Techniques for correlation of drop weight impact testing and numerical simulation for composite GFRP crash boxes using Ls-Dyna

Composite materials are being widely used in automobile industry in order to meet the ever increasing demand to achieve significant weight reduction by maintaining the performance levels. Additionally, there are a number of challenging crashworthiness criteria set by the regulating authorities in various countries for vehicle safety, which the vehicle has to qualify. In these cases composites can be very useful as they exhibit high strength to weight ratios. But, deformation modes of composites are more complex when compared to metals, due to which it is more important to analyse the behaviour of composites under impact. In this study four types of cross sections (square, cylindrical, hexagonal and decagonal) were considered for the geometric shapes of GFRP (glass fiber reinforced plastic) crash boxes. Drop weight impact testing is widely used for studying the behaviour of parts subjected to impact loadings. The drop weight impact testing is highly dynamic in nature and in addition various factors like data acquisition signal noise (disturbance) requires the user to test 3-4 samples for each test specimen, for repeatability and minimisation of errors, which thereby increases the total number of testing samples. Therefore, in practise only a limited number of specimens can be actually tested using drop weight impact test considering the number of specimen samples to be manufactured and the cost involved in it. In this paper the development of numerical simulation model for drop weight impact test and its correlation with the real time drop weight impact testing was discussed in detail. *PART_COMPOSITE model along with *MAT58 material card was used for laminated GFRP composite specimen in Ls-Dyna along with calibration of parameters like longitudinal compression stress limiting factor (SLIMC1) and softness factor (SOFT) for simulation. The numerical simulation was correlated with the experimental data using correlation factors. The deformation modes as well as the F-D (force versus displacement) curves and SEA (specific energy absorption) values were found to have a close match between experiments and finite element simulations.