Discrepancies between calculated and measured residual stresses in quenched pure iron cylinder

W. Mitter,F. G. Rammerstorfer,O. Gründler,G. Wiedner
DOI: https://doi.org/10.1179/mst.1985.1.10.793
IF: 1.8
1985-10-01
Materials Science and Technology
Abstract:The residual stresses in a pure iron cylinder 50 mm in diameter, after water quenching from 850°C were calculated using the finite element method. This rather simple case without phase changes was chosen so that the calculated results could be compared with those from experiments. The comparison, however, showed discrepancies which could not be explained solely by the inaccuracy of the data or by the mathematical approach used for the calculation. These differences were found to be the result of plastic deformations that occurred during measurement of the residual stresses by the boring–turning method. This was borne out by a numerical simulation of the experimental procedure. Beginning the calculation at the residual–stress state after quenching, the boring and turning steps were simulated by a stepwise elimination of the contributions of the removed material to the total stiffness matrix. Thus, the changes in the residual stresses and the corresponding changes in the measured strain at the surface due to removal of material could be calculated. The boring–turning technique assumes pure elastic stress states throughout the process of material removal. However, the calculated results revealed that this technique caused unanticipated plastic deformations which resulted in erroneous experimental results in this special case. Stress values measured by mechanical methods are therefore unreliable if the corresponding equivalent stress in parts of the cross–section of the cylinder approaches the yield point at room temperature. MST/7
materials science, multidisciplinary,metallurgy & metallurgical engineering
What problem does this paper attempt to address?