Professional Documents
Culture Documents
(Asce) Be 1943-5592 0001460
(Asce) Be 1943-5592 0001460
Abstract: The simple for dead load and continuous for live load (SDCL) steel bridge system has proven to be economical for nonseismic
application. Although the system has been used in different methods of construction, this paper concentrates on accelerated bridge construction
(ABC) application. The application of the SDCL steel bridge system in high seismic areas has been nonexistent, mainly due to lack of
appropriate details. The extension of the SDCL system to high seismic areas was initiated by conducting a detailed numerical analysis to
comprehend the types of forces that the connection must resist. The results of the numerical study led to the development of connection details
over the middle pier. In this research, a component testing of the proposed SDCL connection was conducted to find the ultimate limit states,
verification of numerical results, and merits of the proposed detail in light of established seismic design provisions. The developed connection
behaved as designed and prevented damage to capacity-protected elements. The column showed sufficient ductility during cyclic tests, which
indicated the connection should perform well under high levels of displacement. DOI: 10.1061/(ASCE)BE.1943-5592.0001460. Ó 2019
American Society of Civil Engineers.
Introduction and Background pier and joined together over the middle pier using a concrete di-
aphragm. Fig. 1(b) shows a schematic of an SDCL system.
Application of an SDCL steel bridge system has many advantages
Nonseismic SDCL
including eliminating field splices, eliminating expansion joints, re-
The simple for dead load and continuous for live load (SDCL) steel duced negative moment over the pier, and minimized traffic inter-
bridge system has been used in conventional and accelerated con- ruption. Further, encasing the ends of the girder in concrete protects
struction methods of building bridges, mainly in nonseismic areas. the girder ends and results in enhanced service life and lower in-
The SDCL steel bridge system for nonseismic areas was developed spection and maintenance costs as compared to conventional steel
at the University of Nebraska–Lincoln (Azizinamini et al. 2005). bridge systems.
The SDCL system is providing new opportunities for developing Conventional steel bridge system construction with lap splices
economical multispan steel bridge systems. The SDCL system is may demand closure for intersecting traffic. However, using an
especially well-suited for multispan bridges with each span having a integrated deck-girder system (modular approach) with an SDCL
maximum length of about 75 m. A complete summary of research, connection will reduce the impact on the feature intersected. Fur-
application, and performance relating to the SDCL steel bridge thermore, for an SDCL bridge system with deck-girder modules, the
system, as applied to nonseismic areas, using conventional and cast-in-place activity is limited to the concrete diaphragm, as op-
accelerated bridge construction (ABC) methods, is provided else- posed to conventional steel bridge construction, in which the entire
where (Lampe et al. 2014; Azizinamini 2014; Farimani et al. 2014; deck may be cast in place. This results in reducing the on-site
Yakel and Azizinamini 2014; Javidi et al. 2014). A brief intro- construction time.
duction to the SDCL system follows. For a detailed description of Fig. 2 schematically shows the ABC application of SDCL in
the system, refer to the references cited previously. nonseismic areas using a modular approach. In this approach the
Fig. 1(a) shows construction details of a conventional two-span contributory width of the deck is cast on top of the girders and
continuous steel bridge girder. For continuity over a pier, the middle shipped to the final site. These modular units are placed over the
segment of the steel girder is placed and then connected to end pier and abutments and then joined together using longitudinal
segments with either a bolted or welded field splice. The launching closure joints and a transverse joint (diaphragm over the pier). The
and placement of end segments usually require two cranes on site key design item is the type of detail that should be used to join the
with possible traffic interruptions. In the SDCL system, for exam- modules over the pier.
ple, for a two-span bridge, girders are placed over the abutment and Joining the girders over the pier using the introduced SDCL
connection makes the superstructure integral with the substructure.
