Least-squares Based Parameter Identification for a Function-related Surface Optimisation in Micro Ball-end Milling

Micro milling is a flexible technique for the production of micro mechanical components like dies and moulds and process control is the key to reach the strong production requirements. Requirements, given by engineers and designers, are addressed mainly to the functional performance of the produced part, therfore topographic features are most decisive. Surface parameters, mainly of statistical origin, have been used for a long time in surface characterisation and process monitoring. Furthermore, it is known that these parameters correlate with the desired functional behaviour, but this knowledge is usually not used for a deterministic process design, uneconomic try and error approaches are still common.Mathematical investigations can use the full process flexibility for an in-process functionalization by selecting optimal conditions and process parameters with respect to a set of relevant surface parameters. In this study, micro ball-end milling is investigated and process parameters in order meet a predefined bearing ratio curve as accurately as possible are identified. Therefore, a mechanistic surface generation model has been developed and is used as a forward model for an iterative optimisation. Static and dynamic process geometry and a micro mechanical material removal operator are the main features of the model. In the first part of the paper the semi-empirical model is calibrated for certain tool and workpiece materials. In the second part optimal feed speed and width of cut are determined. Finally, an experimental validation is presented and the comparison of the predefined, the predicted and the experimental bearing ratio curves shows a good agreement.