Keywords

Reduced order methods, Heat Transfer, Akaike Information Criterion

Abstract

Reduced order methods using spectral representations show promise in facilitating and accelerating heat transfer analyses. This paper proposes a taxonomy for reduced order methods, classifying a method as reduced order compression, modelling, or analysis. The performance of bases formed with analytical and empirical eigenfunctions are compared for representative reduced order tasks. The Akaike Information Criterion is applied in a novel way to compare the performance of these bases. The present study finds that both bases are parsimonious for reduced order compression tasks. Empirical eigenfunctions are more robust to for reduced order modelling with variations in modelling parameters such as thermal diffusivity. Analytical eigenfunctions are more robust for reduced order analysis in the presence of noise. The Akaike Information Criterion is shown to provide an unambiguous optimal truncation criterion for reduced order analysis applications in the presence of noise.

Document Type

Report

Publication Date

2024-01-09

Language

English

College

Ira A. Fulton College of Engineering

Department

Mechanical Engineering

University Standing at Time of Publication

Graduate Student

Share

COinS