MODELLING THE INFLUENCE OF PART GEOMETRY, STACKING SEQUENCE, AND LOADING RATE ON THE ENERGY ABSORPTION OF OPEN NON-CRIMP FABRIC COMPOSITE CHANNELS
Topic(s) : Material and Structural Behavior - Simulation & Testing
Co-authors :
Valerie AJAYI (CANADA), Michael WORSWICK , John MONTESANO (CANADA)Abstract :
While the potential of the application of carbon fibre reinforced plastics (CFRPs) as energy absorbing structures has been widely acknowledged, adoption in the automotive industry has been limited due to the high manufacturing costs and challenges in accurately predicting the macroscopic responses of composites during impact loading.Liquid composite molding processes with highly reactive resins and unidirectional non-crimp fabric (UD-NCF) reinforcements may enable further adoption of CFRP composites as energy-absorbing parts for the high-volume production of vehicles. Components made via high-pressure resin transfer molding (HP-RTM) offer reduced cycle times, while UD-NCF reinforcements provide a reduction in manufacturing cost, high in-plane mechanical properties, and design flexibility. Modelling performance remains challenging due to the complex interaction of damage mechanisms that are difficult to macroscopically capture. Since predictive modelling is extensively used in the automotive industry, additional research on accurately capturing the crashworthiness of more affordable CFRP components is required to improve their adoption as energy-absorbing structures in vehicles. This study focuses on modelling the influence of part geometry, stacking sequence, and loading rate on the impact performance of unidirectional UD-NCF CFRP open channels manufactured via HP-RTM.
In this study MAT54, a material model for the progressive failure of composite laminates available in LS-DYNA, is calibrated to simulate the damage progression and energy absorption of CFRP open channels under dynamic and quasi-static axial loading. Data for the calibration and validation of the simulations was collected from experimental work available in [3] and [4]. The study results in a set of baseline models that enable the evaluation of the fidelity of MAT54 when used to simulate incrementally altered stacking sequences and geometries. It is shown that MAT54 can be calibrated to predict the force-displacement responses and energy absorption of UD-NCF CFRP channels within 15% of the experimental results as is shown in Figure 1. However, it is also observed that while changes in geometry and loading condition do not require significant recalibration, changes in the stacking sequence are greatly influential – similar to the influence of stacking sequence observed experimentally – and require a reconsideration of the non-physical parameters; most notably the direction and mode-dependent failure strains. Lastly, the results serve as a standard to which more physically based models (i.e., considering delamination, strain rate dependencies, etc.) can be confidently compared.
In this study MAT54, a material model for the progressive failure of composite laminates available in LS-DYNA, is calibrated to simulate the damage progression and energy absorption of CFRP open channels under dynamic and quasi-static axial loading. Data for the calibration and validation of the simulations was collected from experimental work available in [3] and [4]. The study results in a set of baseline models that enable the evaluation of the fidelity of MAT54 when used to simulate incrementally altered stacking sequences and geometries. It is shown that MAT54 can be calibrated to predict the force-displacement responses and energy absorption of UD-NCF CFRP channels within 15% of the experimental results as is shown in Figure 1. However, it is also observed that while changes in geometry and loading condition do not require significant recalibration, changes in the stacking sequence are greatly influential – similar to the influence of stacking sequence observed experimentally – and require a reconsideration of the non-physical parameters; most notably the direction and mode-dependent failure strains. Lastly, the results serve as a standard to which more physically based models (i.e., considering delamination, strain rate dependencies, etc.) can be confidently compared.