Low-Velocity Penetration of Aluminium Plates

Abstract

The present thesis describes research on quasi-static and low velocity perforation of rolled aluminium plates, where the main objective has been to gain a better knowledge of the physical processes taking place during this type of structural problem. The objective has been met by a combination of laboratory tests, material modelling and non-linear finite element simulations.

The thesis is organized in a synopsis, giving a brief introduction to the problem and summarising the main findings and conclusions, in addition to four independent papers.

Paper I presents an experimental technique for measuring the deformations the plate undergoes during impact and perforation. This information can be used to validate numerical models and to increase the understanding of how energy is absorbed by the plate.

Paper II presents an experimental and numerical investigation on the quasi-static perforation of AA5083-H116 aluminium plates. In the tests, square plates were mounted in a circular frame and penetrated by a cylindrical punch. A full factorial design was used to investigate the effects of varying plate thickness, boundary conditions, punch diameter and nose shape. Based on the obtained results, both the main and interaction effects on the maximum force, displacement at fracture and energy absorption until perforation were determined. The perforation process was then computer analysed using the nonlinear finite element code LS-DYNA. Simulations with axisymmetric elements, brick elements and shell elements were conducted. Slightly modified versions of the Johnson-Cook constitutive relation and fracture criterion were used to model the material behaviour. It was shown that the FEM models were able to predict the trends observed in experiments.

Paper III evaluates methods for determination of the anisotropic properties of polycrystalline metallic materials. Four calibration methods were evaluated for the linear transformation-based anisotropic yield function YLD2004-18p (Barlat et al., 2005) and the aluminium alloy AA5083-H116. The different parameter identifications are based on least squares fits to combinations of uniaxial tensile tests in seven directions with respect to the rolling direction, compression (upsetting) tests in the normal direction and stress states found using the full-constraint (FC) Taylor model for 690 evenly distributed strain paths. An elastic-plastic constitutive model based on YLD2004-18p has been implemented in a non-linear finite element code and used in finite element simulations of plane-strain tension tests, shear tests and upsetting tests. The experimental results as well as the Taylor model predictions can be satisfactorily reproduced by the considered yield function. However, the lacking ability of the Taylor model to quantitatively reproduce the experiments calls for more advanced texture models.

Paper IV presents an experimental and numerical investigation on low velocity perforation of AA5083-H116 aluminium plates. In the tests, square plates were mounted in a circular frame and penetrated by a cylindrical blunt-nosed projectile. The perforation process was then computer analysed using the nonlinear finite element code LS-DYNA, in order to investigate the effects of anisotropy, dynamic strain aging and thermal softening in low velocity impacts on the present aluminium alloy. Dynamic strain aging has been shown to influence both the predicted force level and fracture, while thermal softening only influences the fracture prediction. No effect of plastic anisotropy was observed.