Cheng Su, Rui Xu. Time-domain method for dynamic reliability of structural systems subjected to non-stationary random excitations[J]. Chinese Journal of Theoretical and Applied Mechanics, 2010, 42(3): 512-520. DOI: 10.6052/0459-1879-2010-3-2009-042
Citation:
Cheng Su, Rui Xu. Time-domain method for dynamic reliability of structural systems subjected to non-stationary random excitations[J]. Chinese Journal of Theoretical and Applied Mechanics, 2010, 42(3): 512-520. DOI: 10.6052/0459-1879-2010-3-2009-042
Cheng Su, Rui Xu. Time-domain method for dynamic reliability of structural systems subjected to non-stationary random excitations[J]. Chinese Journal of Theoretical and Applied Mechanics, 2010, 42(3): 512-520. DOI: 10.6052/0459-1879-2010-3-2009-042
Citation:
Cheng Su, Rui Xu. Time-domain method for dynamic reliability of structural systems subjected to non-stationary random excitations[J]. Chinese Journal of Theoretical and Applied Mechanics, 2010, 42(3): 512-520. DOI: 10.6052/0459-1879-2010-3-2009-042
School of Civil Engineering and Transportation, State Key Laboratory of Subtropical Building Science, South China University of Technology, Guangzhou 510640, ChinaSchool of Civil Engineering and Transportation, State Key Laboratory of Subtropical Building Science, South China University of Technology, Guangzhou 510640, China
Structural dynamic equations are first transformed intothe form of state equations, which are solved by the precise time integralmethod, and then explicit expressions of structural random responses undernon-stationary excitations are deduced in the time domain. The computationaleffort for such explicit formulation is only equivalent to that for twodeterministic time-history analyses of the structure. Based on the aboveexplicit expressions and combined with the first-excursion failurecriterion, a numerical simulation method is proposed for solving dynamicreliability of structural system under non-stationary random excitations. Ascompared with the power spectrum method, the proposed method does notrequire a large amount of numerical integrals in both frequency and timedomains. Furthermore, the assumptions are no longer required in the presentapproach with respect to the probability distribution of the excursionnumber and the correlation between different failure modes. With numericalexamples, the calculation accuracy and efficiency of the proposed method arecompared with those of the conventional Monte Carlo simulation method, thePoisson process method and the Markov process method. Numerical resultsindicate that the proposed method has perfect accuracy and reasonably highefficiency.