1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36
PERFORMANCE EVALUATION OF TRANSVERSE HPC CLOSURE POUR CONNECTION FOR USE IN MODULAR BRIDGE CONSTRUCTION
Yaohua Deng (Corresponding Author) Bridge Engineering Center, Institute for Transportation Iowa State University 2711 S. Loop Drive, Suite 4700, Ames, IA 50010-8664 Phone: 515-294-2882; Email:
[email protected] Brent M. Phares Bridge Engineering Center, Institute for Transportation Iowa State University 2711 S. Loop Drive, Suite 4700, Ames, IA 50010-8664 Phone: 515-597-5879; Fax: 515-294-0467; Email:
[email protected] Andrew J. Putz Bridge Engineering Center, Institute for Transportation Iowa State University 2711 S. Loop Drive, Suite 4700, Ames, IA 50010-8664 Phone: 515-971-3917; Email:
[email protected] Curtis Carter Office of Bridges and Structures, Iowa Department of Transportation 800 Lincoln Way, Ames, IA 50010 Phone: 515-233-7822; Email:
[email protected]
Submitted for possible publication to the Transportation Research Record, Journal of the Transportation Research Board in response to the following approved calls for papers: • Special Topics in Steel Bridge Design, Delivery, and Service Life Engineering • Emerging Technologies in Bridge Construction Word count: 5,165 words text + 9 tables/figures × 250 words (each) = 7,415 words Submission Date: July 31, 2016
TRB 2016 Annual Meeting
Original paper submittal - not revised by authors
Deng et al.
2
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29
ABSTRACT Accelerated Bridge Construction (ABC) techniques have been introduced to rapidly replace bridges to address increasing concerns about the nation’s aging infrastructure. The Simple-MadeContinuous (SMC) concept has been used as a technique for utilizing Prefabricated Bridge Elements and Systems (PBES) for achieving lower construction cost and time. Due to the potential for concrete crushing, the negative moment transfer mechanism of the transverse connections above the pier support are still a point of concern. The objective of the work described in this paper was to evaluate the behavior, strength, and failure modes of a transverse closure pour connection used for the constructed Little Silver Creek Bridge and to validate its adequacy for modular bridge application. The as-designed transverse connection consisted of a concrete diaphragm, deck, longitudinal reinforcement, a compression block, shims, steel girders, and end steel plates. For comparison purposes and to evaluate the need for the complicated compression block, two transverse connection specimens - one with and one without the compression block - were designed, fabricated, instrumented, and tested under a negative bending moment. FE models were also established to carry out performance evaluation of the two connections and to aid in the interpretation of the test results. Additionally, hand calculations were performed to estimate the moment capacity of the connections. It was found that no significant difference existed between the crack patterns of the two specimens and the diaphragm concrete tended to crush in the bottom regardless of the configuration. The established FE models were sufficient at representing the structural behavior of two transverse connection specimens. The connection with a compression block has higher yield and ultimate moment capacity than the connection without a compression block. To design both types of connections per the classic reinforced concrete design theory, an effective width equal to the girder bottom flange width and the centroid of the compression force at the bottom of the girder bottom flange can be assumed for the connections with and without a compression block, respectively. The hand calculations reasonably estimated on the moment capacity of the two specimens. The two connections were safe under the codified, factored loads, and especially, the connection with a compression block can sustain the entire codified, factored loads prior to yield of the deck reinforcement.
30 31
Keywords: Simple-made-continuous connection, Transverse closure pour connection, Compression block, Modular construction, Experimental testing, and Finite element analysis.
32
Deng et al.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45
3
INTRODUCTION Due to the increasing concern about the nation’s aging infrastructure, Accelerated Bridge Construction (ABC) techniques have been introduced to rapidly replace bridges such that traffic interruption and mobility and environmental impacts on the transportation network could significantly be minimized. The ABC techniques are/were being developed under the Second Strategic Highway Research Program (SHRP 2), which targeted strategic solutions to improve several aspects of transportation including, safety, congestion, and renewal methods for roads and bridges. One of the popular ABC techniques is to use Prefabricated Bridge Elements and Systems (PBES) which are commonly connected by closure pour connections. Prefabricated bridge elements fabricated in a controlled environment generally have a high level of quality. However, the closure pour connections commonly used in both the longitudinal and transverse directions are the most critical components in a modular bridge due to issues about serviceability, ductility, strength, and load transfer. For this study, a bridge replacement project consisting of the replacement of a bridge located on Iowa 92 over Little Silver Creek in Pottawattamie County, Iowa (see Figure 1) is presented. The performance of an ultra-high performance concrete (UHPC) longitudinal joint detail used for the bridge was evaluated in the past research by the authors (1). And in this paper, the behavior of a high performance concrete (HPC) transverse joint detail used for the bridge was further studied. To achieve simpler construction and improved performance of continuous steel girder bridges constructed in an ABC environment, transverse connections have been designed based on a concept called Simple for Dead, Continuous for Live Load, which has been used by several states including Nebraska, Colorado, New Mexico, New York, Ohio, and Tennessee (2). A steel bridge constructed using the simple-made-continuous (SMC) concept can result in lower construction costs and time compared to concepts utilizing field-spliced rolled girders (2). The SMC concept is idealized to