Search Results

The sound field in a vehicle is one of the most complex environments being a mixture of multiple, correlated and uncorrelated sound sources. The simulation of vehicle interior sound has traditionally been produced by combining multiple test results where the influence of one source is enhanced while the other sources are suppressed, such as towing the vehicle on a rough surface for road noise, or measuring noise in a wind tunnel. Such methods are costly and provide inherent inaccuracies due to source contamination and lack of synchronization between sources. In addition they preclude the addition of analytical predictions into the simulation. The authors propose an alternative approach in which the component sounds are decomposed or separated from a single operating measurement and which provide the basis for accurate sound synthesis.

Experiencing the results of virtual NVH analysis in an immersive physical simulation is the only accurate method of developing vehicle, system or component targets and designs. This paper describes an engineering approach specifically created to enable physical interaction with test, CAE and hybrid NVH models, at every stage in the vehicle design process from concept to full detail. It explains the need for sound and vibration decomposition and synthesis, and interactive sound and vibration replay. Implementation of this process has led to the development of engineering tools that enable Interactive NVH Simulation. The paper also describes the practical use an engineer can make of a ‘rapid prototyping’ desktop NVH simulator in the design process. A full scale NVH Simulator is then used to allow evaluation of final design alternatives under realistic driving conditions by non-specialists (i.e. the customer) as well as specialists.

Sound quality targets for new vehicles are currently specified by jury evaluation techniques based upon listening studies in a sound laboratory. However, jury testing is costly, time consuming and at present there are no methods to include customer expectations or brand requirements. This paper describes a neural computing approach that is being developed to generate knowledge and tools to enable objective measures of a product's sound to be converted into a prediction of the subjective impression of potential customers without carrying out the traditional jury evaluation tests.