Design And Development Of Time-domain Explicit Method For Stochastic Vibration Based On SiPESC

Randomness,as one of natural basic characters,is very common in natural phenomena and has very important effects on structure safety,durability and amenity.At the beginning of the 1950 s,stochastic vibration theory,as a very important branch of stochastic process theory,was studied mainly for aviation engineering needs,and later used in the architecture and manufacturing industry.In recent years,China's architecture and manufacturing industry has archived remarkable development,the structure of its products are becoming larger,longer and more irregular.As a consequence,structure design and construction are more difficult.Nowadays people often use numerical simulation technology to guarantee products' safety durability and amenity,to short the design cycle.To simplify calculation,common softwares often use stationary model theory,but most problems in engineering are non-stationary.It will be very difficult and complex to analysis this kind of problem by classic theory,which needs the more highly computer performance.And it will be very hard to complete the task for the simulation of huge structure or complex problems.Due to the complexity of classical nonstationary stochastic vibration analysis theory,common softwares can only implement stationary vibration analysis.In this thesis,an effective analysis procedure is designed and developed based on SiPESC platform,according to timedomain explicit method.And a data format for non-stationary random vibration is proposed.Several numerical examples have been given to compare the result of the pseudo-excitation method and the time-domain explicit method.It shows that with the same time history analysis and numerical integration method the results have no differences between the two methodsAs it is very difficult to input power spectrum in common softwares,a user interface is designed for this problem,which can be used to both stationary and non-stationary random vibration analysis.Users can choose proper power spectrums and mode functions according to different requests,and use proper methods to analysis different problem types.