Nuclear Engineering and Design 311 (2017) 50–59

Contents lists available at ScienceDirect

Nuclear Engineering and Design

journal homepage: www.elsevier.com/locate/nucengdes

Probabilistic finite element investigation of prestressing loss in nuclear

containment wall segments
Georgios P. Balomenos ⇑, Mahesh D. Pandey
Department of Civil and Environmental Engineering, University of Waterloo, 200 University Avenue West, Waterloo, ON N2L 3G1, Canada

h i g h l i g h t s

Probabilistic finite element framework for assessing concrete strain distribution.

Investigation of prestressing loss based on concrete strain distribution.

Application to 3D nuclear containment wall segments.

Use of ABAQUS with python programing for Monte Carlo simulation.

a r t i c l e i n f o a b s t r a c t

Article history: The main function of the concrete containment structures is to prevent radioactive leakage to the
Received 10 May 2016 environment in case of a loss of coolant accident (LOCA). The Canadian Standard CSA N287.6 (2011) pro-
Received in revised form 17 November 2016 poses periodic inspections, i.e., pressure testing, in order to assess the strength and design criteria of the
Accepted 22 November 2016
containment (proof test) and the leak tightness of the containment boundary (leakage rate test). During
Available online 30 November 2016
these tests, the concrete strains are measured and are expected to have a distribution due to several
uncertainties. Therefore, this study aims to propose a probabilistic finite element analysis framework.
Keywords:
Then, investigates the relationship between the concrete strains and the prestressing loss, in order to
Prestressed concrete wall segments
Nuclear containment structure
examine the possibility of estimating the average prestressing loss during pressure testing inspections.
Probabilistic finite element analysis The results indicate that the concrete strain measurements during the leakage rate test may provide
Monte Carlo simulation information with respect to the prestressing loss of the bonded system. In addition, the demonstrated
ABAQUS framework can be further used for the probabilistic finite element analysis of real scale containments.
Ó 2016 Elsevier B.V. All rights reserved.

1. Introduction
Nuclear power plants (NPPs) play a major role for the global
energy supplies, while in the province of Ontario (Canada) 50% of
the electricity is generated by NPPs (Mirhosseini et al., 2014).
The CANDU (CANada Deuterium Uranium) nuclear reactors are
housed by the Gentilly-2 type secondary containment structures
(Elwi and Murray, 1980). This containment structure is circular
and consists of a concrete base, a cylindrical perimeter wall, a ring
beam and a dome (Simmonds et al., 1979), while its basic dimen-
sions are shown in Fig. 1 (Murray and Epstein, 1976a; Murray et al.,
1978). The main function of the containment is to prevent any
radioactive leakage to the environment, if a serious failure occurs
to the process system (Pandey, 1997). Thus, the containment is
designed to withstand the loss of coolant accident (LOCA), where
⇑ Corresponding author.
E-mail addresses: gbalomen@uwaterloo.ca (G.P. Balomenos), mdpandey@uwa-
terloo.ca (M.D. Pandey).
both temperature and pressure are increased inside the contain-
ment due to steam release, leading to increased tensile stresses
in the concrete walls (Lundqvist and Nilsson, 2011). Therefore,
the containment is made of prestressed concrete, either using
bonded or unbonded tendons (Anderson et al., 2008), in order to
ensure integrity and tightness in case of an accident (Anderson,
2005). However, the reliability of the containment is significantly
affected by the degradation of the tendon force (Kim et al.,
2013). Thus, the containment integrity is vulnerable to prestress-
ing losses due to actual material deformations, i.e., creep and
shrinkage of concrete and relaxation of tendons, and due to corro-
sion of the tendons (Pandey, 1997).
For the evaluation of the bonded prestressing system, Appendix
A of the CSA N287.7 (2008) provides three types of tests on both
bonded and unbonded test beams, namely flexural tests, lift-off
tests and a destructive test, while a more detailed review on the
above inspection procedures can be found in literature (Pandey,
1996a). In general, flexural tests involve testing of at least 12

http://dx.doi.org/10.1016/j.nucengdes.2016.11.018
0029-5493/Ó 2016 Elsevier B.V. All rights reserved.
 G.P. Balomenos, M.D. Pandey / Nuclear Engineering and Design 311 (2017) 50–59 51

still have a distribution due to uncertainties, where this distribu-

tion is expected to change due to prestressing losses. Therefore,
there is a need to investigate this change in the distribution of
the concrete strain with respect to the prestressing loss in tendons.
Thus, this study first proposes an easy to implement probabilistic
finite element analysis framework. Then, examines if concrete
strain distribution changes can provide information with respect
to the prestressing loss in bonded prestressing systems.

