Journal of Sound and Vibration vol:329 issue:8 pages:1062-1080
In the vehicle design process, design decisions are more and more based on virtual prototypes. Due to competitive and regulatory pressure, vehicle manufacturers are forced to improve product quality, to reduce time-to-market and to launch an increasing number of design variants on the lobal market. To speed up the design iteration process, substructuring and Component Mode Synthesis (CMS) methods
are commonly used, involving the analysis of substructure models and the synthesis of the substructure analysis results. Substructuring and CMS enable efficient decentralized collaboration across departments and allow to benefit from the availability of parallel computing environments. However, traditional CMS methods become
prohibitively inefficient when substructures are coupled along large interfaces, i.e. with a large number of degrees of freedom (DOFs) at the interface between substructures.
The reason is that the analysis of substructures involves the calculation of a number of enrichment vectors, one for each interface degree of freedom (DOF).Since large interfaces are common in vehicles (e.g. the continuous line connections to connect the body with the windshield, roof or floor), this interface bottleneck poses a clear limitation in the vehicle noise, vibration and harshness (NVH) design process. Therefore there is a need to describe the interface dynamics more efficiently. This paper presents a Wave-Based Substructuring (WBS) approach, which allows reducing the interface representation between substructures in an assembly by expressing the interface DOFs in terms of a limited set of basis functions (“waves”).
As the number of basis functions can be much lower than the number of interface DOFs, this greatly facilitates the substructure analysis procedure and results in faster design predictions. The waves are calculated once from a full nominal assembly analysis, but these nominal waves can be re-used for the assembly of modified components. The WBS approach thus enables efficient structural modification
predictions of the global modes targets, so that efficient vibro-acoustic design modification, optimization and robust design become possible. Two case studies are presented in order to demonstrate the efficiency and accuracy of the WBS approach.