NUMERICAL SCHEME INFLUENCE ON PSEUDO DYNAMIC TESTS RESULTS
The pseudo-dynamic test method is a well established and widely spreading experimental technique for structural analysis but the convenience of the chosen time-stepping scheme is not often adequately stressed and reported in experimental results. In fact to obtain results reliable such as those granted by shaking table tests it is necessary to stand a noticeable computational effort to integrate step by step the dynamic equilibrium equation. In this survey a comparison between a number of numerical integration techniques and some numerical examples is presented. The integration methods have been chosen among the most established and newest ones with the aim to investigate their advantages and downsides particularly with regard to computational effort and error propagation for linear and non-linear structural systems. Furthermore some general information and definitions concerning numerical integration tecniques and their stability and accuracy properties are given; Central Difference Method, Newmark Explicit Method, the α-Method, as conceived by Hilber, Hughes and Taylor in 1977, and the Time Discontinuous Galerkin Method, applied by means of predictor-corrector tecnique by Bursi, Bonelli in 2001 and 2002, are discussed.
Pseudo-dynamic tests errors numerical algorithms non-linear systems.
G. Amato L. Cavaleri
Dipartimento di Ingegneria Strutturale e Geotecnica , Palermo. Italy
国际会议
14th World Conference on Earthquake Engineering(第十四届国际地震工程会议)
北京
英文
2008-10-12（万方平台首次上网日期，不代表论文的发表时间）