theory. This indicates
Fig. 5 Poisson’s ratio of hardened cement paste as a
function of relative humidity for long-term drying. The
notation followed those shown in Fig.4. The data are from
Horiguchi et al. (2011) and Maruyama et al. (2014a).
that a number of water molecules adsorbed on the C-S-H
is the key parameter for drying shrinkage of hcp includ-
ing irreversible shrinkage. In Fig. 6 drying shrinkage (a)
and water content (b) of matured hcp 3 mm-thick
specimens dried slowly for 0.5 year are shown including
re-humidifying process (Maruyama 2010). There is a
clear hysteresis above 40% RH in both length-change
and long-term water vapor sorption isotherms. In addi-
tion to these data, water vapor BET area (SH2O) of hcp
samples was obtained by sorption isotherms by volu-
metric method. Then the relationship between incre-
mental statistical thickness of adsorption (origin is satu-
rated condition) and drying shrinkage strain above 40%
RH is plotted in Fig. 6 (c). As it is shown, clear linear
relation is obtained disregards to water to cement ratio.
The range of 100% to 40% RH is corresponding to the
range where the irreversible shrinkage is observed
(Helmuth and Turk 1967; Maruyama et al. 2015), and it
I. Maruyama / Journal of Advanced Concrete Technology Vol. 14, 691-705, 2016
was suggested that another shrinkage mechanism is ac-
tive rather than surface energy shrinkage mechanism in
this range (Maruyama et al. 2015), and concerning this
another shrinkage mechanism, statistical thickness of
adsorption is the key parameter.
Now, the important issue for long-term properties
change of hcp is laid in the first drying process, and it is
related to colloidal behavior and atomic-scale structural
change of C-S-H. All the phenomena related water con-
tent change under drying can be interpreted different
ways as shown in Fig. 7. In the previous models, C-S-H
is considered to have the fixed interlayer distance and
water sorption process in the interlayer space is saturated
under 11 ~ 30% RH. This assumption is based on an
interpretation by Brunauer et al. (1967) from water vapor
sorption isotherms. However, it is possible that distance
of C-S-H mono layers and number of adsorbed water in
this space depend on the drying process. Previous papers
(Maruyama et al. 2015, 2016a) proposed a hypothesis
that there is a movable C–S–H monolayer and the dis-
tance between the C–S–H monolayers is dynamically
changed by drying and re-humidifying. Because there is
no structural limitation in maintaining the distance be-
tween C–S–H monolayers, such as Si-Q3 sites which
should be a role of pillar to fix the distance between CaO
(a) Shrinkage strain
(c) Relationship between incremental statistical thickness of adsorption and drying shrinkage
Fig. 6 (a) Shrinkage strain of ordinary Portland cement paste with water to cement ratio of 0.55 (N55), 0.40 (N40), and
0.30 (N30). (b) Water content in hcp for N55, N40, and N30. (c) Relationship between incremental statistical thickness of
adsorption and drying shrinkage above 40% RH.
695
layers, and randomly attached silicate dimers on the CaO
layer should inflect the structure, crystal growth in the
002 direction is difficult (Gartner 1997). In addition,
water sorption process should be initiated in the space
surrounded by C-S-H monolayers, because surface po-
tentials from each C–S–H monolayer are overlapped.
These structural characteristics must change the distance
between C–S–H monolayers during the sorption process
like montmorillonites (Cases et al. 1992). In other words,
gel water is re-interpreted as interlayer water in movable
C-S-H monolayers.
Complete picture of C-S-H contributes the time de-
pendent properties change of concrete structures under
drying, further investigations are required for this pur-
pose.
3.2 Aggregate behavior under the first drying
Aggregate generally occupied about 60~70% of volume
of normal strength concrete. Therefore, the property of
aggregate has a large impact on mechanical performance
of concrete. Generally, the chemical and physical sta-
bilities for required property as a concrete aggregate are
evaluated by national or international standard before
concrete mixing, however, still many variations are pos-
sible. One of them the shrinkage performance of aggre-
(b) Water content
I. Maruyama / Journal of Advanced Concrete Technology Vol. 14, 691-705, 2016
gate is not fully checked. Drying shrinkage of concretes
which were made by Japanese design for mixture pro-
portion showed high correlation of shrinkage of coarse
aggregate (Fujiwara 2008). The example is shown in Fig.
8 made from the data by Horiguchi et al. (2011) and
Teranishi et al. (2011). From this aspect, the inherent
shrinkage property of aggregate is significant. In our
previous study, shrinkage of sandstone aggregates
available in Japan is found to be a function of amount of
chlorite which was formed in diagenesis (Igarashi et al.
2015), as it is shown in Fig. 9. This is because chlorite
forms in matrix located between quartz or feldspar grains
and the clay minerals can shrink without large restraint
by the grains. The relationship between mineral compo-
sition of aggregate and physical properties are important
issue for aging management of concrete.
Particle size and its distribution, and interfacial zone
Water content
Water
in gel pore
Water in stable C‐S‐H
Relative humidity
(a) F & S model
Water content
Water in
space between
C‐S‐H monolayers
Relative humidity
(b) Proposed model
Fig. 7 Comparison of (a) Feldman and Sereda microstructure model (Feldman and Sereda 1968) and (b) Proposed model
in which water is trapped between calcium silicate mono layers and their distance and number of trapped water molecules
are dramatically changed under the first desorption process (Maruyama et al. 2015).
Fig. 8 Relationship between shrinkage of coarse aggre-
gate and that of concrete (Horiguchi et al. 2011; Teranishi
et al. 2011).
696
will be treated in the next section through restraint per-
formance for shrinkage of mortar or cement paste.
3.3 Young’s modulus, shrinkage, and creep of
concrete under the first drying
Shrinkage strain of matured mortar is more than 5 times
as large as that of aggregate, and Young’s modulus of
mortar is less than quarter of that of aggregate (Maru-
yama et al 2014b). Consequently, large tensile stress is
produced in mortar of concrete under the first drying
process brought by the restraint roll of coarse aggregate
and cracks can be formed around coarse aggregate
(Carlson 1938; Hansen and Nielsen 1965; Hobbs 1974;
Bisschop and van Mier 2008; Idiart et al. 2011;
Maruyama and Sasano 2014). The width of cracks
formed around aggregate becomes wider as mortar
shrinks. Therefore, the crack opening volume accumu-
Calcium silicate
mono layer
Water molecule
in a fixed C‐S‐H layer
Water molecule
in gel pore
Wet Dry
Fig. 9 Relationship between shrinkage strain of sand-
stone aggregate and amount of chlorite (Igarashi et al.
2015).
I. Maruyama / Journal of Advanced Concrete Technology Vol. 14, 691-705, 2016
lated in concrete is a function of the difference of
shrinkage strain between mortar and coarse aggregate.
The increase in voids in concrete reduces the Young’s
modulus of concrete because load bearing area is reduced
(Maruyama et al. 2014b) (see Fig.10). Resultantly, dif-
ference of shrinkage between concrete and mortar rep-
resent the volume of accumulated spaces by opening of
cracks around aggregate. Therefore, the reduction ratio
of Young’s modulus of concrete after drying has linear
relation with the difference of shrinkage between con-
crete and mortar as it is shown in Fig. 11. With this re-
gards, shrinkage of aggregate is related to the reduction
of Young’s modulus through restraint role to the mortar
shrinkage. Shrinking aggregate has a smaller restraint
Fig. 10 Ratio of Young’s modulus of concrete / mortar /
paste after drying to the Young’s modulus of concrete at
sealed condition (E/Eo). Each specimen reached an
equilibrium in drying environment after 0.5 ~ 1.0 year.
Limestone aggregate was very pure and has almost no
shrinkage property, while sandstone had large amount of
chlorite and shrunk much. Mortar has the same type and
volume of fine aggregate as that of concretes. Concrete
and mortar was made by high early strength Portland
cement and paste was made from whilte cement. Details
are shown in Maruyama et al. (2014a) and Maruyama et
al. (2014b).
Fig.12 Shrinkage of paste and mortar with different fine aggregate volume and sizes (Sakaida et al. 2014).
697
role. As it is seen in Fig. 10, concrete with shrinking
sandstone showed less reduction of Young’s modulus
after drying.
Consequently, Young’s modulus change of concrete is
tightly related to the shrinkage of concrete and mortar.
There are plenty of researches on relationship between
concrete shrinkage and mixture proportion, therefore,
some important notes are addressed here.
Relationship between the shrinkage of mortar and the
shrinkage of cement paste is not fully understood. Be-
cause the shrinkage of mortar is only a function of
volume of aggregate (Fig. 12) and not the size of ag-
gregate (Thierry et al. 2008; Sakaida et al. 2014) while,
the size of coarse aggregate has a large impact on
shrinkage of concrete (Carlson 1938; Maruyama and
Sasano 2014). This phenomenon can be partially ex-
plained by a role of interfacial transition zone (ITZ)
formed between cement paste matrix and aggregate
(Neubauer et al. 1996; Maruyama and Sugie 2014) but
the restraint role of small grains on the shrinkage of paste
remains to be solved.
ITZ is a one of the key factor of concrete property. It is
Fig.11 Ratio of Young’s modulus of concrete affected by
drying to that of concrete under sealed condition
(Ec/Eco) as a function of difference of shrinkage between
mortar and concrete.
I. Maruyama / Journal of Advanced Concrete Technology Vol. 14, 691-705, 2016
possible to explain the shrinkage of concrete affected by
the aggregate size by using the presence of ITZ
(Maruyama and Sugie 2014). The large size of aggregate
reduces the shrinkage of concrete. ITZ is generally con-
sidered to be produced by the “wall effect” of the cement
particle packing process on the surface of the aggregate
(Scrivener et al. 2004). And the weak property on the
surface of aggregate is believed by the facts of porosity
distribution findings (Breton et al. 1993; Scrivener et al.
2004; Cwirzen and Penttala 2005; Herve et al. 2010), an
SEM analysis (Scrivener and Gartner 1988; Scrivener et
al. 2004), and physical properties testing (Xie et al. 1991;
Ping and Beaudoin 1992; Mitsui et al. 1994; Ollivier et al.
1995; Hearing 1997). As the porosity in the ITZ is more
than three times that of normal mortar located far from
the surface of the aggregate (Scrivener and Gartner 1988;
Scrivener et al. 2004), resistivity to shrinkage of mortar
matrix is low. Therefore, even there is a report that the
thickness of ITZ depends on the size of aggregates (Xie
et al. 1991; Ollivier et al. 1995; Tasong et al. 1999; Rao
and Raghu Prasad 2004), the resistivity of aggregate to
the shrinkage of mortar matrix can be evaluated by the
ratio of thickness of ITZ to the radius of aggregate. The
larger size of aggregate has a large restraint role to the
mortar shrinkage, while the smaller size of aggregate has
a small restraint role. And this large restraint role is un-
deniably related to the reduction of Young’s modulus.
Reduction of Young’s modulus also contributes to the
increase of creep under the drying. Larger creep coeffi-
cient under the drying environment than that of concrete
without any moisture exchange with environment is
known as Pickett effect (Pickett 1942). If creep is in-
herent nature of cement paste, creep strain at the cement
paste in concrete should be definable. Therefore, this
phenomenon should be explained by the structural nature.
One is the structural condition of mortar and coarse ag-
gregate. Shrinkage strain of cement paste is larger than
the average creep strain of concrete specimen. This leads
that cracks are formed around aggregate and reduction of
Young’s modulus of dried part of concrete is observed
even under the sustained load. This reduction of Young’s
modulus increase the elastic strain under the drying
condition as it is shown in Fig. 13. This is different from
Fig.14 Creep coefficient data of concrete in wet, 50% RH, and 70% RH conditions (left) reported by Raymond and Harmer
(1931) and calculated components of creep coefficients (right).
698
the concept proposed by Neville and Meyers (1964). In
case of Young concrete and concrete block with a large
section, the Neville’s model is applicable, but in matured
stage, the impact of increase in Young’s modulus due to
additional hydration after demolding can be negligible
and drying impact may overcome it. To confirm this, the
components of creep coefficient is evaluated by using the
data reported by Raymond and Harmer (1931) in Fig. 14.
In Fig. 14 right, reduction of Young’s modulus, which is
based on the Fig. 10 is considered for one of the creep
coefficient component. It should be noted that this as-
sumption is applicable only to the condition under the
equilibrium, and this is due to re-distribution of stress in
concrete under equilibrium condition. Schematic stress
re-distribution in concrete is shown in Fig. 15(a). The
residual components is caused by stress concentration
mechanism after drying. One is based on material char-
acters. After formation of cracks around aggregate, stress
bearing passes are limited and resultantly, the stress per
unit area of cement paste in concrete is increased and
creep behavior is enhanced due to this stress localization.
The other is observed in member size. After drying,
surface part of concrete exhibit larger shrinkage than
inner concrete. Therefore, the stress is re-distributed and
localized in central part of concrete (shown in Fig. 15(b)).
The peripheral concrete part shows shrinkage rather than
creep, therefore, the creep behavior should be governed
by central part concrete. This consequence indicates that
measurement procedure of creep deformation, especially
in position, needs special care and drying creep is
structural performance and not material performance.
Fig. 13 Schematic of creep coefficient components of
matured concrete proposed by Neville (A) and proposed
concept (B).
I. Maruyama / Journal of Advanced Concrete Technology Vol. 14, 691-705, 2016 699

4. Impact of volume and Young’s modulus 20 degree Celsius and 60% relative humidity was con-
change of concrete on RC member ducted (Maruyama and Igarashi 2015) and then rela-
performance tionship between water content and concrete shrinkage
(Maruyama et al. 2014b) is used for input of shrink-
Shrinkage of concrete induces stiffness of reinforced
concrete structure (Okada et al. 1986), and it decrease in
a flexural cracking moment and increase in curvature and
maximum crack width of beam (Tanimura et al. 2007;
Maruyama and Teshigawara 2010). This is explained by
tensile stress of concrete induced by shrinkage restraint
and strain difference before and after cracking of con-
crete. From this regards, creep strain of concrete in RC
column under axial load has the same role as that of
shrinkage. Creep strain and shrinkage strain in column (a) Stress re-distribution in local concrete after drying
make re-distribution of stress among concrete and rebars,
and in case of un-loading process such as seismic re-
sponse, crack is easily produced in tension (Komuro et al.
2008). This re-distribution of stress caused by shrinkage
also reduces the moment capacity at the failure of cover
concrete (Collins et al. 1993). Releasing of compressed
stress of rebar due to concrete shrinkage restraint causes
the additional bending movement and crack opening. In
addition, deflection at the moment capacity at failure of (b) Stress re-distribution in concrete member after drying
concrete is increased by creep strain based on our pre- Fig. 15 Schematic of stress re-distribution in concrete
liminary experiment for impact of sustained load on total and concrete member after drying.
stress-strain curve of concrete shown in Fig.16.
A diagonal cracking moment of RC beam as well as a
shear capacity of RC beam is also affected by shrinkage
(Maruta and Yamazaki 1990; Katayose et al. 2006; Sato
and Kawakane 2008; Gebreyouhannes and Maekawa
2011). This is simply because the stress, position, and
direction of diagonal cracking are altered by shrink-
age-induced stress and enlarged crack width. The wider
diagonal cracks also affect the performance of RC col-
umn, the loading capacity in cyclic loading process after
opening of diagonal cracking is decreased by this width
of diagonal cracks (Maruyama et al. 2011).
Decreasing of Young’s modulus of concrete leads re-
duction of stiffness of RC member. In addition, shrinkage
induced cracking has a large impact on stiffness of RC
member. It should be noted that the both two impacts are Fig. 16 Impact of sustained load on total stress-strain
brought by the shrinkage of cement paste in concrete curve of concrete. (Concrete specimens were the same
mainly. One is by way of cracks around aggregate and as G3 specimens in Maruyama et al. (2014b) with more
they are almost invisible, and the other is by way of than 3-year-curing period. W/C was 0.55.)
visible cracks which are localized cracks due to restraint
of concrete by rebars or outer structure. This difference is
induced by aggregate and its ITZ performance. Concrete
containing (pure) limestone coarse aggregate is known to
reduce the shrinkage-induced visible cracking (Mitani et
al. 2016), but this concrete showed relatively large de-
crease in Young’s modulus of concrete under drying (as it
is shown in Fig. 10). This is the trade-off relationship
(Maruyama et al. 2016b). To compare the impacts of
decrease in Young’s modulus and shrinkage-induced
cracking, simplified numerical analysis has been con-
ducted (Sugie et al. 2014). The calculation was made by
2-dimensional finite element method (FEM). The target
RC wall is shown in Fig. 17. In this moisture transfer at
Fig.17 Target concrete wall structure for FE-analysis.
I. Maruyama / Journal of Advanced Concrete Technology Vol. 14, 691-705, 2016
age-induced stress calculation in concrete structure.
Shrinkage strain is considered as equivalent nodal force
to this system. In addition to this, Young’s modulus
change of concrete is considered by using the relation-
ship between water content and Young’s modulus of
concrete (Maruyama et al. 2014b). In FE-analysis, cracks
are considered as smeared model. The results of structure
made from concrete containing limestone are shown in
Fig. 18(b). The reduction of stiffness (Fig.18(a)) as well
as typical calculated cracks in concrete wall (Fig. 18(b))
is shown. Behavior of structure made from concrete
containing limestone is compared with that from con-
crete containing shrinking sandstone concrete and the
results are shown in Fig. 18(b).
In Fig. 19, the ratio of natural frequency to that
without drying is estimated under the condition that there
is no mass change, and the primary mode is intended.
Figure 19 indicated that the shrinkage-induced cracking
has larger impact than that of decrease in Young’s
modulus due to drying. Based on the comparison with a
record of Onagawa NPP-3, the reduction trend obtained
by FE-analysis seems comparable to that of real record
and it suggests that the reason of stiffness reduction of
real concrete structure is caused by drying environment.
5. Discussion for aging management
Based on the review, it is highly possible that the natural
frequency change is caused by reduction of Young’s
modulus and shrinkage-induced cracking in reinforced
concrete member due to drying, in addition to the dam-
age during earthquake. This natural frequency change
may have impact of seismic load of the member in NPP
as well as sympathetic vibration between building and
instruments, such as piping system, cable support system,
turbine, and other implementations installed in NPP
buildings. Even the wide range of target frequency is
considered in the design state, it is important to consider
(a)
Fig. 18 Calculation results of relationship between load and deformation of concrete wall structure after drying (a) and
typical calculated crack patterns in concrete wall (b).
700
the risk of base line trend of natural frequency. In case of
high-rise building or structure, secondary or other modes
should be considered.
The base line trend is also crucial to evaluate the
damage during the earthquake. Although many moni-
toring sensors, such as accelerometer and velocimeter,
are installed to analyze the performance curve (dis-
placement versus load) of the structure, damage due to
seismic response can not be identified without baseline
trend due to drying. These sensors and obtained analysis
results should contribute to post earthquake performance
evaluation of the structure for efficient operation of NPP.
To develop the automated system for post earthquake
integrity analysis, the mechanism of natural frequency
change and its prediction are necessary.
It should be noted that the strength and Young’s
modulus of concrete has no longer strong correlation
after matured state (Maruyama et al. 2014b), because
drying condition changes strength and Young’s modulus
of concrete by different mechanisms. Direct evaluation
of Young’s modulus of concrete is desired, but using the
Fig. 19 Estimated natural frequency change from stiff-
ness reduction results of FEM compared with record of
Onagawa NPP-3.
(b)
I. Maruyama / Journal of Advanced Concrete Technology Vol. 14, 691-705, 2016
data obtained from core-sample needs special care, be-
cause massive concrete member has a distribution of
water content and resultant Young’s modulus. In this
sense, numerical evaluation might have potential to
evaluate the average concrete stiffness.
The most import knowledge is that the reduction ratio
of natural frequency after equilibrium and the rate of
reduction of natural frequency to achieve the equilibrium.
Because aging management is based on both integrity at
the present state and in future until the target life span,
those two parameters are crucial. Regarding the reduc-
tion ratio, it is possible to estimate the value by using
drying shrinkage strain and Young’s modulus of concrete
under the given environment which is equivalent to that
in nuclear power plant. But the rate of reduction of
natural frequency needs more scientific background of
concrete under drying process. The first drying process
in concrete is always accompanied with colloidal altera-
tion of C-S-H, and this process is not considered in ex-
isting research and they always assumes that concrete is
just a porous material, resultantly, the drying process
under high temperature or temperature gradient needs
more data (Shiire and Cheong 1988; England and
Khoylou 1995; Lien and Wittmann 1998).
6. Summary
It is concluded that possible mechanisms of natural fre-
quency change of concrete structures under an ordinary
drying condition are reduction of Young’s modulus of
concrete and stiffness reduction of reinforced concrete
member due to shrinkage-induced cracking. Both two
mechanisms can be provoked by drying, which is natural
environmental condition for all the building structures.
While the Young’s modulus change of concrete is ne-
glected in the present design codes, it is possible that it
shows more than 20% reduction in general drying con-
dition. In addition, drying condition changes strength and
Young’s modulus differently, it should be kept in mind
that there is no longer correlation between strength and
Young’s modulus of concrete in matured state, unlike
young concrete. From this point of view, direct evalua-
tion of Young’s modulus of concrete, but not from a
calculation based on compressive strength and relation-
ship between strength and Young’s modulus, is desired.
Concrete, however, has a distribution in its properties in
the reinforced concrete member under drying condition,
therefore, data treatment of core-sample should take into
account this nature of concrete member.
The natural frequency change is important for the in-
tegrity analysis of nuclear power plants because seismic
response change and sympathetic vibration between
structure and installed implementations are expected.
Second mode of structure is a potential risk for sympa-
thetic vibration of installed implementations. To under-
stand the mechanism should contributed to
post-earthquake integrity evaluation which is necessary
to efficient and continuous operation of nuclear power
701
plants by identifying the damage due to seismic response
from baseline trend of natural frequency change as well
as reduction of stiffness of structure.
The lost parts of scientific knowledge and desired
experimental data to obtain the whole picture of the
natural frequency change under ordinary drying condi-
tion have been pointed out.
Acknowledgement
The present research has been partly sponsored by JSPS
KAKENHI Grant Number 15H04077.
References
Appa Rao, G. and Raghu Prasad, B. K., (2004). “Influ-
ence of type of aggregate and surface roughness on the
interface fracture properties.” Materials and Struc-
tures, 37(5), 328-334.
Bažant, Z. P., (1972). “Thermodynamics of hindered
adsorption and its implications for hardened cement
paste and concrete.” Cement and Concrete Research, 2,
1-16.
Beaudoin, J. J., Raki, L., Alizadeh, R. and Mitchell, L.,
(2010). “Dimensional change and elastic behavior of
layered silicates and portland cement paste.” Cement
and Concrete Composites, 32, 25-33.
Beltzung, F. and Wittmann, F. H., (2005). “Role of
disjoining pressure in cement based materials.”
Cement and Concrete Research, 35, 2364-2370.
Bisschop, J. and van Mier, J. G. M., (2008). “Effect of
aggregates and microcracks on the drying rate of
cementitious composites.” Cement and Concrete
Research, 38, 1190-1196.
Breton, D., Carles-Gibergues, A., Ballivy, G. and
Grandet, J., (1993). “Contribution to the formation
mechanism of the transition zone between
rock-cement paste.” Cement and Concrete Research,
23, 335-346.
Brunauer, S., Mikhail, R. S. and Bodor, E. E., (1967).
“Some remarks about capillary condensation and pore
structure analysis.” Journal of Colloid and Interface
Science, 25(3), 353-358.
Carlson, R. W., (1938). “Drying shrinkage of concrete as
affected by many factors.” American Society Testing &
Materials Proceedings, 38(2), 419-437.
Cases, J. M., Berend, I., Besson, G., Francois, M., Uriot, J.
P., Thomas, F. and Poirier, J. E., (1992). “Mechanism
of adsorption and desorption of water vapor by
homoionic montmorillonite. I. The sodium-exchanged
form.” Langmuir, 8(11), 2730-2739.
Chiang, W.-S., Fratini, E., Baglioni, P., Liu, D. and Chen,
S.-H., (2012). “Microstructure determination of
calcium-silicate-hydrate globules by small-angle
neutron scattering.” The Journal of Physical
Chemistry C, 116( 8), 5055-5061.
Chiang, W.-S., Fratini, E., Ridi, F., Lim, S.-H., Yeh, Y.-Q.,
Baglioni, P., Choi, S.-M., Jeng, U. S. and Chen, S.-H.,
(2013). “Microstructural changes of globules in
calcium–silicate–hydrate gels with and without
I. Maruyama / Journal of Advanced Concrete Technology Vol. 14, 691-705, 2016
additives determined by small-angle neutron and x-ray
scattering.” Journal of Colloid and Interface Science,
398, 67-73.
Collins, M. P., Mitchell, D. and Macgregor, J. G., (1993).
“Structural design considerations for high-strength
concrete.” Concrete International, 15(5), 27-34.
Cwirzen, A. and Penttala, V., (2005). “Aggre-
gate–cement paste transition zone properties affecting
the salt–frost damage of high-performance concretes.”
Cement and Concrete Research, 35, 671-679.
England, G. L. and Khoylou, N., (1995). “Moisture flow
in concrete under steady state non-uniform tempera-
ture states: Experimental observations and theoretical
modelling.” Nuclear Engineering and Design,
156(1–2), 83-107.
Feldman, R. F. and Sereda, P. J., (1968). “A model for
hydrated portland cement paste as deduced from
sorption-length change and mechanical properties.”
Matériaux et Construction, 1(6), 509-520.
Fonseca, P. C. and Jennings, H. M., (2010). “The effect
of drying on early-age morphology of C–S–H as
observed in environmental SEM.” Cement and
Concrete Research, 40, 1673-1680.
Fujiwara, T., (2008). “Effect of aggregate on drying
shrinkage of concrete.” Journal of Advanced Concrete
Technology, 6(1), 31-44.
Gartner, E. M., (1997). “A proposed mechanism for the
growth of C–S–H during the hydration of tricalcium
silicate.” Cement and Concrete Research, 27, 665-672.
Gebreyouhannes, E. and Maekawa, K., (2011). “Nu-
merical simulation on shear capacity and post-peak
ductility of reinforced high-strength concrete coupled
with autogenous shrinkage.” Journal of Advanced
Concrete Technology, 9(1), 73-88.
Goto, S., Daimon, M., Hosaka, G. and Kondo, R., (1976).
“Composition and morphology of hydrated tricalcium
silicate.” Journal of the American Ceramic Society,
59(7-8), 281-284.
Grassl, P., Wong, H. S. and Buenfeld, N. R., (2010).
“Influence of aggregate size and volume fraction on
shrinkage induced micro-cracking of concrete and
mortar.” Cement and Concrete Research, 40, 85-93.
Hansen, T. C. and Nielsen, K. E. C., (1965). “Influence of
aggregate properties on concrete shrinkage.” ACI
Journal, 62(7), 789-794.
Hearing, B., (1997). “Fracture behavior of
mortar-aggregate interfaces in concrete.” Thesis
(MSc), Massachusetts Institute of Technology.
Helmuth, R. A. and Turk, D. H., (1967). “The reversible
and irreversible drying shrinkage of hardened portland
cement and tricalcium silicate pastes.” Journal of the
PCA Research and Development Laboratories, 9(2), 8
- 21.
Herve, E., Care, S. and Seguin, J. P., (2010). “Influence
of the porosity gradient in cement paste matrix on the
mechanical behavior of mortar.” Cement and Concrete
Research, 40, 1060-1071.
Hobbs, D. W., (1974). “Influence of aggregate restraint
702
on the shrinkage of concrete.” ACI Journal, 71(9),
445-450.
Horiguchi, N., Igarashi, G. and Maruyama, I., (2011a).
“Fundamental study on volume change of aggregate
due to loss of evaprable water.” Proceedings of Japan
Concrete Institute, 33(1), 131-136. (in Japanese)
Horiguchi, N., Igarashi, G. and Maruyama, I., (2011b).
“Young's modulus and poisson's ratio of hardened
cement paste under different relative humidity.” J.
Struct. Constr. Eng. , 76(660), 231-236. (in Japanese)
Hunt, C. M., Tomes, L. A. and Blaine, R. L., (1960).
“Some effects of aging on the surface area of portland
cement paste.” Journal of Research of the National
Bureau of Standards, 64A(2), 163-169.
Idiart, A., López, C. and Carol, I., (2011). “Modeling of
drying shrinkage of concrete specimens at the
meso-level.” Materials and Structures, 44(2),
415-435.
Igarashi, G., Maruyama, I., Nishioka, Y. and Yoshida, H.,
(2015). “Influence of mineral composition of siliceous
rock on its volume change.” Construction and
Building Materials, 94, 701-709.
Japanese Government, (2013). “Basic act for national
resilience contributing to preventing and mitigating
disasters for developing resilience in the lives of the
citizenry.” Act No. 95 of 2013.
Jennings, H. M., (2000). “A model for the microstructure
of calcium silicate hydrate in cement paste.” Cement
and Concrete Research, 30, 101-116.
Jennings, H. M., (2008). “Refinements to colloid model
of C–S–H in cement: CM-II.” Cement and Concrete
Research, 38, 275-289.
Jesser, L., (1927). “Kolloide chemische reaktionen des
tonerdezementmörtels.” Zement, 16, 741.
Kashima, T. and Kitagawa, Y., (2006). “Dynamic
characteristics of a building estimated from strong
motion records using evolution strategy.” Journal of
structural and construction engineering, 602, 145-152.
(in Japanese)
Katayose, N., Takamori, N., Nishida, H. and Teraoka, M.,
(2006). “Mechanical properties and autogenous
shrinkage behavior of high strength concrete at early
age.” Proceedings of the Japan Concrete Institute,
28(1), 497-502. (in Japanese)
Kobayashi, K., (2000). “Konkuri-to ga abunai [concrete
crisis].” Tokyo, Japan, Iwanami Shoten. (in Japanese)
Komatsu, Y., (2000). “Life span of dwellings (Special
issue 1: Dwelling stock analysis).” Housing review,
16(2), 5-20. (in Japanese)
Komatsu, Y., (2008). “Life time estimations of Japanese
buildings and houses at the years of 1997 and 2005.”
Journal of architecture and planning, 73(632),
2197-2205. (in Japanese)
Komuro, T., Imai, K., Korenaga, T. and Watanabe, F.,
(2008). “Study on long term performance of
reinforced concrete columns using high strength
concrete.” Proceedings of the Japan Concrete Institute,
30(3), 223-228. (in Japanese)
I. Maruyama / Journal of Advanced Concrete Technology Vol. 14, 691-705, 2016
Korb, J. P. (2009). “NMR and nuclear spin relaxation of
cement and concrete materials.” Current Opinion in
Colloid & Interface Science, 14(3), 192-202.
Kropp, J., Gräfenecker, T. and Hilsdorf, H. K., (1985).
Characterization of the microstructure of hydrated
cement paste by small angle x-ray scattering. In:
Principles and Applications of Pore Structural
Characterization. J. W. Arrowsmith. Bristol, 83-96.
Lagier, F., Jourdain, X., De Sa, C., Benboudjema, F. and
Colliat, J. B., (2011). “Numerical strategies for
prediction of drying cracks in heterogeneous
materials: Comparison upon experimental results.”
Engineering Structures, 33(3), 920-931.
Li, L., Nakamura, A., Kashima, T. and Teshigawara, M.,
(2014). “Earthquake damage evaluation of an 8-story
steel-reinforced concrete building using Sa-Sd curves.”
J. Struct. Constr. Eng., AIJ, 79(702), 1107-1115.
Lien, H. P. and Wittmann, F. H., (1998). “Mass transfer in
inhomogeneous porous media under thermal
gradients.” Nuclear Engineering and Design, 179(2),
179-189.
Litvan, G. G. and Myers, R. E., (1983). “Surface area of
cement paste conditioned at various relative
humidities.” Cement and Concrete Research, 13,
49-60.
Maruta, M. and Yamazaki, M., (1990). “Study on the
shear behavior of RC beams subjected to hight
temperature for long time.” Proceedings of the Japan
Concrete Institute, 12(1), 603-608, (in Japanese).
Maruyama, I., (2010). “Origin of drying shrinkage of
hardened cement paste: Hydration pressure.” Journal
of Advanced Concrete Technology, 8(2), 187-200.
Maruyama, I. and Igarashi, G., (2015). “Numerical
approach towards aging management of concrete
structures: Material strength evaluation in a massive
concrete structure under one-sided heating.” Journal
of Advanced Concrete Technology, 13(11), 500-527.
Maruyama, I., Igarashi, G. and Nishioka, Y., (2015).
“Bimodal behavior of C–S–H interpreted from
short-term length change and water vapor sorption
isotherms of hardened cement paste.” Cement and
Concrete Research, 73, 0, 158-168.
Maruyama, I., Nishioka, Y., Igarashi, G. and Matsui, K.,
(2014a). “Microstructural and bulk property changes
in hardened cement paste during the first drying
process.” Cement and Concrete Research, 58, 20-34.
Maruyama, I., Sakamoto, N., Matsui, K. and Igarashi, G.,
(2016a). “Microstructural changes in white Portland
cement paste under the first drying process evaluated
by WAXS, SAXS, and USAXS.” Cement and Con-
crete Research, doi: 10.1016/j.cemconres.2016.10.002.
Maruyama, I. and Sasano, H., (2014). “Strain and crack
distribution in concrete during drying.” Materials and
Structures, 47(3), 517-532.
Maruyama, I., Sasano, H. and Lin, M., (2016b). “Impact
of aggregate properties on the development of shrink-
age-induced cracking in concrete under restraint con-
ditions.” Cement and Concrete Research, 85, 82-101.
703
Maruyama, I., Sasano, H., Nishioka, Y. and Igarashi, G.,
(2014b). “Strength and young's modulus change in
concrete due to long-term drying and heating up to
90 °C.” Cement and Concrete Research, 66, 48-63.
Maruyama, I. and Sugie, A., (2014). “Numerical study
on drying shrinkage of concrete affected by aggregate
size.” Journal of Advanced Concrete Technology,
12(8), 279-288.
Maruyama, I. and Teshigawara, M., (2010). “Effect of
autogenous shrinkage of UHSC on bending behavior
of RC column.” Sixth International Conference on
Concrete under Severe Conditions: Environment and
Loading Merida, Mexico, CRC press, 2, 879-888.
Maruyama, I., Teshigawara, M., Suwada, H., Tanimura,
M., Mitani, Y., Nakamura, A. and Teramoto, A.,
(2011). “Effect of autogenous shrinkage on the
structural performance of short rc column.”
Summaries of Technical Papers of Annual Meeting
Architectural Institute of Japan., C-2, Structures IV,
Reinforced concrete structures prestressed concrete
structures masonry wall structures, 107-108. (in
Japanese)
McDonald, P. J., Rodin, V. and Valori, A., (2010).
“Characterisation of intra- and inter- C–S–H gel pore
water in white cement based on an analysis of NMR
signal amplitudes as a function of water content.”
Cement and Concrete Research, 40, 1656-1663.
Mitani, Y., Ishii, Y., Tanimura, M. and Maruyama, I.,
(2016). “Evaluation on shrinkage cracks of
expansive/shrinkage reducing concrete.” Cement
Science and Concrete Technology, 69, 341-348. (in
Japanese)
Mitsui, K., Li, Z., Lange, D. A. and Shah, S. P., (1994).
“Relationship between microstructure and mechanical
properties of paste-aggregate interface.” ACI
Materials Journal, 91(1), 30-39.
Muller, A. C. A., Scrivener, K. L., Gajewicz, A. M. and
McDonald, P. J., (2012). “Densification of C–S–H
measured by 1H NMR relaxometry.” The Journal of
Physical Chemistry C, 117, 1, 403-412.
Muller, A. C. A., Scrivener, K. L. Gajewicz, A. M. and
McDonald, P. J., (2013). “Use of bench-top NMR to
measure the density, composition and desorption
isotherm of C–S–H in cement paste.” Microporous
and Mesoporous Materials, 178, 99-103.
Muller, A. C. A., Scrivener, K. L., Skibsted, J., Gajewicz,
A. M. and McDonald, P. J., (2015). “Influence of silica
fume on the microstructure of cement pastes: New
insights from 1H NMR relaxometry.” Cement and
Concrete Research, 74, 116-125.
Neubauer, C., Jennings, H. M and Garboczi, E., (1996).
“A three-phase model of the elastic and shrinkage
properties of mortar.” Advanced Cement Based
Materials, 4(1), 6-20.
Neville, A. M. and Meyers, B. L., (1964). “Creep of
concrete: Influencing factors and prediction.” Special
Publication, 9, 1-33.
Nicoleau, L., Gadt, T., Chitu, L., Maier, G. and Paris, O.,
I. Maruyama / Journal of Advanced Concrete Technology Vol. 14, 691-705, 2016
(2013). “Oriented aggregation of calcium silicate
hydrate platelets by the use of comb-like copolymers.”
Soft Matter, 9(19), 4864-4874.
Nuclear and Industrial Safety Agency, (2011).
“Numerical simualtion of seismic responce of
buildings.” Tohoku Electric Power Co., Inc. Onagawa
Nuclear Power Plant 2 and 3., Nuclear and Industrial
Safety Agency. Buildings/Structure, Tokyo, 4-3-1.
Ogata, Y., Hirotani, K., Aizawa, N., Inoue, N. and
Nozawa, T., (2011). “A study of initial stiffness on the
building changed by passing age at small and
medium-sized earthquake ground motion by arx
model.” Summaries of technical papers of Annual
Meeting Architectural Institute of Japan. B-2,
Structures II, Structural dynamics nuclear power
plants, 2011, 983-984, (in Japanese).
Okada, T., Akino, K., Yoshizaki, S., Inada, Y. and
Sugano, S., (1986). “Model tests for evaluation of
restoring force characteristics of reactor buildings :
Part 6 scale effect test.” Summaries of Technical
Papers of Annual Meeting Architectural Institute of
Japan. B, Structures I, 1986, 1089-1090. (in Japanese)
Ollivier, J. P., Maso, J. C. and Bourdette, B., (1995).
“Interfacial transition zone in concrete.” Advanced
Cement Based Materials, 2(10), 30-38.
Parrott, L. J., Hansen, W. and Berger, R. L., (1980).
“Effect of first drying upon the pore structure of
hydrated alite paste.” Cement and Concrete Research,
10, 647-655.
Pickett, G., (1942). “The effect of change in
moisture-content on the crepe of concrete under a
sustained load.” Journal Proceedings, 38, 333-356.
Pickett, G., (1956). “Effect of aggregate on shrinkage of
concrete and a hypothesis concerning shrinkage.” ACI
Journal proceedings, 52(1), 581-590.
Ping, X. and Beaudoin, J. J., (1992). “Effects of
transition zone microstructure on bond strength of
aggregate-portland cement paste interfaces.” Cement
and Concrete Research, 22, 23-26.
Powers, T. C., (1965). “Mechanism of shrinkage and
reversible creep of hardened cement paste.” The
structure of concrete and its behaviour under load,
319-344.
Raymond, E. D. and Harmer, E. D., (1931). “Flow of
concrete under the action of sustained loads.” ACI
Journal Proceedings, 27(3), 837-901.
Saito, T., (2012). “Change in dynamic characteristics of
a seismically isolated building caused by the 2011 Off
the Pacific Coast of Tohoku Earthquake.” Research
Insitute of Shimizu Corporation. Tokyo, Japan, 89,
13-22. (in Japanese)
Sakaida, T., Sasano, H. and Maruyama, I., (2014).
“Effect of fine aggregate size on drying shrikage of
mortar.” Proceedings of the Japan Concrete Institute,
36(1), 526-531. (in Japanese)
Sato, R. and Kawakane, H., (2008). “A new concept for
the early age shrinkage effect on diagonal cracking
strength of reinforced hsc beams.” Journal of
704
Advanced Concrete Technology, 6(1), 45-67.
Scrivener, K. L., Crumbie, A. and Laugesen, P., (2004).
“The interfacial transition zone (ITZ) between cement
paste and aggregate in concrete.” Interface Science,
12(4), 411-421.
Scrivener, K. L. and Gartner, E. M., (1988).
“Microstructural gradients in cement paste around
aggregate particles.” Materials Research Society
Symposium Proceedings, 114, 77-85.
Sereda, P. J., Feldman, R. F. and Swenson, E. G., (1966).
“Effect of sorbed water on some mechanical properties
of hydrated portland cement pastes and compacts.”
Highway Research Board Special Report, 90, 58-73.
Setzer, M. J., (2007). “The solid-liquid gel-system of
hardened cement paste.” In: M. J. Setzer Ed. Transport
in concrete : nano- to macrostructure, TRANSCON 07,
proceedings of 5th International Essen Workshop.
Essen, Freiburg : Aedificatio Publishers, 3-23.
Shiire, T. and Cheong, S., (1988). “Experimental study
on strength characteristics of massive concrete
structures subjected to high temperature from on side :
In case of subject to temperature of 65℃ from on
side.” Journal of structural and construction
engineering, 387, 8-14. (in Japanese)
Sugie, A., Maruyama, I. and Teshigawara, M., (2014).
“Numerical simulation for stiffness change of RC wall
due to drying.” Summaries of Technical Papers of
Annual Meeting, 2014, 1175-1176. (in Japanese)
Takahashi, K., Morohashi, N., Kudo, K. and Sakurada, T.,
(2012). “Vibration characteristics of school buildings
in the campus of the CIT , Nihon University as inferred
from the strong-motion records from the 2011 Off the
Pacific Coast of Tohoku Earthquake.” Summaries of
Technical Papers of Annual Meeting, 2012, 683-684.
(in Japanese)
Tanimura, M., Sato, R. and Hiramatsu, Y., (2007).
“Serviceability performance evaluation of rc flexural
members improved by using low-shrinkage
high-strength concrete.” Journal of Advanced
Concrete Technology, 5(2), 149-160.
Tasong, W. A., Lynsdale, C. J. and Cripps, J. C., (1999).
“Aggregate-cement paste interface: Part I. Influence of
aggregate geochemistry.” Cement and Concrete
Research, 29, 1019-1025.
Tennis, P. D. and Jennings, H. M., (2000). “A model for
two types of calcium silicate hydrate in the
microstructure of portland cement pastes.” Cement
and Concrete Research, 30, 855-863.
Teranishi, K., Maruyama, I., Saito, K., Hiraiwa T. and
Mori, K., (2011). “Estimation of drying shrinkage
strain of concrete and selection of methods for
shrinkage crack control : Research committee report
on shrinkage crack of concrete.” Concrete Journal,
49(11), 8-16. (in Japanese)
Thierry, J. P., Chanvillard, G., Schlangen, E. and De
Schutter, G., (2008). “Autogenous and drying
shrinkage modeling: From paste to concrete.”
International RILEM Symposium on Concrete
I. Maruyama / Journal of Advanced Concrete Technology Vol. 14, 691-705, 2016
Modelling-ConMod'08, RILEM Publications SARL,
337-344.
Tomes, L. A., Hunt, C. M. and Blaine, R. L., (1957).
“Some factors affecting the surface area of hydrated
portland cement as determined by water-vapor and
nitrogen adsorption.” Journal of Research of the
National Bureau of Standards, 59(6), 357-364.
Toyobe, R., Kojima, H., Tobita, J. and Fukuwa, N.,
(2013). “Long-term seismic response observation for
evaluation of dynamic properties of low and
medium-rise buildings.” Tokai-shibu Kenkyu
Houkokusyu, 51, 177-180. (in Japanese)
Trapote-Barreira, A., Porcar, L., Cama, J., Soler, J. M.
and Allen, A. J., (2015). “Structural changes in C–S–H
gel during dissolution: Small-angle neutron scattering
and Si-NMR characterization.” Cement and Concrete
Research, 72, 76-89.
Uebayashi, H. and Nagano, M., (2012). “Statistical
estimation of stiffness and damping of high-rise
buildings based on the reports concerning the response
analyses of strong motion observation records.”
Summaries of technical papers of annual meeting,
2012, 697-698. (in Japanese)
Völkl, J. J., Beddoe, R. E. and Setzer, M. J., (1987). “The
specific surface of hardened cement paste by
small-angle x-ray scattering effect of moisture content
and chlorides.” Cement and Concrete Research, 17,
81-88.
705
Valori, A., McDonald, P. J. and Scrivener, K. L., (2013).
“The morphology of C–S–H: Lessons from 1H nuclear
magnetic resonance relaxometry.” Cement and
Concrete Research, 49, 65-81.
Winslow, D. N. and Diamond, S., (1974). “Specific
surface of hardened portland cement paste as
determined by small-angle x-ray scattering.” Journal
of the American Ceramic Society, 57(5), 193-197.
Wittmann, F. H., (2008). “Heresies on shrinkage and
creep mechanisms.” Proceedings of the 8th
International Conference on Creep, Shrinkage and
Durability Mechanics of Concrete and Concrete
Structures (CONCREEP 8), Sept, 3-9.
Wittmann, F. H., (1968). “Surface tension skrinkage and
strength of hardened cement paste.” Matériaux et
Construction, 1(6) 547-552.
Wittmann, F. H., (1973). “Interaction of hardened cement
paste and water.” Journal of the American Ceramic
Society, 56(8), 409-415.
Xie, P., Beaudoin, J. J. and Brousseau, R., (1991). “Effect
of aggregate size on transition zone properties at the
portland cement paste interface.” Cement and
Concrete Research, 21, 999-1005.
Zhang, W., Zakaria, M. and Hama, Y., (2013). “Influence
of aggregate materials characteristics on the drying
shrinkage properties of mortar and concrete.”
Construction and Building Materials, 49, 500-510.
Journal of Advanced Concrete Technology Vol. 14, 643-663, October 2016 / Copyright © 2016 Japan Concrete Institute 643

Technical report

Outline of Guidelines for Maintenance and Management of Structures in

Nuclear Facilities
Yoshito Umeki1*, Shohei Sawada2, Shiro Mitsugi3, Toshinobu Maenaka4 and Katsuki Takiguchi5

Received 20 May 2016, accepted 14 October 2016 doi:10.3151/jact.14.643

Abstract
“Guidelines for Maintenance and Management of Structures in Nuclear Facilities” was published by the Architectural
Institute of Japan in 2008 and revised in 2015. This is the first standard that covers the maintenance and management of
structures in Japanese Nuclear Facilities. This has been contributing to the broad activities related to the life-time man-
agement of Nuclear Facilities. This paper outlines these guidelines.

1. Introduction ment of Structures at Nuclear Facilities for use in main-

taining the functions required of these structures
In order to formulate a standard for maintenance and throughout their in-service period.
management of structures in nuclear facilities, “Guide-
lines for Maintenance and Management of Structures in 3. Principle of guidelines
Nuclear Facilities” (hereafter called “Guidelines”) was
published by the Architectural Institute of Japan in 2008 This section outlines “Chapter 2: Basis for mainte-
and revised in 2015 (AIJ 2015). nance”, ”Chapter 3: Required functions and related per-
This paper outlines these Guidelines, which include formance characteristics” and ”Chapter 5: Preparation of
nine chapters as shown in Fig. 1. maintenance plans”, which describe the principles of the
Guidelines.
2. Objective of guidelines
3.1 Basis for maintenance and management
To maintain the functions required of structures in nu- Maintenance and management of structures at nuclear
clear facilities throughout their in-service period, it is facilities shall be performed to maintain their required
important that the functions to be maintained should be functions throughout their in-service periods. The
identified to ensure reliable maintenance based on peri- soundness of these structures shall be evaluated by
odical inspections. These maintenance activities are
expected to prevent trouble likely to occur in long-term
CONTENTS
operation. Maintenance of structures at nuclear facilities
has been based on this idea. The approach to mainte- Chapter 1 General provisions
nance has relied on use of, or on application by analogy 1.1 Objectives
of, related knowledge and standards on degradation due 1.2 Scope of applicability
to aging of general structures and civil engineering 1.3 Terms and definitions
structures. No systematic maintenance guidelines have Chapter 2 Basis for maintenance and management
been established that are commonly applicable to struc- Chapter 3 Required functions and related
tures at nuclear facilities. Thus, the AIJ decided to pre- performance characteristics
pare a set of Guidelines for Maintenance and Manage- Chapter 4 Degradation phenomena and
degradation factors
Chapter 5 Preparation of maintenance plans
5.1 Categories of maintenance
1 5.2 Maintenance planning
Senior Manager, Civil and Architectural Engineering
Chapter 6 Inspection
Department, Nuclear Power Division, Chubu Electric 6.1 Categories of inspections
Power Co., Inc., Nagoya, Japan. *Corresponding author, 6.2 Periodical inspections
E-mail: Umeki.Yoshito@chuden.co.jp 6.3 Extra inspections
2
Deputy Senior Manager, Nuclear Power Department, Chapter 7 Soundness evaluation
Kajima Corporation, Tokyo, Japan. 7.1 Outline of soundness evaluations
3
Acting General Manager, Nuclear Facilities Division, 7.2 Evaluations of current soundness
Taisei Corporation, Tokyo, Japan. 7.3 Evaluations of long-term soundness
4
Manager, Power Facilities, Engineering Department, Chapter 8 Countermeasures
and verification of effectiveness
Takenaka Corporation, Tokyo, Japan.
5 Chapter 9 Records
Professor Emeritus, Tokyo Institute of Technology,
Tokyo, Japan. Fig. 1 Contents of guidelines.
Y. Umeki, S. Sawada, S. Mitsugi, T. Maenaka and K. Takiguchi / Journal of Advanced Concrete Technology Vol. 14, 643-663, 2014 644

verifying satisfaction of the required performance level
of the structural components and members necessary to
fulfill the required structure functions, and shall be se-
cured by taking appropriate countermeasures and veri-
fying their effectiveness.
Information on the relation between required functions,
related performance characteristics, and soundness
evaluations is based on the Nordic Five Level System
(Fig. 2).
When maintenance of structures at nuclear facilities is
applied to this hierarchy, the purpose is to “ensure public
safety through the maintenance of structures at nuclear
facilities.” Specific requirements to achieve this purpose
are functional requirements, which correspond to, e.g.,
functions to support important equipment (supporting
function) and to prevent radioactive material leakage to
outside the structure (leak-proof function). Performance
requirements are specific requirements necessary to
satisfy functional requirements. Hence, performance
requirements refer to, for example, liquid-tightness per-
formance required of the leak-proof function. A per-
formance verification method is for determining the
attainment of a required performance level through
comparison of inspection results, appropriate analysis,
and experimental results with evaluation criteria. A
deemed-to-satisfy provision is a means of securing per-
formance without the need of performance verification
and is established to specify specific structural dimen-
sions, materials, construction methods, etc., deemed to
satisfy the required performance level.
The principle of maintenance and management shown
in the present Guidelines is to conduct the following
activities in accordance with the above concept. The
interrelationship between the activities is shown in Fig. 3.
(1) It is necessary to identify the functions required of
structures at nuclear facilities and the performance
characteristics required for the maintenance of these
functions.
(2) While taking account of the degradation mechanisms
in concrete and steel-framed structures under the
conditions specific to structures at nuclear facilities,
such as environmental, siting, structural, and mate-
rial, the degradation phenomena and degradation
factors to be monitored shall be defined.
(3) Prior to maintenance, a maintenance plan shall be
established. This plan shall describe the basic ap-
proach to inspections, soundness evaluations, coun-
termeasures and their effectiveness verification, and
records.
(4) An inspection plan shall be established based on the
maintenance plan to specify the inspection method,
the parts to be inspected, the timing of inspection, the
inspection implementation scheme, etc. Inspections
shall be performed according to this inspection plan.
Purpose
Functional
requirements
Performance
requirements
Verification Deemed-to-satisfy
methods provisions
Fig. 2 Concept of hierarchical performance design of
buildings.

Required functions and related perform-

ance characteristics (Chapter 3)

Degradation phenomena and degradation

factors
(Chapter 4)
Continuous improvement
of maintenance
Preparation of maintenance plans
(Chapter 5)

Inspection Records
(Chapter 6) (Chapter 9)

Soundness evaluation Countermeasures and verification

(Chapter 7) of effectiveness (Chapter 8)

Fig. 3 Overview of maintenance and management.

Y. Umeki, S. Sawada, S. Mitsugi, T. Maenaka and K. Takiguchi / Journal of Advanced Concrete Technology Vol. 14, 643-663, 2014
(5) An evaluation item shall be set for each degradation
phenomenon and degradation factor to be monitored,
so that the structure can be evaluated for soundness
at the time of inspection and during its in-service
period.
(6) In accordance with the results of soundness evalua-
tion, a countermeasure shall be taken, such as con-
tinuing or improving the current inspection practices
and executing repair work. If any repair work is
executed, the repaired portion shall be checked by
periodical inspection as well as by additional in-
spection as necessary to ensure that the repaired
portion is restored to the performance level expected
at the time of repair planning.
(7) The results of inspection, soundness evaluation,
countermeasure implementation and effectiveness
verification shall be recorded accurately and in detail
and stored with easy access for future maintenance
references.
(8) The cycle from maintenance planning to counter-
measure implementation and effectiveness verifica-
tion shall be repeated for continuous improvement of
maintenance practices to ensure more effective
maintenance.
3.2 Structure performance changes
The structure’s performance affected by degradation due
to aging tends to show a gradual change over time
throughout its in-service period. The degree of this
change depends on the degree of initial performance,
environmental conditions, and maintenance practices.
Meanwhile, the structure’s performance affected by
degradation due to sudden event tends to change within a
short time. Structures at nuclear facilities should be de-
signed to satisfy their required performance level
throughout their in-service period. Even if the required
performance level is changed according to a revision of
relevant codes and standards, structures must satisfy
Performance
Performance decreases fast in the early
stage and slowly in the long term (e.g.,
carbonation)
Initial
performance
Required level of
performance*
Performance increases in
the early stage and stops in-
creasing in the long term (e.g.,
compressive strength)
In-service period
Start of Inspection
operation Soundness
evaluation
(1) Hard-to-replace part
*: The level of performance, which reflects the latest knowledge, required at the point of soundness evaluation (The level of per-
formance required at the time of design is not necessarily used throughout the in-service period of the building.)
Fig. 4 Conceptual graph of changes in performance characteristics of a building prone to degradation due to aging as-
sumed in the maintenance in accordance with the present Guidelines.
645
newly prescribed level. The maintenance covered in the
Guidelines must ensure that the structure’s performance
varied by degradation due to aging or degradation due to
sudden event meets the required level from the present
into the future. In case the required performance level is
changed and some of the inspection results fail to meet
the evaluation criteria, it is reasonably acceptable if the
structure’s soundness as a whole is maintained.
Figure 4 shows a schematic of changes in perform-
ance of a structure affected by degradation due to aging
considered in the Guidelines. The performance of a
hard-to-replace concrete may show different changes
depending on the evaluation item (e.g. The compressive
strength increases in the early stage and stops increasing
in the long term. The carbonation depth proceeds fast in
the early stage and slowly in the long term), but basically
show gradual changes with maintaining the required
performance level. Meanwhile, inherently replaceable
materials like a coating are specified prescriptive re-
quirements (a deemed-to-satisfy provision) such as the
material specifications and the construction methods.
Their performance is maintained by replacing at periodic
intervals throughout its in-service period. The mainte-
nance covered in the Guidelines primarily focuses on
ensuring the non-occurrence of rapid performance deg-
radation in hard-to-replace structures, and is not intended
to restore or improve the structure’s performance, with
the exception of replaceable materials such as coatings.
In other words, in principle, structures shall always be
kept in a sound condition through inspection, soundness
evaluation, and repairs or other countermeasures.
In the seismic diagnosis of a general structure, seismic
performance of the structure is evaluated with consid-
eration of the degradation due to aging. The seismic
performance of a structure is estimated by multiplying a
coefficient according to aging level to its initial per-
formance. Since structures at nuclear facilities are con-
stantly maintained, severe deterioration sometimes ob-
Performance
Repair (replacement)
Initial
performance
Required level of
performance*
Decrease in performance
(without replacement)
In-service period
Time Time
Start of Inspection
operation Soundness
evaluation
(2) Inherently replaceable part
Y. Umeki, S. Sawada, S. Mitsugi, T. Maenaka and K. Takiguchi / Journal of Advanced Concrete Technology Vol. 14, 643-663, 2014
served in general structures is not expected. However, it
is reported that natural frequency of structures decreases
by earthquake effects or by degradation due to aging. In
case the dynamic characteristics of structures at nuclear
facilities show noticeable changes caused by earthquake
effects or degradation due to aging, their structural
soundness should be evaluated with taking account of
such changes. Since the principal cause of these de-
creases in natural frequency due to earthquake effects has
not been clarified, how to specify the dynamic charac-
teristics of structures remains a challenge for analytical
evaluation. It will be necessary in future to promote
studies on soundness evaluation of structures affected by
earthquake and to accumulate scientific knowledge.
3.3 Minimization of damage to a structure dur-
ing maintenance
In an inspection for current soundness evaluation, the
whole structure should be inspected. Accordingly, in
principle, damage to concrete shall be checked by visual
method. For visually unobservable changes in concrete
strength, their tendencies are grasped using
non-destructive methods. There may be cases where
strength is assessed by core sampling for verification and
accuracy enhancement of a non-destructive method.
Core sampling not only causes damage to structures but
also does not necessarily provide more information than
visual methods. Hence, it is assumed that soundness
evaluation will remain valid even if the use of core
sampling is limited.
Meanwhile, inspection for long-term soundness
evaluation is required to measure the degree of effect of
each degradation factor in order to predict the progres-
sion of degradation phenomena. Accordingly, there may
be cases that require use of a local destructive method,
such as core sampling or drilling.
From the above, even in the case of core sampling and
other inspection methods classified as local destructive
methods, inspections for current soundness evaluation
Functions required of buildings at nuclear facilities
Functions specifically required of nuclear facilities
Pressure-
Supporting
proof
function
function
General functions required of buildings
Fig. 5 Functions required of buildings at nuclear facilities.
646
and for long-term soundness evaluation will differ in
specific procedure and/or significance. In any case, the
use of local destructive methods shall be kept to the very
minimum to avoid damage to structures as much as pos-
sible.
3.4 Required functions and related performance
characteristics for nuclear facilities
It is necessary to identify the functions required of
structures at nuclear facilities and the performance
characteristics required for the maintenance of such
functions in accordance with the basic concept described
above in Section 3.1. The required functions of struc-
tures at nuclear facilities are shown in Fig. 5. These are
classified into two functions, one for ordinary structures
and the other specially required for structures at nuclear
facilities.
The functions required of a structure generally include
supporting function for its own dead weight and live and
seismic loads; comfort use-related functions such as
vibration isolation and sound proofing; functions to
withstand secular changes and to resist fire and other
high-temperature phenomena. These functions are simi-
larly required of nuclear facilities. In the Guidelines,
these general functions shall be collectively called the
usage function.
The following additional functions are specially re-
quired for structures at nuclear facilities.
・Supporting function
A function with which to support major installations
and equipment during normal operation or an earth-
quake.
・Pressure-proof function
A function with which to withstand pressures. In the
case of a concrete containment vessel, a function re-
quired of its concrete portion.
・Shielding function
A function with which to reduce the radiation dose to
a predetermined level to prevent radiation exposure
Negative Secondary
Shielding Leak-proof -pressure -accident
function function holding prevention
function function
Usage function
Y. Umeki, S. Sawada, S. Mitsugi, T. Maenaka and K. Takiguchi / Journal of Advanced Concrete Technology Vol. 14, 643-663, 2014
Table 1 Required functions and related performance characteristics.
Required
function Usage Supporting Pressure-proof
Performance function function function
characteristics
Structural safety 3 3
Usability 3*1 - -
Shielding
- -
performance
Durability (3)*2 (3)*2 (3)*2
Fire resistance (3)*3 (3)*3
*1: Aesthetic appearance shall be considered as necessary.
*2: In the present Guidelines, durability shall be considered as a performance characteristic included in structural safety, usability,
and shielding performance.
*3: In the present Guidelines, fire resistance shall be considered as part of structural safety and usability.
impact on the general public and occupationally ex-
posed personnel.
・Leak-proof function
A function with which to prevent leakage of liquid
radioactive materials from a specified area to outside.
・Negative-pressure holding function
A function in which some areas in a nuclear facility
are kept under negative pressure by ventilation
air-conditioning equipment to reduce external dis-
charge of radioactive materials.
・Secondary-accident prevention function
A function with which to prevent any damage, falling,
or overturning of a structure from compromising the
functions of installations and equipment.
While the supporting and secondary-accident preven-
tion functions can be regarded here as part of the usage
function required also of general structures, these func-
tions are intentionally defined as those specifically re-
quired of nuclear facilities, emphasizing that nuclear
facilities are required to have a higher level of safety
against seismic and similar loads than general structures.
Whether a structure’s required functions can be
maintained depends on the performance of its structural
components and members. Therefore, the required func-
tions must be related to the required performance char-
acteristics to define the levels of performance necessary
to keep the structure functional.
Table 1 shows the relation of the required functions
covered in the Guidelines to the performance character-
istics of structural components and members. Supporting,
pressure-proofing, shielding, leak-proofing, nega-
tive-pressure holding, and secondary-accident preven-
tion functions are defined as those required of structures
at nuclear facilities.
Structural safety characteristics shall include
load-bearing capacity for general loads (structure’s dead
weight, live loads, seismic loads, wind loads, and snow
loads) taken into consideration in the design of a struc-
647
Secondary-
Negative-
Shielding Leak-proof accident
pressure holding
function function prevention
function
function
3 - - - 3
3 3
- -
(liquid-tightness) (airtightness)
- 3 - - -
(3)*2 (3)*2 (3)*2 (3)*2
- - - - (3)*3
ture and shall also include load-bearing capacity for
temperature loads, fatigue capacity against cyclic loads,
and impact resistance capacity against impact loads.
Usability is a generic term for performance charac-
teristics required for comfortable use of structures and
refers to, among other things, resistance capacity against
increase in deflection under permanent load due to creep,
vibration isolation performance against harmful vibra-
tion, thermal insulation performance against heat, wa-
terproof performance against ordinary rainfall and water
usage, liquid- and air-tightness of concrete skeletons
constructed by conventional methods, and surface pro-
tection of coatings and surface finishes. In some cases,
aesthetic appearance not directly related to performance
characteristics of structural components and members
may have to be taken into consideration and hence shall
be included in usability.
Shielding performance is specifically required of some
areas of each structure in nuclear facilities and refers
mainly to neutron and gamma ray shielding performance.
Durability shall be regarded as a generic term for re-
sistance to degradation due to aging in various per-
formance characteristics and shall be included in per-
formance characteristics related to structural safety, us-
ability, and shielding performance rather than as an in-
dependent performance category.
For fire resistance, performance characteristics, such
as damage resistance, heat shielding performance, and
flame shielding performance, and the specification code
for their satisfaction are stipulated in, e.g., the Structure
Standards Act and the Order for Enforcement of the same,
the provisions of which shall be complied with for nu-
clear facilities. In the Guidelines, fire damage resistance
shall be considered as part of structural safety
(load-bearing capacity for special loads) while heat and
flame shielding performance shall be considered as part
of usability (protection of the internal environment
against heat).
Y. Umeki, S. Sawada, S. Mitsugi, T. Maenaka and K. Takiguchi / Journal of Advanced Concrete Technology Vol. 14, 643-663, 2014 648

Table 2 Comparison of two maintenance categories.

