RESPONSE OF A TURBO ROTOR FOUNDATION SYSTEM UNDER EARTHQUAKE EXCITATION

Abstract

Various studies have been incorporated for dynamic analysis of a rotor- bearing-foundation system. Analysis has been done on the steady state and transient responses of the Jaffcott rotor for the elastic bearings mounted on damped and flexible supports. In recent years transient response studies have been taken an important place in the subject of rotor dynamics. This becomes necessary in view of the operating speeds approaching or crossing the stability threshold. It is also necessary to analysis of rotors of machinery subjected to earthquake excitation. Very little work has been reported on this field and required more research on this subject. Two different types of modelling approaches are available for turbo generator foundation system. Comparative study of both approaches is required for accurate design and analysis. In the present study a part of TG foundation is taken for comparative analysis of two different finite element modeling approaches. In first case solid tetrahedrov element is used for modeling of foundation and in second case beam element as recommended in IS 2974 (part 3) is used. Rotor is modelled as a beam element in both cases. This rotor is a part of turbo alternator. Rotor is coupled with foundation by means of bearings which is modelled as a spring element. Modal analysis and response spectrum analysis is done by using ANSYS software. In model analysis ten frequencies and mode shapes are obtained and normalized effective masses at each mode are also calculated. Input response spectra are given at base of foundation in axial X and Z directions. The response is taken in form of maximum stresses and maximum displacement. Time history analysis is also done to study exact comparison for both approaches by using COSMOS/M software. Response is taken at two bearing nodes of the rotor. It becomes easy to give a detailed comparison after getting response for both models. In this work it has been shown that maximum stresses and displacements differ substantially not only in terms of values but also in terms of location for both models. Since tetrahedroh elements are known to give results which are much closer to reality in comparison to beam element, the conclusion of this work is that IS 2974 (part 3) may require revision in this regard.