Tool-tip frequency response function prediction of a 3-PRRU parallel spindle head based on Kriging interpolation

Chatter remains a prevalent challenge in the milling of thin-walled aerospace components, especially for the five-axis hybrid machine tools that employ a 3-PRRU parallel spindle head, often leading to degraded surface quality. The stability of the milling process is determined by the tool-tip frequency response function (FRF), which changes with the position of the spindle head within the workspace. However, direct experimental measurement of FRFs across the entire workspace is prohibitively time-consuming and impractical. To address this limitation, this paper proposes an efficient prediction method based on Kriging interpolation to estimate tool-tip FRFs at any position using a limited set of measured data. Modal parameters were identified from experimental FRF measurements taken at multiple tool-tip locations. A Kriging model was then constructed to interpolate FRFs at unmeasured positions. The accuracy of the predicted results was evaluated via leave-one-out cross-validation (LOOCV), demonstrating excellent agreement between predictions and measurements. The proposed approach significantly reduces the experimental effort required for workspace-dependent FRF characterization while maintaining high predictive accuracy. This method facilitates efficient generation of stability lobe diagrams and offers a practical solution for chatter suppression in 3-PRRU equipped hybrid machine tools, thereby supporting advancements in high-performance machining for aerospace applications.