Category of
Maintenance for ensuring current soundness Maintenance for ensuring long-term soundness
maintenance
Item to be monitored Degradation phenomenon Degradation factor*1
Degradation due to
Type of degradation Degradation due to aging Degradation due to aging
sudden event
Periodical inspection for Periodical inspection for long-term soundness
Type of inspection Extra inspection
current soundness evaluation evaluation
Part to be inspected Buildings in their entirety Representative part
Inspection method Mainly visual method Mainly local destructive method
Item to be evaluated Presence/absence or degree of degradation
Degree of effect of degradation factor
for soundness phenomenon
Soundness Comparison between inspection results and Comparison between progression prediction and
evaluation method evaluation criteria evaluation criteria based on inspection results
Timing for inspection
As a situation such as Allowed to be less frequent than the maintenance
and soundness Every 5 years or less*2
an earthquake occurs for ensuring the current soundness
evaluation
*1: Degradation factor leading to a degradation phenomenon that may mainly affect the structural safety and shielding performance
of a building
*2: In the case of a visual method

Thus, the functions required of structures at nuclear performance manifest itself, and that this shall be
facilities shall be related to structural safety, usability, achieved by repairs and other countermeasures. Ac-
and shielding performance in the Guidelines. cordingly, the Guidelines assume the following two
categories of maintenance:
3.5 Two maintenance methods ・Maintenance for ensuring current soundness
The maintenance in the Guidelines assumes that struc- In order to check structures for current soundness,
tures shall be checked for performance by inspection and their current condition shall be assessed by grasping
soundness evaluation and that they shall be kept in sound the presence/absence or degree of any degradation
condition without letting any noticeable decrease in phenomenon, and an appropriate countermeasure(s)
shall be taken.
Chapter 5 Preparation of maintenance plans ・Maintenance for ensuring long-term soundness
Evaluation of In order to ensure that structures’ functions will be
Maintenance planning validity of maintained into the future, the degrees of effect of
maintenance plan degradation factors shall be grasped before the oc-
currence of any degradation phenomenon. At the same
Chapter 6 Inspection time, the long-term effect shall be evaluated by a
Preparation of periodical and extra method such as progression prediction to take an ap-
inspection plans propriate countermeasure(s).
Maintenance for ensuring current soundness shall be
Implementation of periodical and practiced as the basic mode of maintenance and shall be
extra inspections for current
soundness evaluation supplemented by maintenance for ensuring long-term
(mainly visual method) soundness to make it possible to demonstrate that current
structural soundness is maintained and to ensure that no
Chapter 7 Soundness evaluation noticeable decrease in performance will occur during the
Current soundness evaluation rest of the in-service period. Table 2 compares the two
(Evaluation item: presence/absence or maintenance categories.
degree of degradation phenomenon)
(1) Maintenance for ensuring the current
Chapter 8 Countermeasures and verification of effectiveness soundness
The purpose of this maintenance is to ensure that current
Countermeasure implementation
structural soundness is maintained. Hence, this mainte-
and effectiveness verification
nance shall be performed relatively frequently and the
specific activities involved shall include investigating the
Chapter 9 Records presence/absence of degradation likely to lead to reduced
Recording and storage structural performance and taking an appropriate coun-
termeasure(s) in response to the evaluation results.
Figure 6 shows the flow of the specific maintenance
Fig. 6 Implementation flow for the maintenance for en-
activities to be implemented.
suring the current soundness.
Y. Umeki, S. Sawada, S. Mitsugi, T. Maenaka and K. Takiguchi / Journal of Advanced Concrete Technology Vol. 14, 643-663, 2014
In the Guidelines, the causes of a direct decrease in
structural performance are regarded as degradation
phenomena.
Degradation due to aging that occurs over time shall
be investigated by periodical inspection for current
soundness evaluation. Meanwhile, degradation due to
earthquakes, typhoons, fires, and other sudden events
shall be investigated by extra inspection. These degra-
dation phenomena often occur in the form of damage
such as cracking in concrete surfaces and coated surfaces,
except for some phenomena, such as reduced concrete
strength, and can therefore be grasped by visual methods.
This is performed to determine the presence/absence of
any current structural damage and hence in principle
shall be performed for structures in their entirety.
Inspection results shall be checked against evaluation
criteria to perform structure soundness evaluation and
consider countermeasures. If it is necessary to investigate
degradation factors contributory to any degradation phe-
nomenon in order to consider repair methods and other
countermeasures, non-destructive, local destructive, and
other methods shall be used to collect detailed data. After
any repair, its effectiveness shall be verified. At the same
time, the results of inspection, soundness evaluation, and
countermeasure implementation and effectiveness veri-
fication shall be recorded and stored to review the
maintenance plan as necessary.
(2) Maintenance for ensuring long-term
soundness
The purpose of this maintenance is to check structures
for soundness now and into the future. Even if no likely
degradation phenomenon manifests itself now, the de-
grees of effect of degradation factors contributory to
degradation phenomena shall be grasped and specific
activities shall include predicting whether structural
performance would be maintained above the required
level during the assumed period. Figure 7 shows the flow
of the specific maintenance activities to be implemented.
Since this maintenance is intended for degradation that
proceeds over time, its scope does not include degrada-
tion due to sudden event. Hence, necessary data shall be
collected by periodical inspection for long-term sound-
ness evaluation. This data collection shall precede the
manifestation of any degradation phenomenon. Accord-
ingly, a local destructive method may become necessary
rather than a visual method or a non-destructive method.
The parts to be inspected must be specified in advance
and must be representative of each structure. Moreover,
from the perspective of the importance of structures at
nuclear facilities, precautions must be taken to avoid
damaging the skeleton due to sampling of specimens. It
is desirable to minimize the number of inspection points.
This maintenance uses techniques, such as local de-
structive methods that cause damage to skeletons.
Therefore, this maintenance should not be performed for
all degradation factors. Instead, the degradation factors to
be monitored should be extracted with focus on those
649
contributory to any degradation phenomenon that may
affect the structural safety and shielding performance
that are critical for the safety of structures. Degradation
phenomena are expected to proceed at points likely to be
significantly affected by degradation factors. Therefore,
by selecting such points as representative parts, the
number of inspection points can be reduced. It should be
noted that degradation factors that affect the usability of
structures should be deemed as having already been
assessed because the related degradation phenomena are
checked in the maintenance for ensuring current sound-
ness.
The degrees of effect of degradation factors shall be
grasped based on data collected through inspection. At
the same time, a progression prediction shall be made by
checking the predicted values against the evaluation
criteria to perform soundness evaluation and consider
countermeasures. After any repair is performed, its ef-
fectiveness shall be verified. At the same time, the results
of inspection, soundness evaluation, and countermeasure
implementation and effectiveness verification shall be
recorded and stored to review the maintenance plan as
necessary. Since this maintenance is intended for deg-
radation that does not proceed rapidly, the required in-
spection and soundness evaluation may be performed at
intervals longer than in maintenance for ensuring current
soundness. Note that use of longer intervals is effective
in minimizing damage to skeletons due to the local de-
structive method.
Chapter 5 Preparation of maintenance plans
Evaluation of
Maintenance planning validity of
maintenance plan
Chapter 6 Inspection
Preparation of periodical
inspection plans
Implementation of periodical inspection
for long-term soundness evaluation
(mainly local destructive method)
Chapter 7 Soundness evaluation
Long-term soundness evaluation
(Evaluation item: degrees of
effect of degradation factors)
Chapter 8 Countermeasures and verification of effectiveness
Countermeasure implementation
and effectiveness verification
Chapter 9 Records
Recording and storage
Fig. 7 Implementation flow for maintenance for ensuring
long-term soundness.
Y. Umeki, S. Sawada, S. Mitsugi, T. Maenaka and K. Takiguchi / Journal of Advanced Concrete Technology Vol. 14, 643-663, 2014 650

4. Degradation phenomena and b) Selection of degradation factors based on the charac-

degradation factors
(1)Approach to selection of degradation phe-
nomena and degradation factors to be moni-
tored
In the Guidelines, the terms, degradation, degradation
phenomenon, and degradation factors are defined as
follows:
・Degradation:
A decrease in the performance of a structure.
・Degradation phenomenon
A phenomenon that leads to decreased performance
of structural components or members in a structure.
・Degradation factor
A major factor leading to a degradation phenomenon.
Figure 8 shows the flow for selecting degradation
phenomena affecting the performance of structures at
nuclear facilities and the responsible degradation factors.
First, based on domestic and overseas mainte-
nance-related guidelines, data are collected and collated
on the degradation phenomenon, degradation factors,
and degradation mechanisms that affect the performance
characteristics of structures as summarized in Chapter 3
“Required functions and related performance” of the
Guidelines. Then, selection is made of the degradation
phenomena and degradation factors to be monitored for
maintenance, taking into consideration the characteris-
tics of structures at nuclear facilities, such as environ-
mental conditions, site conditions, structural conditions,
and material conditions.
Degradation phenomena and degradation factors for
each of the concrete and steel-framed structures are se-
lected according to the following three steps:
a) Collection and collation of existing knowledge on
degradation phenomena and degradation factors
teristics of nuclear facilities
c) Summary of degradation phenomena and degradation
factors to be monitored
The steps are outlined as follows:
(2) Collection and collation of existing knowl-
edge on degradation phenomena and degrada-
tion factors
Based on domestic and overseas guidelines currently
available, existing knowledge of degradation phenomena
and degradation factors in general structures and struc-
tures at nuclear facilities is collected and collated for
concrete structures and steel-framed structures. There
may be cases where what is regarded in certain guide-
lines as a degradation phenomenon is regarded in others
as a degradation factor. In the Guidelines, what causes a
direct decrease in performance is regarded as a degrada-
tion phenomenon, while what causes a degradation
phenomenon is classified as a degradation factor. For
example, some guidelines regard concrete carbonation as
a degradation phenomenon. Concrete carbonation,
however, does not compromise structural safety. It is not
until rebar corrosion occurs due to the progression of
concrete carbonation that structural safety is compro-
mised. Therefore, carbonation is regarded herein as a
degradation factor. Similarly, chloride penetration itself
does not cause a direct decrease in performance but
causes rebar corrosion, which leads to a decrease in
structural safety. Hence, chloride penetration is also
regarded as a degradation factor. Some guidelines regard
cracking as a degradation factor because water and
oxygen penetration via cracks results in rebar corrosion.
Cracking, however, may cause a direct decrease in
structural safety or usability and hence is regarded in the

Required functions and related performance characteristics (Chapter 3)

Collection and collation of existing knowledge on degradation phenomena and degradation factors
• Degradation phenomena and degradation factors
• Relation between performance characteristics and degradation phenomena
• Degradation mechanisms

Characteristics of buildings at nuclear facilities

• Environmental conditions
• Site conditions
• Structural conditions
• Material conditions, etc.

Summary of degradation phenomena and degradation factors to be

monitored for maintenance of buildings at nuclear facilities

Maintenance planning (Chapter 5)

Fig. 8 Selection flow for degradation phenomena and degradation factors.

Y. Umeki, S. Sawada, S. Mitsugi, T. Maenaka and K. Takiguchi / Journal of Advanced Concrete Technology Vol. 14, 643-663, 2014 651

Table 3 Degradation phenomena and degradation factors specified by existing guidelines to be monitored (for concrete
structures).
Degradation factor

(elevated temperatures)
Heat
Radiation exposure

Carbonation

Chloride penetration

Alkali-aggregate reaction

Machine vibration

Freeze-thaw

Chemical attack

Drying shrinkage

Creep

Insolation

Weathering

Vehicular traffic, etc.

Differential settlement

Overloads
typhoons, etc.
Loads due to earthquakes,

Heat due to fire

Degradation
phenomenon

Reduced concrete
3 3 3 3
strength
Cracking 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3
Spalling and falling 3 3 3 3 3 3 3 3 3
Rebar corrosion 3 3 3 3 3 3
Deflection 3 3
Vibration 3 3
Abrasion 3
Explosive spalling 3
Water leakage 3
Moisture diffusion 3 3 3

Table 4 Degradation phenomena and degradation factors specified by existing guidelines to be monitored (for steel
framed structures).

Degradation factor

(loss of galvanized plating)

Cyclic loads due to winds,

Loads due to earthquakes,

Coating degradation

Plating degradation

Heat due to fire

typhoons, etc.
Degradation phenomenon etc.(fatigue)

Steel corrosion 3 3
Reduced strength due to heat 3
Yielding, buckling, crack, fracture, and excessive deformation of steel 3 3
Slipping, crack, and fracture of joint
3 3 3
Looseness and fracture of bolt
Reduced strength due to fatigue 3

Guidelines as a degradation phenomenon. are coating degradation and plating degradation (loss of
Relations between degradation phenomena and deg- galvanized plating). Cyclic loads due to wind, etc. (fa-
radation factors specified by existing guidelines are tigue) cause reduced strength due to fatigue, a degrada-
summarized for concrete and steel-framed structures as tion phenomenon.
shown in Tables 3 and 4. The selected degradation phenomena may decrease
Degradation factors leading to degradation phenom- the performance characteristics of structures at nuclear
ena of reduced concrete strength are heat (elevated facilities. Tables 5 and 6 show the degradation phe-
temperatures), radiation exposure, weathering, and heat nomena that may decrease performance characteristic
due to fire. Concrete cracking is regarded, as mentioned values (structural safety, usability, and shielding per-
above, as a degradation phenomenon. It has several formance) for concrete and steel-framed structure, re-
degradation factors including heat (elevated temperature), spectively. Degradation phenomena related to each of the
radiation exposure, carbonation, chloride penetration, performance characteristics are classified into degrada-
and machine vibration. tion due to aging and degradation due to sudden event, to
For steel-framed structures, the degradation factors be consistent with the maintenance plan, where earth-
leading to the degradation phenomenon of steel corrosion quake, typhoon, fire, etc. are assumed to cause degrada-
Y. Umeki, S. Sawada, S. Mitsugi, T. Maenaka and K. Takiguchi / Journal of Advanced Concrete Technology Vol. 14, 643-663, 2014 652

Table 5 Relation of performance characteristics of concrete structures at nuclear facilities to degradation phenomena.
Performance characteristic
Structural safety
Usability
Shielding performance
Degradation phenomenon
Degradation due to sudden event
Degradation due to aging
(earthquake, typhoon, fire, etc.)
Reduced concrete strength Reduced concrete strength
Cracking Cracking
Rebar corrosion Spalling and falling
Spalling and falling Explosive spalling
Cracking Cracking
Spalling and falling Spalling and falling
Deflection Explosive spalling
Vibration
Abrasion
Water leakage
Moisture diffusion Moisture diffusion
Cracking Cracking
Spalling and falling Spalling and falling
Explosive spalling

Table 6 Relation between performance characteristics and degradation phenomena in steel framed structures at nuclear
facilities.
Degradation phenomenon
Performance char-
Degradation due to sudden event
acteristic Degradation due to aging
(earthquake, typhoon, fire, etc.)
Structural safety Reduced strength due to heat
Steel corrosion
Yielding, buckling, crack, fracture, and excessive
Slipping, crack, and fracture of joint
deformation of steel
Looseness and fracture of bolt
Slipping, crack, and fracture of joint
Reduced strength due to fatigue
Looseness and fracture of bolt
Usability Yielding, buckling, crack, fracture, and excessive
Steel corrosion
deformation of steel
Slipping, crack, and fracture of joint
Slipping, crack, and fracture of joint
Looseness and fracture of bolt
Looseness and fracture of bolt

tion due to sudden event.
The relations between performance characteristics and
degradation phenomena are as follows:
a) The structural safety of a concrete structure may be
reduced due to cross-sectional area reductions and
cracks resulting from reduced concrete strength, rebar
corrosion, explosive spalling, or spalling and drop-
ping.
b) Usability is a generic term for performance charac-
teristics required for comfortable use of structures and
refers to, among other things, liquid-tightness,
air-tightness, resistance capacity to increases in de-
flection, and vibration isolation performance against
harmful vibration. Liquid-tightness and air-tightness
may be reduced due to cracking, while resistance ca-
pacity against increases in deflection may be reduced
due to creep and overloads.
c) Shielding performance relates to concrete density and
member thickness. Density reductions due to moisture
diffusion, member thickness reductions due to explo-
sive spalling or spalling and dropping, and excessive
cracking may result in decreases in shielding per-
formance.
d) Structural safety of a steel-framed structure may de-
crease due to steel corrosion, reduced strength due to
heat, yielding, buckling, cracking, fracturing or ex-
cessive deformation of any steel member, slipping,
cracking or fracturing of any joint, looseness or frac-
turing of any bolt, or any reduced strength due to fa-
tigue.
e) The usability of a steel-framed structure relates to,
among other things, vibration isolation performance
against harmful vibration or resistance capacity
against increases in deflection. These performance
characteristics may be affected by steel corrosion,
yielding, buckling, cracking, fracturing, or excessive
deformation of any steel member; slipping, cracking
or fracturing of any joint; or looseness or fracture of
any bolt.
It is important for selection of degradation phenomena
and degradation factors, evaluation, and progression
prediction, to understand degradation mechanisms for
each degradation factor. Degradation mechanisms are
summarized for each degradation factor specified by
existing guidelines to be monitored. The degradation
factors are shown in Tables 3 and 4 for concrete and
steel-framed structures, respectively.
(3) Selection of degradation factors based on
characteristics of nuclear facilities
Structures at nuclear facilities exist in a unique envi-
ronment. Degradation mechanics need to be considered
Y. Umeki, S. Sawada, S. Mitsugi, T. Maenaka and K. Takiguchi / Journal of Advanced Concrete Technology Vol. 14, 643-663, 2014
when degradation factors are selected according to con-
ditions such as environmental conditions and site condi-
tions. Moreover, degradation mechanics are also impor-
tant for evaluation of degradation progression processes
and prediction of progression.
Degradation factors to be monitored are selected
through summarization of characteristics of nuclear fa-
cilities. The following conditions of nuclear facilities are
used as selection criteria: a) Environmental conditions,
b) Site conditions, c) Structural conditions, d) Material
conditions and e) Concrete surface protection conditions.
a) Environmental conditions
・ Parts of structures in nuclear facilities are exposed to
elevated temperatures.
・ Nuclear power plants use the nuclear fission reaction
of radioactive materials, and parts of their structures
are exposed to radiation.
・ Many machines are installed in structures at nuclear
facilities, and many parts of structures are exposed
to vibration.
・ Structures in nuclear facilities include
bulk-receiving docks with floors, etc., exposed to
loads due to inbound and outbound vehicular traffic
and equipment deliveries.
・ With fewer people inside than in general structures,
structures at nuclear facilities tend to be low in
carbon dioxide gas concentration.
b) Site conditions
・ Nuclear power plants in Japan are located in coastal
areas, and their facilities may be exposed to airborne
sea salt particles.
・ Existing nuclear facilities are located in either
semi-frost damage prone areas or general frost
damage prone areas with a minor or higher frost
damage risk.
・ No existing nuclear facilities are located in vicinities
of hot-spring areas or acidic river basins.
・ Nuclear facilities are not different from general
structures regarding factors (ultraviolet rays, heat,
moisture, blown sand, mechanical stress, etc.)
leading to carbonation (outdoors), insolation,
weathering, drying, or coating degradation.
c) Structural conditions
・ Many parts of steel reinforced concrete structures at
nuclear facilities have larger cross-sections and
covering thicknesses than those in general struc-
tures.
・ In principle, structures at nuclear facilities are sup-
ported directly on a rock mass or indirectly via piles.
・ Structures at nuclear facilities are equipped with
many pieces of heavy equipment. Therefore, larger
live loads are taken into consideration than in gen-
eral structures.
d) Material conditions
・ When structures at nuclear facilities are built,
stringent control is in place for the selection and
mixing of concrete materials.
653
e) Concrete surface protection conditions
・ Outdoor coverings in structures at newer nuclear
facilities are thicker than those in general structures
and are protected with a finishing material such as a
coating.
・ Moreover, many parts have indoor concrete surfaces
coated for easier removal of radioactive materials
adhering to them (decontamination).
・ Furthermore, parts in contact with liquids are cov-
ered with liners or coating films.
For degradation factors selected according to mainte-
nance-related guidelines, degradation mechanisms are
summarized, and ones to be monitored are selected based
on the characteristics of nuclear facilities. Table 7 and
Table 8 show the selected degradation factors to be
monitored for concrete and steel-framed structures at
nuclear facilities.
Chemical attack, differential settlement, and overloads
are not selected as degradation factors for degradation
due to aging of concrete structures. The reasons include
the facts that no existing nuclear facilities are located in
vicinities of hot-spring areas or acidic river basins and
that differential settlement is unlikely because the struc-
tures at nuclear facilities are supported directly on bed-
rock or indirectly via piles. Freeze-thaw is selected in
areas with a minor or higher frost damage risk as a deg-
radation factor to be monitored only for usability.
All the degradation factors for steel-framed structures
are selected to be monitored.
(4) Summary of degradation phenomena and
degradation factors to be monitored
Based on the results mentioned above, Tables 9 and 10
show the relations between degradation phenomena and
degradation factors to be monitored for concrete and
steel-framed structures, respectively, at nuclear facilities.
In the tables, degradation phenomena are grouped by
performance characteristics, and combinations with
degradation factors are shown. Degradation due to aging
and degradation due to sudden event are shown in re-
spective tables.
5. Inspection and soundness evaluation
Inspection and Soundness evaluation follows the PDCA
cycle (Plan, Do, Check and Action) and plans for im-
provement through the in-service period. Inspection and
Soundness in the PDCA cycle includes collecting basic
information on maintenance, predicting soundness and
contributing to Nuclear Safety.
Inspection should use a general way and the latest way.
This section describes a basic way of thinking about the
choice of an inspection method. The criteria in Sound-
ness evaluation reflects the latest knowledge and a re-
cord.
This section, in addition to the Inspection and
Soundness evaluation, also describes “Countermeasures
and verification of effectiveness” and “Records”.
Y. Umeki, S. Sawada, S. Mitsugi, T. Maenaka and K. Takiguchi / Journal of Advanced Concrete Technology Vol. 14, 643-663, 2014 654

Table 7 Selection results of degradation factors to be monitored for concrete structures at nuclear facilities (Legend: 3: To be
selected as degradation factor and −: Not to be selected as degradation factor)
Selection
Degradation factor Applicable part Remarks
results
Heat (elevated tem- Reactor body foundation,
3
peratures) internal concrete, etc.
Radiation exposure Reactor body foundation, shell
3
wall, and internal concrete
Carbonation Concrete in general • In applicable parts with coating, degradation progression is
3
inhibited.
Chloride penetration Buildings’ external wall • In applicable parts with coating, degradation progression is
3
surfaces inhibited.
Alkali-aggregate Concrete in general
3
reaction
Machine vibration Turbine generator pedestal,
emergency generator
3
foundations, pump and motor
foundations, etc.
Freeze-thaw Local parts around openings • In areas with a minor or higher frost damage risk, only usabil-
3*
and parapets on external walls ity shall be monitored.
Chemical attack • No existing nuclear facilities are located in vicinities of
− − hot-spring areas or acidic river basins.
• Indoor parts are unlikely to undergo chemical attack.
Drying shrinkage Walls, columns, beams, floors, • Structural safety and shielding performance do not seem
3 etc. likely to be compromised. Accordingly, only usability shall be
monitored.
Creep Beams and floors • The effect of creep has been taken into design consideration.
3
Accordingly, only usability shall be monitored.
Insolation Rooftop, etc. • The effect of temperature rise is minor even on a rooftop, with
a waterproof layer and its protective material provided on it.
3
Accordingly, only usability shall be monitored.
• The walls restraining a rooftop floor shall also be assessed.
Weathering Buildings’ external wall sur- • Considering the large concrete covering thicknesses in nuclear
faces facilities, weathering has a minor effect on the structural safety
3 and shielding performance of buildings. Accordingly, only
usability shall be monitored.
• Applicable parts with coating are less prone to degradation.
Vehicular traffic, etc. Floors, etc. • Degradation will be limited to the surface layers of applicable
3
parts. Accordingly, only usability shall be monitored.
Differential settle- • In principle, nuclear facilities are supported directly on a rock
ment − − mass or indirectly via piles on it and are unlikely to undergo
differential settlement.
Overloads • Prior to installing any heavy equipment, etc., exceeding the
load assumed at the design phase, an evaluation is performed
− −
by an architect(s)/engineer(s) as necessary to ensure the
soundness of structural components and that of members.
Loads due to earth- Walls, columns, beams, floors,
3
quakes, typhoons, etc. etc.
Heat due to fire 3 Parts in rooms on fire
*: Dependent on site conditions.

Table 8 Selection results of degradation factors to be monitored for steel-framed structures at nuclear facilities (Legend: 3:
To be selected as degradation factor).
Degradation factor Results of selection Typical applicable part Remarks
Coating degradation 3 Stacks and roof trusses
Plating degradation (loss of gal-
3 Stacks and roof trusses
vanized plating)
Cyclic loads due to winds, etc.
3 Stacks
(fatigue)
Loads due to earthquakes, ty-
3 Stacks and roof trusses
phoons, etc.
Heat due to fire 3 Stacks and roof trusses
Y. Umeki, S. Sawada, S. Mitsugi, T. Maenaka and K. Takiguchi / Journal of Advanced Concrete Technology Vol. 14, 643-663, 2014 655

Table 9(a) Summary of degradation phenomena and degradation factors in degradation due to aging of concrete
structures at nuclear facilities.
Degradation factor

peratures)
Heat (elevated tem-

Radiation exposure

Carbonation

Chloride penetration

action
Alkali-aggregate re-

Machine vibration

Freeze-thaw

Drying shrinkage

Creep

Insolation

Weathering

Vehicular traffic, etc.

Performance Degradation phe-
characteristic nomenon

Structural Reduced concrete

3 3
safety strength
Cracking 3 3 3 3 3 3
Rebar corrosion 3 3 3
Spalling and falling 3 3 3 3
Usability Cracking 3 3 3 3 3 3 3* 3 3 3
Spalling and falling 3 3 3 3 3*
Deflection 3
Vibration 3 3
Abrasion 3
Water leakage 3
Shielding Moisture diffusion 3 3
performance
Cracking 3 3 3 3 3 3
Spalling and falling 3 3 3 3
*: Dependent on site conditions

Table 9(b) Summary of degradation phenomena and degradation factors in degradation due to sudden event of concrete
structures at nuclear facilities.

Performance Degradation factor

Degradation phenomenon
characteristic Loads due to earthquakes, typhoons, etc. Heat due to fire
Structural safety Reduced concrete strength 3
Cracking 3 3
Spalling and falling 3 3
Explosive spalling 3
Usability Cracking 3 3
Spalling and falling 3 3
Explosive spalling 3
Shielding Moisture diffusion 3
performance Cracking 3 3
Spalling and falling 3 3
Explosive spalling 3
Y. Umeki, S. Sawada, S. Mitsugi, T. Maenaka and K. Takiguchi / Journal of Advanced Concrete Technology Vol. 14, 643-663, 2014
Table 10(a) Summary of degradation phenomena and degradation factors in degradation due to aging of steel-framed
structures at nuclear facilities.
Performance
Degradation phenomenon
characteristic
Structural safety Reduced strength due to fatigue
Steel corrosion
Slipping, crack, and fracture of joint
Looseness and fracture of bolt
Usability Steel corrosion
Slipping, crack, and fracture of joint
Looseness and fracture of bolt
Table 10(b) Summary of degradation phenomena and degradation factors in degradation due to sudden event of
steel-framed structures at nuclear facilities.
Performance
Degradation phenomenon
characteristic
Structural safety Reduced strength due to heat
Yielding, buckling, crack, fracture, and
excessive deformation of steel
Slipping, crack, and fracture of joint
Looseness and fracture of bolt
Usability Yielding, buckling, crack, fracture, and
excessive deformation of steel
Slipping, crack, and fracture of joint
Looseness and fracture of bolt
(1) Inspection
a. Categories of inspections
Inspections are divided into periodical inspections per-
formed at periodic intervals and extra inspections per-
formed in response to earthquake, typhoon, fire, etc. The
purpose of periodical inspections is to grasp the state of
degradation due to aging of structures at nuclear facilities.
The purpose of extra inspections is to check for damage
to all parts of structures at nuclear facilities immediately
after an earthquake typhoon, or fire.
b. Periodical inspections
1) Category and method
Periodical inspections are divided largely into periodical
inspections for current soundness evaluation and periodi-
Direct methods
Periodical inspection
for current soundness
evaluation Indirect methods
Periodical
inspection Direct methods
Periodical inspection
for long-term sound-
ness evaluation Indirect methods
Fig. 9 Categories and methods of periodical inspections.
656
Degradation factor
Coating degradation
(loss of galvanized plating)
Plating degradation
winds, etc. (fatigue)
Cyclic loads due to
3
3 3
3
3 3
3
Degradation factor
Loads due to earthquakes, typhoons, etc. Heat due to fire
3
3 3
3 3
3 3
3 3
cal inspections for long-term soundness evaluation. Each
is subdivided into inspections using a direct or indirect
method. In any case, inspections shall be performed by a
direct method represented by visual, non-destructive, and
local destructive methods and alternatively by an indirect
method (an environmental conditions survey-based
method, a nearby inspection results-based method, or a
method using simulated specimens). Periodical inspec-
tions for current soundness evaluation shall be performed
on structures in their entirety, in principle using visual and
non-destructive methods. Meanwhile, periodical inspec-
tions for long-term soundness evaluation may involve the
use of local destructive methods for representative parts.
Figure 9 shows the categories and methods of periodical
inspections.
Visual and non-destructive methods
Environmental conditions survey-based method
Nearby inspection results-based method
Method using simulated specimens
Local destructive method
Environmental conditions survey-based method
Nearby inspection results-based method
Method using simulated specimens
Y. Umeki, S. Sawada, S. Mitsugi, T. Maenaka and K. Takiguchi / Journal of Advanced Concrete Technology Vol. 14, 643-663, 2014 657

Table 11 Periodical inspections for current soundness evaluation (concrete structures).

Degradation Related perform-
Applicable part Inspection item Inspection method
phenomenon ance characteristic
Parts heated to high tem-
Temperature
perature
Indirect methods (temperature analysis and radia-
Parts subjected to
Reduced tion dose analysis)
long-term radiation expo- Radiation dose
concrete Structural safety
sure
strength
Non-destructive methods (rebound number
Concrete structures in Change in com-
method and elastic wave method)
general pressive strength
Local destructive method (core-sampling method)
Moisture Shielding per- Parts required to be
Temperature Indirect method (temperature analysis)
diffusion formance shielded
Structural safety
Usability
Cracking State of cracking Visual method (crack measurement)
Shielding per-
formance
Rebar corro- Concrete structures in Visual method (hammering method)
Structural safety Rust fluid
sion general Local destructive method (chipping method)
Structural safety
Presence/absence Visual method (hammering method)
Spalling and Usability
of spalling and
falling Shielding per- Non-destructive method (infrared method, etc.)
falling
formance
Parts supporting heavy Deformation
Deflection Usability Visual method (deformation measurement)
equipment amount
Parts supporting vibration
Vibration Usability Vibration Non-destructive method (vibration measurement)
generating equipment
Vehicular/heavy object
Abrasion Usability Abrasion Visual method (surface measurement)
passage floor
Roof slabs, aboveground
Water leakage Usability and underground external Water leakage Visual method (hammering method)
walls, and internal walls

2) Periodical inspection for current soundness b) Steel-framed structures

evaluation
a) Concrete structures
Periodical inspections for current soundness evaluation
are repeated at periodic intervals focusing on degradation
phenomena due to aging. Periodical inspection methods
for current soundness evaluation are divided into direct
and indirect methods. Periodical inspections for current
soundness evaluation of a concrete structure are indi-
cated in Table 11.
Degradation phenomena due to aging in concrete
structures become apparent on concrete surfaces, except
for reduced concrete strength and moisture diffusion.
The visual method is suitable for quick inspection for
any damage that becomes apparent on a wide surface
area. It shall be used, as the standard inspection method
for concrete structures, to inspect any degradation phe-
nomenon that becomes apparent on a concrete surface,
such as cracking, rebar corrosion, spalling and falling,
deflection, abrasion, and water leakage. When a concrete
surface is protected with a coating or liner, a check shall
be made for any coating quality defect (discoloration,
reduced gloss, chalking, fouling, etc.) or any surface
damage (swelling, cracking, peeling, wear, etc.). If there
is no surface damage, the concrete surface need not be
directly checked.
Degradation phenomena due to aging in steel-framed
structures become apparent on their surfaces. Therefore,
the standard method of periodical inspections for current
soundness evaluation of steel-framed structures shall be
the visual method, similar to that for concrete structures.
Periodical inspections for current soundness evalua-
tion of a steel-framed structure are indicated in Table 12.
At the same time, for steel-framed structures, a check
shall also be made for surface rust and partial loss of
cross-section in addition to coating quality defect (dis-
coloration, reduced gloss, chalking, fouling, etc.) and
surface layer damage (swelling, cracking, peeling, wear,
etc.).
3) Periodical inspections for long-term soundness
evaluation
a) Concrete structures
Periodical inspections for long-term soundness evalua-
tion are intended to grasp the effect of each degradation
factor. Points with a marked effect shall be selected and
inspected. As with the direct methods in periodical in-
spections for current soundness evaluation, damage to
structures must be avoided. In some cases, however, the
use of a local destructive method may be necessary to
grasp the effect of each degradation factor. Periodical
inspections for long-term soundness evaluation of a
Y. Umeki, S. Sawada, S. Mitsugi, T. Maenaka and K. Takiguchi / Journal of Advanced Concrete Technology Vol. 14, 643-663, 2014 658

Table 12 Periodical inspections for current soundness evaluation (steel-framed structures).

Related performance
Degradation phenomenon Applicable part Inspection item Inspection method
characteristic
Parts (stacks, etc.) Coating quality defect
Reduced strength due to
Structural safety under cyclic loads Plating quality defect Visual method
fatigue
due to winds, etc. Excessive deformation
Steel-framed struc-
Structural safety Coating quality defect
Steel corrosion tures in general Visual method
Usability Plating quality defect
Foundation joints
Slipping, crack, and fracture of Parts (stacks, etc.) Visual method
Structural safety Coating quality defect
joint under cyclic loads (Hammering method
Usability Plating quality defect
Looseness and fracture of bolt due to winds, etc. and marking check)

Table 13 Periodical inspections for long-term soundness evaluation (concrete structures).

Degradation factor Applicable part Inspection item Inspection method
Heat (elevated
Parts heated to high temperature Temperature Indirect method
temperature)
(Environmental conditions
Parts subjected to long-term radiation
Radiation exposure Radiation dose survey-based method)
exposure
Carbonation Concrete structures in general Carbonation depth Local destructive method
Concrete structures’ external walls, etc.,
Chloride penetration Chloride content Local destructive method
exposed to airborne salt particles
Alkali-aggregate
Concrete structures in general Cracking Visual method
reaction
Turbine generator pedestals, emergency
Machine vibration diesel generator foundations, pump/motor Cracking Visual method
foundations, etc.

Table 14 Periodical inspections for long-term soundness evaluation (steel-framed structures).

Degradation factor Applicable part Inspection item Inspection method
Coating degradation Steel-framed structures in Coating quality defect Visual method
Plating degradation general Plating quality defect
(Loss of galvanized plating)
Cyclic loads due to winds, etc. Parts (stacks, etc.) under Coating quality defect Visual method
(fatigue) cyclic loads due to winds, Plating quality defect Non-destructive method
etc. Excessive deformation or (Ultrasonic testing (UT), liquid pene-
deflection trant testing (PT))

concrete structure are indicated in Table 13.
When the relevant part is difficult to access and inspect
by a direct method due to the problem of high altitude or
equipment layout, an indirect method shall be used.
b) Steel-framed structures
Periodical inspections for long-term soundness evalua-
tion of a steel-framed structure are indicated in Table 14.
In periodical inspections for long-term soundness
evaluation of steel-framed structures, points at which a
relatively large stress occurs and points that may be
affected by fatigue under cyclic loads due to winds, etc.
shall be checked by a visual method or a non-destructive
method.
4) Extra inspection
a) Concrete structures
Extra inspections of a concrete structure shall be per-
formed to check the structure in its entirety for any deg-
radation phenomenon for loads due to earthquake, ty-
phoon, etc., or to check around a fire-damaged point for
any degradation phenomenon. A degradation phenome-
non in a concrete structure often manifests itself as sur-
face damage and therefore can be grasped by a visual
method in extra inspections, similar to periodical in-
spections. If damage is detected by a visual method, a
more detailed inspection shall be performed by the
method shown in Table 15 to grasp effects on related
performance characteristics.
b) Steel-framed structures
Extra inspections of a steel-framed structure shall be
performed to check the structure in its entirety for any
degradation phenomenon for loads due to earthquake,
typhoon, etc., or to check around a fire-damaged point
for any degradation phenomenon. A degradation phe-
nomenon in a steel-framed structure often manifests
itself as surface damage and therefore can be grasped by
a visual method in extra inspections, similar to periodical
inspections. If any damage is detected by a visual method,
a more detailed inspection shall be performed by the
method shown in Table 16 to grasp the effects on the
related performance characteristics.
Y. Umeki, S. Sawada, S. Mitsugi, T. Maenaka and K. Takiguchi / Journal of Advanced Concrete Technology Vol. 14, 643-663, 2014 659

Table 15 Extra inspections (concrete structures).

Degradation Applicable Related performance Degradation phe-
Inspection item Inspection method
factor part characteristic nomenon

Loads due to Structural safety Cracking State of cracking Visual method

Damaged
earthquakes, Usability (Crack measurement)
point
typhoons, etc. Shielding performance Spalling and Presence/absence of spalling (Hammering method)
falling and falling
Local destructive method
Carbonation depth (Drilled-hole method)
(Core-sampling method)
Non-destructive method
Reduced concrete
(Rebound number
strength
method)
Change in compressive strength
Structural safety (Elastic wave method)
Heat due to Fire-damag
Usability Local destructive method
fire ed point
Shielding performance (Core-sampling method)
Visual method
Cracking State of cracking
(Crack measurement)
Spalling and Presence/absence of spalling Visual method
falling and falling (Hammering method)
Visual method
Explosive spalling Range of explosive spalling
(Hammering method)

Table 16 Extra inspections (steel-framed structures).

Degradation Applicable Affected performance Degradation phe-
Inspection item Inspection method
factor part characteristics nomenon
Yielding, buckling, Peeling/cracking of coating
crack, fracture, and Any peeling of mill scale
Loads due to excessive deforma- Any buckled, cracked, fractured, or
earthquakes, Damaged Structural safety tion of steel excessively deformed steel Visual
typhoons, point Usability Slipping, crack and Any slipped, cracked, or fractured Non-destructive
etc. fracture of joint, joint
and looseness and
Any loosened or broken bolt
fracture of bolt
Reduced strength Local
Stress-strain curve
due to heat destructive
Peeling/cracking of coating
Yielding, buckling,
Any peeling of mill scale
crack, fracture, and
Measurement of buckling, crack,
Heat due to Fire-damage Structural safety excessive deforma-
fracture, of excessive deformation of
fire d point Usability tion of steel Visual
steel
Non-destructive
Slipping, crack and Any slipped, cracked, or fractured
fracture of joint, joint
and looseness and Measurement of looseness or fracture
fracture of bolt of bolt

5) Inspection plan b. Evaluation of current soundness

An inspection plan consists of an inspection method,
timing of inspection, parts to be inspected and an in-
spection implementation scheme.
(2) Soundness evaluation
a. Outline
A structure soundness evaluation is an act of checking
the results of inspection and progression prediction
against predetermined evaluation criteria to ensure that a
structure’s level of performance necessary to remain
functional is maintained. If the results of the soundness
evaluation reveal that functions are not maintained or
may be compromised, an appropriate countermeasure(s)
shall be taken.
A structure’s performance changes as degradation phe-
nomena proceed. Accordingly, soundness evaluation
performed with focus on a degradation phenomenon
currently underway provides the basis for the structure’s
maintenance. Maintenance flow for current soundness
evaluation is indicated in Fig. 10.
In a current soundness evaluation, the results of peri-
odical and extra inspections for current soundness
evaluation shall be checked against a predetermined set
of evaluation criteria and shall be rated as A1 (Sound),
A2 (Follow-up observation), or A3 (To be discussed).
The rating shall be A1 (Sound) when the inspection
results meet the evaluation criteria, A2 (Follow-up ob-
servation) when degradation is apparent but the inspec-
Y. Umeki, S. Sawada, S. Mitsugi, T. Maenaka and K. Takiguchi / Journal of Advanced Concrete Technology Vol. 14, 643-663, 2014 660

tion results meet the evaluation criteria, and A3 (To be
discussed) when the inspection results fail to meet the
evaluation criteria.
The evaluation criteria shall be set based on, among
other things, existing guidelines, latest knowledge, and
evidential data supported by actual measurement results,
from the perspective of securing the level of performance
required to maintain the necessary function(s).
Evaluation ratings and evaluation criteria for current
soundness evaluation are indicated in Table 17. Evalua-
tion ratings and evaluation criteria for cracking are in-
dicated in Table 18.
Evaluation ratings are judged by the presence of deg-
radation phenomena. Especially, evaluation ratings and
evaluation criteria for cracking are based on degradation
factors.
c. Evaluation of long-term soundness
In a long-term soundness evaluation, the results of a
progression prediction shall be checked against a set of
predetermined evaluation criteria and rated as B1
(Sound), B2 (Follow-up required), or B3 (To be dis-
cussed). Maintenance flow for long-term soundness
evaluation is indicated in Fig. 11.
B1 (Sound) shall be given when the results of a pro-
gression prediction meet the evaluation criteria. B2
(Follow-up observation) shall be given when the results
of an inspection meet the evaluation criteria at the mo-
ment but are unlikely to meet the evaluation criteria
during the in-service period. B3 (To be discussed) shall
be given when the results of an inspection and the like at
the moment do not already meet the evaluation criteria.
Progression predictions can be made using inspection
results or existing formulae for prediction of progression.
As for degradation factors for which a formula for pre-
diction of progression is available, it is possible to predict
the progression of degradation during the in-service
period based on the applicable formula. It is also possible
to predict the tendency of progression of degradation
from the results of inspections performed in the past.
Formulae have been proposed for predicting progres-
sion for carbonation, chloride penetration, and cyclic
loads due to winds, etc. (fatigue). When there are nec-
essary data available, it is possible to make a progression

Periodical and extra inspections for current

soundness evaluation
Chapter 7

Current soundness evaluation

(Rate the inspection results with focus on degradation phenomena.)

A1 A2 A3
Sound Follow-up To be
required discussed

Consider a countermeasure(s) reflecting evaluation results.

Fig. 10 Maintenance flow for current soundness evaluation.

Periodical inspection for long-term soundness evaluation

Chapter 7

Long-term soundness evaluation

(Soundness evaluation by progression prediction on effects of degradation factors)
Prediction based on formula for Prediction based on inspection results
prediction of progression

B1 B2 B3
Sound Follow-up To be
required discussed

Consider a countermeasure(s) reflecting evaluation results.

Fig. 11 Maintenance flow for long-term soundness evaluation.

Y. Umeki, S. Sawada, S. Mitsugi, T. Maenaka and K. Takiguchi / Journal of Advanced Concrete Technology Vol. 14, 643-663, 2014 661

Table 17 Evaluation ratings and evaluation criteria for current soundness evaluation.
Evaluation rating and evaluation criteria
Degradation phenomena A2 (Follow-up obser-
A1 (Sound) A3 (To be discussed)
vation)
compressive concrete compressive concrete
Reduced concrete strength strength is equal to or － strength is below the design
above the design strength strength
Moisture diffusion a soundness evaluation
Cracking Refer to Table 18
No rebar corrosion Any rebar corrosion (rust
Rebar corrosion －
(rust fluid) fluid)
Spalling and falling No spalling and falling － Any spalling and falling
Concrete No deflection － Any deflection
structures Deflection deflection span ratio is less deflection span ratio is 1/300
－
than 1/300 or more
Vibration No vibration － Any vibration
No exposed coarse Any exposed coarse aggre-
Abrasion －
aggregate gate
No water leakage or water Any water leakage or water
Water leakage －
leak mark leak mark
Degradation phenomena due to
No effect of the fire － Any effect of the fire
fire
No damage on coating or Any damage on coating or
Reduced strength due to fatigue －
plating plating
Any damage (tiny rust
spots on all over the Any steel member is cor-
little rust or few coating
Steel corrosion surface , large rust roded and reduced in
plating quality defects
spots) on coating or cross-sectional area
plating
Steel-framed
Yielding, buckling, crack,
structures
fracture, or excessive
deformation of steel members No damage on the coating Any damage on the coating
－
Yielding, buckling, crack, or plating or plating
fracture, or excessive
deformation of steel members
Degradation phenomena due to
No effect of the fire － Any effect of the fire
fire

Table 18 Evaluation ratings and evaluation criteria for cracking.

Affected performance Evaluation rating and evaluation criteria
characteristic A1 (Sound) A2 (Follow-up observation) A3 (To be discussed)
No cracking likely to affect Cracking likely to affect structural
Structural safety −
structural safety safety
Crack width: Crack width: Crack width:
Below 0.3 mm (outdoor) 0.3 mm or more but less than 0.8 mm or more (outdoor)
Usability Below 0.4 mm (indoor) 0.8 mm (outdoor) 1.0 mm or more (indoor)
0.4 mm or more but less than
1.0 mm (indoor)
Crack-free coating*1 − Cracked coating*1
Crack width: Crack width: Crack width:
Liquid-tightness 0.05 mm or less*2 More than 0.05 mm but less 0.2 mm or more*2
than 0.2 mm*2
Shielding performance The evaluation ratings for usability shall apply with necessary modifications.
*1: When the usability (liquid-tightness) is evaluated based on the condition of coating
*2: When the usability (liquid-tightness) is evaluated based on the condition of concrete
Y. Umeki, S. Sawada, S. Mitsugi, T. Maenaka and K. Takiguchi / Journal of Advanced Concrete Technology Vol. 14, 643-663, 2014 662

Table 19 Evaluation criterion for Degradation factors.

Degradation factor Evaluation item Evaluation criterion
Heat In general parts, within 65°C. In local parts (piping penetration, etc.), within
Concrete temperature
(structural safety) 90°C.
Within 88°C when the heat is due to neutrons. Within 177°C when the heat is
Heat Highest internal temperature
due to gamma rays.
(shielding
Ambient environmental Within 71°C when the heat is due to neutrons. Within 149°C when the heat is
performance)
temperature due to gamma rays.
Radiation exposure 1 × 1020 n/cm2 or less for neutron exposure dose during in-service period.
Radiation exposure dose
(structural safety) 2 × 108 Gy (2 × 1010 rad) or less for gamma ray exposure dose.
Carbonation Carbonation depth During the in-service period, the carbonation depth has not increased so much
as to cause rebar corrosion.
Corrosion amount The rebar corrosion amount has not reached the level for cracking to occur.
Chloride The amount of chloride ions at rebars has not reached the level for rebar corro-
penetration Amount of chloride ions
sion to occur.

prediction. Note, however, that coefficients or the like
used in an existing formula for prediction of progression
are often formulated based on the results of accelerated
degradation of specimens. Hence, the results of a pre-
diction on an actual structure may differ from those of its
inspection. Therefore, a formula for predicting progres-
sion must be checked in advance for its applicability.
With a collation of inspection results accumulated
over years, even in the absence of a formula for predic-
tion of progression, it is possible to predict the tendency
of the effect due to the degradation factor. Generally,
degradation due to aging does not proceed rapidly.
Therefore, if the results of inspections performed in the
past reveal that no degradation phenomenon affecting
performance has occurred, the risk of rapid performance
degradation in future shall be estimated as low. Even if
there is a paucity of accumulated inspection results, this
prediction method will be effective if the future period to
be covered is relatively short.
Evaluation criteria for degradation factors are indi-
cated in Table 19. It is difficult to make a direct predic-
tion of the progression of any effect due to machine
vibration and alkali aggregate reaction. Therefore, such a
prediction shall be based on the crack occurrence situa-
tion. Accordingly, in a soundness evaluation for machine
vibration and alkali aggregate reaction, A1 (Sound) and
A3 (To be discussed), the evaluation ratings for cracks
affecting structural safety, shall be read as B1 (Sound)
and B3 (To be discussed), respectively.
In steel-framed structures, the condition of coating or
plating is checked by periodical inspections for current
soundness evaluation and mainly by the visual method.
The evaluation ratings for coating degradation and plat-
ing degradation (loss of galvanized plating) in the present
Guidelines shall be B1 (Sound) for damage-free coating
or plating and B3 (To be discussed) for any coat-
ing/plating quality defect. To ensure long-term preven-
tion of steel member corrosion, it is important to main-
tain the soundness of the coating or plating. The need for
periodic recoating shall be determined based on useful
life evaluation.
It is difficult to make a direct prediction of the pro-
gression of any effect due to cyclic loads due to winds,
etc. fatigue. Therefore, such a prediction shall be based
on the results of an inspection by the visual method on
the condition of the coating or plating and the pres-
ence/absence of excessive deformation. The evaluation
ratings shall be in line with current soundness evaluation.
The rating shall be B1 (Sound) for damage-free coating
or plating and shall be B3 (To be discussed) when any
coating/plating quality defect or any excessive deforma-
tion is detected by the visual method.
(3) Countermeasures and verification of effec-
tiveness
To maintain the functions required of nuclear facilities, it
is necessary to implement countermeasures based on the
results of current or long-term soundness evaluations so
that the current performance of the structures during the
in-service period will exceed the required level. Figure
12 shows the flow of maintenance countermeasure im-
plementation and effectiveness verifications.
Methods of implementing countermeasures shall in-
clude continuation of the current inspection, enhance-
ment of inspections, and repairs. Repairs shall include
countermeasures implemented alternatively, such as
consideration for members or structural components,
when repairs are unavailable, or when the required per-
formance level cannot be achieved by repairs. Any al-
ternative countermeasure implemented shall be checked
for its effectiveness.
(4) Records
For continuous implementation of maintenance, the
following items shall be recorded:
・ Maintenance plan
・ Contents and results of inspections
・ Methods and results of soundness evaluations
・ Contents and results of countermeasures and methods
and results of effectiveness verifications
To enhance the reliability and effectiveness of main-
tenance, it is necessary to record and store information on
maintenance while analyzing records and continuously
making improvements.
The results of inspections, soundness evaluations, and
countermeasures and effectiveness verifications shall be
Y. Umeki, S. Sawada, S. Mitsugi, T. Maenaka and K. Takiguchi / Journal of Advanced Concrete Technology Vol. 14, 643-663, 2014 663

Results of current or long-term soundness evaluation

A1 or B1 A2 or B2 A3 or B3
Sound Follow-up required To be discussed

Chapter 8

Consideration of methods of countermea-

sures based on evaluation results

Countermeasure implementation and

effectiveness verification

Recording and storage of information on maintenance activities

(A1 to A3 show the evaluation ratings for current soundness evaluation, while B1 to B3 show those for long-term soundness evaluation.)
Fig. 12 Flow of maintenance countermeasure implementation and effectiveness verification.

recorded accurately and in detail, using predetermined
forms, to allow easy understanding of maintenance ac-
tivities. Even if the methods of recording and storage are
changed due to improvements in the maintenance
method and technological advancements, arrangements
shall be made to allow continued use of past records.
When records are electronically stored, data leakage or
loss shall be borne in mind. As for data handling, it is
recommended to specify an appropriate control method
and implement measures for ensuring information secu-
rity.
6．Conclusion
Safe long-term use of nuclear facilities depends heavily
on proper design, construction, and maintenance.
There are an increasing number of nuclear power
plants in service for more than 30 years since the start of
operation. Accordingly, the importance of maintenance
of nuclear facilities is increasing, in addition to the tra-
ditionally recognized importance of their design and
construction. So far, however, the maintenance of
buildings at nuclear facilities has been left to the indi-
vidual operators in the absence of standardized criteria
and guidelines.
Under such circumstances, the previous version of the
present Guidelines was established in 2008 as a set of
guiding principles that set standards for maintenance
work to be performed to maintain the functions required
of buildings in nuclear power plants over a long period.
In addition, the Guidelines were revised to reflect the
latest knowledge in 2015. Since then, the present
Guidelines have been useful in a wide range of mainte-
nance activities from daily inspection to building
soundness evaluation with focus on aging.
This paper has outlined the Guidelines applied to nu-
clear facilities in Japan. An English version of the
Guidelines will be released by the Architectural Institute
of Japan.
Reference
AIJ, (2015). “Guidelines for maintenance and manage-
ment of structures in nuclear facilities.” Architectural
Institute of Japan, Tokyo. (in Japanese)
Journal of Advanced Concrete Technology Vol. 15, 1-9, January 2017 / Copyright © 2017 Japan Concrete Institute 1

Technical report

Effective Aging Management of NPP Concrete Structures

Julia Tcherner1*, Lingam Vaithilingam2 and Ming Han3

Received 4 August 2016, accepted 19 December 2016 doi:10.3151/jact.15.1

Abstract
Nuclear concrete structures are typically passive under normal operating conditions. They are, however, play a key role
in mitigating impacts of extreme/abnormal operating and environmental events. As structures age, changes in material
properties arise from continuing microstructural changes and environmental influences. Changes in environmental con-
ditions anticipated during the original design can sometimes lead to unpredicted effects, which may jeopardize the in-
tegrity of the structures.
In order to continue successful operation, continuous integrity of the safety-related concrete structures shall be ensured
by controlling and mitigating aging related degradation. A key element of aging management is the systematic and rig-
orous assessment of structures most commonly referred to as Condition Assessment (CA). Over the years, SNC-Lavalin
personnel have assessed containment structures and safety related structures of Nuclear Power Plants (NPPs), research
reactor and waste storage facilities.
This paper discusses requirements and approach to aging management for existing as well as new NPP concrete struc-
tures, and presents the methodology used to perform the CA.

1. Introduction life is then based on both this current condition and a

Nuclear concrete structures have history of reliable per-
formances. Structures are subject to time-dependent
changes that may impact their ability to withstand vari-
ous demands from operation and the environment
thereby reducing design margins. In accident conditions,
excessive degradation can lead to failure that often af-
fects serviceability of concrete structure. Some compo-
nents of nuclear concrete structures may be difficult to
inspect and repair. Typically concrete structures are not
replaceable except for some of their components (e.g.
joint sealants). As concrete structures age, assurances
need to be provided that the capability of the safety-
related systems to mitigate extreme events has not dete-
riorated unacceptably.
A key element of aging management is the systematic
and rigorous assessment of structures most commonly
referred to as Condition Assessment (CA). CA generally
involves a review of facility data to assess the effect of
age related degradation on safety-related structures,
establish their current condition, and provide prognosis
of future performance. The assessment of the current
condition of the concrete structures is based on the his-
tory of operation and maintenance assessed against the
design basis and the functional, safety, as well as opera-
tional requirements. The health prognosis for extended
1
Senior Civil Engineer, Civil Design and Analysis,
SNC-Lavalin, Canada. *Corresponding author,
E-mail: Julia.Tcherner@snclavalin.com
2
Senior Section Head, Civil Design and Analysis, SNC-
Lavalin, Canada.
3
Director, Civil and Project Engineering, SNC-Lavalin,
Canada.
systematic identification and assessment of Aging Re-
lated Degradation Mechanisms (ARDMs) and their im-
pact on the functional requirements of concrete struc-
tures.
Many assessments were performed by SNC Lavalin
personnel over the years that included concrete,
prestressing system, non-metallic liner system and joint
sealant material. Assessed structures were of different
age. The oldest one being assessed after more than 50
years of service is a nuclear research facility that went
into operation in the mid-1950s.
2. Requirements and guidelines for aging
management
In 2014 Canadian Nuclear Safety Commission (CNSC)
has published the document for fitness for service and
aging management (CNSC 2014) that provides require-
ments for managing aging of Structures, Systems and
Components (SSC) of a power reactor facility. It pro-
vides guidance on how these requirements are to be met
and is intended as a framework within which codes and
standards are to be applied. The document references
CSA N287.7 (2013a) Standard and N291 (2013b) Stan-
dard that provide requirements for in-service examina-
tion of concrete containment structures and safety re-
lated structures respectively. It also references CSA
N287.1 (1993) that provides general requirements for
concrete containment structures. It should be noted that
new revisions of CSA N291 and N287.1 have been is-
sued recently.
Furthermore, the new standard on aging management
of concrete containment structures has been developed
as described in Tcherner (2015a). The CSA N287.8
J. Tcherner, L. Vaithilingam and M. Han / Journal of Advanced Concrete Technology Vol. 15, 1-9, 2017
Standard (2015) was issued in 2015. The standard pro-
vides aging management requirements to ensure that
concrete containment structures satisfy their functional
and performance requirements in all different phases of
their life-cycle including design, construction, commis-
sioning, operation, and decommissioning. Moreover, the
requirements provided in this standard are intended to
ensure that for new concrete containment structures,
aging is addressed during design and construction
phases.
The International Generic Aging Lessons Learned
(IGALL) project was initiated by IAEA a few years ago
in order to provide detailed information on specific pro-
grams to manage aging and degradation of SSCs. This
initiative followed the United States Nuclear Regulatory
Commission (USNRC) Generic Aging Lessons Learned
(GALL) activity (USNRC 2010) but was extended to
summarize international practices for aging manage-
ment of mechanical, electrical and civil SSCs of NPPs.
Phase 2 of the project completed in 2015 provides
collection of proven aging management programs and
Time Limited Aging Analysis (TLAAs). Phase 3 of the
project is currently underway. The aging management
Fig. 1 Systematic Approach to Managing Aging of a Structure or Component (IAEA 2016)
2
programs for concrete structures refer to IAEA TEC-
DOC 1025 (1998) issued in 1998 to address assessment
and aging management of concrete containment struc-
tures. The TECDOC was revised recently as discussed
in Moore et al. (2015) to include state of the art infor-
mation regarding aging management of concrete struc-
tures. The document was also expanded to address other
NPP concrete structures besides containment. The re-
vised document (IAEA 2016) was issued in 2016.
3. Effective aging management of existing
structures
The documents listed in Section 2 require a systematic
approach to aging management to be implemented as
illustrated in Fig. 1. As can be seen from this figure,
understanding aging lies in the heart of the effective
aging management.
Other components of effective aging management in-
clude development of the Aging Management (AM)
plan, operating to minimize expected degradation, per-
forming inspections and monitoring activities to check
for degradation, and implementing mitigative measures.
J. Tcherner, L. Vaithilingam and M. Han / Journal of Advanced Concrete Technology Vol. 15, 1-9, 2017
The lessons learned in implementation of AM plan
along with the applicable Operating Experience (OPEX)
and Research and Development (R&D) activities should
be considered for optimisation of the AM plan. The ma-
jor components of effective aging management as it
pertains to existing nuclear concrete structures are dis-
cussed in this Section.
3.1 Understanding aging and planning aging
management plan
Prior to the development of an AM plan, evaluation for
aging management and condition assessment shall be
performed in order to ensure understanding of aging and
to establish base line condition of the structure. Re-
quirements for condition assessment and evaluation for
aging management are defined by CSA N287.8 (CSA
2015).
3.1.1 Methodology for evaluation for aging
management and condition assessment
As stated in IAEA Safety Guide NS-G-2.12 (IAEA
2009), an aging management review (or an evaluation
for aging management) forms an integral part of a CA.
Thus, the condition assessment and evaluation for aging
management are typically performed in combination
and constitute Life Assessment (although, more com-
1
Fig. 2 Schematic Condition Assessment Process (Tcherner and Aziz 2013)
1
The term Condition Assessment, as more commonly used, refers to Life Assessment as well.
3
monly the process is referred to as Condition Assess-
ment). Figure 2 illustrates schematics of the process.
The methodology was specifically tailored to civil
structures and includes the following:
• Gathering and reviewing of design, manufacturing,
construction, commissioning, operation, inspection,
and maintenance history, and any other relevant
documentation;
• Defining the design basis and identifying any changes
made to the structure during construction and opera-
tion;
• Establishing the structure's physical and functional
boundaries and associated components;
• Undertaking an aging assessment of structures or
components remaining after screening. This includes
an assessment of ARDMs, aging evaluation, and es-
tablishing the structure’s health prognosis; and
• Providing recommendations for a structure in terms of
inspections, monitoring, repair, and maintenance ac-
tivities to ensure plant safety and production goals
over the plant life.
In addition to current condition and future health pre-
dictions, changes necessary to address issues related to
aging effects are identified as a result of CA; economic
improvement opportunities might also be included.
J. Tcherner, L. Vaithilingam and M. Han / Journal of Advanced Concrete Technology Vol. 15, 1-9, 2017 4

The comprehensive data is required to ensure the to prevent forced outages, etc.;
depth and quality of the CA. Since the results of CA are • Medium means the degradation mechanism is known
largely based on the engineering judgement, qualified for this structure or component, either at this facility
and experienced personnel shall be performing the CAs. or a similar one, and is being managed or mitigated;
Engineering judgment comprises of the knowledge of and
the design and functional requirements of the structure • Low means the mechanism is possible for the struc-
being assessed as well as the understanding of how the ture, but is easily managed with the current programs
stressors can influence the onset and development of or has no impact on achieving target life.
degradation mechanisms and how these mechanisms ARDMs not included in the matrix, or those having
may affect ability of the structure to meet its functional no marks against them, are considered unlikely for the
requirements. While there are standards and guidelines structure given the environmental conditions or materi-
documents available to evaluate influence of various als or a combination thereof and are not considered fur-
parameters on performance of structures e.g. (CSA ther.
2015; IAEA 2016), the value of experience should not Conclusions and health prognoses are provided based
be underestimated. on a thorough review of available information with par-
Interviews with the station staff and contractors in- ticular attention to aging related degradation. Where
volved in design, construction, commissioning and op- sufficient information to provide a health prognosis is
eration of the structure as well as the walkdowns are not available, activities to support the CA results are
required in order to identify information not formally recommended. Additionally, recommendations related
documented and locations of deterioration of the struc- to existing plant programs are provided to ensure con-
ture. tinuing structure health.
Based on a review of available information, design AM plan can then be developed based on the under-
and operational exposure conditions of a structure are standing of the design, construction, material properties,
also compared and engineering judgement is used to operating conditions, aging mechanisms and stressors,
document all potential aging mechanisms that might consequences of degradation, inspection, operation and
influence performance of the structure being assessed. maintenance history and taking into consideration rec-
Evaluation process shown in Fig. 3 is used to assess and ommendations for inspection, monitoring and mitigative
prioritize ARDMs. An ARDM matrix is then created to activates. Refer to CSA N287.8 (2015) for contents of
record evaluation results. A structure’s susceptibility to an effective AM plan.
each ARDM is evaluated as high, medium, or low,
based on the following criteria: 3.1.2 Implementation example
• High means the degradation mechanism is occurring Table 1 contains example of ARDM matrix created for
or has occurred, either in this structure or in a similar a safety related reinforced concrete structure with a non-
structure under similar conditions, and that steps have metallic liner containing demineralised water and ex-
not been taken, or it is unclear if any such steps are posed to elevated temperatures and radiation. Based on
adequate to mitigate degradation for the target life or materials and environment of exposure for the structure,

Table 1 Example of Ageing Related Degradation Mechanism Matrix.

