COMPOSITE COMPOSIT E STEEL-FREE DECK BRIDGES: STATE-OF-THE-ART STATE-OF-THE-ART AND PROPOSED NUMERICAL MODELLING Adel H. Salem1, Mohamed A. El-Ag El-Aghoury houry 2 , Ezzeldin Y. Sayed-Ahmed 3, Tarek S. Moustafa 4 ABSTRACT
Steel-Free Deck Composite Bridges system has been investigated during the past two decades. The concept is totally new and innovative. The new structural system enables the construction of a concrete deck that is totally devoid of all internal steel reinforcement. Traditionally, reinforced concrete bridge decks are designed to sustain loads in flexure. The new innovative bridges with steel-free decks develop internal compressive forces “internal arching” which leads to failure by punching shear at substantially higher loads than the flexural design load. Five composite bridges have been recently constructed in Canada adopting this new concept. The deck slab is built using fibre reinforced concrete to control cracking due to shrinkage, creep and temperature variation effects. The deck is attached to the steel girders through flexible shear connectors. The steel girders are transversely tied together by steel steel straps straps and and cross cross frames frames.. In this this paper paper,, the bridg bridges es built built to-date to-date with with this this new new technology are presented. presente d. A 3-D 3- D finite element model model is adopted to study study the behaviour of each structural component of the bridge. The model accounts for both material and geometric non-linearities. The model is verified against experimental results performed earlier by others. The model and its verified results are discussed in this paper.
Key Words: Composite Bridge, Finite Element, Fibre Reinforced Concrete, SteelSteel- free Deck.
1
Professor Emeritus, Structural Engineering Dept., Faculty of Engineering, Ain Shams University, Cairo, Egypt Professor, Structural Engineering Depart., Faculty of Engineering, Ain Shams University, Cairo, Egypt 3 Associate Professor, Structural Engineering Dept., Faculty of Engineering, Ain Shams University,Cairo, Egypt 4 MSc. Student, Structural Engineering Dept, Faculty of Engineering, Ain Shams University, Cairo, Egypt 2
INTRODUCTION
Until early eighties, most of the highway composite bridges were designed according to the American Association of State Highway and Transportation Officials (AASHTO) specification. At this time, the AASHTO specification considers the deck slab of the bridge to behave in pure flexure (Batchelor, 1978) where the term deck slab describes the concrete slab of a composite girder. The Canadian Ministry of Transportation in Ontario sponsored extensive research towards the behaviour of composite bridges. The research was conducted with the help of both laboratory models and field-testing of bridges deck slabs. Internal arching action was found to be more significant than flexure behaviour of the concrete slab deck. The research confirmed that the amount of steel reinforcement in deck slabs could be reduced considerably by taking into account this internal arching action. The earlier edition of the Ontario Highway Bridge Design Code (OHBDC, 1979) permitted the design of deck slabs by an empirical method that is based on the behaviour of deck slabs in pure flexure. According to this specification, reinforcement is required in two meshes with each mesh consisting of orthogonal bars to resist flexure. The current edition of OHBDC (1992) considers the arching action of the deck slabs. It requires the slab to have a minimum thickness of 225 mm and a minimum reinforcement ratio of 0.3% in each direction in each mesh. The minimum thickness of 225 mm relates to the extra concrete cover required to protect the steel reinforcement from corrosion, which leads to concrete spalling exposing the steel reinforcement and resulting-in more corrosion of the steel. After a careful survey of a series of load tests, Agarwal (1990) concluded that the crack patterns on the soffits of deck slabs are independent of their method of design. For example, the crack patterns in a ten-year old, 190 mm thick, deck slab designed for bending are shown in Figure 1a. The slab is supported on steel plate girders spaced at 2.4 m. On the other hand, the crack patterns in another ten-year old, 190 mm thick, slab designed for arching are shown in Figure 1b. The slab is supported on prestressed girders spaced at 2.7 m. It is evident from Figure 1 that the crack patterns in the two slabs were virtually identical. This confirms the economy achieved by the empirical design method of OHBDC (1992) considering the arching action of deck slabs. Thus, the current AASHTO (1994) specification also permits an empirical design method for deck slabs, which takes account of the arching action, and is similar to the empirical method of OHBDC (1992).