2. Wall specimens

2.1. Test description

The selected wall specimens (Fig. 2) are part of a research pro-

gram at the University of Alberta, which was sponsored by the
Fig. 1. Sketch of the prototype containment structure. Atomic Energy Control Board of Canada. The main objective of
the research program was to investigate the overpressure effect
on the Gentilly-2 type secondary containment structures (Elwi
bonded beams to evaluate the concrete cracking but do not quan-
and Murray, 1980). The first report of the series is divided in two
tify the prestress losses. Lift-off tests require the testing of at least
volumes (Murray and Epstein, 1976a, 1976b) and provides the
4 unbonded beams to measure the prestressing loss at the end of
description of the prototype containment structure together with
the tendon but cannot detect corrosion, since the tendons are per-
the main objectives of the research, followed by a second report
manently greased, and cannot evaluate the prestressing loss of
(Murray et al., 1977) and a third report (Murray et al., 1978).
bonded systems. Destructive tests use a sample from the previous
A series of tests were conducted on reinforced concrete wall
flexural test bonded beam to detect corrosion through visual
segments (specimens 4 and 7) and on prestressed concrete wall
examination of the tendon. Thus, a direct assessment of the pre-
segments (specimens 1–3, 5 to 6 and 8–14), leading to 14 tested
stressing loss of the bonded tendons it is not possible. For contain-
specimens in total. All tested specimens except specimen 7, have
ments with unbonded tendons, the lift-off technique is general
dimensions which correspond to a 1:4 scale of the prototype con-
used during regular in-service inspections, in order to assess the
tainment. Thus, each specimen has a width of 266.7 mm, i.e.,
prestressing loss at the end of the tendons (Anderson et al.,
almost one-fourth of the wall thickness, and a tendon duct size
2008), where it was found that the average prestressing loss along
almost one-fourth the size of ducts used in the prototype. Speci-
the tendon is smaller compared to the measured prestressing loss
men 7 was considered in order to be evaluated the scale effects.
in the end of the tendon (Anderson et al., 2005).
Thus, the thickness of specimen 7 was increased 1.5 times, i.e.,
On the other hand, Clause 6 and 7 of the CSA N287.6 (2011) pro-
400.05 mm, which corresponds to a 1:3 scale of the prototype con-
vide the proof test and the leakage rate test requirements, respec-
tainment, while its reinforcement size, reinforcement spacing and
tively. These are non-destructive techniques, which involve the
concrete cover was also increased proportionally. The lateral
pressurizing of an existing containment structure. This predefined
dimensions were chosen as three times the wall thickness
pressure is equal to 1.15 times the design pressure for the proof
(3  266.7 = 800.1 mm), due to laboratory restrictions regarding
test and equal to the design pressure for the leakage rate test
the total lateral applied force and due to crack observations regard-
(CSA N287.6-11). Under this load the stress–strain is measured in
ing allowing the formation of more than one through the wall
order to be assessed the strength and design criteria of the contain-
crack. The technical report No. 81 (Simmonds et al., 1979) provides
ment (proof test) and the leak tightness of the containment bound-
a detailed description and the test results of the specimens 1–9 and
ary (leakage rate test), where a more detailed review can be found
11–13, while the technical report No. 80 (Rizkalla et al., 1979) pro-
in literature (Pandey, 1996b). Concrete wall strains have been mea-
vides a detailed description and the test results of the two addi-
sured during pressure tests (using SOFO fiber-optic gauges), where
tional specimens involving air leakage, i.e., specimens 10 and 14.
these measurements were in general consistent (Hessheimer et al.,
Specimens 1 and 2 are selected in this study, which represent
2003). However, the measured strains during a pressure test will
the prestressing conditions and loading of the cylindrical wall of

Fig. 2. Sketch of the wall specimen with: (a) non-prestressed reinforcement; (b) prestressed reinforcement (tendon orientation in the containment structure).
52 G.P. Balomenos, M.D. Pandey / Nuclear Engineering and Design 311 (2017) 50–59

Table 1 603 Mpa and modulus of elasticity of 200 GPa. During the test
Overview of concrete variables considered in the wall segment tests. the load was applied in both directions with a different loading
Specimen Concrete ratio for specimens 1 and 2. A detailed overview of the variables
Min. cover (mm) Compressive Modulus of that were considered in the selected wall segments is given in
strength (MPa) elasticity (MPa) Tables 1 and 2.
1 12.70 35 25,924
In the prototype containment structure, the 4 tendon direction
2 12.70 31 27,027 represents the horizontal direction, while the three tendon direc-
tion represents the vertical direction (Fig. 2b). The capacity of the
testing machine was bigger in the vertical direction (Simmonds
et al., 1979), thus the testing segment was rotated 90 degrees com-
Table 2
Overview of steel variables and loading ratio considered in the wall segment tests. pared to the corresponding orientation in the prototype structure.
The detailed location of the 3 tendon (Axial) and the 4 tendon
Specimen Non-prestressed Loading Prestressing force in tendons
(Hoop) direction is shown in Figs. 3 and 4, respectively.
reinforcement ratio (effective after losses)
Per layer Axial/hoop fPH (MPa) fPA (MPa)
2.2. Force under internal pressure
1 10 #10 @ 76.2 mm 1:2 931.5 850.1
2 10 #10 @ 76.2 mm 1:2 919.8 855.6
The prototype containment structure has been designed for an
Note: Axial refers to the 3 tendon direction; Hoop refers to the 4 tendon direction; internal pressure equal to 124 kPa (Murray and Epstein, 1976a),
fPH = prestressing force in hoop direction; fPA = prestressing force in axial direction.
while its internal dimeter is equal to 41,452.8 mm  41.45 m and
its wall thickness is equal to 1066.8 mm  1.07 m (Fig. 1). The con-
tainment has an inner-radius to wall-thickness ratio bigger than 10
(r/t = 20, 726.4/1066.8 = 19.43 > 10). Thus, the containment can be
analyzed using thin-wall analysis (Hibbeler, 2011), in order to cal-
culate the developed stresses under any internal pressure inside
the containment. For a thin-wall cylindrical pressure vessel the
developed stresses are calculated as (Beer et al., 2006)

