Correlation of Dominant Noise Transfer Paths in Statistical Energy Analysis Vehicle Model from Test as Basis for Variant Vehicle Development 2013-01-1994

For purposes of reducing development time, cost and risk, the majority of new vehicles are derived strongly or at least generally from a surrogate vehicle, often of the same general size or body style. Previous test data and lessons learned can be applied as a starting point for design of the new vehicle, especially at early phases of the design before definite design decisions have been finalized and before prototype of production test hardware is available. This is true as well of vehicle NVH development where most new vehicles being developed are variants of existing vehicles for which the main noise transfer paths from sources of interest are already understood via test results and existing targets. The NVH targets for new vehicles are defined via benchmarking, market considerations, and other higher-level decisions. The objective is then to bridge the gap between test data from surrogate vehicles to direct support of the NVH development of new vehicle programs. Because of its strength in providing analysis predictions of the effect of design changes on vehicle NVH at higher frequencies, Statistical Energy Analysis (SEA) is an established tool for using available test data to correlate an SEA model that can be adapted for early design phase NVH development of new vehicles. The effect of changes to materials, gage thickness, sound package, source levels, or geometry changes on the interior noise levels can be predicted by SEA with good accuracy to support design decisions that must be made early in the program. This paper illustrates with a concrete example an idealized implementation of this process. The main test plan design considerations for a baseline surrogate vehicle are discussed. Some key test results and their uses are presented. The updating and correlation of an SEA model representing the baseline vehicle are shown. The objective methods for determining the effectiveness of the correlation are given using this vehicle as an example. Finally, the use of a correlated SEA model to effectively support the NVH development of several variant vehicle programs at an early phase of the design process is presented along with suggestions for the best use of this design tool, its advantages and limitations, and the most effective roles it can serve to support the overall vehicle design cycle.