a). Slab designed for flexure
b). Slab designed for arching
Figure 1. Crack patterns on the soffits of 10 years old 190 mm thick deck slabs In an effort to eliminate the source of corrosion, i.e. steel reinforcement bars, from the concrete deck slab, a steel-free deck slab was developed. This new system exploits the
internal arching action in the deck slab to such an extent that the slab does not require any tensile reinforcement (e.g. Mufti et al., 1993 and Bakht and Mufti, 1996). It was found that the arching action could be fully harnessed if the slab is confined adequately in both the longitudinal and transverse directions of the bridge (Bakht et al., 1998 and Bakht and Lam, 2000). In the transverse direction, restraining the relative lateral movements of adjacent girders can produce the required confinement for the deck slab. Figure 2 schematically shows this arching action concept and the transverse confinement, which is provided by welding steel straps to the top flanges of the steel girders. On the other hand, connecting the slab to the girders by means of suitable shear connectors provides the required confinement in the longitudinal direction.
Shear Studs
R/C slab designed for arching
Strap Composite Beams Figure 2. Schematic representation of the composite steel- free deck slab bridge system GENERATIONS OF DECK SLABS OF COMPOSITE BRIDGES
The riding surface in a composite girder bridge is usually the reinforced concrete slab. The First Generation of deck slabs (Figure 3) was customarily designed as flexural members (e.g. OHBDC 1979 and 1983). For this generation, if the girder spacing is 2.0 m, a concrete deck slab designed for flexure usually contains about 30 kg of steel reinforcement per m2 of the slab area.
Figure 3. First Generation deck slab
Research conducted about two decades ago identified the beneficial arching action in deck slabs of composite girder bridges. By taking advantage of this arching action, the amount of reinforcing steel can considerably be reduced in the deck slabs. This step uncovers the Second Generation of deck slabs (Figure 4). The OHBDC (1992) specifies an empirical design method for this generation of deck slabs. This method, which implicitly takes account of arching action in deck slabs, requires a minimum thickness of the slabs equal to 1/15 of the
girder spacing with a minimum thickness of 225 mm. The deck slab should contain two orthotropic meshes of steel reinforcement with a minimum reinforcement ratio in each direction in each mesh of 0.3%. Currently, hundreds of Second Generation deck slabs exist in composite bridges all over the world. For this generation, a girder spacing of up to 3.4 m yields a steel amount equals 20 kg per m2 of the slab area.
Figure 4. Second Generation deck slab
It is estimated that the empirical design method of the Second Generation of deck slabs, which considers the arching action, has resulted-in an annual saving in the cost of reinforcement placed in bridges of Ontario alone of about one million Canadian dollars. This technique was also adopted in other specifications (e.g. AASHTO, 1994). Research conducted over eight years in the Technical University of Nova Scotia, Canada, and the Ministry of Transportation of Ontario (e.g. Newhook et al. 1995) has led to a complete utilization of the arching action in deck slabs. It has been found that by harnessing the arching action suitably, the tensile reinforcement can be eliminated altogether from the deck slabs. The steel –free deck slabs can be regarded as the Third Generation of deck slabs (Mufti et al. 1993 and Bakht and Mufti, 1996). With the help of tests on large, and full- scale laboratory models, it was found that the arching action in the deck slab depends upon confinement in the longitudinal direction, i.e. in the direction of girders axes (Bakht et al. 1998), as well as in the transverse direction (Bakht and Ali, 1998 and Bakht and Lam, 2000). As mentioned earlier, in the longitudinal direction connecting the slab to the girders by means of suitable mechanical shear connecting devices provides the required confinement. On the other hand, restraining the deck slab supports (flanges of the steel beams) Steel-free Cruciform from lateral movement deck Strap normal to the beam axis Shear studs provides the transverse confinement (Figure 2). A third generation deck slab employing an alternative scheme for transverse confinement, which comprises a cruciform strap Composite with the cross bars laying grider loosely between the shear connectors, is shown in Figure 5. Figure 5. Third Generation deck slab The latest edition of the Canadian Highway Bridge Design Code (CHBDC, 1996) contains a new section dealing with fibre reinforced structures: within this section the Third Generation
