< Back to previous page

Project

Integration of damping in wave based simulation models for mid-frequency vibro-acoustic analysis and design.

Growing customer expectations regarding vibro-acoustic performance together with more restrictive regulations on noise emission levels and human exposure to noise and vibrations, have forced design engineers to takethe vibro-acoustic behaviour of their products into account in the development process. Moreover, following the current ecological trends, lightweight designs are emerging to save material costs and to reduce fuel consumption. This may, however, lead to strongly deteriorated noise and vibration insulation properties. Machine and vehicle manufacturers face the challenging task to meet several, often conflicting, design requirements.

Numerical prediction techniques allow the design engineer toevaluate the sensitivity of different parameters on the design criteria, and limit the need for time-consuming and expensive prototypes. Unfortunately, none of the currently available prediction techniques are capable of giving accurate predictions in the mid-frequency range, which is also the region where for many industrial applications the human hearing is most sensitive. At these frequencies, deterministic prediction techniques lead to unfeasible model sizes, and consequently calculation times,while statistical prediction techniques cannot be applied since the underlying assumptions are not yet met. The inclusion of lightweight poroelastic damping materials in vibro-acoustic models even further restricts the applicable frequency range of the classical deterministic predictiontechniques.

This dissertation fits in the development of a deterministic Wave Based Method, which aims at accurately predicting the low and mid frequency response of steady-state dynamic problems. By approximating the dynamic field variables by a set of wave-like basis functions,an affordable procedure is obtained, allowing predictions up to higher frequencies as compared to conventional deterministic approaches. The method has shown its effectiveness and efficiency for bounded and unbounded acoustic, structural and vibro-acoustic problems.

With the increasing importance of lightweight damping materials in mind, the current PhD project focuses on the extension of the Wave Based Method towards the Biot equations, which accurately describe the coupled dynamic behaviour of the fluid and solid phase of those materials. Efficient schemes areobtained for 2D Cartesian and axisymmetric problems. Whereas the formercan be applied for general problems, axisymmetric Wave Based schemes allow to model for instance an acoustic impedance tube in an affordable way.

An important point of attention in the modelling of poroelastic materials, is the existence and treatment of singularities. Infinite values of solid stresses and fluid displacements can occur in the vertices of poroelastic problem domains. Physically, this is of course impossible, but due to linearisations in the theory of elasticity, singularitiescan occur in the mathematical model. The smooth wave-like basis functions have difficulties describing those steep gradients and as a result, the efficiency of the Wave Based Method deteriorates when singularities are present. This dissertation derives criteria to predict when singularities can be present in acoustic and poroelastic problem domains. Specialpurpose enrichment functions are included in the set of basis functionsto counter the adverse effect of singularities.

The developed procedures are supported by a number of validation cases, illustrating theapplicability and the efficiency of the proposed approaches.
Date:1 Oct 2008 →  10 Dec 2012
Keywords:Vibro-acoustic
Disciplines:Mechanics
Project type:PhD project