In the automotive power train there is a lot of parts with very high but contradicting requirements: superior static and fatigue strength associated with low mass and particular inertia and vibration characteristics. In such cases it is very important to investigate residual strain / stress distribution as it influences to high extent the fatigue strength. For that purpose a test methodology consisting of semi-destructive, holographic interferometry, and non-destructive, 3D image correlation, measurements has been established.
Small diameter blind hole drilling was used as the destructive method to relieve residual stresses dormant in the previously deformed material. Introduction of the void into balanced strain / stress fields in the relatively thin layer caused imbalance, residual stress gradient over the hole depth and therefore complex in- and out-of plane displacements. Those minute displacements were measured by means of holographic interferometry. Holographic interferometry makes use of three-dimensional images recorded as holograms. A hologram of an object prior to the hole drilling is taken and is superimposed over the object's real-time image with the hole drilled. Due to the effect of coherent laser light interference, displacements anywhere on the surface of the object generate patterns of bright and dark lines called fringes. These interferometric fringes define lines of equal stress-induced displacement. In some cases it is enough to make one interferogram per drilled location to calculate strain / stress tensor components in the vicinity of the hole. Accuracy of methodology is defined by the test setup and doesn't depend on the field of view.
Obtained distribution of residual stresses was compared to results of 3D image correlation (photogrammetry) measurements. Several blanks of the component prior to the forming were etched with circle grids of known characteristics. After completion of each forming step a circle grid was analyzed by ARGUS photogrammetry system, which enables calculation of strain tensor components as well as sheet thickness changes. This way an incremental change in strain value could be calculated, as used version of ARGUS is capable only to calculate strain using zero deformation as a boundary condition. This data helps in optimizing each part-manufacturing step.
This paper is presenting a comparison of both methodologies for spot-checking of the local strains in manufacturing environment.