Experimental behaviour of reinforced concrete walls under earthquake loading

SUMMARY The experimental work and first results of a recently completed experimental research programme investigating the response of reinforced concrete (RC) walls under earthquake (EQ) loading are discussed in this paper. A brief literature review is given as a prelude to the outline of research objectives. The tests are presented in two groups according to the scale of models. For the 15 scale tests, a modified similitude relation for small scale reinforced concrete dynamic modelling is developed. Based on the chosen model parameters, the design of the isolated RC walls is given. The test-rig set-up and the EQ input signals suitable for testing the model on the Imperial College shake-table are also discussed. Preliminary observations regarding stiffness, strength and failure modes of the RC wall models are given. Experimental results from the shake-table are compared to tests, at the same scale, under static cyclic conditions. For the scale 1:2.5 cyclic tests a different test-rig assembly is designed. The test results are given in three pairs of flexurally similar walls followed with general observations and discussion. Finally, conclusions are drawn regarding experimental procedures and behaviour patterns of the tested models. Observations from field studies of earthquakes’-3 suggest that a level of drift control higher than that demonstrated by moment resisting frames is necessary in order to avoid excessive non-structural damage. This is shown by the extent of non-structural damage sustained by RC structures subjected to strong ground motion, primarily due to excessive storey displacements. Stiff shear resisting members such as ‘shear walls’ not only enhance the integrity of the load bearing and non-structural components but also may reduce damage to service installations. Present code requirements for earthquake-resistant design often underestimate the ductility of RC walls. This is manifested in the increased design base shear coefficient imposed on buildings with walls. The reason for this undue conservatism may be attributed to an attempt to avoid observed brittle modes of failure in walls designed in accordance with the code provisions for flexural behaviour. Reinforced concrete walls are also referred to as ‘shear’ walls because they resist a high proportion of the shear due to lateral loads. However, failures of RC walls are not necessarily dominated by shear deformations. The balance between shear and flexural loading has a very significant role in the deformational and strength characteristics. Walls with a shear ratio (M/ VL--M and V are applied moment and shear force respectively, L is total wall width) of more than about 1.5 possess flexure-dominated deformational characteristics and are termed ‘flexural’ walls. Walls with a shear ratio of less than 1.5 are referred to as ‘squat’ walls and are influenced more by the presence of high shear stresses. Even though RC walls in a building will be loaded in a complex manner according to the overall geometry, stiffness and other structural building characteristics, the highest forces are expected to occur at the lower floors. Ensuring the integrity and energy dissipation capacity of these critical portions of RC walls will lead to a safer EQ-resistant design.