Title:
Physics enabled Data-driven structural analysis for mechanical components and assemblies
Physics enabled Data-driven structural analysis for mechanical components and assemblies
Author(s)
Shah, Aarohi Bhavinbhai
Advisor(s)
Rimoli, Julian J.
Editor(s)
Collections
Supplementary to
Permanent Link
Abstract
Analyzing structures that exhibit nonlinear and history-dependent behaviors is crucial for many engineering applications such as structural health monitoring, wave management/isolation, and geometric optimization to name a few. However, current approaches for modeling such structural components and assemblies rely on detailed finite element formulations of each component. While finite element method serves to be versatile and well-established for nonlinear and history-dependent problems, it tends to be inefficient. Consequently, their computational cost, becomes prohibitive for many applications when time-sensitive predictions are needed. In the present work, we introduce a framework to develop data-driven dimensionally-reduced surrogate models at the component level, which we call smart parts (SPs), to establish a direct relationship between the input–output parameters of the component. Our method utilizes advanced machine learning techniques to develop SPs such that all the information pertaining to history and nonlinearities is preserved. Unlike other data-driven approaches, our method is not limited to any particular type of nonlinearity and it does not impose restrictions on the type of analysis to be performed. This renders its application straightforward for a diverse set of engineering problems, as we show through multiple case studies. We also propose a novel meta learning based approach to enable an extension of this approach to dynamic problems. In addition, we present several ways to enhance this approach in terms of precision and efficiency. Thus, the present work provides an approach that can dramatically boost the computational efficiency and simplicity to analyze large structures without sacrificing accuracy.
Sponsor
Date Issued
2022-05-20
Extent
Resource Type
Text
Resource Subtype
Dissertation