| dc.description.abstract | In this thesis, two material models were studied, and the simulation results were compared to experimental data. The experiments were material tests (uniaxial tension and compression) and component tests (transverse plate deformation) of HDPE and PVC. The material models were the simple elasto-plasticmaterial model MAT24 and the new constitutive material model intended for thermoplastics developed atNTNU, both implemented in the finite element code LS-DYNA. The main intention was to study the differences between these models, and investigate the possibility of using the simple MAT24 instead of morecomplex and specialized material models, like the NTNU material model, without sacrificing too much accuracy.
Using MAT24 in the simulations performed in this thesis produces good results, compared to both theexperiments and the NTNU material model. The behavior is easily controllable, and the parameters are easilyidentified. Both MAT24 and the NTNU material model has their advantages and disadvantages, but whenincluding time and user-control as important factors in an industrial perspective, the MAT24 material model stands out as a viable option for simulations of thermoplastic components. However, understanding the deformation field in the component is vital for obtaining satisfactory results.
The most important features of the mechanical behaviour of thermoplastics is the large strain rate sensitivity of yield stress, the yield stress differences in tension and compression, and the plastic dilatation. MAT24 is by definition not able to capture the last two features, but the strain rate sensitivity of increasing yield stress with increasing strain rate was found to be captured well by an additive method through the VP option. The NTNU material model, however, captures all mentioned features, but predicts dilatation also in compression. | nb_NO |