rH ¼ pr=t ð1Þ

rA ¼ pr=2t ð2Þ

where rH is the hoop stress (also called circumferential), rA is the

Fig. 3. Sketch of the 3 tendon location (axial or meridional direction).
Fig. 4. Sketch of the 4 tendon location (hoop or circumferential direction).
the containment structure (Fig. 1). These are square panels of
800.1 mm with a width of 266.7 mm and are prestressed in both
directions (Fig. 2). The hoop direction (or circumferential) consists
of 4 tendons with 7 smooth wires in each tendon and the axial
direction (or meridional) consists of 3 tendons with 6 smooth
wires in each tendon. Each smooth wire has a diameter of
7.01 mm, yield strength of 1627 Mpa, ultimate strength of 1820
Mpa and modulus of elasticity of 200 GPa. Apart from the tendons,
all the selected specimens are reinforced with two grids, where
each grid consists of 10 #10 (metric units) non-prestressed bars
in each direction (Fig. 2a). Each bar of the non-prestressed rein-
forcement has yield strength of 401 Mpa, ultimate strength of
axial stress (also called meridional), p is the internal pressure, r is
the internal radius and t is the wall thickness.
In general, when r/t = 10 the thin-wall analysis predicts stresses
which are approximately 4% less than the actual maximum stress,
while as the r/t ratio is increased the relative error is decreased
(Hibbeler, 2011). Therefore, the containment can be analyzed in a
simple manner, provided it has a thin wall, and the developed
stresses under the design pressure p = 0.124 MPa are calculated
as rH = 2.41 MPa and rA = 1.205 MPa. The wall segments represent
part of the containment, i.e., specimens 1 and 2 represent part of
the containment’s cylindrical wall. Thus, under internal pressure
the developed stresses result to forces in hoop (FH) and axial (FA)
direction.
Each specimen is a square panel (800.1 mm) with a width of
266.7 mm, resulting to a cross section area A = 213,387 mm2 for both
directions. Therefore, under the design pressure p = 0.124 MPa, the
developed force in the hoop direction is calculated as FH = 514.26 kN
and in the axial direction is calculated as FA = 257.13 kN. For the
proof test the applied pressure is equal to 1.15 times the design pres-
sure; for the leakage rate test the applied pressure is equal to the
design pressure (CSA N287.6-11). Thus, the required hoop prestress-
ing force after losses is FH,proof = 1.15  2.41  213.387 = 591.40 kN
for the proof test and FH,leak = 1.00  2.41  213.387 = 514.26 kN for
the leakage rate test. Similar, the required axial prestressing force
after losses is FA,proof = 1.15  1.205  213.387 = 295.70 kN for the
proof test and FA,leak = 1.00  1.205  213.387 = 257.13 kN for the
leakage rate test.
3. Deterministic finite element analysis
3.1. Modelling of specimens
Deterministic FEA is applied to the selected wall specimens
using the commercial FEA software ABAQUS (2012). Simple
 G.P. Balomenos, M.D. Pandey / Nuclear Engineering and Design 311 (2017) 50–59 53

Fig. 5. Geometry, load and boundary conditions of the specimens.

supports are introduced around the bottom edge and to the one Fig. 6. Reinforcement layout of the specimens.
lateral edge of the specimens, while the load is applied with a small
velocity through the top and the other lateral edge of the speci-
mens (Fig. 5). The summation of the reactions at these supports
gives the total measured load in each direction. ABAQUS has two
main analysis products, i.e., the ABAQUS/Standard and the ABA-
QUS/Explicit. ABAQUS/Standard solves the equations of the system
implicitly at each solution increment, while ABAQUS/Explicit goes
with the solution forward through small time increments without
solving the equations of the system at each increment. ABAQUS/
Standard uses a stiffness-based solution technique that is always
stable, while ABAQUS/Explicit uses an explicit integration solution
technique which is conditionally stable. It can be said herein, that
some analysis problems even if they can be effectively solved in
ABAQUS/Standard, they appear difficulty in converging because
of the material complexity, resulting in many iterations. Such anal-
yses become expensive in ABAQUS/Standard because each of the
‘‘many iterations” demands many equations that have to be solved. Fig. 7. Tensile stress-crack displacement curve of concrete.
On the other hand, ABAQUS/Explicit determines the solution with-
out iterating by explicitly taking advantage of the kinematic state
of the previous increments, making the analysis more efficient equal to rb0/rc0 = 1.16. The tensile behavior of concrete is consid-
and less time consuming. Taking into account the required compu- ered using the tension stiffening approach, where the fracture
tational cost for the following probabilistic FEA in this study, energy of concrete should be first specified (Fig. 7). The fracture
Quasi-static analysis in ABAQUS/Explicit is performed using small energy (Gf) is obtained from the CEB-FIP Model Code 90 (1993),
velocity, leading to a smaller computational cost per increment depending on the maximum aggregate size and the compressive
compared to the implicit method (Genikomsou and Polak, 2015). strength of concrete. Thus, the fracture energy for the specimen 1
The concrete is modeled using 8-noded hexahedral elements and 2 is calculated as 0.0703 N/mm and 0.0655 N/mm, respec-
with reduced integration (C3D8R), while the reinforcement is tively. The compressive behavior of concrete is modeled with the
modeled using 2-noded 3D linear truss elements (T3D2). The Hognestad parabola (Fig. 8). The elastic behavior of both non-
embedded option is adopted which assumes perfect bond between prestressed reinforcements and tendons is defined through the
the concrete and the reinforcement, while the reinforcement lay- modulus of the elasticity (Es) and the Poisson’s ratio (v) with values
out for the specimens is shown in Fig. 6. A mesh sensitivity study equal to 200 GPa and 0.3, respectively. The plastic behavior of both
was performed in advance, indicating that the results are almost non-prestressed reinforcement and tendons is defined based on an
non mesh sensitive. Thus, the concrete part of the specimens is input stress-strain relationship and it is shown in Table 3 (Elwi and
meshed using only one brick element, in order to decrease the total Murray, 1980).
computational time for each FE analysis.
The behavior of the concrete is simulated using the concrete 3.2. Modelling of the prestressing in tendons
damaged plasticity model which is offered in ABAQUS (2012). Con-
crete damaged plasticity model was developed by Lubliner et al. The prestressing in tendons is modeled by applying either
(1989) and then modified by Lee and Fenves (1998). This model initial stress or initial temperature to the tendons. Using the first
considers both the tensile cracking and compressive crushing of approach, i.e., initial stress, the prestressing to the tendons is
concrete as possible failure modes. In this study the Poisson’s ratio introduced in the initial step. In the following step, the end of
is set equal to v = 0.2, the dilation angle is set equal to w = 36o, the the tendons are fully restrained (Ux = Uy = Uz = 0), while these
shape factor is set equal to Kc = 0.667, and the stress ratio is set boundary conditions are deactivated in the subsequent step and
54 G.P. Balomenos, M.D. Pandey / Nuclear Engineering and Design 311 (2017) 50–59

Fig. 8. Compressive stress-strain relationship of concrete.

