Open Collections

Saadatmand, Fatemeh

University of British Columbia

This thesis calculates and examines response sensitivities in nonlinear dynamic finite element analysis. The application is earthquake engineering, and the focus is on parameters of stochastic ground motion models. The direct differentiation method is employed to compute the response sensitivities. The derivatives are calculated with respect to ground motion intensity, soil frequency, and soil damping for two synthetic ground acceleration models representing spectral and filtered white noise approaches. The response sensitivities are compared for different soil types, structural natural periods of vibration, and yield displacements. Another aspect of the research presented here involves comparing two different Rayleigh damping models. Specially, both structural-level and element-level Rayleigh damping models are considered, in order to assess their pros and cons in nonlinear dynamic analysis. The considerations include spurious damping forces that may occur in certain cases of seismic loading. The first finding of the research is that the displacement response of a nonlinear structure is more sensitive to ground motion parameters than that of a linear structure. Furthermore, in case of nonlinear structural response, the parameters related to softer soil are more important than those for firmer soil in both spectral and filtered white noise approaches. For soft soil, it is the natural frequency of the soil that is the leading parameter. Conversely, for stiff soil, it is the total energy in the ground motion, per unit time, that is the most important parameter. However, at the time of maximum peak displacement response, the ground motion intensity is observed to be the parameter with the highest response sensitivity value among all. Lastly, element-level Rayleigh damping model is shown to reduce the spurious damping forces, in part by assembling the damping matrix at the element-level using the natural frequency of vibration for each element.

Text

2024-10-24

Attribution-NonCommercial-NoDerivatives 4.0 International

http://hdl.handle.net/2429/89483

Civil Engineering

University of British Columbia

UBCV

http://creativecommons.org/licenses/by-nc-nd/4.0/