Vibro-Acoustic Reduced Order Modelling: A Modular Approach

Master Thesis (2025)
Author(s)

L. Costa Arslanian (TU Delft - Mechanical Engineering)

Contributor(s)

PG Steeneken – Mentor (TU Delft - Precision and Microsystems Engineering)

Bart van der Aa – Mentor (ASML)

Haydar Dirik – Mentor (ASML)

J. Yang – Graduation committee member (TU Delft - Mechatronic Systems Design)

Faculty
Mechanical Engineering
More Info
expand_more
Publication Year
2025
Language
English
Graduation Date
02-07-2025
Awarding Institution
Delft University of Technology
Programme
['Mechanical Engineering']
Faculty
Mechanical Engineering
Reuse Rights

Other than for strictly personal use, it is not permitted to download, forward or distribute the text or part of it, without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license such as Creative Commons.

Abstract

Acoustic-structural interactions present significant engineering challenges, particularly in the domains of noise reduction and vibration control. At ASML, measurement-based analyses have revealed that acoustic disturbance paths often dominate the dynamic behavior of atmospheric lithography machines. This project focuses on enhancing ASML’s current one-way coupled acoustic-structural modelling approach by developing a two-way coupled, known as vibro-acoustic, modelling framework. However, this advancement introduces substantial computational complexity, necessitating effective model reduction techniques. The primary objective of this work is to reduce vibro-acoustic models in a way that preserves their ability to be modularly coupled with other system components. To this end, three Component Mode Synthesis (CMS)-based reduction methods were evaluated, with only one proving suitable for both academic and industry-scale models. The resulting reduced-order models successfully retained the dynamic fidelity of the full system and enabled efficient coupling with other substructures. When applied to harmonic excitation analyses, the reduced models achieved a dramatic reduction in computational time, from several hours to approximately one minute, while accounting for the cost of model reduction.

Files

MSc_Thesis_Lucas.pdf
(pdf | 10.9 Mb)
License info not available