Fig. 9. Curves of load-strain: Hoop direction of specimen 1.

Table 3
Stress-strain relationship for steel.

Non-prestressed bars Prestressed tendons

r (Mpa) e (103) r (Mpa) e (103)
0 0 0 0
401 2.04 1413 6.97
480 40.00 1572 8.40
– – 1634 10.00
– – 1655 12.00
– – 1724 20.00
– – 1732 41.00

Note: r = stress; e = strain.

simple supports are introduced to the bottom and to the one lateral
edge of the specimen (Fig. 5). In that way, the prestressing action is
taking place. In the final step, the load is applied to the top and to Fig. 10. Curves of load-strain: Axial direction of specimen 1.
the other lateral edge of the specimen (Fig. 5).
Using the second approach, i.e., initial temperature, the temper-
ature of the environment (20 °C) is introduced to the tendons in
the initial step, together with the simple supports to the bottom
and to the one lateral edge of the specimen. In the following step,
the prestressing action is taking place by applying a new tempera-
ture value to the tendons calculated as DT = rpe/(aEs), where rpe is
the prestressing in tendon, a is the thermal coefficient of linear
expansion of the tendon and Es is the modulus of elasticity of the
tendon. Here, the coefficient of linear expansion of steel is consid-
ered as 105 (1/°C). In the final step, the load is applied to the top
and to the other lateral edge of the specimen.
In general, the initial stress approach requires 4 steps in total
with a computational cost equal to 114 s per FEA trial and the ini-
tial temperature approach requires 3 steps in total with a compu-
tational cost equal to 78 s per FEA trial. Each FEA trial is executed
on a personal computer with Intel i7-3770 3rd Generation Proces-
sor and 16 GB of RAM.
Fig. 11. Curves of load-strain: Hoop direction of specimen 2.

3.3. Finite element analysis results

The FEA results of specimen 1 (Fig. 9, Fig. 10) and specimen 2
(Fig. 11, Fig. 12) are in good agreement compared to the test
results, in terms of load-strain curves, where temperature refers
to the initial temperature approach and stress refers to the initial
stress approach for modelling the prestressing in tendons. For
the leakage rate test, it is found that the required hoop prestressing
force is FH,leak = 514.26 kN and the required axial prestressing force
is FA,leak = 257.13 kN. Considering that the load-strain curves for all
specimens are linear in that range, linear interpolation is applied
for the previous mentioned forces. In this way, we are able to esti-
mate the hoop and the axial concrete strains based on the leakage
rate test (Table 4). In a similar manner, the hoop and the axial con-
crete strains are estimated based on the proof test, i.e., strains
which correspond to FH,proof = 591.40 kN and FA,proof = 295.70 kN,
respectively (Table 5). Both ways of modelling the prestressing
provide similar strain results, but the initial stress approach
requires more computational time. Thus, the initial temperature
approach for modelling the prestressing is chosen for the subse-
quent probabilistic analysis.
 G.P. Balomenos, M.D. Pandey / Nuclear Engineering and Design 311 (2017) 50–59 55

4. Probabilistic finite element analysis

4.1. General

Probabilistic FEA is applied using the Monte Carlo simulation

(MCS). The material properties (Tables 6–8) and the prestressing
loss in hoop and axial direction (Table 9) are considered as uncer-
tain, leading to 15 random variables in total for each specimen. The
mean value of the tensile strength of concrete is calculated based
on its compressive strength (Table 6). Although there is some cor-
relation between the compressive and the tensile strength of con-
crete (Hariri-Ardebili and Saouma, 2016), for sake of conservatism
this study considers that these two random variables are totally
independent.
The probability distribution of each random variable has been
Fig. 12. Curves of load-strain: Axial direction of specimen 2. selected based on relevant literature (Table 6). Note that quality
control measures implemented during the construction would
avoid the use of materials with exceptionally low strength. To
Table 4 reflect this fact in the simulation, the probability distributions of
Calculated concrete strains based on the loading used for the leakage rate test. the material properties could be suitably truncated. Since this
Specimen Hoop strain (l) Axial strain (l) information is not readily available, the continuous (or non-
Approach 1 Approach 2 Approach 1 Approach 2 truncated) probability distributions have been used in the simula-
tion process presented in the paper.
1 76.05 75.97 31.06 31.04
2 73.14 73.21 29.89 29.87
The coefficient of variation (COV) of the prestressing force is
considered as 4% for a new structure and as 12% for an old struc-
Note: Approach 1 refers to the initial stress approach for modelling the prestressing; ture (Ellingwood, 1984). The increased COV reflects the added vari-
Approach 2 refers to the initial temperature approach for modelling the pre-
stressing; l is the micro symbol for denoting a factor of 106.
ability due to highly uncertain time-dependent creep and
shrinkage of concrete (Pandey, 1997). Thus, for the hypothetical
cases of 20%, 25% and 30% prestressing loss, the COV is slightly
increased to 15% (Table 9). Since ABAQUS is a deterministic FEA
software, python development environment is used for developing
Table 5
Calculated concrete strains based on the loading used for the proof test. the deterministic FE model and then for updating the uncertain
input parameters for each FE simulation (Balomenos et al., 2015).
Specimen Hoop strain (l) Axial strain (l)
Approach 1 Approach 2 Approach 1 Approach 2
4.2. Probability distribution of concrete strains
1 87.47 87.39 35.74 35.71
2 84.14 84.21 34.40 34.37
MCS is applied with 103 trials for each specimen and prestress-
Note: Approach 1 refers to the initial stress approach for modelling the prestressing; ing loss scenario (2 specimen  4 scenarios/specimen = 8 scenar-
Approach 2 refers to the initial temperature approach for modelling the pre- ios). MCS is selected because is a well understood and easy to
stressing; l is the micro symbol for denoting a factor of 106.
implement method. Considering that this study is a first attempt
of checking the relation between the concrete strain and the

