Analytical Life Prediction Modelling of an Automotive Timing Belt 2008-01-1207

This paper presents a methodology that makes use of computer based analytical simulation methods combined with statistical tools to predict timing belt life. This allows timing belt life to be estimated with no requirement for running test engines and associated test equipment, which is both very time and expense exhaustive. A case study on a belt driven primary drive for a V6 Diesel engine was used to illustrate the methodology. A computer based dynamic model for the belt drive system was developed and validated, and a belt life prediction model was developed, which uses tooth load predictions from the analytical model. Statistical modeling of predicted damage accumulated to failure was used to estimate the model parameters given a limited set of belt life results from a motored rig test. The practical use of the model is illustrated by predicting belt life under customer usage.