produce: (a) a hinged connection for permanent dead loads during girder erection, and (b) a continuous connection for live loads and superimposed dead loads through transverse closure pours consisting of the diaphragm, deck, and longitudinal reinforcement. The SMC connections, commonly placed above the pier supports in the negative moment region, are composed of a tension region resisted by the longitudinal deck reinforcement, a compression region at the bottom of the diaphragm, and a horizontal shear region resisted by typical shear studs. Due to the potential for concrete crushing in the compression region, several types of SMC transverse connections have been introduced with the goal of designing a sufficient negative moment transfer zone above the pier support. Three types of connections were developed by Azizinamini et al. (3): Type I − the two girder bottom flanges of a connection are longitudinally connected by a steel plate prior to diaphragm concrete placement; Type II − the two steel girders are simply encased by the diaphragm concrete; and Type III − similar to Type II but end bearing plates are welded to the girder ends. Full-scale specimens using the three connection types were fabricated and tested until concrete crushing at the diaphragm bottom. They concluded that the Type I and III connections were superior to the Type II connection in terms of moment capacity, ductility, and the amount of yielded steel bars at failure. Niroumand (4) further improved the Type II connection developed by Azizinamini et al. through welding steel blocks and stiffeners at the girder ends to transfer the compression force at the diaphragm bottom. Talbot (5) also developed another type of connection which has a single shear bolted connection on the girder top flanges, a cover plate welded on the bottom of each bottom flange, two trapezoidal wedges placed between the bottom flanges, and a concrete diaphragm. Additionally, van de Lindt (6) proposed a non-
Deng et al.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46
4
concrete diaphragm connection which consists of a compression plate welded to the bottom flanges, a reinforced rod in the deck, an elastomeric bearing pad, and steel diaphragms. Based upon information contained in the technical literature at the time, a transverse connection with a concrete diaphragm and a steel block was designed for the Little Silver Creek Bridge (see Figure 1(c)) located in Pottawattamie County, IA by the Iowa Department of Transportation. To complete the transverse joints, prefabricated modules were placed against compression blocks, which were anchored into the piers. Shims were used to fill the gap between the compression blocks and the girder ends, and the girder ends were welded with steel plates. The compression block was used to ensure the transfer of the compressive force longitudinally from the one adjacent module to the next. To address premature deterioration due to environmental effects, HPC is used to form both the diaphragm and a 10-ft portion of the bridge deck over the pier location as shown in Figure 1(c). The objective of the work described in this paper is to evaluate the behavior, strength, and failure modes of the HPC transverse connection of the Little Silver Creek Bridge and to validate its adequacy for modular bridge application. Additionally, it was desired to understand the importance of including the somewhat complex compression block in future designs. To achieve the goals of this work, a pair wise comparison of the capacity and ductility at the failure limit state between connections with and without a compression block was made. EXPERIMENTAL PROGRAM An experimental program consisting of strength tests on transverse connections was designed and implemented in this work. Two specimens were designed, fabricated, instrumented, and tested to evaluate the performance of the transverse joints with and without a compression block subject to a negative bending moment. Specimen Design and Fabrication The test specimens were designed based on the actual details of the Little Silver Creek Bridge. The bridge has six prefabricated deck modules at each span and two transverse HPC transverse connections as shown in Figure 1(b). The girders of each prefabricated module are spaced at 4 ft6 in. and the adjacent girders between two modules are spaced at 3 ft-4 in. A specimen (referred to as “Specimen I”) was designed to exactly replicate the bridge details and included a compression block between the two longitudinally aligned girders as shown in Figure 2(a). Another specimen (referred to as “Specimen II”) with no compression block has the same design details as Specimen I minus the compression block and is shown in Figure 2(b). Each of the specimens consists of two steel girders, a deck panel, and a concrete diaphragm as shown in Figure 2. The steel girders have a W40x149 cross-section, a length of 7 ft-6 in, and a 9 in. longitudinal gap between them. Each girder has stiffeners on each side of the web over the support location as well as underneath the loading point as shown in Figure 2. The girder top flange has shear studs with a length of 6 in. and a diameter of 7/8 in. The shear studs have longitudinal and transverse spacing of 8 in. and 4-3/8 in, respectively. The diaphragm has a length of 2 ft-9 in., a width of 3 ft-11 in., and a depth of 4 ft-1.5 in. The deck panel has a depth of 8 in. and a width the same as the diaphragm. Due to the uneven girder spacing, the steel girders are offset 3.5 in. from the deck centerline resulting in one overhang being 7 in. wider than the other shown in Figure 2(c). The deck panel of each specimen has two identical layers of steel reinforcement as shown in Figure 2. Each layer contains eight #7 bars in the longitudinal direction and sixteen #6 bars in the transverse direction, evenly spaced at 5.5 and 12 in., respectively, as shown in Figure 2(e). The longitudinal and transverse reinforcement run continuously along the whole length and width of
Deng et al.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46
5