deck slabs are recognized for the first time. The CHBDC technical Committee 6 (1996) and the CHBDC (1996) requires: 1. The deck slab thickness should not be less than 1/15 of the girder spacing with a minimum thickness of 175 mm. 2. The maximum spacing between girders composite with deck slabs is 3.7 m. 3. The top flanges of all the adjacent girders are transversely connected by an external confining system, comprising straps, with a minimum axial stiffness (MN/m) in the transverse direction of (135 x strap spacing) for outer panels and (100 x strap spacing) for inner panels. 4. The maximum spacing between the straps is 1/2 the girder spacing. 5. The strap connection to the girder must withstand the ultimate strength of the strap according to its yield stress and cross sectional area. 6. Composite diaphragms having adequate flexural rigidity in plane of the slab should support the transverse edges of the slab. The minimum flexural rigidity of this edge beam is 3.5 Lu4 (MN.m2 ), where Lu is the unsupported length of the edge beam. 7. The deck slab may have negative tensile reinforcement to resist moments resulting from loads on deck slab overhangs and load on railing and barrier walls. For bridge decks with overhangs or barrier walls, to satisfy the seventh requirement, a mesh of Glass Fibre Reinforced Polymers (GFRP) is usually provided as negative reinforcement. GFRP is commonly used due to its lower cost, easier handling and to hold the main concept of the steel-free deck slab. From their experimental investigation and based on the above requirements of the CHBDC (1996), Bakht et al. (1998) specified a minimum cross sectional area (A in mm) for each strap as: A
Fs S 2 Sl 109 ⋅
=
⋅
⋅
E t
(1)
⋅
Where Fs is a factor equals 5.0 for internal panels and 6.0 for external panels, S is the spacing between the girders in m, Sl is the spacing between the straps n i m, E is the modulus of elasticity in MPa and t is the thickness of the deck slab in mm. COMPOSITE BRIDGES WITH STEEL-FREE DECK SYSTEM CONSTRUCTED TO DATE
The third generation deck slabs is considered as an innovative bridge system. It enables the concrete bridge deck to be constructed totally devoid of all the internal steel reinforcement bars. The design of these structures was based on empirical models, which emanate from eight years of experimental investigations and tests (Newhook et al 1995, Bakht and Ali 1997, Mufti et al 1999, and Bakht and Lam 2000). Five bridges have been constructed in Canada adopting this new concept (Bakht et al. 1998 and Bakht and Mufti, 1998). The deck slab is built using fibre reinforced concrete to control cracking due to shrinkage, creep and temperature variation effects. The deck is attached to the steel girders through flexible shear connectors. The steel girders are transversely tied together by steel straps and cross frames. In the following section, the bridges built to date with this new technology are outlined.
Salmon River Bridge
The first steel-free deck-slab was cast on the Salmon River Bridge, part of the Trans Canada 104 Highway near Kemptown in Nova Scotia – Canada (Newhook and Mufti, 1996). Construction of the bridge, which consists of two, 31.2 m spans, includes a steelfree deck over one simple span and a conventional steel reinforced deck over the other. Figures 6 and 7 show a schematic layout of the bridge, the steel girders, details
Figure 6. The Salmon River Bridge: Schematic layout (above) and steel girders (below) showing shear studs and strap connection details
Figure 7. Cross section of the steel- free deck part of the Salmon River Bridge of shear studs typically used with steel-free deck composite bridges and half cross section of the steel-free deck part. Internal arching in the slabs helps in transferring the loads to the girders. The load is transferred from the deck to the supporting girders in the same way that an arch transfers loads to supporting columns. Steel straps are welded to the top flanges of the girders thereby resisting any lateral movement and tie the girders together producing the
required lateral confinement (Figure 7). With no steel inside the concrete, thinner deck could be designed. The Salmon River steel-free bridge deck has withstood a number of Canadian winters, and it appears to be defying the conventional approach of building steel-reinforced bridge decks. Newhook and Mufti (1996) and Newhook et al. (1996a,b) gave more details about the bridge design and construction. Field assessment of the bridge has been performed and outlined by Mufti et al. (1996). Chatham Bridge
In cooperation with ISIS (Intelligent Sensing for Innovative Structure) Canada, the Ontario Ministry of Transportation has rehabilitated the Chatham Bridge crossing Highway 401 to include the world’s second steel-free concrete deck slab (Figures 8 and 9). The roadway has
Figure 8. Half cross section of the Chatham Bridge. two lanes of traffic. The bridge has four spans of 13, 20, 20 and 13 m. It is composed of five girders spaced at 2.134 m. The two outer spans of the bridge have 175 mm thick steel-free deck slabs while the two inner spans have 225 mm thick steel-free deck GFRP Grid slabs. The slabs are transversely confined with steel straps welded to the top flanges of the steel girders. A grid of Glass Fibre Reinforced Polymer reinforces the 0.82 m cantilever portions of the deck slab. With no steel inside the bridge deck to Shear studs corrode, the concrete is immune to the damaging effects of chlorides. This in turn means that bridge decks can be virtually maintenance free. Mufti et al. Figure 9. Formwork for the Chatham Bridge (1997) and Ali et al. provided details about the design, construction and field assessment of the Chatham Bridge.
