Frequency domain fatigue under non-Gaussian, correlated vibration loads

• Applicant: Professor Dr.-Ing. Peter Wolfsteiner
• Subject Area: Mechanics
• Term: since 2024
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 528049351

The evaluation of the fatigue strength of mechanical structures is of central importance in mechanical engineering and has a significant influence on their reliability and economic efficiency. Corresponding models and procedures are well researched and well developed in many areas. However, important fundamental methods are lacking in the area of fatigue strength of vibrating structures loaded by random vibrations. Such random vibrations occur, for example, when a wheel rolls on the road or rail; the resulting loads can significantly damage the affected structures due to resonances. Previously known models are based on the assumption of Gaussian random processes. On their basis, a consistent, statistically sound analysis of fatigue strength in the frequency domain is possible on the basis of power spectra densities (PSD). This spectral description has important advantages in terms of modelling and computing time, especially for structural dynamics problems. PSDs are even used in technical standards to define vibration loads. However, practical problems rarely fulfil the requirement of Gaussian distributions, which leads to considerable design errors, especially in fatigue processes. In order to solve this open problem, the applicant has developed an innovative calculation method based on higher-order spectra, which allows a well-founded solution for a practically relevant class of non-Gaussian signals. This method allows the transformation of non-Gaussian signals to several Gaussian parts and thus opens up the possibility of further using existing PSD-based calculation methods with all their advantages. However, this method is currently limited to a single load channel. The aim of this research project is the extension and generalisation to multiple, correlated excitation channels. This will enable a statistically sound and computationally efficient solution to numerous practical problems that are hardly solvable so far.

• DFG Programme: Research Grants (Transfer Project)
• Application Partner: Engineering Center Steyr GmbH & Co KG; (Magna Powertrain); Siemens Mobility GmbH