According to the seismic design philosophy of bridges, the super-
1
Research Assistant, Dept. of Civil and Environmental Engineering, structure including the integral connection should remain elastic
Florida International Univ., Miami, FL 33174 (corresponding author). during a seismic event. To date, the application of SDCL steel
ORCID: https://orcid.org/0000-0001-5172-9393. Email: asade010@fiu.edu bridge systems in high seismic areas has been nonexistent, due to
2
Research Assistant, Dept. of Civil and Environmental Engineering, lack of suitable details for seismic areas. In a seismic event, the
Florida International Univ., Miami, FL 33174. connection joining the girders over the pier may experience positive
3
Professor and Chair, Dept. of Civil and Environmental Engineering, moments. Under positive moment the bottom flange will have a
Florida International Univ., Miami, FL 33174.
tendency to pull out of the concrete diaphragm. Therefore, the
Note. This manuscript was submitted on May 30, 2018; approved on
March 28, 2019; published online on July 19, 2019. Discussion period open nonseismic SDCL detail (Fig. 2) may not be suitable for seismic
until December 19, 2019; separate discussions must be submitted for in- application (Taghinezhadbilondy et al. 2018).
dividual papers. This paper is part of the Journal of Bridge Engineering, Ó The focus of this paper is to suggest a suitable seismic detail and
ASCE, ISSN 1084-0702. verify its merits using an experimental test. The extension of the
© ASCE 04019094-1 J. Bridge Eng.
(b)
Fig. 1. Steel bridge construction: (a) conventional continuous bridge girder; and (b) SDCL bridge system.
SDCL steel bridge system to high seismic areas was initiated by ments are to remain elastic during an entire seismic event. These
conducting detailed numerical analysis. The analysis was performed elements are called capacity-protected elements. The inelasticity is
to understand the types of forces that the connection detail over forced to form in the predefined locations, which have sufficient
middle piers, where girders are joined together, must resist in high ductility. These damage locations in bridges are located at the ends
seismic areas. The end result was the development of a detail for of columns (forming plastic hinges). In the SDCL steel bridge
connecting the girders over the middle pier and making them inte- system, the integral connection of the superstructure and substruc-
gral with bridge columns. The next section provides an overview of ture causes the damage location to be at the end of the column near
the suggested detail for connecting the modules over the pier. the cap beam.
In the initial stages of the numerical study, several connection
details were considered based on similar studies on prestressed
Connection Detail for SDCL System
concrete girders (Vander Werff et al. 2015; Sritharan et al. 2014).
in High Seismic Areas
Following preliminary analysis of these connections, one of them
The design philosophy for bridges in seismic areas is to predefine was chosen for further development (Taghinezhadbilondy 2016).
locations for damage to take place and design them for adequate The chosen connection was best suited for steel girders, with a slight
levels of ductility. In this design approach, the superstructure ele- modification to nonseismic SDCL (Fig. 3).
© ASCE 04019094-2 J. Bridge Eng.
Fig. 3. Developed SDCL connection details for seismic areas: (a) tie bars; (b) steel blocks; (c) end stiffeners; (d) dowel bars; and (e) live load
continuity reinforcement.
their function in an SDCL steel bridge system under high seismic testing was to verify the performance of the suggested detail before
activities. carrying out the shake table test. This project was a joint investi-
The tie bars and shear studs on the compression flange [Fig. 3(a)] gation between FIU and University of Nevada, Reno (UNR). FIU
in the proposed detail are the main difference between the details was responsible for developing the detail for extending the appli-
for nonseismic and seismic application of SDCL. These ties are cation of the SDCL steel bridge system to high seismic areas, fol-
to accommodate possible tension forces between the girders’ lowed by a shake table test of a scaled two-span steel bridge system
bottom flanges. The tension may occur under positive moments at UNR incorporating the FIU detail.
in the pier area, resulting from high vertical seismic excitations.
The area of the tie bars should be designed to resist a positive
moment induced by 25% of the dead load acting upward. Experimental Program
Steel blocks at the end of the compression flanges [Fig. 3(b)] are
used to transfer the compression forces between the girders’ This section provides details of an experimental testing program for
bottom flanges. This compression force is generated by super- verifying the recommendations of numerical studies. A prototype
imposed dead and live load negative moment. The width of the two-span steel I-girder bridge was selected for finding the demand
block is equal to the width of the bottom flange, and the height of side of the detail over the pier under seismic loads. The prototype
it is suggested by the previous work to be one-sixth of the height bridge was designed and scaled down to one-third for the purpose of
of the girder. The block size should be checked for negative this research. The scaled bridge was designed to undergo the same
moment generated from the governing live load combination; stresses as the prototype bridge.
