Development of a Fatigue Analysis Tool Chain for Automotive Structural Applications

The drive toward increased usage of light metal alloys in automotive structural applications creates new design challenges. In some cases fatigue performance may become an important design consideration. This paper examines the integration of existing finite element pre & post processors, fatigue analysis codes & software robots; for the purpose of providing the designer with parameter sensitivity studies and robustness analysis through Monte Carlo Simulation (MCS).

These tools are aimed at exploiting ever growing high performance computing (HPC) resources to provide more robust designs while reducing the design cycle time. To provide meaningful results such analysis can only be conducted on a single element for a given analysis cycle; to this end, methods of generating localized regions within a part are discussed.

The analysis and subsequent improvement, of the fatigue performance of a cast aluminum part was used as a case study. Significant improvements in design turnaround time and potential reduction of errors were made through the automation of the fatigue analysis process. This was achieved through reducing the analyst's workload (for each design iteration) and providing quantitative performance indicators, reducing the number of required iterations and highlighting which design inputs were major contributors to fatigue in critical regions. The MCS provided results of the statistical robustness of the design, unavailable in many design processes.