Crystallization of Chlorides & Other Salts

Elevated Temperature/Thermal Cycling

Abrasion/Erosion/Cavitation
Alkali-Aggregate Reactions
Leaching and Efflorescence

Loss of Prestressing Force

Stress Corrosion Cracking

Hydrogen Embrittlement
Freeze-Thaw Attack

General Corrosion
Fatigue/Vibration
Acid/Base Attack

Creep/Shrinkage
Sulphate Attack

Delamination
Carbonation

Weathering
Settlement
Irradiation

Wear

Concrete M M L M
Reinforcing Steel M
Embedded Parts M
Joint Sealant M M M M
Non-metallic Liner M L L L M
J. Tcherner, L. Vaithilingam and M. Han / Journal of Advanced Concrete Technology Vol. 15, 1-9, 2017 5

the following ARDMs were considered potentially rele- gate Reaction (AAR), elevated temperatures/thermal
vant: cycling, and shrinkage;
• For concrete – leaching/efflorescence, Alkali Aggre- • For reinforcing steel – corrosion;

Fig. 3 Assessment of ARDMs (Tcherner and Aziz 2013).

J. Tcherner, L. Vaithilingam and M. Han / Journal of Advanced Concrete Technology Vol. 15, 1-9, 2017
• For embedded parts – corrosion;
• For joint sealant – elevated temperature/thermal cy-
cling, irradiation, weathering, and wear.
• For non-metallic liner - elevated temperature/thermal
cycling, irradiation, weathering, and wear.
Each ARDM listed above was systematically evalu-
ated following the process illustrated in Fig. 3. For ex-
ample, evidences of efflorescence and leaching were
observed on the outside (dry) surface of the structure.
There was no program in place to detect and manage
leaching. Leaching increases porosity and the perme-
ability of concrete, thus lowering its strength and mak-
ing concrete more vulnerable to aggressive environ-
ments. However, since there were no evidences of a
widespread degradation, this was considered a local
phenomenon that was not likely to affect facility target
life. Therefore, medium ranking was given. It should be
noted however that as part of the life assessment, rec-
ommendations are provided including those related to
enhancements to the facility’s aging management pro-
gram. Thus, although it is stated that the mechanism is
not expected to affect facility’s target life, this may
change in the long term as local phenomenon may be-
come a widespread phenomenon if aging management
is not implemented.
As far as corrosion of reinforcement and embedded
parts is concerned, there were no evidences of corrosion
damage observed during inspection of the structure. It is
however considered possible given the material and
environment of exposure including exposure of concrete
to demineralised water (discontinuities in the liner were
evidenced by observed efflorescence) and elevated tem-
peratures. Leaching may cause local reduction in alka-
linity of concrete that might lead to corrosion of rein-
forcement and embedded parts. Furthermore, elevated
temperature tends to accelerate the rate of corrosion
reaction. Corrosion has a potential to impact facility’s
target life, therefore medium ranking is given.
3.2 Minimizing expected degradation
In order to minimize expected degradation, the opera-
tors should understand the importance of aging man-
agement and be involved in the implementation of AM
plan. Operating conditions that affect aging assumptions
should be monitored and trended to ensure that the as-
sumptions made during design are valid. Any excursions
or events causing the change in operating conditions
should be evaluated and their effects on aging of con-
crete structures should be assessed.
Aging must be considered when implementing any
change in operations or modifications to the structure or
its components.
3.3 Checking the degradation
Periodic visual inspections are typically used to ensure
adequate performance of the structures. However, nu-
clear concrete structures often have limited accessibility
for visual inspection. In addition, the internal degrada-
6
tion may not necessarily be manifesting itself on the
surface of the structure. Therefore, monitoring instru-
mentation is often relied upon to provide assurance with
the integrity of the structure and to confirm its predicted
response to the applied loading (e.g. during containment
leakage rate test).
Monitoring of environmental and operating condition
enables assessing and trending the degradation stressors.
While it is imperative for inaccessible structures to en-
sure that design assumptions are valid, it has also
proven to be useful to monitor environment of exposure
of accessible structures particularly in between periodic
inspections. The change in exposure environment might
facilitate existing degradation mechanisms and may
cause origination of the new ones, i.e. the aging man-
agement activities may need to be refocused on other
mechanisms than those identified as per the original
condition assessment. Depending on the structure, its
location, geometry, function and possible degradation
mechanisms, monitoring of the following parameters
has proven to be simple (as far as monitoring instrumen-
tation involved) and informative: ambient and operation
temperature and humidity, chemistry of the ground wa-
ter and soil, elevation of the ground water table and its
fluctuations, monitoring chemistry, temperature, level of
the liquid inside the structure (for liquid retaining struc-
tures) and the amount of a make-up water.
As aging is a dynamic process, it is important to trend
performance of the structure in order to establish the
rate of degradation. Experience shows that in order to
track performance of the structure, condition indicators
– a characteristic of the structure that may infer or di-
rectly indicate the current and future ability of the struc-
ture to function within acceptance criteria. Examples of
condition indicators include cracks, spalls, delaminations,
deflections, ingress of harmful substances (e.g. chlorides),
depth of carbonation, leak rate, prestressing force).
Furthermore, environmental and operating conditions
(e.g. temperature, humidity, radiation, chemistry of the
water and soil that the structure is exposed to) should be
trended and any excursions from the parameters consid-
ered in the design should be noted to establish the rela-
tionship between degradation and associated stressors.
Predictions for future performance of the structure may
then be made based on the analysis of past performance,
the rate of degradation and the changes in associated
stressors. Knowledge of the rate of degradation and in-
fluencing parameters will also allow selecting appropri-
ate mitigative measures if necessary.
3.4 Mitigating degradation
Maintenance activities may be required to mitigate deg-
radation. These might include preventative maintenance,
corrective maintenance, spare parts management, and
replacement. Typically, corrective maintenance is used
for nuclear concrete structures.
OPEX shows that majority of the issues with concrete
structures are either directly caused by or facilitated by
J. Tcherner, L. Vaithilingam and M. Han / Journal of Advanced Concrete Technology Vol. 15, 1-9, 2017
the inadequate design and construction practices (refer
to Fig. 4). The design challenges include:
- Inadequate structural design;
- Poor detailing;
- Improper material selections; and
- Inadequate specifications.
Construction issues include:
- Not following specifications/ procedures;
- Lack of attention to details; and
- Poor QC / QA.
This should be considered when designing and im-
plementing repairs to mitigate the effects of degradation.
Specifically, the importance of adequate technical speci-
fication cannot be overemphasized. Although the indus-
try is moving towards performance based specification,
specifications for repair of nuclear concrete structures
tend to contain some prescriptive requirements as well
to ensure that specific restrictions as required by the
standards or dictated by the OPEX are incorporated.
Figure 5 illustrates the list of necessary inputs and out-
puts of the specification including both performance
based and prescriptive requirements.
Inadequate repairs may facilitate degradation of the
structure and are often impractical to mitigate. There-
fore in order to ensure satisfactory and long-lasting re-
pair, the importance should be placed on the pre-
qualification tests as opposed to the verification tests.
Materials qualified by tests and satisfactory field per-
Fig. 4 Where Problems Arise (adapted from Mailvaganam (1992)).
Fig. 5 Technical Specification for Repair – Performance Based and Prescription Based (Tcherner 2015b).
7
formance in similar application and environment should
be specified. The pre-qualification tests should be per-
formed to ensure proper application of the specified
material in the anticipated conditions. Mock-up tests
should be considered for first-of-a-kind repair activities.
Additional information related to waterproofing of nu-
clear concrete structures is provided (Tcherner 2015b).
3.5 Optimising aging management
As illustrated in Fig. 6, the inputs for optimisation of the
AM plan include the lessons learned from implementa-
tion of the AM Plan including operating, inspection and
mitigation activities.
Repeating evaluation for aging management and con-
dition assessment would determine if there are any
changes in the condition of the structure that may neces-
sitate changes to operation, inspection/monitoring and
maintenance activities. It would also help ensuring that
operating and environmental conditions that were ac-
counted for in design have not changed. As aging is a
dynamic process, the benefits of trending the stressors
as well as the performance of the structures to improve
understanding of aging as discussed in Section 3.3 can-
not be overemphasised. This information needs to be
addressed when optimising AM plan.
Furthermore, it is important to consider applicable
OPEX and R&D activities thus help addressing any
“surprises” that tend to manifest themselves more often
J. Tcherner, L. Vaithilingam and M. Han / Journal of Advanced Concrete Technology Vol. 15, 1-9, 2017
Fig. 6 Development and Optimization of Aging Management Plan (Tcherner 2015a).
as the knowledge and experience in NPP operation and
nuclear concrete structures’ performance grows.
Thus, optimising includes improvements/revisions to
AM Plan addressing the following:
• Evaluation of changes in performance of the structure
via monitoring and trending of condition indicators;
• Evaluation of changes in the environmental and oper-
ating conditions via monitoring and trending of the
stressors and comparing them to those considered in
the design;
• Evaluation for aging management and condition as-
sessment (i.e. life assessment);
• Relevant OPEX including OPEX with implementa-
tion of AM Plan as well as OPEX with performance
of other structures subjected to similar environmental
and operating conditions; and
• Relevant R&D.
4. Conclusions
OPEX indicates that concrete structures have a history
of reliability and durability, but there have been occur-
rences of degradation. As OPEX increases, more “sur-
prises” are to be expected. Degree of degradation de-
pends on quality of the structure as constructed and the
aggressiveness of environment. Analysis and design
assumptions should be confirmed during the life of the
structure to account for changes in operating environ-
ment and material properties. If changes are identified
(such as higher operating temperatures, greater loading),
8
their effect should be evaluated and analyses need to be
updated so that the structures can perform their func-
tions safely. Material properties used for analysis should
consider aging of the structure. Specifically for concrete
structures, samples should be obtained if possible and
tested to determine actual properties that account for
aging. Concrete, unless subjected to degradation is
known to gain strength with age. This should be con-
firmed by physical testing and should be accounted for
when analysis of mature structures are performed.
An evaluation for aging management and condition
assessment for nuclear safety-related structures per-
formed as part of periodic safety review, in preparation
of life extensions, to establish a baseline condition for
periodic inspection, prior to modification of the struc-
ture, or for other purposes provide an effective tool in
assessing aging of the structures, evaluating their cur-
rent condition, providing a health prognosis, recom-
mending remedial measures and identifying opportuni-
ties for improvements including design, operation and
aging management.
It is important to consider instrumentation as well as
ability to inspect and monitor the structures during de-
sign and construction. Concrete containment structures
particularly those with grouted post-tensioning system
should be monitored to ensure that they will continue
meeting their functional requirements and have ade-
quate margins available during the life of the plant (in-
cluding life extensions). Monitoring of the stressors
(environmental and operating conditions) and perform-
J. Tcherner, L. Vaithilingam and M. Han / Journal of Advanced Concrete Technology Vol. 15, 1-9, 2017
ance of the structure is paramount.
In spite of recently issued documents providing re-
quirements and guidance for aging management of nu-
clear concrete structures, the decisions are largely based
on engineering judgement necessitating involvement of
knowledgeable and experienced personnel.
References
CNSC, (2014). “Fitness for service – Aging
management.” REGDOC 2.6.3, Canadian Nuclear
Safety Commission.
CSA, (1993). “General requirements for concrete
containment structures for Candu nuclear power
plants.” N287.1, Canadian Standards Association.
CSA, (2013a). “In-service examination and testing
requirements for concrete containment structures for
nuclear power plants.” N287.7-08, Canadian
Standards Association.
CSA, (2013b). “Requirements for safety-related
structures for Candu nuclear power plants.” N291-08,
Canadian Standards Association.
CSA, (2015). “Aging management for concrete
containment structures for nuclear power plants.”
N287.8-15, Canadian Standards Association.
IAEA, (1998). “Assessment and management of aging
of major nuclear power plant components important
to safety: Concrete containment buildings.”
9
TECDOC-1025, The International Atomic Energy
Agency.
IAEA (2009). “Aging management for nuclear power
plants.” IAEA Safety Standards Series Safety Guide,
NS-G-2.12, Vienna.
IAEA (2016). “Aging management of concrete
structures in nuclear power plants.” IAEA NP-T-3.5,
The International Atomic Energy Agency.
Mailvaganam, N. P., (1992). “Repair and protection of
concrete structures.” CRC Press.
Moore, J. et al., (2015). “Aging management of
concrete structures in nuclear power plants.” In: Proc.
SMiRT 23, Manchester, UK.
Tcherner, J. and Aziz, T., (2013). “Candu energy
experience in condition assessment of nuclear safety
related concrete structures.” In: Proc. SMiRT 22, San
Francisco, USA.
Tcherner, J., (2015a). “Aging management of concrete
containment structure – Standard development.” In:
Proc. SMiRT 23, Manchester.
Tcherner J., (2015b). “Waterproofing of nuclear
concrete structures.” In: Proc. SMiRT 23, Manchester,
UK.
USNRC, (2010). “Generic aging lessons learned
(GALL) report.” NUREG-1801 Rev 2, United States
Nuclear Regulatory Commission.
Journal of Advanced Concrete Technology Vol. 14, 590-606 September 2016 / Copyright © 2016 Japan Concrete Institute 590

Scientific paper

Maintenance Management of Turbine Generator Foundation Affected by

Alkali–Silica Reaction
Ryosuke Manabe1*, Hiroshi Kawae2, Kiyoshi Ogawa3 and Masayuki Matsuura4

Received 31 March 2016, accepted 16 September 2016 doi:10.3151/jact.14.590

Abstract
A change in distance between the generator rotor axis and the bearing of the table deck in the turbine generator (TG)
foundation of Unit #1 of the Ikata nuclear power plant was observed during an inspection in 1979 after the start of op-
eration. This led to measurements of the structure expansion (from 1981), the cracking condition (from 1982), and core
sampling of the concrete member (from 1986), based on which deformation and cracking was determined to be the result
of an alkali–silica reaction (ASR). However, as the strength of the TG foundation was still in excess of its design value,
ASR was not considered to have affected its ability to support equipment. Subsequent analysis of the structural stability of
the TG foundation found no change in deformation or concrete strength over a period of continued monitoring, and hence,
the TG foundation has been used up to the present time. The aim of this paper is to establish a maintenance management
method for similar structures exhibiting cracks caused by ASR.

1. Introduction 2. Overview of turbine generator foundation

Concrete degradation by an alkali–silica reaction (ASR)
has been reported in recent years at nuclear power plants
such as the Gentilly facility in Canada and Seabrook
facility in the United States (Chénier et al. 2012). In
Japan, the rupture of reinforcing bars by expansive forces
has also been reported in outdoor structures affected by
ASR, such as piers and abutments of bridges (Minato and
Torii 2010). As long-term-service is essential to major
infrastructure, there has been an increasing need for
maintenance management plans for structures affected
by ASR.
This paper discusses the maintenance management
method that has been implemented for the turbine gen-
erator (TG) foundation of Unit #1 of the Ikata nuclear
power plant, in which concrete cracks caused by ASR
have been observed. This approach includes an evalua-
tion of the current and long-term soundness of the
structure, and is aimed at establishing a method that can
be used to manage the cracks caused by ASR in any
major structure.
The TG foundation of Unit #1 was constructed in 1975 to
provide a supporting structure for the turbine and gen-
erator, and has been in commercial operation since 1977.
This reinforced concrete (RC) structure consists of a
table deck, columns, walls and a foundation mat meas-
uring 50 × 15 × 18.7 m in length, width, and height,
respectively. The dimensions of this structure and the
materials used in its construction are shown in Fig. 1.
Part of the top surface of the table deck is covered by
plastic tiles, while the rest of the concrete surface is
exposed.
As the TG foundation is located within the turbine
building, it is not exposed to direct sunlight, wind, or rain.
Furthermore, no moisture is supplied by any of the other
equipment in the building. However, the table deck is in
an atmosphere that is affected by the heat of the turbine.
From environmental measurements taken in 2003–2011,
it is known that the annual average temperature and
relative humidity around the TG foundation during op-
eration is approximately 35°C and 30%, respectively.

3. Maintenance management
1
Construction Project Group, Civil & Architectural
3.1 Concept
Engineering Dept., Shikoku Electric Power Co., Inc.,
The current condition of the TG foundation was assessed
Japan. *Corresponding author,
(to understand the extent of ASR) through measurements
E-mail: manabe16561@yonden.co.jp
2 of its deformation, the mechanical properties of the RC
Subleader of Construction Project Group, Civil &
member, and a structural analysis. These results have
Architectural Engineering Dept., Shikoku Electric Power
suggested that its support function can be maintained;
Co., Inc., Japan.
3 however, a long-term evaluation has also been imple-
Leader of Construction Project Group, Civil &
mented that will continue these measurements so as to
Architectural Engineering Dept., Shikoku Electric Power
predict the future should any change in the ASR process
Co., Inc., Japan.
4 occur (Table 1, Fig. 2).
Leader of Equipment and technology Group (PLM),
Nuclear Power Dept., Shikoku Electric Power Co., Inc.,
Japan.
R. Manabe, H. Kawae, K. Ogawa and M. Matsuura / Journal of Advanced Concrete Technology Vol. 14, 590-606, 2016 591

Table 1 Classification of TG foundation maintenance items.

Measurement for understand the extent of ASR Structural analytical
Purpose
Deformation Mechanical properties evaluation
･The measurement of the expansion of
the table deck
Concrete core sampling tests
･Visual inspection of concrete cracks
･Compressive strength and
･The measurement of the internal tem-
static elastic modulus
perature of the concrete
The current ･Reactivity of the aggregate
･The measurement of reinforcing bar
soundness (Observation of cut surface ･Structural analysis
strain
evaluation of concrete core)
The soundness ･The measurement of the axis vibration
･Accelerated expansion test
evaluation level
･The rebound hardness meas-
･The column inclination measurement
urement of concrete
･The investigation for reinforcing bar
rupture
The ･Structural analysis
･Accelerated expansion test
long-term based on an assump-
･Continuation of the above measurement ･The rebound hardness meas-
soundness tion of ASR recom-
urement of concrete
evaluation mencing

3.2 History
A change in the distance between the generator rotor axis
and bearing was discovered in the table deck during an
equipment overhaul in 1979, but no abnormality was
found in the axis vibration and the deformation was able
to be corrected by adjusting the equipment. Cracks later
appeared on the lateral surface of the beam of the table
deck in 1982, and it is thought that these may have been
caused by deformation of the concrete resulting in the
change in the distance between the generator rotor axis
and bearing. Observation of these cracks and measure-
ment of the deformation of the table deck therefore
commenced. In order to clarify this, strain gauges were
installed on the TG foundation in 1986 to measure the
strain in the reinforcing bars. Concrete core samples
were also taken and subjected to compressive strength
testing. These studies concluded that the deformation of
the TG foundation was caused by ASR based on the

Sea Side
50

Low-Pressure Steam
11
Axle High-Pressure Housing

Generator
2.5
No.2 Low-
Pressure Turbine
No.1 Low-
4.0 Pressure Turbine
3.0 15
2.5 High-Pressure High-Pressure
3.0 Turbine Steam
2.5 Foundation Table
5.0 Deck
2.5
4.76
2.5 18.7
Mountain Side
Reinforcing bar
・ Deformed bar SD345 (Former JIS SD 35) Unit : m
Concrete
・ Specified Strength 2
: 20.6 N/mm (210 kg/cm )
2 ・ Length : 50.0 m
・ Cement : moderate heat portland cement ・ Width (Turbine Side) : 15.0 m
・ Fine Aggregate : washed sea sand ・ Width (Generator Side) : 11.0 m
: Ehime-Pref. Imabari-City ・ Height : 18.7 m
・ Coarse Aggregate : crashed stone (from top of foundation mat)
: Yamaguchi-Pref. Yanai-City Andesite
Fig. 1 Schematic of turbine generators and their foundation.
R. Manabe, H. Kawae, K. Ogawa and M. Matsuura / Journal of Advanced Concrete Technology Vol. 14, 590-606, 2016 592

developing table deck expansion, distribution of concrete might affect its operation. Furthermore, the compressive
cracks with directionality along the main reinforcement strength of the concrete has remained in excess of its
direction, a decrease in the static elastic modulus, and the design strength and ongoing structural analysis com-
fact that andesite containing silicate mineral was used as menced in 1988 showed enough supporting function.
a coarse aggregate. A gelled substance extracted from the As ASR is a progressive process, it is necessary to
surface of the core samples was identified as alkaline evaluate the long-term soundness of the TG foundation
silica gel, with the alkali content of the concrete as much based on the measurements that have been taken up to the
as 9 kg/m3. Despite this, it was concluded that there was present time. In 2002, the development of ASR was
no problem in terms of the TG foundation providing verified to have settled based on the fact that table deck
sufficient support, as there was no induced vibration that expansion stopped around 1991, no new cracks indica-

Cracks appeared
on the lateral surface
of the beam of table deck
(1982)
Start operation judging that ASR settled
(1977) (2002) year

the change bearing judging that ASR Operation stop

portions gap was found (1987) (2011)
(1979)

Verified that there is no problem in supporting function

Measurement for understand the extent of ASR Measurement for understand the extent of ASR

The measurement of the axis vibration level (start in 1977)

The measurement of the expansion of the table deck (start in 1982)

Visual inspection of concrete cracks (start in 1982)

(Deformation)

The measurement of reinforcing bar strain (start in 1986)

The measurement of the internal temperature of the concrete (start in 1986)

The column inclination measurement (start in 2008)

The investigation for reinforcing bar rupture (start in 2015)

Concrete core sampling tests (start in 1986 by proper frequency)

（Mechanical properties of the RC member）

(1986) (1989) (1992) (2002)

･The compressive strength
･The static elastic modulus
･Reactivity of the aggregate
(Observation of cut surface of concrete core)
･Accelerated expansion test
･The alkali content of the concrete

The rebound hardness measurement of concrete (start in 2008)

(1988) (2005)
structuralanalyses

The two dimensional The three dimensional

frame model frame model
(1998)
The two dimensional
FEM
Fig. 2 History of maintenance management items.
R. Manabe, H. Kawae, K. Ogawa and M. Matsuura / Journal of Advanced Concrete Technology Vol. 14, 590-606, 2016 593

tive of ASR were observed, and that the rate of remaining to understand the current ASR conditions and determine
expansion had been decreasing since 1989. The structure whether they had settled, as well as the results of a sub-
was therefore considered to have a low possibility of sequent structural analysis. Chapter 5 outlines the state
expansion in the future (Takakura and Ishikawa et al. monitoring that has been continued or added since de-
2005). Also in 2005, structural analysis by a termining that the ASR conditions had settled.
three-dimensional frame model of the TG foundation
deformation that took into account the mechanical 4. Soundness evaluation
properties of its concrete verified that its supporting
function would be maintained. Structural analysis as- 4.1 Analysis of ASR condition
suming redevelopment of ASR was carried out to show 4.1.1 Table deck expansion
enough supporting function in future. Since then, no Expansion of the table deck was measured to determine
major changes in the deformation conditions or com- the extent of any deformation of the TG foundation after
pressive strength of the concrete that would indicate a measuring the change of the distance between the gen-
change in ASR conditions have been observed through erator rotor axis and its bearing. Ten benchmarks were
ongoing column inclination measurements, rebound set for this measurement on both the mountain side and
hardness tests and reinforcing bar rupture investigations. sea side of the TG foundation, as shown in Fig. 3.
As a result of this verification of the current and Starting in December 1981, the distances between these
long-term soundness of the TG foundation, it has re- benchmarks were measured using a micrometer and steel
mained in continuous use up to the present time. rod, and the total of these distances was used to deter-
Chapter 4 presents the main measurement results used mine the displacement. As shown in Fig. 4, expansion in

Sea Side : Benchmark

R6 R5 R4 R3 R2 R1
R10 R9 R8 R7

No.2 No.1 High-

Generator Low-Pressure Low-Pressure Pressure Turbine
Generator
Side Turbine Turbine Turbine Side

L10 L9 L8 L7
L6 L5 L4 L3 L2 L1
2×5,600 5,100
3×5,600 2×5,600
Low-Pressure Turbine High-Pressure
Generator Section Section Turbine Section
(21,778.14) (16,800) (11,200)
(49,778.14)
Mountain Side Unit : mm

The value in ( ) means Setting Distance between the Benchmark

Fig. 3 Location of benchmarks and setting distance between them.

40

35
Expansion Values of Axial Direction (mm)

"0" is as December 1982

30
25

20
15

10
5

0
-5 Sea Side Mountain Side
-10
'78 '80 '82 '84 '86 '88 '90 '92 '94 '96 '98 '00 '02 '04
(Year)

Fig. 4 Expansion of table deck in axial direction (until 2004).

R. Manabe, H. Kawae, K. Ogawa and M. Matsuura / Journal of Advanced Concrete Technology Vol. 14, 590-606, 2016 594

the axial direction due to ASR increased up until the function, and to better understand the extent of ASR and
early 1980s, but appears to have stopped around 1991. the expansion rate. The compressive strength of the
Since then, the expansion has remained at the same level concrete, its static elastic modulus, observations of the
as the bench mark measurement, which fluctuates by cut surface of the cores and the results of accelerated
approximately 10 mm as a result of temperature varia- expansion tests carried out up until 2002 are all presented
tions due to seasonal change and turbine generator op- here.
eration.
(1) Compressive strength and static elastic
4.1.2 Crack conditions modulus
Visual inspection of concrete cracks in the TG founda- The locations on the TG foundation from which concrete
tion has been undertaken since 1982 to monitor the de- core was sampled are shown in Fig. 6, and the compres-
formation of the TG foundation. In 2002, major cracks sive strength of these cores is shown in Fig. 7. This re-
were observed in the main reinforcing bar direction on veals that the actual compressive strength of the table
the mountain side surface of the TG foundation (Fig. 5),
and these were accompanied by cracks developed in a
perpendicular direction to the reinforcing bar at the lower
portion of the turbine support beams. The maximum
local crack width was approximately 3.0 mm. As only a
few cracks were observed in the main reinforcing bar
direction and top of the column during the early stage of
crack monitoring (1982), these later cracks are assumed
to have developed as a result of expansion of the TG
foundation. Once ASR was determined to have settled,
and crack repair work was carried out, no new cracks
have been observed.

4.1.3 Concrete core sampling

Concrete core sampling of the TG foundation was carried
in order to determine if it could still provide a supporting
Fig. 6 Core sampling position.

Fig. 5 Cracking on mountain side of table deck (2002).

Fig. 7 Compressive strength of concrete core samples.

R. Manabe, H. Kawae, K. Ogawa and M. Matsuura / Journal of Advanced Concrete Technology Vol. 14, 590-606, 2016 595

Fig. 8 Static elastic modulus of concrete core samples.

deck area was well beyond the specified design strength

of 20.6 N/mm2, though was slightly less than the strength
in the column area. The static elastic modulus of the table
deck area (Fig. 8) was also reduced compared to other
areas, which is consistent with existing knowledge that
the static elastic modulus decreases with the develop-
ment of ASR. The relation between compressive strength
and the static elastic modulus (Fig. 9) had no clear
time-dependent tendency.

(2) Cut surface of core samples

Reaction rims and/or cracks in the aggregate are known
to occur when ASR develops, and so preliminary ob-
servation of cores taken from different sections was
carried out to clarify the extent of ASR in each area. The
core samples were cut into 2 cm-thick disks, the surfaces
of which were polished for more detailed observation
using a stereomicroscope. Through this analysis of six
surface points per core sampling point, the surface crack
and reaction rim condition was classified into one of four
categories, and number of reaction rims and/or cracks
was counted (Fig. 10). When compared to other areas of
the TG foundation, the table deck had more Type 2 (with
reaction rims) and Type 4 (with reaction rims and cracks)
features, which clearly indicates that this area was af-
fected by ASR (Japan Association for Building Research
Promotion 2006, Takakura and Watanabe et al. 2005).
This would also explain why cracks were more con-
spicuous in this area in Fig. 5.
(3) Accelerated expansion tests
The time-dependent behavior of the concrete cores ob-
tained through accelerated expansion testing using the
JCI-DD2 method proposed by the Japan Concrete Insti-
tute (JCI) is shown in Fig. 11. This method uses cylin-
drical cores (100 mmφ × 250 mm) kept at 20°C and 95%
relative humidity to measure the released expansion, and
Fig. 9 Correlation between compressive strength and
static elastic modulus.
these are subsequently exposed to an environment of
40°C and >95% relative humidity to measure the residual
expansion.
The possibility for future expansion of the table deck
section is considered low because maximum expansion
of the concrete was reached in 1989, and the expansion
rate has remained low since then. Any alkali in the con-
crete has since been consumed by the production of silica
gel due to aging, and so the expansion rate is less than
that in 1986 and still below the reference value (<0.1%
after 6 months) specified in JIS A1146 (mortar bar
method). Expansion possibility of the concrete other than
the table deck is considered to be small because its ex-
pansion rate has been smaller than that of the table deck
since the measurement started.
R. Manabe, H. Kawae, K. Ogawa and M. Matsuura / Journal of Advanced Concrete Technology Vol. 14, 590-606, 2016 596

4.2 Structural evaluation
The structural stability of the TG foundation then was
verified in 2005 by an analytical evaluation that took into
account the compressive strength and static elastic
modulus values obtained from the table deck expansion
and concrete core tests conducted once it was determined
that ASR had settled. Although it is quite unlikely that
ASR may reoccur in the future, this possibility was con-
sidered in order to fully examine the long-term sound-
ness of the structure (Shimizu and Asai et al. 2005; Ta-
kakura et al. 2009).
4.2.1 Analysis model
A three-dimensional frame model with beam elements,
as shown in Fig. 12, was adopted to analyze the TG
foundation so as to

Fig. 10 Classification of concrete core sample cut surface.

1986 1989 1992 2002

0.15 0.15 0.15 0.15
Table
deck
0.10 0.10 0.10 0.10
Ｍ１
Ｍ２ (%) (%) (%) (%)
Ｍ２’ 0.05 0.05 0.05 0.05
Ｍ３
Ｔ 0.00 0.00 0.00 0.00
Ｃ 0 3 6 9 12 0 3 6 9 12 0 3 6 9 12 0 3 6 9 12
Ｇ (M) (M) (M) (M)

0.15 0.15
Column･
intermediate 0.10 0.10
beam (%) (%)
0.05 0.05
Ｍ４
Ｍ６
Ｍ１０ 0.00 0.00
0 3 6 9 12 0 3 6 9 12
Ｍ１１ (M)
(M)

0.15 0.15

Foundation
0.10 0.10
mat
(%) (%)
0.05 0.05

Ｍ９ 0.00 0.00
0 3 6 9 12 0 3 6 9 12
(M) (M)

Fig. 11 Time-dependent behavior of accelerated expansion tests.

R. Manabe, H. Kawae, K. Ogawa and M. Matsuura / Journal of Advanced Concrete Technology Vol. 14, 590-606, 2016
(1) Take into account the difference in elongation be-
tween the sea side and mountain side of the table
deck and the nonlinearity of some of the applied load.
(2) To model the entirety of the TG foundation structure
above the basemat in order to properly evaluate the
stresses in the columns and intermediate beams.
In the beam element shown in Fig. 12, the mechanical
properties of the concrete and reinforcing bar were de-
fined individually. This was then used to simulate the
stress states in the reinforced concrete caused by expan-
sion due to ASR. The expansion behavior of the concrete
was simulated to fit the measured deformation on
horizontal bidirection by applying a linear thermal ex-
pansion coefficient and an equivalent temperature in-
crement. Here, it was assumed that the thermal expansion
of concrete can be used to simulate expansion by ASR
(Takakura and Shimizu et al. 2005).
4.2.2 Material properties
From previous studies of material properties obtained
through investigations of core specimens taken from
structures affected by ASR, it is known that the strength
and static elastic modulus deteriorate greatly when
compared to unaffected concrete (Kobayashi et al. 1991).
However, it has also been reported that there is very little
change in the stiffness and strength of concrete affected
by ASR if it contains reinforcing bar restraints resulting
in chemical prestress (Kobayashi 1986). The material
properties of the table deck used in the analysis therefore
Fig. 12 Analytical model and modeling of a section of beam.
597
took into account the material property reduction ratio,
i.e., the ratio of the material properties of concrete af-
fected by ASR to those of unaffected concrete. This ratio
was defined on the basis of a literature review (Koyanagi
et al. 1998, Tanahashi et al. 1996, Yamura et al. 1994),
model testing (Murazumi and Watanabe et al. 2005,
Murazumi and Hosokawa et al. 2005), vibration testing,
measurements of elastic velocity (Takakura and Wata-
nabe et al. 2005) and testing of core specimens taken
from the actual TG foundation, as shown in Table 2. The
relationship between the reinforcement ratio, compres-
sive strength ratio and modulus ratio obtained using these
methods is shown in Figs. 13 and 14.
The material properties used for analysis of the con-
crete were those obtained from core samples of the TG
foundation based on an average value from five different
methods, as shown in Table 2 (Shimizu and Watanabe et
al. 2005). In order to simulate long-term forced defor-
mation by elongation of the table deck section, the static
elastic modulus in the sound section was reduced by 1/3
from that of the core sample (Takakura and Shimizu et al.
2005). The material properties of concrete were defined
using these values for the uniformly sound section and
ASR-affected section.
The material properties of the reinforcing bars were
defined using average values for the uniformly obtained
through tensile testing of reinforcing bars taken from the
TG foundation (Table 3).
For the stress-strain curve of the concrete and rein-
R. Manabe, H. Kawae, K. Ogawa and M. Matsuura / Journal of Advanced Concrete Technology Vol. 14, 590-606, 2016 598

Table 2 Study procedure.

Item Preocedure Data studied
Using the compressive test results of ASR reinforced concrete, shown in avail-
able reports, stress-strain curves of the concrete have been calculated by deduc- Ratio of Compressive
Literature Review ing the forces borne by rebar from load-displacement curves of the tests. The Strength and Elastic
ratios of compressive strength and elastic modulus of ASR concrete to sound modulus
concrete have calculated from the stress-strain curves for certain rebar amounts.
Performing compression tests on the ASR specimens, the material properties
Ratio of Compressive
have been calculated from the test results through the same treatment of the data
Model Test Strength and Elastic
described above obtained. Comparison of the material properties obtained from
modulus
the tests on cast cylinder specimen and core specimen has been performed.
Vibration tests(vibrator test, microtremor measuring test) have been performied
In situ tests of actual foundation

on the structures with and without ASR. Elastic modulus of the structures have Ratio of Elastic
Vibration Test
been estimated by parametric analysis to obtain eigen value and mode to coincide modulus
the test results.
Measurements of elastic wave velocity(Logging Method, Transmission Method)
Measurement of Ratio of Elastic
have been performed on the structures with and without ASR. Elastic modulus
elastic velocity modulus
has been obtained from the velocity.
Test of Core
Ratio of Compressive
specimens taken Material property tests of core specimens taken from the structures with and
Strength and Elastic
from actual foun- without ASR have been performed.
modulus
dation

Table 3 Material properties.

Concrete Reinforcing bar
Scope Compressive Strength Fc Elastic Modulus Ec Elastic Modulus Es Yield Strength σy
(N/mm2) (N/mm2) (N/mm2) (N/mm2)
Soundness 30.4 (1.0) 2.49×104 (1.0)
2.04×105 409
ASR 28.8 (0.67) 1.67×104 (0.95)
( ) Raitio of the properties

forcing bars, the non-linear curve shown in Fig. 15 was
used so as to appropriately reflect the nonlinearity on
compression side of the concrete, the cracks produced on
the tensile side, and the yielding of the reinforcing bars.
For structural analysis assuming the redevelopment of
ASR, the compressive strength and static elastic modulus
of the ASR-affected region were assumed to be 80 and
50% that of the respective sound portion of the structure.
These values were considered the lower limit based on
the study results shown in Figs. 13 and 14 (Shimizu and
Watanabe et al. 2005).
4.2.3 Applied load and deformation
Dead load, operating load (live load, equipment load,

Fig. 13 Reinforcing bar ratio vs. compressive strength of concrete.

R. Manabe, H. Kawae, K. Ogawa and M. Matsuura / Journal of Advanced Concrete Technology Vol. 14, 590-606, 2016 599

Table 4 List of analytical conditions.

Case 1 Case 2
1-1 1-2 1-3 1-4 1-5 1-6 2-1 2-2 2-3 2-4 2-5 2-6 2-7 2-8
① dead + operating loads as applied to actual foundation (see 4.2.3)
② expansion of
X + Z direction Constant expansion
TD*1
X direction ＋ - Sc - - - - ○ - - -
Local Condition

(Axial direc-
－ - - Sc - - - - ○ - -
tion)
Horizontal Z direction ＋ - - - Sc - - - - ○ -
③ seismic load
(Perpendicular
(static)*2
to axis direc- － - - - - Sc - - - - ○
tion)
＋ - - - - - - ○ - ○ - ○ - ○ -
Vertical Y direction
－ - - - - - - - ○ - ○ - ○ - ○
④ short circuit torque*3 - - - - - T - - - -
*1 TD : Table Deck
*2 Sc : seismic load equivalent to design level
○ : gradually increasing seismic load
Load applied direction X direction (axial direction) : plus direction is from generator to turbine
Y direction (Vertical direction) : plus direction is from lower to upper
Z direction (perpendicular direction) : plus direction is from sea side to mountain side
*3 T : short circuit torque

etc.), seismic load and short circuit torque load were all
considered in this analysis (Japan Association for
Building Research Promotion 2006; Shimizu and Asai et
al. 2005). Expansion in the axial direction of the table
deck due to ASR was assumed to be 50 mm on the sea
side of the TG foundation and 34 mm on the mountain
side. These values were determined by adding the dis-
placement at the axial center of the turbine generator
obtained from adjustment records of the turbine genera-
tor from the start of its operation until December 1982, at
which point the expansion difference between the sea
side and mountain side was 4 mm (Fig. 16).
Expansion in the perpendicular direction was assumed
to be 6.7 mm for the generator side and 9.7 mm for the
turbine side based on the notion that a strain (840 μ)
equivalent to the average expansion (42 mm) in the axial
direction should occur in the perpendicular direction.
The average value for the expansion rate of the table deck
in 2002 was determined to be 0.02% based on the ac-
celerated expansion tests of core samples discussed in
4.1.3(3). The expansion value was assumed to be 10 mm
(i.e., 0.02% of the total length of the table deck), which
increased the length of expansion on the sea side from 50
to 60 mm. Thus, for structural evaluation assuming the
redevelopment of ASR, the expansion value was set to
1.2 times the total value.
4.2.4 Evaluation methods
To evaluate the margin of safety with regards to the
aseismic strength of the structure, the material strengths

Fig. 14 Reinforcing bar ratio vs. elastic modulus of concrete.

R. Manabe, H. Kawae, K. Ogawa and M. Matsuura / Journal of Advanced Concrete Technology Vol. 14, 590-606, 2016 600

of the reinforcing bar and concrete obtained through
testing were used to determine an appropriate reference
(i.e., a standard value for margin evaluation). For this, the
following values were adopted as standards:
2
• Reinforcing bar yield strength = 409 N/mm (as de-
termined by tensile strength testing)
• Concrete compressive strength (determined from core
specimens taken in 2002)
2
• ASR-affected region = 28.8 N/mm
2
• Unaffected concrete = 30.4 N/mm
To calculate the shear strength capacity, the
Ono–Arakawa lower bound equation was used to get
conservative results, as outlined by the standards of the
Architectural Institute of Japan (AIJ). The influence of
twisting was taken into account by using an equation
proposed by Article 22 of “AIJ Standard for Structural
Calculation of Reinforced Concrete Structures”.
4.2.5 Results of analysis
Analytical evaluations of the TG foundation based on
monitoring of its condition up until 2004, and assuming a
recommencement of ASR, were undertaken for two
separate cases. In Case 1, conditions under normal op-
eration (a dead load + operating load and expansion of
the table deck) were combined with seismic load or load
due to short circuit torque to determine if the stress state
remained within designed levels. In Case 2, the seismic
load was gradually increased to determine the safe mar-
gin. A list of the analytical conditions used to evaluate
the safety margin for strength in both cases is provided in
Table 4.
(1) Evaluation of structure based on measurements
a) Case 1
The tensile stress of the reinforcing bar, compressive
stress of the concrete, and occurrence of stress due to
shear force are shown in Fig. 17 in relation to the stan-

Fig. 15 Material property curves.

60

50

40
Displacement (mm)

30

20

10

0
'75'76'77'78'79'80'81'82'83'84'85'86'87'88'89'90'91'92'93'94'95'96'97'98'99'00'01'02'03'04
Sea Side Year
Mountain Side
Difference between Sea Side and Mountain Side
Sea Side ( on annual inspection )
Mountain Side ( on annual inspection )
Difference between Sea Side and Mountain Side ( on annual inspection )
End Play
Fig. 16 Expansion of turbine generator foundation.
R. Manabe, H. Kawae, K. Ogawa and M. Matsuura / Journal of Advanced Concrete Technology Vol. 14, 590-606, 2016 601

dard value. A representative set of results for this analysis
is given by the deformation mode of Case 1-1 and Case
1-3 shown in Fig. 18. Note that in Case 1-1, the maxi-
mum ratio of working stress to the standard value of
tensile stress for the reinforcing bar is 60%, which pro-
vides a sufficient safety margin relative to the reference
value (Fig. 17). In Case 1-3, this ratio increases to 70%,
but this is still considered within an appropriate margin.
A comparison of the deformation between these two
cases suggests that ASR was more critical than seismic
load (Fig. 18).
b) Case 2
Figure 19 compares the seismic loads corresponding to
yielding of the reinforcing bars, compressive fracture of
the concrete and shear failure to the designed seismic

Fig. 17 Results of analysis (Case 1: Analytical evaluation of structure based on measurements).

Fig. 18 Deformation (Case 1: Analytical evaluation of structure based on measurements).

Fig. 19 Results of analysis (Case 2: Analytical evaluation of structure based on measurements).

R. Manabe, H. Kawae, K. Ogawa and M. Matsuura / Journal of Advanced Concrete Technology Vol. 14, 590-606, 2016 602

40
35

Expansion Values of Axial Direction (mm)

"0"
"0" isis as
as December
December 1982
30
25
20
15

10
5
0
-5 Sea Side Mountain Side
-10
'78 '80 '82 '84 '86 '88 '90 '92 '94 '96 '98 '00 '02 '04 '06 '08 '10 '12 '14
(Year)
Fig. 20 Expansion of table deck in axial direction (until 2014).

load. This reveals that the actual seismic load is within a

suitable margin, being approximately 4.7 times of the
designed load.

(2) Evaluation of structure with recommencing ASR

a) Case 1
An analysis similar to Case 1-3, where the ratio of the
working stress to the standard value was maximum, was
conducted based on an assumption of ASR recommenc-
ing using the measured conditions up to 2004 as a basis.
The maximum ratio of the working stress to the standard
value was found to increase in this case to 77%, but was
still considered to provide a sufficient margin of safety.

b) Case 2
An analysis similar to Case 2-3, where member safety
margin to seismic load was minimum, for the recom-
mencement of ASR using the measured conditions up to
2004 found that the strength of the structure remained 4.4
times that of the gradually increasing seismic load, which Fig. 21 Expansion between benchmark distances in each
section.
was considered a sufficient margin.

5. Continuous monitoring 5.1 Monitoring of deformation conditions

(1) Table deck expansion
Continuous/additional monitoring for deformation in the The time-dependent behavior of the table deck expansion
TG foundation and of the compressive strength of its up until 2014 shown in Fig. 20 reveals that a gradual
concrete has been used to verify that ASR has not re- decrease in occurred in the expansion of both the sea and
commenced after reaching a settled state of structural mountain side after about 2007. Following the cessation
stability (Taniguchi et al. 2015). The column inclination of operations in 2011, this decrease became much more
has been measured along with any expansion of the table pronounced, which is believed to be the result of the
deck to determine deformation, while the concrete has structure only being affected by the temperature of the
been assessed by monitoring any cracking, its internal atmosphere once heat generation in the steam pipe to the
temperature and the strain in the reinforcing bars. This turbine stopped.
has verified that there has been no further development The expansion of unit between the benchmark dis-
of ASR as of 2016, nor has there been any rupture of the tances of each section of the high- and low-pressure
reinforcing bars based on magnetic flux density meas- turbines and generator over time is shown in Fig. 21.
urements. Rebound hardness testing has also ascertained Note that the unit expansion value used here represents
that there has been no major change in the compressive the sum of the expansion values for each section divided
strength of the concrete. by the set benchmark distance for each section. In the
high-pressure turbine section, the internal temperature of
the concrete during turbine operation in summer was
R. Manabe, H. Kawae, K. Ogawa and M. Matsuura / Journal of Advanced Concrete Technology Vol. 14, 590-606, 2016
approximately 40°C on the sea side and approximately
45°C on the mountain side. These temperatures dropped
by approximately 10°C during winter. A similar seasonal
change was observed in other sections, but during regular
inspection, the internal temperature was generally ap-
proximately 25°C all year. The positions at which the
internal temperature was measured are shown in Fig. 22,
while the time-dependent behavior of each section for a
given temperature is shown in Fig. 23.
From the relationship between the table deck expan-
sion (Fig. 21) and internal temperature of the concrete
(Fig. 23), it is evident that the internal temperature on the
Fig. 22 Location of internal temperature measurement
points.
Fig. 24 Location of reinforcing bar strain measurement
points.
Fig. 23 Temperature in each section.
603
sea side was lower (maximum temperature of approxi-
mately 40°C) than on the mountain side, and that the unit
expansion of the mountain side was greater. As the unit
expansion on the mountain side was highest in areas of
low internal concrete temperature (maximum tempera-
ture was approximately 40°C), the temperature most
conducive to promoting ASR is considered to be in the
vicinity of 40°C. The shrinkage observed in the table
deck axial direction can also be explained by a decrease
in the internal temperature of the concrete after opera-
tions were stopped.
(2) Reinforcing bar strain
The strain in the reinforcing bars has been measured
since 1986 using meters that were installed following
cutting of the main reinforcement of the columns and
beams, where the initial strain at the cut-off was intro-
duced to the meter. The measurement points are shown in
Fig. 24 and the time-dependent behavior of the main
reinforcing bar (M-1) in the sea side of the high-pressure
turbine section is shown in Fig. 25. We see from this that
there was a downward trend in the reinforcing bar strain
following a peak in the early 1990s, which is consistent
with the change in expansion behavior between the
benchmarks.
(3) Column inclination
The column inclination has been monitored since 2008
Fig. 25 Reinforcing bar strain (M-1).
R. Manabe, H. Kawae, K. Ogawa and M. Matsuura / Journal of Advanced Concrete Technology Vol. 14, 590-606, 2016 604

by measuring the difference between the displacement of
the table deck relative to the position of the column base
(a difference in height of approximately 11 m) using
pendulums situated at each corner of the TG foundation.
The measurement points and starting displacement is
shown in Fig. 26, and the time-dependent behavior of the
table deck displacement along its axis is shown in Fig. 27.
Note that prior to 2011, the displacement on the moun-
tain side varied between 56 and 62 mm depending on
temperature, while the sea side displacement was be-
tween 50 and 56 mm. Once operations ceased, however,
this displacement decreased owing to the decrease in the
internal temperature of the concrete.
5.2 Compressive strength of concrete
In order to determine the compressive strength of the
concrete, rebound hardness has been measured since 2008
near the portion where concrete cores were taken from the
TG foundation in 2002. These tests have verified that there
has been no major change in this time (Fig. 28).
5.3 Reinforcing bar rupture
The possibility of reinforcing bar rupture by ASR was
considered to be low in the TG foundation because of its
indoor environment. However, as reinforcing bars of
similar diameter have ruptured in outdoor structures, the
magnetic flux density of the bars has been measured at
the corners of the columns and beams of the TG founda-
tion since 2015. This has been achieved using a mag-
netizing unit based on a permanent magnet, with rupture
being determined by whether the peak magnetic flux
density exceeds a predetermined value obtained through
a series of experiments (Hirose et al. 2012). No rupture
has been detected since measurements began, with an
example of the peak value for the magnetic flux density
measured in the TG foundation provided in Fig. 29.

Fig. 26 Location of column inclination measurement and

starting displacement.

Fig. 27 Trend of column inclination.

Fig. 28 Trend of rebound hardness.

R. Manabe, H. Kawae, K. Ogawa and M. Matsuura / Journal of Advanced Concrete Technology Vol. 14, 590-606, 2016
Fig. 29 Peak value of magnetic flux density distribution (example).
6. Conclusion
This paper has presented an overview of the maintenance
management method that has been applied to the TG
foundation of Unit #1 at the Ikata nuclear power plant
ever since concrete cracks caused by ASR were first
identified. This approach has included an evaluation of
both the current and long-term soundness, and can be
applied to other structures affected by ASR. The main
conclusions drawn from this are as follows:
• The continued use of a structure subject to ASR is
dependent on defining its required function and
evaluating its current and long-term soundness. In the
case of the TG foundation in this study, no problems
were found in its ability to provide a support function
because there was no induced vibration that might af-
fect its operation.
• Outdoor structures such as piers and bridge abutments
are more susceptible to ASR, as the surface of their
concrete is more readily exposed to moisture and alkali,
whereas the architectural structures are protected.
However, the movement of water held within a con-
crete structure, and the heat generated within the
structure, can affect the likelihood of ASR. There is
therefore still a need to clarify the relationship between
deformation caused by ASR and the surrounding en-
vironment in terms of the temperature, humidity, in-
vasion of moisture, and others because these are not
clarify.
References
Chénier, J. O., Komljenovic, D., Gocevski, V., Picard, S.
and Chrétien, G., (2012). “An approach regarding
aging management program for concrete containment
structure at the gentilly-2 nuclear power plant.” 33rd
Annual Conference of the Canadian Nuclear Society
and 36th Annual CNS/CNA Student Conference, 10-13
June 2012, Saskatchwan, Canada.
Hirose, M., Maeda, T., Matsuda, K., Yokota, M., Hattori,
A. and Miyaawa, T., (2012). “Establishment of the
criterion in non-destructive test method for fracture of
reinforcing steel bar by measuring magnetic flux
605
density.” Journal of Structural Engineering A, 58A,
867-878. (in Japanese)
Japan Association for Building Research Promotion,
(2006). “Report of the investigative commission for
soundness evaluation and maintenance method of
reinforced concrete structures on Alkali aggregate
reaction.” (in Japanese)
Kobayashi, K., (1986). “Load carrying capacity of the
members and structures affected by alkali silica
reaction-beam member.” Concrete Technology, 24(11),
70-78. (in Japanese)
Kobayashi, K., Maru, A. and Tatematsu, H., (1991).
“Series, diagnosis of concrete structure durability No.
2, Diagnosis of alkali aggregate reaction.” Morikita
Publishing, 42-41. (in Japanese)
Koyanagi, W., Uchida, Y., Iwanaga, T. and Asano, Y.,
(1998). “Restraint effect of ASR expansion in RC
members with small reinforcement ratios.” Cement
Science and Concrete Technology, 52, 786-791. (in
Japanese)
Minato, T. and Torii, K., (2010). “ASR-deteriorated
bridges and rock type of reactive aggregate on Noto
expressway.” Journal of the Society of Materials
Science Japan, 59 (10), 781-786. (in Japanese)
Murazumi, Y., Watanabe, Y., Matsumoto, N., Mitsugi, S.,
Takiguchi, K. and Masuda, Y., (2005). “Study on the
influence of alkali-silica reaction on structural
behavior of reinforced concrete members.” 18th
International Conference on Structural Mechanics in
Reactor Technology (SMiRT18), SMiRT18-H03-2,
Beijing, China, August 7-12.
Murazumi, Y., Hosokawa, T., Matsumoto, N., Mitsugi, S.,
Takiguchi, K. and Masuda, Y., (2005). “Study on the
influence of alkali-silica reaction on mechanical
properties of reinforced concrete members.” 18th
International Conference on Structural Mechanics in
Reactor Technology (SMiRT18), SMiRT18-H03-3,
Beijing, China, August 7-12.
Shimizu, H., Asai, Y., Hosokawa, T., Sekimoto, H., Sato,
K. Oshima, R., Takiguchi, K., Masuda, Y. and
Nishiguchi, I., (2005). “Investigation of safety margin
for turbine generator foundation affected by alkali
R. Manabe, H. Kawae, K. Ogawa and M. Matsuura / Journal of Advanced Concrete Technology Vol. 14, 590-606, 2016
silica reaction based on non-linear structure analysis.”
18th International Conference on Structural
Mechanics in Reactor Technology (SMiRT18),
SMiRT18-H03-4, Beijing, China, August 7-12.
Shimizu, H., Watanabe, Y., Ishikawa, T., Sekimoto, H.,
Oshima, R., Takiguchi, K., Masuda, Y. and Nishiguchi,
I., (2005). “Study on material properties in order to
apply for structural analysis of turbine generator
foundation affected by alkali-silica reaction.” 18th
International Conference on Structural Mechanics in
Reactor Technology (SMiRT18), SMiRT18-H03-5,
Beijing, China, August 7-12.
Tanahashi, K., Iwanaga, T., Koyanagi, W. and Asano, Y.,
(1996). “Structural behavior of RC beam and columns
deteriorated by ASR.” Proceedings of the Japan
Concrete Institute, 18(1), 843-848. (in Japanese)
Taniguchi, T., Ogawa, K., Kawae, H. and Takagi, S.,
(2015). “Investigation on the long-term changes of
turbine generator foundation affected by alkali-silica
reaction.” Proceedings of the Japan Concrete Institute,
37(1), 871-876. (in Japanese)
Takakura, M., Watanabe, Y., Hosokawa, T., Ishi, T.,
Takiguchi, K. and Masuda, Y., (2005). “Vibration
measurement and simulation analysis on a reinforced
606
concrete structure with alkali-silica reaction.” 18th
International Conference on Structural Mechanics in
Reactor Technology (SMiRT18), SMiRT18-H03-1,
Beijing, China, August 7-12.
Takakura, T., Ishikawa, T., Mitsugi, S., Matsumoto, N.,
Takiguchi, K., Masuda, Y. and Nishiguchi, I., (2005).
“Investigation on the expansion value of turbine
generator foundation affected by alkali-silica
reaction.” 18th International Conference on Structural
Mechanics in Reactor Technology (SMiRT18),
SMiRT18-H03-7, Beijing, China, August 7-12.
Takakura, T., Shimizu, H., Masuda, Y. and Matsumoto,
N., (2009). “Consideration of deformation for
reinforced concrete structure affected by alkali-silica
reaction.” Concrete Research and Technology, 20(3),
1-10. (in Japanese)
Takakura, T., Shimizu, H., Masuda, Y., Matsumoto, N.,
Oshima, R. and Sekimoto, H., (2005). “Investigation
of safety margin for turbine generator foundation
affected by alkali-silica reaction (Part3) Simulation of
model and actual structure.” Preprints of Annual
Convention of the Architectural Institute of Japan,
Osaka, Japan, September 1-3. (in Japanese)
Journal of Advanced Concrete Technology Vol. 14, 464-474 August 2016 / Copyright © 2016 Japan Concrete Institute 464

Technical report

Maintenance of Prestressed Concrete Containment Vessels in a Nuclear

Power Plant
Yoshihiro Yamaguchi1, Takashi Kitagawa2, Hideki Tanaka3*, Yasumichi Koshiro4, Hideo Takahashi5,
Shinichi Takezaki6 and Masazumi Nakao7

Received 31 March 2016, accepted 31 July 2016 doi:10.3151/jact.14.464

Abstract
The main feature of a prestressed concrete containment vessel (hereinafter referred to as “PCCV”) is that unbonded
tendons, which are not bonded to the structure, are used because the tension of the tendon needs to be measured in pe-
riodical in-service inspection (hereinafter referred to as “ISI”) to ensure that the performance of the power plant is
maintained during the service life. This report describes the results of measurement of the tension of the tendon, which has
been performed as an activity to maintain the PCCV in Ohi Nuclear Power Plant (hereafter referred to as Ohi “NPP”)
Units 3 and 4 of Kansai Electric Power in the past twenty some years since the construction of the PCCV, the changes in
the standards and guidelines for the maintenance of the PCCV in Japan, as well as the discussion of non-destructive
test-based methods for estimating the strength of high-strength concrete used for the PCCV and the development of a
hydraulic shim-type load cell-based system for the measurement of the tension of the tendon, taking into account the ease
of non-destructive testing, etc.

1. Introduction 2. Outline of the PCCVs (Ozaki et al. 2004)

A NPP has a component called a containment vessel to
contain basic components such as the nuclear reactor
vessel loaded with fuel. In conventional pressurized
water reactors (hereinafter referred to as “PWR”), a steel
containment vessel (hereinafter referred to as “SCV”) is
used as the reactor containment vessel. However, a
PCCV has been used to meet the requirements as a result
of the increase in the output power and has been twenty
some years since its construction.
This report describes the maintenance activities
mainly for the prestressed members of the PCCV in Ohi
NPP Units 3 and 4 of The Kansai Electric Power Co., Inc.
1 fuel and cool the fuel. It is a potential condition in which
Leader, Nuclear Power Division,The Kansai Electric
Power Co.,Inc., Japan.
2 steam is released instantaneously. The containment vessel
Manager, Nuclear Power Division, The Kansai Electric
Power Co.,Inc., Japan.
3 two functions: (1) the pressure-retaining (internal pres-
Deputy General Manager, Nuclear Facilities Division,
Obayashi Corporation, Japan. *Corresponding auther,
E-mail: tanaka.hideki@obayashi.co.jp
4 the increase in the output power, the need arose for a
Chief Research Engineer, Technical Research Institute,
Obayashi Corporation, Japan.
5
Manager, Planning & Design Section, Nuclear Facilities
Engineering II, Nuclear Facilities Division, Taisei
Corporation, Japan.
6
Manager, Building System and Material Research
Section, Building Engineering Research Institute,
Technology Center, Taisei Corporation, Japan.
7
Manager, Wakasa Branch, The General Environmental
Technos Co., Ltd., Japan.
The preparation construction work of Ohi NPP Units 3
and 4 of The Kansai Electric Power Co., Inc. started in
1985. Unit 3 went into commercial operation in De-
cember 1991 and Unit 4 in February 1993. Both of them
are NPPs with an output power of 1,180 MW and use
PCCVs for the containment vessels. Table 1 shows the
outline of Ohi NPP Units 3 and 4.
The PCCV can achieve the requirements for the func-
tions of pressure-retaining and leakage prevention
mainly in the event of an accident at a NPP with an
output power of more than 1,000 MW. Its details are
described below.
One of the severest conditions assumed in the design of
a PWR is a loss of coolant accident (hereinafter referred to
as “LOCA”) due to a pipe break in the primary coolant
system, which is the main system to remove heat from the
due to a primary coolant pipe break, very high-pressure
is a structure to prevent the steam from leaking out and has
sure) in a LOCA and (2) leakage prevention of the steam.
A 40-mm thick SCV had been used previously. With
larger containment vessel to reduce the pressure in a
LOCA. The PCCV can be much smaller in volume than
Table 1 Outline of Ohi NPP Units 3 and 4 of The Kansai
Electric Power Co., Inc.
Commencement of
Unit Reactor Rated Power
Operation
Unit 3 Pressurized Water 1,180 MW December 1991
Unit 4 Reactor (PWR) 1,180 MW February 1993
Y. Yamaguchi, T. Kitagawa, H. Tanaka, Y. Koshiro, H. Takahashi, et al. / Journal of Advanced Concrete Technology Vol. 14, 464-474, 2016 465

Table 2 Basic specification of the PCCV (Ozaki et al. 2004).