the specimen, respectively. Concrete cover is 1 and 2.75 in. at the deck bottom and top, respectively. The diaphragm has three #5 bars placed in the longitudinal direction through holes drilled in each girder web and four #5 bars used as shear reinforcement (i.e., the bent bars) as shown in Figures 2(a) and 2(b). The concrete was specified to use the standard Iowa DOT High Performance Concrete (HPC-D) with a nominal compressive strength of 5 ksi. The compression block and end steel plates are made of 1-in. thick steel plates and its details are shown in Figures 2(a), 2(b) and 2(e). It should be noted that only incorporating a compression block into one specimen, but keeping the rest of the design exactly the same made it possible to directly compare the results and to determine the effect the compression block had on the performance of the transverse connection. The specimens were fabricated in the Iowa State University Structural Engineering Laboratory. The shear studs were attached to the top flange as shown in Figure 3(a). The compression block was fabricated by the same fabricator that would provide them for the Little Silver Creek Bridge and the bearing stiffeners were full-length welded on each side. Construction of the formwork was started once the girders were delivered to the lab. The sides for the diaphragm were constructed and slid into place forming a tight fit with the bottom flange as shown in Figure 3(b). The deck was formed and steel braces were used to support the overhangs of the decks as shown in Figure 3(c). After all of the formwork was assembled, the reinforcing steel mats were lifted into place and the vertical diaphragm reinforcement was installed as shown in Figure 3(d). The concrete for the two specimens was placed using concrete from a single source and placed at the same time. Fifteen uniaxial strain gauges were installed on the top layer of the longitudinal bars near the diaphragm location prior to concrete placement. The specimens were tested after 28days of curing and the 28-day compressive strength of the concrete was measured to be 5.5 ksi. Strength Tests Strength tests were conducted on both specimens using the test setup shown in Figure 4. Supports were placed under the bearing pads to simulate the contact points between the diaphragm and the pier. Two point loads were applied 6 in. from the outside edges of the specimens to produce a negative moment region near the diaphragm. A load cell was placed on each side to measure the applied load throughout the testing process. The instrumentation plans for both specimens were exactly the same except for the gauges on the compression block. For each specimen, there were fifteen embedded strain gauges located in the deck panel and twelve surface mounted strain gauges located on the girders. For the strain gauges embedded in the deck panel, five gauges were mounted in the center of the diaphragm and five gauges were installed 6 in. away from each side of the diaphragm as shown in Figure 4(b). The surface mounted strain gauges on the girder were positioned on the bottom of the top flange, at the center of the web, and at the top of the bottom flange. Locations of these gauges were 6 in. outside the diaphragm and midpoint of the girder overhang for both the west and east side as shown in Figure 4(c). Note that the gages with the letter “A” were installed on Specimen I and the gages with the letter “B” were installed on Specimen II as shown in Figure 4. Additionally, two strain gauges were also installed on the outside face and mid-depth of the compression block of Specimen I. During the testing process loading was halted periodically to monitor crack formations. Conventional crack mapping techniques were used to document the failure patterns of each specimen. This was performed until the loading on each side reached approximately 300 kips and it was deemed no longer safe to approach the specimen. Cracks that formed after this point were marked after the load was taken off. The specimens were unloaded at a stage when the deflection increased significantly without adding much extra loading.
Deng et al.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
6
FINITE ELEMENT ANALYSIS Three-dimensional nonlinear FE models (see Figure 5) were established to further investigate the performance of the specimens. The steel beams and compression block were both modeled using a four-noded shell element with three translational and three rotational degrees of freedom at each node. An elastic–plastic uniaxial material model including bilinear kinematic hardening was used for the steel. The yield strength, elastic modulus, and Poisson’s ratio of the steel were set to 50 ksi, 29,000 ksi, and 0.3, respectively. The strain hardening modulus was set to 5% of the elastic modulus. The concrete deck and diaphragm were modeled using an eight-noded solid element which has three translational degrees of freedom at each node and incorporates cracking and crushing capabilities. The concrete material properties were assigned with multi-linear isotropic hardening in combination with the von Mises yield criterion. The stress-strain relationship of the concrete proposed by Hognestad (7) was utilized for the concrete constitutive model:
ε ε 2 fc = fc' 2 − ε o ε o
(1)
where, fc and ε are stress and strain on concrete respectively; and strain at peak stress (εo) is expressed as (8): ε o = 0.00078( f c' )1/4 (in MPa) (2) The smeared fixed crack model and Rankine maximum stress criterion were utilized to determine the initiation and development of concrete cracking. According to the AASHTO LRFD Bridge Design Specifications (9), maximum concrete tensile strength can be derived by:
ft ' = 0.63 fc'
(in MPa) (3) The steel bars were modeled using a uniaxial tension-compression element with three translational degrees of freedom at each node. An elastic–plastic uniaxial material model was used for the steel. The yield strength, elastic modulus, Poisson’s ratio, and tangent modulus of the steel were set to 60 ksi, 29000 ksi, 0.3, and 5% of the elastic modulus, respectively. The steel bars were perfectly connected to the concrete through sharing of common nodes. The concrete diaphragm was also perfectly connected to the steel beams and compression block through sharing common nodes. The shear studs between the concrete deck and the steel beams were modeled using a unidirectional spring element, which can incorporate a nonlinear generalized force-deflection relationship. The spring element was active along the slip direction and the other two directions of the two nodes were coupled together. The shear force-slip relationship proposed by Ollgard et al. (10) was incorporated in the spring element and can be expressed as:
