Abstract
NVH represents a significant contribution to the Brand Image of vehicles. With today´s more quiet engines, annoying noises coming from the gearbox are more and more a source of concern. At the same time, gearboxes become more complex and have to transmit higher torque and horsepower, while packaging and fuel consumption requirements demand more compact and lightweight designs.
In view of shorter vehicle development cycle engineers have to maximize their understanding of gearbox NVH phenomena and to rely more and more on simulation in the early stages as well as during the final refinement, to minimize lengthy and costly prototype testing. In this paper, new approaches for gearbox testing and simulation are described and illustrated by concrete examples. In a first section it is explained how the complex source-transferreceiver chain is identified in detail on a gearbox test bench. Specific methods enable to find the root cause of major noise phenomena in terms of excitation forces at the bearings and noise radiating surfaces. The outcome is a detailed understanding of the noise generation mechanism, which can guide the engineer in solving remaining issues and help him to avoid critical issues in next generation products.
In a second section numerical methods are introduced, which aim at evaluating the noise radiated by the gearbox prior to the prototype availability. The goal is to compare the result of the simulation with predefined targets and to eliminate the main weakness of the design early in the development process. In order to achieve this, fully numerical methods can be used, whereas sometimes a combination of the models with predecessor test data is more efficient. Weak spot detection tools enable to understand the weaknesses in detail, while Fast or Detailed modification prediction tools are combined to come to an optimal solution. The results are illustrated in the case of the development of automatic transmission for a rear wheel drive vehicle.
Keywords: Gearbox, NVH, mechanical Transfer Path Analysis, Panel Contribution Analysis, numerical optimization