the resulting moment arm is the distance between the deck The prototype bridge has two 28.96-m spans 9.75 m wide and
tension reinforcement and the center of the steel blocks. The was designed based on AASHTO provisions (AASHTO 2012). The
blocks are welded to the end of the compression flanges. superstructure consists of four steel I-girders of 101.6 cm depth and
End stiffeners [Fig. 3(c)] from the nonseismic version of the SDCL 31.4 N/cm (W40 215) and a 19-cm deck. Simply supported end
connection had to be modified for placing the tie bars between the abutments and a middle pier bent supported by two columns were
compression flanges. These stiffeners stiffen the webs at the beam assumed for this bridge. In this research, for the purpose of com-
end against vertical buckling and provide a smoother transfer of ponent testing, a pier bent column and girders on both sides were
forces from beam to concrete diaphragm over the middle pier considered in an inverted test setup.
when beam ends are subjected to positive or negative moments.
Dowel bar [Fig. 3(d)] reinforcements, similar to the available Column
detail for the integral cap beam, are designed for the torsion and The column was designed to sustain large inelastic deformations
shear in a cap beam. Torsion and shear in the bent cap occur prior to failure. Caltrans specifies target upper limits of displace-
under longitudinal (along-traffic) excitations, and load transfer ment ductility to reduce demand imparted to capacity-protected
from girders to columns. These bars are also the main mecha- elements (Caltrans 2010). Thus, the longitudinal reinforcement of
nism to resist the forces developed as a result of moment reversal the scaled column and volumetric ratio and spacing of spirals were
during the longitudinal component of the ground motion (par- designed in order to meet seismic provisions. As a result of scaling
allel to traffic flow). Results of numerical studies indicated that down, the size of the prototype column was scaled while the rein-
the design of dowel bars could be based on established Caltrans forcement ratio was kept constant. An axial load of 418 kN, as-
(2010) design provisions for capacity-protected elements (Ta- sumed to be 10% of the product of the gross cross-section area and
ghinezhadbilondy 2016). concrete compressive strength, was applied to the column. The
Live load continuity reinforcements [Fig. 3(e)] are placed to
length of the column, from the face of the cap beam to the line of
provide continuity for live load and superimposed dead loads. action of the lateral load, is 1.63 m. The lateral load was applied to
The live load continuity deck reinforcement is incorporated in the the connection through a rectangular cap at the end of column.
deck design. In ABC application of SDCL, the deck reinforce-
ments need to be developed in the diaphragm. One approach is by
Girders and Deck
hooking them inside the concrete diaphragm as shown in Fig. 3.
The global and local behavior of the numerical model of the seismic The superstructure was designed as simply supported for the dead
detail under push-down loading were similar to those of the non- loads of deck concrete and steel girders, and as continuous girders
seismic detail. Under push-up loading, finite-element results showed under live loads and superimposed dead loads. The steel I-girders
that continuity of the bottom flange increases the ductility and ca- were scaled down to A709 GR50 steel plate girders, with a
pacity of the connection. Because the bottom flange was not con- 45.7 0.95 cm web and 15.2 1.59 cm flanges. The deck was
tinuous, tie bars helped the system to increase the ultimate moment scaled down to 7.6 cm thick with #13 (12.7 mm) Grade 60 rein-
capacity. Under reversal loading, dowel bars were the most critical forcing bars at 12.7-cm spacing. The composite action between the
elements of the connection. The results demonstrated that tie bars girder and deck was achieved by providing four 1.9-cm-diameter
over the bottom flanges were unable to provide additional moment shear studs spaced at 45.7 cm on center. The details of the super-
capacity for the system under the moment-reversal loading config- structure are shown in Fig. 5.
© ASCE 04019094-4 J. Bridge Eng.
excitation (parallel to traffic), column deformations result in double strength capacity of the column and capacity of the cap beam in
curvature, and under transverse excitation, column deformations re- both directions.
sult in single curvature or double curvature, as shown in Fig. 6. After designing the cap beam and diaphragm based on the ca-
The prototype bridge used for this project has a two-column pacity approach, different details of the cap beam, vertical bars
bent in the middle pier, so the cap beam was designated and (dowel), horizontal bars, side bars, etc., were designed and checked
Fig. 7. Capacity design of cap beam: (a) column moment curvature and calculated overstrength capacity; (b) cap beam moment-curvature in
transverse direction compared to column overstrength capacity; and (c) cap beam moment-drift in longitudinal direction [from nonlinear finite-element
(NFE) analysis in Taghinezhadbilondy (2016)] compared to column overstrength capacity.
according to Caltrans procedures. The cap beam, designed as a are shown in Fig. 8 and they meet the requirements defined by
whole, consisted of a 20.3 61.0 cm (8 24 in.) dropped cap, a Caltrans.