Q Qn (1 − e =
2 −18 s 5
)
(4) where, Q = shear force in a shear stud; s = slip at the weld point of the stud; and according to AASHTO (9), nominal shear resistance, Qn, is determined by (5) = Qn 0.5 Asc f c' Ec ≤ Fu Asc where, Asc = cross-sectional area of the stud; f′c= compressive strength of concrete; Ec = elastic modulus of the concrete; Fu = minimum specified tensile strength of the stud (60 ksi). Due to the symmetry of the geometry, loading, and boundary conditions, a half model was established. Loads were applied on the nodes of each FE model at the loading location. Boundary conditions were defined taking into account both geometric symmetry at the symmetric sections
Deng et al.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46
7
and simply supported condition at the support locations. Convergence criteria and tolerances were set for the displacement and force. The following strategies were utilized to facilitate convergent computations (1,11,12,13,14): (1) concrete compressive stress was constrained to a constant value after reaching its peak value; (2) shear transfer coefficients of 0.15 and 0.9 were used for open and closed cracks, respectively; (3) capability of concrete crushing was deactivated in the analysis, but the failure of the model was determined when the concrete reached the maximum compressive strain of 0.003 (i.e., concrete failure strain); (4) suppression gf extra displacement shapes and tensile stress relaxation after cracking were applied to solid elements. RESULTS AND DISCUSSIONS Crack and Failure Patterns For both specimens, the first cracks were found on the top of the deck and located at the crosssection near the edge of the diaphragm as shown in Figure 6(a). As the load increased, the cracks propagated across the width of the specimens and additional cracks developed both on the deck top and within the diaphragm. For the short side, the cracks finally propagated to the bottom region of the diaphragm as shown in Figure 6(b). However, for the long side, the crack did not propagate down from the top to bottom of the diaphragm as shown in Figure 6(c). No significant difference was found between the two specimens in terms of the crack patterns. At the end of loading, cracks were also found at the diaphragm bottom as shown in Figure 6(d), which indicated the potential for concrete crushing at this region. After the testing, the concrete in the deck was removed and no fracture was observed in the longitudinal steel reinforcing bars. Comparisons of Measured and Predicted Results Load-strain relationships for the steel bars were developed based on the measured strains in the gages for the two specimens. Typical load-strain relationships for the cross-sections at the edge and the center of the diaphragm illustrated in Figure 7(a). For each load-strain relationship, a plateau (see Figures 7(a)) can be observed at an early loading stage due to the initiation of concrete cracking. For each specimen, the load at concrete cracking was determined based on the smallest plateau of the load-strain relationships. Likewise, the load at steel yield was determined based on the smallest load when the steel bars reached the yield strain (i.e., 2,069 micro-strains for 60-ksi steel). For each specimen, the failure load was determined to be the maximum load applied to the specimen. For consistence and comparison purposes, the cracking load, yielding load, and failure load for each specimen were further converted to the cracking moment, yielding moment, and ultimate moment, respectively, with respect to the location at the diaphragm center, which are summarized in Table 1. The failure load of each specimen as predicted by the FE model was determined based on concrete crushing in the compression region of the critical section of the diaphragm. Note that, for Specimen I, the critical section was located at the interface between the steel girder end and the compression block; and for Specimen II, the critical section was located at the steel girder end which was embedded within the diaphragm. Concrete crushing was determined based on the longitudinal compressive strain in the critical section exceeding 0.003 as shown in Figure 8(a). The load-strain relationships for the steel bars at the diaphragm center obtained from the FE models were compared with the test results and are shown in Figures 7(b). The other load-strain relationships for the steel bars showing similar patterns were omitted due to the word limitation. And, the cracking load and yielding load applied to the FE models were also determined based on the same approaches used for processing the experimental test results. Figure 7(b) indicates that the predictions from the FE models follow the similar orders as those obtained from the test results,
Deng et al.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46
8
however, the FE models show a higher stiffness and less ductility. Table 1 also indicates that cracking and yielding moments were slightly under-estimated and over-predicted by the FE models compared to the test results, respectively. The discrepancies in the prediction are due to several reasons: (1) the smeared crack model utilized in the FE models did not provide accurate prediction of the deformation caused by the cracks; (2) the bond stress-slip relationships between the concrete and steel bars, between the concrete and steel girders, and between the concrete and compression block were not taken into account. However, the ultimate moments of the two specimens predicted using the FE models are in good agreement with the test results as shown in Table 1. Furthermore, the load-strain relationships for the steel girders predicted using the FE models are in good agreement with the test results as shown in Figure 7(c). Other load-strain relationships for the steel girders showing similar patterns were omitted. And the load-strain relationships for the compression block predicted using the FE model compare well with the test results as shown in Figure 7(d). Consequently, the adequacy of the FE models was verified for representing the structural behavior of two transverse connections. The axial stress in the steel bars of the two specimens were extracted from the FE models as shown in Figure 8(b). Figure 8(b) indicates