Travail de fin d'études et stage[BR]- Travail de fin d'études : Improvement and validation of the numerical model of a 30-ton truck used in crash simulations[BR]- Stage d'insertion professionnelle : GD Tech

Abstract

This thesis aimed to enhance and validate a numerical model of a 30-ton truck for crash simulations using LS-DYNA software. The software uses dynamic principles and methods such as multibody simulations and finite element analysis to simulate the behavior of vehicles in realistic scenarios. The main objective was to create a vehicle model that accurately reflects reality. LS-DYNA is one of the best software packages on the market for handling complex problems that involve intricate geometries and highly non-linear behavior, such as vehicles crashing into road safety barriers. The thesis was conducted at GDTech, a company that specializes in crash dynamics with numerical vehicle models. The work involved analyzing the existing model and the previous work, proposing improvements to the model based on the analysis, and following the EN 16303 validation procedure for numerical models in crash simulations. The main improvements were the geometry of the mudguard, which increased the maximum compression for front suspensions; and the stability of the model by constraining the ballasts. These improvements were crucial for the validation roadmap that tested the model's performance and reliability. The model passed all the tests except the curvilinear track test, which faced some problems due to plastic hourglass deformation of the axles. A methodology was adopted to avoid these problems. Further simulations should be done to validate this test. We also suggested some future work on the model, such as using cruciform structures for the chassis; and using new dampers for the rear suspensions that have more damping force in rebound than in compression. This thesis, therefore, contributes to the creation and use of realistic and reliable numerical models for crash simulations in accordance with the EN 16303 standard