Optimisation of the finish hard turning process for hardened 52100 steel with PCBN tools. (13th September 2007)

• Record Type: Journal Article
• Authors: Zhang, JingYing; Liang, Steven Y.; Yen, David
• Is Part Of: International journal of manufacturing research. Volume 2:Number 4(2007)
• Page Start: 428
• Page End: 447
• Publication Date: 2007-09-13
• Subjects: hard turning -- surface integrity -- tool wear -- genetic algorithms -- GA -- back-propagation neural networks -- optimisation -- hardened 52100 steel -- PCBN tools -- tool geometry -- cutting conditions -- white layer thickness -- residual stress -- process planning -- process modelling -- temperature -- force -- surface roughness -- optimal design; Manufacturing processes -- Periodicals; Manufacturing processes -- Automation -- Periodicals; Production engineering -- Periodicals; Factory management -- Periodicals; 670.5
• Languages: English
• ISSNs: 1750-0591

Abstract

This study develops a methodology for the planning of optimal tool geometry and cutting conditions to achieve specific performance goals, a satisfactory surface finish, and other practical constraints, while using the example of finish turning of hardened AISI 52100 bearing steel with Low CBN tools. A hierarchical structure modelling scheme for predicting white layer thickness and residual stress profiles is pursued to integrate the existing experimental, numerical and analytical knowledge into the process planning in an effective manner. Other hard turning process models, including temperature, force, surface roughness and tool wear, are also presented. Some of those models are non-explicit and are not analytically differentiable from the design variables, which renders traditional nonlinear optimisation techniques difficult to apply. An evolutionary optimisation scheme, a Mixed Integer Evolutionary Algorithm (MIEA), has been applied for this process optimisation problem. The optimal design outcomes derived from the methodology in this study were compared to other experimental and analytical work results for validation.