Identification of defect locations in forged products using the finite-element method

In the first of the metalforming operations considered, that of plane-strain side-pressing, the finite-element predicted deformation behaviour of 60-40 brass and 7075 aluminium alloy has been compared with experiment. Both visioplasticity and hardness studies have been performed. The macrohardness survey was found to be the most appropriate in validating the use of the finite-element technique. The numerical calculations of the local stress and strain distributions are then used with a number of previously published continuum fracture criteria to predict the fracture initiation sites. For certain of the successful criteria the level of deformation at fracture has also been predicted using critical values of fracture found experimentally from an axisymmetric tensile test. For the operation of plane-strain side-pressing, very good agreement with both the experimental fracture initiation site, and the level of deformation at fracture, is reported for the criterion of a critical value of generalised plastic strain at fracture, and the equivalent criterion of generalised plastic work. The second metalforming operation considered is simple upsetting. The fracture initiation behaviour of 60-40 brass has been examined for specimens of the four initial aspect ratios of 0.5, 1.0, 1.5 and 2.0. Again, the generalised plastic strain/work criterion has successfully predicted the experimental fracture initiation sites in all cases. However, satisfactory levels of deformation at fracture predictions have only been found for the two lower aspect ratio specimens. A small range of strip compression and tension operations on 60-40 brass compose the third type of metalforming operation considered in this thesis. Yet again the generalised plastic strain/work criterion has successfully predicted the fracture initiation site found experimentally but not the level of deformation at fracture. Finally, the axisymmetric extrusion of 60-40 brass and 7075 aluminium alloy is examined. For the brass using the generalised plastic strain/work criterion, good agreement with experiment has been found for both the fracture initiation site and level of deformation at fracture found experimentally. However, for the aluminium alloy only the correct fracture initiation site has been found. In summary, it appears that the fracture criterion of a critical value of generalised plastic strain/work has successfully predicted the fracture initiation site found experimentally in all the operations considered in this thesis but has been unable to consistently predict the correct level of deformation at fracture. These differences in the predicted and experimental level of deformation at fracture cannot be explained with reference to the finite-element calculated levels of hydrostatic stress. Further work is necessary to explain this difference.