Avoiding ghost stress on reconstruction of stress- and composition-depth profiles from destructive X-ray diffraction depth profiling

The present paper addresses the accuracy of the reconstruction of depth profiles from X-ray diffraction lattice-strain analysis, combined with successive sublayer removal. In order to test the accuracy of reconstruction irrespective of experimental inaccuracies, X-ray diffraction was simulated for model stress-depth profiles and/or composition-depth profiles, reflecting those obtainable with surface engineering of materials, i.e. shot peening, carburizing of austenite and low temperature nitriding of stainless steel. Two principally different methods for the reconstruction of the actual stress and composition profiles were compared:(I)assuming that the lattice parameter determined at a specific depth, for a specific value for ψ is a weighted average over the actual lattice spacing profile for this ψ-direction;(II)assuming that the stress/strain determined at a specific depth is a weighted average over the actual stress/strain depth profile.On the basis of the results it is concluded that method (I) virtually avoids the occurrence of ghost stresses (stress artefacts) upon data evaluation. Substantial ghost stresses may occur upon data analysis using method (II) for strongly compositionally graded materials.