Analytical Matlab/Simulink model of pyrotechnical gas generators for airbags

Abstract

One of the key components of inflatable restraint systems is the gas generator (inflator). This relevance of the generator is mainly due to the fact that it is the only active component on most airbag systems. Each new car model requires a new process of developement and system integration to set up the performance of the restraint system. This process usually demands a performance adjustment of the gas generator to optimise the overall behaviour of the whole system. Often, the number of gas generator configurations available from serial or prototype manufacturing does not cover the whole range of each design parameter. Also, the time needed to manufacture a prototype definition of a gas generator can be delayed from some days to several months, depending on the extent of the required modifications. This process can be significantly shortened by a simulation approach that provides the performance at the output of a particular gas generator. The present work describes a new analytical model of a pyrotechnical gas generator developed in Matlab/Simulink code. The advantage of the selection of this software is that it can be linked with most of the vehicle crash simulation softwares currently available in the market. Also, the utilisation of the block modelling technique of Simulink allows the easy implementation of new capabilities to the model. The model has been built to reproduce the performance of the inflator measured on the widely accepted closed tank test. Finally, to guarantee the predictability of the model it has been validated by means of (1) the comparision, made at three different test temperatures, of the theoretical results with the baseline of the results of the tank pressure tests performed during one year of serial production of a particular gas generator model; (2) comparing gas generator variability limits for tank test performance, calculated with the same serial production gas generators, with the variability limits obtained by means of a Monte Carlo analysis using the analytical model. It is demonstrated by means of this validation of robustness of the presented model.