Table 6
Concrete statistics for each specimen.

Specimen Random variable Distribution Mean COV Reference

0
1 fc (MPa) Normal 35 0.135 Nowak et al. (2012)
0
ft (MPa) Normal 1.95 COVf 0c Ellingwood et al. (1980)
Ec (MPa) Normal 25,924 0.08 Rajashekhar and Ellingwood (1995)
0
2 f c (MPa) Normal 31 0.14 Nowak et al. (2012)
0
f t (MPa) Normal 1.85 COVf 0c Ellingwood et al. (1980)
Ec (MPa) Normal 27,027 0.08 Rajashekhar and Ellingwood (1995)
1, 2 cc (kN/m3) Normal 24 0.03 Ellingwood et al. (1980)
qffiffiffiffiffi
0 0 0
Note: f c = compressive strength of concrete; f t ¼ 0:33 f c = tensile strength of concrete; Ec = modulus of elasticity of concrete; cc = density of concrete.

Table 7
Non-prestressed reinforcement statistics for each specimen.

Specimen Random variable Distribution Mean COV Reference

1, 2 fy (MPa) Normal 401 0.04 Nowak and Szerszen (2003)
Es (GPa) Normal 200 0.033 Mirza and Skrabek (1991)
As (mm2) Normal 71.2 0.015 Rakoczy and Nowak (2013)
cs (kN/m3) Normal 78 0.03 Assumed

Note: fy = yield strength of steel; Es = modulus of elasticity of steel; As = cross-section area of steel (#10 Bars in metric units); cs = density of steel.
56 G.P. Balomenos, M.D. Pandey / Nuclear Engineering and Design 311 (2017) 50–59

Table 8
Prestressed reinforcement statistics for each specimen.

Specimen Random variable Distribution Mean COV Reference

1, 2 fy (MPa) Normal 1627 0.025 Nowak and Szerszen (2003)
Es (GPa) Normal 200 0.033 Mirza and Skrabek (1991)
As (mm2) Normal 38.6 0.015 Rakoczy and Nowak (2013)
cs (kN/m3) Normal 78 0.03 Assumed

Note: fy = yield strength of steel; Es = modulus of elasticity of steel; As = cross-section area of steel per wire (hoop direction consists of 7 wires per tendon; axial direction
consists of 6 wires per tendon); cs = density of steel.

Table 9
Prestressing loss statistics for each specimen.

Specimen APLS (%) Random variable Distribution Mean COV Reference

1, 2 15 fPH (MPa) Normal 0.85fPH 0.12 Ellingwood (1984)
fPA (MPa) Normal 0.85fPA 0.12 Ellingwood (1984)
20 fPH (MPa) Normal 0.80fPH 0.15 Assumed
fPA (MPa) Normal 0.80fPA 0.15 Assumed
25 fPH (MPa) Normal 0.75fPH 0.15 Assumed
fPA (MPa) Normal 0.75fPA 0.15 Assumed
30 fPH (MPa) Normal 0.70fPH 0.15 Assumed
fPA (MPa) Normal 0.70fPA 0.15 Assumed

Note: APLS = average prestressing loss scenario; fPH = prestressing force in hoop direction (Table 1); fPA = prestressing force in axial direction (Table 1).

Fig. 13. Histogram and distribution fitting of the hoop strain: Leakage rate test for Fig. 15. Normal probability paper plot of the hoop strain: Leakage rate test for
specimen 1 with 30% loss of prestressing. specimen 1 with 30% loss of prestressing.

Fig. 14. Histogram and distribution fitting of the axial strain: Leakage rate test for
specimen 1 with 30% loss of prestressing.
prestressing loss, 103 trials are selected for each scenario mainly
due to the high computational cost. However, variance reduction
techniques such as the Latin hypercube sampling can be used, for
better accuracy of the results. Especially for the case of the applied
framework on real scale containments, where the computational
cost is expected to increase.
After each MCS trial, the ABAQUS results are stored in terms of
load-strain values. Similar to the previous section, for each MCS the
concrete strain can be calculated using linear interpolation, for
either proof or leakage rate test. In this study, for the probabilistic
analysis we consider only the leakage rate test. Thus, the linear
interpolation is performed for each trial and the strains are calcu-
lated, i.e., the hoop strain is calculated for FH,leak = 514.26 kN and
the axial strain is calculated for FA,leak = 257.13 kN. This results to
a vector of 103 values of the hoop strain for each scenario and to
the same amount of values for the axial strain. Based on his-
tograms, the calculated hoop strains and axial strains are consid-
ered to follow a Normal distribution. Indicatively, this is clearly
shown for the hoop strain (Fig. 13) and the axial strain (Fig. 14)
of the specimen 1 for the 30% prestressing loss scenario.
In addition to histograms, the probability papers can also be
used in order to determine whether the observed data follow a par-
ticular distribution (Ang and Tang, 2007). Therefore, the normal
portability paper is plotted for the calculated hoop and axial
strains. Indicatively, the linearity of the normal probability paper
 G.P. Balomenos, M.D. Pandey / Nuclear Engineering and Design 311 (2017) 50–59 57

