摘要
The occurrence of chatter in milling processeswas investigated in this study. The prediction of the sta-bility lobes of metal cutting processes requires a model ofthe cutting force and a model of the dynamic machine toolbehavior. Parameter uncertainties in the models may leadto significant differences between the predicted and mea-sured stability behavior. One approach towards robuststability consists of running a large number of simulationswith a random sample of uncertain parameters and deter-mining the confidence levels for the chatter vibrations,which is a time-consuming task. In this paper, an efficientimplementation of the multi frequency solution and theconstruction of an approximate solution is presented. Theapproximate solution requires the explicit calculation of themulti frequency solution only at a few parameter points,and the approximation error can be kept small. This studyfound that the calculation of the robust stability lobe dia-gram, which is based on the approximate solution, is sig-nificantly more efficient than an explicit calculation at allrandom parameter points. The numerically determinedrobust stability diagrams were in good agreement with theexperimentally determined stability lobes.
The occurrence of chatter in milling processeswas investigated in this study. The prediction of the sta-bility lobes of metal cutting processes requires a model ofthe cutting force and a model of the dynamic machine toolbehavior. Parameter uncertainties in the models may leadto significant differences between the predicted and mea-sured stability behavior. One approach towards robuststability consists of running a large number of simulationswith a random sample of uncertain parameters and deter-mining the confidence levels for the chatter vibrations,which is a time-consuming task. In this paper, an efficientimplementation of the multi frequency solution and theconstruction of an approximate solution is presented. Theapproximate solution requires the explicit calculation of themulti frequency solution only at a few parameter points,and the approximation error can be kept small. This studyfound that the calculation of the robust stability lobe dia-gram, which is based on the approximate solution, is sig-nificantly more efficient than an explicit calculation at allrandom parameter points. The numerically determinedrobust stability diagrams were in good agreement with theexperimentally determined stability lobes.
