Numerical prediction of noise sources : facts, fears, future

Abstract

Numerical procedures, such as the finite element and boundary element methods, are known to be very suitable tools for the investigation of noise sources. It turned out, however, that early formulations of these methodologies are rather computer time consuming and therefore not used very often in practical applications. The contribution aims to give an overview of the current possibilities of numerical methods to model the noise generated by technical systems. It will show some "facts" of today's real live applications. Also the "fears" that computer simulations will replace measurements completely, will be addressed by means of representative examples. Finally, it will be discussed how in the "future" modern computations will be performed and which topics will be of particular interest in current research. All methodologies discussed in this contribution are introduced briefly and illustrated by examples which show how the approaches can be used and how accurate they are.