Vibration isolation concept in aseismic design of structures

Abstract

The phi losophy of conventional design principles is reviewed for earthquake resistant structures. It is emphasized that in order to achieve a higher degree of rei iabil ity and also to prevent the occurence of any cracking at critical points, especially for important structures, like nuclear power plants, a new technique of aseismic design should be developed. The concept of vibration isolation, although used extensively in connection with machine foundations and bridge supports during the last one hundred-years, is introduced for important structures as an indispensable measure of security against earthquakes. A comprehensive review is presented about the types of vibration isolation available and their specific appl ications on real structures. The mathematical formulation is introduced for the time history dynamic response analysis of structures with base isolation. In order to assess the degree of accuracy as welI as to determine the validity of various assumptions of analytical studies, a steel model frame is experimentally investigated at the shaking table facilities of the Institute of Earthquake Engineering and Engineering Seismology, Skopje, Yugoslavia. For each floor level, peak response values of accelerations and displacements have been calculated analyticalIy and also have been measured instrumentally at the shaking table. A wide spectrum of peak input accelerations, ranging from 0,05 g to 0.70g9 have been considered in association with the 1940 EI Centro, U.S.A. and the 1979 Montenegro, YugosIavia earthquakes. The frequency contents of these earthquakes have been also varied by using a reduced time scale. The ratio of reduction has been taken as t corresponding to the square root of the ratio of the geometric scale of the model. Generally, the analytical investigations produced in very close agreement results to those obtained by laboratory measurements. Some discrepancies occured however, in the response values of the fixed base case, under real time earthquakes, on account of nonlinear behaviour of the modeI at the shaking tabIe tests. It has been established both by analysis and experiments that the peak response values of the structure are significantly reduced. Moreover the structure practically moves only in the rigid body modes, remaining always in the elastic range, when vibration isolation is used. It has been also concluded that the rubber elements are unable to provide any vibration isolation in the vertical direction, thereby being very susceptible to high degree of acceleration amplifications. Finally, the influence of damping on the displacement response has been also investigated. The use of viscodampers in connection with hellcal springs as recommended by the writer, proved to be very successful in supplying adequate energy absoprtion capacity as welI as in providing necessary isolation in alI possible modes of vibrations.