A New Model for Simulation of Fabric Leakage in LS-DYNA

Uncoated fabrics will continue to have a major market share in airbag applications in the future. The reasons are smaller package size, easier recycling and reduced material cost. In order to achieve proper prediction of performance and restraint function of an airbag module for occupant protection, a detailed and physical representation of all significant factors contributing to gas management within the airbag cushion is required. Inflator output and gas losses through vent holes as well as leakage through seams and the fabric itself must be clearly identified and described. The FE-Code LS-Dyna has been used for many years to support the development of airbag systems and to optimize the restraint function in the specific vehicle environment. The code provides several options for the description of gas outflow including fabric leakage. But so far the available models for fabric permeability have shown limited predictivity in conjunction with standardized measured permeability data. Therefore a new model for fabric leakage has been developed by Autoliv. It is capable of calculating the massflow through the fabric, depending on strain in the fabric. Each fabric is characterised by a set of constants. As for orifices the equation for the flow through nozzles and diffusors after St. Venant / Wantzel [1] has been used. This model is able to distinguish subsonic and critical flow. Autoliv’s approach to the necessary but unknown effective area consists of a pressure dependent and a strain dependent part. It also takes the angle between warp and weft into account. This model has recently been implemented in LS-Dyna by Dynamore/LSTC. Data from Autoliv’s leakage testing device (GES) for testing of fabric samples has been used to obtain the relevant input data for validation at fabric level. Reference tests as well as tests with fabrics having high, low and very low permeability have been performed. All tests have been carried out both with nominal strain (bulge) and with reduced strain (grid). It could be shown that even with reduced strain this in-house device delivers significantly different results for the fabrics with high and low permeability. After the simulation model of the test device had been validated according reference tests using a leakproof plate, sets of constants describing the fabrics’ permeability could be determined. Very good correlation to the test results could be achieved. In order to verify the model on module level whole passenger airbag modules have been tested on a linear impactor. Airbags with the above validated fabrics and two different ventsizes were used. Significant differences in the results for the different fabrics were obtained for the airbags with 2x8mm vents. The simulation model of the airbag has been tuned according the test results with coated fabric in order to account for all other leakage besides fabric permeability. Simulation of the impactor tests with prevalidated fabric leakage data showed the validity of the new approach.

application/pdf D-III-02.pdf — 4.9 MB