Crowchild Trail Bridge
Many bridges require upgrading because they were not built to handle the weight of today’s increased traffic loads. Calgary’s Crowchild Bridge is one such case. The project involved replacing the existing concrete superstructure with a new composite steel girder bridge (Tadros et al., 1998). The new 90-m long, 11-m wide bridge carries two lanes of traffic over its three continuous spans. For the bridge to carry the higher vehicle weight within the existing limited space, five steel plate girder spaced 2.0 m apart were used (Figures 10 and 11). Initially 225 mm deck slab was designed to Ontario Highway Bridge Design Code (OHBC). As an alternative, a 185 mm thick steel-free deck slab is used as shown in Figure 10. The deck slab itself is free of reinforcing; five steel girders with welded external steel
Figure 10. Half cross section of the Crowchild Bridge. straps and cross frames support it. Glass Fibre Reinforced Polymer (GFRP) C-bars were used to provide the continuity and to minimize the transverse cracks of the steel-free deck over the intermediate bridge piers. Based on the results of a full-scale model test at the University of Manitoba, glass FRP C bars were also used to reinforce the Strap cantilever slabs of the bridge. To reduce surface cracks, the bridge deck concrete contains short random polypropylene fibres. The bridge is also outfitted with remote monitoring technology: 81 strain gauges, 19 embedded gauges, five thermisters, three smart glass rebars and two fibre Figure 11. Straps welded with the shear optic gauges. Ashrami et al. (1998) connectors: Crowchild Bridge. discussed in details the field assessment of the Crowchild Bridge.
Waterloo Creek Bridge
Based on the Mufti et al. (1999) investigation, the Waterloo Creek Bridge was constructed. The bridge is part of a series of bridges being constructed by the British Columbia Ministry of Transportation and Highways as part of the new Vancouver Island highway, in the northern part of the island. The bridge consists of two separate single-span decks, one for the northbound lanes and one for the southbound lanes, with common abutments. The northbound structure has a steel-free deck, while the south has a conventional, reinforced concrete deck. Each deck is about 25 m long and 12 m wide. Figure 12 shows the 190 mm thick steel-free deck slab, which rests on 5 precast concrete girders spaced at 2.8 m. Transverse confinement of the steel-free deck was provided by means of studded straps spaced at 1.25 m. Fifty-three sensors were installed at various locations of the northbound structure and eleven sensors were installed in the southbound structure.
Figure 12. Half cross section of the Waterloo Creek Bridge Lindquist Creek Bridge
Forestry bridges in Canada are usually single-lane, single-span structure with two steel plate girder and a deck of precast reinforced concrete panels, which are made composite with the girder by means of clusters of studs (Figure 13). The deck panels are provided with circular
Figure 13. Cross section of the Lindquist Creek Bridge.
holes to accommodate the clusters of studs. The holes containing the studs are filed with a quick setting grout. The bridge can be opened to traffic within 24 hours after the erection of the steel works. The concept of arching action in deck slabs has led to an alternative to the reinforced concrete precast panels for such forestry bridges. The alternative, entirely devoid of tensile reinforcement, is illustrated in Figures 13 and 14 (Sargant et al., 1999). Panels with a 150 mm crown thickness were used for the Lindquist Bridge on a gravel forestry road in British Columbia. The spacing between the two steel girders is 3.5 m. The transverse confinement to the panels is provided by 25x50 mm studded steel straps at a spacing of 1.0 m. At their ends, the straps are embedded in the precast panels. Figure 14 shows the construction of this small bridge. In Table 1, a brief comparison between the previously mentioned five bridges is outlined.