Fig. 16. Normal probability paper plot of the axial strain: Leakage rate test for Fig. 17. Probability distribution of the hoop strain: Leakage rate test for specimen 1.
specimen 1 with 30% loss of prestressing.

plot for the hoop strain (Fig. 15) and the axial strain (Fig. 16) of the
specimen 1 for the 30% prestressing loss scenario, indicates that
the calculated strains are represented very well by the Normal
distribution. Both histograms and normal probability paper plots
indicate that the probability distribution of the calculated hoop
and axial strain is following the Normal distribution, with mean
and standard deviation as reported in Table 10 (specimen 1) and
Table 11 (specimen 2). The probability paper plots also show that
some dots deviate from the straight line, especially in the tails of
the simulated distribution. This may correspond to extreme cases
in which the panels have cracked. However, this investigation is
out of the scope of the current study.
Based on the previous observations, the probability distribution
of the concrete strain in each direction (hoop and axial) can be
plotted for each specimen, following the Normal distribution with Fig. 18. Probability distribution of the axial strain: Leakage rate test for specimen 1.
mean and standard deviation as reported in Tables 10 and 11. The
concrete hoop strain distribution for each prestressing loss sce-
nario is shown in Fig. 17 (specimen 1) and Fig. 19 (specimen 2). while the next section examines the probability of having
The concrete axial strain distribution for each prestressing loss sce- increased average concrete strains due to increased prestressing
nario is shown in Fig. 18 (specimen 1) and Fig. 20 (specimen 2). It is loss. In addition, the standard deviation of the strain seems to
observed that the mean value of the concrete strain is increased decrease with the increase of the prestressing loss. This, results
with the increase of the prestressing loss, resulting to the strain to a decreased coefficient of variation of the concrete strain, indi-
distribution shifting to the right. Thus, the average concrete strain cating less variability to the concrete strains as the prestressing
is expected to increase with the increase of the prestressing loss, loss is increased.

Table 10
Statistics of concrete strains: leakage rate test for specimen 1.

APLS (%) Hoop strain Axial strain

Mean (l) Stdev (l) COV Mean (l) Stdev (l) COV
15 97.14 18.56 0.1911 41.95 11.39 0.2716
20 104.73 21.06 0.2011 45.58 13.07 0.2868
25 113.71 20.42 0.1795 48.31 11.85 0.2452
30 119.58 18.74 0.1567 52.50 10.76 0.2050

Note: APLS = average prestressing loss scenario; Stdev = standard deviation; COV = coefficient of variation; l is the micro symbol for denoting a factor of 106.

Table 11
Statistics of concrete strains: leakage rate test for specimen 2.

APLS Hoop strain Axial strain

Mean (l) Stdev (l) COV Mean (l) Stdev (l) COV
15% 95.44 17.26 0.1808 39.61 10.84 0.2735
20% 101.25 20.15 0.1990 42.78 13.02 0.3045
25% 108.33 18.57 0.1715 46.93 11.94 0.2544
30% 116.82 17.71 0.1516 49.89 10.92 0.2188

Note: APLS = average prestressing loss scenario; Stdev = standard deviation; COV = coefficient of variation; l is the micro symbol for denoting a factor of 106.
58 G.P. Balomenos, M.D. Pandey / Nuclear Engineering and Design 311 (2017) 50–59

Table 13
Probability of average concrete strain exceeding the average concrete strain of the
15% base case: Leakage rate test for specimen 2.

APLS (%) Hoop strain Axial strain

b p = U(b) b p = U(b)
15 N/Α N/Α N/Α N/Α
20 0.2187 0.587 0.1866 0.574
25 0.5082 0.694 0.4539 0.675
30 0.8646 0.806 0.6678 0.748

Note: APLS = average prestressing loss scenario; p = probability of the average

concrete strain (due to prestressing loss) exceeding the average concrete strain of
the 15% base case.

concrete strain for the base case, rY is the standard deviation of the
calculated average concrete strain for each hypothetical case and
Fig. 19. Probability distribution of the hoop strain: Leakage rate test for specimen 2.
qXY = 0 since they are assumed uncorrelated. It should be noted that
Eq. (3) is applied when both variables (X, Y) follow a normal distri-
bution. Otherwise, the variables should be first transferred to the
normal space, while further details can be found in Nowak and
Collins (2000).
The probability is then calculated as

p ¼ ½X 6 Y ¼ p½X  Y 6 0 ¼ UðbÞ ð4Þ

where p is the probability of the average concrete strain for each

Fig. 20. Probability distribution of the axial strain: Leakage rate test for specimen 2.
Table 12
Probability of average concrete strain exceeding the average concrete strain of the
15% base case: Leakage rate test for specimen 1.
hypothetical case exceeding the average concrete strain of the base
case and U is the standard Normal distribution function with mean
equal to zero and standard deviation equal to one. Table 12 (speci-
men 1) and Table 13 (specimen 2) shows the probability of the aver-
age concrete strain exceeding the average concrete strain of the 15%
base case. The results indicate that the probability of having an
average concrete strain bigger than the 15% base case is increased
with the increase of the prestressing loss. For the hoop strain, this
probability ranges from 0.59 to 0.61 for the case of 20% prestressing
loss and from 0.70 to 0.73 for the 30% case. Thus, comparing the
average concrete strain, e.g., measured during the leakage rate test,
with the average concrete strain of a selected base case can provide
us with information with respect to the average prestressing loss.

