3rd ICAI 2024

International Conference on Automotive Industry 2024

Mladá Boleslav, Czech Republic

specified ranges of input parameters, the shoulder force is in most cases significantly lower than the required limit, while the upper rib compression is unsatisfactory high in many cases. Therefore, more attention has been paid to such combinations of input parameters that significantly reduce the upper rib compression and at the same time do not increase the shoulder force. Fig. 7 shows the resulting shape of the airbag. Although there has been an increase in shoulder force, it still meets the requirements safely. However, the compression of the upper rib has decreased significantly. Figure 7: Comparison of the original (a) and modified (b) airbag shape and its influence on rib compression and shoulder force (a) (b)

4. Conclusion The conducted study demonstrated the feasibility of airbag design optimisation using the AI principles, which consisted in the automatic creation of a surrogate model for the studied problem. In addition to the optimized solution itself, the surrogate model offers the opportunity to test the influence of input parameters on the result and to define the range of input parameters where the best or worst results can be expected. It is also possible to test the interaction of input parameters on output data. Procedures based on similar principles may be applied to other types of tasks related to the development of the car body. It is necessary to pay more attention to the stochastic character of explicit FEM computations, i.e. slightly different results may be computed for identical input data. From the preliminary analyses, it seems that suitable approximation functions should be used instead of Lagrange interpolation polynomials for the mathematical description of the surrogate model. Such formulation can cover better the results scatter. Then, the resulting surrogate model shows rather trends than the exact solution. The resulting optimized solution have to be verified by a FEM calculation.

99