Item Specification
Upper
Shape: Hemispherical cylindrical
Construction: Prestressed concrete
Diameter: About 45 m
Shape, structure and Height: About 65 m
dimensions of PCCV Bottom
Shape: Flat base mat
Construction: Reinforced concrete
Base mat thickness: About 10 m
Number of Buttresses: 2
Prestressing system: VSL (Unbonded type: Filled with a corrosion protection medium)
Prestressing members (tendon)
Capacity: 10 MN class
Dimension of Inverted U tendon: 90 (@2º)
Quantity
prestressing members Hoop tendon of cylinder: 90 (@450)
(spacing)
Hoop tendon of dome: 18 (@2.5º)
Tensile force during Inverted U tendon: 7.05 MN (719 tf)
prestressing Hoop tendon: 7.19 MN (733 tf)
Concrete Design strength 44.1 N/mm2 (450 kgf/cm2)

SCV by prestressing the concrete and thus can be more
resistant to the pressure in a LOCA. The steam leakage
can be prevented by applying a steel liner plate to the
inner surface of the concrete. In addition, the SCV re-
quires a reinforced concrete outer shielding wall, which
has the function to shield radiation, on the outer side of
the containment. The concrete of the PCCV has the
function to shield radiation. Thus, the PCCV is a rea-
sonable structure with the functions of the SCV and the
outer shielding wall combined.
Table 2 shows the basic specification of the PCCV.
Figure 1 shows a schematic cross-sectional elevation
view of the PCCV and Fig. 2 shows the details of the
elevation. The PCCV consists of about 10 m thick base
mat and about 65 m high cylindrical shell and a dome
with a diameter of about 45 m. The cylindrical shell and
the dome, the upper structure, are prestressed concrete.
The steel liner plate on the inner surface has the function
to prevent leakage, and the load capacity of the
Dome part
prestressing members (hereinafter referred to as “ten-
don”) is 10 MN (1000 tf).
Figure 3 shows the overall arrangement of the tendons
and a schematic cross-sectional plan view of the PCCV.
The tendons consist of hoop tendons (90 in the cylin-
drical part and 18 in the dome part), which apply a force
in a manner to hoop the PCCV, and inverted U vertical
tendons (90), which apply a force in a manner to press
down the PCCV. The hoop tendons, which circulate
around the circumference of the PCCV, are anchored at
both ends to the anchorages on the buttresses. There are
two buttresses, one each at diagonal positions on the
circumference. The hoop tendons are arranged so that
they are anchored alternately to each buttress. A group of
45 inverted U tendons are arranged in an orthogonal grid
pattern in a plan view and anchored to the ceiling of the

Inverted
U tendon
Cylindrical part

Equipment hatch
Shell

Piping
penetration

Bottom
Fig. 1 Schematic cross-sectional elevation of the PCCV. Fig. 2 Wall structure of PCCV (Yonezawa et al. 2003).
Y. Yamaguchi, T. Kitagawa, H. Tanaka, Y. Koshiro, H. Takahashi, et al. / Journal of Advanced Concrete Technology Vol. 14, 464-474, 2016 466

tendon gallery in the base mat. During prestressing, the bonded to the structure, are used and the sheath for the
prestressing force applied to the ends of the tendon is tendon is filled with grease as the corrosion protection
7.19 MN (733 tf) for the hoop tendon and 7.05 MN (719 medium.
tf) for the inverted U tendon, so that the membrane stress The prestressed system uses a 10 MN-class VSL
during a LOCA is compressive, even taking into account (Vorspann System Losinger) system. One tendon con-
the loss of the prestressing force due to the creep of sists of 55 PC steel strands with a diameter of 12.7 mm
concrete and the relaxation of PC steel strands during the and has a rated tensile strength of 10.09 MN (1028.5 tf).
service life of the plant. The anchorage is such that the reaction is taken by the
To ensure that the performance of the PCCV is main- anchor head; the tendons are pulled by a jack; and the PC
tained during the service life, periodical ISI needs to be steel strands are anchored to the anchor head by wedges.
performed to measure the prestressing force of the ten- Figure 4 shows a schematic of the VSL system. As a
don. For this reason, unbonded tendons, which are not compressing force, the prestressing force applied to the
tendons is transferred to the concrete structure via the
anchor heads and the bearing plates. The inner diameter
of the trumpet sheath is extended at one end into the
Inverted U concrete structure so that the anchor head moves into the
tendon Dome sheath to release the prestressing force. Shims are in-
Hoop tendon serted between the anchor head and the bearing plate to
of dome transfer the stress during the service life. The anchorage
Cylinder
is covered by the end cap and filled with the corrosion
protection medium to prevent entry of moisture and
Equipment Hoop tendon protect the system from corrosion.
hatch of cylinder
3. In-service Inspection (ISI)
3.1 Outline
A plan for ISI of the PCCVs of Ohi NPP Units 3 and 4
Tendon gallery has been developed in accordance with the Technical
Base mat Guidelines for Prestressed Concrete Containment Ves-
sels for Ohi NPP Units 3 and 4 (February 1987 (partially
revised in 1988), Agency for Natural Resource and En-
Tendon arrangement ergy, Ministry of International Trade and Industry). (1)
Visual inspection, (2) inspection of the prestressed sys-
Hoop tendon
tem and (3) inspection of the corrosion protection me-
dium were performed in the 1st, 3rd and 5th year of op-
eration (the liner is excluded from this report). 4% and 4
Buttress or more of the total number of tendons in each group
were selected for (2) inspection of the prestressed system
from each group of tendons (hoop and inverted U ten-
dons). To evaluate the behavior, etc. of the tendons dur-
Hoop tendon ing the service life of the plant and find a correlation
between the observation data from the previous per-
Schematic cross-sectional plan view of the PCCV
formed ISIs, more than one “common” tendon (history
inspection tendon) to be inspected in every ISI was se-
Fig. 3 Overall arrangement of the tendons and a sche- lected from each group.
matic cross-sectional plan view of the PCCV (Ozaki et al. Based on the results of the inspection performed in the
2004). 1st, 3rd and 5th year of operation, Regulatory Guide 1.35

Corrosion inhibitor Corrosion inhibi-

injection port tor injection port Wedge
End cap Sheath Trumpet End cap
Trumpet sheath
Sheat

PC steel strands PC steel strands Anchor head

Anchor head
Bearing plate Shim
Bearing plate
Fig. 4 Outline of the prestressing system by VSL (Ozaki et al. 2004).
Y. Yamaguchi, T. Kitagawa, H. Tanaka, Y. Koshiro, H. Takahashi, et al. / Journal of Advanced Concrete Technology Vol. 14, 464-474, 2016
of USNRC and ASME Section XI, it was decided to
perform the inspection every five years between the 10th
to the 15th year of operation. The quantity of tendons to
be inspected was reduced to 2% and 3 or more of the total
number of tendons in each group. For the common ten-
dons, tendons same with the ones inspected in the 1st, 3rd
and 5th year of operation were inspected.
Some of the tendons to be inspected in the 20th and
subsequent years of operation were changed based on the
plan to incorporate the Rules on Fitness-for-Service for
Nuclear Power Plants into the Codes for Nuclear Power
Generation Facilities - Rules on Concrete Containment
Vessels for NPPs (hereinafter referred to as “the CCV
Rules”) enacted by Japan Society of Mechanical Engi-
neers. More specifically, there is no change to the plan to
inspect 2% and 3 or more of the total number of tendons
in each group in the 5th and subsequent years of operation.
However, regarding hoop tendons, a total of two com-
mon hoop tendons, one dome hoop tendon and one cyl-
inder hoop tendon, were selected. In the plan to incor-
porate the Fitness-for-Service Rules into the CCV Rules,
one common tendon will be selected from all hoop ten-
dons and two arbitrarily selected tendons (hereinafter
referred to as “random sample tendon”) will be selected
as tendons to be inspected. Considering the above, as for
hoop tendons to be inspected, a total of four tendons (two
common tendons and two random sample tendons) were
selected. As for the common tendons, two tendons con-
tinued to be inspected, taking into account the accumu-
lated data. For the random sample tendons, tendons that
had not been inspected were selected from the perspec-
tive of avoiding duplication, which is based on the plan
to incorporate the Fitness-for-Service Rules into the
CCV Rules.
The inspection of the prestressing force is called a
lift-off test to verify the residual tension of the tendon.
More specifically, a filler gauge (a 0.3 mm thick stainless
steel sheet) is used to measure the lift-off load. In this
measurement, a jack is placed at the end of an anchorage;
reaction is taken by the bearing plate; and the tendons are
pulled by the jack; and the lift-off load is the load when
the filler gauge between the anchor head and the bearing
plate can be removed. Figure 5 shows the method for the
Fig. 5 Method for the measurement of the lift-off load (Abe
et al. 2003). (Lift-off Load measurement by filler gauge)
467
measurement of the lift-off load. In a lift-off test, it is
verified that the prestressing force of the tendon is more
than the design requirement.
3.2 Summary of the observed results and the
evaluation
In the Ohi NPP, the 25th year ISI of Unit 3 PCCV was
performed in August 2015 and the 20th year ISI of Unit 4
was performed in February 2012.
In each ISI, as a visual inspection, a general visual
inspection of the concrete surface and a detailed visual
inspection of the controlled portions were performed;
and as testing of the prestressed system, a tension test of
the tendons, a detailed visual test of the anchorages of the
tendons and a test of the corrosion protection medium
were performed.
The results and evaluation of the ISI of Ohi NPP Units
3 and 4 are summarized below.
(1) Visual inspection
i) General visual inspection of the concrete surface
For a general visual inspection of the concrete surface, a
general visual inspection was performed on the outer
surface of the cylindrical part and the buttresses. The
results of the previously performed ISIs showed no
cracks and damage that could adversely affect the
structure of the containment vessel.
ii) Detailed visual inspection of the controlled areas
For a detailed visual inspection of the controlled areas,
distinctive areas of the surface of the containment vessel
were selected and defined as control areas, and a visual
inspection of the concrete surface at the following areas
was performed in each ISI. The results of the previously
performed ISIs showed only very small cracks in the
concrete surface.
i Cylindrical part near the base mat
ii General cylindrical part with no pipe penetration in
the vicinity
iii Boundary between the cylindrical part and the
dome
iv Boundary between the cylindrical part and the but-
tress
v Near the air lock
vi Near the equipment hatch
vii Local area of the cylindrical part with a pipe pene-
tration
(2) Inspection of the prestressed system
i) Tension test of the tendons
For a tension test of the tendons, the lift-off load is
measured to verify the prestressing force which is the
subject of inspection is more than the design requirement
(5.36 MN for the hoop tendon and 5.66 MN for the in-
verted U tendon). Figure 6 shows the results of the ten-
sion test for Ohi NPP Unit 3. The results of the previ-
ously performed ISIs showed that the prestressing force
was more than the design requirement.
Y. Yamaguchi, T. Kitagawa, H. Tanaka, Y. Koshiro, H. Takahashi, et al. / Journal of Advanced Concrete Technology Vol. 14, 464-474, 2016
9.0
8.0
7.0
Lift-off load (MN)
6.0
5.0
Actual Lift-off Force(Dome)
4.0 Actual Lift-off Force(Cylindrical)
Minimum Required Value
3.0
1 10
Time (Years)
(a) Hoop tendons in the cylindrical part and in the dome part
Fig. 6 Results of the tension test for Ohi NPP Unit 3.
ii) Detailed visual test of the anchorages of the ten-
dons
In a detailed visual test of the anchorages of the tendons,
for the tendon which is the subject of inspection, a visual
inspection for cracks in concrete at the anchorages of
the tendons, a visual inspection of the anchorage (anchor
head, bearing plate, etc.), and an inspection of the num-
ber of effective strands are performed. The results of the
previously performed ISIs showed no problem.
(3) Inspection of the corrosion protection medium
For an inspection of the corrosion protection medium, a
visual inspection of the extracted corrosion protection
medium, a chemical composition analysis of the ex-
tracted corrosion protection medium, and a volume
measurement of the newly injected corrosion protection
medium were performed. In the visual examination, it
was checked that there was no infiltration of water from
the outside, and water-soluble impurities etc. were in-
spected in the chemical component analysis of the ex-
tracted corrosion protection medium. The results of the
previously performed ISIs showed all clear with the
results of the visual inspection of the extracted corrosion
protection medium. Also, for the results of the chemical
composition analysis of the extracted corrosion protec-
tion medium, water-soluble impurities (Chloride ion etc.)
were below the detection limit. The results of the volume
measurement of the newly injected corrosion protection
medium showed that the amount of the newly injected
corrosion protection medium was more than the amount
of the removed corrosion protection medium.
4. Changes over the year in the standards
and guidelines for the maintenance of
concrete containment vessels in Japan
The PCCV of Tsuruga NPP Unit 2 of The Japan Atomic
Power Company is the first concrete containment vessel
(hereinafter referred to as “CCV”). Since at that time,
there were no technical standards for the CCV, internal
review rules (Approval of Specially Designed Facilities
468
9.0
8.0
7.0
Lift-off load (MN)
6.0
5.0
Actual Lift-off Force
4.0
Minimum Required Value
3.0
100 1 10 100
Time (Years)
(b) Inverted U tendon
under Article 3 of Ministerial Ordinance No. 62) were
individually made by the Agency for Natural Resources
and Energy, Ministry of International Trade and Industry.
The aforementioned Technical Guidelines for
Prestressed Concrete Containment Vessels for Ohi NPP
Units 3 and 4 were used as the internal review rules for
the Ohi NPP.
The internal rules provide technical guidelines for the
material, design, construction, pre-service structural
integrity testing (hereinafter referred to as “SIT”) and ISI
of a PCCV. The guidelines provide only a brief descrip-
tion of periodic ISIs: visual inspection for cracks in the
surface of concrete and corrosion of PC steel members,
liner plates, etc. tension testing, and material properties
testing of tension member fittings. As reference literature,
Regulatory Guide 1.90 Rev.1 “Inservice Inspection of
Prestressed Concrete Containment Structures with
Grouted Tendons”, Aug. 1977 and Proposed Revision 3
to Regulatory Guide 1.35 “Inservice Inspection of Un-
grouted Tendons in Prestressed Concrete Containments”,
Apr. 1979 are cited.
Subsequently, a reinforced concrete containment ves-
sel (hereinafter referred to as “RCCV”) was added to the
guidelines, and it became necessary to develop technical
standards for CCVs. In September 1987, the Ministry of
International Trade and Industry made an inquiry to the
Thermal and Nuclear Power Engineering Society con-
cerning “Development of Technical Standards for Con-
crete Containment Vessels.” In response to the inquiry,
the Committee on the Development of Standards for
Concrete Containment Vessels prepared the report
“Proposed Technical Standards for Concrete Contain-
ment Vessels” in February 1989. The above standards
specify technical standards for the design of concrete
components, the materials used in reinforced and
prestressed concrete, liners, knuckles, body anchors, and
pressure testing, etc. The Committee also presented
proposed guidelines for the verification of the structural
performance and ISI of CCVs as well as for the design
and construction of diaphragm floors in the containment
vessel.
Y. Yamaguchi, T. Kitagawa, H. Tanaka, Y. Koshiro, H. Takahashi, et al. / Journal of Advanced Concrete Technology Vol. 14, 464-474, 2016
The proposed technical standards were incorporated
into the Technical Standards for the Structure of Concrete
Containment Vessels (Ordinance No. 452 of the Ministry
of International Trade and Industry enacted in October
1990), but those for SIT and ISI were not.
In an effort to develop commercial standards in a
systematic manner, the Power Generation System Stan-
dards Committee was established in October 1996 by the
Japan Society of Mechanical Engineers (JSME), and
rules such as “Rules on Design and Construction for
Nuclear Power Plants (JSME S NC1-2001)”, “Rules on
Fitness-for-Service for Nuclear Power Plants (JSME S
NA1-2000)” and “Rules on Welding for Nuclear Power
Plants (JSME S NB1-2001)” were developed under the
Nuclear Power Expert Committee. However, the provi-
sions for class-MC vessels in JSME S NC1-2001 were
applicable only to SCVs and not applicable to CCV.
Therefore, it was necessary to develop technical stan-
dards for CCV as the academic society and association
standards. Thus, the Concrete Containment Vessel
Standards Subcommittee was established in September
2002 under the Nuclear Power Expert Committee, and
“Rules on Concrete Containment Vessels for Nuclear
Power Plants (JSME S NE1-2003)” consistent with
JSME S NC1-2001 were developed. To make the CCV
Code consistent with the international standards, the
Codes were developed based on Boiler and Pressure
Vessel Code, Section III, Division 2, Code for Concrete
Containments formulated by American Society of Me-
chanical Engineers (hereinafter referred to as “ASME
Sec. III, Div. 2”), which was used to prepare Ordinance
No. 452.
In accordance with the policy presented on July 22,
2002 in the report, “Performance Specification in Tech-
nical Standards for Nuclear Power Generation Facilities
and Use of Commercial Standards” prepared by the Nu-
clear Reactor Safety Subcommittee of the Nuclear and
Industrial Safety Subcommittee, Advisory Committee
for Natural Resources and Energy, the Ministerial Ordi-
nance Partially Revising the Ministerial Ordinance of
Establishing Technical Standards for Nuclear Power
Generation Equipment (Ministerial Ordinance No. 68 of
October 31, 2005 of the Ministry of International Trade
and Industry) was put into effect; and, as a result, the
Technical Standards for Structure etc. of Nuclear Power
Generation Equipment (Ministerial Ordinance No. 501
of the Ministry of International Trade and Industry) and
the Technical Standards for Structure, etc. of Concrete
Containment Vessels (Ministerial Ordinance No. 452 of
the Ministry of International Trade and Industry) were
removed and the CCV Code (JSME S NE1-2003) be-
came the technical standards for CCVs after the Code
was endorsed in 2005.
However, the CCV Code (JSME S NE1-2003) pro-
vided technical standards only for the material and de-
sign of CCVs. Subsequently, in 2011, the CCV Code was
revised. Construction, testing and inspection during
construction, the SIT and SIS were added to the CCV
469
Code (JSME S NE1-2014), and the Code was issued in
2015.
Structures in nuclear facilities have been maintained
by nuclear operators. Since there were no unified codes
and guidelines for nuclear facilities, in 2008, the Archi-
tectural Institute of Japan (AIJ) issued the Guidelines for
Maintenance and Management of Structures in Nuclear
Facilities (hereinafter referred to as “the AIJ Mainte-
nance Guidelines”), which can be also used for the Plant
Life Management (hereinafter referred to as PLM), to
provide standards to maintain the required function of
structures in nuclear facilities during the service life.
CCVs were excluded from the application of the AIJ
Maintenance Guidelines for the reason that they are
within the scope of application of JSME’s CCV Code.
Since the CCV Code is applicable only to ISI, subse-
quently, the need arose to develop guidelines for the
maintenance of CCVs as with other structures in nuclear
facilities. Since response to the PLM was required, par-
ticularly for PCCVs, a provision for the maintenance of
PCCVs was added to the Appendix when the AIJ Main-
tenance Guidelines were revised in 2015. The provision
for the maintenance of PCCVs, which does not apply to
ISI, states that the ISI set forth in the CCV Code should
be used as needed.
Thus, domestic standards and guidelines were put in
place nearly 30 years after the first NPP that uses a CCV
went into operation.
5. Development of new technology
5.1 Strength estimation methods based on
non-destructive testing of high-strength con-
crete (Tsuzuki et al. 2014)
A non-destructive testing method based on the rebound
number is often used to evaluate the compressive
strength of concrete. The commonly used formula of the
Society of Materials Science, Japan (JSMS) for strength
estimation and the formula defined in AIJ’s
non-destructive testing manual tend to underestimate the
actual strength of a high-strength concrete structure. The
compressive strength of concrete used in PCCVs is ex-
pected to be more than 50 N/mm2 from the results of
long-term material age strength testing of control
specimens and monitoring specimens prepared at the
time of construction. Therefore, the current formula to
estimate the strength of concrete from the rebound
number is not applicable. To estimate the compressive
strength of concrete used in PCCVs, it is desirable to use
a strength estimation formula that takes into account the
effects of aggregate used in PCCVs, mixing, concrete
placement, and ambient conditions. Methods for esti-
mating the compressive strength of concrete that can be
applied to high-strength mass concrete structures were
evaluated using a mass concrete member simulating a
PCCV. Table 3 shows the preparation of the concrete
used in the full-scale simulated member. Figure 7 shows
a schematic of the simulated member and its shape.
Y. Yamaguchi, T. Kitagawa, H. Tanaka, Y. Koshiro, H. Takahashi, et al. / Journal of Advanced Concrete Technology Vol. 14, 464-474, 2016 470

Table 3 Preparation of concrete for the full-scale simulated member (Tsuzuki et al. 2014).
Unit Quantity (kgf/m3)
Mixing Ratio Percentage of Fine
Water Cement Fly ash
(W/B) Aggregate (%) Fine aggregate Coarse aggregate
Admixture
(W) (B)
50% 49.7 165 264 66 889 921 2.145
40% 46.6 165 330 83 800 937 2.478
30% 44.5 170 454 113 698 888 7.371
Water: groundwater, Cement: moderate-heat Portland cement (density: 3.21 g/cm3)
Fly ash: JIS Type II (density: 2.33 g/cm3, specific surface area: 3890 cm2/g)
Fine aggregate: mixture of crushed sand (density in saturated surface-dry condition: 2.66 g/cm3) and mountain sand (density in
saturated surface-dry condition: 2.59 g/cm3)
Coarse aggregate: crushed tight sand (density in saturated surface-dry condition: 2.70 g/cm3), Admixture: polycarboxylic
acid-based high-performance AE water-reducing agent (standard type)

The non-destructive testing methods used in the
evaluation were the rebound hammer, ultrasonic (surface
wave velocity) and elastic wave methods (P-wave ve-
locity, S-wave velocity). The results obtained by these
methods were compared with the compressive strength
of the core samples taken from the full-scale simulated
member. Figure 7 shows the position where the core
samples were taken and the position where the
non-destructive test was conducted.
A single regression analysis was performed on the
results obtained by each non-destructive test-based
strength estimation method and the compressive strength
of the core samples. The square of the coefficient of
correlation (= r) expressed by an exponential function
was defined as the “percentage of contribution,” and the
correlation between the compressive strength of the core
samples and each non-destructive testing method was
evaluated based on the percentage of contribution. As a
result, the percentage of contribution of the rebound
number was the highest, and the material age had the
least effect on the percentage of contribution. A rebound
number-based strength estimation formula was evaluated
that can estimate the compressive strength (About 30 to
90 N/mm2) of a high-strength mass concrete structure
that meets the conditions for the full-scale simulated
member.
F = 16.4 × exp(2.91×10−2 × R)
where F is estimated compressive strength (N/mm2) and
R is rebound number determined by a rebound hammer.
Figure 8 shows the correlation between the rebound
number and the compressive strength of the core samples.
The figure shows the evaluated strength estimation
formula, the 95% confidence interval, the formula of the
JSMS, and the formula in AIJ’s non-destructive testing
manual. At the lower limit of 95% confidence interval
calculated by the evaluated strength estimation formula,
the compressive strength calculated by the evaluated
method is higher than the one calculated by the formula
of the JSMS and the formula in AIJ’s non-destructive
testing manual when the rebound number is 40 or higher.
Figure 8 also shows the results of the compressive
strength test performed at a material age of 22 years on

1300mm
2325mm :コア採取位置
Core sampling point
Mate- :反発度測定位置
Rebound number measurement point
材齢
rial
age
44週 :表面波速度測定位置
Surface wave velocity measurement point

13
:P波,S波速度測定位置
P-wave and S-wave velocity measurement point
13週
Arrangement
配筋： of reinforcement:
26週
26
横筋：D38@225(かぶり厚:96㎜)
Horizontal reinforcement: D38@225 (covering depth: 96 mm)
2250mm

1年
1 Vertical reinforcement: D38@310 (covering depth: 186 mm)
縦筋：D38@310(かぶり厚:186mm)
2年
2 Placement: winter (mid-January, 2006)
打設：冬期(平成18年1月中旬)
During the formwork installation, 200 mm insu-
反発度測定面およびその裏面以
3年
3
lation was placed on surfaces other than the
外は，型枠時に断熱材(200㎜)を
5
5年 surface for rebound number measurement and its
設置している。
backside.
L C R
*※材齢4年目に，
The top of the simulated member
採取位置
Sampling point was cut off at a material age of 4 years.
上部を切断撤去。

Fig. 7 Schematic and shape of the full-scale simulated member (Tsuzuki et al. 2014).
Y. Yamaguchi, T. Kitagawa, H. Tanaka, Y. Koshiro, H. Takahashi, et al. / Journal of Advanced Concrete Technology Vol. 14, 464-474, 2016 471

the PCCV monitoring specimens (Monitoring Specimen
B) that were fabricated with the same mixing ratio on the
site during the construction of Ohi NPP Units 3 and 4 and
exposed to the same ambient conditions. This indicates
that the evaluated strength estimation formula gives a
rough estimate of the compressive strength of the actual
PCCV.
A full-scale model test was performed to verify the
accuracy of the prestressing force measurement method
based on the change in the amount of oil. Figure 13
shows the test specimen.
In the test, a 10 MN-class VSL system, which is the

5.2 Development of a new prestressing force

measurement system for PCCVs (Abe et al.
2003; Ozaki et al. 2002)
Since the aforementioned lift-off test in ISI requires work
that involves the use of large equipment in a small space,
the test is less efficient and very time-consuming. A more
effective and efficient prestressing force measurement
system was developed.
This measurement system uses a lightweight hydraulic
load cell (hereinafter referred to as a “hydraulic shim
load cell”). The load cell has the same diameter and
thickness as the shim between the bearing plate and the
anchor head and can be installed in the existing PCCVs
without releasing the prestressing force. The load cell
can be split into halves and does not require the removal
of the corrosion protection medium in the end cap when
taking measurements. Figures. 9 and 10 show the hy- Fig. 8 Relationship between rebound number and core
strength.
draulic shim load cell.
The hydraulic shim load cell is permanently installed
and usually serves as a shim. In the measurement of the
prestressing force, the oil is pumped into the hydraulic
shim load cell to produce hydraulic pressure, and the
hydraulic pressure produced when the anchor head is
pressed up by the hydraulic shim load cell is measured by
an external pressure gauge to obtain the residual
prestressing force.
Figure 11 shows the measurement system for the ac-
tual PCCV. The equipment and systems required for the
measurements are a hydraulic shim load cell, a hydraulic
pump, a hydraulic booster cylinder, a displacement
gauge, a pressure gauge, a data logger, and a laptop PC.
All of them are lightweight and easy to carry around.
The change in the amount of oil pumped into the hy-
Fig. 9 Shape of the hydraulic sim load cell.
draulic shim load cell is used to detect the prestressing
force by the hydraulic shim load cell, instead of the
currently used filler gauge. The change in the amount of
oil can be obtained by measuring the stroke displacement
of the hydraulic booster cylinder shown in Fig. 11. This
makes it possible to take measurements with the end cap
on. Figure 12 shows the concept of the prestressing force
measurement based on the change in the amount of oil.
Since the volume of the oil chamber remains constant
until the oil is pumped into the hydraulic shim load cell
and the ram is retracted, the increase in the amount of oil is
equal to the amount of the elastic deformation of the oil
and the hose. After the applied force is exceeded and the
ram starts to retract, the increase in the amount of oil is
significant due to the increase in the volume of the oil
chamber in addition to the aforementioned elastic defor-
mation.
Fig. 10 Hydraulic shim load cell (Abe et al. 2003).
Y. Yamaguchi, T. Kitagawa, H. Tanaka, Y. Koshiro, H. Takahashi, et al. / Journal of Advanced Concrete Technology Vol. 14, 464-474, 2016 472

same as the one used in the actual PCCV, was reproduced Table 4 Results of the tests.
in a concrete block (1.1 × 1.1 × 1.6 m) and a hydraulic Prestressing Evaluated
shim load cell was installed at the tendon anchorage. B／A
force A (MN) load B (MN)
Three levels of prestressing force, 5.5 MN, 6.5 MN and No.1 5.52 5.62 1.02
7.3 MN, were applied in the test. Three measurements 5.5MN No.2 5.55 5.53 1.00
were taken by the hydraulic shim load cell at each load No.3 5.58 5.68 1.02
level and compared with the reference load cell load. No.1 6.42 6.56 1.02
The displacement of the hydraulic booster cylinder 6.5MN No.2 6.37 6.42 1.01
(the amount of oil) increased nearly linearly until the No.3 6.38 6.40 1.00
reference load cell load was exceeded and thereafter No.1 7.34 7.40 1.01
7.3MN No.2 7.38 7.22 0.98
significantly increased. This trend was observed in all
No.3 7.38 7.21 0.98
tests performed, and thus the measurement method based
on the amount of oil was determined to be valid. Table 4
shows the results of the tests. tember 2001 during ISI of Ohi NPP Unit 3. The lift-off
As a result of the tests, the ratio of the evaluated force load is the load of the intersection of the two straight
to the reference load cell load was between 0.98 and 1.02, lines when the data is separated on the point where the
with a maximum evaluation error of 2%. tangent stiffness of each step becomes a half initial
Figure 14 shows an example of measurements of the stiffness, and linear approximation of the first slope and
change in the amount of oil pumped into the lift-off jack the second slope is carried out by the least-squares
and the lift-off load in a lift-off test performed in Sep- method.

Fig. 11 Measurement system for the actual PCCV.

Fig. 12 Concept of the prestressing force measurement based on the change in the amount of oil.
Y. Yamaguchi, T. Kitagawa, H. Tanaka, Y. Koshiro, H. Takahashi, et al. / Journal of Advanced Concrete Technology Vol. 14, 464-474, 2016
Fig. 13 Full-scale model test (Abe et al. 2003).
The difference in the error in the evaluated lift-off load
between the use of hydraulic shim load cell and the use of
filler gauge was 1.2% at a maximum. It was confirmed
that the workability of the load cell was satisfactory.
6. Conclusions
This report describes the maintenance activities mainly
for the prestressed members of the PCCV in Ohi NPP
Units 3 and 4 of The Kansai Electric Power Co., Inc. The
description in this report can be summarized as follows.
(1) As an outline of the PCCV, the required functions,
the basic specifications, and the prestressed system,
etc. of the PCCVs of Ohi NPP Units 3 and 4 of The
Kansai Electric Power Co., Inc. are shown.
(2) For the maintenance management of the PCCV, the
outline of the plan, the observed results and the
evaluation for ISI of the PCCV of Ohi NPP Units 3
and 4 of The Kansai Electric Power Co., Inc. are
shown. The results of the previously performed ISIs
showed that the prestressing force was more than the
design requirement, and that it was all clear with the
results of other visual inspections and of the inspec-
tion of corrosion protection medium. Therefore, we
confirmed that the functions have been maintained
since the service was started.
(3) Changes over the year in the standards and guide-
lines for the maintenance of concrete containment
vessels in Japan are shown. The domestic standards
and guidelines have been put in place nearly 30
years after the first NPP that uses a CCV went into
operation.
(4) The development of new technology for PCCVs is
outlined as follows.
- A rebound number-based strength estimation
formula is proposed that can estimate the com-
pressive strength of PCCVs as a result of the
discussion of non-destructive test-based methods
for estimating the strength of high-strength that
can be applied to high-strength area and mass
concrete structures specific to PCCVs.
- The more effective prestressing force measure-
ment system using a hydraulic shim load cell in
the measurement of the prestressing force was
developed.
473
7.0
6.35 MN
Lift-off jack load ( MN )
6.0
6.34 MN
5.0
Lift-off load by filler gage
Injected oil quantity
4.0 First slope
Second slope
Lift-off load
3.0
0 50 100 150 200 250 300
Projection length of the cylinder （mm）
Fig. 14 Relationship between the lift-off load and the
amount of oil pumped into the lift-off jack (Abe et al.
2003).
The future compliances based on the regulation trends
in Japan are described in the following paragraph.
Under the Ministerial Ordinance for Commercial Nu-
clear Power Reactors concerning the Installation， Op-
eration, etc. (hereinafter referred to as “the Commercial
Reactor Rules”), Japan has the PLM in place, which
requires nuclear facilities that have been in operation for
30 years to perform a technical evaluation of the degra-
dation of components and structures due to aging every
10 years. The program requires a technical evaluation of
the degradation of components and structures due to
aging that are important for safety and hard to repair or
replace so that such a technical evaluation concerning
PLM can determine if they are appropriately managed to
protect against the degradation events to be considered; if
there is any possibility of occurrence and progress of
such events; or if they have sufficient margin to the in-
tegrity during the service life. Accordingly, maintenance
activities for PLM are evaluated and reviewed. The
program also applies to the concrete structures in NPPs,
including the PCCV. In July 2013, the new regulatory
standards for NPPs, which were developed by the Nu-
clear Regulation Authority while taking into account the
lessons learned from the accident at the Fukushima
Daiichi NPP and the suggestions from home and abroad,
were put into effect. Ohi NPP Units 3 and 4 have been in
operation for 24 years and 22 years respectively, as of the
Y. Yamaguchi, T. Kitagawa, H. Tanaka, Y. Koshiro, H. Takahashi, et al. / Journal of Advanced Concrete Technology Vol. 14, 464-474, 2016
end of December 2015. Before the passing of 30 years
from the start of operation, a technical evaluation of
degradation of the PCCV due to aging needs to be per-
formed in accordance with the PLM, taking into account
the results of ISI and the results of technology devel-
opment, and the evaluation needs to be reviewed. The
PCCV has been maintained appropriately as described
above. To ensure the integrity of the PCCV, the mainte-
nance activities will be continued with the understanding
of the purpose of the new regulatory standards.
References
Abe, T., Kawai, I., Ohashi, H., Murazumi, Y. and
Takezaki, S., (2003). “New prestressing force meas-
urement system for prestressed concrete containment
vessels.” SMiRT-17, August, Paper #H03-3.
Ozaki, M., Kawai, I., Ohashi, H., Murazumi, Y. and
Takezaki, S., (2002). “New prestressing force
measurement system for prestressed concrete
474
containment vessels.” The first fib Congress, Session
15, Monitoring, 233-238.
Ozaki, M., Takeuchi, K. and Kitagawa, T., (2004).
“Construction and maintenance of a Nuclear Power
Plant Prestressed Concrete Containment Vessel
(PCCV).” Journal of Prestressed Concrete, 46(6),
41-48. (in Japanese)
Tsuzuki, M., Koshiro, Y., Fushimi, M. and Nakano, M.,
(2014). “Study on estimating strength of high strength
concrete − Experimental studies of several
non-destructive inspection for concrete and concrete
core strength using mock-up mass concrete specimen
at 5 years age −.” AIJ J. Technol. Des. 20(45), 487-490.
(in Japanese)
Yonezawa, K., Imoto, K., Ohba, M., Ikeuchi, T., Kozuma,
M., Murazumi, Y. and Sato, K., (2003) “Analytical
study on structural failure mode of 1/4 PCCV test
model.” SMiRT-17, August, Paper #H02-1.
Journal of Advanced Concrete Technology Vol. 14, 384-396 July 2016 / Copyright © 2016 Japan Concrete Institute 384

Scientific paper

The Strengthening of an ASR-Affected Water Intake Tower in

a Hydro-Electric Dam by Using Post-tensioned Tendons and
the Long-term Monitoring of the Tower
Kazuyuki Torii1*, Tetsuji Kubo2, Chikao Sannoh3 and Maki Kanitani4

Received 1 April 2016, accepted 18 June 2016 doi:10.3151/jact.14.384

Abstract
In the Hokuriku district in Japan, large numbers of concrete structures have been suffering from damage caused by alkali
silica reaction (ASR). Some huge volcanoes are located within this district, and the headwaters of main rivers prompt the
outflow and spreading of volcanic rocks such as andesite, rhyolite and tuff stones, which are the main volcanic reactive
stones causing the serious damage of ASR in the entire area. To solve this problem effective countermeasures such as
repair and strengthening methods should be established.
One intake tower in this area had deteriorated due to ASR, and deformation had occurred as a consequence of ASR ex-
pansion of the concrete. Countermeasures were carefully considered by academic experts, and post-tensioned tendons
were inserted into the intake tower concrete (vertically oriented) so that the power station can continue to operate safely in
the future. This is thought to be the first challenge of its type anywhere in the world for which the deformation of a real
structure caused by ASR expansion must be controlled.
In this paper, the results obtained from laboratory tests using reactive aggregate and the overview of the investigation and
the method of the reinforcement including the effect of the countermeasures will be discussed.

1. Introduction structures has been indeterminate, and a method of con-

trol has not been confirmed. Furthermore, while there
have been reports of structural deformities, methods for
In the Hokuriku district in Japan, large numbers of RC relieving the pressure have been employed, but methods
(Reinforced Concrete) and PC (Pre-stressed Concrete) of controlling the deformity as of yet have not been
bridges have been suffering from the damage caused by confirmed (Shayan et al., 2000).
alkali silica reaction (ASR) (Torii 2010 ; Torii et al. 2012;
Nomura et al. 2012). Some huge volcanoes are located
within this district, and the headwaters of main rivers
prompt the outflow and spreading of volcanic rocks such
as andesite, rhyolite and tuff stones, which are the main
volcanic reactive stones causing the serious damage of
ASR in the entire area. Figure 1 shows typical features
of seriously deteriorated bridge piers as a result of ASR,
where cracking, deformation and ASR gel extrusion
occurred at the same time, and in the most serious cases,
the steel reinforcements also were ruptured at their point
of bending.
In order to clarify the influence on structures caused
by ASR expansion, measurements of ASR expansion
related pressure were conducted (Diamond 1989;
Kawamura et al. 2004; Binal 2004; Multon et al. 2006).
However, the actual amount of deformation on actual

1
Professor, Kanazawa University, Kanazawa, Japan.
*Corresponding author, E-mail: torii@t.kanazawa-u.ac.jp
2
Manager, Hokuriku Electric Power Company Co. Ltd,
Ushijima-machi, Toyama, Japan.
3
Assistant manager, Hokuriku Electric Power Company
Co. Ltd, Ushijima-machi, Toyama, Japan.
4
Hokuriku Electric Power Company Co. Ltd, Fig. 1 Typical features of seriously ASR-deteriorated
Ushijima-machi, Toyama, Japan. bridge piers.
K. Torii, T. Kubo, C. Sannoh and M. Kanitani / Journal of Advanced Concrete Technology Vol. 14, 384-396, 2016 385

Table 1 Results of chemical method according to JIS A1145 and accelerated mortar bar method according to ASTM C1260.
Chemical method (JIS A1145) Mortar bar method (ASTM C1260)
Name of river
Sc*1 Rc*2 Sc/Rc ASR Reactivity Expansion at 14day(%) ASR Reactivity
Kurobe 22 37 0.59 Innocuous 0.137 Innocuous or Deleterious(Unclear)
Jouganji 242 72 3.36 Not Innocuous 0.720 Deleterious
Jinzu 27 41 0.66 Innocuous 0.306 Deleterious
Shou 26 54 0.48 Innocuous 0.139 Innocuous or Deleterious(Unclear)
*1: Amount of Dissolved Silica (m mol/l), *2: Amount of Alkalinity Reduction (m mol/l)

This paper discusses an assessment of alkali-silica 2. Mineralogical properties and Alkali-Silica

reactivity in volcanic gravel from Jouganji River in the
Hokuriku district, which is considered to be the most
reactive in Japan. Secondly we will present the investi-
gation of ASR-caused deformities to a water intake tower
which was constructed by using river aggregates pro-
duced from Jouganji river and the attempts to reinforce
the tower using post-tensioned tendons (Bianchi et al.
1992). Concerning the reinforcements, laboratory test
results indicated that a small restraining stress of 0.2～
0.3N/mm2 could decrease the ASR expansion, and this
was the basis for the determination of the initial stress
(Torii et al. 2000; Ishii et al. 2005). The construction was
carried out in 2011, and in 2012 after applying the ten-
sion, it was determined that the deformity had been ef-
fectively controlled. This case does not only concern the
use of post-tensioned tendons for large scale structural
reinforcement and controlling of deformity, but also has
provided a model for the understanding of ASR deform-
ity in other hydro-engineering structures, and as such has
proven to be very useful (Silveira et al. 1996; Cavalcanti
et al. 1996; Blaikie et al. 1996; Huang et al. 1996; Sei-
gnol et al. 2009; Sellier et al. 2009; Pan et al. 2013).
reactivity of volcanic river aggregates
Mineralogical properties and alkali-silica reactivity of
volcanic river aggregates which had been used in To-
yama prefecture are presented in this chapter. Table 1
shows the results of the chemical method (JIS A1145)
and the mortar bar method (ASTM C1260) used on river
gravel produced in the river of Toyama prefecture. As
shown in this table, gravel of Jouganji river which was
used in the intake tower described in later chapter is the
most reactive one. It is necessary to grasp the character-
istics of each type of river gravel, because the ASR re-
activity for each type is different.
Incidentally, in this region, the ASR suppressing effect
of locally produced fly ash has been confirmed and the
ASR control measures by fly ash have been recom-
mended in recent years (Sannoh et al. 2008; Hashimoto
et al. 2013; Torii et al. 2016).
The photomicrograph of reactive minerals and other
materials in the Jouganji river gravel is presented in Fig.
2. In the petrographic survey on mineralogical properties
of the river gravel, it was found that they were mainly

Pl

PX
PX

Crs
Pl Crs
Op
0.1mm 0.1mm

(a) Opl from altered Andesite: plane polarized light (b) Cristobalite from Andesite: plane polarized light
PX
Gls
Gls
Gls Pl PX
Gls Pl
Trd Trd Pl
Gls
Trd
0.1mm Pl 0.1mm
Gls
(c) Tridymite from Andesite: crossed polarized light (d) Volcanic glass from Andesite: plane polarized light
[Op : opl; Crs : cristobalite; Trd : tridymite; Gls : volcanic glass; Pl : plagioclase; Px : pyroxene]
Fig. 2 Photomicrographs of reactive minerals and non-reactive minerals in the river gravel used in the intake tower taken
with a polarizing microscope observation.
K. Torii, T. Kubo, C. Sannoh and M. Kanitani / Journal of Advanced Concrete Technology Vol. 14, 384-396, 2016
Table 2 Lithology of the Jouganji gravel determined by petrographic observation.
Rock type Vol.%
Granitic rocks 39
Rock fragment
Andesite 36
Basalt 2 Plagioclase, Pyroxene, Volcanic glass, Opaque mineral, Cristobalite
Mineral fragment 23
constituted of granitic rocks and volcanic stones, in
which the reactive minerals in typical andesite stones
were cristobalite and/or tridymite, opl, and small
amounts of volcanic glass, as shown in Table 2. The
composition ratio of 30% andesite stones was roughly
equivalent to the most pessimistic results gained by the
mortar bar or concrete bar test of andesite stones in the
river. Thus, it has become very clear that the reactive
river sand and river gravel contain reactive minerals,
especially opl, the most reactive mineral, and that its
amount may be near the most pessimistic volume. A
recent survey shows that in the bridges and buildings
using this river sand and gravel, a severe ASR has still
occurred after 2010 even in the case of total alkali con-
tent of 2 kg/m3 and less, which is considerably smaller
than the JIS A5308 required value of less than 3 kg/m3.
3. Overview of the investigation
3.1 History of the intake tower
This intake tower is located at a dam in the upstream
region of Jouganji river in Hokuriku District. Figure 3
shows the actual site. 74m3/s of water can pass through
the intake tower and 400 thousands kW can be generated
by using this water at 3 power stations located down-
stream of the tower. This electricity corresponds ap-
proximately to the power needs of about 130 thousand
households. Thus, this intake tower is a critical structure
for the area and the company.
The intake tower is a SRC structure. Figure 4 shows a
cross-sectional view and a side view of the intake tower.
The water level of the dam varies depending on the
Intake tower
High water level
Connecting bridge
Front gate
Fig. 3 Overview of actual site of the intake tower.
386
Main constituents
Plagioclase, Quartz, Hornblende, Biotite, Alkali feldspar, Chlorite, Epidote,
Sphene, Prehnite, Opaque mineral, Pyroxene
Plagioclase, Cristobalite, Tridymite, Volcanic glass, Pyroxene, Opaque min-
eral, Quartz, Opal, Smectite, Biotite, Hornblende, Olivine, Apatite
Plagioclase, Quartz, Alkali feldspar, Biotite, Pyroxene, Hornblende, Chlorite
power demand. Usually it reaches the highest level
around May, and decreases to about 1070m in the sum-
mer months. Then it decreases further to about 1060m in
the winter (February - March), returning to the highest
water level due to snowmelt in the spring. Temperatures
during the summer rise to about 25 degrees, and in winter
they fall to about minus 10 degrees. In the winter, there is
snow cover of about 5m. Therefore, between November
and April it is not possible to go to the site area.
Twenty years after the construction of the tower, an
inclination with concomitant expansion of the structure
was discovered due to the formation of a clearance of
40mm between the top of the intake tower and the con-
necting bridge. After the detection of this separation,
investigations were conducted to find the cause.
3.2 Displacement measurement
Distance from vertical line to both the interior wall and
the exterior wall were measured at each elevation. Fig-
ure 5 shows the results of the investigation of the interior
wall and the exterior wall. Measured distance shows the
average of measured values on the interior wall and the
exterior wall. With the bottom of the tower’s base located
at 1,045 meters above sea level, we noted that from about
an elevation of 1,065 meters the intake tower wall began
to exhibit an inclination in the direction of the lake. At
the top of the tower (elevation 1,090.7 meters above sea
level), there was a deviation from the vertical of 80mm. It
was not clear at the time of construction how accurately
the vertical orientation of the tower wall was, so it is
unclear exactly to what degree the wall’s configuration
has deformed.
Figure 6 shows the horizontal change in distant over
time at the top of the tower, and Fig. 7 shows the vertical
change over time at the top of the tower, which horizontal
and vertical measurements were both conducted from 22
years after construction. During the 8-year period of 22
to 30 years following construction the rate of horizontal
change at the top of the tower was 5.6mm per year. As-
suming that the average rate of change was the same
prior to the commencement of measurements and taking
into account the currently measured change of 80mm, we
estimate that the beginning of this deformation started
about 15 years following construction. During the same
8-year period, the rate of vertical change was calculated
to be 2.1mm per year. Thus, assuming that the deforma-
tion along the vertical axis of the tower also began 15
years after construction and assuming that the rate of
deformation was uniform over time, we estimate that the
K. Torii, T. Kubo, C. Sannoh and M. Kanitani / Journal of Advanced Concrete Technology Vol. 14, 384-396, 2016 387

overall height of the tower at its top has increased by El 1060m after 30 years of the construction. As shown in
33mm on the side of the connecting bridge. Fig. 8, a slightly expansive tendency was observed at the
heights of El 1070m and El 1060m.
3.3 Results of the residual expansion test Therefore, it has been determined that it will be nec-
Figure 8 shows the results of the residual expansion test essary to watch for the future expansion.
according to the JCI DD2 method . Each core was taken
from the side of the tower at the heights of El 1070m and

(Lake side) (Mountain side)

12,000 1,100

2,500 9,500

9,000 500
H-beam or BH-beam
EL.1,090,700
300

3,900 300
1,500

2,200

中間壁柱断面
側壁柱断面
2,000
7,850

2,500

Specification of H-beam and

5,000

H-582×300×12×17
H-582×300×12×17
Built-up-H-beam(BH-beam)
7,500

24,700
① BH-1400×400×16×28mm
H-582×300×12×17
H-588×300×12×20
② BH-1400×400×19×32mm
13@2,500=32,500

6@3,750=22,500
7,500
37,700

③ H-588×300×12×20mm
Ａ
Ａ Ａ
Ａ
FPL-17×350 WPL-12

FPL-20×350 WPL-12

④ H-582×300×12×17mm
7,500

FPL-25×350 WPL-16

FPL-28×350 WPL-16

10,000
62,700

5,500

Ｂ
Ｂ Ｂ
Ｂ
FPL-28×350 WPL-16

FPL-32×350 WPL-19

Auxiliary arranged steel

5,500

8,500

EL.1,051,500
for construction
8@2,750=22,000

Ｃ
Ｃ
Ｃ
Ｃ
EL.1,047,500
8,500

EL.1,045,500
3,750
7,500

FPL-32×400 WPL-19
FPL-28×400 WPL-16

EL.1,043,000
3,750
9,000

3,000

EL.1,038,000
7,500

2,500

19,500

EL.1,034,000
4,500

8,200
5,500
6,000

Direction
3,000
550

EL.1,028,000
2,400 2,200 2@1,050=2,100
7@1,000=7,000 5@900=4,500 6@800=4,800 500

6,500 8,250 2,250

6,500 17,000
Gate metal material North
23,500
(Channel steel:200×90×2～8mm)
Ａ－Ａ Ｂ－Ｂ Ｃ－Ｃ
9,000 Rebar 10,000 11,500

6,500 2,500 (D16～D29) 6,500 3,500 6,500 5,000

①
2,500

2,500

2,500

④ ① ③
7,600

7,600

7,600

②
22,700

22,700

22,700
2,500

2,500

2,500

③ ② ③
7,600

7,600

7,600

①
1,400 2,500

1,400 2,500

1,400 2,500

④ ① ③

5,050 3,400 550 5,050 3,400 1,550 5,050 3,400 3,050

9,000 10,000 11,500

Fig. 4 Cross-sectional view and side view of the intake tower.

K. Torii, T. Kubo, C. Sannoh and M. Kanitani / Journal of Advanced Concrete Technology Vol. 14, 384-396, 2016 388

Measured distance from vertical line

Fig. 5 Investigation of the interior wall and exterior wall.

Horizontal change (mm) summer and autumn

soon after the inclination 86mm

Horizontal change
was discovered
Connecting bridge

(measured in 2007.8)
41mm 5.6mm/year
Intake tower

(22 years after construction)

1993 2000.8 2008.6

(15 years after construction) (30 years after construction)
(year) Time (years)
Fig. 6 Horizontal change in the position of the top of the tower over time.

Vertical change (mm)

Vertical change
Connecting bridge

Intake tower 33mm

15mm
2.1mm/year
(22 years after construction)
1993 2000.8 2008.6
(15 years after construction) Time (years) (30 years after construction)
(year)
Fig. 7 Vertical change in the position of the top of the tower over time.
K. Torii, T. Kubo, C. Sannoh and M. Kanitani / Journal of Advanced Concrete Technology Vol. 14, 384-396, 2016 389

Expansion(%) Expansion(%)
(EL 1070m) (EL 1060m)
0.07 0.07

0.06 0.06
Judgment line Judgment line
0.05 0.05

残存膨張量（％）
残存膨張量（％）
0.04 0.04

0.03 0.03

0.02 0.02

0.01 0.01

0.00 0.00

-0.01 -0.01
0 7 14 21 28 35 42 49 56 63 70 77 84 91 0 7 14 21 28 35 42 49 56 63 70 77 84 91
日 日
Time (days) Time (days)
(immersed in the temperature of 40℃ and the humidity of 95% or more)
Mountain
Lake side
side
Cores were taken from the inside of the mountain side
× Cores were taken from the outside of the mountain side
Cores were taken from the inside of the lake side

inside outside

Cross section of the tower

Fig. 8 Result of the residual expansion test according to JCI DD2.

3.4 Cause of the inclination
Figure 9 shows the condition of the surface of the intake
tower’s exterior wall. On the surfaces of the exterior and
interior walls extensive mapping cracks were observed.
The cracks were approximately 0.2mm in width at 20 to
30 cm intervals and were not severe.
The intake tower concrete (at the levels of EL1035, 1047,
1060, 1070, and 1087) was investigated using an
SEM-EDS. Figure 10 shows an example of an image at
the level of EL1060. This investigation discovered ASR
gel in all samples taken from all locations.
As a result of the research, ASR was observed in the
entirety of the intake tower, causing expansion. Fur-
thermore, from the elevation of 1065m, the tower was
leaning towards the lake. The height of the tower is about
63m, and in the cross section of the tower the water in-
take gates are shown. Due to the gate there is less ex-
posure to sunlight on the lake side. On the mountain side
there is a solid wall. Compared to the lake side, the
mountain side wall has more exposure to sunlight which
is believed to be a large influence on the ease of ASR
development. What’s more, on the mountain side in the
vertical direction, the structure is not reinforced. Thus, it
is believed that this permits a large ASR influence on this
side of the tower.
The amount of ASR expansion is said to be affected by
the percentage of the steel that constitutes the cross sec-
tion. Table 3 shows the value of the cross-sectional area
of the steel (steel, rebar, gate metal material) as a fraction
of the cross-sectional area of the concrete. Comparing
the mountain side wall to the lake side wall, the propor-
tion of steel in the mountain side wall is small, thus it is
presumed that the expansion of the mountain side is
larger than that of the lake side. Comparing the fractional
value of steel in each section, the value of the A-A sec-
tion is small, thus it is presumed that the expansion of the
upper portion is larger than that of the lower portion.
Furthermore, in March the lake level is at its lowest, so

ASR gel

1m
Fig. 9 Condition of the surface of the intake tower’s exte- Fig. 10 Example of the observation of the concrete by
rior wall. SEM-EDS.
K. Torii, T. Kubo, C. Sannoh and M. Kanitani / Journal of Advanced Concrete Technology Vol. 14, 384-396, 2016 390

Table 3 Values of the steel as a fraction of the cross-sectional area of the concrete (%).
Values of the steel as a fraction of the cross-sectional area of the concrete (%)
Section (See Fig. 4)
Lake side wall Mountain side wall
A-A 0.0065 0.0028
B-B 0.0079 0.0038
C-C 0.0079 0.0039

between May and August as the lake fills, the upper
portion of the tower is exposed periodically to more
sunlight followed by lake water inundation which we
believe promotes the formation of ASR gel.
In this way the comparative differences between the
concrete environments and the different degrees of re-
inforcement on the mountain side and lake side result in
unequal amounts of ASR expansion on opposite sides of
the tower. This is the fundamental cause of the tower’s
increasing inclination in the direction of the lake. Figure
11 is a graphic illustration of the primary causes of the
ASR. Incidentally, this figure is obtained by charting the
estimated, qualitative effects taking into account the
basic mechanisms of the ASR (Larive,C., et al., 2000).
4. The Method of the reinforcement
whether or not the connecting bridge might possibly drop.
In fact, the stopper installed to prevent the such bridge
collapses has already been repaired once due to the de-
formation of the tower.
As described above, concerning the overall usage of
the intake tower, there have been no safety or usage
problem up to now except repairing the bridge’s con-
necting members. Although after 30 years of ASR, the
alkalinity and the residual expansion tests do not indicate
the faster development of the tower inclination, in fact
the intake tower change is continuing so we must con-
sider that it may continue. If this occurs the concrete and
metal gates could separate resulting in the gates being
rendered unusable. Thus, even if there is little physical
deformity or expansion, we believe ASR countermea-
sures are valid.

4.2 Chosen countermeasure

4.1 Countermeasures to ASR
There is an affect on the operability of the gate when the
top of the tower has deformed horizontally by 1700mm.
Therefore, even if there are further horizontal deforma-
tions of 5.6mm annually (the deformation rate based on
previous measurements), there will be no problem for
300 years. However, if partial peeling between the gate
attachment and concrete occurs due to the deformation,
there may exist a risk that the gate will be rendered un-
usable that cannot be clearly predicted. If this were to
happen, power generation would be stopped and there
would likely be great difficulty in carrying out repairs.
Thus, we decided to implement the countermeasures.
Another potential problem that has been considered is
Usually measures to control ASR eliminate water source
which is necessary for the reaction. However, being as
this tower stands inside of the lake, its concrete surfaces
are necessarily exposed to water as a matter of course.
Due to this, it is necessary to effect a countermeasure for
all of the interior and exterior surfaces, but it is imprac-
tical to effect a countermeasure completely in the limited
time available. This is the difficulty we face concerning
the elimination of the reaction’s water source. In the case
of this structure, the ASR was discovered 15 years fol-
lowing its construction and close to 30 years later the
reaction is gradually progressing. For now we need not
consider urgent measures, and we must consider that the
gradual halt of the ASR may be possible. With this in

Fig. 11 Graphic illustration of the primary causes of the ASR.

K. Torii, T. Kubo, C. Sannoh and M. Kanitani / Journal of Advanced Concrete Technology Vol. 14, 384-396, 2016 391

Table 4 Applicability of normal repairs and reinforcement.

Aim Example of Method
ASR Suppression Paint application for moisture blocking
Reinforcement using PC anchor
Controlling deformation caused by ASR Bound by PC wire inside the precast concrete or the attachment of a steel plate to the
concrete surface
Rebuilding Complete or partial rebuilding

mind, we think it is proper to consider, even if just a little,
the control of the ASR expansion and deformation.
Table 4 shows the applicability of normal repairs and
reinforcement for this intake tower given the ASR situa-
tion (Torii et al. 2004; Kubo et al. 2004; Daidai et al.
2012). Generally, deformation bound by PC wire inside
the precast concrete or the attachment of a steel plate to
the concrete surface are the countermeasures of choice.
However, as this intake tower has a complicated shape
due to the gate on the front, restraining further
deformation in the horizontal direction is difficult.
Further to this, since the inclination of the intake tower is
due to the vertical expansion of the ASR, this time it was
decided to adopt a method for restraining the further
vertical expansion using the PC steel wire. Of course,
complete rebuilding of the structure would be the most
reliable method. However, since such a construction
project is exorbitantly large in scope and requires
long-term shut-down of the power plant’s operation, that
option was not adopted.
(Pre-stressed Concrete) developed by SEEE
Corporation.
Figure 12 shows the cross-section of the hole after
construction and Fig. 13 shows the situation of the
construction. The anchors are inserted into the drilled
holes. The anchor rods consist of 19 cables, which are
each made from 7 PC steel wires. They have a diameter
of 89mm and a length of 45.7m. The cable has a flexible
structure. That is to say the cable (see cross sectional
figure) is coated with anti-rust oil. This together with the
polyethylene resin is inserted into the outer slide pipe.
The lower parts of the rods are made from steel to fix the
anchors at the bottoms of the holes. Throughout the holes
cement milk is injected. First the upper parts of the rods
Cement milk
Polyethylene resin
Slide pipe
Cable (19×φ12.7mm)

4.3 The reinforcement using post-tensioned

tendons
63.5mm
Concerning the specifications of the PC steel wire, the
(Diameter of anchor)
tension-adjustable SEEE anchor A type, 360TA (19× 165mm
Φ12.7mm, yield point load of 2964kN, drilling diameter (Drilled hole)
of Φ165mm), which was the maximum size available, Fig. 12 Cross-section of the holes drilled for the
was adopted. SEEE anchor is an anchor for PC post-tensioned tendons.

Hydraulic
system

Tendon(φ89mm)
Anchor fixing part
(φ120mm, 4.7m)
During the insertion of the anchor During the tensioning work
Fig. 13 Photographs documenting the construction.
K. Torii, T. Kubo, C. Sannoh and M. Kanitani / Journal of Advanced Concrete Technology Vol. 14, 384-396, 2016
are hydraulically pulled, and then they are fixed in place
with a nut. In other words, it is possible to create a
compression force on the concrete by introducing a
tensioning force to the rods. The anchors adopted this
time have a screw on their tops. They are a type of anchor
with adjustable tensioning force. Additionally, after
considering margins corresponding to potential ASR
development in the future, the largest anchor available in
our country was adopted.
Figure 14 shows the construction location plan view,
and Fig. 15 shows the drilling depth view. Drilling on the
mountain side was, as much as possible, evenly spaced
with 10 places, each place spaced at a distance of about
2m from the next. Furthermore, in order to balance the
whole structure, two anchors were placed on the
lake-side. The anchors’ base positions of immobilization
are set at near EL1045m close to the bottom portion of
the tower structure, and in this way tension was applied
to the entire structure. Thus, 45.7m drilling depth was
needed.
Construction was first carried out for only one of the
candidate locations in 2009. This was a test construction
for the purpose of confirming the concrete properties and
the workability of anchors in advance. Then, the re-
maining construction for the rest of the eleven locations
was carried out in 2011. Drilling was accurate to within ±
0.5°. Tension work was carried out in the spring of 2012.
The initially applied force was set at 1500kN (50% of the
yield point load, 0.26N /mm2). The value of 0.26MPa
was derived from the following calculation.
Tension per anchor : 1500kN
Total number : ten
The cross-sectional area of the intake tower’s
mountain side : 2.5 × 22.7m = 56.75m2
Thus, 1500 × 10 ÷ 56.75 = 0.26Mpa. In the history of
6.5m 2.5m
Metal gate about 2m
plate
Mountain
Lake side side
Front gate
Fig. 14 Top view of the locations of the holes for the
post-tensioned tendons (holes represented by red dots).
392
research to date concerning the expansive pressures of
ASR gels and the required forces required to suppress
that expansion there are reports about such topics. (e.g.,
Kawamura, M.)
According to some results of the investigations,
pressure of 0.2 ～ 0.3MPa is said to be effective to
suppress the expansion due to ASR (Torii et al. 2000;
Ishii et al. 2005; Okuyama, et al. 2007; Ishii et al. 2008).
However, the actual phenomena observed in the field
have been much more complex. To this end we are
conduct creep tests on cores taken from the intake tower
by introducing a small, long-term load to the test sample
in order to confirm the long-term influences of such
loads on deformability(Torii et al. 2010).
Furthermore, the behavior of the actual structure is
unclear. Thus, the monitoring is being continued, and,
further to the behavior observed, we will adjust the
anchors’ introduced forces accordingly.Incidentally,
when the tension in the longitudinal direction is
introduced, there is a possibility that deformation will
proceed in the transverse direction (Gravel et al. 2000),
so follow-up measurements and visual inspections of the
structure’s appearance become increasingly important.
5. Testing of the boring core concrete
5.1 Compressive strength and static modulus of
elasticity
Figure 16 shows the test results for the compressive
strength of a core sample from the tower, and Fig. 17
shows the test results for static modulus of elasticity.
Please note that in Figs.16 and 17, the green outline of
the tower on the left illustrates the relative locations in
the tower structure where the various compressive
strength measurements were taken. The compressive
EL 1,090.7m
45.7m Tendon
EL 1,066m
EL 1,056m
EL 1,045m
Anchor fixing part
Fig. 15 Side view indicating the depth of the holes for the
post-tensioned tendons.
K. Torii, T. Kubo, C. Sannoh and M. Kanitani / Journal of Advanced Concrete Technology Vol. 14, 384-396, 2016 393

strength results were on average higher than the planned
average standard strength(σ28:24N/mm2). Also, the re-
sults for compressive strength and static modulus of
elasticity were both even higher at the 1,090 meter level.
The compressive strength of the core from the intake
tower’s center showed a relatively high value. This is
because the restraining force of the steel is stronger at the
center of the tower.
Although variations of the compressive strength and
the static modulus of the concrete in which ASR occurs
tends to be large, the durability of the structure is be-
lieved to be no problem, because the actual concrete
interior is bound with re-bars. In the lower part of the
tower below 1066m, the fact that the static modulus of
elasticity is reduced indicates that ASR should have
progressed. However, the constraining conditions are
significant, and we believe this is why not so much ex-
pansion was observed. Furthermore, it is believed that an
additional reason why the progress of the ASR in the
lower part of the tower below 1066m differs from the
upper part is the lower part is under water most of the
year.
5.2 Observation of the thin section by the po-
larizing microscope
Some thin pieces made from cutting the core (40mm ×
25mm) were observed under a polarizing microscope.
These pieces were made one by one, each from the core
in the vicinity of EL1086.5m, EL1080.6m, EL1076.1m,
EL1071.3m, EL1068.4m, EL1061.9m, EL1057.1m,
EL1052.4m, EL1050.5m and EL1046.4m. Figure 18
shows an example of a micrograph (EL1061.9m). As a
result, since cracks through the fine aggregate were
observed in slices from all the elevations, it has become
clear that ASR has occurred throughout the whole
interior of the intake tower.