that the stress in steel bars gradually increases from girder ends to the diaphragm center, most of steel bars (except the two steel bars at the farthest of the long side) yielded at the ultimate state, and the stress in some steel bars was much higher than 50 ksi due to strain-hardening. Figure 8(c) shows the von Mises stress in steel girders of the two specimens. The steel yielded at the bottom flange in the vicinity of the edge of the diaphragm but most of the steel of the steel girder did not yield as shown in Figure 8(c), which verifies the conclusion that for the two specimens the critical section was located at the girder end within the diaphragm. The longitudinal compressive stress in the compression block was plotted in Figure 8(d). Figure 8(d) indicates that the compression block transferred significant amount of compression force at the bottom region of the diaphragm and the steel of the compression block did not yield under the ultimate load. And due to influence of the compression block, for Specimen I the maximum compression strain was located at the bottom of the concrete diaphragm, while for Specimen II the maximum compression strain in the concrete diaphragm was located in the vicinity of the steel girder bottom flange. For Specimen II, the neutral axis was located at 10.9 in. from the diaphragm bottom as shown in Figure 8(a). For conservative consideration, hand calculations of the ultimate capacity for Specimen II were based on the classic reinforced concrete design theory and an effective width equal to the bottom flange width. The hand calculations resulted in a compression depth 10.5 in. which was close to that predicted using the FE model (i.e., 10.9 in.). For Specimen I, the centroid of the compression force at the bottom region of the diaphragm was assumed at the bottom of the bottom flange of the steel girders. For both specimens, the tension force was estimated based on the yield strength of the steel bars. According to force equilibrium and the moment arm between the tension and compression forces, the ultimate moments for the two specimen were estimated and are summarized in Table 1. The crack moments for the two specimens were calculated based on the classic reinforced concrete beam design theory and are also summarized in Table 1. Additionally, the conventional design of the Little Silver Creek Bridge according to the AASHTO LRFD bridge design specifications (9), the bridge girder demand including the unfactored service moment and factored ultimate moment were calculated and are summarized in Table 1. The results summarized in Table 1 suggests that Specimen I with a compression block had higher yield and ultimate moment capacity than Specimen II. The FE models under-estimated the cracking moment, over-estimated the yielding moment, and predicted well the ultimate moment
Deng et al.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46
9
compared with the test results. The moment capacity of the two connections were reasonably predicted using hand calculations although the moment capacity were slightly under-estimated due to the fact that the strain hardening effects of steel were not taken into account. Concrete cracking in the deck of both specimens should occur under both the unfactored service and factored ultimate loads, but the two transverse connections are safe under the factored loads per the AASHTO LRFD bridge design specifications (9). Especially, due to the conservative design, Specimen I with a compression block can sustain the entire factored load prior to yield of the deck reinforcement. CONCLUSIONS Negative moment flexural strength tests were conducted to evaluate the behavior of the transverse closure pour connection used to longitudinally connect longitudinally adjacent prefabricated elements of the Little Silver Creek Bridge recently constructed in Pottawattamie County, IA. For comparison purposes, two specimens with and without a compression block were designed, fabricated, instrumented, and tested. FE models were also established to carry out a performance evaluation of the two specimens and to aid in the interpretation of the test results. Additionally, hand calculations were performed taking advantage of the classical reinforced concrete beam design theory and the FE results to estimate the moment capacity of the specimens. The following conclusions were drawn: • No significant difference was found between the two specimens in terms of the crack patterns. For both specimens, cracks were initially found on the deck top and near the edge of the diaphragm and ultimately propagated to the bottom region of the diaphragm. At the end of loading, the concrete of the diaphragm bottom tended to crush and no fracture was observed in the longitudinal steel bars. • The established FE models were sufficient at representing the structural behavior of the two transverse connection specimens. Overall, the FE models showed a higher stiffness and less ductility, under-estimated the cracking moments, and over-predicted the yielding moments. However, the FE models very well predicted the ultimate moment capacities of the two specimens. • The connection with a compression block has higher yield and ultimate moment capacity than the connection without a compression block. For both specimens, most of the longitudinal steel bars in the deck yielded at the ultimate state and the critical section was located at the girder end which was embedded within the diaphragm. The maximum compression strain was located at the bottom of the concrete diaphragm and in the vicinity of the steel girder bottom flange for connections with and with without a compression block, respectively. • To design both type of connections per the classic reinforced concrete design theory, an effective width equal to the girder bottom flange width can be assumed for the connection without a compression block and the centroid of the compression force at the bottom of the girder bottom flange can be assumed for the connection with a compression block. The hand calculations provide reasonable estimations on the moment capacity of the two connections. • The two transverse connections are safe under the factored loads per the AASHTO LRFD bridge design specifications. And the connection with a compression block can sustain the entire factored load prior to yield of the deck reinforcement.