Figure 14. Construction of the Lindquist Creek Bridge using precast panels.
Table 1. Steel- free deck composite bridges constructed to-date (Mufti and Bakht, 1999).
Salmon River
Girders, Spacing Steel plate
Slab Thickness 200 mm
Chatham
2.7m Steel plate
175 mm
Bridge
2.1 m
Crowchild Trail
Steel plate
185 mm
Waterloo Creek
2.0 m Precast concrete
190 mm (Crown)
Lindquist
2.8 m Steel plate
150 mm
3.5 m
Features First steel-free deck slab in new construction. Transverse Confinement: welded steel straps. 6% more expensive than conventional slab. First steel-free deck slab used in rehabilitation. Transverse confinement: welded steel straps. GFRP for transverse negative bending. Significantly more expensive than conventional slab. First steel-free deck slab on continuous span. Transverse confinement: studded straps. (GFRP) for transverse negative moments. First steel-free deck slab on precast concrete girders. Transverse confinement: studded straps Nearly the same coast as conventional slab. First steel-free deck precast panel Transverse confinement: studded straps embedded in the precast panel Record girder spacing to minimum thickness ratio, of 23.3. 30% cheaper than conventional panel.
NUMERICAL MODELLING OF COMPOSITE STEEL-FREE DECK BRIDGE SYSTEMS
In order to investigate the behaviour of the new composite bridge system, an expensive, time consuming and elaborate experimental analyses are required. On the contrary, if a suitable and trustable finite element model is developed, a wider parametric investigation may be performed. In the same time, a cut-down on the cost and the time will be achieved using this numerical model. A finite element model is developed and proposed here for the analysis of the steel free-deck bridge systems. The model results are verified against results from the experimental investigation performed earlier by Mufti et al (1993). Experimental Investigation Used for Finite Element Model Verification
Mufti et al. (1993) tested to failure four half-scale composite bridge superstructures (Figures 15 to 17). The bridge decks composed of cast-in-place steel-free fibre concrete, supported on steel girders. The dimensions of the 100 mm thick concrete deck are 3.66 m x 2.13 m for the first three test series and 3.66 x 3.19 m for the fourth series. A hydraulic jack bearing on a rectangular 127 mm x 254mm steel plate and rubber bearing pad applied a concentrated load at the centre of the deck slab according to the configurations shown in Figures 15 to 17. These loading configurations represent a half scale model of the truck tire footprint. 1
Hydraulic jack
Hydraulic jack
Steel plate
Neoprene pad
Steel plate Neoprene pad 100 mm
C 200x17 530 mm
W 460x82
1067 mm 2127 mm 1
530 mm
C 200x17
W 460x82 915 mm
915 mm
915 mm
915 mm
3660 mm Sec 1-1
Figure 15. Schematic presentation of Mufti et al. (1993) test series 1,2 and 3. In the first test series three intermediate diaphragms connected the steel girders with no diaphragms existing near the supports (Figure 15). The model failed under a concentrated load of 173 kN: the mode of failure was not that of pure flexure, nor did it conform to the punching shear type of failure. In the second series, Mufti et al. (1993) used diaphragms at the supports. The model failed at 222 kN in practically the same mode as that as the first model. Mufti et al. (1993) provided no steel straps between the steel girders in series 1 and 2. Thus, there is a lack of transverse confinement for the deck slab in those two series. For the third series, diaphragms were made even more substantial. Eight steel straps, spaced at 457 mm centre to centre, were also welded to the underside of the top flanges of the steel
girder. This diaphragm and straps configuration is very close to the ideal lateral restraint, which would permit the development of the maximum compressive stress during the development of the C.L. Transverse free edge internal arching mechanism. The first Girder support Steel girder test on the third series 530 resulted-in a failure by punching shear at a load of 4l8 KN. The angle of the shear cone was Location 1 Location 2 Location 3 1067 close to the typical C.L. 257 value of 30o . Two other tests were conducted on 914 457 127 the third series to replicate punching failure at different load Strap: 64x10 mm 530 positions (Figure 16). Steel-free deck