Probabilistic Analysis Of Shape Memory Alloy Modeling And Its Implication For Seismic Hazard Analysis

Superelastic shape memory alloys(SMAs)are cable of recovering large inelastic deformation and meanwhile dissipating energy under loading reversals therefore presents promising application in vibration control of buildings and bridges subjected to seismic loads.Properties of SMA are often determined through optimization and treated as deterministic for dynamic analysis of structures with SMA based devices.This study utilizes the Metropolis-Hasting algorithm to characterize the uncertainties within the SMA material model and to provide insight into SMA model parameter uncertainty.A series of cyclic tests of SMA bars with the same geometric size and heat treatment were first conducted and the experimental data is analyzed using the Markov Chain Monte Carlo(MCMC)method.The statistical properties of model parameters are calculated based on the posterior parameter samples.The influence of SMA model uncertainty on seismic application is explored on the energy dissipation capacity under cyclic tests.The first four L-moment method including parameter correlation are utilized to establish an efficient sampler based on MCMC-generated parameter samples to sample parameters for nonlinear response time-history analysis of a six-layer frame equipped with SMA support.The influence of the uncertainty of the SMA model parameters on the time-history analysis of the nonlinear response of a structure equipped with an SMA support member is studied based on the uncertainty of the response results including the maximum roof drifts and the maximum interstory acceleration,etc.