Deng et al.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43
10
REFERENCES 1. Deng, Y., Phares, B., Putz A.J., Carter C., Michael N., and Dean B. (2016). Performance Evaluation of Longitudinal UHPC Closure Pour Connection for Use in Modular Bridge Construction – A Pairwise Comparison of Capacity and Ductility at Failure Limit State. Transportation Research Record: Journal of the Transportation Research Board, No. 2592, 10.3141/2592-08, In Press. 2. Azizinamini A. and Veen L.V. (2004). Bridges Made Easy. Roads & Bridges. http://www.roadsbridges.com/bridges-made-easy, November 08, 2004. 3. Azizinamini A., Yakel A., and Farimani M. (2005) Development of a Steel Bridge System Simple for Dead Load and Continuous for Live Load (Volume 1- Analysis and Recommendations). NDOR Research Project Number P542, University of Nebraska Lincoln, Lincoln, Nebraska. 4. Niroumand S.J. (2009). Resistance mechanism of simple-made-continuous connections in skew and non-skew steel girder bridges using conventional and accelerated types of construction. Ph.D. Thesis, University of Nebraska – Lincoln, Lincoln, Nebraska. 5. Talbot, J. (2005), Simple Made Continuous. Steel Bridge News, Vol. 6, No. 4, pp. 1, 4–5, National Steel Bridge Alliance. 6. van de Lindt, J. W., Stone, A. J., & Chen, S. (2008). Development of Steel Design Details and Selection Criteria for Cost-Effective and Innovative Steel Bridges in Colorado (No. CDOT-2008-12). 7. Hognestad, E. (1951). A study of combined bending and axial load in reinforced concrete members. Bull. Ser. No. 399, Univ. of lllinois Engrg. Experimental Station, Champaign, Ill, 1951. 8. Wee T. H., Chin M. S., and Mansur M. A. Stress-Strain Relationship of High-Strength Concrete in Compression. Journal of Materials in Civil Engineering, 1996, Vol. 8, No. 70, 70–76. 9. AASTHO. LRFD Bridge Design Specification, Customary U.S. Units, 7th Edition, with 2015 and 2016 Interim Revisions. Washington, D.C.: American Association of State Highway and Transportation Officials, 2015. 10. Ollgard, J. G., Slutter, R. G., and Fischer, J. W. Shear strength of stud connectors in lightweight and normal concrete. AISC Engineering Journal, 1971, 8, 55-64. 11. Deng, Y., Norton, T.R., & Tuan, C.Y. Numerical Analysis of Concrete-Filled Circular Steel Tubes. Structures and Buildings, Vol. 166, No. 1, 2013, pp. 3-14. 12. Deng, Y. & Morcous, G. Efficient Prestressed Concrete-Steel Composite Girder for Medium-Span Bridges: II - Finite Element Analysis and Experimental Investigation. Journal of Bridge Engineering, ASCE, Vol. 18, No. 12, 2013, pp. 1358-1372. 13. Deng, Y., Phares, B., and Owen, S.W. Experimental and Numerical Evaluation of a Folded Plate Girder System for Short-Span Bridges - A Case Study. Engineering Structures, 2016, Vol. 113, pp. 26-40. 14. Deng, Y., Phares, B, Dang, H., and Dahlberg J. Impact of Concrete Deck Removal on Horizontal Shear Capacity of Shear Connections. Journal of Bridge Engineering, ASCE, 2016, Vol. 21, No. 3, 04015059.
Deng et al.
1 2
11
TABLE 1 Summary of Results Bridge Girder Demand
Test Results
FE Predictions
Hand Calculations
Specimen Unfactored Factored No. Cracking Yielding Ultimate Cracking Yielding Ultimate Cracking Ultimate Service Ultimate Moment Moment Moment Moment Moment Moment Moment Moment Moment Moment (kip-ft) (kip-ft) (kip-ft) (kip-ft) (kip-ft) (kip-ft) (kip-ft) (kip-ft) (kip-ft) (kip-ft) Specimen 423 1456 2672 325 1729 2697 348 2136 I 816 1428 Specimen 358 1183 2434 325 1684 2482 348 2028 II
3
Note: All moments refer to the location at the diaphragm center.
Deng et al.
N
12
1 2
Specimen design
Longitudinal Joint
S
Prefabricated Module (a) Cross-section W
Longitudinal Joint
Transverse Joint
E
Specimen design
91'-0''
3 4
92'--0''
51'-0''
Prefabricated Module
(b) Plan view 10'-0"
Deck Concrete diaphragm
Steel girder
Shims
End steel plate
Steel block Pier
5 6 7 8
(c) Transverse Joint FIGURE 1 LITTLE SILVER CREEK BRIDGE ON IOWA 92
Deng et al.
13
2.75" 1.00" 1.00"
Ø0.75"
Bridge Deck
Diaphragm
#7 Bars
0.875" 0.875"
1'
#6 Bars
4.5"
5 8"
Ø0.625"
1'
Bearing Stiffener
4.5"
5e1 1'
W40x149 3.00"
W40x149
1.00"
#5 Bars
1' 5e3
4.00" 3.00" 6'-6.00"
1 2
1 2"
9.00" 1'
1'
Neoprene Bearing
N
Steel Plate
2'-9.00"
6'-6.00"
(a) Specimen I 2.75" 1.00" 1.00"
Ø0.75"
#7 Bars
0.875" 0.875"
1' 4.5"
Bridge Deck
Diaphragm
#6 Bars
5 8"
Ø0.625"
1'
Bearing Stiffener
4.5"
5e1 1'
W40x149 3.00"
W40x149
1.00"
#5 Bars
1' 5e3
4.00" 3.00"
3 4
6'-6.00"
1 2"
9.00" 1'
1'
Neoprene Bearing Steel Plate
2'-9.00"
6'-6.00"
(b) Specimen II 1'-8.00"
2'-3.00" 3'-11.00"
8.00"
1.50"
4.38"
Short side
Long side
3'-2.25"
5 6 7 8
11.75"
(c) Cross-section of composite beam FIGURE 2 SPECIMEN DETAILS (CONTINUED)
N
Deng et al.