slab @ 457 mm As the load position approached the edge of 1830 1830 the deck slab, the mode of failure changed to a hybrid mode of flexural Figure 16. Loading configuration of series 3. and punching failure 1 Hydraulic jack due to the lack of confinement in the Steel plate Neoprene pad longitudinal direction. Mufti et al used the fourth test series (Figure 17) to investigate the ability of the steel-free deck slab to sustain a pair of concentrated loads, which straddle the girder and cause tensile stress in the concrete above it. The crosssectional area of the straps, the strap spacing and the loading plate dimensions were identical to the previous three models. The series resulted in simultaneous punching shear failure under the two loads, with each loading pad carrying 4l8 kN. Thus, punching failure of
100 mm 64x10 mm strap C 200x17 W 460x82 530 mm
1067 mm
1067 mm
530 mm
3194 mm 1 Location 3
Location 1
Location 2
457
C 200x17
W 460x82
915 mm
915 mm
915 mm
915 mm
3660 mm Sec 1-1
Figure 17. Schematic presentation of test series 4
multi-girder steel-free deck bridges subjected to multi-point loads is possible as long as adequate lateral confinement is provided. Subsequent tests performed on this series with the load closer to the transverse free edge yielded a reduced ultimate load due to the lack of longitudinal confinement (Mufti at al. 1993). The Proposed Finite Element Model
The proposed finite element model employs four types of elements. The first type is 4-node shell elements with 6 degrees of freedom per node. The second element type is 2-node beam elements with 6 degrees of freedom per node. The third element type is 2-node truss elements with 3 degrees of freedom per node. The fourth element type is 8-node continuum solid elements with 3 degrees of freedom per node. Shell elements were used to model the steel girders while continuum solid elements model the steel-free concrete deck. The beam elements and the truss elements are used to model the cross frames/diaphragms and the straps respectively. MultiApplied Load 4352 solid elements point constraint for the concrete dec k equations are introduced to account for the 18 truss elements for the steel stra s missing degrees of freedom (rotation) between shell/beam elements and solid/truss elements. The model considers both the material and geometric nonlinearties of the 40 beam elements for concrete deck, the the steel diaphragms steel girders, the 2304 shell elements straps and the cross for the steel beams frames/diaphragms. Diaphagms A typical model used to analyse series 4 of Mufti et Finite element model with al (1993) tests is the concrete deck removed shown in Figure 18. The concrete deck was modelled using an initial elastic modulus of 43 GPa and Poisson’s ratio of 0.2. The nonlinear behaviour of the concrete is Steel considered based on bems an assumed relation Straps for the axial stressFigure 18. Proposed finite element model of Series 4 tests strain of concrete in performed by Mufti et al (1993)
compression. An elastic-perfectly plastic model was used for the material behaviour of the steel girders, the straps and the diaphragms. All of them have an initial elastic modulus of 200 GPa, Poisson’s ratio of 0.3 and yield strength of 300 MPa. The thickness of the concrete deck is identical to that used by Mufti et al (1993) and equals to 100 mm. The strap cross sectional area is 640 mm2 and the spacing between them is 457 mm. The finite element package COSMOS/M is used for pre-processing of the model, equation solution of the finite element technique and post-processing of the model results. Verification of the Proposed Finite Element Model
The developed finite element model is verified against results experimentally obtained from test series 3 and 4 (Mufti et al., 1993). In order to investigate the behaviour of the model throughout the loading history, the load displacement curves shown in Figures 19 to 21 are introduced and compared to those obtained from the experimental tests executed by Mufti et al (1993). Series 4 with Load Location 1 and Series 3 with Load Locations 1 and 3 are shown as typical examples of the verification analyses in Figures 19 to 21 (refer to Figures 15 to 17 for Loading locations of each series). The figures show a good correlation between the finite element model results and the test data.
