Residual Stress Characterization in Cross-Sections of Small Parts By Combining the Contour Method and Scanning White-Light Interferometry

Background The contour method (CM) allows characterizing the distribution of the normal component of the residual stresses that acts in the entire cross-section of mechanical parts. The technique starts with the well-controlled cutting of the part to be measured, usually carried out by wire electric discharge machining (WEDM). The geometric shape of the cross-section is carefully measured and used to determine the displacements resulting from the elastic relaxation of the residual stresses arising from the cut. Finally, this data is used to indirectly determine the residual stresses acting on the part before cutting. This method has been successfully applied to relatively large parts, where shape deviations can be properly measured with tactile coordinate measuring machines. When parts with small cross-sections are involved, the shape deviations on the cut surfaces are too small to be properly measured by tactile methods. Objective The objective is to provide a way to apply the contour method in cross-sections of small size parts using an alternative optical method. Method In this paper the authors use a tailor-made scanning white-light interferometer (SWLI) to measure the contours of the cut surfaces. On the one hand, interferometric measurement achieves sub-micrometer resolution over a large cloud of points. On the other hand, the measurements are strongly influenced by the natural roughness resulting of the EDM cutting process, which, on this scale, give rise to considerable noise levels. Different data processing strategies are compared to reduce noise effects and extract useful shape deviation information to calculate residual stresses. To evaluate the procedure, a steel part with a cross-section of 6.0 × 19.0 mm 2 was measured. In order to make a reliable comparison, the sample was submitted to thermal treatment to reduce residual stresses. After that, a permanent four-point bending load was applied. Two WEDM cuts were made: (a) the first one in the central region of the specimen with bending-induced residual stresses and (b) the second one in a region of the bar where no residual stresses are expected. Results The obtained residual stress distribution of the central region showed good agreement with values resulting from the analytical prediction. The results from the stress-free region were closer to zero, which is good evidence that the WEDM cutting process was correctly made. All steps of the contour method needed to be done carefully, aiming to