NVH engineers are faced with the challenge of designing trim parts for vehicle interior and exterior, like inner dash insulators, carpets, underbody shields or engine encapsulations, which can be made with very different Bills of Materials (BOMs) including among others foams, felts or heavier layers. The measurables commonly used to rank various solutions are Transmission Loss (TL) and absorption. Depending on the numerical analysis method, different approaches may be considered for the evaluation of the TL of an automotive component. In particular, in Statistical Energy Analysis (SEA), automotive components are modeled as an assembly of panels having a simple shape, e.g. flat panels and/or panels with single or double curvature. Furthermore, in SEA the trim is normally modeled by means of the Transfer Matrix Method (TMM), which is essentially a 2-dimensional methodology. This paper intends to analyze in some depth the level of approximation that these practices bring with themselves, specifically in relation to the modelling of an automotive floor.
More in detail, the aim of the paper is first to investigate what impact has the presence of the tunnel on the TL of a vehicle floor in bare and trimmed conditions and then to evaluate if the presence of the tunnel can be better modeled by using a semi-cylinder or three flat plates welded together in a trapezoidal shape, both shapes considered as a reasonable simplification of the actual geometry of a typical tunnel.
The analysis is carried out at simulation level using FE. To investigate both air bone noise and structure borne noise transmission, two types of excitations are used: a diffuse acoustic pressure field applied to the entire floor surface and an imposed displacement applied to the edge of the floor surface. Furthermore, 3 different kind of trims are taken into consideration in order to analyze if and how the tunnel modeling strategy may influence the evaluation of the trim effectiveness.