APLS (%) Hoop strain Axial strain

5. Conclusions
b p = U(b) b p = U(b)
15 N/Α N/Α N/Α N/Α This study presents a framework which combines probabilistic
20 0.2703 0.607 0.2095 0.583
analysis in conjunction with finite element analysis. The demon-
25 0.6004 0.726 0.3870 0.651
30 0.8507 0.803 0.6733 0.750 strated framework is then used for the analysis of the effect of
prestressing losses on the distribution of concrete strains. Two
Note: APLS = average prestressing loss scenario; p = probability of the average
3D containment wall segments, already tested at the University
concrete strain (due to prestressing loss) exceeding the average concrete strain of
the 15% base case. of Alberta, are modeled and analyzed using ABAQUS. The deter-
ministic finite element analysis (FEA) results indicate that the frac-
ture energy approach, used for modelling the tensile strength of
4.3. Probability of increased average concrete strains the concrete within the concrete damage plasticity model offered
by ABAQUS, captures effectively the behavior of the concrete in
Since the concrete strain distribution seems to be affected from terms of load-strain curves. The prestressing force of the tendons
the prestressing loss, the magnitude of this effect can be quantified is modeled using two different approaches, i.e., by introducing
using a parameter b, which is similar to the reliability index either initial stress or initial temperature variation to the tendons.
(Hasofer and Lind, 1974), as Deterministic FEA results indicate the accuracy of the two
modelling techniques, in terms of load-strain curves. However,
lX  lY the initial strain technique requires slightly more computational
b ¼ qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ð3Þ
ðrX Þ þ ðrY Þ2  ð2qXY rX rY Þ
2 time, since one extra step has to be introduced. Thus, the adopted
initial temperature can be considered as a computational economic
where X denotes the calculated strain distribution for the base case technique for modelling the prestressing force.
of 15% loss, Y denotes the calculated strain distribution for each The Monte Carlo simulation (MCS) is chosen for the probabilis-
hypothetical case of 20%, 25% and 30%, lX is the mean value of tic analysis, because it is an easy to use method. However, the com-
the calculated average concrete strain for the base case, lY is the putational cost will increase for probabilistic finite element studies
mean value of the calculated average concrete strain for each hypo- of real scale containments. Thus, a more efficient probabilistic
thetical case, rX is the standard deviation of the calculated average method should be used. MCS is implemented in ABAQUS with
 G.P. Balomenos, M.D. Pandey / Nuclear Engineering and Design 311 (2017) 50–59
the use of the python programing and it is performed for 8 cases in
total with 103 trials per case. The load-strain curve is stored after
each MCS and the concrete strain is estimated based on the leakage
rate test and using linear interpolation. Alternatively, the proof rate
test can also be used for the estimation of the concrete strains, fol-
lowing the already presented probabilistic framework.
The concrete strain measurements during periodical inspec-
tions, i.e., proof test or leakage rate test, may provide information
with respect to the prestressing losses, since the direct assessment
of prestressing loss of bonded tendons is not possible for a concrete
containment structure. The results indicate a high probability of
increase in the average concrete strain with the increase of the
average prestressing loss. For example, a 0.80 probability of the
average hoop concrete strain exceeding the average hoop concrete
strain of the 15% base case approximates a 30% average prestress-
ing loss in tendons. This probabilistic framework can be further
applied for real scale concrete containment structures.
Acknowledgements
The authors are grateful to the University Network of Excel-
lence in Nuclear Engineering (UNENE) and to the Natural Sciences
and Engineering Research Council (NSERC) of Canada for the finan-
cial support of this study.
References
ABAQUS 6.12-3, 2012. Abaqus Analysis User’s Manual. Dassault Systèmes Simulia
Corp, Providence, RI.
Anderson, P., Berglund, L.-E., Gustavsson, J., 2005. Average force along unbonded
tendons: a field study at nuclear reactor containments in Sweden. Nucl. Eng.
Des. 235 (1), 91–100.
Anderson, P., 2005. Thirty years of measured prestress at Swedish nuclear reactor
containments. Nucl. Eng. Des. 235 (21), 2323–2336.
Anderson, P., Hansson, M., Thelanderson, S., 2008. Reliability-based evaluation of
the prestress level in concrete containments with undonded tendons. Struct.
Saf. 30 (1), 78–89.
Ang, H.-S.A., Tang, H.W., 2007. Probability Concepts in Engineering: Emphasis on
Applications in Civil & Environmental Engineering. John Wiley and Sons Inc,
Hoboken, NJ.
Balomenos, G.P., Genikomsou, A.S., Polak, M.A., Pandey, M.D., 2015. Efficient
method for probabilistic finite element analysis with application to reinforced
concrete slabs. Eng. Struct. 103 (8), 85–101.
Beer, F.P., Johnston, R.E., DeWolf, J.T., 2006. Mechanics of Materials. McGraw-Hill,
Boston, MA.
CSA N287.6, 2011. Pre-Operational Proof and Leakage Rate Testing Requirements
