Search Results

Car floor structures typically contain a number of smaller-scale features which make them challenging for vibro-acoustic modelling beyond the low frequency regime. The floor structure considered here consists of a thin shell floor panel connected to a number of rails through spot welds leading to an interesting multi-scale modelling problem. Structures of this type are arguably best modelled using hybrid methods, where a Statistical Energy Analysis (SEA) description of the larger thin shell regions is combined with a finite element model (FEM) for the stiffer rails. In this way the modal peaks from the stiff regions are included in the overall prediction, which a pure SEA treatment would not capture. However, in the SEA regions, spot welds, geometrically dependent features and directivity of the wave field are all omitted. In this work we present an SEA/FEM hybrid model of a car floor and discuss an alternative model for the SEA subsystem using Discrete Flow Mapping (DFM).

Modelling the vibro-acoustic properties of mechanical built-up structures is a challenging task, especially in the mid to high frequency regime, even with the computational resources available today. Standard modelling tools for complex vehicle parts include finite and boundary element methods (FEM and BEM), as well as Multi-Body Simulations (MBS). These methods are, however, robust only in the low frequency regime. In particular, FEM is not scalable to higher frequencies due to the prohibitive increase in model size. We have recently developed a new method called Discrete Flow Mapping (DFM), which extends existing high frequency methods, such as Statistical Energy Analysis or the so-called Dynamical Energy Analysis (DEA), to work on meshed structures. It provides for the first time detailed spatial information about the vibrational energy of a whole built-up structure of arbitrary complexity in this frequency range.

Predicting the distribution of vibro-acoustic energy in complex built-up structures in the mid-to-high frequency regime is often a difficult task because “numerically exact” results obtained by Finite Element Method (FEM) may be of little practical value; the vibro-acoustic response of “identical” structures assembled as part of a manufacturing process is very sensitive to small changes in material parameters and/or variability in the shape of the structure. These differences may lead to large changes in the resonance spectrum and a full (and time expensive) FEM calculation for an individual sample has at best statistical significance. This problem becomes severe in the mid-frequency regime where the high-frequency techniques, such as Statistical Energy Analysis (SEA), are not yet available. Mid-frequency problems usually occur in structures with large variation of local wavelengths and/or characteristic scales.