48.3 61.0 cm (19 24 in.) diaphragm, and a 7.6-cm (3-in.) deck. Live load continuity reinforcement [Fig. 3(e)] are placed in the
The final dimensions of the cap beam were 76.2 61.0 cm (30 24 longitudinal (parallel to traffic flow) direction in the deck to
in.), shown in Fig. 8. As mentioned previously, the dropped cap achieve live load continuity. The design of this reinforcement is
section [20.3 61.0 cm (8 24 in.)] was designed to support the based on the negative moment over the pier, developed by the
weight of steel girders, deck concrete, and construction loads before governing live load combination. For the test specimen 10 #13
casting the concrete diaphragm over the pier. Details of the cap (12.7 mm) bars were provided in the deck.
beam and column reinforcements are shown in Fig. 8.
The elements of the suggested SDCL detail for high seismic
Test Setup
areas for the test specimen were designed based on the detailed
finite-element analysis described elsewhere (Taghinezhadbilondy The results of the numerical study indicated that the behavior of the
2016) and resulted in the following details: SDCL connection is more critical under along-traffic excitations. In
Tie bars between the shear studs on the compression flange the case of an integral bridge, the longitudinal component of earth-
[Fig. 3(a)] in the form of two U-shaped /10-mm bars were quake load results in a double curvature deformation with the in-
added on each side of each girder between two shear studs flection point at the middle of the column (Fig. 9), when column ends
1.9 cm (0.75 in.) in diameter on each compression flange. The are fixed. The length of the superstructure is taken between inflection
total area of tie bars was designed based on the numerical work points of two adjacent spans, calculated by placing uniform load
and as explained before. over the two spans. Similarly, the length of the column represents the
Steel blocks [Fig. 3(b)] 5.1 5.1 15.2 cm (2 2 6 in.) were location of the inflection point. After a thorough review of various
welded to the ends of the compression flanges. The sizing of test setups used in previous similar investigations (Wassef et al.
these steel blocks was based on previous work on the develop- 2004; Patty et al. 2001; Moustafa and Mosalam 2015), an inverted
ment of an SDCL detail for nonseismic areas. The height of the test setup as shown in Fig. 10 was selected for experimental inves-
blocks was slightly reduced so as not to obstruct the tie bars. tigation. In the inverted setup, the ends of the girders were simply
End stiffeners [Fig. 3(c)] in the form of two 1.3-cm plates were supported, and the lateral load was applied to the end of the column.
welded to the web and tension flange, 10 cm from the end of The length of the test specimen was adjusted to approximately
each girder. It is recommended that the thickness of end stiff- account for the fact that the inverted test setup caused an additional
eners be equal to or greater than the thickness of the web. The negative moment in the pier due to the dead weight of the specimen.
height of the web stiffeners was based on detailing to accom- Figs. 9 and 10 show the test setup and resulting shear and moment
modate the placement of tie bars. diagrams. The application of axial compressive load to the column,
Dowel bar [Fig. 3(d)] reinforcements, as mentioned previously, created by the dead weight of the structure, was achieved by using
are designed for torsion and shear in the cap beam. The details threaded bars as shown in Fig. 10. The lateral load was applied
© ASCE 04019094-6 J. Bridge Eng.
Fig. 9. Deflected shape, moment diagram, and shear diagram of bridge system and inverted specimen under longitudinal excitations.
(a) (b)
Fig. 10. Test specimen and setup: (a) details and dimensions; and (b) test specimen constructed in FIU structures laboratory.
