Uncertainty analysis of inelastic response of high-rise buildings to wind using a reduced-order building model

The influence of various uncertainties on wind-induced inelastic building response is quantified through response time history simulation of a reduced-order building model of a 60-story steel building as an example. The reduced-order building model and its uncertainties are established through modal push-over analysis of a highfidelity nonlinear finite element building model with distributed plasticity. The basic uncertain parameters are modal mass, stiffness, damping ratio and yielding displacement as well as loading power spectra. The results demonstrated that the uncertainty of response standard deviation (STD) is mainly contributed by uncertainty of modal damping ratio for elastic response, while it is further contributed by the uncertainty of the yielding displacement for inelastic response. The uncertainty of inelastic response with a higher level of yielding is less than that of elastic response. The uncertainty in extreme response is also determined in terms of the influences of randomness, sampling uncertainty and parameter uncertainty. The results showed that the influence of randomness can be accurately estimated using the crossing rate theory with consideration of non-Gaussian response character. The influence of sampling and parameter uncertainties on extreme response reduces with increasing level of yielding. The uncertainty in the time-varying mean alongwind displacement is also quantified, which is very sensitive to building yielding displacement.

Automated summary

The influence of uncertainties on wind-induced inelastic building response is quantified using a reduced-order building model. The basic uncertain parameters include modal mass, stiffness, damping ratio, yielding displacement, and loading power spectra. The results showed that the uncertainty in response standard deviation is mainly caused by the uncertainty of modal damping ratio for elastic response, while it is further contributed by the uncertainty of the yielding displacement for inelastic response. The influence of randomness, sampling uncertainty, and parameter uncertainty on extreme response is also determined.