Date of Award

August 2023

Degree Type

Thesis

Degree Name

Master of Science

Department

Engineering

First Advisor

Ilya Avdeev

Committee Members

Woo-Jin Chang, Konstantin Sobolev

Keywords

Numerical Simulation, Reduced-Order Modeling

Abstract

Numerical simulations have been proven to be a powerful tool for predicting, testing, and validating the capabilities of new designs. However, given the high demand for simulating extremely complicated geometries and nonlinear physical phenomena, simulations can often be significantly time consuming. Consequently, the development of high-precision reduced-order models becomes indispensable to reduce computational time. In this study, we simplified and characterized an industrial motion system based on linear magnetic motor technology using accurate full 3-D numerical model. The system behavior was explored through various scenarios, including extreme conditions, to gain a deeper understanding of its thermal behavior during operation. The simulation results were then compared with experimental measurements. To achieve model order reduction, the initial and boundary conditions, along with temperature distributions derived from the simulation results, were translated into excitations and outputs for constructing robust reduced-order models. Subsequently, the reduced order model was thoroughly tested and validated against new scenarios derived from the 3-D simulation results.

Download

Share

COinS