On establishment and validation of a new predictive model for material removal in abrasive flow machining

Abrasive flow machining (AFM) is a non-traditional surface finishing method by extruding abrasive media through workpieces' surfaces. It is difficult to implement uniformity and controllability for material removal in AFM, which makes predictions of material removal and its distribution along media flow direction significantly imperative. In this paper, a new predictive model for material removal in AFM is proposed based-on combination of material removal model for single abrasive and statistics model of active abrasives. To validate its applicability, two experiments are conducted to calibrate the material removal model, and another two experiments are carried out to validate the calibrated material removal model. Results show that this model can predict material removal in terms of profile height variation Delta H and variation trend of mass variation AM precisely, with predictive error (delta(error)) of 6.4% and 6.9% achieved in the validation experiments. In the case that the workpiece and flow channel are given, the variation of profile height Delta H on the workpiece surface shows consistent distribution trend with the distribution of the pressure and media flow velocity on the workpiece surface.