Abstract

Surface topography after slurry erosion contains meaningful information concerning the process of erosive wear. This paper describes a procedure of inverse analysis to determine particle impact velocity from collective erosive features through image processing and further modeling. Slurry pot is used to produce wear damage on copper in exposure to a highly turbulent flow of dilute slurry of glass beads. Topography data for image processing is acquired by non-contact 3D profilometry. The analysis is carried out using two empirical models (Oka and Huang) and one analytical model (Cheng). The results reveal that the back-calculated velocity depends significantly on the model used. Whereas Huang's and Cheng's models produce values representative of the nominal flow conditions, the Oka's model results in unrealistic values. Finally, monotonic correlations between experimental variables associated with individual erosive features are discussed in the context of collective erosion models.