450.0 400.0
Finite element model results
350.0 N300.0 k n i d 250.0 a o L d 200.0 e i l p p A150.0
Experimental results of Series 3 - load location 1 (Mufti et al ., 1993)
100.0 50.0 0.0 0.0
1.0
2.0
3.0
4.0
5.0
6.0
7.0
Displacement under the load in mm
Figure 19. Verification of the finite element model results for Series 3 at Load Location 1
450.0 400.0 350.0 Finite element model results
N300.0 k n i d 250.0 a o L 200.0 d e i l p 150.0 p A
Experimental results of Series 4 - load location 1 (Mufti et al ., 1993)
100.0 50.0 0.0 0.0
1.0
2.0
3.0
4.0
5.0
6.0
7.0
Displacement under the load in mm Figure 20. Verification of the finite element model results for Series 4 at Load Location 1
250.0 Experimental results of Series 3 - load location 3 (Mufti et al ., 1993)
200.0 N k n i
150.0
d a o L d e i l 100.0 p p A
Finite element model results
50.0
0.0 0.0
1.0
2.0
3.0
4.0
5.0
Displacement under the load in mm Figure 21. Verification of the finite element model results for Series 3 at Load Location 3
In Table 2, a comparison is also made between the results obtained from the proposed finite element model and the experimental analyses of Series 3 and 4 with loading location 1 and Series 3 only with loading location 3 (refer to Figures 15 to 17 for Loading locations 1 and 3). It is evident from Table 2 that the numerical analysis can predict both the failure load and the displacement of the new system with acceptable accuracy.
Table 2. Numerical model results versus experimental result (Mufti et al 1993). Load Series Series no. 3 Location 1 Series no. 4 Location 1 Series no. 3 Location 3
Finite element Test Results Finite element Test Results Finite element Test Results
Failure Load (KN) 406 418 406 418 199 204
FE to Test Results 97.1 % 97.1 % 97.5 %
Deformation Displacement FE to under the load Test Results at Failure (mm) 5.75 96.6 % 5.9 6.45 94.8 % 6.8 4.55 96.4 % 4.72
SUMMARY
Research on the steel-free deck slabs composite with steel girders proved that the internal arching action in the deck slab could be harnessed if adequate transverse and longitudinal confinements are provided to these slabs. Transverse confinement is achieved by connecting the girders upper flanges by steel straps while longitudinal confinement can be achieved by using adequate shear studs and cross frames and/or diaphragms. Thus, a new innovative bridge system is born: steel-free deck composite bridges. The bridges constructed during the last decade employing this innovative deck were outlined in this paper. Furthermore, a numerical model based on the finite element technique has been introduced. The model is verified against the experimental investigations performed by others on the new system. The model will be used in the future to perform a wide parametric study on steel-free deck composite bridges to verify the currently used empirical design equations and to study the general behaviour and failure criteria of the new system of bridges. REFERENCES
Ali, A., Bakht, B., and Schaefer, J. 1997. Design ad Construction of a Steel-free Deck Slab In Ontario. Proceedings, Annual Conference of the Canadian Society for Civil Engineering, Sherbrooke, Quebec, Canada. Ashrami, S., Cheng, J.J.R., and Alexander, S.D.B. 1998. Field Measurements and Assessment of the Crowchild Bridge. ISIS Alberta Workshop: Design and Rehabilitation of Structures Using Fibre Reinforced Polymers, Edmonton Alberta, Canada. AASHTO, American Association of State Highway and Transportation Officials 1994, LRFD Bridge Design Specification, Washington, D.C., USA, 1994.
Agrawal, A.C. 1990. Load Testing of New Concrete Bridge Deck Slabs. Developments in Short and Medium Span Bridge Engineering CSCE, Montreal, Canada, 1990. ,
Bakht, B. and Ali, A. 1997. Testing in Isolation of Transverse Confining Systems for Steelfree Deck Slabs. Canadian Journal Of Civil Engineering. Vol. 25, No. 3, pp. 789-796. Bakht, B., and Lam, C. 2000. Behaviour of Transverse Confining System for Steel-free Deck Slabs. Journal of Bridge Engineering, ASCE, Vol. 5, No. 2, pp. 139-147. Bakht, B., and Mufti, A.A. 1996. FRC deck Slabs without Tensile Reinforcement. Concrete International Vol 18, No. 2, pp. 514-530. .