Test Observations
The observed damages at the end of each third cycle at different
displacement ductility ratios are shown in Fig. 13. The first cracks
were seen during the first cycle on the column in the plastic hinge
region. The first signs of crushing in the cover concrete were also
observed during the first cycle. It is believed that this behavior
occurred because the specimen was subjected to load levels causing
yielding of column reinforcement during the first cycle to establish
the first yield displacement. Most of the observed cracks were in the
Fig. 12. Experimental definition of first yield displacement. plastic hinge region of the column. Cover concrete started to spall at
a displacement ductility ratio of 2. The spalling was extended to
Fig. 13. Observed damage in the plastic hinge region at different levels of testing.
larger areas in later cycles but was limited to the expected plastic ductility ratio 6, the column longitudinal reinforcement on both
hinge region. sides exhibited signs of buckling. As the testing regime was con-
Limited superficial cracking was observed on the cap beam near tinued to a displacement ductility ratio of 7, one of the column
the column on both sides. The first crack on the cap beam formed at longitudinal reinforcement fractured. This rebar fracture resulted in
a displacement ductility ratio of 3. By the end of the test, a few a 20% strength reduction and the testing was halted. The hysteresis
cracks were observed on either side of the cap beam, as shown in loops did not show any strength degradation before the fracture of
Fig. 14. The cracked concrete was removed after the test and the rebar.
depth of cracking was estimated to be less than 5 cm, and it was
limited to the cap beam cover concrete. The limited cracking in the
cap beam is believed to be mostly a result of scaling. The minimum Test Results and Discussion
cap beam width suggested by Caltrans is 0.6 m larger than the The resulting lateral load versus lateral displacement information
column diameter. The scaling down of the connection resulted in a as obtained from testing is shown in Fig. 16. Fig. 16 also shows
cap beam 0.2 m wider than the column diameter. This reduced width the load-displacement push-over curve as obtained from numer-
is insufficient for joint shear transfer and thus believed to be the ical analysis, demonstrating good agreement with test results. The
reason for cracking on the cap beam. During the test, the diaphragm overall behavior of the specimen was symmetrical under cyclic
and the deck remained intact without any visible cracks or yielding. loading.
The cracking and crushing patterns as obtained from numerical The distribution of maximum tensile strains on column longi-
results are shown in Fig. 15. Although monotonic loading was tudinal reinforcement during each displacement ductility ratio is
applied to the numerical models, the patterns match the experi- plotted in Fig. 17. The graph corresponds to strains on the longi-
mental observations. Limited cracking was also visible on the cap tudinal bar on the right, which experienced maximum strains in the
beam near the column, similar to the experimental results. Non- section. The longitudinal bar on the left showed similar peaks,
linearity in the models was limited to the parts shown in Fig. 15 and which are not shown in this figure. The results are only plotted for
connection details remained elastic. four ductility ratios. At higher ductility ratios, the steel strains in the
plastic hinge region exceeded the range of the sensors. It should be
noted that the loading direction was right to left in this figure.
Mode of Failure
As shown in Fig. 17, the longitudinal reinforcement, within 5 cm
The specimen was able to resist three cycles at a displacement below the face of the cap beam, experienced strains exceeding yield
ductility ratio of 6 (6.5% drift). In the last cycle of displacement strain. However, at about 20 cm below the cap beam face, strains
© ASCE 04019094-9 J. Bridge Eng.
(a) (b)
Fig. 15. (a) Crushing and (b) cracking patterns in numerical model.
were well below the yield strain. Higher measured strain inside the difference in the loading scheme: cyclic loading causes development
cap beam was expected due to the conventional detail used for the of more cracks and distribution of nonlinearity in concrete.
column-to-cap-beam connection. This also explains cracking around The strain measurements validate the fact that the cap beam and
the column on the cap beam (Fig. 14). diaphragm remained undamaged. The peak strain values increased
The strain measurements on the vertical legs of closed stirrups at higher displacement ductility ratios, which correspond to the
within the concrete diaphragm, or dowel bars [Fig. 3(d)], as they are extent of cracking in the cap beam and damage in the column.
referred to in this paper, are shown in Fig. 18. The strains are plotted Although the cracking in the cap beam was only visible near the
for all the ductility ratios that the specimen experienced. The max- column, strains on the bars were an indication of microcracks in the
imum measured strain on the dowel bars was about 900 microstrain, region. Existence of these cracks does not contradict the fact that
as indicated in Fig. 18. Fig. 19 shows the strain values from mono- the cap beam remained protected because the measured strains
tonic push-over numerical models compared with the experimental were well below the steel yield strain.
results at two displacement ductility ratios. The maximum values and Strain measurements on the tie bars [Fig. 3(a)] showed a gradual
the pattern obtained from the experiment correspond to the numer- increase in strain during the test. The maximum strain observed
ical results; however, the discrepancy between them is due to the in the tie bars during entire loading cycles was about 300 micro-
© ASCE 04019094-10 J. Bridge Eng.
