FatiRAN – Optimized Generation of Spectra for Random Vibration Testing*

Abstract

Abstract Random vibration testing with stochastic broad band noise signals is a commonly used procedure in qualification testing of vehicle parts. However, there are no standard procedures for deriving adequate test power spectral densities from road load data; hence test spectra can be defined far too low leading to tests not covering the loads of service life. In this paper, a new method is presented that gives consideration to the aspects of material fatigue during the whole process of windowing, overlapping, amplitude correction, and averaging that all have vital influence on the test spectra obtained. This allows for deriving test spectra with calculated damage equivalent to a defined set of design load histories, and thus helps minimizing the effort for calibration runs of test spectra on a test rig. The theoretic derivation will be shown as well as a comparison with the commonly used procedures using an exemplary case study.