Experimental And Finite Element Based Investigations Of Shear Behavior In Reinforced Concrete Box Culverts

Abstract

This study evaluates the shear behavior and capacity of the precast concrete box culverts subjected to HS 20 truck wheel load. The most critical culvert behavior was considered by studying culverts subjected to zero depth fill and placed on rigid laboratory floor without bedding material. Three major phases were considered to complete the study, which included: (1) experimental program; (2) finite element modeling; and (3) development of distribution width and the determination of shear capacity.

A full-scale experimental testing program was undertaken to perform tests on 24 typical precast concrete box culverts designated as per ASTM C 1433-05 with and without distribution steel (As6) which were produced by two different precast manufacturers. The wheel load was simulated by a 25 cm x 51 cm (10 in x 20 in) load plate, which was placed, at the distance "d" from the tip of the haunch to the edge of the load plate. An incremental loading history was adopted to capture the culvert's nonlinear behavior. The test results include load-deflection plots as well as the step-by-step description of the events. The test results further indicated that flexure governed the behavior up to and beyond AASHTO 2005 factored load. Independent shear cracks formed before the ultimate load at approximately twice the AASHTO 2005 factored load for most of the test specimens. Complete detailed three-dimensional finite element models (FEM) of the test specimens were developed and analyzed to simulate the experimental results. Three- dimensional shell and solid elements were used to model the culvert systems. The welded wire fabrics were modeled by using the rebar elements placed on the surface-elements provided by the ABAQUS Software. The contact surface between the outside face of the bottom slab and reaction floor was modeled by using non-linear node-to-surface contact algorithm. The analysis algorithm consisted of an incremental loading history to capture the problem non-linearity. Smeared crack model along with the Risk Algorithm were incorporated to analyze the system for micro-cracks and to stabilize the solution, respectively. The converged solution was obtained by using H-convergence coupled with the difference between the external work done and the strain energy density of the system. The load-deflection plots obtained from the FEM analyses were compared with those obtained from the experimental results, which showed close correlation. All the forty-two standard cases of the ASTM C 1433 were modeled using the verified FEM model developed in Chapter 3. A regression analysis was conducted to develop equations for the calibrated FEM parameters. The 3-D volumetric shear force distribution on the top slab of the boxes was obtained. The peak shear force in each of the plots was identified and a vertical plane was passed through it parallel to the box's joint length. This yielded to a 2-D shear force distribution diagram along the box joint length from which the distribution width was calculated by dividing the area under the 2-D diagram with the peak shear force. The distribution width for the ASTM C1433 boxes were compared with those reported in the AASHTO 2005. The calculated values of the distribution width were used to calculate the critical factored shear force for all the boxes, which were then compared with the American Concrete Institute (ACI) shear capacity equations. It was shown that the shear capacity exceeded the factored critical shear force for all the ASTM C1433 boxes. This study concludes that the AASHTO 2005 provision with regard to the shear transfer device across the joint is unsupported and it needs to be revisited.