Average 30.1 N/mm2

1100

1090
(m)

1080
Elevation

1070
標高(m)

1060

1050

1040

1030
0 10 20 30 40 50
圧縮強度(N/mm2)
Compressive strength (N/mm2)
Fig. 16 Compressive strength measurements of the cores taken at various locations in the tower.

Average 18.6k N/mm2

1100

1090
Elevation (m)

1080

1070
標高(m)

1060

1050

1040

1030
-10 0 10 20 30 40 50
静弾性係数(kN/mm2)
Static modulus of elasticity (N/mm2)
Fig. 17 Static modulus of elasticity measurements of the cores taken at various locations in the tower.
K. Torii, T. Kubo, C. Sannoh and M. Kanitani / Journal of Advanced Concrete Technology Vol. 14, 384-396, 2016 394

sub-zero temperatures in winter, standard measuring

devices cannot be used (Jensen 2000; Salles 2012; Zoilo
2012). Furthermore, due to the deep snow drifts in the
Andesite area, it is not possible to go to the site in the winter. For
this reason, measurements of displacement are conducted
between spring and autumn.
In these figures, if the horizontal displacement had
ASR gel continued to progress in the same manner as had been
shown through previous measurements between June
2012 and November 2015, it is estimated that the hori-
zontal displacement at the top of the tower should have
3mm progressed an additional 16mm. However, after coun-
Fig. 18. Reactivity of Andesite in a micrograph (EL termeasures, only 5mm of further progression in the
1061.9m) : plane polarized light. horizontal displacement were observed. Thus, we con-
firmed that the suppression effect of the countermeasures
was the reduction of further horizontal displacement by
6. Effect of the countermeasure 11mm, which corresponds to 3mm per year. In brief,
horizontal changes to the top of the intake tower after the
6.1 The measurement of the displacement countermeasure was implemented have decreased.
Figure 19 shows the horizontal change in distant over As described above, although the horizontal dis-
time at the top of the tower, and Fig. 20 shows the ver- placement greatly suppressed, the suppression of the
tical change over time at the top of the tower, including vertical displacement appears to be less. Essentially the
after the countermeasure was implemented. As described horizontal displacement is due to the difference between
above, because of the severe environmental conditions at the vertical expansions on the mountain side and on the

Horizontal change (mm)

After countermeasures
120
120
were implemented
Horizontal change Before
Connecting 102mm
100
100
bridge
97mm 1.4mm/year
80
80

Intake tower 4.7mm/year (2012.6)

60
60
41mm
40
40

20
20

00
2000.8
H12.8.10 H13.9.14 H14.10.19 H15.11.23 H16.12.27 H18.1.31
2008.8
H19.3.7
H20.4.10 2015.11
H21.5.15 H22.6.19 H23.7.24 H24.8.27 H25.10.1 H26.11.5 H27.12.10

Time (years)
Fig. 19 Horizontal change in distant over time at the top of the tower.

Vertical change (mm)

60
60
After countermeasures
50 Vertical change were implemented
50 Connecting
bridge
Before
39mm 42mm
40
40
1.0mm/year
30
30 Intake tower
2.0mm/year (2012.6)
20
20
15mm
10
10

00
2000.8
H12.8.10 H13.9.14 H14.10.19 H15.11.23 H16.12.27 H18.1.31 2008.8
H19.3.7
H20.4.10 2015.11
H21.5.15 H22.6.19 H23.7.24 H24.8.27 H25.10.1 H26.11.5 H27.12.10

Time (years)
Fig. 20 Vertical change in distant over time at the top of the tower.
K. Torii, T. Kubo, C. Sannoh and M. Kanitani / Journal of Advanced Concrete Technology Vol. 14, 384-396, 2016
lake side of the tower. Although the results of this rein-
forcement have had a small inhibiting effect on the
amount of vertical displacement, the difference in verti-
cal expansion rate between the mountain side and the
lake-side of the tower has been reduced, and this has
effectively suppressed the horizontal displacement.
Please note that the primary purpose of the rein-
forcement is to suppress the horizontal deformation.
Therefore, it can be said that the effect of the
countermeasure has been confirmed.
7. Concluding remarks
In this paper the assessment of alkali-silica reactivity of
volcanic gravel from Jouganji river in the Hokuriku
district in Japan, which is considered to be the most
reactive one in Japan, has been introduced. Secondly, the
investigation of ASR-caused deformities to a water in-
take tower which was constructed by using river aggre-
gates produced from Jouganji river and the attempts to
reinforce the tower using post-tensioned tendons has
been introduced. Although an inclination was confirmed
on the intake tower, it has been shown that this is because
of the difference in the ASR expansion due to the
difference in environmental conditions and structural
conditions. As a method for controlling the deformation
of the intake tower caused by ASR, the method of using a
PC anchor was adopted. Consequently it was possible to
effectively suppress the deformation from the applied
force of 0.2 ~ 0.3N/mm2.
References
Bianchi, R. H. and Bruce, D. A., (1992). ”The use of
post-tensioned tendons on Stewart Mountain dam,
Arizona; A case study involving precision drilling.”
Second Interagency Symposium on Stabilization of
Soils and Other Materials, Metairie, LA, 15 p.
Binal, A., (2004). ”A new experimental method and
device for measuring alkali silica gel pressure in
mortar.” Proc. 12th Inter. Conf. on Alkali-aggregate
Reaction in Concrete, 266-272.
Blaikie, N. K. and Bowling, A. J., (1996). “The
assessment and management of alkali-silica reaction
in the gordon river power development intake tower.”
Proc. 10th Inter. Conf. on Alkali-aggregate Reaction
in Concrete, 500-507.
Cavalcanti, A. and Silveira, J., (1989). “Investigation on
the Moxoto powerhouse concrete affected by
alkali-silica reaction.” Proc. 8th Inter. Conf. on
Alkali-aggregate Reaction in Concrete, 797-802.
Daidai, T., Andrade, O. and Torii, K., (2012). ”The
maintenance and rehabilitation techniques for
ASR-affected bridge piers with fracture of steel bars.”
Proc. 14th Inter. Conf. on Alkali-aggregate Reaction
in Concrete, 10 p.
Diamond, S., (1989). ”ASR-another look at
mechanisms.” Proc. 8th Inter. Conf. on Alkali-
aggregate Reaction in Concrete, 83-94.
395
Grvel, C., Ballivy, G., Khayat, K., Quirion, M. and
Lachemi, M., (2000). ”Expansion of AAR concrete
under triaxial stress: Simulation with instrumented
concrete block.” Proc. 11th Inter. Conf. on
Alkali-aggregate Reaction in Concrete, 959-968.
Hashimoto, T. and Torii, K., (2013). ”The development
of highly durable concrete using classified fine fly ash
in Hokuriku district.” Journal of Advanced Concrete
Technology, 11, 312-321.
Huang, M. and Pietruszczak, S., (1996). ”Numerical
analysis of concrete structures subjected to
alkali-aggregate reaction.” Mechanics of
Cohesive-frictional Materials, 1, 305-319.
Ishii, K., Okuda, Y. and Torii, K., (2005). “ASR
mitigation-effect of repair and retrofitting methods
against ASR damaged reinforced concrete columns.”
Concrete Journal, 43(7), 42-50. (in Japanese)
Ishii, K., Okuda, Y., Wasada, S. and Torii, K., (2008).
“Countermeasure to the ASR-affected bridge piers.”
Concrete Journal, 46(1), 55-60. (in Japanese)
Jensen, V., (2000). “In-situ measurement of relative
humidity and expansion of cracks in structures
damaged by AAR.” Proc. 11th Inter. Conf. on
Alkali-aggregate Reaction in Concrete, 849-858.
Kawamura, M. and Iwahori, K., (2004). ”ASR gel
composition and expansive pressure in mortars under
restraint.” Cement and Concrete Composites, 26,
47-56.
Kubo, Y., Yamanashi, T. and Torii, K., (2004). “Study on
repair method with FRP sheet for spalled concrete
structure caused by alkali-silica reaction.” Proc. 12th
Inter. Conf. on Alkali-aggregate Reaction in Concrete,
1254-1261.
Larive, C., Laplaud, A. and Coussy, O., (2000). “The role
of water in alkali-silica reaction.” Proc. 11th Inter.
Conf. on Alkali-aggregate Reaction in Concrete,
61-69.
Multon, S. and Toutlemonde, F., (2006). ”Effect of
applied stress on alkali-silica reaction-induced
expansions.” Cement and Concrete Research, 36,
912-920.
Nomura, M., Komatsubara, A., Kuroyanagi, M. and Torii,
K., (2012). “Evaluation of the residual expansion of
cores due to alkali-silica reaction in Hokuriku
district.” Proc. 14th Inter. Conf. on Alkali-aggregate
Reaction in Concrete, 10 pages
Okuyama, K., Ishii, K., Okuda, Y. and Torii, K.,
(2007). ”The strengthening and monitoring of an
ASR-affected bridge footings.” Concrete Research
and Technology, 29 (1), 1263-1268. (in Japanese)
Pan. J., Feng. Y. T., Jin. F. and Zhang. C.,
(2013). ”Numerical prediction of swelling in concrete
arch dams affected by alkali aggregate reaction.”
European Journal of Environmental and Civil
Engineering, 17 (4), 231-247.
Salles, F., Pinfari, J., Kuperman, C., Mizumoto, C. and
Bernardes, H., (2012). ”Jaguari hydropowerplant-
evaluation, diagnosis and control of a structural
K. Torii, T. Kubo, C. Sannoh and M. Kanitani / Journal of Advanced Concrete Technology Vol. 14, 384-396, 2016
affected by alkali-aggregate reaction.” Proc. 14th Inter.
Conf. on Alkali-aggregate Reaction in Concrete, 10
pages
Sannoh, C. and Torii, K., (2008). ”Suppressing effect of
various mineral admixtures on combined deterioration
caused by ASR and chloride attack.” Proc. 13th Inter.
Conf. on Alkali-aggregate Reaction in Concrete, 9
pages.
Seignol, J. F., Boldea, L. I., Leroy, R., Godart, B. and
Hammerschlag, J. G., (2009). ”Hydro-power structure
affected by alkali-aggregate reaction: a case-study
involving numerical re-assessment.” International
Conference on Computational Technologies in
Concrete Structures, 21p.
Sellier, A., Bourdarot. E., Multon. S. and Grimal. L. E.,
(2009). ”Combination of structural monitoring and
laboratory tests for assessment of alkali-aggregate
reaction swelling: Application to Gate structure dam.”
ACI Materials Journal, 106 (3), 281-290.
Shayan, A., Wark, R. E. and Moulds, A., (2000).
“Diagnosis of AAR in Canning dam, characterisation
of the affected concrete and rehabilitation of the
structure.” Proc. 11th Inter. Conf. on Alkali-aggregate
Reaction in Concrete, 1383-1392.
Silveira, J., Degaspare, J. and Cavalcanti, A., (1989).
“The opening of expansion joints at the Moxoto
powerhouse to counteract the alkali-silica reaction.”
Proc. of 8th Inter. Conf. on Alkali-aggregate Reaction
in Concrete, 747-751.
Torii, K., Kumagai, Y., Okuda, Y., Ishii, K. and Sato, K.,
396
(2000). “Strengthening method for ASR affected
concrete piers using prestressing steel wire.” Proc.
11th Inter. Conf. on Alkali-aggregate Reaction in
Concrete, 1225-1233
Torii, K., Sannoh, C., Kubo, Y. and Ohashi, Y., (2004).
“Serious damages of ASR affected RC bridge piers
and their strengthening techniques.” Proc. 12th Inter.
Conf. on Alkali-aggregate Reaction in Concrete,
1283-1288.
Torii, K., Inagaki, T., Wakita, K. and Sannoh, C., (2010).
“Creep behavior of ASR deteriorated concrete under
the humid environment.” Concrete Research and
Technology, 32(1), 971-976. (in Japanese)
Torii, K., (2010). “The characteristic feature of fracture
of steel reinforcement in ASR-deteriorated concrete
structures.” J. Corrosion Engineering, 59 (4), 59-65.
Torii, K., Prasetia, I., Minato, T. and Ishii, K., (2012).
“The feature of cracking in prestressed concrete bridge
girders deteriorated by alkali-silica reaction.” Proc.
14th Inter. Conf. on Alkali-aggregate Reaction in
Concrete, 10 p.
Torii, K., Hashimoto, T., Kubo, T. and Sannoh, C.,
(2016). ”The alkali-silica reactivity of volcanic
aggregates and its preventive measure using fine fly
ashes.” Proc. 14th International Conference on
Structural Engineering and Construction, 310-318.
Zoilo, C., Bernardes, H., Salles, F. and Kuperman, C.,
(2012). “Monitoring of displacements due to
alkali-aggregate reaction.” Proc. 14th Inter. Conf. on
Alkali-aggregate Reaction in Concrete, 8 p.
Journal of Advanced Concrete Technology Vol. 14, 408-420 August 2016 / Copyright © 2016 Japan Concrete Institute 408

Scientific paper

Numerical Strategy for Forecasting the Leakage Rate of Inner

Containments in Double-Wall Nuclear Reactor Buildings
Mehdi Asali1*,2, Bruno Capra3, Jacky Mazars4 and Jean-Baptiste Colliat5

Received 9 April 2016, accepted 31 July 2016 doi:10.3151/jact.14.408

Abstract
In the context of life extension of nuclear reactor buildings while ensuring safety and regulatory requirements, a nu-
merical strategy is proposed to compute and forecast the air leakage rate evolution of inner containments in double-wall
reactor buildings under standard long-term operation. In order to reduce the numerical cost of such complex computa-
tions, the proposed strategy is based on a coarse but adapted mesh together with a chained weakly-coupled thermo-
hydro-mechanical modeling. The total leakage rate is then computed with a specially-designed tridimensional finite
element based on a non-linear thermal analogy.
This methodology is used to model the behavior of the VeRCoRs mock-up, a simplified nuclear reactor building at scale
1:3, built and monitored by EDF. Results obtained until the first pre-operational pressurization test have been discussed
in a dedicated benchmark organized by EDF. The proposed methodology provides delayed strains and leakage results in
good agreement with available experimental data. A blind prolongation until the first decennial test of the mock-up is
presented and analyzed.

1. Introduction expensive and time-consuming mitigation and repair

The containment building represents the third and last
protection barrier of nuclear reactor buildings (NRB),
after the fuel sheath and the casing of the primary cool-
ant system. Thus, it plays a major role for ensuring the
containment of radioactive elements in case of an acci-
dent and potential failure of the first two barriers.
This study concerns the double-wall reactor building
technology without steel liner, which represents for in-
stance 24 out of the 58 assets of EDF (French nuclear
operator). In that case, most of the leak tightness is en-
sured by an inner containment in prestressed and rein-
forced concrete. Ageing mechanisms in concrete and
prestressing losses are phenomena, among others, that
could modify this tightness capacity over time. To en-
sure safety, internal leakage rate (ILR) tests are per-
formed every ten years to measure the air leak rate of
NRB and to check it stays below an imposed regulatory
value (1.5 % of the internal volume per day). If not,
1 ior characterization. The concrete composition is also
R&D Consultant, OXAND France, Avon-Fontaine-
bleau, France. *Corresponding author,
E-mail: mehdi.asali@oxand.com
2
Doctoral student, Univ. Lille, CNRS, Centrale Lille,
Arts et Métiers Paris Tech, FRE 3723 – LML – Labora-
toire de Mécanique de Lille, Lille, France.
3
Technical and Innovation Manager, OXAND Group,
Avon-Fontainebleau, France.
4
Professor Emeritus, Grenoble Institute of Technology,
Grenoble, France.
5
Professor, Dept. of Mechanics, Univ. Lille, CNRS,
Centrale Lille, Arts et Métiers Paris Tech, FRE 3723 –
LML – Laboratoire de Mécanique de Lille, Lille, France.
techniques must be set up and outage of the tested unit
may be extended, leading to important losses of income.
In the current context of life extension of its assets
beyond the initially designed 40 years of operation,
EDF has built an experimental containment vessel
mock-up at scale 1:3 within the frame of the VeRCoRs
project (VErification Réaliste du COnfinement des Ré-
acteurS, which means realistic verification of reactor
containments, see https://fr.amiando.com/EDF-vercors-
project.html). Considering that drying of concrete is the
main phenomenon governing the leakage rate evolution,
the scale of the mock-up accelerates the ageing of the
structure by a factor 9 compared to a full-size NRB,
enabling EDF to have access to data representative of a
60-year-old NRB in 2021.
The behavior of the mock-up is finely monitored
since the beginning of its construction with more than
500 sensors and 2 km of fiber optic cables acquiring
daily measurements by default and hourly measure-
ments during ILR tests. More than 1000 samples of on-
site concrete are or will be used for material and behav-
available for extra tests on specimens conducted by
laboratories.
From January to October 2015, EDF has proposed a
first international benchmark dedicated to the behavior
of the mock-up until its first ILR test. Three main
themes have been benchmarked:
- Early age, with the prediction of the gusset behavior
from pouring to ten months;
- Containment history, with the prediction of strains,
stresses and cracking in the whole structure at the end
of prestressing and at the end of the 5.2 bar absolute
pressurization plateau;
M. Asali, B. Capra, J. Mazars and J-B. Colliat / Journal of Advanced Concrete Technology Vol. 14, 408-420, 2016
- Leakage, with the prediction of the total air leakage
rate at the end of the pressurization plateau.
The results of the benchmark (Corbin and Garcia
2016) have been presented and discussed during a dedi-
cated workshop, held beginning of March 2016 at EDF
R&D center “Les Renardières”, in France.
To predict accurately the air leakage rate of NRB, the
coupling of thermal, hydric and mechanical (THM) ef-
fects has to be taken into account. Although a wide
range of more or less sophisticated models are already
available to represent each physical phenomenon occur-
ring in concrete (such as drying, delayed strains, crack-
ing or evolution of permeability), global strategies aim-
ing at computing the final leakage rate of a full-size
construction are less common. Indeed, a good compro-
mise between the refinement of discretization, the com-
plexity of behavior laws and the strength of couplings
has to be reached to tackle the global problematic of air
leakage rates for industrial purposes. From the point of
view of computational cost, this leads either to fully-
coupled non-linear strategies (Wang and Hutchinson
2005; Nicklash et al. 2005; Jason and Masson 2014)
only applicable to structural elements or to full-structure
computations based on weakly-coupled elastic strategies
and post-treatments with the help of design codes for-
mulae (Mozayan et al. 2012). In those strategies, the
representation of localized cracking is often one of the
most complex steps.
In that context, the present paper introduces a nu-
merical strategy aiming at forecasting the leakage rate
evolution of inner containments during operation. It
focuses on a chained weakly-coupled THM modeling
used to represent, at lower numerical cost, the physical
behavior of the structure (Fig. 1). First, the models im-
plemented for each physical phenomenon are described,
with a 3D finite element (FE) specially designed for air
leakage computation through unsaturated porous and
cracked concrete. Then, the proposed strategy is applied
to the VeRCoRs mock-up. Main obtained results are
analyzed and compared with experimental data when
available and benefit from discussions and feedback
from the first VeRCoRs benchmark.
2. Modeling principles of the numerical
strategy
After having justified (Asali et al. 2014) that the main
assumption of a weak coupling between THM effects is
valid for inner containments under standard operation, a
1. Temperature 2. Saturation 3. Strains
(section 2.1) (section 2.2) (section 2.3)
4. Air Leak
(section 2.4)
Fig. 1 Representation of the THM strategy.
409
3D numerical strategy is proposed to compute the air
leak rate. The underlying philosophy is to use simple
physically representative models to describe accurately
those complex structures. At that stage, no early age
behavior of concrete is considered.
Despite the weak coupling of phenomena, computa-
tions are chained in the following order to take into ac-
count their relative influence.
2.1 Thermal model
First, the spatiotemporal temperature field T ( x, t ) is
computed (in K) with the classical non-linear heat equa-
tion,
∂T ( x, t )
ρb C p + ∇ ⋅ [−λ (T )∇T ( x, t )] = 0 (1)
∂t
where ∇ is the differential “nabla” operator ( ∇ being
the gradient operator and ∇ ⋅ the divergence operator),
ρb is the concrete density (in kg·m-3), C p is the con-
crete thermal capacity (J·kg-1·K-1) and λ is the concrete
thermal conductivity field (in W·m-1·K-1).
2.2 Hydric model
In a second step, the water saturation field Sl ( x, t ) is
computed (values between 0 and 1). Considering that
drying of concrete is mainly due to liquid water transfer,
the hydric problem reduces to a single diffusion equa-
tion in (Eq. 2a) (Verdier 2001). Thermal dependency of
the hydric diffusion coefficient D ( Sl ( x, t ), T ( x, t )) is
modeled with an Arrhenius-type law as proposed by
Granger (1995) in (Eq. 2b). For the sake of concision in
all equations, the notation D( Sl , T ) for fields depending
on other computed fields is now used.
∂Sl ( x, t )
+ ∇ ⋅ [ D ( Sl , T )∇Sl ( x, t )] = 0 (2a)
∂t
E ⎛ 1 1⎞
l
K int krl ( Sl ) ∂Pc T ( x, t ) Ra ⎜⎝ T ref − T ⎟⎠
D ( Sl , T ) = e (2b)
μl φ ∂Sl T ref
l
where K int is the water intrinsic permeability of con-
crete (in m2), krl is the water relative permeability field
(values between 0 and 1), μl is the water viscosity (in
Pa·s), φ is the concrete porosity, Pc is the capillary
pressure field (in Pa), T ref is the reference temperature
(in K) at which all hydric properties of (Eq. 2) are de-
fined or measured, Ea is the thermal activation energy
of concrete (in J·mol-1), R is the universal gas constant
(in J·mol-1·K-1).
The water retention model of Van Genuchten (1980)
is used to define the evolution of the capillary pressure
field with the saturation degree,
Pc ( Sl ) = Pr ( Sl−1/ m − 1)
1n
(3)
where Pr (in Pa) and n (> 1) are model parameters and
m = 1 − 1/ n . Mualem’s law (Mualem 1976) gives the
evolution of the water relative permeability field with
M. Asali, B. Capra, J. Mazars and J-B. Colliat / Journal of Advanced Concrete Technology Vol. 14, 408-420, 2016 410

saturation, evolution of creep strains with time (Eqs. 11, 13) and
2
biaxial creep effects with a creep Poisson’s ratio differ-
krl ( Sl ) = Sl ⎡⎢1 − (1 − Sl1/ m ) ⎤⎥
m
(4) ent from elasticity (Eqs. 11, 12). Evolution laws of the
⎣ ⎦ model are the following,

ε bc ( x, t ) = ε kv ( x, t ) + ε am ( x, t ) (9)
2.3 Mechanical model
The third step of the methodology aims at computing σ α ( x, t ) = α bc σ + σ (10)
+ −
the total strain state of concrete. Prestressing losses in
tendons are modeled according to ETC-C formulae ηˆam (t )ε am = (1 +ν bc )σ α −ν bc tr (σ α )I 3 (11)
(AFCEN 2010) with a perfect bond between concrete
and steel cables. Passive reinforcement bars are not ex- ηˆkv ε kv + kˆkv εkv = (1 + ν bc )σ −ν bc tr (σ )I 3 (12)
plicitly taken into account.
In order to model delayed effects in prestressed con- where ε kv is the reversible basic creep strain field, ε am
crete over time, the concrete total strain tensor field is the non-reversible creep strain field, σ α is the stress
ε ( x, t ) is split into five main components, field taking into account a decrease or an increase of
creep strain in tension compared to compression
ε ( x, t ) = ε el + ε bc + ε dc + ε ds + ε th (5)
through the use of the coefficient α bc and vbc is the
where ε el is the elastic strain tensor, ε bc is the basic basic creep Poisson’s ratio. Moisture and temperature
creep tensor, ε dc is the drying creep tensor, ε ds is the impact creep strains as follows,
drying shrinkage tensor and ε th is the thermal expan-
kam t ⎛E ⎛1 1 ⎞⎞
sion tensor. ηˆam (t , T , Sl ) = exp ⎜ a ⎜⎜ − ⎟⎟ ⎟ (13)
Sl ⎜ R ⎟
⎝ ⎝ T Tref ⎠⎠
(a) Thermal expansion model
The thermal expansion tensor is considered as isotropic ⎛ Ew ⎛ 1 ⎞⎞
η kv 1
and proportional to the temperature, ηˆkv (T , Sl ) = exp ⎜ a ⎜ − ⎟⎟ ⎟ (14)
Sl ⎜ ⎜ ⎟
⎝ R ⎝ T Tref ⎠⎠
ε th ( x, t ) = α th (T ( x, t ) − T ( x, 0))I 3 (6)
k ⎛ Ew ⎛ 1 1 ⎞⎞
where α th is the thermal expansion coefficient (in K-1) kˆkv (T , Sl ) = kv exp ⎜ a ⎜ − ⎟⎟ ⎟ (15)
Sl ⎜ ⎜ ⎟
and I 3 is the 3D identity matrix. ⎝ R ⎝ T Tref ⎠⎠
(b) Drying shrinkage model where kam is the stiffness associated to the non-
The drying shrinkage tensor is considered as isotropic reversible ageing dashpot (in Pa), kkv is the stiffness
and is proportional to the saturation degree, associated to the reversible creep strains (in Pa), η kv is
the viscosity associated to the reversible creep strains
ε ds ( x, t ) = κ ds ( Sl ( x, 0) − Sl ( x, t ))I 3 (7) (in Pa·s) and Eaw is the thermal activation energy asso-
ciated to water viscosity (in J·mol-1).
where κ ds is the drying shrinkage coefficient (without
unit).
(e) Elastic-damageable model
Isotropic damage evolution is described with the μ-
(c) Drying creep model
model (Mazars et al. 2015) in order to take into account
Concrete’s total creep is higher when drying compared
the unilateral effect (crack closing and reopening) oc-
to basic creep in autogenous conditions (Pickett 1942).
curring with cycling inflating tests. Main evolution laws
In order to take this Pickett’s effect into account, the
are,
drying creep strain is considered as proportional to the
stress and hydric states of concrete, E ε el ( x , T ) = (1 + v)σ − vtr(σ )I 3 (16)
ε dc ( x, t ) = κ dc Sl ( x, t ) σ ( x, t ) (8)

where κ dc is the drying creep coefficient (in Pa-1), and

σ is the effective stress tensor field (in Pa). Kkv ηam(t)

(d) Basic creep model σ ηkv σ

Basic creep is described through a Burger’s rheological
model (Fig. 2) modified by Hilaire (2014) in order to
take into account a possible dissymmetry between creep
in tension and in compression (Eq. 10), a logarithmic εkv εam
Fig. 2 Hilaire’s model for basic creep.
M. Asali, B. Capra, J. Mazars and J-B. Colliat / Journal of Advanced Concrete Technology Vol. 14, 408-420, 2016
σ ( x, t ) = (1 − d )σ ( x, t )
(1 − A)Y0 B (Y −Y ( ε ))
d ( x, t ) = 1 − − Ae 0 eq
Y (ε eq )
ε eq ( x, t ) = ε el + χ (ε bc + ε dc )
where E is the Young’s modulus of concrete (in Pa), v
its Poisson’s ratio, d is the scalar effective damage
variable field (values between 0 and 1, a non-null value
meaning active cracking either in tension or compres-
sion), σ is the total stress tensor field (in Pa), A and
B are coefficients depending on the stress state of con-
crete and defined from post-peak responses of concrete
in pure tension and pure compression, Y is a driving
variable for damage evolution depending on the equiva-
lent strain tensor ε eq , Y0 is the initial threshold of vari-
able Y and χ is a coupling coefficient between elastic
and total creep strains (value between 0 and 1).
To avoid mesh sensitivity due to the softening behav-
ior of concrete, parameter B is adjusted with element
sizes (Hillerborg et al. 1976).
2.4 Air leakage model
Air leakage computation is the final step of the method-
ology. Cracks within concrete being a localized but im-
portant path for leakage, a dedicated 3D finite element
(FE) has been developed to consider an equivalent
cracking state without explicitly meshing the defects. Its
principle, shown in Fig. 3, consists in superimposing a
Darcy’s flow within the unsaturated porous concrete
matrix and a Poiseuille’s flow within a perfectly plane
crack.
Under the assumption of small perturbations, all
quantities are expressed in relation to the system’s initial
configuration and the crack is seen as a 2D plan of area
Σ ce (in m2) associated to a uniform opening we (in m)
in the 3D porous matrix of volume Ω e (in m3). The air
flow is considered stationary and compressible, air be-
ing a perfect gas. Air flux field q D within the matrix (in
kg·s-1·m-2) is following Darcy’s law,
Mg K intg krg ( Sl )
qD = − ∇Pg2 ( x, t )
2 RT ( x, t ) μg
where K intg is the air intrinsic permeability of concrete
θ
Pint we
Fig. 3 Principle of the cracked FE for leakage computa-
tion.
411
(17) (in m2), krg its air relative permeability field (values
between 0 and 1), μ g is the air viscosity (in Pa·s), M g
is the air molar mass (in kg·mol-1) and Pg is the air
(18) pressure field (in Pa). The evolution of the air relative
permeability field follows Mualem’s law as modified by
(19) Verdier (2001),
krg ( Sl ) = 1 − Sl (1 − Slm )m / 2 (21)
Air flux field q P within the crack (in kg·s-1·m-2) is
following Poiseuille’s law with no friction or tortuosity
coefficient to reduce the flow, which is a conservative
assumption,
M g we
2
qP = − (I 3 − n ec ⊗ n ce )∇Pg2 ( x, t ) (22)
RT μ g 24
where nec is a unit crack surface normal. Adding the two
contributions (Eqs. 20, 22) to mass conservation leads to
the strong formulation of the problem to solve,
∇ ⋅ qD + ∇ ⋅ qP = 0 (23)
Complete details concerning the final FE formulation
and implementation can be found in (Asali et al. 2014).
Crack openings and orientations in each FE, needed
to compute the total leakage rate, are obtained according
to (Jourdain 2014) following (Matallah et al. 2010). In
Fig. 4, a crack-effective strain tensor field ε cod is de-
fined from the elastic-damageable behavior of concrete
(Eq. 24). Its maximal eigenvalue in tension (if any) is
used to compute the crack opening (Eq. 25) and the as-
sociated eigenvector defines the crack surface normal.
1
ε ecod ( x, t ) = ε el − [(1 + v)σ − vtr (σ )I 3 ] (24)
E
we = 3 Ω e max ⎡⎣ max ( Sp(εecod ) ) , 0 ⎤⎦ (25)
where Sp(ε ecod ) is the spectrum of ε ecod , i.e. the set of its
eigenvalues.
Total air leakage is then defined as the quantity of air
leaving the internal volume of the mock-up. Thus, nodal
air flux vectors are projected and integrated over the
(20)
whole internal surface, giving a mass leak rate value
Stress
Q
Pext
Strain
Fig. 4 1D principle to define the crack-effective strain.
M. Asali, B. Capra, J. Mazars and J-B. Colliat / Journal of Advanced Concrete Technology Vol. 14, 408-420, 2016
QM expressed in kg·s-1. In order to compare leakage
rates of all NRB independently of each ILR test tem-
perature and pressure conditions, EDF converts mass
flow rates into volumetric flow rates expressed at nor-
mal temperature and pressure conditions,
R T norm
QVnorm = 3600 QM
M g P norm
where QVnorm is the volumetric total leakage rate (in
Nm3·h-1) at normal temperature T norm (0 °C) and normal
pressure P norm (101315 Pa).
3. Application to the VeRCoRs mock-up
3.1 Used tools
The numerical strategy presented in section 2 is imple-
mented within the general framework of Code_Aster
(http://code-aster.org/). Code_Aster v12.3 is used as the
FE solver for all THM and air leak computations. The
cracked FE of section 2.4 is developed as a new non-
linear thermal model within Code_Aster. The concrete
delayed strain model of section 2.3 is implemented as an
external user behavior law through MFront v2.0.1
(http://tfel.sourceforge.net/).
3.2 Mesh
The coarser mesh proposed for the benchmark is lin-
earized before being used for all computations. It in-
cludes 34694 nodes in 3D concrete elements (penta- and
hexahedra) and 16066 nodes in 1D cable elements
(bars). 295 cables (horizontal, vertical, gamma and
dome) and their deviations are taken into account (Fig.
5(a)). Despite vertical and circumferential coarse mesh-
ing of concrete (Fig 5(b)), the 10-element progressive
fineness within the 40 cm thickness is suitable to repre-
sent phenomena linked to drying (Fig. 5(c)).
(a) Prestressing cables
Fig. 5 Overview of the coarse mesh of the mock-up.
412
3.3 Boundary conditions and loadings
As the overall structure is subject to a changing envi-
ronment as well as to a complex construction phasing
process, the on-site evolutions of loadings and boundary
conditions (BC) are simplified before being applied
over the whole considered time frame. The main con-
struction planning of the mock-up is considered as such:
(26)
- Erection is done at once, the starting date of all com-
putations is on 30 November 2014;
- Due to water aspersion on the mock-up’s walls, the
structure begins to dry only from 1 April 2015;
- Prestressing occurs between 6 May and 12 August
2015 with a reduced number of tensioning sequences;
- Two main life phases of the structure are occurring:
pre-operational until 28 January 2016 during which
the mock-up is subject to its external environment,
then operational starting on 1 April 2016. The transi-
tion between both phases is linear.
- The first three ILR tests of the mock-up are consid-
ered in the present study: two pre-operational tests
(PT0 and PT1) occurring between 10-14 November
2015 and 24-28 January 2016 respectively, and the
first decennial test (DT1) between 20-24 April 2017.
(a) Thermal BC and loadings
Soil temperature applied at the bottom of the raft is kept
constant during the whole computation at 10.5 °C (ini-
tial value). During pre-operation, measurements within
concrete are available to calibrate imposed temperatures.
No daily temperature variations are considered, only
linear evolutions between extremal values.
Operating temperature conditions are 35 °C at the in-
ner side and 15 °C at the outer side of the wall. Tem-
perature is kept constant during each decennial test at
10 °C (Fig. 6).
(b) Hydric BC and loadings
Hydric conditions are naturally expressed in terms of
40 cm
(b) Concrete (c) Zoom on thickness
M. Asali, B. Capra, J. Mazars and J-B. Colliat / Journal of Advanced Concrete Technology Vol. 14, 408-420, 2016 413

relative humidity (RH, value between 0 and 1). In order
to convert RH to saturation, Van Genuchten’s model (Eq.
3) is completed with Kelvin’s law,
ρl RT
Pc ( RH , T ) = −
Ml
where ρl is the water density (in kg·m-3) and M l is the
water molar mass (in kg·mol-1).
As for temperature, soil RH is kept constant at 0.98
(initial value). During pre-operation weather station
measurements are imposed at the inner and outer sides
with linearization between extremal values. Operating
HR conditions are 0.45 at the inner side and 0.6 at the
outer side of the containment, with no interruption dur-
Pre‐op
40
35
30
Temperature (°C)
ing decennial tests (Fig. 7).
(c) Mechanical BC and loadings
As no early age behavior is considered, initial stress and
strain states are null. The bottom of the raft is embedded
ln( RH ) (27)
during the whole computation. Reinforcement bars are
not explicitly meshed, but participate in the dead weight
of the structure (extra 100 kg·m-3 for concrete density).
The raft and anchoring ribs remain visco-elastic without
damage. Prestressing is simplified compared to the real
construction planning to only 14 grouped steps (instead
of 41 sequences in 16 phases), with an initial tension of
848 kN for each cable. All ILR tests are identical: they
last four days, with a 24-hour ramp from standard atmos-
phere to maximal pressure (5.2 bar abs.), a 24-hour pla-
teau at maximal pressure and 48 h of deflating (Fig. 8).
Operational phase
(d) Air leakage BC and loadings
During ILR tests, inner and outer air pressure follows
the evolution of Fig. 8. The raft being submerged during
ILR tests, it is not considered as a possible leakage

25
20 pathway (null flux).
15
10 3.4 Modeling and material parameters
5 Inner side Concerning concrete, the mean value of parameters
0 Outer side measured on-site for each lift is used when available
‐5 (density, porosity, Young’s modulus, tensile strength,
thermal expansion coefficient, air apparent permeabil-
ity). Other concrete parameters are either provided
Fig. 6 Evolution of imposed temperature. (thermal capacity, linear thermal conductivity with tem-
perature, activation energy, fracture energy) or identified
1
on specific lab tests (water intrinsic permeability, Van
Inner side
Genuchten’s model parameters, drying shrinkage coeffi-
0,9
Outer side cient, basic and drying creep parameters) (Fig. 9).
Relative humidity

0,8
Due to lack of some characterization tests, biaxial
creep effects, distinction between creep in tension and in
0,7 compression and coupling of creep with damage have
0,6 not been taken into account.
Cable parameters come from design and manufac-
0,5
turer values. The complete set of parameters used for
0,4 the present study is available in Appendix A.

4. Results and discussions

Fig. 7 Evolution of imposed relative humidity.
In this section, obtained numerical results are presented
and analyzed. They are compared to available experi-
6 0,5 mental data at four observation points, considered as
5
Inner side
0,4
representative of the behavior of the structure in stan-
Absolute Pressure (bar)

Outer side dard areas (far from any change in concrete geometry or
4 0,3 tendon deviation). Points H1, H2, H5 and H6 are moni-
Stress (MPa)

tored with vibrating wire and temperature sensors to

3 0,2
measure and correct vertical and tangential strains. They
2 0,1 all lie on the same horizontal plane at mid-height of the
cylindrical part. Groups H1-H2 and H5-H6 are diamet-
1 0
rically opposed, H1 and H5 being located at the outer
0 ‐0,1 side, H2 and H6 at the inner side of the wall (see Fig.
0 1 2 3 4 10).
Days

Fig. 8 Evolution of imposed air pressure during ILR tests.

M. Asali, B. Capra, J. Mazars and J-B. Colliat / Journal of Advanced Concrete Technology Vol. 14, 408-420, 2016 414

47,4 0

47,2 VeRCoRs data VeRCoRs data

‐100
47,0 Model fit Model fit

Strain (µm/m)
Weight (kg)

46,8 ‐200

46,6 ‐300
46,4
‐400
46,2

46,0 ‐500
0 30 60 90 120 150 180 210 240 90 120 150 180 210 240 270 300 330 360 390
Time (days) Time (days)
(a) Specimen weight loss (b) 1D drying shrinkage
0 0

‐100 VeRCoRs data ‐200 VeRCoRs data

‐200 Model fit ‐400 Model fit

Strain (µm/m)

Strain (µm/m)
‐300 ‐600

‐400 ‐800

‐500 ‐1 000

‐600 ‐1 200

‐700 ‐1 400
90 120 150 180 210 240 270 300 330 360 390 90 120 150 180 210 240 270 300 330 360 390
Time (days) Time (days)
(c) 1D basic creep (d) 1D total creep
Fig. 9 Identification procedure for some concrete parameters.
H6
tests DT2 to DT6 occur in May 2018, June 2019, July
H5
2020, August 2021 and September 2022 respectively.
As for temperature, saturation at the inner and outer
sides of the mock-up is following seasonal variations
during the first year. With the identified Van Genuch-
ten’s parameters, imposed RH values of 0.6 and 0.45
correspond to imposed saturation values of 0.64 and
0.57 respectively.
In the dome and cylindrical part, the saturation level
at the core of concrete thickness is decreasing with little
impact of imposed pre-operational RH cycles, meaning
global drying of the structure.
H2
Lacking concrete sorption isotherms data to identify
H1 Van Genuchten’s parameters independently of the con-
Fig. 10 Localization of observation points H1, H2, H5 crete water intrinsic permeability on Fig. 9(a) and some
and H6. RH measurements within the mock-up, drying kinetics
of the model is not yet experimentally confirmed.
4.1 Temperature in concrete
Temperature of concrete follows external seasonal 4.3 Evolution of concrete total strains
variations during the first year of the structure. Slowly Figure 12 shows the evolution of total strains within the
changing BC and loadings associated to the used ther- mock-up at the four observation points. All measured
mal parameters lead to an almost homogenous tempera- strains of Fig. 12 are corrected from temperature effects
ture of concrete within the cylindrical part and within according to the following formula (with null initial
the dome (less than 1 °C-difference between mid- strains),
thickness and outer sides).
ε cor (t ) = ε raw (t ) + α s (T (t ) − T (0)) (28)
4.2 Drying of concrete where ε cor is the corrected measured strain, ε raw is the
Figure 11 shows the evolution of the saturation degree raw measured strain and α s is the wire thermal expan-
along line H1-H2 in the thickness of the cylinder sion coefficient (in K-1, steel material). Until the first
(40 cm) before each ILR test. During operation, leakage
M. Asali, B. Capra, J. Mazars and J-B. Colliat / Journal of Advanced Concrete Technology Vol. 14, 408-420, 2016 415

1 to experimental data. The gap between H2 and H5 verti-

0,9
cal strains is around 25 % for the computation and 45 %
for measurements. For tangential strains, the gap be-
Saturation degree

0,8 tween H1 and H6 is around 10 % for the computation

0,7
0,6
Initial state
0,5
DT1
DT4
0,4
0 0,05
Fig. 11 Evolution of saturation in concrete thickness.
ILR test, computed total strains (dashed lines) are com-
pared with available corrected strain measurements
(plain lines). The mechanical computation is then ex-
tended to the first decennial test DT1.
According to Fig. 12(a), kinetics of delayed effects as
well as strain levels are well reproduced in the vertical
direction. According to Fig. 12(b), kinetics of delayed
effects is well reproduced in the tangential direction, but
compressive strain levels are globally underestimated by
the model, especially at points H5 and H6. This gap
may be explained by the fact that only uniaxial creep
tests are available to calibrate a biaxial creep model
leading to a homogeneous behavior of concrete in the
vertical and tangential directions.
To be more precise in the level of short-term strains,
the drying shrinkage model may have to be modified.
As shown in Fig. 9(b), the chosen model is preferen-
tially able to fit the long-term constant level of strains
and is less accurate for the first months. Its impact on
the total strain in Fig. 9(d) remains noticeable for the
mock-up during the first months after erection. However,
keeping in mind that the strategy aims at representing
the long-term behavior of full-size NRB, this effect
could be accepted in a first approach.
According to Fig. 12, numerical results between the
four considered points show less discrepancy compared
and 25 % for measurements. This observation may be
due to the homogeneous mechanical properties taken
into account in the computation, which are the mean
PT0 PT1 values of the real scattering of on-site concrete proper-
DT2 DT3 ties between the different lifts. Lower scattering of nu-
DT5 DT6
merical and experimental tangential strains compared to
0,1 0,15 0,2 0,25 0,3 0,35 0,4 the vertical ones is due to the fact that all four points
Thickness (m)
were taut together in the horizontal direction (they are at
the same height), which was not the case in the vertical
direction (they are diametrically opposed).
4.4 Damage of concrete
Figure 13 shows the computed damage field in the
mock-up at the end of each maximal pressure plateau.
For the sake of clarity, only maximal values (between
0.7 and 1) are kept. Areas where cracks are connected
and going through the concrete thickness are also high-
lighted.
During PT0, a small number of FE is damaged. The
gusset and its junctions to anchorage ribs are lightly
damaged at the outer side of the containment (value
lower than 0.2, in the first layer of FE only). According
to Fig. 13(a), areas at the top and bottom of the material
hatch are more highly damaged, from the inner side to
the middle of the thickness. Associated crack openings
range from 10 to 51 µm. Compared to visual inspection
in Fig. 14 (not able to detect crack openings lower than
100 µm), the computed results are in good agreement in
the cylinder, dome and hatch areas where no crack is
identified but only defects like porosity lines, corrosion
or honeycombing.
Nevertheless, vertical cracks are observed at the outer
side of the gusset during the inflation test (in thick green
lines in Fig. 14). Over the 30 identified cracks, only 6
have an opening exactly equal to 100 µm, the other ones
having no measured value. This cracking originates

0 0
End of tensioning
DT1

DT1
PT0

PT1

PT0

PT1

Measure Computation Measure Computation

H1 ‐100 H1
‐100
H2 H2
H5 ‐200 H5
‐200 H6 H6
Strain (µm/m)

Strain (µm/m)

‐300

‐300 ‐400

‐500
‐400
End of tensioning

‐600
‐500
‐700
Operation Operation
‐600 ‐800

(a) Vertical strains (b) Tangential strains

Fig. 12 Evolution of measured and computed total strains at points H1, H2, H5 and H6.
M. Asali, B. Capra, J. Mazars and J-B. Colliat / Journal of Advanced Concrete Technology Vol. 14, 408-420, 2016 416

Damage scale Going-through cracks

0 – 200 gr 200 – 400 gr

(a) PT0

0 – 200 gr 200 – 400 gr

(b) PT1

0 – 200 gr 200 – 400 gr

(c) DT1

Fig. 13 Evolution of damage in the mock-up during ILR tests.

M. Asali, B. Capra, J. Mazars and J-B. Colliat / Journal of Advanced Concrete Technology Vol. 14, 408-420, 2016 417

3 -1
Table 1 Evolution of total air leakage rate and distribution (in Nm ·h ) at the inner side (starred values are measured at
the outer side of the wall).
VeRCoRs measurements Computation results
ILR test Total no crack
Total Hatch Gusset Dome Cylinder Total Hatch Gusset Dome Cylinder
PT0 7.7 0.67* 2.44* 0.17* 1.08* 6.69 0.09 0.03 1.45 5.12 6.58
PT1 9.5 - - - - 9.25 0.09 0.02 3.30 5,84 7.79
DT1 - - - - - 55.6 0.09 7.14 2.74 45.6 7.71

from the early-age behavior of the structure and, despite
being closed by prestressing, reopens during ILR tests.
As the present computation starts with a null initial state
and a structure built in only one step, this phenomenon
could not be represented in associated results. But the
capacity of the model to close and reopen cracks en-
ables to take into account another initial stress-strain-
damage state, coming from previous early-age computa-
tion for instance.
During PT1, the first thin layer of FE at both sides of
the containment is damaged at a low average level
(value around 0.4). This generalized surface behavior,
not encountered in other computed ILR tests, may be
due to temperature and humidity gradients that are still
evolving during the pre-operational phase. The maximal
crack opening at mid-thickness within the material
hatch increases to 90 µm. According to Fig. 13(b), more
highly damaged areas on the inner and outer sides (val-
ues greater than 0.7) are linked to geometrical accidents
in concrete and to tendons deviations: anchorage ribs,
material and personal hatches. As it is not connected
through the concrete thickness, this diffuse surface
cracking does not impact significantly the total leakage
rate of the mock-up.
During DT1, the first layer of FE at the inner side of
the containment is damaged at a medium average level
(value around 0.5), which may be due to the tempera-
ture gradient imposed between the internal and external
sides during operation. Cracking progresses in the mate-
rial hatch: 70 % of its thickness is continuously cracked
from the inner side, with a maximal opening of 160 µm.
According to Fig. 13(c) and contrary to previous pre-
operational tests, two cracks connected through the
thickness are computed at the two junctions between the
gusset and anchorage ribs. The crack opening within the
gusset and the first lift above it between 200 and
400 grades evolves from 150 µm at the inner side to
40 µm at the outer side.
4.5 Air leakage of the mock-up
Table 1 shows the evolution of air leakage rate com-
puted at the end of maximal pressure plateau for the first
three ILR tests (PT0, PT1 and DT1). The total leakage
is compared to in-situ measurements for the two pre-
operational tests. The distribution of air leakage by area
is also shown. However this distribution is provided at
the inner side for the computation while experimental
results are only collected at the outer side of the inner
containment. It can be noticed that only 57 % of the
total internal leak is measured at the outer side: it is in-
deed difficult to inspect 1100 m2 of concrete during the
24 h of maximal pressure and the focus is put on visu-
ally leaking areas (soap aspersion). The influence of
modeled cracking on the total leakage rate is quantified
by comparing computed results to an extra computation
taking into account the unsaturated-concrete darcean
flow only.
A first computation has been launched using the mean
value of air apparent permeabilities (3.13·10-16 m2) as
the intrinsic permeability of concrete. In that case, no

400 gr 300 gr 200 gr 100 gr 0 gr

400 gr 300 gr 200 gr 100 gr 0 gr

Fig. 14 Visual inspection of the outer side after PT0 (cracks in green, Corbin and Garcia 2016).
M. Asali, B. Capra, J. Mazars and J-B. Colliat / Journal of Advanced Concrete Technology Vol. 14, 408-420, 2016
Klinkenberg’s effect (Klinkenberg 1941) is taken into
account leading to overestimating the total leakage rate
during PT0 (42.9 Nm3·h-1, +557 % compared to the ex-
perimental measure). Thus the intrinsic permeability
value for the proposed computation is calibrated on this
first ILR test (to 5·10-17 m2). For PT0 and PT1, this
value enables to compute total air leak rates that are
very close to experimental measurements (respectively -
13 % and -3 %).
For the proposed computation, air leakage is mainly
due to concrete porosity during PT0 and PT1 (98 % and
84 % of the total leak respectively). Even if cracking
occurs at that stage, it is not sufficiently going through
concrete thickness to participate significantly to the total
leakage, as stated in the previous section. This is cor-
roborated by analyzing the leakage rate distribution by
areas, which is globally proportional to the surface of
each zone. The influence of the hatch and gusset does
not evolve between the first two tests and all the leakage
increase is due to the cylinder and to the dome, which
are thinner and on which drying has more effect during
those three months.
During DT1, cracking of concrete is contributing sig-
nificantly to the total leakage rate. Indeed, leakage
through porosity does not evolve between PT1 and DT1
(and only represents 14 % of the total leak), contrary to
leakage through the gusset and the cylinder which in-
creases strongly (respectively 357 and 7.8 times). Asso-
ciated to the damage (Fig. 13(c)) and air flux mapping,
the leaking increase is mainly caused by the continuous
cracks at the junctions between the gusset and the two
anchorage ribs.
The computed total leakage value of 55.6 Nm3·h-1
may be high due to the conservative assumption of not
considering any friction coefficient for the air flow
within cracks. For instance, using a classical reduction
factor of 0.1 (Picandet 2001) on the Poiseuille’s part of
the flow during DT1 would lead to a total leakage rate
of 12.5 Nm3·h-1. In that case, the leak through concrete
porosity would still be prevailing (62 %). This coeffi-
cient could be identified on experimental measures
when available.
Considering that air leak distribution measured during
PT0 at the outer side could be extrapolated at the inner
side of the containment, experimental data show a local
behavior different from the one predicted by the compu-
tation. The gusset area participates significantly to the
total leak. This observation confirms that cracks identi-
fied at the outer side of the gusset are most certainly
radial and going through the whole concrete thickness.
5. Conclusions and perspectives
This paper presents a numerical strategy that aims at
computing and forecasting the leakage rate of double-
wall NRB (without steel liner) under operating condi-
tions. Based on (1) a coarse but adapted mesh, (2) a
chained weakly-coupled THM modeling and (3) a spe-
418
cially-designed 3D FE to compute air leakage through
unsaturated porous and cracked concrete, the proposed
methodology is applied and benchmarked to the scale
1:3 NRB mock-up VeRCoRs built by EDF. After show-
ing good overall agreement with experimental observa-
tions until the first pre-operational ILR test, the compu-
tation is extended to the first decennial test of the mock-
up. Under the considered assumptions of operating con-
ditions with no early-age behavior effects, the proposed
methodology predicts a proper physical evolution of the
total leakage rate coupled to delayed strains, drying and
cracking of concrete.
Compared to available experimental results on the
VeRCoRs mock-up, some improvements of the pro-
posed strategy are already identified: (1) A drying
shrinkage model able to represent short-term kinetics as
well as long-term delayed drying strains levels could be
introduced (based on a Burger rheological chain for
instance). The same applies to the drying creep model if
the drying shrinkage improvement is not sufficient. (2)
In all cases, the drying kinetics of core concrete needs to
be validated with future experimental data. (3) The
mesh refinement in the thickness may need to be further
customized if temperature and humidity effects are in-
deed the cause of surface cracking during ILR tests. (4)
The identification procedure, based on lab results for the
mock-up, could be adapted to in-situ total strains meas-
urements, for instance to define a representative creep
Poisson’s coefficient. As only few characterization tests
are available on full-size NRB compared to VeRCoRs,
this feature seems necessary. (5) The 3D FE designed
for air leak computations is currently modeling one
crack related to the mechanical state of concrete. But its
formulation enables patching as many independent
cracks as desired in the same FE. This feature could also
be used to consider initial defects of the structure such
as porosity lines or leaking lift joints. (6) Finally, the no-
early-age effect assumption of the strategy may need to
be reconsidered in light of experimental observations.
Lack of initial early-age cracks leads to underestimating
the gusset leaking during PT0, which would be prob-
lematic only if crack openings in this area evolve with
time. If not, the simplest way to take into account the
initial damage of the gusset could be to locally increase
the intrinsic permeability to recover experimental results.
If crack openings increase in the gusset, an initial early-
age computation would be necessary to define a non-
null strain-stress-damage state for the proposed strategy
which is able to close them during prestressing and re-
open them during ILR tests.
Considering conservative assumptions, using simple
modeling techniques (non-linear thermal analogy for
drying and for air-leak computations) and a delayed
strain model representative of important phenomena
with few parameters to identify (compared to other
models available in the literature and FE codes), the
proposed methodology is numerically cost-effective and
thus enables taking into account variabilities and uncer-
M. Asali, B. Capra, J. Mazars and J-B. Colliat / Journal of Advanced Concrete Technology Vol. 14, 408-420, 2016
tainties of main parameters. It could be used as a quali-
tative and quantitative decision-support tool to help op-
erators: (1) Pre-empting and optimizing leak mitigation
actions before each decennial ILR test and consequently
avoiding outage extensions and associated losses of
income; (2) Forecast the behavior of their assets in the
context of long-term operation and life extension be-
yond design in order to demonstrate their compliance
with safety and regulatory requirements.
Acknowledgements
Carried out within the RSNR MACENA project, this
study received government assistance handled by the
French National Research Agency (ANR) under the
“investments for the future” program, part number
ANR-11-RSNR-0012-01.
The authors gratefully acknowledge the VeRCoRs
project team for analyzing and sharing large amounts of
data as well as the AMA team of EDF Lab Paris-Saclay
for numerous discussions and for helping with
Code_Aster and MFront modeling.
References
AFCEN, (2010). “EPR technical code for civil works.”
Lyon: Association Française pour les règles de
Conception, de construction et de surveillance en
exploitation des matériels des Chaudières Electro
Nucléaires.
Asali, M., Capra, B., Colliat, J.-B. and Mazars, J.,
(2014). “Elements for computing and forecasting the
leakage rate of the inner containment of reactor
buildings.” In: Fontevraud 8 – Contribution of
Materials Investigations and Operating Experience to
LWR’s Safety, Performance and Reliability, Avignon
14-18 September 2014, France.
CEA, “Mfront, a code generation tool dedicated to
material knowledge.” released under GNU GPL and
CECILL-A license. Website:
http://tfel.sourceforge.net/ [accessed on 15 March
2016].
Corbin, M. and Garcia, M., (2016). “International
benchmark VeRCoRs 2015 – Overview, synthesis and
lessons learnt.” Villeurbanne: EDF SEPTEN.
EDF Lab, “Code_Aster, Strucutres and
Thermomechanics Analysis for Studies and
Research.” released under GNU GPL. Website:
http://code-aster.org/ [accessed on 15 March 2016].
Granger, L., (1995). “Comportement différé du béton
dans les enceintes de centrales nucléaires : analyse et
modélisation.” Thesis (PhD). Ecole Nationale des
Ponts et Chaussées.
Hilaire, A., (2014). “Etude des déformations différées
des bétons en compression et en traction, du jeune
âge au long terme.”Thesis (PhD). Ecole Normale
Supérieure de Cachan.
Hillerborg, A., Modeer, M. and Petersonn, P. E., (1976).
“Analysis of crack formation and growth in concrete
by mean of fracture mechanics and finite elements.”
419
Cement and Concrete Research, 6, 773-782.
Jason, L. and Masson, B., (2014). “Comparison between
continuous and localized methods to evaluate the
flow rate through containment concrete structures.”
Nuclear Engineering and Design, 277, 146-153.
Jourdain, X., (2014). “Etude numérique méso-macro des
propriétés de transfert des bétons fissurés.” Thesis
(PhD). Ecole Normale Supérieure de Cachan.
Klinkenberg, L. J., (1941). “The permeability of porous
media to liquids and gases.” API Drilling and
Production Practice, 200-213.
Matallah, M., La Borderie, C. and Maurel, O., (2010).
“A practical method to estimate crack openings in
concrete structures.” International Journal for
Numerical and Anaytical Methods in Geomechanics,
34, 1615-1633.
Mazars, J., Hamon, F. and Grange, S., (2015). “A new
3D damage model for concrete under monotonic,
cyclic and dynamic loadings.” Materials and
Structures, 48, 3779-3793.
Mozayan, M., Davenne, L. and Ghavamian, S., (2012).
“Modelling of leak tightness degradation of
reinforced containment vessels due to aging and
cracking mechanisms.” In: SSCS 2012 – Strategies
for Sustainable Concrete Structures, Aix-en-Provence
29 May – 1 June 2012, France.
Mualem, Y., (1976). “New model for predicting
hydraulic conductivity of unsaturated porous media.”
Water Resources Research, 12 (3), 513-522.
Nicklasch, C., Coudert, L., Heinfling, G., Hervouet, C.,
Masson, B., Herrmann, N. and Stempniewski, L.,
(2005). “Numerical investigation of the leakage
behavior of reinforced concrete walls.” In:
NURETH11 – International Topical Meeting on
Nuclear Reactor Thermal-Hydraulics, Avignon 2-6
October 2005, France.
Picandet, V., (2001). “Influence d’un endommagement
mécanique sur la perméabilité et sur la diffusivité
hydrique des bétons.” Thesis (PhD). Université de
Nantes.
Pickett, G., (1942). “The effect of change in moisture
content on the creep of concrete under a sustained
load.” ACI Journal Proceedings, 38, 333-355.
VeRCoRs project website. https://fr.amiando.com/EDF-
vercors-project.html [accessed on 15 March 2016].
Wang, T. and Hutchinson, T. C., (2005). “Gas leakage
through reinforced concrete shear walls: numerical
study.” Nuclear Engineering and Design, 235, 2246-
2260.
Van Genuchten, M. T., (1980). “A closed-form equation
for predicting the hydraulic conductivity of
unsaturated soils.” Soil Science Society of America
Journal, 44, 892-898.
Verdier, J., (2001). “Contribution à la caractérisation de
l’évolution du taux de fuite des enceintes de
confinement du parc nucléaire.” Thesis (PhD).
Université Paul Sabatier Toulouse.
Journal of Advanced Concrete Technology Vol. 14, 408-420 August 2016 / Copyright © 2016 Japan Concrete Institute 420

Appendix A: Modeling and material parameters

Parameter Symbol Value Unit Source

Concrete density ρb 2395 kg·m -3
Mean value of on-site lifts
Concrete thermal capacity Cp 880 J·kg ·K-1 -1
Lab test result
Concrete thermal conductivity λ (T ) 6.77T – 239 10 W·m ·K -3 -1 -1
Lab test result
Concrete water dynamic viscosity μl 1.002·10-3 Pa·s
Concrete porosity φ 0.146 - Mean value of on-site lifts
Concrete thermal activation energy Ea 28000 J·mol -1
Lab test result
-1 -1
Universal gas constant R 8.314 J·mol ·K
ref
Reference temperature T 20 (293.15) °C (K) Lab test result
Liquid water density ρl 998.3 kg·m-3
Water molar mass Ml 18·10-3 kg·mol-1
l
Concrete water intrinsic permeability kint 6.0737·10-20 m2
n 1.3014 - Identification on weight loss lab test curve
Van Genuchten’s parameters
Pr 17.607·10 6
Pa
Concrete Young’s modulus E 36.849 GPa Mean value of on-site lifts
Concrete Poisson’s ratio v 0.2 - Default value (no info)
Reinforced concrete density 2495 kg·m -3
ρb + 100
Concrete thermal expansion coeff. α th 1.22·10 -5
K -1
Mean value of on-site lifts
Basic creep Poisson’s ratio vbc 0.2 -
Default value (no info)
Basic creep tension/compression coeff. α bc 1 -
Water viscosity activation energy Eaw 18700 J·mol-1 (Hilaire 2014)
-1
Concrete fracture energy (tension) 100 N·m Lab test result
Concrete yield stress (tension) ft 3.89 MPa Mean value of on-site lifts
Damage parameter (tension) At 0.9 -
Ac 1.25 -
Damage parameters (compression) Default values (no info)
Bc 600 -
Coeff. Coupling creep and damage χ 0 -
Drying shrinkage coefficient κ ds 1.0265·10 -3
- Identification on 1D shrinkage lab test curve
kkv 168.22 GPa
Basic creep coefficients η kv 141.84·106 GPa·s Identification on 1D basic creep lab test curve
kam 81.63 GPa
Drying creep coefficient κ dc 6.8968·10 -11
Pa-1 Identification on 1D total creep lab test curve
Air molar mass Mg 29·10 -3
kg·mol -1

Air dynamic viscosity μg 1.8·10 -5

Pa·s
g -17
Concrete air intrinsic permeability k int 5·10 m2 Identification on first ILR test
Wire steel thermal expansion coef. αs 1.15·10 -5
K -1
(Corbin and Garcia 2016)
Journal of Advanced Concrete Technology Vol. 14, 634-642, October 2016 / Copyright © 2016 Japan Concrete Institute 634

Scientific paper

Residual Static Strength of Concrete Cylinder Specimen and Stud

Anchor Specimen after Cyclic Loadings
Osamu Kontani1*, Naoki Ishitobi2, Junya Kawada2, Noriaki Taogoshi3, Makoto Koge4 and Yoshito
Umeki5

Received 5 April 2016, accepted 14 October 2016 doi:10.3151/jact.14.634

Abstract
The residual strength tests after cyclic loadings on concrete cylinder specimens and stud anchor specimens were per-
formed to obtain data regarding effects of cyclic loadings on concrete and stud anchor specimens for proposing a new
method capable of evaluating the soundness of equipment support using the S-N curves. More specifically, cyclic load-
ings may lead to a reduction in the strength and stiffness of concrete and also lead to the pullout and a reduction of the
strength of stud anchor sections, which may reduce the rigidity of the equipment support, and this situation is not pref-
erable from the viewpoint of long-term soundness evaluation. Then, in order to investigate whether or not cyclic loadings
reduce the strength and the rigidity of the equipment support, cyclic compressive loadings with constant amplitude were
applied to cylinder specimens, then static compression tests were performed. Also, cyclic tensile loadings with constant
amplitude were applied to stud anchor specimens, then static pullout tests were performed. As a result, it was found that
the cyclic loading would not lead to the pullout of stud anchor and a reduction of the strength of stud anchor if the
maximum stress ratio is less than the long-term allowable stress.

