Earthquakes worldwide have clearly demonstrated the vulnerability of reinforced concrete members to degradation in shear strength when subjected to cyclic loading. Such degradation can lead to significant damage to the structure and, possibly, even collapse. With the advancement of performance-based earthquake engineering, where the response of the structure must be traced through all levels of damage, there is a significant need to accurately define the deformation capacity and shear strength for such members. This symposium publication represents an effort from researchers across the globe trying to address this challenging problem. Although at the time of publication there are some methodologies that can be used in performance-based earthquake engineering, there is a significant need for improved methods better suited for these types of applications. Furthermore, one of the concerns often expressed by researchers is that test data used in the past to develop and calibrate existing models consisted of relatively small data sets. This problem is compounded by differences between experimental studies in aspects such as the type of load history used, the manner in which deformations were recorded during tests, and the definition of displacement and strength at failure. The recent development of the PEER column database, hosted by the University of Washington, provided a valuable resource to overcome some of these problems. It presented researchers with a larger pool of data, which included the full hysteretic response of every column in the data set. Although this represented a very significant step forward, efforts of this kind should continue to improve the ability of researchers to calibrate and evaluate models for shear strength and deformation capacity. A joint technical session was organized by Joint ACI-ASCE Committees 441, Reinforced Concrete Columns, and 445, Shear and Torsion, during the American Concrete Institute’s Fall 2004 Convention in San Francisco, CA. The goal of the technical session was to showcase recent developments in this area, with the hope that continued discussion will lead to improved models that are suitable for performance-based engineering. Note: The individual papers are also available. Please click on the following link to view the papers available, or call 248.848.3800 to order. SP236

The new provisions in the 2004 Canadian code for flexural displacement capacity of concrete walls, and the new provisions for seismic shear design of slender concrete walls are presented. To facilitate explanation of the seismic shear provisions, general expressions for shear design are first presented, and the non-seismic shear design provisions in the Canadian and ACI 318 building codes are briefly reviewed. According to the new seismic shear design provisions presented here, the maximum shear force and concrete contribution depend on the inelastic rotation demand in the plastic hinge, and the compression stress (critical crack) angle used to determine the quantity of horizontal reinforcement depends on the axial compression stress applied on the wall. The 2004 Canadian code provisions generally require more horizontal reinforcement than the ACI 318 provisions except when inelastic rotational demand is small and axial compression stress is large; however, the Canadian provisions permit significantly higher shear stress for high-strength concrete walls. The new provisions can be used to design concrete walls given the expected level of drift demand or, as demonstrated in this paper, can be used to estimate drift capacity of walls accounting for the significant influence of shear.

The sequence of failure in reinforced concrete (RC) prismatic members is used as a tool in estimating dependable deformation capacity. Response mechanisms that may limit the response leading to damage localization are identified (web diagonal cracking, bar buckling, disintegration of compressive struts due to load reversal, and anchorage failure of primary reinforcement). Deformation components are additive only if stable hysteretic response controlled by flexure prevails. In all other cases, the deformation component associated with the controlling mode of failure dominates the overall deformability of the member. Because the sequence of failure depends to a large extent on load history, deformation attained at any particular level of load is also load history dependent. This is why experimental values for deformation capacity reported in international literature are characterized by excessive scatter. The proposed methodology is applied to a number of published column tests. Analytical estimates are evaluated through comparisons with experimental results and by parameter studies conducted in order to examine the sensitivity of the estimated displacement limit at compression bar buckling to important design variables.

A probabilistic capacity model for the drift at shear failure for reinforced concrete columns with light transverse reinforcement is developed using a Bayesian methodology. Both the aleatory and epistemic uncertainties are properly incorporated in the probabilistic model. During the model formulation the key parameters influencing the drift capacity at shear failure are identified as the shear stress, the axial load ratio, the ratio of the spacing of the transverse reinforcement to the effective depth, and the aspect ratio. The drift capacity model is employed to formulate a fragility curve, with confidence bounds, for a column damaged during the Northridge Earthquake. Fragility curves developed for a range of parameters suggest that the spacing of the transverse reinforcement is the most important parameter in the determination of the probability of sustaining shear failure at a given drift demand.

Six nominally identical reinforced concrete columns were subjected to a variety of lateral displacement histories to evaluate the effects of cycling on their failure displacement and failure mechanism. The columns, typical of bridges constructed before the mid-1970s, had circular cross-sections, low axial loads, and little transverse reinforcement. Shear failure caused five of the six columns to lose their axial load carrying ability at drift ratios between 3% and 5%. The sixth column failed in an axial-flexure mode at a drift ratio of 6%. Increasing the number of cycles at each displacement level from one to fifteen decreased the maximum displacement preceding flexure-shear failure by approximately 35%. The effect of cycling on damage accumulation was modeled with the Park-Ang damage model, a Modified Park-Ang damage model, and a Cumulative Plastic Deformation damage model. The Cumulative Plastic Deformation model correlated best with the observed damage, and it was the easiest to implement.