Identifying stable cutting conditions that minimise cycle times is essential for manufacturers to maximise return on investment (ROI). However, current industrial practices still rely heavily on trial and error and operator experience, leading to conservative parameter selection and increased production costs. Although computational tools can generate stability lobe diagrams (SLDs), they seldom provide stability predictions or parameter guidance when only geometric descriptions of the cutting tool are available. Therefore, this paper presents DIGICUT, a computational tool developed by Mondragon Unibertsitatea that predicts cutting stability and determines cutting conditions to reduce cycle time. Tool tip dynamics are predicted from STL files using a machine learning-assisted slicing algorithm combined with receptance coupling substructure analysis (RCSA). The predicted frequency response functions serve as inputs for constructing SLDs and calculating feed forces in milling operations. A genetic algorithm then searches for chatter-free machining conditions that meet the machine limits and achieve the shortest machining time. In this search, the algorithm adjusts machining parameters such as the number of teeth, spindle speed, feed per tooth, axial depth of cut, and radial depth of cut. Evaluated constraints include chatter, machine head power limits, and maximum axis feed forces. Computational validation studies in pocketing operations demonstrate that the software can predict the most suitable cutting conditions in 90-200 seconds with acceptable accuracy.
Chaux et al. (Thu,) studied this question.