Optimizing crash box design to meet injury criteria: a protocol for accurate simulation and material selection

• Published in: Structural and multidisciplinary optimization Vol. 67; no. 8; p. 156
• Main Authors: Rostro-González, Horacio, Puigoriol-Forcada, Josep Maria, Pérez-Peña, Armando, Menacho, Joaquin, Garcia-Granada, Andrés-Amador
• Format: Journal Article
• Language: English
• Published: Berlin/Heidelberg Springer Berlin Heidelberg 01.08.2024; Springer Nature B.V
• Subjects: Bumpers; Computational Mathematics and Numerical Analysis; Criteria; Design optimization; Design standards; Energy levels; Engineering; Engineering Design; Initial fit; Injuries; Materials selection; Metal sheets; Physical properties; Statistical distributions; Theoretical and Applied Mechanics; Thickness; Yield strength; Yield stress; Monte Carlo; Crash; FEM
• ISSN: 1615-147X; 1615-1488
• DOI: 10.1007/s00158-024-03855-2

The design of a deformation element or crash box that meets a given injury criterion based on deceleration requires careful consideration of physical properties and space requirements. Variations in material yield stress or geometry can result in statistical variations in the injury criterion output. Optimizing the crash box to fulfil two different injury criteria and two different energy levels may require more space than initially specified. In this study, we propose a protocol where the crash box is collapsed, and force–displacement is fitted to an equation. This fit is carried out with just two simulations and compared to 30 possible scenarios, obtaining a maximum error of 38.9%. With this initial fit, the appropriate thickness and yield stress can be chosen to perform crashes with two energy levels and monitor four injury values. With the ideal yield stress and sheet metal thickness, we introduce real statistical distributions using Monte Carlo design to perform 200 simulations and obtain 400 injury values for each design proposal. This technique ensures that the design will meet injury requirements for any possible combination of thickness and yield stress accepted by quality inspection. If only one simulation is performed, all designs meet the requirements, but only the last proposed design decreased the average injury to 9.2 g with a standard deviation of 2.68 g and a maximum value of 14.4 g, which is less than the required 15 g. This technique minimizes the risk of finding combinations of yield stress and thickness that produce an undesirable injury criterion.