ECCOMAS 2024

Projection Based Reduced Order Model of Elasto-Acoustic Vibrations Computed with Isogeometric Analyses

  • Landi, Tommaso (University of Salerno)
  • Hoareau, Christophe (Conservatoire National des Arts et Métiers)
  • Deü, Jean-François (Conservatoire National des Arts et Métiers)
  • Citarella, Roberto (University of Salerno)
  • Ohayon, Roger (Conservatoire National des Arts et Métiers)

Please login to view abstract download link

In this study, we focus on the modelling of vibro-acoustic phenomena [1], employing Isogeometric analyses (IGA) [2,3,4,5] and projection-based reduced-order models (PROM) [6]. The chosen numerical approach relies on the use of NURBS (Non-Uniform Rational B-splines) to describe the geometry, unknown displacements and pressure fields. The objective of this work is to apply IGA methodology to the study of coupled fluid-structure vibrations in the context of automotive applications [7,8]. This type of study holds significant relevance for automotive design and noise prediction within vehicles, addressing the computation of elasto-acoustic vibrations and providing valuable insights for optimizing and reducing noise in the interior acoustic environment of cars. The discretization of both the fluid and structure significantly influences the coupled dynamic behavior. This study explores the advantages of employing NURBS, known for preserving geometries from computer aided design models, and PROM to reduce the dimensionality of numerical systems. Comparisons with iso-parametric finite element method approaches (FEM) are provided, particularly for 3D geometries. Our methodology involves: (i) generating a reference test case, using FEM with robust industrial software, such as Nastran [9], to compute coupled eigenmodes and frequency response functions within a specified frequency range; (ii) constructing a PROM of the problem using FEM; (iii) replicating the same test case using IGA [10]; and (iv) constructing a PROM of the problem using IGA. Comparisons in terms of errors and convergence rate will be presented to demonstrate the effectiveness of this methodology in the context of automotive interior acoustic prediction, whether using IGA or FEM.