Performing dynamic stress analysis on large-scale models within a practical timeframe by leveraging superelements.

This paper addresses the challenge of high hardware requirements for large full-vehicle models with millions of degrees of freedom, which can make dynamic stress simulations for durability impractical due to hardware shortages. To tackle this issue, an approach that reduces runtime without compromising accuracy is essential. The super element technique is proposed, particularly when certain modules remain unchanged across iterations. By leveraging previously solved results for these unchanged modules, the super element technique eliminates the need for solving the entire model in each iteration, thus significantly reducing analysis runtime. The objective of this paper is to outline a modeling strategy that incorporates the super element method and to evaluate its runtime efficiency and accuracy compared to conventional analysis methods. Modal frequency response function analysis is used to demonstrate the methodology. Instead of the traditional Lanczos approach for extracting eigenmodes, Altair's AMSES algorithm is employed. The paper compares the regular analysis approach with the super element approach in terms of modeling and run time, demonstrating the effectiveness of the latter in optimizing computational efficiency.