An improved discrete system model for form error control in surface grinding

Abstract Grinding remains as one of few choices being able to machine very hard materials to deliver ultra high precision at high material removal rate for efficiency. Effective models are needed for precision control of the machining process. So far, few studies on form error prediction have been reported. Machining usually begins with partial removal of workpiece surface. Without in-process sensing, system parameters could not be accurately determined nor surface form information thus preventing us from modeling for precision control. In this study, an improved discrete system model and an in-process sensing technique have been proposed to address the partial removal and precision control problems. Models for partial removal, full removal, and sparking out conditions have been established. Form error assessment in the partial removal stage has been investigated. It is found that the grinding constant is able to reflect changes in machining conditions and is able to represent machining capability. A larger grinding constant will mean a reduced size reduction. Further studies of the grinding constant are necessary. For the accurate estimation of the grinding constant, two approaches are proposed. The iterative approach was found more suitable and convergent. The proposed models and in-process sensing technique were validated through experimental testing in terms of workpiece surface form profile y n ( x , z 0 ), average size reduction c n , surface form error E pv n and normal grinding force F n n . Through detailed examination and comparative studies, the proposed models and in-process sensing technique offered significant improvements ranging from approximately 16.9% to 23%, compared with the existing models. Except the grinding force, which was indirectly measured through a voltage measurement approach, the overall relative errors between the theoretical results and the experimental results under full removal conditions were found ranged from 2.08% to 6.87%, indicating the improved precision prediction capabilities of the proposed system model. The experimental results can be used as a set of references for further studies to offer performance assessment, precision prediction, process planning, and process condition monitoring for this important precision machining process.