1. Background of research evaluating soundness is investigated using S-N curves in

The degradation factors considered in the evaluation of
long-term soundness of structures in nuclear facilities are
heat (high temperature), irradiation, carbonation, chloride
penetration, alkali silica reaction, and machine vibration
(AIJ 2015). Since the degradation state due to machine
vibration cannot be identified until cracks are visually
identified as degradation phenomena, the soundness
evaluation of the equipment support against machine
vibration is seemed to be very rudimentary compared with
the evaluation against other factors.
On the other hand, although the number of load cycles
due to vibration equipment, such as turbines, is very high
in the normal operation of a nuclear power plant, since the
stress ratio acting as the cyclic load is very low, it is con-
sidered that the fatigue caused by machine vibration
would never damage concrete.
In the cooperative research performed by electric in-
dustries for the common purpose, a method capable of
1
General Manager, Nuclear Power Department, Kajima
Corporation, Tokyo, Japan.
*Corresponding author, E-mail: kontani@kajima.com
2
Deputy Manager, Nuclear Nuclear Power Department,
Kajima Corporation, Tokyo, Japan.
3
Assistant Manager, Power Generation Division, The
Chugoku Electric Power Co., Inc., Hiroshima, Japan.
4
Associate Manager, Power Generation Division, The
Chugoku Electric Power Co., Inc., Hiroshima, Japan.
5
Senior Manager, Civil and Architectural Engineering
Department, Nuclear Power Division, Chubu Electric
Power Co., Inc. Nagoya, Japan
the stage before cracks due to machine vibration can be
visually confirmed. This paper reports the results of re-
sidual static strength tests of concrete cylinder specimens
and headed stud anchor specimens conducted to obtain
basic data for the purpose of establishing an evaluation
method. In this paper, the equipment support means the
reinforced concrete section with steel anchorage embed-
ded for installing and holding equipment. A headed stud
anchor was selected as steel anchorage because a large
volume of its experimental data is available.
Plain concrete specimens undergo changes of material
properties with increasing number of load cycles. Plain
concrete specimens exhibit increases in residual com-
pressive strength and stiffness after cyclic loading. As the
number of load cycles is getting closed to fatigue life, the
residual strength and stiffness are decreased and become
less than those without cyclic loadings.
Bennett and Muir (1967) measured the residual
strength after cyclic loadings of the plain concrete using
8×3×3 inch prism specimens with compressive strength of
40MPa to 64MPa. The tests were performed only on the
11 specimens which survived one million load cycles with
speed of 4Hz to 8Hz at the maximum stress ratios of 0.67
and the minimum ratios varying 0.15 and 0.2 of the of
static strength of prism specimens without cyclic loadings.
The increase in the residual strength varied between 1%
and 23% with average of 11%.
Awad and Hilsdorf (1974) performed fatigue tests and
creep tests using the 12×4×4 inch prism specimens with
compressive strength of 28MPa. The maximum stress
ratios were quite high at 0.90 and 0.95 of the static
strength without cyclic loadings. There was an increase of
5% in the residual strength after cyclic loading or sus-
O. Kontani, N. Ishitobi, J. Kawada, N. Taogoshi, M. Koge and Y.o Umeki / Journal of Advanced Concrete Technology Vol. 14, 634-642, 2016 635

Table 1 Mix proportion for cylinder specimens.

Unit Mass（kg/m3）
ID W/C(%) s/a(%)
water cement sand gravel
WC59 59 42 158 268 781 1095
Additives : air entraining water reducing agent (1% of cement weight)
air entraining agent (0.003% of cement weight)
Slump :15±2.5cm, Air Contents：4.0±1.0％

tained loading with the maximum stress ratio of 0.9 at life
ratio of 0.3 which means at 30% cycles of fatigue life or
30% of duration to cause the creep failure.
Ballatore and Bocca (1997) measured the residual
strength after cyclic loadings using the plain concrete core
specimens of 45.5 mm diameter and 140 mm long with
compressive strength of 19MPa and 59MPa. The maxi-
mum stress ratios were at 0.1 and 0.2 and the minimum
ratio was 0.02 of the static strength without cyclic load-
ings. The cyclic loadings of at most 7200 cycles were
applied with loading speed of 1Hz. The increases in the
residual strength and stiffness were 0.11 and 0.87 respec-
tively for concrete with 19MPa and were 0.14 and 0.22
respectively for concrete with 59MPa.
Bocca and Crotti (2003) measured the residual strength
after cyclic loadings using the plain concrete core speci-
mens of 60 mm diameter and 120 mm long with com-
pressive strength of 24MPa, which were similar tests
performed by Ballatore and Bocca (1997). The maximum
stress ratios were at 0.2, 0.6 and 0.8 of the static strength
without cyclic loadings. The cyclic loadings of 3600 cy-
cles were applied with loading speed of 1Hz. The in-
creases in the residual strength and stiffness were 5% and
6% respectively for the maximum stress ratio of 0.2, were
7.6% and 36% respectively for the maximum stress ratio
of 0.6, and were 11.4% and 36.1% respectively for the
maximum stress ratio of 0.8.
Ballatore and Bocca (1997) stated that dynamic loading
reduces pore space to improve material compaction and
strain hardening takes place all the same time as is re-
flected in an increase in strength and stiffness after cyclic
loadings. Bocca and Crotti (2003) stated that the test
pieces subjected to preloading yielded more homogenous
results, with a reduction in standard deviation, compared
to non-preloaded test pieces probably because of a de-
crease in the stress peaks occurring in the material and a
redistribution of stresses due to cyclic load induced creep.
For the headed stud anchor, only one datum of the re-
sidual tensile strength after cyclic loadings was found in
Matsuzaki, et al. (1981). With the stud diameter of 19 mm,
load cycles was increased by 5% compared with the static
tensile strength without cyclic loadings. The maximum
and minimum stress ratio was 0.5 and 0.1 respectively
with loading speed of around 1Hz.
2. Concrete cylinder specimen
This section reports the residual strength test results of the
concrete cylinder specimens in order to obtain the very
basic information of concrete under cyclic compressive
stresses.
2.1 Outline of tests
Table 1 shows the mix proportion of concrete. The con-
crete was manufactured using gravel and sand produced
in Shizuoka Prefecture and ordinary Portland cement with
water to cement ratio of 0.59. Compressive strength of
35MPa was the target.
The cylinder specimens were 10 cm in diameter and 20
cm long and were cast in steel molds. The specimens were
demolded the next day, and were cured and sealed in
aluminum bags at 30°C for 4 months to prevent further
strength development during fatigue tests. After 4 months
of curing, cylinder specimens were still kept in aluminum
bags until those were submitted to tests. It took 3 months
to finish all the tests. During cyclic loadings and residual
strength tests, the specimens were exposed to an atmos-
phere in the testing laboratory.
Table 2 shows the maximum stress ratios and number
of load cycles before residual strength tests for each
specimen. Three levels of the maximum stress ratios (0.75,
0.65 and 0.33) were considered in the fatigue tests. The
minimum stress ratio was 0.05 and the loading speed was
5 Hz in all cases. Figure 1 shows test cases in terms of the
number of load cycles and the maximum stress ratios in
the S-N curves. During tests, the compressive strains were
Cyclic Loading+Static Test
1 (10 specimens)
Fatigue Test Max. Stress Ratio
0.8
(5 specimens) 0.75
Maximum Stress Ratio

0.65
0.75
embedded depth of 80 mm and concrete strength of Fatigue Test
(5 specimens) 0.33
0.65
23.5MPa, the residual tensile strength after two million of 0.6 Cylinder Specimen
load cycles was increased by 27% compared with the Cyclic Loading+Static Test (Uni-Axial Compression)
(10 specimens)
static tensile strength without cyclic loadings. The 0.4 Stud Anchor Specimen
0.33 (Corn Failure)
maximum and minimum stress ratio was 0.75 and 0.11
Cyclic Loading+Static Test
respectively with loading speed of 10Hz. 0.2 (15 specimens)
The residual bond strength between deformed bar and
0
concrete after cyclic loadings was found in Rehm and 10 0 10 4 10 8 10 12 10 16 1020
Eligehausen (1979). With the rebar diameter of 14 mm, Nunber of Load Cycles
embedded length of 42 mm and concrete strength of Fig. 1 Test Cases in terms of number of load cycles and
23.5MPa, the residual tensile strength after two million of the maximum stress ratio.
O. Kontani, N. Ishitobi, J. Kawada, N. Taogoshi, M. Koge and Y.o Umeki / Journal of Advanced Concrete Technology Vol. 14, 634-642, 2016 636

Table 2 Maximum stress ratios and number of load cycles.

Specimen ID+No. Max. Stress Ratio Fatigue Life Fatigue Life Ratio Type of Test
CL-075-A-1～5 316 1/100 STCL
CL-075-B-1～5 0.75 3,160 1/10 STCL
CL-075-C-1～5 31,600 1/1 FT
CL-065-A-1～5 20,000 1/100 STCL
CL-065-B-1～5 0.65 200,000 1/10 STCL
CL-065-C-1～5 2,000,000 1/1 FT
CL-033-A-1～5 20,000 1/100 STCL
CL-033-B-1～5 0.33* 200,000 1/10 STCL
CL-033-C-1～5 2,000,000 1/1 STCL
Fatigue Life Ratio shows ratio of number of load cycles to fatigue life.
* : No fatigue failure is expected. The max. load cycles are set at two million.
FT：Fatigue Test (Fatigue life was obtained from Eq. (1).)
STCL：Static Test after Cyclic Loadings

Table 3 Residual compressive test results after cyclic loadings.

Maximum Residual Elastic Strain at Residual Elastic Strain at Residual Elastic Strain at
Stress No. Strength Modulus Failure No. Strength Modulus Failure No. Strength Modulus Failure
Ratio (ID) (MPa) (kN/mm2) (μ) (MPa) (kN/mm2) (μ) (MPa) (kN/mm2) (μ)
(1/100 of loading cycles=20,000) (1/10 of loading cycles=200,000) (1/1 of loading cycles=2 millions)
A-1 35.5 29.0 1805 B-1 38.7 30.8 2051 C-1 37.8 30.7 2174
A-2 38.1 30.1 2213 B-2 40.1 32.3 2059 C-2 40.0 32.0 2190
0.33 A-3 34.5 30.5 1958 B-3 38.9 30.8 2202 C-3 40.2 30.3 2340
(CL-033) A-4 39.2 33.5 1946 B-4 37.6 32.1 1843 C-4 39.1 31.9 2115
A-5 41.5 31.0 2013 B-5 39.2 34.5 1866 C-5 41.8 30.9 2149
Ave. 37.8 30.8 1987 Ave. 38.9 32.1 2004 Ave. 39.8 31.2 2194
SR 1.01 SR 1.04 SR 1.07
(1/100 of fatigue life=20,000) (1/10 of fatigue life=200,000) (1/1 of fatigue life=2 millions)
A-1 38.8 25.1 2074 B-1 36.2 21.9 2199 C-1 651437*1 ―
A-2 39.9 26.4 1970 B-2 40.3 25.1 1966 C-2 623876*1 ―
0.65 A-3 40.7 26.7 2164 B-3 37.7 23.6 2131 C-3 278233*1 ―
(CL-065) A-4 40.2 25.8 2156 B-4 40.0 23.6 2340 C-4 36.2 18.3 2855
A-5 39.5 28.8 1613 B-5 44.8 27.7 2300 C-5 40.2 17.6 3374
Ave. 39.8 26.6 1995 Ave. 39.8 24.4 2187 Ave. 38.2 18.0 3115
SR 1.07 SR 1.07 SR 1.02
(1/100 of fatigue life=316) (1/10 of fatigue life=3,160) (1/1 of fatigue life=31,600)
A-1 40.4 30.7 1759 B-1 40.4 25.5 2029 C-1 17061*1 ― ―
A-2 41.4 28.3 2102 B-2 39.9 23.3 2203 C-2 4187*1 ― ―
0.75 A-3 40.4 29.7 2019 B-3 40.2 26.3 1967 C-3 9602*1 ― ―
(CL-075) A-4 40.4 29.4 1951 B-4 40.2 24.8 1972 C-4 9687*1 ― ―
A-5 39.8 28.6 1887 B-5 40.5 25.5 2100 C-5 20368*1 ― ―
Ave. 40.5 29.3 1944 Ave. 40.2 25.1 2054 Ave. ― ― ―
SR 1.09 SR 1.08 SR ―
*1: Number of load cycles when fatigue failure occurred.
SR : Strength Ratio of residual static strength over static strength without cyclic loadings

measured by two strain gages with a gage length of 9cm der uniaxial compression as shown in Eq. (1) (Kishitani
attached on the opposite side of cylinder specimens. The and Nishizawa 1987). The number of load cycles was set
loads applied to specimens were also measured. at two million for the maximum stress ratio of 0.33 be-
The maximum stresses were determined based on the cause the stress level of 0.33 was too low to expect fatigue
compressive strength obtained from the static tests of the failure within feasible time frame.
same cylinder specimens. In the tests described in Table 2,
the number of load cycles was applied to the cylinder s = 1 − log N /18 (1)
specimens with constant amplitude and then residual
where s： maximum stress ratio (Strass ratio to fatigue
compression tests were performed in order to investigate
failure at number of load cycles of N) and N：fatigue life
whether or not cyclic loadings reduce strength and stiff-
(number of load cycles to fatigue failure at the maximum
ness of concrete. The number of load cycles before com-
stress ratio of s)
pression tests was set to be 1/10 and 1/100 of the fatigue
life estimated from S-N curve of cylinder specimens un-
O. Kontani, N. Ishitobi, J. Kawada, N. Taogoshi, M. Koge and Y.o Umeki / Journal of Advanced Concrete Technology Vol. 14, 634-642, 2016
2.2 Test results
Table 3 shows the list of results of residual compressive
tests after cyclic loadings. The compressive strength was
37.3MPa, and the elastic modulus was varying between
31.1 kN/mm2 and 34.4 kN/mm2 from the compression
tests without cyclic loadings. The fatigue life was evalu-
ated to be 31,600 and all specimens exhibited fatigue
failure for the maximum stress ratio of 0.75, and to be
2,000,000 and three specimens out of five exhibited fa-
tigue failure for the maximum stress ratio of 0.65. The
cylinder specimens that did not exhibit fatigue failure
during cyclic loadings showed a larger average residual
compressive strength than the compressive strength of
specimens without cyclic loadings. The increase ratios are
from 1% to 9%. These tendencies are similar to the pre-
vious research found in Ballatore and Bocca (1997) and
Bocca and Crotti (2003). The explanation of the phe-
nomena can be also found in Ballatore and Bocca (1997)
and Bocca and Crotti (2003). The cyclic loadings reduce
pore space to improve material compaction and strain
hardening takes place. The cyclic loadings also decrease
(CL-033-B-4)
(CL-065-B-4)
(CL-075-B4)
Fig. 2 Compressive load-displacement relationship.
637
the stress peaks occurring in the material and redistribute
stresses due to cyclic load induced creep. These might be
the reasons of increase in strength and stiffness after cy-
clic loadings.
The average elastic moduli were subjected to little ef-
fects of cyclic loadings at the maximum stress ratio of
0.33, and were decreased as the number of load cycles
increased at the maximum stress ratios of 0.65 and 0.75.
Figure 2 shows the stress-strain relationships of the
static test without cyclic loadings, the cyclic loadings and
residual tests after cyclic loadings. The only representa-
tive cycles are shown for the cyclic loadings in Fig. 2. The
figure shows the tendency that the residual strength after
cyclic loadings is higher than that in the static test without
cyclic loadings. Even at the stress level of 0.33, there is
inelastic compressive strain. This can be explained by
cyclic load induced creep and material compaction.
Figure 3 shows the changes of the compressive strains
at the maximum stresses with respect to the number of
load cycles. The figure shows that although the compres-
sive strain of cyclic loadings is proportional to logarithm
(CL-033-C-4)
(CL-065-C-4)
(CL-075-C-4)
O. Kontani, N. Ishitobi, J. Kawada, N. Taogoshi, M. Koge and Y.o Umeki / Journal of Advanced Concrete Technology Vol. 14, 634-642, 2016
of the number of load cycles up to 1/100 of fatigue life at
the maximum stress ratios of 0.65 and 0.75, it suddenly
increases when the number of load cycles exceeds 1/10 of
fatigue life. Increasing number of load cycles increase
compressive strain, especially in cases with the same
maximum stress levels. Compared with cases of the
maximum stress levels of 0.65 and 0.75, compressive
strains of the maximum stress levels of 0.65 are greater
than those of the maximum stress levels of 0.75 because
the number of load cycles of 0.65 cases is much larger
than 0.75 cases. For cases with fatigue failure, the com-
pressive strain might not be measured up to the last load
cycles because the strain gage came off due to fatigue
damages on the surface. In case of the maximum stress
ratio at 0.33, the compression strain increases very
gradually compared with cases of 0.65 and 0.75.
Figure 4 shows the changes of the elastic modulus ra-
tios with respect to the number of load cycles. The
modulus ratios were obtained from the hysteresis curves
of cyclic loadings. The stiffness of the cylinder specimens
from the second hysteresis curve is higher than that the
first hysteresis curve. Although the stiffness decreases as
the number of load cycles increases, the rate of decreasing
stiffness at the maximum stress ratio of 0.33 is very low
compared with cases of 0.65 and 0.75.
3. Stud anchor specimen
This section reports the results of the residual strength
tests conducted using the stud anchor specimens in order
to establish the method for evaluating the soundness of
equipment support against machine vibration.
Fig. 3 Changes of the compressive strains at the max.
stresses with respect to the number of load cycles.
Fig. 5 Stud anchor specimen.
638
It is preferable that the failure mode of the anchor sec-
tion is not brittle fracture relating to concrete fracture but
yielding of the anchor enabling the expectation of
toughness. Therefore, the concrete of the stud anchor
section seems to hardly be damaged by cyclic loadings
under the long-term allowable tensile load during the
normal operation of a nuclear power plant.
However, the pullout of anchors and the reduction of
the anchor strength due to cyclic loadings may reduce
rigidity of the equipment support, and is not preferable
from the viewpoint of long-term soundness evaluation.
Therefore, the cyclic tensile loadings with constant am-
plitude were applied to the stud anchor section specimens,
and then static pull-out tests were performed in order to
investigate the reduction of rigidity and anchor strength of
the equipment support section due to cyclic loadings.
3.1 Outline of tests
Figure 5 shows a stud anchor specimen. No reinforce-
ment was arranged in the cone region with an angle of 30°
in reference to the top surface of the specimen so that the
reinforcement doesn’t disturb formations of conical
cracks. The diameter of the headed stud anchor (material:
JIS SCM435) was 19 mm, and the embedment depth of 90
mm was adopted so that the failure mode due to tensile
load became the cone failure of concrete. Table 4 shows
the mix proportion of concrete. The concrete was manu-
factured using gravel and sand produced in Shizuoka
Prefecture and ordinary Portland cement with water to
cement ratio of 0.6.
The anchor specimens were cast in the wood form and
cured at 20°C for one month and then cured at ambient
Fig. 4 Changes of the elastic modulus ratios with respect
to the number of load cycles.
O. Kontani, N. Ishitobi, J. Kawada, N. Taogoshi, M. Koge and Y.o Umeki / Journal of Advanced Concrete Technology Vol. 14, 634-642, 2016 639

Table 4 Mix proportion for anchor specimens.

Unit Mass（kg/m3）
ID W/C(%) s/a(%)
water cement sand gravel
WC60 60 43 155 258 806 1085
Additive : air entraining water reducing agent (0.8% of cement weight)
air entraining agent (0.0025% of cement weight)
Slump :15±2.5cm, Air Contents：4.0±1.0％

Table 5 Test conditions and test results.

Maximum Stress Ratio Residual Strength after Ratio of Resuidual Strength
Specimen # Number of Load Cycles
(Maximum Load) Cyclic Loadings (kN) to Static Test of #7
1 0.75 (69.6kN) 466,797 107.7 1.16
3 719,413 101.3 1.09
0.65 (60.4kN)
4 20,000 92.6 1.00
2 10,000,000 108.5 1.17
5 0.22 (20.2kN) 2,000,000 100.2 1.08
6 20,000 98.1 1.06
7 Static Test － 92.8
The Max. load was determined from the stress ratio and static test of #7.
The pullout strength of the headed stud anchor is 60.6kN according to (AIJ 2010).
The long-term allowable tensile load is 20.2kN, which is 1/3 of the strength from (AIJ 2010).
The static test means static anchor strength without applying any cyclic loadings.

temperatures varying between 3°C and 28°C for 6 months
to prevent further strength development during fatigue
tests. The specimens were cured in sealed condition until
those were submitted to tests. It took one month to finish
all the tests. During cyclic loadings and residual strength
tests, the specimens were exposed to an atmosphere in the
testing laboratory. Compressive strength of 35MPa was
the target, and the compressive strength was 32–33MPa at
the test.
Since the influence of loading speed on fatigue strength is
very small (Inoue et al. 1992), the speed of 10 Hz was
adopted only for the case of 10 million load cycles to
reduce the test time. The pullout displacement of the stud
anchor was measured by displacement transducers. The
tensile loads applied to the anchor were also measured.
S-N curve of the stud anchor specimen can be found in
Eligehausen et al. (2006) and was expressed in Eq. (3) in
the same manner as Eq. (1).

3.2 Conditions of tests s = 1 − log N / 20 (3)

The maximum stress ratios of cyclic loading were set at
where s ： maximum stress ratio (strass ratio to fatigue
0.75, 0.65 and 0.22. The maximum tensile loads of 69.6
failure at number of load cycles of N) and N：fatigue life
kN and 60.4 kN were corresponding to the cases of the
(number of load cycles to fatigue failure at the maximum
maximum stress ratios of 0.75 and 0.65 based on the result
stress ratio of s)
of the static test of the specimen #7 (92.8 kN).
The pullout strength of stud anchor is 60.6 kN when the
3.3 Test results
cone angle of 45° is assumed according to Eq. (2) in AIJ
Table 5 shows the test results. Figure 6 shows the test
(2010). The definition of the cone angle is shown in Fig. 5.
results compared with S-N curves that are the same curves
The long-term allowable tensile load of the anchor is 20.2
found in Fig. 1 (Kishitani and Nishizawa 1987; Elige-
kN that is 1/3 of the pullout strength of the stud anchor.
hausen et al. 2006). The S-N curve is a line graph ob-
Then, the case of the maximum stress ratio of 0.22 (the
tained by linear regression with the vertical axis for ar-
maximum load of 20.2 kN) was regarded as the case
corresponding to the long-term allowable tensile load. 1.2 5.68×1010 : Number of Load
Cycles due to Machine
p = σ c × Ac (2) 1.0 Vibration of 60Hz for 60 years
mum Stress Ratio

Stud Anchor Specimen

where p : the pullout strength of the headed stud anchor 0.8 0.75 (Corn Failure)
Cylinder Specimen
with concrete cone failure, σ c : tensile strength of con- 0.65 (Uni-axial Compression)
0.6
crete (MPa) expressed by σ c = 0.31 Fc , Fc : compres- ：Static Loading
：Cyclic Loading
sive strength of concrete (MPa), Ac : horizontally pro-
Maxim

0.4
jected area of concrete cone (mm2) expressed by 0.22
0.2
Ac = π ⋅ lc ⋅ (lc + D) , lc : embedment depth (mm) and D:
diameter of head of stud anchor (mm) 0.0
The minimum stress ratio of 0.05 was adopted for all 100
1.E+00 104
1.E+04 108
1.E+08 1012
1.E+12 1016
1.E+16 1020
1.E+20

cyclic loadings. Loading speed of 5 Hz was adopted. Number of Load Cycles

Fig. 6 Test results compared with S-N curves.
O. Kontani, N. Ishitobi, J. Kawada, N. Taogoshi, M. Koge and Y.o Umeki / Journal of Advanced Concrete Technology Vol. 14, 634-642, 2016
ranging the maximum stress ratio and the horizontal axis
for arranging the logarithm of load cycles, and is used for
assessing fatigue life corresponding to the maximum
stress ratio. Figure 6 shows the maximum stress ratio and
the residual strength ratio against the static test of the
specimen #7. In addition, the cyclic loading is expressed
by the arrow in the X-direction and the residual anchor
strengths are shown by the arrow in the Y-direction. All
the residual anchor strength after cyclic loadings is equal
or greater by up to 17% than the static strength of the
specimen #7. This phenomenon has a tendency similar to
the results of compression tests of cylinder specimens
after cyclic loadings. It is considered according to Bocca
and Crotti (2003) that there must be conical cracking
surface will be developed by pulling stud anchor and that
the stress singularities in crack tips may be decreased and
redistribution of stress occur due to cyclic loading in-
duced creep. In this way, the residual anchor strength
might be greater than the static strength without any cyclic
loadings. The residual anchor strength is getting greater
with increasing number of loading cycles.
Figures 7 and 8 show the results of tests of the speci-
men #3 at the maximum stress ratio of 0.65 (the number of
load cycles of 719,413). In this case, the inelastic dis-
placement is observed from the first cycle as shown in the
load-displacement relationship and the pullout displace-
ment of stud anchor monotonously increases as the
number of load cycles increases. The inelastic displace-
ment of stud anchor of 0.5 mm is remained in the end of
load cycles. The residual anchor strength and the pullout
displacement at peak load after cyclic loadings are greater
than those of the static test of the specimen #7.
Fig. 7 Tensile load vs. pullout of stud anchor. (Specimen
#3)
Fig. 8 Pullout of stud anchor vs. number of load cycles.
(Specimen #3)
640
Figures 9 and 10 show the test results of the specimen
#1 at the maximum stress ratio of 0.75. In this case, the
inelastic displacement is also observed from the first cycle
as shown in the load-displacement relation and the pullout
displacement of the anchor monotonously increases as the
number of load cycles increases. Although cyclic loadings
stopped just before fatigue failure judging from the pro-
gress of the pullout displacement, the residual strength of
the stud anchor was still higher than the static strength of
the specimen #7. The inelastic displacement of 1.1mm
was remained and the displacement at the peak load is
higher than that of the static test of the specimen #7. In
Fig. 9, the stiffness of residual anchor strength is greater
than the last hysteresis curve of cyclic loading. The reason
might be recovery of remaining strain occurring in 15
hours of rest period between the cyclic loadings and the
residual strength test. The same kind of strain recovery of
plain concrete can be found in Bennett and Muir (1967).
In case of the maximum stress ratios of 0.65 and 0.75,
although the anchor strengths after cyclic loadings are
higher than the static strength of the specimen #7, the
pullout displacement monotonously increases as the
number of load cycles increases, which might cause the
reduction in the rigidity of the equipment support section.
Figures 11 and 12 show the results of tests of the
specimen #2 at the maximum stress ratio of 0.22 (load
cycles of 10 million). In this case, no inelastic displace-
ment is observed in the load-displacement relationships in
spite of a large number of load cycles, and moreover the
load-displacement relationships up to 40kN is similar to
that of the static test of the specimen #7.
The anchor specimen was designed so that cone failure
Fig. 9 Tensile load vs. pullout of stud anchor. (Specimen
#1)
Fig. 10 Pullout of stud anchor vs. number of load cycles.
(Specimen #1)
O. Kontani, N. Ishitobi, J. Kawada, N. Taogoshi, M. Koge and Y.o Umeki / Journal of Advanced Concrete Technology Vol. 14, 634-642, 2016
of concrete occurred when tensile loads are applied. The
cone angle is approximately 35° on average according to
Eligehausen et al. (2006). In these tests, the some concrete
cones were broken into two to four pieces. It seems that
the cone angle became smaller when conical cracking
surface is extended outward with increasing number of
load cycles. When pullout displacement is increased at
residual strength test, the concrete cone might be broken
into pieces because bending moment might be generated
between pullout of anchor and unseparated periphery of
the concrete cone.
When the maximum stress ratio is less than the
long-term allowable tensile load defined by (AIJ 2010),
the pullout displacement of the anchor is not observed at
all even if subjected to load cycles up to 10 million. Then,
it is considered that no reduction of rigidity and strength
will be observed in the anchor section.
The intercept of the X-axis of the fatigue strength
formula of the anchor shown in Fig. 6 demonstrates that
fatigue failure occurs at the number of load cycles of 1020
even if the maximum stress ratio is zero. This is not real-
istic. However, according to Bennett (1967) concrete has
no fatigue limit under 10 million of load cycles differing
from steel materials. The fatigue limit of concrete wasn't
confirmed in the tests also. It is impossible to set fatigue
limit for concrete. However, it might be reasonable to
employ S-N curve for predicting fatigue life of the
equipment support if S-N curve is extrapolated in the
region of load cycles more than 10 million as shown in
Fig. 6.
For example, the turbine generator at 60Hz region in
Japan will generates load cycles of 5.68 x 1010 for 60
years of operation. If the stress level is 0.22 equal to the
Fig. 11 Tensile load vs. pullout of stud anchor. (Specimen
#2)
Fig. 12 Pullout of stud anchor vs. number of load cycles.
(Specimen #2)
641
long-term allowable tensile load level, the fatigue life of
the equipment support of would be 3.98 x 1015, which is
much greater that of turbine support explained the above.
In this way, soundness of equipment support can be ac-
cessed.
4. Summary
The residual compressive strength tests are performed on
cylinder specimens with the number of load cycles and the
maximum stress ratio as parameters. The residual strength
of cylinder specimens after cyclic loadings is higher by
1% to 9% than the static strength without cyclic loadings.
The stiffness of the cylinder specimens from the second
hysteresis curve is higher than that the first hysteresis
curve. The stiffness decreases as the number of load cy-
cles increases. The compressive strain of cyclic loadings
is proportional to logarithm of the number of load cycles
up to 1/100 of fatigue life, it suddenly increases when the
number of load cycles exceeds 1/10 of fatigue life. In-
creasing number of load cycles increase compressive
strain, especially in cases with the same maximum stress
levels.
The residual strength tests are performed on headed
stud anchor specimens with the number of load cycles and
the maximum stress ratio as parameters. All the residual
anchor strength after cyclic loadings is equal or greater by
up to 17% than the static strength without cyclic loadings.
The residual anchor strength is getting greater with in-
creasing number of loading cycles.
In case of the maximum stress ratios of 0.65 and 0.75,
although the anchor strengths after cyclic loadings are
higher than the static strength without cyclic loadings, the
pullout displacement increases monotonously as the
number of load cycles increases.
In case of the maximum stress ratio of 0.22, the anchor
strengths after cyclic loadings are higher than the static
strength without cyclic loadings and no residual dis-
placement is observed in the load-displacement relation-
ships in spite of a large number of load cycles and fur-
thermore the load-displacement relationships of the re-
sidual strength tests is similar to that of the static test
without cyclic loadings.
The results of residual strength tests of anchor speci-
mens after cyclic loadings revealed that no reduction of
rigidity and anchor strength is observed in the anchor
section if the maximum load is not more than the
long-term allowable tensile load.
Acknowledgment
This research was performed as a part of the cooperative
research performed by electric industries for the common
purpose. We sincerely thank Professor Emeritus Katsuki
Takiguchi of Tokyo Institute of Technology, Dr. Hideo
Kasami, Professor Shigemitsu Hatanaka of Mie Univer-
sity, Professor Ichizo Kishimoto of Kinki University and
Professor Ippei Maruyama of Nagoya University who
instructed us in the implementation of this study.
O. Kontani, N. Ishitobi, J. Kawada, N. Taogoshi, M. Koge and Y.o Umeki / Journal of Advanced Concrete Technology Vol. 14, 634-642, 2016
References
AIJ, (2010). “Guidelines for various composite structure
design.” The Architectural Institute of Japan. (in
Japanese)
AIJ, (2015). “Guidelines for maintenance and manage-
ment of structures in nuclear facilities.” The Architec-
tural Institute of Japan. (in Japanese)
Awad, M. E. and Hilsdorf, H. K., (1974). “Strength and
deformation characteristics of plain concrete subjected
to high repeated and sustained loads.” In: Abeles
Symposium Fatigue of Concrete, ACI SP-41-1, l-14.
Ballatore, E. and Bocca, P., (1997) “Variations in the
mechanical properties of concrete subjected to low
cycle loads.” Cement and Concrete Research, 27(3),
453-462.
Bennet, E. W. and Muir, S. E. St J., (1967). “Some fatigue
tests of high-strength concrete in axial compression.”
Magazine Concrete Research, 19, 113-117.
Bocca, P. and Crotti, M., (2003). “Variations in the
mechanical properties and temperature of concrete
subjected to low cycle loads, including high loads.”
642
Materials and Structures, 36(1), 40-45.
Eligehausen, R., Mallée, R. and Silva, J. F., (2006).
“Anchorage in concrete construction.” Ernst & Sohn.
Inoue, S. Nishibayashi, S. and Yoshino, A., (1992).
“Study on the compressive fatigue characteristics of
concrete and the characteristic values of fatigue
strength.” Journal of the Japan Society of Civil
Engineers, 451, 59-67. (in Japanese)
Kishitani, K. and Nishizawa, N. eds., (1987). “Fatigue
-Durability series of concrete structures-.” Gihodo
Shuppan, Tokyo. (in Japanese)
Matsuzaki, I., Abe, Y. and Usami, Y., (1981), “Study on
the bearing capacity of embedded steel plates used for
supporting machines and pipes (Part 2) -Tensile or
shear fatigue strength of headed steel anchor-.” Annual
Report of Kajima Technical Research Institute, 29,
61-66. (in Japanese)
Rehm, G. and Eligehausen, R., (1979). “Bond of ribbed
bars under high cycle repeated loads.” ACI Journal,
76(1), 297-313.
Journal of Advanced Concrete Technology Vol. 14, 299-310 June 2016 / Copyright © 2016 Japan Concrete Institute 299

Scientific paper

Accelerated Aging of Concrete Dry Cask Storage Systems for Nuclear

Waste
Arezou Attar1, Bora Gencturk2*, Mohammad Hanifehzadeh3 and Kaspar Willam4

Received 1 April 2016, accepted 18 June 2016 doi:10.3151/jact.14.299

Abstract
Dry cask storage systems (DCSS) are widely used worldwide for storage of spent nuclear fuel (SNF). Particularly, in
the United States, other than the SNF pools, DCSS are the only means for storage of SNF. In the United States, the
DCSS are licensed for an initial 20 years (with a possible extension of 40 years). The absence of a long-term (or perma-
nent) storage facility has brought up concerns regarding the long-term performance of DCSS, which may now have to
be used for extended durations reaching over 100 years. The DCSS with an exposed concrete overpack account for ap-
proximately 61% of the DCSS inventory in the United States. The corrosion of the steel reinforcing bars (rebar) and the
alkali-silica reactivity (ASR) of concrete have been identified as two of the main degradation mechanisms. In this paper,
the accelerated aging of reinforced concrete (RC) overpacks of a vertical DCSS is evaluated experimentally at the struc-
tural scale. Three 1/3-scale specimens were fabricated. The first specimen was built using a conventional self-
consolidating concrete to serve as a control. The second and third specimens were prepared using special concrete mix-
tures, designed to accelerate the corrosion of rebar and ASR. All three casks were observed for 2 years for aging-
induced deterioration using various non-destructive approaches including visual inspection, half-cell potential, Schmidt
hammer, and ultra-sonic pulse velocity (UPV) measurements. The RC overpacks have been observed to exhibit signifi-
cant distress due to these aging mechanisms. The overall conclusion is that accelerating ASR and corrosion through use
of reactive aggregates and/or addition of chemicals (NaOH and CaCl2 in this particular case) is a viable and practical
approach for large-scale studies. Although accelerated aging of concrete structures have been extensively studied in the
literature, this is one of the first studies on the long-term degradation in DCSS due to corrosion and ASR.

1. Introduction reprocessing plant or a permanent geological repository.

Most nuclear power plants (NPP) in the United States
were constructed between 1965 and 1980. In these
NPPs, the fuel assemblies are typically replaced every
four to six years, and the spent nuclear fuel (SNF) is
stored in storage pools temporarily for radiation shield-
ing and cooling. The pools have limited capacity; there-
fore, in a decade or so after becoming operational, addi-
tional storage of SNF was needed for most NPPs. In
mid-1980’s the utilities have started building sites for
dry storage as an interim solution (Rigby, 2010). The
initial plan was to store the SNF in temporary (pools)
and interim storage (dry casks), then move them to a
1 day (U.S. NRC 2015b). These delays and uncertainty in
PhD candidate, Department of Civil and Environmental
Engineering, University of Houston, Houston, Texas,
USA.
2
Assistant Professor, Sonny Astani Department of Civil
and Environmental Engineering, University of Southern
California, Los Angeles, CA, USA.
*Corresponding author, E-mail: gencturk@usc.edu.
3
PhD candidate, Department of Civil and Environmental
Engineering, University of Houston, Houston, Texas,
USA.
4
Professor, Department of Civil and Environmental
Engineering, University of Houston, Houston, Texas,
USA.
A reprocessing plant, near Buffalo, New York, operated
in the United States from 1966 until 1972; however, it
was shut down permanently for not satisfying revised
regulatory requirements of the time (CBO 1977;
Andrews 2008). Further interest in commercially re-
processing the SNF diminished in the United States
(Chopra et al. 2013). The plan to permanently store the
highly reactive nuclear waste has been debated since
1954. In 1960, the Atomic Energy Commission (AEC)
proposed the use of a salt mine in Lyons, Kansas, which
got suspended in 1972 due to technical difficulties
(Andrews 2006; YuccaMountain 2016). A geologic re-
pository at Yucca Mountain, Nevada, was proposed as
an alternative option in 1986. However, the prospects of
this second alternative also remain unclear up to this
long-term storage and no reprocessing of the used fuel
render the dry storage the only option currently in the
United States. A complete cycle of nuclear fuel is shown
in Fig. 1. As of writing of this paper, stages 9 through 11
do not take place in the United States.
In the United States, currently there are over 1,850
dry cask storage systems (DCSS) in 34 states at 75 sites
(Howard and Akker 2014; U.S. NRC 2015a). Over next
few decades, SNF, at increasing numbers, will have to
be transferred from the pools to DCSS, and if the SNF
production rate stays the same, by the time the last op-
erating reactor is decommissioned in 2055, there will be
A. Atta, B. Gencturk, M. Hanifehzadeh and K. Willam / Journal of Advanced Concrete Technology Vol. 14, 299-310, 2016 300

Table 1 Options for dry storage of spent nuclear fuel (IAEA 2007a, b).
Option Heat Transfer Containment Shielding Feature
Double lid metal gasket
Metal cask Conduction through cask wall Metallic wall Dual purpose
(inert gas)
Air convection around Cavity lining / seal
Concrete cask/silo Concrete and steel overpack Vertical
canister welding (inert gas)
Concrete Air convection around Canister sealing
Concrete wall Horizontal
module canister (inert gas)
Air convection around
Vault Thimble tube (inert gas) Concrete wall Several cases
thimble tube
Drywell/ Heat conduction
Canister (inert gas) Earth Below ground
tunnel through earth

approximately 140,000 metric tons of waste in DCSS
(U.S. GAO 2012), see Fig. 2. This will increase the
number of DCSS to 9,300. DCSSs may be categorized
into two: non-canistered designs where the fuel assem-
blies are directly loaded into the cask and canister-based
storage systems where fuel assemblies are contained in
large multi-assembly canisters and inserted into an over-
pack (Chopra et al. 2013). Certain cask configurations
are designed for dual purpose (both storage and trans-
portation) while others can only be used for storage.
These storages rely on concrete or steel, or both for
shielding the nuclear radiation and heat. Both vertical
the long-term performance of DCSS, particularly, the
impacts of material aging. The majority of commercial
SNF is stored in concrete storage casks or modules. As
of 2012, 60.6% of casks had an exposed reinforced con-
crete (RC) overpack. Therefore, this study focused on
DCSS with an exposed RC overpack. Out of all DCSS
with an exposed RC overpack, 62.84% are in a horizon-
tal and 37.26% are in a vertical configuration (Lambert
4
x 10
15
Spent Nuclear Fuel (ton)

and horizontal configurations are available. For shield-

ing, the horizontal configurations rely on a metal canis-
Accumulation of

ter inside a concrete storage module, while the vertical 10

configurations use metal, concrete or metal and concrete Wet storage
overpacks. The categorization of DCSS by the Interna- Dry storage
tional Atomic Energy Agency (IAEA) is shown in Table 5 Total spent fuel
1 (IAEA 2007a, b).
The initial licensing term approved by the U.S. Nu-
clear Regulatory Commission (NRC) for dry storage is
20 years, which may be extended another 40 years (NEI 0
2011 2020 2030 2040 2050 2060 2067
2014; U.S. NRC 2016). However, in the absence of a Year
permanent storage facility, these casks might be used for Fig 2 Trends in accumulation of spent nuclear fuel in the
longer periods. Therefore, it is important to understand United States (U.S. GAO 2012).

Power Temporary
Generation Storage
6 7

Fresh Fuel
Uranium Fuel
Mining Refining Conversion Enrichment Fabrication
Interim
1 Uranium Ore 2 Concentrate 3 Natural UF6 4 Enriched 5 8 Storage
Uranium

Recovered
Plutonium
Recovered
Uranium

11 9 Reprocessing
Recovered Plutonium
& Uranium

10
Permenant Storage &
Disposal Facilty
Fig. 1 A typical cycle of nuclear fuel. Note that the dashed lines indicate the cycle for SNF.
A. Atta, B. Gencturk, M. Hanifehzadeh and K. Willam / Journal of Advanced Concrete Technology Vol. 14, 299-310, 2016 301

Table 2 Degradation factors that can impact the performance of safety related RC structures for NPPs (IAEA 1998; Naus
2007).
Aging stressors
Percolation of fluid through concrete due
to moisture gradient
Exposure to alkali and magnesium sulfates
present in soils, seawater or groundwater
Exposure to aggressive acids and bases
Combination of reactive aggregate,
high moisture levels, and alkalis
Cyclic loads/vibrations
Exposure to thermal cycles
at relatively low temperatures
Exposure to flowing gas or liquid carrying
particulates and abrasive components
Thermal exposure/ thermal cycling
Irradiation
Exposure to water containing dissolved salts
(e.g., seawater)
Depassivation of steel due to carbonation or
presence of chloride ions
Aging mechanism
Leaching and efflorescence
Sulfate attack
Conversion of hardened cement to
soluble material that can be leached
Alkali-aggregate reactions
leading to swelling
Fatigue
Freeze/thaw
Abrasion; erosion; cavitation
Moisture content changes and
Cracking; spalling; strength loss; reduced
material incompatibility due different
thermal expansion values
Aggregate expansion; hydrolysis
Salt crystallization
Compression or concentration
cells leading to corrosion
Aging effect
Increased porosity and permeability;
lower strength
Expansion and irregular cracking
Increased porosity and permeability
Cracking; gel exudation; aggregate
pop-out; disintegration
Cracking; strength loss
Cracking; scaling; spalling
Section loss
modulus of elasticity
Cracking; loss of mechanical properties
Cracking; loss material
Concrete cracking and spalling; loss of
reinforcement cross section

et al. 2012; Leduc 2012). In this study, we focused on a
vertical configuration for its higher vulnerability to
other events such as tip-over impact in the case of a
hazard.
Like any other structure, DCSS should be capable of
performing their intended function (i.e., safe storage of
SNF) throughout their lifetime. DCSS are subjected to
harsh environmental conditions including high internal
temperatures, radiation, and exposure to weathering
effects from outside; therefore, aging management is
necessary to maintain a high level of safety, and contin-
ued operation. The degradation of cask materials de-
pends on the environmental conditions. As concrete
matures, its microstructure changes constantly, starting
with cement hydration and continuing with the crystalli-
zation of amorphous constituents, and reactions between
cement, paste, aggregate and reinforcing steel. Concrete
may also deteriorate because of its interaction with the
environment and exhibit distress due to shrinkage, creep,
corrosion of reinforcing steel, and alkali-silica reactivity
(ASR) of aggregates. These mechanisms affect the con-
crete, the steel and the interface properties, and reduce
structural performance under mechanical loading such
as impact. IAEA (1998) and Naus (2007) address poten-
tial aging mechanisms, age-related degradation, and
aging management for the concrete containment struc-
tures for NPPs, which can be extended to DCSS. The
degradation factors and their effects are shown in Table
2 (IAEA 1998; Naus 2007).
One of the most common degradation mechanism in
RC DCSS is the corrosion of steel reinforcement (rebar)
(Naus et al. 1996). Most NPP (hence the sites for
DCSS) are located close to a marine environment and
the wind carries the chloride ion into the land (Cicek
2013). ASR may also cause serious safety problems for
DCSS due to higher temperatures inside the concrete
resulting from the decay heat of the SNF and being ex-
posed to high levels of relative humidity in outdoor
conditions. As mentioned earlier, these two degradation
mechanisms are investigated in this paper for RC over-
packs of DCSS.
ASR in concrete were studied extensively using ac-
celerated aging methods (Diamond and Ong 1994;
Smaoui et al. 2005; Poole 2007; Temuujin et al. 2009;
Garci Juenger and Jennings 2011; Nuruddin et al. 2011;
Bahadure and Naik 2013; Gao et al. 2013; Kupwade-
Patil and Allouche 2013; Memon et al. 2013). However,
these studies were conducted at the material level (or in
other words on small-scale specimens). In these studies,
in addition to use of high levels of alkali, the specimens
were conditioned in high temperature with or without
the application of sodium hydroxide solutions. Similarly,
to accelerate corrosion, at the material level, different
method such as submerging small specimens in saline
solutions, passing a current or mixing salts with con-
crete were used (Diamond 1986; Helmuth et al. 1993;
Pakshir and Esmaili 1998; Kelestemur and Yildiz 2006;
Wang et al. 2006; Anacta 2013). However, except for
the last one, these methods are not applicable for large-
scale investigations. There have been a handful of stud-
ies in literature on accelerated aging of structural mem-
bers for ASR and corrosion (Folliard et al. 2006;
Deschenes et al. 2009; Talley et al. 2016). These studies
were focused on structural components (e.g., bridge
bent caps and columns). Therefore, the significance of
this study is in the assessment of an important structural
type under accelerated aging. To our knowledge, there
are no other similar studies performed on DCSS over-
packs where aging induced deterioration due to corro-
sion and ASR were studied at this scale.
A. Atta, B. Gencturk, M. Hanifehzadeh and K. Willam / Journal of Advanced Concrete Technology Vol. 14, 299-310, 2016 302

Table 3 Dimensions of the prototype and scaled models.

Property Prototype (mm) Scaled (mm) Final Dimension (mm)
Height 5722.6 2019.7 2032.0
Outer diameter 3454.4 1219.2 1219.2
Concrete thickness 673.1 237.6 238.1
Liner length 5557.5 1961.5 1968.5
Liner inner diameter 2019.3 712.7 711.2
Liner thickness 44.5 15.7 15.9
Air outlet 1300 × 114 459.5 × 40.4 457.2 × 44.5
Air outlet plate thickness 6.35 2.11 3.2
Air inlet 1300 × 114.0 459.5 × 40.4 457 × 38.1
Air inlet top thickness 50.8 18.0 19.1
Air inlet side thickness 19.1 6.35 6.35

2. Experimental program setting time (37% reduction compared to control) and

Three different mixtures were used to study the effect of
aging on the DCSS. The first mixture was a regular self-
consolidating concrete (SCC) and used as the control.
Based on a literature review (Stanton 1943; Jones and
Tarleton 1958; Ferraris 1995; Smaoui et al. 2005;
Bahadure and Naik 2013), it was decided to accelerate
ASR by adding 0.8% NaOH by weight of cement and
by removing the fly ash from the mixture. The corre-
sponding Na2O equivalent in cement was 0.49%. In
addition, reactive fine aggregate (river sand) and non-
reactive coarse aggregate (pea gravel) were used in all
the mixtures. The addition of 0.8% NaOH to the con-
crete mixture decreased the setting time by 16% and
compressive strength at 28 days by 36% in comparison
to the control mixture. To accelerate corrosion, first the
influence of adding different amounts of chloride to
concrete on the fresh (i.e., setting time) and hardened
properties (i.e., compressive strength) was studied. It
was determined that 4% chloride ion by weight of ce-
ment is the highest amount that results in a reasonable
acceptable concrete strength (24% reduction compared
to control at 28 days). These mixtures with 0.8% of
NaOH and 4% chloride ion by weight of cement are
referred to as NaOH-8 and CaCl2-4, respectively.
The dimensional information of the prototype and the
scaled (2.83 scale-factor) physical cask are provided in
Table 3. The prototype cask information is based on a
review of most commonly used commercial cask de-
signs with an exposed RC overpack. The cask geometry
is shown in Fig. 3 where the main components are iden-
tified. Note that the rebar cage is removed from the
drawings for clarity and it is shown separately in Fig. 4.
The concrete cover was equal to 40 mm in the scaled-
casks and #10 rebar (9.5 mm diameter) was used as both
vertical and horizontal reinforcement. Only one layer of
circumferential (horizontal) reinforcement was used
(note the difference in Fig. 4 in the hoop diameter below
and above the air outlet). The yield strength of the rebar
was found as 551.4 MPa. The rebar were fabricated at a
precast concrete plant and shipped to the University of
Houston Structural Research Laboratory where the test-

Lid

Air
Outlet

Liner

Air
Inlet
Base
Plate

a) b) c)

Fig. 3 Cask model (a) steel components, (b) concrete shield, and (c) cask.
A. Atta, B. Gencturk, M. Hanifehzadeh and K. Willam / Journal of Advanced Concrete Technology Vol. 14, 299-310, 2016 303

Table 4 Properties of the concrete mixtures. models is shown in Fig. 5(c). The temperature and rela-
Mix SCC NaOH-8 CaCl2-4 tive humidity were continuously measured during the
Slump (mm) 258 229 275 testing period of 1.5 years, the temperature varied be-
Density (g/cm3) 2.328 2.320 2.349 tween 4°C and 37°C, and the relative humidity varied
Setting time (min) 285 240 180 between 30% and 100%.
Hydration temperature (°C) 60.71 59.21 63.95 The self-consolidating concrete mixtures were de-
signed to be cohesive enough to fill spaces of almost
ing was conducted. They were placed around the steel any size and shape without segregation or bleeding. The
liner in the Laboratory, see Fig. 5(a), before the con- results of the tests performed on the fresh concretes are
crete molds are assembled as shown in Fig. 5(b). The shown in Table 4. The results presented here are the
three concrete mixtures: two identical SCC mixtures average of at least three measurements. The slump, den-
with 25% flay ash replacement of cement and one SCC sity and setting time were measured based on ASTM
mixture with no fly ash replacement of cement were C1611 (2009c), ASTM C1688 (2013a), and ASTM
delivered separately in three ready-mix trucks. NaOH C403 (2008), respectively. The differences in the slump
powder was added to the fresh SCC mixture with no fly values were mainly attributed to differences in the con-
ash to accelerate ASR, and mixed inside the truck for 15 crete batches and mixture proportions; however, the
min before casting. Similarly, CaCl2 flakes were added addition of chemicals may have also altered the slump.
to one of the SCC mixtures with fly ash to accelerate It is noted here though that the latter is not explicitly
corrosion. Thereafter, casks were poured and formwork studied in this research. The density of the three mix-
was sealed with plywood and plastic sheets to prevent tures was found to be very similar. It was seen that the
rapid loss of moisture. Twenty-eight days after casting, setting time decreased substantially for the CaCl2 mix-
the formwork was removed. A picture of the three cask ture. The effect of NaOH on the setting time was less

#10

#10 76.2 mm
101.6
mm #10
#10

1016 mm
1828.8
mm

1930.4 mm

1143 mm

Fig. 4 Rebar cage.

CaCl2‐4 SCC NaOH‐8

a) b) c)
Fig. 5 (a) Steel parts and rebar cage, (b) mold for concrete casting, and (c) three concrete casks.
A. Atta, B. Gencturk, M. Hanifehzadeh and K. Willam / Journal of Advanced Concrete Technology Vol. 14, 299-310, 2016 304

Table 5 Test plan for concrete casks.

Standard ASTM C805 (2013b) , ASTM C876 (2009a), ASTM C597 (2009b)
Tests performed Schmidt hammer, half-cell potential, UPV, visual inspection
Schmidt hammer: A single measurement at the end of first year
Visual inspection for NaOH-8 cask: 517 and 591 days
Test dates* Visual inspection for CaCl2-4 cask: Monthly inspections after 163 days of age
UPV: Monthly measurements after 163 days of age
Half-cell potential: Monthly measurements after 168 days of age
*UPV and half-cell potential measurement was performed on all three casks

drastic. During casting of the concrete, the temperature ity index test ASTM C 1611(2014) was performed and
inside the concrete (at mid-point along the wall thick- all the mixtures were determined as highly stable. A
ness and close to the bottom of the cask) was measured, summary of material mechanical properties for the three
see Fig. 6. Adding calcium chloride to the mixture concrete mixtures are presented in Fig. 7.
slightly increased maximum hydration temperature from The aging-induced deterioration of the casks was
60.71°C to 63.95°C in comparison to SCC cask. The monitored through a series of non-destructive tests: vis-
sodium hydroxide did not have a significant effect on ual inspection, ultrasonic pulse velocity (UPV), and
the hydration temperature. Furthermore, a visual stabil- half-cell potential over a period of 1.5 years. The test
plan is given in Table 5.

70 3. Results and discussion

SCC
60 NaOH
To determine the variation of compressive strength over
Temperature (C)

CaCl2
the height of the casks, a Schmidt hammer test was per-
50
formed at 370 days. The Schmidt hammer measure-
40
ments were calibrated using the actual compressive
strengths obtained from material testing such that the
30
mean value along the height be the same as the average
of the material test results at 365 days. It is noted here
20 that the purpose of the Schmidt hammer test was to do a
0 1000 2000 3000 4000 5000 6000
relative comparison of the strength along the height of
Time (min)
the cask. To convert the Schmidt hammer index ob-
Fig. 6 Hydration temperature.

