CAE Transfer Path Analysis and Its Accuracy Evaluation Using a Validation Method

Abstract

<div class="section abstract"><div class="htmlview paragraph">In-cabin Noise at low frequency (due to engine or road excitation) is a major issue for NVH engineers. Usually, noise transfer function (NTF) analysis is carried out, due to absence of accurate actual loads for sound pressure level (SPL) analysis. But NTF analysis comes with the challenge of having too many paths (~20 trimmed body attachment locations: engine and suspension mounts, along with 3 directions for each) to work on, which is cumbersome. Physical test transfer path analysis (TPA) is a process of root cause analysis, by which critical contributing paths can be obtained for a problem peak frequency. In addition to that, loads at the attachment points of trimmed body of test vehicle can be derived. Both these outputs are conventionally used in CAE analysis to work on either NTF or SPL. The drawback of this conventional approach is that the critical bands and paths suggested are based on the problem peak frequency of test vehicle which may be different in CAE. Secondly, the force that is derived from measured acceleration, in physical test, when applied to CAE trimmed body will not produce the similar acceleration to that of Physical test. This is due to the correlation differences in transfer functions between physical test and CAE in trimmed body. The study presented in this paper replaces the conventional TPA carried out in physical testing with CAE TPA by using the transfer functions of CAE trimmed body. Hence, the critical bands and ranking of critical paths along with the loads derived through this method would be more accurate in carrying out the NTF and SPL analysis, respectively.</div></div>