Automotive drivetrain model for transmission damage prediction

This paper presents the development and the validation of an automotive drivetrain model aimed to investigate the impact of true-to-life operating conditions on the damage prediction of transmission gearwheels. The model is based on a torsional and non-linear multi-body system representing the rotating parts between the engine and the driven wheels. The boundary conditions are performed with an engine model and a tire model coupled with a suspension & pitch model. The generation of the driving inputs is fully automated through a tunable driver model relying on predefined driving cycles. The model allows the simulation of start-up processes, gear shifts and tire grip loss, as well as of the transmission dynamics and the eigenbehavior of the gearwheels. The damage estimation is computed continuously and in parallel, by means of a linear damage accumulation method. The drivetrain dynamics was validated with temporal and frequency analyses based on measurements, data from the transmission manufacturer and literature. The estimated damages are also reality-conform. They were validated with comparative data recorded on vehicles during a measurement campaign on open roads.