Mass flow determination of airbag inflators

The aim of this study in cooperation between Fraunhofer ICT and AUDI AG is to develop and optimize a new mass flow and temperature determination method for performance characterization of airbag inflators. The mass flow of an inflator is essential information for prediction of inflator's performance within a multitude of crash scenarios. It is based on the thrust characterization method in combination with emission spectroscopic temperature determination using near infrared (NIR) spectroscopy. The intended scope of this study is to apply this newly developed characterization method on a pyrotechnic airbag inflator, which is currently in use on several vehicles, and for which there is a considerable amount of hardware test data available. The proposed characterization method comprises two different experimental setups, from which it can be generated two different mass flow data sets respectively. These data sets can be subsequently cross correlated as a validation method. The first experimental setup consists in a device to perform internal NIR-spectroscopy and pressure measurements to directly determine mass flow and temperature during inflator's outflow after ignition. The second experimental setup consists in a modified ballistic pendulum on which the airbag inflator is mounted. There, using high-speed video recordings and particle image velocimetry, the inflator´s thrust can be calculated. A second mass flow data set can be then obtained combining the thrust measurements and the temperature data from the first experimental setup. After obtaining the two mass flow data sets a cross correlation is possible. To validate the obtained inflator data, these are applied stepwise as input data for simulation models of different complexity. It starts with simple models to be able to directly evaluate the data without the influence of other parameters. As second step the data impact on more complex airbag deployment models and, for example, OoP simulations are evaluated. In all considerations, a comparison is made between simulation with conventional inflator data from tank tests as well as with ICT data and, where available, hardware tests. The aim is to identify the impact of this new approach, in comparison to conventional data, on the simulation results of different load cases and which outcomes reflect reality better.