Abrasive flow finishing (AFF) is one of the widely used advanced finishing processes in which a small quantity of work material is removed by flowing semisolid abrasive-laden putty over the workpiece surface to be finished. AFF is popular for finishing and deburring of difficult-to-access areas. This process is also used for radiusing, producing compressive residual stresses, and removal of recast layer. In order to enhance productivity of the process, several modifications in AFF process are being tried. In this paper, a concept of rotating the medium along its axis has been introduced to achieve higher rate of finishing and material removal. This process is termed as drill bit-guided abrasive flow finishing (DBG-AFF) process. In order to provide random motion to the abrasives in the medium and to cause frequent reshuffling of the medium, the medium is pushed through a helical fluted drill, which is placed in the finishing zone. The experiments are carried out to compare AFF and DBG-AFF processes with AISI 1040 and AISI 4340 as workpiece materials. The performance of DBG-AFF as compared to AFF is encouraging, specifically with reference to percentage change in average surface roughness (% ΔRa) and amount of material removed. Modeling using non-linear multi-variable regression analysis and artificial neural networks are carried out to conduct parametric analysis and to understand, in depth, the DBG-AFF process. The simulation data of neural network show a good agreement with experimental results.