Bakht B., Mufti, A.A., 1998. Five Steel-Free Bridge Deck Slabs in Canada. Structural Engineering International, IABSE, Vol. 8, No 3, pp. 196-200. Bakht, B., Mufti, A.A., Jaeger, L. G. 1998 Design Provision for Steel-Free Deck Slabs. Proceeding, Fifth International Conference on Short and Medium Span Bridges, Calgary, Canada, October, 1998. CD proceedings. Batchelor, H. and Csagoly, P. F. 1978. Investigation of the Ultimate Strength of Deck Slabs of Composite Steel/Concrete Bridges. TRR Record NO 644. Transportation Research Board. pp. 162-170. .
CHBDC 1996. Canadian Highway Bridge Design Code. Canadian Standard Association, Raxdale, Ontario, Canada. CHBDC Technical Committee 6, 1996. Design Provision for Fibre Reinforced Structures in the Canadian Highway Bridge Design Code. Second International Conference on Advanced Composite Materials in Bridges and Structures, Montreal, Canada. pp. 767776. Mufti, A.A., and Bakht, B. 1999. Innovative Deck Slabs for Highway and Forestry Bridges. IABSE Symposium in Rio de Janeiro, Paper 1026. Mufti A.A., Jaeger, L.G., and Bakht, B. 1997. Field Performance of Steel-free Deck Slab of Girder Bridges. Proceedings, US-Canada-Europe Workshop on Bridge Engineering, Zurich, Switzerland, pp. 236-246. Mufti A.A., Jaeger, L.G., Bakht, B. and Wegner, L.D. 1993. Experimental Investigation of FRC Slabs without Internal Steel Reinforcement. Canadian Journal of Civil Engineering, Vol. 20, No. 3. pp.398-406. Mufti, A.A., Newhook, J.P., Jalali, S.J. 1999. Investigation of the Flange Splitting Stresses in Composite Bridge Beams. Proceedings, Canadian Society for Civil Engineering Annual Conference, General Conference, Structural Engineering, Vol. 1, pp. 41-50. Mufti, A.A., Newhook, J.P., Mahoney, M.A. 1999. Salmon River Bridge field Assessment. Proceedings, Canadian Society for Civil Engineering Annual Conference, General Conference, Structural Engineering, Vol. 1 pp. 51-61.
Newhook, J.P., Mufti, A.A., and Wegner, L.D. 1995. Fibre Reinforced Concrete Deck Slabs Without Steel Reinforcement: Half Scale Testing and Mathematical Formulation. Research Report No. 1-1995. Department of Civil Engineering, Technical University of Nova Scotia, Halifax, Nova Scotia, Canada. Newhook, J.P. and Mufti, A.A. 1996. A Reinforcing Steel-free Deck Slab for the Salmon River Bridge. Concrete International. ACI, Vol. 18, No. 6, pp. 30-34. Newhook, J.P., Mufti, A.A., and Jaeger, L.G. 1996a. Steel-free Concrete Deck- the Salmon River Project – Design and Construction. Proceedings, 1 st Structural Speciality Conference, Canadian Society for Civil Engineering, Edmonton, Alberta, Canada. pp. 203-214. Newhook, J.P., Mufti, A.A., and Jaeger, L.G. 1996b. Steel-free Concrete Deck- the Salmon River Project – Experimental Verification. Proceedings, 1 st Structural speciality Conference, Canadian Society for Civil Engineering, Edmonton, Alberta, Canada. pp. 123-133. Sargant, D.D., Mufti, A.A., and Bakht, B. 1999. Design, Construction and Field Testing of Steel-free Arch Panel Bridge Deck for Forestry Bridges. Proceedings, Canadian Society for Civil Engineering Annual Conference, General Conference, Structural Engineering, Vol. 1 pp. 95-104. OHBDC Ontario Highway Bridge Design Code 1979, 1983 and 1992. Transportation of Ontario, Downsview, Ontario, Canada.
Ministry of
Tadros, G., Tromposch, E., and Mufti, A.A. 1998. Superstructure Replacement of the Crowchild Trail Bridge. Proceedings, Development in Short and Medium Span Bridge Engineering, Advanced Materials in Bridges, Calgary, Alberta, Canada. CD-Proceedings.
