Application of high-performance computing to a bolt static tensile test

Abstract : Faced with an increasingly fast-changing market, airplane manufacturers have to reduce the development time of their new programs whilst dealing with more and more complex composite structures. The European project Maaximus (More Affordable Aircraft Structure Lifecycle through eXtended, Integrated & Mature nUmerical Sizing) aims at providing methods and tools to enable the fast development and right-first-time validation of a highly optimized composite airframe. In this context, accurate models of fasteners are crucial to ensure correct structure modeling at a higher level. The present study focuses on the mechanical behavior of a threaded assembly subjected to an axial tensile load. After a review of the state-of-the-art in this domain, the first part of the paper shows what benefits can be expected from explicit simulation for large quasistatic models. Parallelization strategies for High-Performance Computing (HPC) are presented, bringing out the need for consistency between the parallelization strategy and the computation structure. These simulation techniques are then applied to the modeling of a tensile test of a fastener. The results of corresponding experiments for two fastener diameters are presented. A successful correlation between simulations and experiments is found, both qualitatively and quantitatively. This study shows the efficiency of a well-chosen resolution algorithm in a coherent HPC environment for solving large non-linear models.
Complete list of metadatas

https://hal.archives-ouvertes.fr/hal-02048303
Contributor : Bruno Castanié <>
Submitted on : Monday, February 25, 2019 - 2:43:04 PM
Last modification on : Tuesday, October 22, 2019 - 5:20:44 PM

Identifiers

Citation

Louis Adam, Alain Daidié, Bruno Castanié, Elodie Bonhomme. Application of high-performance computing to a bolt static tensile test. International Journal on Interactive Design and Manufacturing, Springer Verlag, 2012, 6 (3), pp.195-203. ⟨10.1007/s12008-012-0153-0⟩. ⟨hal-02048303⟩

Share

Metrics

Record views

43