A dynamic modelling approach to computer aided optimum selection of grinding parameters

• Published in: Journal of materials processing technology Vol. 38; no. 1; pp. 227 - 245
• Main Authors: Zhu, C.B., Midha, P.S., Trmal, G.J.
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 1993
• Subjects: Computer aided analysis; Database systems; Knowledge based systems; Mathematical models; Optimization; Selection
• ISSN: 0924-0136
• DOI: 10.1016/0924-0136(93)90199-G

The traditional approach to the specification of grinding parameters is experiential rather than scientific. Whilst some research work has been done in the grinding optimisation area, most of the work dealt with quite simplified optimisation problems. In particular, these research findings are fragmented, not readily available for production engineers, often not user-friendly and the validity of existing grinding optimisation models is subject to the experimental environment in which the models were developed. The present paper introduces a dynamic modelling approach which was proposed as a computerised methodology to generalise grinding research findings and production expertise as far as possible and to make them available for the production grinding environment. A dynamic model has a flexible structure and can be formalized automatically by determining its modifying factors using a knowledge based system according to the grinding conditions input by the user. A number of objective functions has been developed to dynamically describe relationships between input and output parameters for the cylindrical plunge grinding process, such as grinding forces, specific energy, grinding temperature, surface roughness, grinding G ratio, etc. The paper describes also a comprehensive computer aided system developed for optimum selection of grinding parameters with consideration of multiple criteria and constraints. The system consists of multiple dynamic models, modifier databases, knowledge bases and a user interface. The process of optimum specification of grinding parameters is divided strategically into two layers. Layer I is for determination of grinding conditions and Layer II for optimum selection of equivalent chip thickness and operating parameters. An example of application is given in the paper.