14
4.50" 4.25" 5.50" 5.50"
1'
1'
1'
N
1.50" 1'
1'
1'
5.50" 5.50" 4.25"
1
*All bars are symmetrical on top and bottom of deck slab
2
3 4 5 6 7
(d) Rebar layout of deck
(e) Compression block FIGURE 2 SPECIMEN DETAILS
1.50"
4.50" 1.50"
Deng et al.
15
Diaphragm formwork
Compression block
Shear Studs
Bearing Stiffeners
1 2
(a) Steel girders
(b) Diaphragm formwork
Deck formwork Reinforcement
Steel Braces 3 4 5 6 7
(c) Deck formwork FIGURE 3 FABRICATION PROCEDURE
(d) Reinforcement
Deng et al.
16
Actuators Load Cells
Test Frames
Deck
Concrete Diaphragm
Steel Girder
Data Acquisition System
1 2
(a) Laboratory Test Setup 4.50" 4.25" 5.50" 5.50"
1'
1'
1'
3 4
1'
A15 (B15)
A10 (B10)
A14 (B14)
A9 (B9)
A4 (B4)
A8 (B8)
A3 (B3)
A7 (B7) A6 (B6)
A2 (B2)
A13 (B13)
5.50" 5.50" 4.25"
1.50"
N
A12 (B12) A11 (B11)
*All bars are symmetrical on top and bottom of deck slab
1'
1'
A5 (B5)
A1 (B1)
1.50"
Centerline of Girder
(b) Strain gages on steel bars
G1-A (G1-B)
G7-A (G7-B)
G2-A (G2-B)
G8-A (G8-B)
6.00"
6.00" G9-A (G9-B) G10-A (G10-B)
5 6 7 8
G3-A (G3-B)
G4-A (G4-B)
G5-A (G5-B)
G6-A (G6-B)
7'-6.00"
G11-A (G11-B) C1 & C2
G12-A (G12-B)
3'-9.00"
6.00" 7'-6.00"
(c) Strain gages on steel girders and compression block FIGURE 4 INSTRUMENTATION AND TEST SETUP
N
4.50" 1.50"
Deng et al.
17
Y
1 Loads
Deck Panel
Z
X
Steel girder
2
Diaphragm
Stiffener Support
3 4 5 6 7 8
FIGURE 5 FINITE ELEMENT MODEL OF SPECIMENS
Support
Symmetric plane
Deng et al.
18
First Crack First Crack Specimen II
Specimen I
1 2
(a) First crack
Specimen II
Specimen I
3 4
(b) Short side
Specimen I
5 6
Specimen II (c) Long side
Cracks
Specimen I
Cracks
Specimen II
7 8 9 10
(d) Diaphragm bottom FIGURE 6 CRACK PATTERNS IN SPECIMENS
Deng et al.
19
6 in. west of Diaphragm
1 2
Center of Diaphragm
(a) Strains in Steel Bars - Specimen I Specimen II
Specimen I
3 4
(b) Strain Comparison in steel bars at the diaphragm center Specimen I
5 6 7 8 9 10
Specimen II
(c) Strains in steel girders 6 in. away from diaphragm FIGURE 7 LOAD-STRAIN RELATIONSHIPS FOR DIFFERENT COMPONENTS OF SPECIMENS (CONTINUED)
Deng et al.
1 2 3 4 5 6
20
(d) Strain Comparison in Compression Block FIGURE 7 LOAD-STRAIN RELATIONSHIPS FOR DIFFERENT COMPONENTS OF SPECIMENS
Deng et al.
