Diagonal cracking in reinforced concrete deep beams : an experimental investigation

Concrete deep beams with a shear span to depth ratio of less than 2.32 [73] will work as tied arches after flexural cracking, provided there is sufficient reinforcement. The compression strut formed between the support and the loading points is under biaxial compressive and tensile stresses. The current Canadian Code [5] stipulates that deep beams and corbels should be designed using the Strut-and-Tie Method. This method incorporates the work done by Collins and Mitchell [8][9] where the cracked concrete behaves as a new material and that the compressive strength of concrete is reduced due to strain-softening. Here-in lies an area of discrepancy. The work done by Collins and Mitchell utilizes beam theory which requires that plane sections remain plane. However, deep beams and corbels are classified as "regions of discontinuity" consequently beam theory does not apply to these structures. An area of the Canadian code which needs to be examined is the dimensioning of the compression strut. To date there is no clear explanation as to how the design guidelines of the compression strut were developed. A weakness of the design code is that numerous assumptions must be made. The designer first assumes that the compression strut reaches a maximum concrete strain of 0.002 [5] , and then must assume the strains in the tension ties. The focus of this research has been to investigate diagonal splitting strength of reinforced concrete deep beams. In conducting this study, twelve deep beams, categorized in four groups were tested. The test variables included the shear span, the amount of web reinforcement and the concrete compressive strength. Surprisingly, no researcher has published measured strain incurred by the compression strut in deep beams. In our research, a single beam from each of the four test groups was fitted with strain gauges to measure the tensile strain in the main tensile reinforcement. As well, the concrete strains along the main diagonal formed between the support and the loading points as well as perpendicular to the strut were measured. The experimental work demonstrated the development of diagonal cracking. These cracks appeared above the supports and propagated towards the loading points. The strain gauges on the concrete surface confirmed that the stresses along the compression strut were under biaxial compression tension stresses. A finite element analysis determined that the compression stress acting parallel to the diagonal were uniform in distribution and symmetrical. Perpendicular to the diagonal, high compressive stresses were seen at the supports and the loading points. However, the stresses in between these areas were uniformly distributed in tension. The measured compressive strains were much less than the recommended value of 0.002, and the compression strut was found to be much wider than that defined by the Canadian Code. As a consequence of these the findings, a truss model was defined using a biaxial concrete strength envelope. This truss model was applied to the test beams of this study as well as too ninety-nine test beams available in literature. In all cases, the truss model was able to accurately predict the strength of these test beams.