Towards large-scale acoustic shape optimization for industrial applications using the Boundary Element Method

Research output: Chapter in Book/Report/Conference proceedingArticle in proceedingsResearchpeer-review

Abstract

Recent developments within the area of acoustic shape optimization show huge potential for the improvement and design of acoustic devices. So far, acoustic shape optimization has been mostly applied to simplified academic setups. However, its application to industrial sized problems is still limited. In this work, we explore the possibilities of extending current research on acoustic shape optimization based on the Boundary Element Method to large-scale acoustic setups. The efficiency of a recently proposed semi-analytical adjoint sensitivity analysis is investigated in the context of large problems. The adjoint approach is applied to optimize and evaluate the radiation performance of acoustic setups. Moreover, an important factor for successful shape optimization is how the geometry is allowed to deform, i.e., the choice of parametrization. This choice is especially difficult in the case of 3D problems. Therefore, the work will also discuss different implementations of the parametrization of 3D surfaces.
Original languageEnglish
Title of host publicationProceedings of Internoise
Number of pages11
PublisherInternational Institute of Noise Control Engineering
Publication date2020
Publication statusPublished - 2020
Event49th International Congress and Exposition on Noise Control Engineering - Virtual event, Seoul, Korea, Republic of
Duration: 23 Aug 202026 Aug 2020
https://internoise2020.org/

Conference

Conference49th International Congress and Exposition on Noise Control Engineering
LocationVirtual event
Country/TerritoryKorea, Republic of
CitySeoul
Period23/08/202026/08/2020
Internet address
SeriesINTER-NOISE and NOISE-CON Congress and Conference Proceedings
ISSN0736-2935

Fingerprint

Dive into the research topics of 'Towards large-scale acoustic shape optimization for industrial applications using the Boundary Element Method'. Together they form a unique fingerprint.

Cite this