Taguchi-controlled knowledge-based assistant for improving quality of hollow cylindrical sections

Abrasive reaming has been successfully applied for internal finishing of cylindrical sections to produce highly accurate bore characteristics, i.e. diameter, cylindricity, surface finish, straightness and roundness. The process is an alternative to honing and cylindrical grinding whose mechanics are more complex and expensive. This success has, however, been manually orientated with little use of computer aids. To achieve a given bore geometry and maintain low scrap rates, a knowledge-based system may be used to optimize the reaming process. This paper proposes a generic knowledge-based system, called SAR using parameters which significantly influence the reaming process, identified from Taguchi-based experimental analysis. The main objective of the system is to determine process limits, detect uncontrollable situations, identify causes and suggest corrective actions, in order to achieve a better overall bore profile.