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

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