Implementation of power semiconductor aging behavior in an electric vehicle powertrain model

Thermal stress in power semiconductor modules causes mechanical wear effects, which eventually lead to failure. A common approach to estimate the time to failure is to utilize power cycling tests. The results of these tests can not directly be transferred to real-life applications. In this paper, we propose a way to implement an on-line lifetime estimation in a dynamic electric vehicle simulation model. Results from power cycling tests are combined with simulation results of the application dependent temperature characteristics. They are processed using rainflow counting and Palmgren-Miners rule in order to calculate the progression of lifetime on-line. This information is fed back into the semiconductor simulation model.