Advances in the contour method for residual stress measurement

Abstract

The aim of this PhD thesis is to extend the capability of the contour method for residual stress measurement in metallic components by resolving two of its main limitations. The contour method involves sectioning a body into two equal parts that have mirror-symmetric geometry, stiffness and residual stress field. The deformations of the cut surfaces introduced by sectioning are then measured. These raw measured data are processed using filtering and smoothing techniques. The last step involves back calculating the residual stress distribution acting out of the cut plane that has been relaxed. This is equal to the original residual stress present at the cut plane before the body was cut. A major limitation of the contour technique is that it is strictly only applicable to flat cuts along a symmetry plane of a body. Another fundamental assumption in the contour method is that residual stresses re-distribute elastically during the sectioning cut. However, this assumption can be violated if the residual stress magnitude is close to the material yield strength value and lead to plasticity cutting induced errors in the contour method results. Thus another major limitation of the technique is the risk of cutting induced plasticity that can introduce significant stress measurement errors. This PhD thesis contributes to knowledge in the field through first presenting a novel contour data analysis approach for the more general case of sectioning at an arbitrary plane where the cut parts do not possess mirror-symmetry. This greatly extends the types of structure, and the volume of material within structures, where residual stresses can be measured using the contour method. The second contribution to knowledge is the invention of an incremental contour measurement method involving multiple cuts. This new approach can be applied to sequentially reduce residual stresses in the structure of interest and thereby lower or eliminate the risk of inducing plasticity during cutting and reduce consequent measurement errors. Both new approaches proposed in this PhD thesis are successfully demonstrated through numerical simulation using the finite element method and experimentally on benchmark steel specimens and against neutron diffraction measurements.