50 5
Splitting Tensile Strength (MPa)
Compressive Strength (MPa)

SCC
40 NaOH-8
4
CaCl -4
2
30 3

20 2
SCC
10 1 NaOH-8
CaCl2-4
0 0
a) 1 3 7 28 210 365 b) 1 3 7 28 210 365
Age (Days) Age (Days)
6
Modulus of Rupture (MPa)

2 SCC
NaOH-8
1 CaCl -4
2
0
c) 28 365
Age (Days)
Fig. 7 (a) Compressive strength, (b) splitting tensile strength, and (c) modulus of rupture for the three concrete mixtures.
A. Atta, B. Gencturk, M. Hanifehzadeh and K. Willam / Journal of Advanced Concrete Technology Vol. 14, 299-310, 2016 305

tained from the instrument, the cylinder strength was
used. It is not intended here to study the accuracy of
Schmidt hammer testing or to relate the cylinder tests to
actual field strength. The strength using a Schmidt
hammer was measured at ten equally spaced locations
along the height of the cask. Ten measurements were
collected around each location and the results were av-
eraged. The results of these measurements are shown in
Fig. 8. Before calibration, the SCC cask showed the
highest strength and NaOH-8 cask shows the lowest
strength, which was in line with the material tests. In all
three casks, the strength was found to decrease along the
height of the casks from bottom to top. This is attributed
to the settlement of the coarse aggregate towards the
bottom of the casks since the casting was done from the
top and a self-consolidating mixture was used. Although
some variation along the height is observed, the strength
at the bottom was approximately 17%, 28% and 30%
higher than that at the top for the SCC, NaOH-8 and
CaCl2-4 casks, respectively.
Cracks have been observed on the NaOH cask ap-
proximately 90 days after casting. Similar crack pattern
to those from ASR can also be seen from different proc-
esses such as drying shrinkage, freezing/thawing cycles
and sulfate attack. For the given exposure conditions in
this study, the only possible mechanisms left for such
cracks are the drying shrinkage and ASR (no presence
of sulfates or temperature below freezing). To ensure
and verify that the source of cracks on the casks is ASR,
the following was performed. (1) The casks were kept in
the forms for 28 days after pouring and therefore, the
loss of moisture was been prevented during this period,
which significantly reduces the drying shrinkage. (2)
The concrete was reinforced with steel at a ratio higher
than minimum shrinkage reinforcement requirement
which reduces the likelihood of the cracks due to
shrinkage. (3) SEM/EDS analysis were performed on
concrete samples and gel formation was observed
around the aggregates. Observation of ASR cracks at 90
days of age indicates that the ASR process was acceler-
ated considerably given that in regular structures it takes
approximately 10 years until the appearance of first
cracks (IStructE, 2010). The crack maps for ASR accel-
erated cask at 517 and 591 days are shown in Figs. 9(a)
and (b), respectively. It is seen that the density of crack-
ing increased with time and higher crack density was
observed towards the top of the cask potentially due to
lower concrete strength in that region as explained
above. The crack widths remained between 0.05 mm
and 0.1 mm over the measurement period.
On the CaCl2-4 cask, the first cracks were visually
observed after around 120 days. The crack propagation
over time on the CaCl2-4 cask is shown in Fig. 10. After
less than a year, there was extensive cracking on the
casks and several of the cracks had coalesced. The
propagation of the cracks were in the same direction
with the rebar, which indicates that the cracks were due
to pressure generated by the corrosion products. The
crack width was measured at 23 randomly chosen points
on the casks, and the change over time in the average

SCC NaOH-8 CaCl2-4 Height from the SCC Cask NaOH-8 CaCl2-4
base (mm) (MPa) Cask (MPa) Cask (MPa)
2000 2042 37.3 24.3 24.7
Height from the Base (mm)

1838 36.8 21.1 27.4

1500
1634 38.9 23.1 29.4
1430 37.3 24.3 27.4
1225 38.9 24.7 28.4
1000 1021 43.8 23.9 25.0
817 39.5 27.9 28.4
613 42.0 24.7 27.7
500
408 44.6 30.3 34.7
204 45.0 33.9 35.4
0
20 30 40 50 Avg. 40.42 25.82 28.9
Strength (MPa)
Fig. 8 Schmidt hammer measurement. Note that the vertical dashed lines in the figure shows the average compressive
strength at 365 days as obtained from cylinder tests.

a) b)
Fig. 9 Crack propagation on cask with NaOH-8 mixture (a) 517 days (b) 591 days.
A. Atta, B. Gencturk, M. Hanifehzadeh and K. Willam / Journal of Advanced Concrete Technology Vol. 14, 299-310, 2016
crack width is shown in Fig. 11. It is seen that at 569
days (i.e., the last measurement) the cracks were 0.33
mm wide on average with the largest and smallest
measurements being 0.1 mm and 1.0 mm.
ASR and corrosion crack propagation is affected by
the size, distribution and amount of the reinforcement.
Lollipop specimens were prepared with both #10 (bar
diameter of 9.52 mm) and #19 (bar diameter of 19.05
mm) reinforcing bars with concrete covers of 38.1 mm
and 63.5 mm, respectively. In both cases of ASR and
corrosion, the crack density was lower in specimens
with more concrete cover and larger size of reinforce-
ment. The increase in crack width due to corrosion in
specimens with #19 rebar was less compared to those
with a #10 rebar. These findings are in line to what has
been reported in the literature (Houston et al. 1972;
Rasheeduzzafar et al. 1992; Du et al. 2014).
The half-cell potential was measured to evaluate the
a)
c)
e)
Fig. 10 Crack propagation on cask with CaCl2-4 mixture (a) 163 days (b) 186 days (c) 225 days (d) 295 days (e) 512
days (f) 569 days.
0.7
0.6
Crack Width (mm)
0.5
0.4
0.3
0.2
0.1
0 168 199 248 291 334 469 499 532 569
163 186 225 195 512
Age (Days)
Fig. 11 Crack width growth on CaCl2 cask.
306
probability of corrosion in each of the casks and the
results are presented in Fig. 12. At each measurement
date, more than 150 points were read on one-half of the
cask along a grid of lines. The measurements were con-
sistent at different locations on the casks; therefore, in
Fig. 12, only the average of these measurements (in a
1.5 year time period) is shown. Half-cell potentials in
SCC and NaOH-8 casks were similar and always higher
than -200mV/CSE, which indicates that the probability
of corrosion is less than 10%. On the other hand, this
measurement for CaCl2-4 cask was always lower than -
350mV/CSE that shows that the probability of corrosion
is more than 90%. As an overall trend, the half-cell po-
tential has decreased over time, which is an indicator of
increased corrosion activity in the CaCl2-4 cask.
The UPV was measured at three different levels and
on two sides of each cask as shown in Fig. 13. The av-
erage of UPV measurements over time is shown in Fig.
b)
d)
f)
-600
-500
Potential (mV/CSE)
-400
SCC
-300 NaOH-8
CaCl2-4
-200
-100
0
569 Age (Days)
Fig. 12 Half-cell potentials of reinforcing steel in three
casks.
A. Atta, B. Gencturk, M. Hanifehzadeh and K. Willam / Journal of Advanced Concrete Technology Vol. 14, 299-310, 2016 307

14 along with the temperature and relative humidity approximately 36% at 28 days compared to control
measurements during the same period. These results self-consolidating concrete (SCC).
indicated a higher velocity in the SCC cask compared to • Similarly, it was observed that the direct addition of
NaOH-8 and CaCl2-4 casks, which had similar UPV CaCl2 is an effective way of accelerating the corro-
values. These findings were consistent with the material sion-induced aging. In this study, corrosion cracks
testing and Schmidt hammer measurements. These were observed at approximately 120 days of age with
measurements also indicate that there is a strong corre- 4% addition of CaCl2 by weight of cement into the
lation between the temperature and the UPV values. fresh concrete. However, it was seen that the concrete
However, since the variation of relative humidity was strength reduced by 24% at 28 days in comparison to
almost constant through time, its effect could not be the control SCC mixture. The CaCl2 and SCC mix-
quantified. The UPV results were not able to capture tures were identical except for the addition of chlo-
any degradation in concrete properties due to aging at ride; therefore, all of this reduction was due to the
this scale in contrast to what was initially thought be- presence of chloride.
cause of their sensitivity to moisture content, tempera- • The average width of cracks in CaCl2 cask increased
ture and the presence of rebar and other minor defects from about 0.1 mm to 0.3 mm in approximately 1 year,
inside the concrete around the measurement locations. and the maximum crack width reached to 1.0 mm.
• Rebound hammer tests showed an increase in strength
4. Conclusions towards the bottom of the cask. This is attributed to
the casting of the self-consolidating concretes from
Based on the findings of this study, the following con- the top, which resulted in a higher coarse aggregate
clusions are drawn: content towards the bottom due to settlement. This
• Adding NaOH to the concrete mixture, removing any variation in strength has also resulted in the observa-
supplementary cementitious materials with alkali- tion of higher crack density at the top of both NaOH
silica reactivity (ASR) suppression characteristics and CaCl2 casks.
(e.g., fly ash) and using reactive fine aggregate could • Adding chloride or alkaline to the mixture decreased
be used as an effective method to accelerate ASR at the setting time of the concrete by 37% and 16％, re-
the structural level. An 8% addition of NaOH by spectively in comparison to the control SCC mixture.
weight of cement into the fresh concrete reduced the Addition of chloride also increased the hydration tem-
time to first observation of ASR distress from tens of perature by 5%.
years to approximately 90 days. However, the con- • Although not studied directly as a part of this paper, it
crete strength with this special mixture reduced by is expected that environmental aging due to ASR and
corrosion observed in these tests will impact the struc-
tural performance of the casks, particularly, in the
Side 1
Level 1
case of impact (e.g., a tip-over event). On the other
hand, it is not anticipated that the thermal or radiation
shielding performance will be effected from environ-
mental aging. Testing and quantitative analysis on
Level 2 combined effects of environmental aging and me-
chanical or shielding performance remain as a future
45° research topic.
Level 3
Acknowledgements
Side 2
The financial support for this project was provided by
the United States Department of Energy through the
Fig. 13 Locations of UPV measurement on casks. Nuclear Energy University Program under the Contract
60
Temperature (C)
Ultrasonic Pulse Veloity (m/sec)

4600 SCC 40
NaOH-8
4400
CaCl2-4 20
4200
0
100 200 300 400 500 600
4000 b)
Age (Days)
Relative Humidity

100
3800

3600 50
3400
100 200 300 400 500 600
a) 0
Age (Days) 100 200 300 400 500 600
c) Age (Days)
Fig. 14 (a) UPV measurement on casks (b) temperature from 100 to 600 days (c) relative humidity from 100 to 600 days.
A. Atta, B. Gencturk, M. Hanifehzadeh and K. Willam / Journal of Advanced Concrete Technology Vol. 14, 299-310, 2016
No. 00128931. The findings presented herein are those
of the authors and do not necessarily reflect the views of
the sponsor.
References
Anacta, E. T., (2013). “Effect of salt-contaminated
mixing water and aggregates on time-to-initiate rebar
corrosion in concrete.” International Journal of
Scientific & Engineering Research, 4(4), 1524-1527.
Andrews, A., (2006). “Radioactive waste streams:
Waste classification for disposal.” Report No.
RL32163, Congressional Research Service (CRS),
Report for Congress.
Andrews, A., (2008). “Nuclear fuel reprocessing: U.S.
Policy Development.” Report No. S22542,
Congressional Research Service (CRS), The Library
of Congress.
ASTM, (2008). “Standard test method for time of setting
of concrete mixtures by penetration pesistance.”
ASTM C403/C403M-08, American Society for
Testing of Materials (ASTM), West Conshohocken,
PA.
ASTM, (2009a). “Standard test method for corrosion
potentials of uncoated reinforcing steel in concrete.”
ASTM C876-09, American Society for Testing of
Materials (ASTM), West Conshohocken, PA.
ASTM, (2009b). “Standard test method for pulse
velocity through concrete.” ASTM C597-09,
American Society for Testing of Materials (ASTM),
West Conshohocken, PA.
ASTM, (2009c). “Standard test method for slump flow
of self-consolidating concrete.” ASTM
C1611/C1611M-09b, American Society for Testing of
Materials (ASTM), West Conshohocken, PA.
ASTM, (2013a). “Standard test method for Density and
Void Content of Freshly Mixed Pervious Concrete.”
ASTM C1688/C1688M-13, American Society for
Testing of Materials (ASTM), West Conshohocken,
PA.
ASTM, (2013b). “Standard test method for rebound
number of hardened concrete.” ASTM C805/C805M-
13a, American Society for Testing of Materials
(ASTM), West Conshohocken, PA.
ASTM, (2014). “Standard test method for slump flow of
self-consolidating concrete.” C1611/C1611M − 14,
American Society of Testing and Materials Academic
& Science (ASTM), West Conshohocken, PA.
Bahadure, B. M. and Naik, N. S., (2013). “Effect of
alkaline activator on workability and compressive
strength of cement concrete with RHA.”
International Journal of Computational Engineering
Research, 3(5), 15-20.
CBO, (1977). “Nuclear reprocessing and proliferation:
Alternative approaches and their implications for the
federal budget.” Report No. O-89-442, The Congress
of the United States, Congressional Budget Office
(CBO), Washington, D.C.
Chopra, O. K., Diercks, D., Ma, D., Shah, V. N., Tam,
308
S.-W., Fabian, R. R., Han, Z. and Liu, Y. Y., (2013).
“Managing aging effects on dry cask storage systems
for extended jong-term storage and transportation of
used fuel.” FCRD-UFD-2013-000294, Argonne
National Laboratory (ANL), Department of Energy
(DOE), Lemont, IL.
Cicek, V., (2013). “Cathodic protection: Industrial
solutions for protecting against corrosion.” Wiley-
Scrivener, Beverly, MA.
Deschenes, D. J., Bayrak, O. and Folliard, K. J., (2009).
“ASR/DEF-damaged bent caps: Shear tests and field
implications.” Report No. 12-8XXIA006, Texas
Department of Transportation (TxDOT), Austin, TX.
Diamond, S., (1986). “Chloride concentrations in
concrete pore solutions resulting from calcium and
sodium chloride admixtures.” Cement, Concrete and
Aggregates, 8(2), 97-102.
Diamond, S. and Ong, S,. (1994). “Effects of added
alkali hydroxides in mix water on long-term SO42-
concentration in pore solution.” Cement & Concrete
Composites, 16(3), 219-226, DOI: 10.1016/0958-
9465(94)90019-1.
Du, X., Jin, L. and Zhang, R., (2014). “Modeling the
cracking of cover concrete due to non-uniform
corrosion of reinforcement.” Corrosion Science, 89,
189-202.
Ferraris, C. F., (1995). “Alkali-silica reaction and high
performance concrete.” NISTIR 5742, National
Institute of Standards and Technology (NIST),
Gaithersburg, MD.
Folliard, K. J., Barborak, R., Drimalas, T., Du, L.,
Garber, S., Ideker, J., Ley, T., Williams, S., Juenger,
M., Fournier, B. and Thomas, M. D. A., (2006).
“Preventing ASR/DEF in new concrete: Final report.”
FHWA/TX-06/0-4085-5, Center for Transportation
Research at The University of Texas at Austin, Austin,
TX.
Gao, X. X., Multon, S., Cyr, M. and Sellier, A., (2013).
“Alkali-silica reaction (ASR) expansion: Pessimum
effect versus scale effect.” Cement and Concrete
Research, 44, 25-33.
Garci Juenger, M. C. and Jennings, H. M., (2011).
“Effects of high alkalinity on cement pastes.” ACI
Materials Journal, 98(3), 251-255.
Helmuth, R., Stark, D., Diamond, S. and Moranville-
Regourd, M., (1993). “Alikali-silica reactivity: An
overview of research.” SHRP-C-342, Strategic
Highway Research Program (SHRP), Washington,
DC.
Houston, J. T., Atimtay, E. and Ferguson, P. M., (1972).
“Corrosion of reinforcing steel embedded in
structural concrete.” 112-1F, Texas Highway
Department, US Department of Transportation
Federal Highway Administration (FHWA), Austin,
TX.
Howard, R. and Akker, B. v. d., (2014). “Considerations
for disposition of dry cask storage system materials at
end of storage system life.” Symposium on Recycling
A. Atta, B. Gencturk, M. Hanifehzadeh and K. Willam / Journal of Advanced Concrete Technology Vol. 14, 299-310, 2016
of Metals arising from Operation and
Decommissioning of Nuclear Facilites, April 8-10,
Nyköping, Sweden, D-4.
IAEA, (1998). “Assessment and management of ageing
of major nuclear power plant components important
to safety: Concrete sontainment buildings.” IAEA-
TECDOC-1025, International Atomic Energy Agency
(IAEA), Vienna, Austria.
IAEA, (2007a). “Operation and maintenance of spent
fuel storage and transportation casks/containers.”
IAEA-TECDOC-1532, International Atomic Energy
Agency (IAEA), Vienna, Austria.
IAEA, (2007b). “Selection of away-from-reactor
facilities for spent fuel storage, a guidebook.” IAEA-
TECDOC-1558, International Atomic Energy Agency
(IAEA), Vienna, Austria.
IStructE, (2010). “Addendum, The Institution of
Structural Engineers.” London, UK.
Jones, F. E. and Tarleton, R. D., (1958). “Reaction
between aggregates and cements.” Parts I-IV, DSIR
National Building Studies Research, HMSO, London,
UK.
Kelestemur, O. and Yildiz, S., (2006). “Effect of various
NaCl concentration on corrosion of steel in concrete
produced by addition of styrofoam.” Gazi University
Journal of Science, 19(3), 163-172.
Kupwade-Patil, K. and Allouche, E., (2013). “Impact of
alkali silica reaction on fly ash-based geopolymer
concrete.” ASCE Journal of Materials in Civil
Engineering, 25(1), 131-139.
Lambert, J. D., Bakhtiari, S., Bondar, I., Kot, C. and
Pence, J., (2012). “NRC job code V6060: Extended
In-situ and real time monitoring.” ANL/NE‐12/18,
Argonne National Laboratory (ANL), Lemont, IL.
Leduc, D. R., (2012). Dry Storage of Used Fuel
Transition to Transport.” FCRD-UFD-2012-000253,
Savannah River National Laboratory (SRNL), Aiken,
SC.
Memon, F. A., Nuruddin, M. F., Khan, S., Shafiq, N.
and Ayub, T., (2013). “Effect of sodium hydroxide
concentration on fresh properties and compressive
strength of selfcompacting geopolymer concrete.”
Journal of Engineering Science and Technology, 8(1),
44-56.
Naus, D. J., (2007). “Primer of durability of nuclear
power plant reinforced concrete structures - A review
of Pertinent Factors.” NUREG/CR-
6927, ORNL/TM-2006/529, Oak Ridge National
Laboratory (ONRL), Washington, DC.
Naus, D. J., Oland, C. B. and Ellingwood, B. R., (1996).
“Report on aging of nuclear power plant reinforced
concrete structures.” NUREG/CR-6424 BNL-
NUREG-13148, Oak Ridge National Laboratory
(ONRL), Oak Ridge, TN.
NEI, (2014). “Guidance for operations based aging
management for dry cask storage.” Report No. NEI
14-03, Nuclear Energy Institute (NEI), Washington
D.C.
309
Nuruddin, M. F., Demie, S., Ahmed, M. F. and Shafiq,
N., (2011). “Effect of superplasticizer and NaOH
molarity on workability, compressive strength and
microstructure properties of self-compacting
geopolymer concrete.” International Journal of
Environmental, Chemical, Ecological, Geological
and Geophysical Engineering, 5(3), 187-194.
Pakshir, M. and Esmaili, S., (1998). “The effect of
chloride ion concentration on the corrosion of
concrete.” Scientia Iranica, 4(4), 201-205.
Poole, J. L., (2007). “Modeling temperature sensitivity
and heat evolution of concrete.” Thesis (PhD). The
University of Texas at Austin, Austin, TX.
Rasheeduzzafar, Al-Saadoun, S. S. and Al-Gahtani, A.
S., (1992). “Corrosion cracking in relation to bar
diameter, cover, and concrete quality.” Journal of
Materials in Civil Engineering, 4(4), 327-342.
Rigby, D. B., (2010). “Evaluation of the technical basis
for extended dry storage and transportation of used
nuclear fuel.” United States Nuclear Waste Technical
Review Board (NWTRB), Arlington, VA,
http://www.nwtrb.gov/reports/eds_rpt.pdf.
Smaoui, N., Bérubé, M. A., Fournier, B., Bissonnette, B.
and Durand, B., (2005). “Effects of alkali addition on
the mechanical properties and durability of concrete.”
Cement and Concrete Research, 35(2), 203-212.
Stanton, T. E., (1943). “Studies to develop an
accelerated test procedure for the detection of
adversely reactive cement-aggregate combination.”
American Society of Testing and Materials
Proceeding, 43, 875-893.
Talley, K. G., Kapitan, J. G. and Breen, J. E., (2016).
“Methods for approximation of ASR/DEF damage in
concrete columns.” ACI Structural Journal, 113(1),
105-110.
Temuujin, J., Williams, R. P. and Riessen, A. v., (2009).
“Effect of mechanical activation of fly ash on the
Properties of gopolymer cured at ambient
temperature.” Journal of Materials Processing
Technology, 209(12-13), 5726-5280.
U.S. GAO, (2012). “Spent Nuclear fuel, accumulating
quantities at commercial reactors present storage and
other challenges.” GAO-12-797, United States
Government Accountability Office (GAO),
Washington, D.C.
U.S. NRC, (2015a). “Backgrounder on dry cask storage
of spent nuclear fuel.” United States Nuclear
Regulatory Commission (U.S. NRC), Available from:
http://www.nrc.gov/reading-rm/doc-collections/fact-
sheets/dry-cask-storage.html, Accessed on February
10, 2016.
U.S. NRC, (2015b). “Backgrounder on licensing Yucca
Mountain.” United States Nuclear Regulatory
Commission (U.S. NRC), Available from:
http://www.nrc.gov/reading-rm/doc-collections/fact-
sheets/yucca-license-review.html, Accessed on
February 10, 2016.
U.S. NRC, (2016). “NRC regulations, title 10 Code of
A. Atta, B. Gencturk, M. Hanifehzadeh and K. Willam / Journal of Advanced Concrete Technology Vol. 14, 299-310, 2016 310

federal regulations (10 CFR).” U.S. Nuclear 173-188.

Regulatory Commission (U.S. NRC), Rockville, MD. YuccaMountain, (2016). “Nuclear waste policy
Wang, K., Nelsen, D. E. and Nixon, W. A., (2006). delimma.” Available from: http://www.yuccamoun
“Damaging effects of deicing chemicals on concrete tain.org/, Accessed on February 10, 2016.
materials.” Cement & Concrete Composites, 28(2),
Journal of Advanced Concrete Technology Vol. 14, 335-343 July 2016 / Copyright © 2016 Japan Concrete Institute 335

Scientific paper

Review of Research on the High Temperature Resistance of Concrete

Structures in Chinese NPP
Jianzhuang Xiao1*, Wengang Xie2 and Qinghai Xie3

Received 1 April 2016, accepted 27 June 2016 doi:10.3151/jact.14.335

Abstract
In line with the strategic energy goals and sustainable development, considerable investment has been made on nuclear
energy, leading to construction of a growing number of nuclear power plants (NPPs) across China. Concrete is the main
structural materials applied in many key elements of a typical NPP including reactors. Among other stringent require-
ments, resistance to thermal loads both at early ages and during operating life is crucial for the concrete used in nuclear
facilities. This paper provides an overview of high temperature resistance of concrete structures in Chinese NPP, in-
cluding information on construction and operation. Early-age thermal cracking of mass concrete structures due to dif-
ferential thermal stresses induced by hydration heat has been highlighted in literature as a major issue. The common
approaches to tackling this issue are based on limiting the maximum temperature in the concrete and temperature dif-
ference between hotter interior and cooler exterior of the concrete. These approaches include the optimum mix ratio
method and the reasonable design of the construction technology. During operation, the temperature control of con-
tainment under steady or accidental case is an important prerequisite to ensure the safety of NPP, especially after a loss of
coolant accident (LOCA). In addition, various coatings used for thermal insulation in China’s NPP are introduced and
compared. Based on the existing literature, further studies involving longtime monitoring of temperature, strain and
displacement are found necessary to obtain a better understanding of thermal resistance, long-term performance and
safety of concrete containment used in NPP.

1. Introduction the share of nuclear power in China’s total power gen-

As a new energy which is clean and efficient, nuclear
power has gained worldwide attention and been widely
used in Europe, America and other developed countries
and regions. It is also the important strategic direction of
China's energy in the future. By 1 July 2015, there are 30
countries and regions operating nuclear power plants
(NPPs) in the world (Schneider and Froggatt 2015).
Table 1 shows 15 countries and regions with the largest
amount of nuclear electricity generation in 2014. In ad-
dition, although no unit in Japan operated or generated
power due to the Fukushima accident, the number of
reactors is also presented to remind the readers about the
Fukushima accident. Based on the evaluation of the
Fukushima accident, the aging degradation has little
influence on the loss of function in systems and com-
ponents important to safety. Additional investigation
needs to be launched to confirm the status of equipment
in the future when the further understanding is gained
(Sekimura 2012).
As seen in Table 1, despite the growing investment on
nuclear energy and increasing number of NPPs in China,
1 measures adopted in Chinese NPP to control the tem-
Professor, Department of Structural Engineering, Tongji
University, Shanghai, China.
*Corresponding author, E-mail: jzx@tongji.edu.cn
2
Master student, Department of Structural Engineering,
Tongji University, Shanghai, China.
3
PhD student, Department of Structural Engineering,
Tongji University, Shanghai, China.
eration is still small. Based on the Medium- and
Long-term Development Plan of the State on Nuclear
Power issued by the State Council in 2007, it plans to
increase China's total nuclear power capacity to 40 mil-
lion kW in 2020, making the under-construction capacity
not less than 18 million kW. This is expected to lead to an
increase proportion of China's nuclear power installed
capacity from 1.7% in 2005 to about 4% in 2020 (Ye
2012). As shown in Table 2, the total capacity of 28
commercial nuclear power units operating in China
mainland is 168.99 TWh, accounting for 3.01% of the
national total power generation in 2015. However, the
Chernobyl nuclear accident in the former USSR and the
leak of Japan's Fukushima NPP have attracted attentive
focus around the world on the safety of nuclear power.
During the process of construction and operation of
NPP, the concrete structure and facility might be affected
by exposure to elevated temperatures. From an engi-
neering point of view, a number of measures have been
investigated to guarantee the safety of NPP, including the
application of a composite liner glued onto the inner side
of the containment wall (Costaz and Danisch 1997; Wu
et al. 2013). This paper begins with reviewing the
perature in the concrete structure during the construction
phase. The paper then continues by analyzing the be-
havior of containment in the natural state as well as under
the loss of coolant accident (LOCA). Furthermore,
various high temperature coatings used in NPP are
summarized.
J. Xiao, W. Xie and Q. Xie / Journal of Advanced Concrete Technology Vol. 14, 335-343, 2016 336

Table 1 Number of nuclear reactors and nuclear electricity generation.

Nuclear reactors (1 July 2015) Nuclear electricity (2014)
Country or Region
Operating (units) Constructing (units) Generation (TWh) Share in electricity mix
USA 99 0 797 20.2%
France 58 0 415.9 76.9%
Russia 34 3 171 18.6%
South Korea 24 5 150.4 30.1%
China 27 18 130.5 2.39%
Canada 19 0 100.9 16.0%
Germany 8 0 91.7 15.8%
Ukraine 15 0 82.0 48.6%
Sweden 9 0 62.2 41.2%
UK 16 0 57.8 16.6%
Spain 7 0 54.8 20.4%
Taiwan, China 6 0 40 18.8%
India 20 7 33 3.40%
Belgium 6 0 32 50.0%
Czech Rep. 6 0 28.6 33.3%
Japan 43* 0 - -
* All reactors in Japan had been shut down.

Table 2 Power production statistics of Chinese NPP in 2015*.

NPP Reactor type Installed capacity (MWe) Generation (TWh)
Qinshan NPP CNP300 310 2.57
The 2nd Qinshan NPP CNP600 2620 20.29
The 3rd Qinshan NPP CANDU 6 1456 11.24
Fang Jiashan NPP CNP1000 2178 15.168
Hongyanhe NPP CPR1000 3356.37 14.47
Tianwan NPP VVER-1000 2120 16.62
Changjiang NPP CNP600 650 0.07
Daya Bay NPP M310 1968 15.43
Lingao NPP phase I M310
4152 32.28
Lingao NPP phase II CPR1000
Yangjiang NPP CPR1000 2172 12.95
Ningde NPP CPR1000 3267 19.59
Fuqing NPP CPR1000 2178 8.34
Total 26427.37 168.99
*The data come from http://www.china-nea.cn/html/2016-01/35018.html.

2. Temperature control during construction
The hydration of Portland cement is an exothermic re-
action. In mass concrete, (e.g., concrete containment and
foundation), where the surface area of concrete is rela-
tively small compared to its total volume, the hydration
heat may result in considerable temperature difference
between interior and exterior of concrete. The thermal
stresses caused by such temperature gradients can readily
induce the development of early-age thermal cracks in
concrete, which is highly undesired from both aesthetic
and durability points of view (Acker and Ulm 2001;
Wang 2011). Temperature control of concrete after its
casting is considered to be the key to limit the concrete
thermal cracking (Ling et al. 2005; Schindler 2004). The
delayed ettingite formation (DEF) because of high tem-
perature and the resulting stresses generated by the ex-
panding ettringite have been particularly identified as a
cause of concrete cracking, especially in European
pressurized-water reactor (EPR) nuclear power plants
(NPPs) (Zhao and Zhou 2012). In order to limit the
maximum temperature and the temperature difference
between inside and outside of the concrete within the
prescribed standard, the concrete temperature should be
reduced from two aspects as the optimum mix design and
the reasonable construction technology.
2.1 Concrete mix design
Hydration heat is the main reason that the volume of
concrete grows larger when the temperature increases.
Medium or low heat cement can be chosen to control the
hydration heat of binding materials in concrete strictly.
Meanwhile, adding mineral admixture and chemical
additive should also be helpful (Xu 2010). The mineral
admixture can partly replace the Portland cement to
lower and delay the peak temperature of hydration.
Furthermore, some chemical additive such as Polycar-
boxylate water-reducing agent (PCA) can retard cement
hydration and increase the dosage of slag or fly ash in
concrete. The concrete proportion of the reactor building
J. Xiao, W. Xie and Q. Xie / Journal of Advanced Concrete Technology Vol. 14, 335-343, 2016
Table 3 The concrete proportion (Xu 2010; Wang et al. 2012).
The reactor building raft of Taishan NPP
Material
Specification quantity
Sand Medium sand
Coarse 5~16
aggregate 16~31.5
Cement PII 42.5
Slag powder S95
Fly ash I 100
Additive PCA*
Water Potable water
*Polycarboxylate water-reducing agent.
Table 4 Examples of aggregates based on density (Kaplan 1989).
Type aggregate Density (kg/ m3)
Normal weight 1520-1680
Light weight < 1120
Heavy weight > 2080
raft of Taishan and Yangjiang NPPs can be taken as a
typical example, which is listed in Table 3. The concrete
unit weights of both Taishan and Yangjiang NPPs meet
the standard of normal-weight aggregates in Table 4,
which have been used in the biological shields of pres-
surized-water reactors (PWRs) (Kaplan 1989; Willam et
al. 2013).
2.2 Reasonable construction scheme
During the maintaining process of the mass concrete,
temperature measurement is the main gist of temperature
control to decide the maintenance time of concrete, and
to adjust the layers of the proof material. This procedure
can limit the temperature difference and cooling rate
inside the concrete, and thereby can prevent the devel-
opment and growth of cracks. During the construction of
the foundation in Hongyanhe NPP, a temperature moni-
toring and controlling system was applied to avoid the
negative influence of winter weather and to assure the
program on schedule. In this strategy, 54 temperature
sensors and 57 strain sensors were set to monitor the
temperature and stress development, along with an alert
system being triggered when stress got close to the ten-
sile strength of concrete (Yang et al. 2009; Wei et al.
2008).
In the southern coastal area of China, the high tem-
perature and humid environment brings many difficulties
to the pouring and maintaining process of NPP founda-
tion, such as overhigh central temperature and great
temperature gradient, which are very difficult for con-
trolling cracks. Gu et al. (2010) undertook an intensive
research on controlling cracks during the maintaining
period by taking some NPPs in south China as examples.
A finite element numerical simulation model was built by
ANSYS to compare with the monitoring data of tem-
perature and strain during the pouring of concrete. Ac-
cording to the numerical calculation results, the control
337
The 3rd reactor building raft of Yangjiang NPP
Usage /(kg/ m3) Specification quantity Usage /(kg/ m3)
780 Medium sand 785
5~16
1050 1075
16~31.5
240 PII 42.5 280
50 - -
I 100
3.9 PCA 3.8
151 Potable water 148
Examples Concrete density (kg/ m3)
Limestone, granite, sandstone 2400
Volcanic pumice, blast furnace slag, cinder < 1850
Magnetite, barytes, ferrophosphorus > 4000
section of concrete and stress orientation in this section
are determined to provide the guidance of the position
design of sensors. The data from the sensors could pro-
vide useful information for the effective adjustment of
maintaining measures. When the temperature in concrete
is overhigh, some cooling measures can be adopted like
ice bag cooling, rebar covering, spraying on pouring pipe
during stratified slope concreting and adjusting the cov-
ering layer thickness based on the real-time monitoring
data in the conservation process. The research shows that
there can be no obvious cracks even when the tempera-
ture reaches 80 °C.
Considering the enormous pouring volume of con-
tainment of Taishan NPP with a height of 48.95 m, the
area is divided into 16 layers, where the thickest one is
1.8 m. The outer containment has an inner diameter of 53
m and an outer diameter of 55.6 m, with the C45 massive
concrete of 1237.6 m³ in all (Liu 2014). Through the
numerical simulation, Liu (2014) found that the center of
containment wall had the highest temperature over 70 °C
while the surfaces were much cooler. The temperature of
inside surface was the same as the outside surface, and
they were symmetrical by the center. According to the
existing construction layering arrangement, the thickness
of construction layer rather than the height could have a
significant influence on the temperature field. For a 1.8
m thick layer, the center temperature of containment wall
could reach 75 °C, which was 5 °C higher than that of the
1.3 m one. The maximum vertical temperature gradient
would reach 71.8 °C/m, and the maximum radial tem-
perature gradient would reach 50.3 °C/m. These posi-
tions with the maximum temperature gradient in two
directions should be the key point during the construction.
It is advised that maintaining time should be increased to
10 days since increasing the maintaining time is one of
the effective ways to reduce the temperature difference.
The worse the insulation is, the much bigger the differ-
J. Xiao, W. Xie and Q. Xie / Journal of Advanced Concrete Technology Vol. 14, 335-343, 2016
ence between the inside and outside surface before
hardening will be. Because the thermal conductivity of
concrete suddenly rises after hardening, the surface
temperature reduces more rapidly with a higher tem-
perature before hardening. The better the insulation ma-
terial is, the much bigger the difference between the
inside and outside surface after hardening will be (Liu
2014; Qian et al. 2014). Therefore, adopting proper in-
sulation material can help to prevent the temperature
difference from exceeding the limit in the China stan-
dards of GB 50496-2009.
As the temperature rises after pouring, it is necessary
to cure the concrete by wet maintenance methods to
guarantee the hydration reaction of binding material
inside concrete. This can reduce the peak temperature
inside the concrete and avoid the loss of strength in the
later period. When concrete begins to cool down, the
structure shrinks and the compression state of concrete
turns into tension state. Penetrating cracks may generate
due to the restraints, which can break the integrality and
durability of structures (Ariyawardena et al. 1997).
Therefore, heat and humidity preservation measures
should be taken to avoid the occurrence of penetrating
cracks during the cooling stage.
3. Thermal safety of containment
The containment is the retaining structure of a nuclear
reactor, which is the third safety barrier in addition to fuel
rod cladding and pressure vessel. According to the
composition of materials, the containment structure
system of nuclear reactor can be classified as steel con-
tainment, reinforced concrete containment and
prestressed concrete containment. The function of con-
tainment is to restrict and eliminate the release of ra-
dioactive fission products to ensure the safety of NPP and
the public when the loss of coolant accident (LOCA)
caused by the rupture of primary coolant pipe happens
(Zhao et al. 2003; Costaz and Danisch 1997). The tem-
perature effect has great influence on overall mechanical
properties of the containment structure, and calculation
of the containment temperature field is an important
prerequisite to ensure the safety of NPP under steady and
accident situation.
3.1 Thermal response of containment in at-
mospheric environment
In the process of testing the overall performance of
containment structures, the test results of deflection and
strain can be intervened by the temperature effect due to
the change of the atmospheric environment (Lin and Yan
2003). Xie et al. (1993) concluded that the thermal
conduction of the containment was a three-dimensional
problem according to the theory of heat transfer. While
the containment was relatively thin and the environment
was relatively uniform, therefore it could be simplified
into a one-dimensional heat transfer problem. They also
selected several representative parts of containment of
338
Qinshan NPP and established the differential equations.
The temperature field was established according to the
calculated temperature at each time point. By compari-
son, the calculated temperature were consistent with the
measured temperature. For example, both the measured
temperature and the calculated temperature of 62.5 m
elevation on the surface of containment fluctuated in the
range of 20~45 °C.
According to weather records of several years in the
local experiment place and the existing data of air tem-
perature in the containment, Lin and Yan (2003) revealed
the variation, distribution pattern and the magnitude of
the effects of temperature field based on the finite dif-
ference method. When the maximum of temperature
difference reached 20 °C on the surface, the temperature
varied within 1 °C in the middle thickness. Due to the
influence of projection angle, duration of sunlight and
other factors, the surface temperature on the top of con-
tainment was higher than that of the tube wall, and the
variation range was larger (Fig. 1). The calculation re-
sults of temperature field were then introduced to the
structure analysis program, and results showed that the
structure deflection was less than 0.3 ~ 0.5 mm under the
atmospheric environment for less than 10 days. In the
atmospheric environment, the containment temperature
decreases rapidly and distributes nonlinearly along
thickness. The temperature difference of containment
surface has little effect on the temperature and thermal
stress in the inner (Singh and Heller 1980).
With the increasing operation time of NPP, the con-
crete structure and facility are bound to deteriorate.
Therefore, the reactor must undergo in-depth inspection
and test at intervals while the containment is evaluated
by monitoring and inspecting. In order to ensure the
security and safety of NPP, the research on the behavior
of containment after withstanding longtime radiation and
temperature variation need to be conducted by meas-
urements of temperature, strain, and displacement.
3.2 Thermal resistance of containment after
LOCA
LOCA releases high temperature steam and water, in-
ducing an equivalent pressure of fluid with 150 °C acting
50
In the top of containment
At 47.8 m of containment
40
Temperature (°C)
30
20
10
0 24 48 72 96 120 144 168 192 216
Time (h)
Fig. 1 Temperature-time curve at two places.
J. Xiao, W. Xie and Q. Xie / Journal of Advanced Concrete Technology Vol. 14, 335-343, 2016 339

Table 5 Initial parameters of containment.

Value
Parameters
Wu et al. (2010) Sun et al. (2015)
Volume of containment (m3) Approx. 300000 Approx. 90000
Height of containment (m) 76 58.5
Strength grade C50
Elastic modulus (MPa) 3.45×104
Poisson ratio 0.2
Concrete
Thermal expansion coefficient (/ °C) 1×10−5
Thermal conductivity (kJ/(m·h·°C)) 8.28 10.6
Density (kg/m3) 2500

on the containment surface. Then the temperature rises
suddenly in the containment, resulting in a greater tem-
perature difference between inside and outside, which
brings about considerable temperature stress and influ-
ences the distribution of temperature and stress (Li and
Wang 2010; Kanzleiter 1976). Therefore, the design of
containment should take the effect of LOCA into con-
sideration, especially in the type of PWR NPP which is
applied widely in China.
Li and Wang (2010) from the Central Research Insti-
tute of Building and Construction of Metallurgical Cor-
poration of China Group analyzed the temperature field
and stress field of a 1.2 m-thick containment structure
under the temperature effect of the LOCA by ANSYS.
Assuming that the inner surface of containment was
adiabatic and the heat transmits from the inner surface to
the outer surface, the instantaneous temperature of
150 °C was applied to node on the inner wall of the
containment. According to the analysis results, the tem-
perature of each node at different distance from the inner
wall is shown in Fig. 2. It can be found in Fig. 2 that the
temperature only influences a thin part of the tube wall.
Affected by accidental instantaneous temperature, the
high stress is mainly concentrated at the fixed bottom of
the cylinder, at the top of the dome and around the
opening, which does not reach the tensile strength of
concrete. So the effect of accidental instantaneous tem-
perature does not cause destructive damage to concrete
structures.
The distribution of temperature field in the concrete
containment becomes obviously nonlinear after LOCA.
Based on the analysis by Zhao et al. (2003) about the
containment temperature change for PWR, Wu et al.
(2010) and Sun et al. (2015) respectively carried out the
finite element simulation of concrete containment after
LOCA. The parameters are listed in Table 5 and the
results are shown in Fig. 3. Due to the differences of
factors including containment wall thickness, thermal
conductivity of concrete and temperature difference
inside and outside the containment, the extent of tem-
perature increase in outer wall of containment is quite
different when it is 6 h after the occurrence of LOCA,
while the outside temperature of containment with
thickness of 1.1 m does not obviously change. After
obtaining the temperature field in containment, the inner
force distribution and the deformation of containment in
different time can be derived from theoretical analysis or
numerical simulation:
(1) 6 h after the occurrence of LOCA, the bending mo-
ment reaches its maximum value according to the
theory of elasticity, ignoring the confinement effect
on structure bottom. Thus the temperature field at
this time can act as a benchmark for internal force
design. In practical engineering, a reduction coeffi-
cient ranging from 0.3 to 0.5 should be adopted for
the calculation of thermal stress to consider the
concrete cracking and creep effect (Wu et al. 2010).
(2) The temperature effect has a great influence on the

160 160
0m Thickness 1.1m/10min(Wu et al. 2010)
140
Thickness 1.1m/6h (Wu et al. 2010)
120 0.1m 120 Thickness 0.9m/10min(Sun at al. 2015)
Temperature(°C)

Temperature (°C)

0.2m 100 Thickness 0.9m/6h (Sun et al. 2015)

80 0.6m 80
60
40 40
20
0 0
0 1800 3600 5400 7200 9000 10800 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Time(s) Relative thickness
Fig. 2 Distance and temperature curve (Li and Wang Fig. 3 Variation of temperature along relative thickness at
2010). different time after LOCA.
J. Xiao, W. Xie and Q. Xie / Journal of Advanced Concrete Technology Vol. 14, 335-343, 2016
Table 6 Main types of coatings for Qinshan I NPP Project (Gong 2009).
Category Concrete structure
Red polyurethane+
Containment
polyurethane
Radiation control area Epoxy
Security equipment Epoxy, polyurethane
deformation pattern of the overall containment de-
formation. When a LOCA occurs, the maximum
expansion displacement of containment structure
increases nearly 14 mm and the overall expansion
displacement also increases 6 ~ 10 mm under the
combination of internal temperature and pressure
(Sun et al. 2015).
If the pressure vessel fails, the integrity of containment
will be jeopardized. The task of top priority is to cool
down and seal the corium inside the pressure vessel. As
an important material used in core catch, sacrificial con-
crete can melt and be mixed with the corium, which will
bring the benefits of reducing solidus and liquids tem-
perature of corium. Therefore, research on the prepara-
tion, function mechanism of nuclear sacrificial concrete
is of great significance to ensure that containment re-
mains intact. The sacrificial concrete designed by
Southeast University in China has been applied in
Taishan NPP (Chu et al. 2016).
4. High temperature resistant coating for
NPP
Coating is a continuous film formed by spraying all paint
at once to the surface of solid, usually referred to as the
substrate, such as metal, fabric and plastic. It is a plastic
thin layer for protection, thermal insulation, decoration
etc. Coating of NPP is mainly used for steel and concrete
structure, pipes and other components on the nuclear
island and the conventional island. Although the corro-
sion resistance is one of the most basic requirements of
the coating performance, strict process control has
guaranteed the basic requirements of anticorrosion for
the special nuclear power coating.
During the operation period, the temperature of the
concrete containment is about 50 °C, and the radiation is
0.1 Gy/h. However, the temperature is likely to rise to
more than 150 °C under accidental conditions with nu-
merous of radiation particles generating. Therefore, the
coating will experience a continuous aging process due
to the high temperature and radiation, which may jeop-
ardize its original function in the normal operation of
NPP or accidental conditions (Chen 2013).
A great deal of research and development work have
been done by the research institutes and suppliers in
China. The coatings for Qinshan I NPP Project are pro-
vided by the Shanghai Kailin paint factory and the
Changzhou Coating Chemical Research Institute (Ni et
al. 2005). Table 6 gives the main types of coatings for
340
Steel structure Steel equipment
Inorganic zinc silicate + Inorganic zinc silicate +
polyurethane polyurethane
Epoxy primer +
alkyd
polyurethane
Epoxy zinc-rich +
Alkyd, epoxy asphalt etc.
alloprene
Qinshan I NPP Project. Based on the successful applica-
tion of these coatings, corresponding standard of special
coating for NPP is compiled referring to relevant con-
tents in the American Society for Testing and Materials
and Norme Française standards. At present, there are few
researches on the aging and life prediction of the coating
on the NPP in China, while the focus is placed on the
production and application of the coating (Liu 2015).
In the test of the performance of the coating, simulated
Design Basis Accident (DBA) is included as well as
coating irradiation test and coating decontamination test.
DBA simulates the effect of reactor on the coating for
containment under the condition of design basis accident.
The coating sample is placed in a high temperature and
high pressure vessel sprayed with boric acid aqueous
solution with pH=9 ~ 10. The temperature, pressure,
period and other conditions can be determined according
to different standards or specific design parameters. The
typical condition is a 4 ~ 7 d period experiment under
153 °C with a certain pressure. After the test, the sample
coating sample will be considered as not qualified if the
cracking, delamination, bubble or breakage of the coat-
ing occurs (Zhang 2007).
Chi et al. (2006) made a comprehensive check and
evaluation on the aging condition of No.1 reactor factory
in Daya Bay NPP after an operation time of over 10 years.
Through field test, the coating adhesion of steel structure,
walls, floors and ceilings surface conformed to the re-
quirement of the original design of the NPP. The coatings
of the steel panels and concrete blocks had a poor con-
dition with destructive defects caused by scratch, grind,
hit, etc. Those test boards were prepared during the con-
struction of Daya Bay NPP, which were later used as
specimens for LOCA tests. Specimens were examined at
2 hours and 2 weeks after LOCA tests respectively, and
the adhesion test were performed after 2 weeks. Ex-
cluding the effect of preexisting defects, the steel panels
and concrete blocks reached the requirements for ac-
ceptance in Table 7.
The passive containment used in the third-generation
NPP is composed of reinforced concrete structure and
steel containment. The inner and outer surfaces of steel
containment are covered by inorganic zinc coating with
excellent heat transfer performance (Kindred and Wright
2014). Different from the aging characteristics of ordi-
nary epoxy coatings, the inorganic zinc coating will
generate zinc oxide during aging, and the oxidation sub-
stances can continue to fill pores in the coatings. Because
the thermal conductivity of zinc oxide is better than that
J. Xiao, W. Xie and Q. Xie / Journal of Advanced Concrete Technology Vol. 14, 335-343, 2016 341

Table 7 Results after LOCA tests.

Requirements for Acceptance Specimens Changes of coating Adhesion Conclusion
No spalling, peeling or powering;
Allow the coating to blister slightly, Steel
but each bubble is not broken and Except the existing pinhole and 3.1 MPa Excluding the preexisting
panels
the diameter is less than 2 mm; crack expansion, there were no defects, the specimens satis-
The crack length of specimen is not spalling, bubble or new crack in the fied the requirements for ac-
more than 10 mm; Concrete surfaces of specimens. ceptance.
The adhesion is not less than 0.2 1.2 MPa
blocks
MPa.

Table 8 Results of adhesion test (Fan et al. 2015).

* Before accident After accident
Location of Sample Points
Dry film thickness (µm) Adhesion (MPa) Dry film thickness (µm) Adhesion (MPa)
95 12.01 101 10.21
Region A 85 13.24 93 11.2
90 11.13 88 13.13
96 12.35 93 11.34
Region B 90 10.09 90 12.78
100 11.65 98 10.11
90 10.61 87 10.88
Region C 100 9.21 100 8.79
96 9.67 94 10.02
*The sample points are selected in region A, B, C (The specimen is rectangular, A and B are located on the same side of the
specimen, bilateral symmetry; the C area is located on the other side, in the central area of the specimen). 3 sample points are se-
lected from each region.

of coating pores, the heat transmission of inorganic zinc
coating will not be affected adversely by aging (Liu et al.
2015). According to the physical environment that might
form in the LOCA accident, Fan et al. (2015) conducted
a simulation experiment on test panel with inorganic zinc
coating. Dry film thickness and adhesion before and after
the LOCA accident are shown in Table 8. A single
simulation of the LOCA accident does not have a sig-
nificant impact on the dry film thickness and adhesion of
inorganic zinc coating, and the coating still meets the
requirements of serviceable capacity, such as the integ-
rity, heat transfer and radiation resistance.
5. Summary
This paper reviewed the researches and technical meas-
ures for minimizing the risk associated with concrete in
Chinese NPP exposure to high temperatures. The main
conclusions are summarized in the following:
(1) The concrete component should be optimized for the
construction of foundation and main structure of NPP.
While temperature and strain sensors are used for
real-time monitoring of the temperature of each
measuring point, appropriate insulation measures can
be taken to prevent cracking due to excessively high
temperature or excessive temperature gradient in
concrete.
(2) Finite element analysis can be used to predict the
temperature field of the containment vessel under the
conditions of normal temperature and LOCA accident,
which can provide reference for the internal force
analysis and the structural design of the containment.
(3) To complete further analysis (such as the research on
the whole life appraisal of containment), the longtime
measurements of temperature, strain, and displace-
ment are advised to adopt for the safety of contain-
ment.
(4) In China, some institutions already have the ability to
research and produce the special coating for NPP.
They have compiled the norms of NPP coating by
referring to the norms of France and USA to provide
technical support for the safe application of the NPP.
Moreover, the damage of coating is inevitable and
needs alleviation during the operation of generator
units. Therefore it is necessary to compile the norms
of coating maintenance and assessment based on the
situation of the coating of NPP.
Acknowledgements
The authors would like to gratefully acknowledge all the
national and international investigators related to the
references and the research grants from the Chinese
National 973 Plan (2012CB719703).
Reference
Acker, P. and Ulm, F., (2001). “Creep and shrinkage of
concrete: physical origins and practical
measurements.” Nuclear Engineering and Design,
203(2), 143-158.
Ariyawardena, N., Ghali, A. and Elbadry, M., (1997).
“Experimental study on thermal cracking in
reinforced concrete members.” ACI Structural
Journal, 94(4), 432-441.
Chen, Z., (2013). “Study of aging and lifetime for the
J. Xiao, W. Xie and Q. Xie / Journal of Advanced Concrete Technology Vol. 14, 335-343, 2016
epoxy coatings in nuclear power plant.” Shanghai
Jiao Tong University. (in Chinese)
Chi, Z. H., Tang, M. L. and Cui, L., (2006). “Evaluation
of aging of inner coating of containment vessel of
Daya Bay nuclear power plant.” National Nuclear
Material Academic Communication 2006. (in
Chinese)
Chu, H., Jiang, J., Sun, W., and Zhang, M., (2016).
“Thermal behavior of siliceous and ferro-siliceous
sacrificial concrete subjected to elevated tempera-
tures.” Materials & Design, 95, 470-480.
Costaz, J. L. and Danisch, R., (1997). “Discussion on
recent concrete containment designs.” Nuclear
Engineering and Design, 174(2), 189-196.
Fan, H. C., (2015). “Study on the characteristic of
inorganic zinc coating for passive containment.”
Surface Technology, 6, 7-10. (in Chinese)
GB 50496-2009, (2009). “Code for construction of mass
concrete.” Beijing: Standards Press of China. (in
Chinese)
Gong, Y. Z., (2009). “Review on the research and
development of China nuclear power coatings.”
China Paint, 24(6), 5-8. (in Chinese)
Gu, H. M., Cheng, D. Y. and Zhang, X. B., (2010).
“Monitoring of temperature strain during pouring raft
foundation for a nuclear island of a nuclear power
plant in hot and humid environment.” Industrial
Construction, 40(1), 17-20. (in Chinese)
Kanzleiter, T. F., (1976). “Experimental investigations
of pressure and temperature loads on a containment
after a loss-of-coolant accident.” Nuclear
Engineering and Design, 38(1), 159-167.
Kaplan, M. F., (1989). “Concrete radiation shielding:
nuclear physics, concrete properties, design and
construction.” Essex: Longman Scientific &
Technical.
Kindred, T. A. and Wright, R. F., (2014). “Effects of
aging on the thermal conductivity of the AP1000®
containment vessel inorganic zinc coating.” 2014 22nd
International Conference on Nuclear Engineering,
American Society of Mechanical Engineers,
V001T03A028-V001T03A028.
Li, J. W. and Wang, Y. H., (2010). “Calculation and
analysis of temperature and stress fields for nuclear
power plants under temperature effect of LOCA
accident.” Industrial Construction, S1, 300-305. (in
Chinese)
Lin, Z. S. and Yan, J. H., (2003). “The temperature field
of PCCV of nuclear power station under atmosphere
circumstance.” Journal of Building Structures, 36(6),
12-17. (in Chinese)
Ling, D. S., Xu, D. S. and Shen, Y. Y., (2005). “Model
for hydration heat of cement in concrete and its
application.” Journal of Zhejiang University, 39(11),
1695-1698. (in Chinese)
Liu, G. Z., (2015). “Research on immersed thermal
aging of coatings in nuclear power plant.” Shanghai
Jiao Tong University. (in Chinese)
342
Liu, S. H., (2014). “Research on construction
temperature field characteristics and effect of nuclear
island outer containment.” Harbin Institute of
Technology. (in Chinese)
Liu, X. Q., Xu, X. L. and Meng, F. J., (2015).
“Reliability and design analysis of coatings used in
passive containment vessel of nuclear power plant.”
Paint & Coatings Industry, 45(4), 74-78. (in Chinese)
Ni, A. B., Ding, W. Z. and Wang, L. F., (2005).
“Development of heat-resistant coatings for nuclear
facilities and equipment.” Paint & Coatings Industry,
35(11), 7-9. (in Chinese)
Qian, H. L., Qin, F. and Fan, F., (2014). “Analysis on
temperature filed of outer containment of the Taishan
nuclear power plant.” Journal of Building Structures,
35(7), 115-122. (in Chinese)
Schindler, A. K., (2004). “Effect of temperature on
hydration of cementitious materials.” ACI Materials
Journal, 101(1), 72-81.
Schneider, M. and Froggatt, A., (2015). “World nuclear
industry status report 2015.”
Sekimura, N., (2012). “Securing the stability of ageing
NPP considering Fukushima Dai-Ichi lessons
learning: improving inspection system and ageing
management and planning.” 3rd International
Conference on NPP Life Management (PLIM) for
Long Term Operations (LTO), Salt Lake City, U.S.A.
Singh, M. P. and Heller, R. A., (1980). “Random
thermal stress in concrete containments.” Journal of
the Structural Division, 106(7), 1481-1496.
Sun, F., Pan, R. and Chai, G. H., (2015). “Preliminary
study on response law of prestressed concrete
containment after LOCA in certain NPP.” Atomic
Energy Science and Technology, 10, 1815-1820. (in
Chinese)
Wang, G. T., (2011). “Preparation of self-compacting
concrete used in nuclear power plant and its
application in cut-off cone structure.” South China
University of Technology. (in Chinese)
Wang, H. C., Chen, L. S. and Lv, G. Y., (2012).
“Preparation and construction technology of nuclear
power plant high-performance mass concrete.”
Nuclear Power Engineering, 32(z2), 47-50. (in
Chinese)
Wei, J. G., Zhang, Z and Zhang, X. B., (2008).
“Temperature-controlling monitoring and emulation
analysis of mass-concrete of raft foundation of
nuclear power plant.” Industrial Construction, z1,
1033-1035. (in Chinese)
Willam, K., Xi, Y. and Naus, D., (2013). “A review of
the effects of radiation on microstructure and
properties of concrete used in nuclear power plants.”
US NRC, NUREG/CR-7171.
Wu, C., Meng, S. P. and Zhou, Z., (2010). “Analyses on
temperature field distribution and internal force of
concrete containment vessel after LOCA.”
Engineering Mechanics, 27(12), 206-212. (in
Chinese)
J. Xiao, W. Xie and Q. Xie / Journal of Advanced Concrete Technology Vol. 14, 335-343, 2016
Wu, Y., Shi, L. and Liu, J. Z., (2013). “Gas tightness
evaluation of structural concrete applied to nuclear
containment vessels.” Applied Mechanics &
Materials, 405, 19-23. (in Chinese)
Xie, Y. J., Xue, J. L. and Fu, B. L., (1993). “The study
of the structural integrity test on security containment
volumn of Qinshan nuclear power station.” Industrial
Construction, 3, 36-41. (in Chinese)
Xu, C., (2010). “Temperature control measure used in
HRA reactor building of Taishan NPP.” Concrete, 12,
114-117. (in Chinese)
Yang, Z. Y., Wei, J. G. and Ji, L. B., (2009). “Integral
pouring construction technology of raft foundation A
and B layers in Hongyanhe nuclear power station
reactor workshop.” Construction Technology, 38(8),
90-93. (in Chinese)
Ye, Q. Q., (2012). “China’s nuclear power development
after Fukushima nuclear power plant accident.”
343
Chinese Society for Electrical Engineering, 32(11),
1-8. (in Chinese)
Zhang, Y., (2007). “Analysis on application of special
coating for nuclear power plant.” 1st Symposium on
Special Coating and Application, 210-214. (in
Chinese)
Zhao, J., Yu, H. X. and Li, F., (2003). “Calculation and
analyses of containment pressure and temperature for
PWR1000XL.” Nuclear Power Engineering, 24(5),
409-411. (in Chinese)
Zhao, P. and Zhou, C. G., (2012). “Optimization of
construction mix proportion based on raft foundation
concrete of DEF nuclear power plant.” Henan
Technology, 17, 88-90. (in Chinese)
Zhao, S. M., Lin, S. T. and Wang, Y. H., (2003). “The
structural integrity test of containment for Qinshan
nuclear power plant.” Industrial Construction, 33(9),
38-40. (in Chinese)
Journal of Advanced Concrete Technology Vol. 14, 261-270 May 2016 / Copyright © 2016 Japan Concrete Institute 261

Scientific paper

Experimental Study on Hydraulic and Macro-Mechanical Property of

a Mortar under Heating and Cooling Treatment
Zhinan Lin1, Weiya Xu2, Wei Wang3*, Jiuchang Zhang4, Huanlin Wang5 and Rubin Wang6

Received 4 March 2016, accepted 12 May 2016 doi:10.3151/jact.14.261

Abstract
A research program was carried out to investigate the effect of heating and cooling treatment upon hydraulic property and
macro-mechanical property of cement based material. Those treatments were performed on cylindrical mortar specimens
(water to cement ratio of 0.5), subjected to 5, 10, 15 heating-cooling cycle times and three levels of maximum heating
temperatures of 105°C, 145°C and 200°C. The hydraulic property tests were performed on a gas permeability and po-
rosity test system. The initial permeability and porosity value of the material were measuring before the heating and
cooling treatment under slightly confine (3MPa) and show the material has a good homogeneity. The damage of mortar
due to the effect of heating and cooling treatment were evaluated the variation of permeability, porosity and
macro-mechanical parameters by compare before and after heating and cooling treatment. Significant changes in the
macro-mechanical property and hydraulic property before and after heating and cooling treatment in mortar were ob-
served. The compressive strength and elastic modulus gradually decrease with the increase of maximum heating tem-
perature and the number of heating cycles, while the porosity and permeability increases gradually.

1. Introduction used to isolate the waste and fill the concrete containers
in the nuclear wastes repository. As the last protection
The safety of radioactive waste storage has become a barrier, its role is to isolate the waste from water or hy-
major concern in the development of nuclear energy drologic isolation (Plecas 2010), so it plays a significant
since it had been generated in the early 1950s. Mortar has role on the protection system (Plecas et al. 2015). But on
been widely used for radioactive waste storage because account of the porousness of mortar, its permeability and
of its low permeability and could be further used for the porosity should be minimized in the long-term operation
formation of disposal engineering barriers (McCulloch et of the nuclear waste storage. So it is important to predict
al. 1985; Wilding 1992; Glasser et al. 1994). It is usually the damage of mortar in the operation of long-term waste
storage. During its long service life, mortar damage is
developed both under the change of internal parameters
1 (e.g. the chemical decompositions as a result of hydration
Ph.D. candidate, Research Institute of Geotechnical
processes (Mainguy et al. 2000; Vydra et al. 2001), and
Engineering, Key Laboratory of Ministry of Education
also under the change of external conditions (e.g. tem-
for Geomechanics and Embankment Engineering, Hohai
perature (Lim and Mondal 2014; Vodák et al. 2004),
University, Nanjing, China.
2 moisture (Saemann et al. 2000), leading to the significant
Professor, Research Institute of Geotechnical
degradation of material properties. Those processes can
Engineering, Key Laboratory of Ministry of Education
seriously affect durability of the material (Chen et al.
for Geomechanics and Embankment Engineering, Hohai
2010) so as to reduce the safety of radioactive wastes
University, Nanjing, China.
3 storage.
Assistant Professor, Research Institute of Geotechnical
It is generally acknowledged that high temperature can
Engineering, Key Laboratory of Ministry of Education
increase mortar’s porosity due to the generation of mi-
for Geomechanics and Embankment Engineering, Hohai
cro-cracks in the structure (Chen et al. 2009; Gao et al.
University, Nanjing, China.
* 2002; Vydra et al. 2001). The effect of temperature on
Corresponding author, E-mail: wwang@hhu.edu.cn
4 strength properties of concrete has been extensively
Postdoc, Research Institute of Geotechnical
investigated (Bastami et al. 2011; Chan et al. 2000; Cül-
Engineering, Key Laboratory of Ministry of Education
fik and Özturan 2002; Ergün et al. 2013; Fall and Samb
for Geomechanics and Embankment Engineering, Hohai
2009; Gardner et al. 2005; Lin et al. 2015), and it was
University, Nanjing, China.
5 concluded that the reduction in strength of a concrete is
Professor, Key Laboratory of Coastal Disaster and
disastrous if heated up to 800°C (Chan et al. 2000). As
Defence, Ministry of Education, Hohai University,
reported in many publications(Caré 2008; Castellote et al.
Nanjing, China.
6 2004; Farage et al. 2003; Lim and Mondal 2014; Rostásy
Assistant Professor, Research Institute of Geotechnical
et al. 1980; Zhang et al. 2013), the change of mi-
Engineering, Key Laboratory of Ministry of Education
cro-structure and pore structures of mortar subjected to
for Geomechanics and Embankment Engineering, Hohai
high temperatures was studied, revealing that the port-
University, Nanjing, China.
Z. Lin, W. Xu, W. Wang, J. Zhang, H. Wang and R. Wang / Journal of Advanced Concrete Technology Vol. 14, 261-270, 2016
Table 1 Chemical compositions and physical properties of used cement.
PC Oxide compositions (wt. %)
SiO2 Al2O3 CaO SO3 MgO Na2O
21.58 5.16 63.05 2.57 1.57 0.27
landite in mortar was decomposed abruptly and trans-
formed into lime when it was heated up to 620°C before
cooling (Castellote et al. 2004). The transport properties
of mortar subjected to temperature loading have been
investigated (Caré 2008; Chen et al. 2010; Chen et al.
2010; Pulkrabek and Ibele 1987; Vydra et al. 2001). The
correlation between heating-induced crack porosity (and
crack aspect ratio) and permeability has been investi-
gated (Chen et al. 2010). The variation of porosity and
permeability for mortar under heat cycling were studied
(Chen et al. 2010; Chen et al. 2009). Fire tests were
carried out to study the mechanical behavior of rein-
forced concrete frame at elevated temperatures (Raouf-
fard and Nishiyama 2015). Damage evaluation in mortar
with history of high temperature was investigated by
Henry (Henry et al. 2016). Furthermore, the durability of
an Ultra High Performance Fiber Reinforced Concrete
after accelerated aging was studied through gas perme-
ability test (Wang et al. 2014c).
Recently, many researches focused on the strength
properties, the micro-structure and the permeability of
mortar subjected to high temperature effects and heat
cycling (Bastami et al. 2011; Cülfik and Özturan 2002),
to a certain extent, they can provide useful information
related to the macro-mechanical property and hydraulic
property of mortar subjected to heating treatment.
However, few of them studied the relationship between
the variation of macro-mechanical, permeability and
porosity of mortar subjected to heat cycling. Besides,
those studies often aimed at a few heating-cooling cycles
or only once, and they did not study the influence of
multiple heating-cooling cycles on the mortar during the
long-term operation of engineering. In this work, a spe-
cial mortar was selected as the experiment object, dif-
ferent multiple heating-cooling cycles were adopted in
order to simulate the real heating treatment during the
long-term operation of engineering. And during the
heating process, different maximum temperatures were
selected in order to simulate the different heating treat-
ments. During the tests, the variations of permeability,
porosity and macro-mechanical of mortar were measured.
The purpose of the test was to study the damage of
mortar subjected to different heating treatments and
heating-cooling cycles.
2. Testing equipment and methods
2.1 Specimen preparation and curing
The experiments were undertaken on a normalized
mortar with a water/cement (w/c) ratio of 0.5. The ce-
ment was made of China commercial PⅡ52.5 Portland
cement. According to the Chinese National Standard
(GB 175-2007), this cement comprised more than 95%
262
Loss (%) BS (m2/kg)
K2O Fe2O3
3.10 316
0.31 3.39
Table 2 Mass of the mortar constituents.
Constituent Mass (g)
Cement 720
Standard sand 1220
Water 310
Bubble reducing agent 5
clinker and gypsum, and 0-5% limestone. Its 3-day and
28-day compressive strengths are 31.6 MPa and 65.6
MPa, respectively. Its 3-day and 28-day tensile strengths
are 6.4 MPa and 9.8 MPa, respectively. The chemical
compositions and physical properties are presented in
Table 1.
In order to simulate the cement-based material cov-
ering the nuclear waste, the material whose compositions
are similar to those existing in real nuclear waste re-
pository was adopted. According to Ilija Plecas (Plecas et
al. 2015), mortar samples were prepared with the 0.5
W/C ratio and the 1/1.694 cement-to-sand ratio, which
could minimize the permeability and leaching rate and
maximize the mechanical strength. The standard sand
(mud mass content<0.2%, mass content of SiO2 is
96.0%) was used with 0-1 mm in diameter. The mass of
the various constituents are presented in Table 2.
Firstly, the mixing procedure consisted in the mixing
sand and cement during 30 s. Then water and bubble
reducing agent were added and the mixing process was
maintained for further 3 minutes. Subsequently, the
mixtures were put into a cylindrical organic glass mold
(50 mm high, 50 mm in diameter) until no air bubbles
were produced. After about 3 h, they were sealed into a
plastic bag with some water at the bottom. The sealed
mold remained in the laboratory at the temperature of 24
± 1°C. After 24-h curing, the mortars were demount and
stored in a freshwater tank with water temperature at 20±
5°C. After 28 days curing, the saturated mortars were
taken out from the water tank.
2.2 Heating and cooling cycle procedure
The experiment was designed to simulate the in-situ
conditions of the long-term storage of the radioactive
wastes, therefore multiple heating-cooling cycles and
different maximum temperatures during the heating
process could be used. Based on the methods by Lion et
al. (2005), the reference temperature was set at 60ºC,
since it could cause evaporation of free porous water
without any notable effect on the CSH gel. In this study,
each reaction cycle included a heating and cooling
process. The heating was performed between 105°C and
200°C. During this process, the specimens were divided
into three batches for different tests, which were heated
up to 105, 145 and 200°C, respectively. The initial tem-
perature was 24ºC (room temperature) and the incre-
Z. Lin, W. Xu, W. Wang, J. Zhang, H. Wang and R. Wang / Journal of Advanced Concrete Technology Vol. 14, 261-270, 2016 263

Table 3 Experimental design of heating and cooling curing.