[Fig. 3(b)], which are close to the tie bars, were not affected in this
experiment. These steel blocks are mostly critical under gravity type
loadings.
The distribution of maximum tensile strains in the longitudinal
reinforcement in the deck [Fig. 3(e)] during the displacement duc-
tility ratio of 6 is shown in Fig. 20. As shown in this figure, the
observed maximum tensile strains in the longitudinal reinforcement
was 205 microstrain, well below the yield strain. Comparing the
flexural capacity of the column to the flexural capacity of the su-
perstructure, the girders and deck were expected to be in the linear
region, and the deck strains verified this fact. The strain in the deck
Fig. 18. Strain measurements on dowel bars. correlated with the lateral load applied. The lateral load did not in-
crease significantly in higher ductility ratios (Fig. 16), so the mea-
surement for the last ductility ratio is plotted.
strain. Comparing the strain measured on the dowel bars (about 900 The maximum principal strains on the girders and cap-beam re-
microstrain) and the tie bars (about 200 microstrain) verified the inforcement, at the maximum displacement ductility ratio (7) applied
aforementioned conclusion from the numerical results—namely, to the column, as obtained from the numerical model, are shown in
that the dowel bars play the main role under along-traffic excitations Fig. 21. The figure shows that all the capacity-protected elements
(reversal loading). The positive strains on the tie bars correspond to remained below yield strain at this displacement ductility ratio.
tension forces in these elements. As anticipated, the steel blocks The curvature distributions along the height of the column for
© ASCE 04019094-11 J. Bridge Eng.
Fig. 21. Maximum principal strain obtained from numerical model on cap-beam reinforcement and girders.
Conclusions
Lincoln, NE: NDOR. Taghinezhadbilondy, R. 2016. “Extending use of simple for dead load and
Caltrans. 2010. Seismic design criteria. Version 1.6. Sacramento, CA: continuous for live load (SDCL) steel bridge system to seismic areas.”
Caltrans. Ph.D. thesis, Dept. of Civil and Environmental Engineering, Florida Int.
Farimani, R., S. Javidi, D. Kowalski, and A. Azizinamini. 2014. “Numerical Univ. https://doi.org/10.25148/etd.fidc001238.
analysis and design provision development for the simple for dead load- Taghinezhadbilondy, R., A. Yakel, and A. Azizinamini. 2018. “Deck-pier
continuous for live load steel bridge system.” Eng. J. 51 (2): 109–126. connection detail for the simple for dead load and continuous for live
Javidi, S., A. Yakel, and A. Azizinamini. 2014. “Experimental investigation, load bridge system in seismic regions.” Eng. Struct. 173: 76–88. https://
application and monitoring of a simple for dead load-continuous for live doi.org/10.1016/j.engstruct.2018.06.086.
load connection for accelerated modular.” Eng. J. 51 (3): 177–198. Vander Werff, J., R. Peggar, Z. Cheng, and S. Sritharan. 2015. Seismic
Lampe, N., N. Mossahebi, A. Yakel, R. Farimani, and A. Azizinamini. performance of precast girder-to-cap connections for accelerated
2014. “Development and experimental testing of connections for the bridge construction of integral bridges. Rep. No. CA16-2265. Ames,
simple for dead load-continuous for live load steel bridge system.” Eng. IA: Dept. of Civil, Construction and Environmental Engineering, Iowa
J. 51 (2): 83–106. State Univ.
Moustafa, M. A., and K. M. Mosalam. 2015. “Seismic response of bent caps Wassef, W. G., D. Davis, S. Sritharan, J. Vander Werff, R. E. Abendroth, J.
in as-built and retrofitted reinforced concrete box-girder bridges.” Eng. Redmond, and L. F. Greimann. 2004. Integral steel box-beam pier caps.
Struct. 98: 59–73. https://doi.org/10.1016/j.engstruct.2015.04.028. NCHRP Rep. 527. Washington, DC: Transportation Research Board.
Park, R. 1989. “Evaluation of ductility of structures and structural assem- Yakel, A., and A. Azizinamini. 2014. “Field application case studies and
blages from laboratory testing.” Bull. N. Z. Nat. Soc. Earthquake Eng. long-term monitoring of bridges utilizing the simple for dead-continuous
22 (3): 155–166. for live bridge system.” Eng. J. 51 (3): 155–175.