3rd ICAI 2024

International Conference on Automotive Industry 2024

Mladá Boleslav, Czech Republic

Figure 5: Workflow of side airbag optimisation

When a sufficiently accurate replacement HDMR model was available, it was possible to proceed with a detailed mapping of the system’s response to any change in input parameters. Using the Monte Carlo method, a set of 100,000 combinations of input parameters was generated, for which HDMR calculated the expected output values – maximum shoulder force and maximum upper rib compression. In the next step, statistical methods were used to analyse these results and to obtain a lot of interesting information. Fig. 6 shows how the data can be processed. For example, it is possible to evaluate the sensitivity of output values to input parameters changes and to assess which parameters have the greatest influence on the result. Fig. 6a shows this analysis for upper rib compression. At the same time, it is possible to separate the input parameter ranges into sub-areas, in which the input parameters lead to minimum or maximum values of the output values. Figure 6b shows this separation for the scaling factor of the gas generator power and its influence on the upper rib compression. For values around 0.7, the minimum rib compression can be expected, while around 1.4, the highest rib compression values can be expected. Figure 6: Upper rib compression: sensitivity analysis (a) and regions of minimal or maximal values for gas generator power (b) (a) (b)

However, the main goal of the study was to find such combinations of input parameters that lead to the lowest load on the dummy. The results of the study showed that for the

98