Neural network approach for estimating the hole damage in drilling process of reactive powder concrete (rpc)

This study deals with the analysis of hole damage in the drilling of reactive powder concrete (RPC). Attention was first focused on the definition of an original approach on the basis of digital image processing (DIP) to quantify the extent of the damage. Then, an experimental approach, on the basis of the design of experiments (DoE), was performed to evaluate the influence of drilling operational parameters (namely, cutting speed, feed hole diameter) on the thrust force, torque RPC hole accuracy. The extent of hole damage was found to be influenced by the settings of feed and hole diameter, with cutting speed playing a minor role. Tool vibrations and troublesome centering procedures of the tools during drilling further complicated the interpretation of the hole damage.