Crash Modeling of High-Pressure Wet Wound Composite Vessels

Limitations of fossil fuels and concerns surrounding global warming favor the introduction of new powertrain concepts with higher efficiency and low greenhouse gas emissions. Fuel cell vehicles offer the highest potential for sustainable mobility in the future. One major component of fuel cell vehicles is the hydrogen storage system. The most-used approach is to store hydrogen in carbon fiber reinforced plastic (CFRP) vessels manufactured by a filament winding process with an operating pressure up to 70 MPa (hereafter referred as H2 vessel). Accurate and reliable failure prediction of such thick composite structures with numerical methods in case of impact events is important.
The objective of this paper is the evaluation of the commercial fiber reinforced plastics material model MAT162 in LS-DYNA [1] to describe both the onset and the progression of damage of the H2 vessel. MAT162 has the capability of modeling progressive damage of composites. It is based on Hashin's quadratic failure criteria [2] and Matzenmiller's progressive damage model in terms of damage variables [3], and it is able to discern various composite failure modes.
Based on the introduced comprehensive equation set of MAT162 a consistent set of test procedures and the associated composite specimens are defined to quantify the material properties which are needed for solving the equations. The properties quantify the non-isotropic stressstrain relationship, the non-isotropic strength as well as continuum damage theory related values. The test procedures cover ASTM standardized tests as well as non-standardized tests. The material properties are either directly measured or quantified by calibration simulations using MAT162.
Finally, a quasi-static punch test using the cylindrical part of the vessel with a representative fiber layup and a comparable composite thickness derived from a real hydrogen storage vessel is reported in terms of strain, displacement, load, and damage propagation to validate the MAT162 FE crash model.