Maximum Number of
Sample Height (mm) Diameter (mm) Mass (g) Density (g/cm3) Type of tests
temperature (°C) heating cycle
S-1 100.55 50.02 401 2.030
S-2 100.41 50.08 406.23 2.054 60 0
S-3 100.25 50.11 410.23 2.075
S-11 100.02 50.05 400.55 2.036
S-12 100.39 50.01 407.33 2.066 105 5
S-13 100.33 50.01 402.9 2.044
S-21 100.35 49.91 400.9 2.042
S-22 100.65 50.06 405.46 2.047 105 10
S-23 98.74 49.98 406.1 2.096 Macro-mechanical
S-31 100.69 49.92 401.73 2.039 properties tests
S-32 100.17 50.09 398.72 2.020 105 15
S-33 100.43 49.95 398.79 2.026
S-41 100.22 49.95 401.21 2.043
S-42 100.73 49.95 404.36 2.049 145 5
S-43 100.54 49.91 397.03 2.019
S-51 100.34 49.95 401.73 2.043
S-52 99.44 50.16 402.09 2.046 200 5
S-53 100.27 49.91 405.32 2.066
q-1 50.07 50.12 203.71 2.062 105 5
q-2 50.07 50.43 203.47 2.035 145 5
Gas porosity and
q-3 50.01 50.07 201.24 2.044 200 5
permeability tests
q-8 50.07 50.31 203.47 2.044 105 10
q-9 50.01 50.51 201.24 2.008 105 15

mental rate was set at 5ºC/minute by a high temperature
oven. When the peak temperature was attained, samples
were maintained in the oven for more 12 h until constant
weights were accessed (Lion et al. 2005). During the
cooling process, samples were transferred into a fresh-
water tank with a temperature of 20± 5°C for 3h. In order
to simulated the real heating treatment during the
long-term operation of engineering, the multiple heat-
ing-cooling cycles was adopted. The specimens were
processed with 5, 10 and 15 times of heating and cooling
cycles, respectively. The experiment design of heating
treatment for each batch of specimens are presented in
Table 3.
2.3 Testing equipment of macro-mechanical
parameters
The macro-mechanical parameters tests were performed
on the automatic triaxial servo-controlled instrument.
The lateral pressure is controlled by a high-precision
pump and automatically regulated by the computer. The
measurement system consists of a pair of LVDTs (linear
variable displacement transducer) and a chain transducer.
Axial displacement is measured by the LVDTs installed
inside the triaxial chamber. During the experiments, the
deformation data was recorded by the corresponding
sensors in an acquisition system.
2.4 Measurements of permeability and porosity
The permeability and porosity subjected to the heat-
ing-cooling cycles were evaluated by using a gas per-
meability and Porosity Test System which was jointly
developed by our laboratory and Lille University of
Science and Technology. The diagram of the test system
is shown in Fig. 1 (Wang et al. 2014a). The confining
pressure Pc is generated and controlled by an oil pump
and its maximum capacity is up to 60MPa. During the
gas permeability test, the interstitial pressure is con-
trolled and monitored by two manometers, which lo-
cated at the ends of both side of samples. In this test, the
argon with 99.9% purity was used as permeable media
instead of water. The test system can be also used to
perform the porosity test through adding a pipe and a
precision pressure gauge with the pressure range of 0～
2 MPa and the resolution of 0.0001 MPa (Wang et al.
2014a).
The porosity φ of a porous medium describing the
fraction of void space in the material, it is defined by the
following formula:
φ = ⎛⎜ Vn V ⎞⎟ ×100% (1)
⎝ ⎠
where Vn is the effective volume of the pores (m3), and
V is the total volume of sample (m3).
In this study, the porosity of mortar is measured
through the gas expansion method. The procedures is
described as follows. A sample with a known bulk vol-
ume is enclosed in a container of known volume, as
shown in Fig. 1. It is connected to another container with
a known volume which is evacuated (i.e., near vacuum
pressure). When a valve connecting the two containers is
opened, gas passes from the first container to the second
until a uniform pressure distribution is attained. By using
the ideal gas law, the volume of the pores can be calcu-
lated,
Z. Lin, W. Xu, W. Wang, J. Zhang, H. Wang and R. Wang / Journal of Advanced Concrete Technology Vol. 14, 261-270, 2016
Fig.1 Schematic diagram of test system for permeability and porosity measurements.
P − PB
Vn = A VA − VB − VC
PB
where PA is the initial pressure in volume VC, PB is the
final pressure in the entire system, VA, VB, VC are the
known volumes shown in Fig. 1.
The test samples were performed initially under the
dry condition (every sample is dried at 60°C until the
mass stabilization), and all tests were conducted at the
room temperature of 24 ± 1°C. The permeability test
were performed by using the steady state flow test
method. During the test, a continuous gas flow through
the specimen was monitored until a steady conditions,
which is shown schematically in Fig. 1. The inert gas
(argon) was injected through the cylindrical specimen.
The sample was subjected to a constant gas pressure PA
through a buffer reservoir to the bottom side, and at-
mospheric pressure P0 was at the upper side (Wang et al.
2014b; Chen et al. 2010).
According to Darcy’s law, the permeability of a mortar
can be expressed as:
2Qs μ LP0
k=
A ( P 2 − P02 )
P = ( P0 + PA ) / 2
where A is the cross section of sample (m2), L is the
length of sample (m), P0 is the atmospheric pressure(Pa),
P is the constant mean injection pressure(Pa), Qs is the
mean entry gas flow (m3/s), and μ is the gas viscosity
(Pa·s).
3. Experimental results and analysis
As mentioned above, for the purpose of evaluation of
damaged mortar subjected to heating and cooling cycles,
a series of uniaxial tests, permeability and porosity tests
264
have been conducted on the samples at different maxi-
mum heating temperatures (T=105°C, 145°C and 200°C)
(2)
and different heating and cooling cycles (N=5, 10 and 15).
All the uniaxial tests were carried out with a displace-
ment rate of 0.075 mm/min. In the permeability and
porosity tests, the confining pressure is 3MPa.
The initial porosity of the tested mortar heated to 60°C
varies from 16.54% to 17.48% and the initial perme-
ability varies between 1.32·10-17 and 3.67·10-17. This
variation is relatively small and shows a good reproduc-
tively of the material.
3.1 Permeability and porosity
The targeted mortar was composed of cement hydrates,
un-hydrated cement particles and sand. Permeability and
porosity are two important evaluation parameters of
microstructure. The Permeability and Porosity Test Sys-
tem was used to estimate the damage effects when mor-
tar was subjected to heating and cooling cycles. The
results of porosity tests were shown in Fig. 2.
Regardless of different heating treatment, the porosity
of samples always increased after the treatment. How-
ever, CSH gel inside mortar could suffer desiccation
when the heating temperature exceeded 105ºC (Harma-
(3) thy 1970). The study by Vydra et al. (2010) also showed
the maximum loss of free water in endotherm at 90ºC for
TEMELIN and at 95ºC for PENLY. As Lion et al. (2005)
explained, aggregates expanded upon heating, whereas
cement paste did not expand until 150ºC and it shrank
again when temperature continued to rise. As a conse-
quence, the strain incompatibilities between them could
cause micro-cracks and resulted in high variability of
porosity and permeability. What’s more, water cooling
method was adopted in this study and it may result in a
sharp thermal shock, which could cause a severer
strength deterioration when compare to furnace cooling
and nature cooling (Luo et al. 2000). During the heating
process, the outer layers of the tested specimen expands
Z. Lin, W. Xu, W. Wang, J. Zhang, H. Wang and R. Wang / Journal of Advanced Concrete Technology Vol. 14, 261-270, 2016 265

a b
Fig. 2 Variation of porosity of sample under different heating and cooling cycle curing conditions (confining pressure=3
MPa).

more than inner layers. At the beginning of the cooling 8

q-1(before heating cycle) q-1(after heating cycle)
process, suddenly cooling would lead to an internal un- 7
q-2(before heating cycle) q-2(after heating cycle)
q-3(before heating cycle) q-3(after heating cycle)
balance stresses developed inside the mortar since the
outer layers attempt to temperature balance with inner. 6
Permeability(1E-17m2)

So the cracks inside the mortar occurred due to sudden

cooling. Besides, during each heating-cooling cycle
treatment, some chemical reactions and physical reac-
tions occurred inside the mortar, which will increase
internal damage. As the number of heating-cooling
treatment increased, the damage inside mortar accumu-
lated, so the hydraulic property and macro-mechanical
property of mortar changed significantly. Take q-1 as an
example, the porosity of mortar increased by 11.56%
under 5 heating and cooling cycles curing (T=105ºC).
This phenomenon is related to combination of strain
incompatibilities, physical-chemical phenomena and
cracks induced by thermal shock which results in damage
development within the mortar, such as crack extension,
mineral dehydration, and containing moisture or free
water evaporation.
Figure 2 (a) shows the variation of porosity under
three maximum heating temperatures (T=105°C, 145°C
and 200°C) after 5 heating and cooling cycles curing.
With the increasing max temperatures in five heating
cycles, the porosity gradually increases by 11.56%,
18.72% and 27.01%. It means the damage of mortar
increased with the maximum heating temperature, in
other words, the micro-crack gradually increased due to
the increased heating and cooling cycles. So the bigger
the maximum heating temperature is, the higher the
porosity growth is, which means that the damage of the
mortar becomes much larger. When the maximum heat-
ing temperature was 105ºC, the mortar would lose free
water and some adsorbed water then forming the gap
inside the mortar, and with heating and cooling cycles
increased, due to the different thermal expansion coeffi-
cients between aggregate and cement paste, the damage
increased with the increase of porosity. When the
maximum heating temperature was 145ºC, the
non-bounded water was totally lost, then the porosity
5
4
3
2
1
0
0.4 0.6 0.8 1.0 1.2 1.4 1.6
Seepage pressure Pc(MPa)
Fig. 3 Variation of permeability under different maximum
heating temperatures (confining pressure=3MPa).
continued to increase. When the maximum heating
temperature was 200ºC, CSH gel desiccation were ob-
served due to the loss of bounded water, then the porosity
continued to increase.
Figure 2 (b) shows the variation of porosity under
different heating and cooling cycles (T=105ºC). With the
increasing number of heating-cooling cycle, the porosity
gradually increases by 11.56%, 16.67% and 19.90%.
With the same maximum heating temperature, the heat-
ing-cooling cycle treatment caused new micro-cracks in
mortar which appeared at the beginning of cooling for
each cycle. So with the increasing number of heating and
cooling cycles, the micro-cracks inside the mortar ac-
cumulated gradually. It was concluded that with the
number of heating and cooling cycles for the same
maximum heating temperature, the cracks inside the
mortar accumulated continuously, so the damage of
mortar were developed and the porosity of mortar in-
creased.
Figure 3 shows the variation of sample’s permeability
before and after 5 heating-cooling cycle treatment under
different max temperatures. Take the sample q-2(before
Z. Lin, W. Xu, W. Wang, J. Zhang, H. Wang and R. Wang / Journal of Advanced Concrete Technology Vol. 14, 261-270, 2016 266

heating) as an example, it shows that with the increase in
gas pressure, the permeability of mortar decreases. This
may be explained by the Klinkenberg effect which de-
scribes the collision phenomenon between gas molecules
and pore walls. The increasing pressure leads to smaller
effective permeability, showing the good mechanical
state of the samples: damaged specimens tend to give
increasing pressure (Farage et al. 2003).
It also shows that the permeability of sample is higher
after the heating-cooling cycle treatment than before.
The permeability increase of mortar indicates a modifi-
cation of the connected porous network happened. Under
the same fluid pressure of 0.5MPa, the permeability
increases by 61.4%, 136.03% and 149.80% for the sam-
ples q-1, q-2, and q-3, respectively. It means that with the
maximum temperature of heating process (N=5) in-
creased, the permeability of mortar increases continu-
ously. As can be seen from Figs. 2 (a) and 3, the poros-
ity and permeability of mortar consistently increased.
Compare with Fig. 2, the permeability change is much
larger than the porosity change in the same samples. In
means that permeability is more sensitive to mi-
cro-cracks aperture and connection than porosity.
Figure 4 shows the variation of porosity under dif-
8
q-1(before heating cycle)
q-8(before heating cycle)
7 q-9(before heating cycle)
6 stage, and reduces slowly with the increase of pressure
Permeability(1E-17m2)
ferent heating-cooling cycles treatment (T=105ºC). It
shows that the permeability gradually increases with the
number of temperature heating-cooling cycle increased.
For example, under the same fluid pressure of 0.5MPa,
for the 5, 10, 15 times of heating-cooling cycle treatment,
the permeability increases by 61.40%, 258.20% and
391.34%, respectively.
Figure 5 shows the variation of permeability during
the loading and unloading process for the samples q-2
and q-3 under two different fluid pressures (0.5MPa and
1MPa). Regardless of different fluid pressures, during
the loading and unloading process of confining pressure,
the permeability was bigger in the loading stage than that
in the unloading stage for the same confining pressure.
Take the sample of q-2 for example, under the confining
pressure of 15MPa, the permeability was 5.2·10-17 in the
loading stage while it was 4.35·10-17 in the unloading
stage. It means that when the confining pressure is re-
leased, the deformation of its pores cannot be completely
restored, and the mortar sample underwent the plastic
deformation during the loading and unloading process.
This phenomenon was owing to the stress equilibrium
state of mortar granule structure was destructed when the
confining pressure increase after the heating and cooling
cycles curing, then the granule structure of mortar hap-
q-1(after heating cycle)
pen to irreversible damage.
q-8(after heating cycle) It is shown from those figures that the permeability of
q-9(after heating cycle)
mortar decreases sharply at the beginning of the loading

confining. It is noted that with the increasing of the con-

5
fining pressure, the micro-defects would be significantly
4 decreased when the confining pressure is small, and the
micro-defects would be difficult to be further com-
3
pressed with the increasing of confining pressure when
2 the pressure much higher.
Figure 6 illustrates the porosity variations of the sam-
1
ples q-2/q-3 during the loading-unloading process after
0 the heating cycle. Compared with Fig. 5, Fig. 6 shows
0.4 0.6 0.8 1.0 1.2 1.4 1.6 that the porosity and permeability of mortar have the
Seepage pressure Pc(MPa)
Fig. 4 Variation of permeability under different heating-
same variation tendency in the loading and unloading
cooling cycles treatments (confining pressure=3MPa).
process.

7 7

q-2 q-3 6 q-2 q-3

Fluid pressure 0.5MPa
Permeability(1E-17m2)

Permeability(1E-17m2)

6 Fluid pressure 1.0 MPa

5
lo
ad
ngi

lo
pr

ad

5
o

4
in
ce

g
ss

pr
oc
es
s

3
4

unloading process 2
unloading process

0 5 10 15 20 25 30 0 5 10 15 20 25 30
Confining pressure(MPa) Confining pressure(MPa)
a b
Fig. 5 Variation of permeability during loading and unloading under different fluid pressures.
Z. Lin, W. Xu, W. Wang, J. Zhang, H. Wang and R. Wang / Journal of Advanced Concrete Technology Vol. 14, 261-270, 2016 267

Table 4 Compressive strengths of mortar samples under different heating and cooling cycle times.
Number of heating cycles Compressive strength (MPa) Mean compressive strength (MPa)
50.25
Initial (60°C) 55.49 50.37
45.38
43.89
0 times 40.5 41.80
41.02
36.56
10 times 34.51 34.94
33.75
33.19
15 times 33.94 33.5
33.38

Table 5 Compressive strengths of mortar samples under 105°C, 145°C and 200°C maximum heating temperatures.
Maximum heating temperature Compressive strength (MPa) Mean compressive strength (MPa)
50.25
Initial (60°C) 55.49 50.37
45.38
43.89
105°C 40.5 41.80
41.02
29.25
145°C 31.13 29.81
29.06
25.88
200°C 20.25 24.42
27.13

3.2 Macro-mechanical properties elastic modulus) happened due to the accumulated

This section is devoted to the evaluation of mechanical
response of the mortar when it is subjected to the heating
and cooling cycles. Table 4 shows the uniaxial com-
pressive stress of mortar under 0, 5, 10 and 15 times of
heating and cooling cycles at the maximum treatment
temperature of 105°C. It can be concluded that the mortar
exhibits a heating and cooling cycle effect sensitive be-
havior. With the increasing number of heating-cooling
cycles, the cracks gradually increased due to stain in-
compatibilities. The developed damage could decrease
the macro-mechanical properties, such as compressive
strength and elastic modulus. For 5, 10 and 15 times of
heating-cooling cycle treatment, the average compres-
sive strength respectively decreases by 17.01%, 31.01%
and 33.12%. Figure 7 shows the typical stress-axial
damage of internal structure. In contrast to the reference
value, the elastic modulus and compressive strength for
15 cycle times decrease by 60.41 and 37.94%, respectively.
Table 5 and Figs. 9-10 show that variation of
macro-mechanical properties of mortar for the maximum
heating temperatures of 105°C, 145°C and 200°C under
5 heating and cooling cycles. The mortar exhibits a
temperature-dependent sensitive behavior. With the in-
crease of max temperature, the compressive strength
22
q-2 q-3
21 Fluid pressure 0.5MPa
lo
ad

strain curve of the sample under 0, 5, 10, 15 (only four 20

in
Porosity(%)

tests are presented for the sake of clarity) respectively.

pr
oc
es

All stress-strain curves are basically similar, and it can be

s

19
divided into five stages: compaction stage, elastic stage,
elastic-plasticity stage and failure stage. However, it can
un
lo

18
ad

be concluded that when the number of heating- cooling

in
g

cycle is larger, the peak stain is larger and compressive

pr
oc

strength is lower. Fig. 8 illustrates the variation of

e
ss

17
macro-mechanical properties of typical sample under
different heating-cooling cycle treatments (T=105°C). It 0 105 15 20 25 30
is shown that with the increasing number of heat- Confining pressure (MPa)
ing-cooling cycles (T=105°C), more severe deterioration Fig. 6 Variation of porosity during loading and unloading
in macro-mechanical parameters (peak strength and process under fluid pressure of 0.5MPa.
Z. Lin, W. Xu, W. Wang, J. Zhang, H. Wang and R. Wang / Journal of Advanced Concrete Technology Vol. 14, 261-270, 2016 268

60 60
S-1 (initial) S-12(105°, 5times) S-1 (initial) S-12(105°, 5times)
S-22(105°， 10times) S-32 (105°， 15times) S-22(105°， 10times) S-32 (105°， 15times)
50 50

40 40

Stress (MPa)
Stress (MPa)

30 30

20
20

10
10

0
0 0 3 6 9 12 15 18
0 3 6 9 12 15 18
Strain(‰)
Strain(‰)
Fig. 9 Typical stress-axial strain curves of mortar sam-
Fig.7 Typical stress-axial strain curves of mortar under
ples under 105°C, 145°C and 200°C maximum heating
different heating-cooling cycle treatments.
temperatures.

Fig. 8. Variation of macro-mechanical parameters under different heating-cooling cycle treatments.

Fig. 10.Variations of macro-mechanical properties with the maximum heating temperature.

gradually decreases. The average compressive strength
respectively decreases by 17.01%, 40.81%, and 51.61%
for 105°C, 145°C and 200°C in comparison with the
referenced sample strength. It is noted that under dif-
ferent maximum heating temperatures, different physical
reactions occurred, resulting in different damage degrees,
so that the macro-mechanical properties showed differ-
ent degrees of deterioration. Compared with Fig. 7, Fig.
9 shows that the mortar is more ductile with the maxi-
mum temperature of heating process.
Figure 10 shows the comparisons of compressive
strength and elastic modulus for four different maximum
heating temperatures (60°C, 105°C, 145°C and 200°C).
With the increase of maximum heating temperature, the
compressive strength and the elastic modulus gradually
decrease. In contrast to the referenced value, the elastic
modulus and compressive strength for 200°C decrease
by 66.45% and 50.30%, respectively.
Z. Lin, W. Xu, W. Wang, J. Zhang, H. Wang and R. Wang / Journal of Advanced Concrete Technology Vol. 14, 261-270, 2016
4. Conclusions
In the present study, the measurements of porosity, per-
meability and macro-mechanical property were con-
ducted to evaluate the effects of heating and cooling
cycles upon a mortar. The tests were performed on the
mortar samples subjected to 5, 10, 15 cycling times and
maximum heating temperatures of 105°C, 145°C and
200°C. The results indicate that the degradation of the
mortar is due to the evaporation of free water and the
formation of pore gaps in the process of the heating and
cooling cycles. And on account of the different thermal
expansion coefficients between aggregate with cement
paste, the pore gaps continue to develop.
With the increase of maximum heating temperature
(N=5) and with the increasing number of heating and
cooling cycles (T=105 ºC), the porosity and permeability
of the mortar increases gradually, and the compressive
strength and the elastic modulus gradually decrease
while the peak strain gradually increases. The porosity
and the permeability of the mortar consistently decrease,
and the permeability was more sensitive than the porosity
to heating and cooling cycles curing.
Another conclusion is drawn from the permeability
tests. The permeability of the mortar reduces gradually
with the increase of confining pressure in the loading
stage, and increases with the decrease of confining
pressure in the unloading stage. Moreover, the perme-
ability of mortar is larger in the loading stage than in the
unloading stage under the same confining pressure.
Acknowledgments
The work presented in this paper is financially supported
by the National Natural Science Foundation of China
(No. 11272113) and Fundamental Research Funds for the
Central Universities (No. 2015B35714, No.2016B08014).
References
Bastami, M., Chaboki-Khiabani, A., Baghbadrani, M.
and Kordi, M., (2011). “Performance of high strength
concretes at elevated temperatures.” Scientia Iranica,
18(5), 1028-1036.
Caré, S., (2008). “Effect of temperature on porosity and
on chloride diffusion in cement pastes.” Construction
and Building Materials, 22(7), 1560-1573.
Castellote, M., Alonso, C., Andrade, C., Turrillas, X. and
Campo, J., (2004). “Composition and microstructural
changes of cement pastes upon heating, as studied by
neutron diffraction.” Cement and Concrete Research,
34(9), 1633-1644.
Chan, Y. N., Luo, X. and Sun, W., (2000). “Compressive
strength and pore structure of high-performance
concrete after exposure to high temperature up to
800°C.” Cement and Concrete Research, 30(2),
247-251.
Chen, X. T, Rougelot, T., Davy, C. A., Chen, W., Agostin,
F., Skoczylas, F. and Bourdon, X., (2009). “Experi-
mental evidence of a moisture clog effect in
269
cement-based materials under temperature.” Cement
and Concrete Research, 39(12), 1139-1148.
Chen, X. T, Davy, C. A., Shao, J. F. and Skoczylas, F.,
(2010). “Experimental and micro-mechanical analysis
of the mechanical and transport properties of mortar
containing heat-induced micro-cracks.” Cement and
Concrete Composites, 32(9), 678-685.
Chen, X., Davy, C. A., Skoczylas, F. and Shao, J. F.,
(2009). “Effect of heat-treatment and hydrostatic
loading upon the poro-elastic properties of a mortar.”
Cement and Concrete Research, 39(3), 195-205.
Cülfik, M. S. and Özturan, T., (2002). “Effect of elevated
temperatures on the residual mechanical properties of
high-performance mortar.” Cement and Concrete
Research, 32(5), 809-816.
Ergün, A., Kürklü, G., M. Serhat, B. and Mansour, M. Y.,
(2013). “The effect of cement dosage on mechanical
properties of concrete exposed to high temperatures.”
Fire Safety Journal, 55, 160-167.
Fall, M. and Samb, S. S., (2009). “Effect of high
temperature on strength and microstructural properties
of cemented paste backfill.” Fire Safety Journal, 44(4),
642-651.
Farage, M. C. R., Sercombe, J. and Gallé, C., (2003).
“Rehydration and microstructure of cement paste after
heating at temperatures up to 300 °C.” Cement and
Concrete Research, 33(7), 1047-1056.
Gao, X. F., Lo, Y. T. and Tam, C. M., (2002).
“Investigation of micro-cracks and microstructure of
high performance lightweight aggregate concrete.”
Building and Environment, 37(5), 485-489.
Gardner, D. R., Lark, R. J. and Barr, B., (2005). “Effect
of conditioning temperature on the strength and
permeability of normal- and high-strength concrete.”
Cement and Concrete Research, 35(7), 1400-1406.
Glasser, F. P. and Atkins, M., (1994). “Cements in
radioactive waste disposal”. MRS Bulletin, 19(12),
33-38.
Harmathy, T. Z., (1970). “Thermal properties of concrete
at elevated temperatures.” Journal of Materials, 5(1),
47-74.
Henry, M., Hashimoto, K., Darma, I. S. and Sugiyama, T.,
(2016). “Cracking and chemical composition of
cement paste subjected to heating and water
re-curing.” Journal of Advanced Concrete Technology,
14(4), 134-143
Lim, S. and Mondal, P., (2014). “Micro- and nano-scale
characterization to study the thermal degradation of
cement-based materials.” Materials Characterization,
92, 15-25.
Lin, M., Itoh, M. and Maruyama, I., (2015).
“Mechanism of change in splitting tensile strength of
concrete during heating or drying up to 90°C.” Journal
of Advanced Concrete Technology, 13(2), 94-102.
Lion, M., Skoczylas, F., Lafhaj, Z, and Sersar, M., (2005).
“Experimental study on a mortar. Temperature effects
on porosity and permeability. Residual properties or
direct measurements under temperature.” Cement and
Z. Lin, W. Xu, W. Wang, J. Zhang, H. Wang and R. Wang / Journal of Advanced Concrete Technology Vol. 14, 261-270, 2016
Concrete Research, 35(10), 1937-1942.
Luo, X., Sun, W., Chan, S. Y. N., (2000). “Effect of
heating and cooling regimes on residual strength and
microstructure of normal strength and high-perfor-
mance concrete.” Cement and Concrete Research,
30(3), 379-383.
Mainguy, M., Tognazzi, C., Torrenti, J. and Adenot, F.,
(2000). “Modelling of leaching in pure cement paste
and mortar.” Cement and Concrete Research, 30(1),
83-90.
McCulloch, C. E, Angus M. J., Crawford R. W., Rahman,
A. A. and Glasser, P., (1985). “Cements in radioactive
waste disposal: some mineralogical considerations.”
Mineralogical Magazine, 49, 211-221.
Plecas, I., Dimovic, S. and Arbutina, D., (2015).
“Modeling of physico-chemical characteristics of
mortar-waste mixture in radioactive waste manage-
ment.” Progress in Nuclear Energy, 80, 86-89.
Plecas, I. B., (2010). “Mathematical modelling of
immobilization of radionuclides 137Cs and 60Co in
concrete matrix.” Progress in Nuclear Energy, 52(7),
685-688.
Pulkrabek, W. W. and Ibele, W. E., (1987). “The effect of
temperature on the permeability of a porous material.”
International Journal of Heat and Mass Transfer,
30(6), 1103-1109.
Raouffard, M. M. and M. Nishiyama, (2015). “Fire
resistance of reinforced concrete frames subjected to
sService load: Part 1. Experimental study.” Journal of
Advanced Concrete Technology, 13(12), 554-563.
Rostásy, F. S., Weiβ, R. and Wiedemann, G., (1980).
“Changes of pore structure of cement mortars due to
temperature.” Cement and Concrete Research, 10(2),
270
157-164.
Saemann, J. C. and Washa, G. W., (1957). “Variation of
mortar and concrete properties with temperature.”
Journal of the American Concrete Institute, 29(5),
385-395.
Vodák, F., Trtı́k, K., Kapičková, O., Hošková, Š. and
Demo, P., (2004). “The effect of temperature on
strength – porosity relationship for concrete.”
Construction and Building Materials, 18(7), 529-534.
Vydra, V., Vodák, F., Kapičková, O. and Hošková, Š.,
(2001). “Effect of temperature on porosity of concrete
for nuclear-safety structures.” Cement and Concrete
Research, 31(7), 1023-1026.
Wang, H., Xu, W. and Zuo, J., (2014a). “Compact rock
material gas permeability properties.” Physica B:
Condensed Matter, 449, 10-18.
Wang, H., Xu, W., Shao, J. and Skoczylas, F., (2014b).
“The gas permeability properties of low-permeability
rock in the process of triaxial compression test.”
Materials Letters, 116, 386-388.
Wang, W, Liu, J., Agostini, F., Davy, C. A., Skoczylas, F.,
Corvez, D., (2014c). “Durability of an Ultra High
Performance Fiber Reinforced Concrete (UHPFRC)
under progressive aging.” Cement and Concrete
Research, 55(1), 1-13.
Wilding, C. R., (1992). “The performance of cement
based systems.” Cement and concrete research,
22(2-3), 299-310.
Zhang, Q., Ye, G. and Koenders, E., (2013).
“Investigation of the structure of heated Portland
cement paste by using various techniques.” Const-
ruction and Building Materials, 38(0), 1040-1050.
Journal of Advanced Concrete Technology Vol. 14, 502-510 September 2016 / Copyright © 2016 Japan Concrete Institute 502

Scientific paper

Effect of Mineral Admixture and Fibers on Shrinkage Crack of Sacrificial

Concrete
Zuquan Jin1*, Xia Zhao2*, Tiejun Zhao3, Yuanchao Cao4

Received 1 April 2016, accepted 25 August 2016 doi:10.3151/jact.14.502

Abstract
The ferro-siliceous sacrificial concrete is an important protection material to Nuclear power plant, high density aggregate
and high fluidity can lead to occurrence of shrinkage crack. In this paper, the early-age crack and restrained shrinkage
crack of sixteen types of sacrificial concrete containing different content of cementitious materials and two kinds of fibers
were researched. Results showed that the workability of sacrificial concrete decreased with increasing addition of fiber,
and the slump loss resulted from polypropylene (PP) fiber was much higher than that of basalt fiber. The fibers had less
influence on the compressive strength, while had greater impact on the flexural strength, which was improved by 10-20%.
The reduction of total cementitious materials and replacement ratio increment of mineral admixtures including fly ash and
Ground Granulated Blastfurnace Slag (GGBS) in sacrificial concrete could decrease its early-age crack obviously. The
utilization of 1.0-1.5kg/m3 PP fiber could extend the cracking time and reduce the crack width of sacrificial concrete and
total cracking area. The improving crack resistance capacity of sacrificial concrete resulted by basalt fiber was weaker
than PP fiber. The optimized mix proportion of sacrificial concrete mixture was 450kg/m3 total cementitious materials,
50% replacement ration of mineral admixtures and 1.0kg/m3 PP fiber.

1. Introduction When sacrificial concrete is present in the core catcher,

corium active reducers react with Fe2O3 and SiO2 (Re-
Nuclear power provides about 2% of the electricity action 5-8), and there are no active reducers present in
generation in China, and the proportion will increase to corium and Reactions 1-4 will not begin. A small amount
8-10% by the end of 2030. In general, the nuclear power of hydrogen can form during the interaction of Zr (U)
plant has achieved a high level of safety in operation. with water vapor released from the sacrificial concrete
However, the major risk to the public associated with (Komlev et al. 2015 ). Therefore, the explosion induced
potential radioactive releases from nuclear power plants by molten core can be avoided. Recently, a new class of
induced by substantial degradation and melting of the function materials, i.e., an oxide Sacrificial materials
core, cannot be completely eliminated. A nuclear reactor (SM) and a sacrificial steel or concrete, located in cru-
core melt mainly exists in the metallic and oxide forms. cible-type core catchers, were used in VER-12OO NPPs
The main constituents of core melt include Fe, Cr, Ni, Zr, and the third generation of Nuclear power plant (NPP),
UO2 and ZrO2. Reactor core melting point is approximate European Pressurized Water Reactor (EPR) (Komlevet
ranged from 1350°C to 3000°C. If sacrificial concrete is al. 2015; Fischer 2004).
absent from the core catcher, corium active reducers
Zr + 2H 2 O → ZrO 2 + 2H 2 + 6.3MJ/kg (1)
react with external cooling water flooder on the surface
of nonfully oxidized corium (Reactions 1 and 2), and
U + 2H 2 O → UO 2 + 2H 2 + 2.4MJ/kg (2)
atmospheric oxygen (Reactions 3 and 4). Heat generation
temperature in these Reactions is calculated at 1716°C, Zr + O 2 (atmospheric) → ZrO 2 +11.8MJ/kg (3)
and explosion induced by H2 will occur subsequently.
U + O 2 (atmospheric) → UO 2 + 4.5MJ/kg (4)
1
Professor, College of Civil Engineering, Qingdao Zr + SiO 2 → ZrO 2 + Si +1.6MJ/kg,
University of Technology, Qingdao, China. (5)
(T < 2298°F or T < 1870°C)
*Corresponding author, E-mail: jinzuquan@126.com,
zhxiakk@163.com
2 Zr + 2SiO 2 + 4.7MJ/kg → ZrO 2 + 2SiO(g),
Research assistant, Institute of Oceanology, Chinese (6)
Academy of Science, Qingdao, China. (T < 3398°F or T > 1870°C)
*Corresponding author, E-mail: zhxiakk@163.com
3
Professor, College of Civil Engineering, Qingdao 3Zr + 2Fe2 O3 → 3ZrO 2 + 4Fe + 5.8MJ/kg (7)
University of Technology, Qingdao, China.
4
Research engineer, College of Civil Engineering, 3U + 2Fe 2 O3 → 3UO 2 + 4Fe + 2.5MJ/kg (8)
Qingdao University of Technology, Qingdao, China.
Z. Jin, X. Zhao, T. Zhao and Y. Cao / Journal of Advanced Concrete Technology Vol. 14, 502-510, 2016
Table 1 Chemical composition of cementitious materials.
Constituent (wt%) Loss SiO2 Al2O3 Fe2O3
Cement 0.58 21.07 3.79 3.19
GGBS 0.37 32.87 13.40 1.34
Fly ash 0.73 51.67 33.41 4.33
When molten core penetrates the reactor pressure
vessel and reacts with sacrificial concrete, the concrete in
the floor and sidewalls of the cavity starts to melt. Nor-
mally the erosion rate of the concrete is limited by the
heat transfer rate from the melt to the concrete. During
the first pouring of the melt, spalling of concrete might
take place (Kalifa et al. 2000), and it can cause faster
erosion of the sacrificial concrete than expected and
damage the containment vessel concrete. Therefore,
sacrificial concrete possesses high temperature stability
and excellent melting properties during a nuclear
incident. Molten organic fibers can generate release
passage for water vapor in concrete, and thus reduce
burst crack of concrete (Jin et al. 2010). The basalt fiber
can keep about 90% of the normal temperature strength
after exposure at 600 °C for 2 h（Sim et al. 2005; Fiore et
al. 2015）. The steel fiber can improve the strength and
crack-resistance capacity, but decrease the workability of
concrete (Pikus 2016). Therefore, the PP and basalt fiber
were often used to improve the high temperature resis-
tance capacity of sacrificial concrete.
But as time goes on, the protective properties of these
sacrificial concrete will degrade due to environmental
loads including thermal effect, irradiation, neutralization
and penetration of salinity (Kitsutaka and Tsukagoshi
2014). Safety guide for the monitoring of nuclear power
stations and ageing management for nuclear power plants
has been published by IAEA and French Electricity
Board (Stephan and Salin 2012). As mentioned, Fe2O3
and SiO2 in sacrificial concrete will react with corium
active reducers and reduce the explosion risk. Therefore,
the sacrificial concrete in the EPR reactor is a special
type “FeSi” concrete, which includes both high density
aggregates, hematite (Fe2O3) and siliceous (SiO2). And
the density of sacrificial concrete used in EPR is around
2600 kg/m3 (Sevón et al. 2010). High density aggregate
prone to sinking in fresh mortar and induced dry
shrinkage in the surface layer of sacrificial concrete.
Additionally, the size of sacrificial concrete core catcher
zone is φ6m×0.45m, the workability of fresh concrete
will be great in order to cast in small space (Chu et al.
2016). Consequently, early-age autogenous shrinkage
cracking and dry shrinkage crack will easily generate in
EPR sacrificial concrete due to segregation and high
content of mixing water. When excessive shrinkage
cracks occur in sacrificial concrete, its durability will be
threatened seriously. From an aging management per-
spective, the presence of concrete cracks is an impor-
tance threat because they provide possible avenues of
access for corrosion ions (Naus et al. 1999; Dai et al.
2010; Jin et al. 2012). In containment building, CO2,
water steam and O2 will penetrate into sacrificial con-
503
CaO MgO K2O Na2O TiO2 SO3 P2O5 Cl
61.85 3.05 0.61 0.26 0.24 2.26 0.08 0.039
41.06 5.64 0.54 0.45 0.59 2.68 0.04 0.01
4.87 0.77 1.04 0.39 1.46 0.69 0.34 0.018
crete through shrinkage crack and induce the corrosion of
steel bar and result in seriously durability problem. Ad-
ditionally, these shrinkage cracks in sacrificial concrete
will also promote the thermal conduction and the leakage
of radioactive during the melting of the core. Therefore,
crack control is very important for durability and safety
of sacrificial concrete.
Mineral admixtures including fly ash and GGBS are
commonly used to replace cement in high performance
concrete to improve the durability of concrete under
severe environments (Ouda 2015; Seleem et al. 2010).
Additionally, low pozzolanic reactivity of fly ash and
GGBS decreases the autogenous shrinkage at early age,
and the drying shrinkage of self-compacting concretes
incorporating fly ash, GGBS and metakaolin is lessened
(Mermerdas and Arbili 2015; Guneyisi et al. 2010). The
use of fibers is effective in resisting shear forces in con-
crete structures, and the addition of fibers into conven-
tional concrete increases the ductility and energy dissi-
pation of wall concrete structures subjected to strong
ground motions (Choun and Park 2015). And fibers
increase the resistance to crack formation and propaga-
tion, therefore reduce crack width and length (Afrough-
sabet, and Ozbakkaloglu 2015; Saeid et al. 2012). There
is very scarce experimental data about the shrinkage
crack behavior of this special concrete type. This is one
of the reasons for optimizing crack resistance capacity of
sacrificial concrete by use of mineral admixtures and
different type of fibers.
The main objective of the current research is to inves-
tigate the suitability of some concrete components for
producing EPR sacrificial concrete using mineral ad-
mixture and fibers that could enhances its shrinkage
crack resistance capacity.
2. Experiment and materials
2.1 Materials and specimen preparation
P.I.52.5 Portland cement in accordance with Chinese
standard GB175-2007, with a compressive strength of
59.8 MPa at an age of 28 days, was used in this study.
Class I fly ash (as per Chinese standard GB1596-2005)
and S95 GGBS (Chinese standard GB/T18046-2008)
were employed to replace Portland cement. The chemical
composition of cement, fly ash, and GGBS is shown in
Table 1. The specific gravity of cement, Fly ash and
GGBS is 3.13, 2.30 and 2.20, respectively.
Considering the various phenomena that occur during
core melt localization, it is important to minimize the
highly exothermic corium oxidation by water/air and to
limit additional hydrogen production to ensure hydrogen
Z. Jin, X. Zhao, T. Zhao and Y. Cao / Journal of Advanced Concrete Technology Vol. 14, 502-510, 2016 504

Table2 Chemical composites of hematite.

Constituent (wt%) Fe2O3 SiO2 Al2O3 CaO K2O P2O5 MgO SrO MnO
hematite 71.48 14.3 4.48 1.49 1.97 1.82 0.65 1.33 0.71

Table 3 Grain composition of hematite.

Size (mm) (wt %) >9.5 9.5~4.75 4.75~2.36 2.36~0.63 0.63~0.3 0.3~0.15 <0.15 total
hematite 0 26.6 17.4 36.8 8.9 4.7 5.3 99.7

safety in the containment vessel. The amount of Zr in 2.2 Experimental program

core is about 15000kg in European Pressurized Water
Reactor, the mass fraction of Fe2O3 and SiO2 in sacrifi-
cial concrete will be more than 59.3% in order to oxi-
dized all the Zr element in the melt core in sever acci-
dent. So the quartz with SiO2 content more than 95% was
used for providing the silica element, and the mid-grade
hematite with Fe2O3 content being 70~80% was for
providing the iron element. And the size of aggregates
including quartz and hematite is 0-4mm and 4-8mm. The
chemical composition of hematite was analyzed by XRF
and shown in Table 2, and its grain composition was
shown in Table 3. Therefore, the quartz and hematite
amount in sacrificial concrete was adjusted according to
the target value of 59.3% of Fe2O3 + SiO2 in sacrificial
concrete.
The polypropylene (PP) fiber with elasticity modulus
of 3500MPa and melting point of 160℃, the basalt fiber
with elasticity modulus of 11000MPa and melting point
of 700°C were used in this test, and the morphology of
fibers are shown in Fig. 1. A polycarboxylic acid-type
superplasticizer was used and its dosage was adjusted to
according to the slump of fresh mixture.
Based on considerable trials and mechanical, chemical
composites tests, the mixture proportion of plain sacrifi-
cial concrete was determined. Then the PP fiber and
basalt fiber was added to plain concrete for improving
the crack resistance capacity of sacrificial concrete. A
total of 16 concrete mixtures were designed with total
cementitious materials content of 500, 450 and 400
kg/m3, respectively. Table 4 summarizes the mixture
proportions of sacrificial concrete.
Cubes of 150 x 150 x 150mm3 and prisms of 100 x 100 x
400mm3 in dimension were casted and placed at room
temperature with mould. The mould was removed after
24hours. Then, all specimens were cured in a condition
of 20±3℃ and 95% of relative humidity for a period of
28 days. The cubes were used for the determination of
compressive strength, and the prisms were used for
flexural strength test in accordance with Chinese stan-
dard GB/T 500821-2002.
The early-age crack test suggested by ‘Standard for
test methods of long-term performance and durability of
ordinary concrete’ (Chinese standard GB/T 50082-2009)
was used to evaluate the crack resistance capacity of
sacrificial concrete. As shown in Fig. 2, the size of con-
crete formwork is 800mm×600mm×100mm, and in-
ducing groove of 78mm height is added to accelerate
crack of concrete. Before moulding, the PVC film was
spread on the mould as a sealing layer. Then, the sacri-
ficial concrete specimens were cast in the mould. The
specimens were exposed to the wind with the velocity of
5m/s after 30 min. When specimens were cast and cured
for 24h, the number of cracks was recorded; and the
crack length and width were tested by steel ruler and
numerical reading microscope, respectively.
The Mean cracking area of each crack can be de-
scribed as Eq.(9).
N
1
a=
2N
∑ (W × L )
i =1
i i (9)
And the Crack number can be expressed as Eq.(10).

(a) basalt fiber (b) PP fiber

Fig. 1 Morphology of fibers used in sacrificial concrete.
Z. Jin, X. Zhao, T. Zhao and Y. Cao / Journal of Advanced Concrete Technology Vol. 14, 502-510, 2016 505

Table 4 Mix proportion of sacrificial concretes.

Concrete component/kg·m-3 Fiber
No.
Cement Fly ash GGBS Hematite Quartz Water Type Quantity /kg·m-3
MF0 325 105 70 1255 814 162 / 0.0
MF1 325 105 70 1255 814 167 SF 6.0
MPF1 325 105 70 1255 814 167 PP 1.5
N1 360 54 36 1258 842 167 / 0.0
NF1 360 54 36 1258 842 167 SF 6.0
Y0 225 75 150 1261 944 175 / 0.0
Y4 225 75 150 1261 944 175 PP 1.0
Y5 225 75 150 1261 944 175 PP 1.5
Y6 225 75 150 1261 944 175 SF 3.0
Y8 225 75 150 1261 944 180 SF 9.0
U0 320 32 48 1261 861 167 / 0
UF0 320 32 48 1261 861 167 SF 6.0
UPF0 320 32 48 1261 861 167 PP 1.5
T3 260 47 93 1263 970 170 / 0
T4 260 47 93 1263 970 175 PP 1.0
T5 260 47 93 1263 970 175 PP 1.5
PP is polypropylene (PP) fiber, and SF is basalt fiber.

sions of the ring mould was shown in Fig. 3. Immedi-

N ately after demoulding the external cardboard ring, at the
b=
A age of 1 day, the top surface of the ring specimens was
The Cracking area on unit area can be calculated by
Eq.(11).
c = a×b
Where Wi is the maximum crack width of No. i crack
(mm), Li is the length of No. i crack (mm), N is the
total number of cracks, A is the area of formwork (m2),
a is mean crack area of each crack (mm2), b is the
number of cracks in unit area, c is total crack area on
unite area (mm2/m2).
The ring test was performed to evaluate the cracking
behavior of concretes in restrained shrinkage conditions.
The test was carried out following the suggestion of the
standard ASTM C1581–04, however, some geometry
modifications were made, the scheme and the dimen-
(10)
sealed with silicone rubber. And then the specimens were
allowed to dry within in the chamber maintained at the
temperature of 20°C±3°C and relative humidity of
50±10％. Steel ring strain measurements were monitored
(11) from the casting time, having the subsequent readings
taken every half-an-hour until the concrete ring cracked.
After that, measurements of the cracking widths were
taken every day for at least 4 weeks.
3. RESULTS and DISCUSSION
3.1 Slump and density
Slump and unit weight test results of 16 different con-
crete mixes are shown in Table 5. The Slump value of
concretes varies between 140 and 225 mm. A loss of
workability was observed with the addition of fiber to

1.Lateral side panel; 2. Vertical side panel; 3.Bolt; 4. Reinforcing rib; 5. Inducing groove; 6. Bottom plate
Fig. 2 Diagram of test equipment for early-age crack.
Z. Jin, X. Zhao, T. Zhao and Y. Cao / Journal of Advanced Concrete Technology Vol. 14, 502-510, 2016 506

Table 5 Properties of fresh sacrificial concrete mixtures.

Mix. No Slump(mm) Density (kg/m3)
MF0 225 2736.2
MF1 180 2707.2
MPF1 140 2659.5
N1 185 2686.2
NF1 165 2645.6
Y0 230 2723.1
Y4 200 2692.7
Y5 160 2645.3
Y6 180 2637.3
Y8 160 2612.3
U0 195 2641.7
UF0 220 2602.4
UPF0 160 2592.9
T3 220 2641.1
T4 185 2634.4
T5 160 2613.5

the concrete matrix, in fact as the fiber content increased,
the slump slightly decreased. Due to the density of ba-
salt fiber and PP fiber are 2650kg/m3 and 910kg/m3,
respectively, the specific surface area of basalt fiber is
smaller than that of PP fiber, therefore, concrete slump
loss resulted from PP fiber is much greater than that of
basalt fiber. Test results showed that, the application of
fibers decreases the unit weight of all tested concrete
slightly. And the density of sacrificial concrete is about
2600-2700kg/m3.
3.2 Compressive strength and flexural strength
The part results of the compressive strength tests are
given in Fig. 4. The compressive strength of sacrificial
concrete with total cementitious materials content of
500kg/m3 is about 50MPa, and 37-42MPa for sacrificial
concrete with 400kg/m3cementitious materials. In gen-
eral, the influence of fibers on compressive strength of
sacrificial concrete can be ignored. The flexural strength
of sacrificial concrete was tested and shown in Fig. 5.
The flexural strength values of fiber reinforced sacrifi-
cial concrete increased by 10-20%, comparing to the
respective plain concrete specimens. Additionally, the
increase of the concrete flexural strength by basalt fiber
is stronger than 1.5kg/m3 PP fiber. And the flexural
strength of sacrificial concrete containing 1kg/m3 PP

Fig. 3 Restrained shrinkage ring test equipment.

60 50
(a)
3d 28d (b)
3d 28d
Compressive strength (MPa)

50
Compressive strength (MPa)

40
40
30
30
20
20

10 10

0 0
MF0 MF1 MPF1 U0 UF0 UPF0
3 3
Fig. 4 Compressive strength of fiber reinforced sacrificial concrete (a) 500kg/m cementitious materials, (b) 400kg /m
cementitious materials.
Z. Jin, X. Zhao, T. Zhao and Y. Cao / Journal of Advanced Concrete Technology Vol. 14, 502-510, 2016 507

Table 6 Early-age crack parameters of sacrificial concretes.

Maximum crack length Maximum crack width Mean cracking area Crack number Cracking area on
Mix. No
(mm) (mm) (mm2) (Number/mm2) unit area (mm2/m2)
MF0 600 0.95 97.52 8.33 812.66
MF1 600 0.62 98.57 6.25 616.04
MPF1 600 0.42 42.8 10.42 445.83
N1 600 0.62 98.57 6.25 616.04
NF1 600 0.49 147 2.08 305.76
Y0 600 0.52 147 3.38 496.86
Y4 420 0.14 23.47 6.25 146.69
Y5 0 0 0 0 0
Y6 480 0.55 58.08 8.33 483.81
Y8 500 0.37 74.8 6.25 466.8
U0 600 0.59 108.05 6.25 675.31
UF0 600 0.45 92.25 5.08 468.63
UPF0 410 0.38 34.6 10.42 360.53
T3 600 0.6 100.9 6.25 630.625
T4 600 0.72 216 2.08 449.28
T5 380 0.2 26.65 4.17 111.13

fiber is higher than that containing 1.5kg/m3 PP fiber,
because increased content of fiber would result in de-
creased dispersion and more defects in concrete.
3.3 Early-age crack of sacrificial concrete
The crack number, length and width of each crack were
observed and recorded after concrete were cast and cur-
ing for 24h. And then the maximum crack length and
width, the mean crack area, the crack number and the
total crack area on unit area were calculated and sum-
marized in Table 6.
The cracking degree of the sacrificial concrete in-
creased with the amount of cementitious material as it is
400~500kg/m3. When the amount of cementitious mate-
rials is 500kg/m3, the maximum crack width of the sac-
rificial concrete is 1.6 times as much as that of the
400kg/m3, and the total cracking area on unit area in-
creased by over 20% than that of 450 and 400kg/m3. This
is due to the increase amount of cementitious materials,
the hydration heat would aggravate, and then, the risk of
cracking improved. In addition, comparing Y0 with N1,
when the replacement rate of mineral admixture in ce-
mentitious materials increased from 20% to 50%, the
maximum crack width of sacrificial concrete decreased
5 the basalt fiber, the crack resistance ability for the sacri-
4 ture increasing fiber addition. As the basalt fiber content
Flexural strength (MPa)
from 0.62mm to 0.49mm and the total cracking area on
unit area also decreased by 50%. For the sacrificial con-
crete with 400kg/m3 cementitious materials, the re-
placement ratio of mineral admixture was increased from
20% (U0) to 35% (T3), and the total cracking area on unit
area was also decreased by about 6.7%. Therefore, with
the increment of the replacement ratio of mineral ad-
mixture in cementitious materials, when the hydration
rate of fly ash and GGBS was slow and hydration heat
was still low, the cracking risk of the sacrificial concrete
was reduced accordingly.
For sacrificial concrete with 500 kg/m3cementitious
materials, 6kg/m3basalt fiber decreased the maximum
crack width by 0.33mm, reduced the total cracking area
on unit area by 24.2%. 1.5kg/m3 PP fiber reduced the
maximum crack width and the total cracking area on unit
area by 56% and 45%, respectively. However, because
the amount of cementitious materials is too high, even if
the basalt fiber and PP fiber were mixed in sacrificial
concrete, the cracking risk of the sacrificial concrete was
much higher than that of the others.
For the sacrificial concrete with 450kg/m3 cementi-
tious materials, the cracking risk decreased with incre-
ment content of PP fiber. When the content of PP fiber
reached 1.5kg/m3, early cracking would not occur. For
ficial concrete is not significantly increased with its fu-

reached 9kg/m3, the total cracking area on unit area de-

creased by only 6%. The fiber blocking effect also ap-
3
peared in the sacrificial concrete with 400kg/m3 cemen-
titious material. As the content of PP fiber reached
2
1.5kg/m3, the maximum crack width was only 0.2mm
and the total cracking area on unit area reduced to
1 111.13mm2/m2.

0
MF0 MF1 MPF1 Y0 Y4 Y5 U0 UF0 UPF0
3.4 Restrained shrinkage of sacrificial concrete
Fig. 5 Flexural strength of fiber reinforced sacrificial con- The cracking behavior of PP fiber reinforced sacrificial
crete. concrete in restrained shrinkage conditions was tested
Z. Jin, X. Zhao, T. Zhao and Y. Cao / Journal of Advanced Concrete Technology Vol. 14, 502-510, 2016 508

Table 7 Cracking parameters of sacrificial concrete in restrained shrinkage conditions.

PP fiber Initial cracking Mean cracking Maximum cracking Total cracking
Group Cracking number
(kg/m3) time(d) width(mm) width(mm) area(mm2)
Y0 0 11 1 0.71 0.82 106.5
Y4 1.0 15 1 0.45 0.51 67.5
Y5 1.5 13 1 0.35 0.37 55.5

and shown in Table 7. strength of fiber reinforced sacrificial concrete increased

Figure 6 (a) indicates the development of crack width
with the age of cracking for the ring specimens of plain
sacrificial concrete and PP fiber reinforced concrete. It
shows that the plain sacrificial concrete cracked at 11
days, whereas the cracking time of sacrificial concrete
incorporating 1.0 kg/m3 and 1.5 kg/m3 of PP fiber were
the 15th day and the 13thdays, respectively. And the
maximum crack width in plain sacrificial concrete is
about 0.82mm, while the maximum crack widths in
concrete with 1.0 and 1.5kg/m3 PP fiber are 0.51mm and
0.37mm, respectively. With the PP fiber increased from 0
to 1.5kg/m3 in sacrificial concrete, the total crack area
also decreased with addition of PP fiber and its crack area
reduced by 48%. Obviously, the PP fiber in sacrificial
concrete allows bridging the crack by preventing growth
of the crack. Early-age crack results also indicated the
number of cracks increased but the crack width de-
creased with the volume fraction of PP fiber, which
suggested that PP fiber dispersed the shrinkage stress and
avoid the stress concentration.
The evolution of strain in inner steel ring was tested
and shown in Fig. 6 (b). Considering the elastic modulus
of steel ring is constant, Fig. 6(b) indicated that shrink-
age stress induced by plain sacrificial concrete is the
maximum, and this stress increased with drying time.
When the sacrificial concrete cracked, the stress was
released. Even after crack, the shrinkage stress applied
by plain sacrificial concrete is about twice as much as the
concrete with 1.0 kg/m3 PP fiber. Since the tensile
strength of sacrificial concrete is low, when restrained,
concrete becomes vulnerable to cracking due to shrink-
age strains. Improvement in tensile strength of the con-
crete provides resistance against shrinkage cracking and
delayed the cracking time. The experiment results of
flexural strength in Fig. 5 indicated that the flexural
by 10-20%, compared to that of plain sacrificial concrete.
And the flexural strength of sacrificial concrete with
1.0kg/m3 PP fiber is 1.05 times of concrete with 1.5
kg/m3 PP fiber. Therefore, the increased tensile strength
improved the crack resistance of concrete with PP fiber,
but once the content of PP fiber is excessive, it would
increase the macro-defects of sacrificial concrete and
move up the cracking time.
3.5 Numerical simulation about influence of
crack on molten depth of sacrificial concrete
In order to further reveal the influence of crack on tem-
perature evolution and molten depth of sacrificial con-
crete layer, the numerical modeling and performance of
sacrificial concrete is established by using engineering
simulation software. The parameters used in this section
are based on the following assumptions:
a) The temperature of molten core is 3000°C. And
the initial temperature, melting temperature,
thermal conductivity and thermal expansion coef-
ficient of sacrificial concrete is 40°C, 1200°C,
1.22 W/m·K and 30 × 10-6 m/K, respectively.
b) The size of sacrificial concrete layer is 500×
200mm. And the crack width and depth is 1mm
and 50mm, respectively.
The numerical simulation results about influence of
crack on the inner temperature evolution and molten
depth of sacrificial concrete were shown in Figs. 7 and 8.
Obviously, the crack zone of sacrificial concrete layer
begins to melt at the 7.3 second, when it initially contacts
with the melt of the reactor core. After exposed for 1 hour,
the melting depth in crack zone increased with a melting
depth of 70.4mm, which is higher than 25.6mm for
non-crack sacrificial concrete. After 6.9 hours, all sacri-
ficial concrete layers melted completely. And this time is

1.0 60
3
0kg/m PP fiber (a)
0.9 (b)
3
1.0kg/m PP fiber 40
0.8 3 3
Maxmium crack width (mm)

1.5kg/m PP fiber 0kg/m PP fiber

0.7 20 3
1.0kg/m PP fiber
Strain (µm/mm)

3
0.6 1.5kg/m PP fiber
0
0.5
0.4 -20
0.3
-40
0.2
0.1 -60
10 12 14 16 18 20 22 24 26 28 30 32 0 4 8 12 16 20 24 28 32
Time (days) Time (days)
Fig. 6 Cracking behavior of sacrificial concrete in restrained shrinkage conditions, (a) crack width, (b) strain in inner steel
ring.
Z. Jin, X. Zhao, T. Zhao and Y. Cao / Journal of Advanced Concrete Technology Vol. 14, 502-510, 2016 509

(a) 7.3s (b) 1h (c) 6.9h

Fig. 7 Temperature distribution of sacrificial concrete contacted with molten core for different time.

3. The numerical simulation about influence of crack

200 Non-crack with width of 1mm on the inner temperature evolution
180 Crack
and molten depth of sacrificial concrete indicated that the
160
molten depth in crack zone is higher than that of
140
non-crack sacrificial concrete by 50% after contacting
Molten depth (mm)

120
100
80
60
40
20
0
0 1 2 3 4 5
Time (h )
Fig. 8 Influence of crack on molten depth of sacrificial
concrete.
1 hour ahead of that of the non-crack sacrificial concrete
layer. Therefore, the crack control of sacrificial concrete
is very important for the design of sacrificial concrete
layer in core spreading compartment and core catcher
zone.
4 Conclusions
The workability, density, compressive strength, flexural
strength, early-age crack and restrained shrinkage crack
behavior of sacrificial concretes with total cementitious
materials of 400-500kg/m3, 20-50% replacement ration
of mineral admixtures and two types of fiber were in-
vestigated. Based on the findings of this study, the fol-
lowing conclusions were drawn:
1. The influence of basalt fiber and PP fiber on the
density and compressive strength of sacrificial concrete
could be ignored. The slump of fresh concrete decreased
with the volume fraction of fibers, and the flexural
strength of fiber reinforced sacrificial concrete was in-
creased by 10-20% comparing to plain sacrificial con-
crete.
2. The compressive strength of sacrificial concrete
increased with total cementitious materials, while the
early-age crack risk obviously increased with total ce-
mentitious materials content. Replacement of cement by
high volume of fly ash and GGBS caused a significantly
reduction of early-age shrinkage crack of sacrificial
concrete.
with molten core for1 hour.
4. PP fiber with the content range from 1.0 kg/m3 to
1.5 kg/m3 can reduce the early-age crack of sacrificial
concrete, and the compressive strain in the steel ring of
the concrete with PP fiber was reduced by 50%, the total
crack area decreased by 37-48% comparing to the plain
6 7 sacrificial concrete.
5. The slump loss of fresh sacrificial concrete caused
by basalt fiber is less than that of PP fiber, 3-6 kg/m3
basalt fiber decreased the cracking area on unit area of
sacrificial concrete by 30-50%, but the improving crack
resistance capacity of basalt fiber was weaker than that of
PP fiber.
Acknowledgments
This work is part of a series project financially supported
by the Chinese National Natural Science Foundation
(NSF) Grant No. 51378269 and No.51420105015, and
the Chinese National 973 project Grant No.
2015CB655100. The authors gratefully appreciate the
financial support provided by the NSF and other Foun-
dations.
References
Afroughsabet, V. and Ozbakkaloglu, T., (2015). “Me-
chanical and durability properties of high-strength
concrete containing steel and polypropylene fibers.”
Construction and Building Materials, 94, 73-82.
Choun, Y. S. and Park, J. H., (2015). “Evaluation of
seismic shear capacity of perstressed concrete con-
tainment vessels with fiber reinforcement.” Nuclear
Engineering Technology, 47, 756-765.
Chu, H. Y., Jiang. J. Y., Sun, W. and Zhang, M. Z.,
(2016). “Thermal behavior of siliceous and
ferro-siliceous sacrificial concrete subjected to
elevated temperatures.” Materials and Design, 95,
470-480.
Dai, J. G., Akira, Y., Wittmann, F. H., Yokota, H. and
Zhang, P., (2010). “Water repellent surface impregna-
tion for extension of service life of reinforced con-
Z. Jin, X. Zhao, T. Zhao and Y. Cao / Journal of Advanced Concrete Technology Vol. 14, 502-510, 2016
crete structures in marine environments: The role of
cracks. “ Cement and Concrete Composites, 32,
101-109.
Fiore, V., Scalici, T., Di Bella, G. and Valenza, A.,
(2015). “A review on basalt fibre and its composites.”
Composites Part B, 74, 74-94.
Fischer, M., (2004). “The severe accident mitigation
concept and the design measures for core melt
retention of the European Pressurized Reactor
(EPR).” Nuclear Engineering and Design, 230(1-3),
169-180.
Guneyisi, E., Mehmet, G. and Ozbay, E., (2010).
“Strength and drying shrinkage properties of
self-compacting concretes incorporating multi-system
blended mineral admixtures.” Construction and
Building Materials, 24, 1878-1887.
Jin, Z. Q., Sun, W., Hou, B. R. and Jiang, J. Y., (2010).
“Deformation and microstructure evolution of
concrete subjected to elevated temperature.” Journal
of Southeast University (Natural Science Edition)
40(3), 619-623.
Jin, Z. Q., Zhao, T. J., Chen, H. S. and Zhuang, Q. C.,
(2012). “Chloride corrosion of splitting cracked
concrete under marine environment.” Journal of
Central South University (Science and Technology),
43, 2821-2826.
Kalifa, P., Menneteau, F.-D. and Quenard, D., (2000).
“Spalling and pore pressure in HPC at high
temperatures.” Cement and Concrete Research, 30,
1915-1927.
Komlev, A. A., Almjashev, V. I. and Bechta, S. V.,
(2015). “New sacrificial material for ex-vessel core
catcher.” Journal of Nuclear Materials, 467, 778-784.
Kitsutaka, Y. and Tsukagoshi, M., (2014). “Method on
the aging evaluation in nuclear power plant concrete
structures.” Nuclear Engineering and Design, 269,
286-290.
510
Mermerdas, K. and Arbili, M. M., (2015). “Explicit
formulation of drying and autogenous shrinkage of
concretes with binary and ternary blends of silica
fume and fly ash.” Construction and Building
Materials , 94, 371-379.
Naus, D. J., Oland, C. B. and Ellingwood, B. R., (1999).
“Summary and conclusions of a program addressing
aging of nuclear power plant concrete structures.”
Nuclear Engineering and Design, 194, 73-96.
Ouda, A. S., (2015). “Development of high-perform-
ance heavy density concrete using different aggre-
gates for gamma-ray shielding.” Progress in Nuclear
Energy, 79, 48-55.
Pikus, G. A., (2016). “Steel fiber concrete mixture
workability.” Procedia Engineering, 150, 2119 -
2123.
Stephan, P. and Salin, J., (2012). “Ageing management
of concrete structure: Assessment of EDF
methodology in comparison with SHM and AIEA
guides.” Construction and Building Materials, 37,
924-933.
Seleem, H., Rashad, A. M. and El-Sabbagh, B. A.,
(2010). “Durability and strength evaluation of
high-performance concrete in marine structures.”
Construction and Building Materials, 24, 878-884.
Saeid, K., Hazizan, M. A., Morteza, J. and Jalal, R.,
(2012). “The effects of polypropylene fibers on the
properties of reinforced concrete structures.” Const-
ruction and Building Materials, 27(1), 73-77.
Sevón, T., Kinnunen, T., Virta, J., Holmström, S.,
Kekki, T. and Lindholm. I., (2010). “HECLA experi-
ments on interaction between metallic melt and
hematite-containing concrete.” Nuclear Engineering
and Design, 240, 3586-3593.
Sim, J., Park, C. and Moon, D. Y. (2005). “Characteris-
tics of basalt fiber as a strengthening material for con-
crete structures.” Composites: Part B, 36, 504-512.