Material models and crash simulations in LS-DYNA: Development of virtual test laboratory
Examensarbete för masterexamen
Automotive engineering (MPAUT), MSc
Rao Sudarshan, Mayur
Viswanath Shankar Rao, Sriharsha
Extensive developments in the field of computational mechanics led to substantial improvements in the assessment of vehicles in a virtual environment. Vehicle safety is one of the prime areas where virtual simulations are comprehensively used in the product development phase. In order to obtain effective results, it is very important that the material model predicts the failure behaviour accurately in crash analysis. The main goal of the current study is to contribute to the development of a virtual material test laboratory where different materials can be tested in multiple stress states in LS-DYNA using a calibrated input flow curve (obtained from an uniaxial tension test simulation). The study was carried out in two phases. Initially, coupon tests were simulated using solid fine elements which were finally validated against the experimental results in order to achieve better accuracy. A input flow curve was calibrated from the uniaxial tension test simulations using optimisation technique and the same curve was finalized for all the further simulations. The same coupon tests were then simulated using shell elements and a damage model (GISSMO) was coupled to a constitutive material model, as damage models play an important role in predicting the fracture in the material. Necessary information like triaxiality values, failure strain, and critical strains computed from the solid element models was used to calibrate the GISSMO model. The calibrated model was then subjected to mesh regularization for element sizes varying from 0.5mm to 5mm to determine the sensitivity of mesh size on failure prediction. The calibrated material model for shells showed promising results and was in good agreement with experiments and in a full scale Offset Deformable Barrier crash simulation. The material model was computationally efficient and, using this methodology, different stress states can be predicted quite effectively using a single flow curve (derived from simple tension test). This methodology can also be verified for other metals and polymers.
LS-DYNA , *MAT_024 , GISSMO , Optimisation , triaxiality , critical strain , failure strain , flow curve