for Concrete Containments Structures for Nuclear Power Plants. Canadian
Standard Association, Mississauga, Ontario, Canada.
CSA N287.7, 2008. In-Service Examination and Testing Requirements for Concrete
Containment Structures for CANDU Nuclear Power Plants. Canadian Standard
Association, Mississauga, Ontario, Canada.
CEB-FIP MC 90, 1993. Design of Concrete Structures. CEB-FIP-Model Code 1990.
Telford, London, UK.
Ellingwood, B., 1984. Probability Based Safety Checking of Nuclear Plant Structures,
NUREG/CR-3628. Brookhaven National Laboratory, Upton, NY.
Ellingwood, B.R., Galambos, T.V., MacGregor, J.G., Cornell, C.A., 1980. Development
of Probabilities Based Load Criterion for American National Standard A58. NBS
Special Publication No. 577, National Bureau of Standards, Washington, DC.
59
Elwi, A.E., Murray, D.W., 1980. Nonlinear Analysis of Axisymmetric Reinforced
Concrete Structures. Structural Engineering Report No. 87. Department of Civil
Engineering. University of Alberta, Edmonton, Alberta, Canada.
Genikomsou, A.S., Polak, M.A., 2015. Finite element analysis of punching shear of
concrete slabs using damaged plasticity model in ABAQUS. Eng. Struct. 98 (4),
38–48.
Hariri-Ardebili, M.A., Saouma, V.E., 2016. Sensitivity and uncertainty quantification
of the cohesive crack model. Eng. Fract. Mech. 155 (2), 18–35.
Hessheimer, M.F., Klamerus, E.W., Lambert, L.D., Rightley, G.S., 2003.
Overpressurization Test of a 1:4-Scale prestressed Concrete Containment
Vessel Model. NUREG/CR-6810, SAND2003-0840P, Sandia National
Laboratories, Albuquerque, NM.
Hasofer, A.M., Lind, N.C., 1974. Exact and invariant second-moment code format.
ASCE J. Eng. Mech. Div. 100 (1), 111–121.
Hibbeler, R.C., 2011. Statics and Mechanics of Materials. Pearson Prentice Hall Inc,
Upper Saddle River, NJ.
Kim, S.-H., Choi, M.-S., Joung, J.-Y., Kim, K.-S., 2013. Long-term reliability evaluation
of nuclear containments with tendon force degradation. Nucl. Eng. Des. 265,
582–590.
Lee, J., Fenves, G.L., 1998. A plastic-damage concrete model for earthquake analysis
of dams. Earthquake Eng. Struct. Dynam. 27 (9), 937–956.
Lubliner, J., Oliver, J., Oller, S., Onate, E., 1989. A plastic-damage model for concrete.
Int. J. Solids Struct. 25 (3), 299–326.
Lundqvist, P., Nilsson, L.-O., 2011. Evaluation of prestress losses in nuclear reactor
containments. Nucl. Eng. Des. 241 (1), 168–176.
Mirhosseini, S., Polak, M.A., Pandey, M.D., 2014. Nuclear radiation effect on the
behavior of reinforced concrete elements. Nucl. Eng. Des. 269, 57–65.
Mirza, S.A., Skrabek, B.W., 1991. Reliability of short composite bean-column
strength interaction. ASCE J. Struct. Eng. 117 (8), 2320–2339.
Murray, D.W., Epstein, M., 1976a. An elastic stress analysis of a Gentilly type
containment structure Volume 1. Structural Engineering Report No. 55.
Department of Civil Engineering, University of Alberta, Edmonton, Alberta,
Canada.
Murray, D.W., Epstein, M., 1976b. An elastic stress analysis of a Gentilly type
containment structure Volume 2 (Appendices B to F). Structural Engineering
Report No. 56. Department of Civil Engineering, University of Alberta,
Edmonton, Alberta, Canada.
Murray, D.W., Rohardt, A.M., Simmonds, S.H., 1977. A classical flexibility analysis
for gently type containment structures. Structural Engineering Report No. 63.
Department of Civil Engineering, University of Alberta, Edmonton, Alberta,
Canada.
Murray, D.W., Chitnuyanondh, L., Wong, C., Rijub-Agha, K.Y., 1978. Inelastic analysis
of prestressed concrete secondary containments. Structural Engineering Report
No. 67. Department of Civil Engineering, University of Alberta, Edmonton,
Alberta, Canada.
Nowak, A.S., Collins, K.R., 2000. Reliability of Structures. McGraw-Hill, NY.
Nowak, A.S., Szerszen, M.M., 2003. Calibration of design code for buildings (ACI
318): Part 1 – Statistical models for resistance. ACI Struct. J. 100 (3), 377–382.
Nowak, A.S., Rakoczy, A.M., Szeliga, E.K., 2012. Revised Statistical Resistance Models
for R/C Structural Components. ACI Special Publication, SP284-6, pp. 1–16.
Pandey, M.D., 1996a. Reliability-based Inspection of Prestressed Concrete
Containment Structures. INFO-0639, Atomic Energy Control Board of Canada,
Ottawa, Ontario, Canada.
Pandey, M.D., 1996b. Proof Testing of CANDU Concrete Containment Structures.
INFO-0646, Atomic Energy Control Board of Canada, Ottawa, Ontario, Canada.
Pandey, M.D., 1997. Reliability-based assessment of integrity of bonded prestressed
concrete containment structures. Nucl. Eng. Des. 176 (3), 247–260.
Rajashekhar, M.R., Ellingwood, B.R., 1995. Reliability of reinforced-concrete
cylindrical shells. ASCE J. Struct. Eng. 121 (2), 336–347.
Rakoczy, A.M., Nowak, A.S., 2013. Resistance model of lightweight concrete
members. ACI Mater. J. 110 (1), 99–108.
Rizkalla, S.H., Simmonds, S.H., MacGregor, J.G., 1979. Leakage Tests of Wall
Segments of Reactor Containments. Structural Engineering Report No. 80.
Department of Civil Engineering, University of Alberta, Edmonton, Alberta.
Simmonds, S.H., Rizkalla, S.H., MacGregor, J.G., 1979. Tests of Wall Segments From
Reactor Containments Volume 1, Structural Engineering Report No. 81.
Department of Civil Engineering. University of Alberta, Edmonton, Alberta,
Canada.