EARTHQUAKE PROTECTION OF MEDIUM-RISE REINFORCED CONCRETE BUILDINGS BY BASE ISOLATION

Abstract

Earthquake protection by base isolation of buildings has attracted considerable attention in recent years. The main concept here is to isolate the structures from ground instead of the conventional techniques of strengthening the structural members. This new design methodology appears to have considerable potential in preventing damages to the structures and non-structural elements. Loose contents in the building are also protected. In the present study, earthquake protection of medium rise reinforced concrete framed building by base isolation has been studied - both analytically and experimentally. Model laminated rubber bearing (LRB) has been designed for seismic isolation of a three storeyed r.c. framed building during Shake Table test. Analytical studies are carried out to asses the suitability of pure friction isolator (P-F), lead rubber bearing (LLRB), siiding-elastomer bearing (EDF) for seismic isolation of medium rise r.c. framed buildings. Stability theories of LRB, proposed by different investigators, have been studied in connection with design of model LRB. Static testing of model bearing has been carried out to determine bearing parameters. Compressive test has been carried out to determine the vertical stiffness of the model bearing, which should have a very large value to avoid rocking and other unwanted modes of vibration. Shear test of model bearing under reversible lateral load shows that shear forcedisplacement relationship is non-linear in nature. Shake Table Tests of the three storeyed base isolated l/6th scale model has been carried out to assess the effectiveness of the base isolation in controlling the response of the superstructure. Additional loads are attached to each floor level for gravity load simulation. Simulated earthquake motions generated from a time scaled average spectra for alluvial soil, have been used as the input table acceleration histories. The model did not suffered any damage, even when it was subjected to high peak table acceleration of the order of 0.55g. Higher modes contribution in the seismic response of isolated structure are effectively filtered out by model bearing. Relative performances of P-F bearing, LRB, LLRB and EDF bearing in controlling the response of a three storeyed r.c. structure, subjected to unidirectional earthquake motion is studied. Geometry and post yielding stiffness of all LRB based isolation system are kept same, while coefficient of friction in P-F and EDF isolators are considered to be 0.1. The bilinear hysteretic behaviour of LRB and lead rubber bearing are represented by equivalent linear stiffness and damping factor. Frictional force developed at the sliding interface is modelled by rigid plastic model. Superstructure is idealized as a rigid body and flexible model to understand the effects of flexibility of superstructure on overall response of base isolated buildings. Koyna (Long., 1967) and El-Centro (N-S, 1940) accelerogram is used as input excitations to understand the behaviour of base isolated building subjected to earthquakes with different characteristics. A unified solution algorithm has been developed for analysis of base isolated building supported over selected isolation systems, based on Newmark's method in predictorcorrector form. Effects of superstructure flexibility on the base displacement for LRB, LLRB and EDF isolation systems are not significant, although acceleration response increases slightly, when the flexibility of superstructure has been taken into account. But, both base displacement and acceleration response for P-F bearing are largely influenced by flexibility of the superstructure. It is observed that base isolation technique is more effective in controlling the response of the structure for earthquake excitation with most of its energy contents in high frequency range. Fourier decomposition of roof acceleration for both excitations show that LRB acts as low pass filter and higher modes contribution is lowest for this system as compared to LLRB and EDF bearing. P-F is not able to filter out high frequency contribution in the response. When isolated structure experiences multidirectional motion due to asymmetry in the structure and/or due to multidirectional excitation, it becomes very difficult to compute the response of the base isolated structure by modelling the hysteretic behaviour of LRB based isolation systems by bilinear model and that of sliding systems by rigid plastic model. In the present study, hysteretic non-linear model developed by Wen and modified visco-plastic model developed by Constantinou el al. IV have been considered for modelling different types of isolation system under both unidirectional and bidirectional motions. Close form solution of stiff differential equation of hysteretic model for forces mobilized in non-linear elements of different isolation systems are obtained. Experimental shear force-displacement relationship obtained from uniaxial test of present study and that obtained by other investigator from both uniaxial and biaxial tests, have been simulated. Simulated hysteresis loops of different isolation system under both uniaxial and biaxial motion are found to be in good agreement with experimentally obtained hysteresis loops. A unified solution algorithm has been developed for computation of response of medium-rise base isolated structures, considering non-linear behaviour of isolation systems, subjected to general plane motion. This solution algorithm is based on Newmark's method in predictor-corrector form. The centre of mass of all the floors and the base are assumed to be on the same vertical axis. The response of a three storeyed r.c. framed building supported over either - P-F bearing, LRB, LLRB or EDF system, subjected to bidirectional motion of Koyna earthquake (1967), has been studied. Comparison of response of isolated structure in a particular direction for unidirectional and bidirectional excitations reflects the effects of biaxial interaction between orthogonal components of restoring force of isolation bearings. Response of solution algorithm and the computer programs developed in this study are in good agreement with that obtained from the more complex numerical studies reported in the literature. Slender shape of model LRB used for isolation of the test structure during Shake Table test, induced rocking mode of vibration In the overall response of the isolated structure. When, an additional rocking degree of freedom is incorporated only at the rigid base of the superstructure, computed response obtained from flexible model is found to be in close agreement with the experimental response of the model. Thus, response of the base isolated medium-rise framed buildings can be predicted reliably by solution algorithm and computer programs developed in the present study.