Title:
Degradation of Shear Strength of Reinforced Concrete Members with Inelastic Cyclic Displacements
Author(s):
Dionysis E. Biskinis, George K. Roupakias, and Michael N. Fardis
Publication:
Structural Journal
Volume:
101
Issue:
6
Appears on pages(s):
773-783
Keywords:
column; deformation; loading; reinforced concrete; test
DOI:
10.14359/13452
Date:
11/1/2004
Abstract:
A database of 239 tests on reinforced concrete (RC) specimens of beams, columns with rectangular or circular sections, rectangular or barbelled walls, and hollow rectangular piers that failed in shear after initially yielding in flexure are used to fit two alternative models for the degradation of shear resistance, as controlled by transverse reinforcement with cyclic displacement ductility demand. Both models include a distinct, physically-based mechanism for the effect of axial compression on shear resistance and base the contribution of web reinforcement to shear strength Vs on the classical 45-degree truss analogy. Between the two models, the one that provides for degradation with cycling of both the web steel and concrete contributions—Vs and Vc , respectively—gives a slightly superior fit to the data. Although the data for large ductility demands are limited, they suggest that shear resistance does not degrade further beyond a displacement ductility limit demand of approximately 6. The fit to the data achieved by the two models is better than that of previous ones, which were fitted to much smaller sets of data of columns with either rectangular or circular sections. On the basis of the present, much larger, dataset, earlier models are found to give very good average fit for the type of cross section for which they were developed—albeit with larger scatter than that of the two models proposed herein—but significant bias for the other type of section. The sensitivity of shear resistance to cyclic ductility demand is not sufficient to allow inverting the proposed models to estimate the deformation capacity of RC members as controlled by shear after flexural yielding. The scatter associated with such an inversion renders such an exercise almost meaningless. An attempt to cast the present work within the framework of shear design advocated by present European codes for monotonic loads, namely that of a variable strut inclination approach without Vc , made it clear that such an approach is not well suited for design against seismic loads. For such loads, the classical 45-degree truss model supplemented with Vc that depends on cyclic displacement ductility lends itself better to the shear design of RC members. Using the results of 45 tests on squat walls and of another 44 tests on squat columns failing in shear by diagonal compression, empirical models are developed for the degradation of their shear strength—as controlled by diagonal compression—with cyclic displacement ductility demand after flexural yielding. The model for squat walls is based on the classical 45-degree truss analogy, while that for squat columns postulates a concrete compression strut along the column diagonal. According to both models shear strength due to diagonal compression increases with ÷fc¢ as in the ACI 318 limits for shear resistance before flexural yielding.