21
Specimen II
Specimen I
1 2
(a) Z-strain of concrete in the critical section Specimen II
Specimen I
3 4
(b) Axial strain in the steel bars Specimen II
Specimen I
5 6
(c) von Mises stress in the steel girder Y X Z
7 8 9 10
(d) Z-Stress contour in compression block FIGURE 8 STRAINS OR STRESS IN DIFFERENT COMPONENTS OF SPECIMENS
From: To: Subject: Date:
Mr. Hussam Fallaha
[email protected] TRB Paper 17-02665 Review Results Friday, September 30, 2016 5:24:00 AM
Thank you for submitting your paper for presentation at the Transportation Research Board 2017 Annual Meeting and/or for publication in the 2017 Transportation Research Record series (Journal of the Transportation Research Board). Yours was one of nearly 5800 papers submitted. Approximately half of these papers will be presented at the Annual Meeting, and about 20 percent will be published in the Journal of the Transportation Research Board. Your paper, number 17-02665, "PERFORMANCE EVALUATION OF TRANSVERSE HPC CLOSURE POUR CONNECTION FOR USE IN MODULAR BRIDGE CONSTRUCTION", submitted for Presentation and Publication, was reviewed by TRB's Standing Committee on Concrete Bridges (AFF30). Based on the peer review results, the committee is making the following recommendation to TRB. Presentation: The committee recommends that your paper be considered for presentation at the TRB Annual Meeting. Please note that papers submitted for presentation consideration may be eligible for placement in a poster or lectern session, a workshop, or a committee meeting. TRB will review this recommendation and will notify you in late October regarding whether they are able to schedule your paper presentation in the Annual Meeting program and include it in the Annual Meeting Compendium of Papers. Please note, inclusion of your paper in the Compendium of Papers does not assure subsequent publication of your paper in the Journal of the Transportation Research Board. You may revise your paper in light of the review comments below and submit it to the website (https://annualmeeting.mytrb.org/Paper/RevisedFileUpload/12401) between October 15 and November 15. The latest version of the paper will be included on the Annual Meeting Compendium of Papers. Publication: The committee recommends that your paper be considered for publication in the Journal of the Transportation Research Board. TRB will notify you if your paper is accepted for publication and request the final draft of the manuscript prepared in a different format. The final manuscript for publication must address the review comments. Submit your final manuscript only after hearing from TRB. The peer review results are summarized below. Please contact me if you have any questions. Sincerely, Hussam Fallaha Paper Review Coordinator Standing Committee on Concrete Bridges (AFF30)
[email protected]
------------------------------REVIEW RESULTS Averages based on a scale from 1-Poor to 5-Excellent. 1. Objectives appropriate and clearly stated: Average=4.3 2. Methodology technically sound: Average=4.7 3. Data valid: Average=4.7 4. Conclusions valid and properly supported: Average=4.3 5. Existing work adequately described and properly referenced: Average=4.0 6. Study effort adequately described: Average=4.7 7. Overall contribution to the state-of-the-art or practice: Average=4.3 8. Originality and timeliness: Average=4.3 9. Ready for implementation by practitioners (practice-ready): Reviewer 1 : Yes Reviewer 2 : No Reviewer 3 : No 10. Usefulness to researchers: Average=4.0 11. Long-term value as a research reference or description of practice: Average=4.0 12. Paper organization: Average=4.7 13. Abstract clearly conveys meaning of paper: Average=4.3
14. Well written and easily understood: Average=4.7 Comments for corresponding author: REVIEWER 1 The manuscript presents results from an experimental program where two specimens mimicking connection details for an existing bridge. The difference between the two specimens was in the existence of a compression block that helps transfer forces between bottom (compression) flanges at negative moment regions over intermediate supports. Finite element models were also built for both specimens. Experimental and numerical results obtained for both specimens were compared and conclusions were drawn. The presented research is interesting, however, the authors should address the following comments in the opinion of this reviewer: General Comments: ---------------------1. The title and keywords do not reflect that the manuscript deals with steel girder beams. 2. Abstract exceeds TRB limit of 250 words. Rewrite. 3. Continuous bridges such as the one described in this study are subjected to positive moments due to daily temperature variations, in particular thermal gradients. These positive moments can cause adverse cracking, which should be brought to the attention of the reader even though it is not the main focus of the study.
Technical Comments: ----------------------1. (p.7,l.5-7) Why was disabling concrete crushing needed if analyses were stopped when concrete strains reached 0.003? Elaborate. 2. (Fig. 6d) It is not clear what the nature of the diaphragm bottom cracks is? They seem to be vertical cracks, which is confusing since this area should be under compression. Elaborate. 3. (p.7,l.29) The nominal yield stress (60ksi) is just that; i.e., nominal. What is the actual yield strength, or where is the actual stress-strain relationship for the tested steel section, rebar and concrete? 4. (p.9,l.42-44) The last bullet in the conclusions is not substantiated in the body of the paper. Design moments for the bridge need to be provided before such a statement can be included.
REVIEWER 2
This is an excellent paper on a highly relevant topic. My comments are primarily editorial. 1. The authors should consider adding recommendations regarding the use/omission of the compression block. In my opinion this would move the paper in the "practice ready" category. 2. The conclusions were qualitative. It is suggested that quantitative comparisons be made. How much stronger was specimen I? To what degree did the models agree with the tests? 3. "Short side" in Figure 2 should be moved 4. Distance scales would be useful in Figure 8. 5. Figure 8b is confusing. Is the grey box from a screen shot or is it concrete? Suggest showing from the top view? 6. Confirm that figures will look OK when printed in black and white. 7. You should consider explicitly mentioning that the limit states you used to compare the FE models were based on theoretical strains (concrete crushing and steel yielding) not on tested properties. Your approach is reasonable because it is used for relative comparisons. 8. Review for typos and missed words. Such as lines 7 and 25 on page 7. 9. Page 5 lines 45 and 46. Revise awkwardly worded sentence. 10. Page 5 line 2: Use you "longitudinal" here in a manner that is inconsistent with the rest of the paper. Suggest rewording.
REVIEWER 3 Very well written and conceived paper. Excellent job. I have no major concerns as a conference paper and publication. I think more work could be done with the FEA to make this a TRR paper. -------------------------------
