A verified and efficient approach towards fatigue validation of safety parts
Abstract
In the automotive industry, safety parts must be designed according to the state of the art of science and technology such that they do not fail as long as the vehicle is used according to its purpose and misuse of the vehicle does not exceed a reasonably expectable degree. Due to scatter in customer loads and component properties, fatigue validation needs to be based on statistical methods. Mathematically sound methods are devised in order to make the validation process as efficient as possible. They allow considering all test results, including censored test data (e.g. tests suspended due to premature failure of components which are not under consideration). Furthermore, these methods permit adapting the success run criterion successively to the testing process.
