Abstract
This paper presents a successful implementation of a methodology for NVH phenomenon evaluation carried out on the example of the gear rattle. The main purpose is to determine the NVH properties of a 5-speed manual transmission in view of gear rattle tendency and its effect on an entire system, especially the vehicle interior noise, in the early stages of product development process. Based on the X-in-the-loop approach for drive train systems, investigations and assessments of transmission noise can be performed in different test environments, such as test benches or fields depending on the applied evaluation criterion. The main objective is to study the phenomenon itself and to verify the gearbox performance. The dynamic drive train test bench allows the investigation with comparable boundary conditions without interference of other sources like the combustion engine. To predict the tendency of the gear rattle presence and evaluate the resulted annoying level, the user-friendly human sensation modeling tool is used. This has been developed based on the Artificial Neural Networks (ANN) and is already successfully applied for vehicle start-up analysis and ratings. In comparison to an investigation in the field, one on the roller test bench enables the same boundary conditions like runway friction coefficient or aerodynamic drag by simulation of the environment. From the experimental results, the relationship between the objective parameters captured in different X-in-the-loop layers and the subjective comfort ratings is determined. Consequently, the elaborated human sensation models can be effectively applied to compare different gearbox solutions for future product by estimation of the gear rattle tendency and by prediction of the comfort rating based on the experimental as well as simulation data.
Keywords: comfort objectification, gear rattle, NVH, human sensation modeling, Artificial Neural Networks (ANN)