Real-time composite smoothing interpolation for improved contour accuracy and surface quality in five-axis CNC machining
摘要
This study presents global and local composite smoothing interpolation (GLCSI), a real-time segment-wise detect-and-replace smoothing interpolator for five-axis computer numerical control machine tools. The core contribution is a dual-domain criticality criterion that unifies WCS tolerance violations and MCS kinematic infeasibility (via curvature/torsion thresholds derived from axis acceleration and jerk limits) into a segment-wise switching mask, enabling selective local regeneration without global re-planning. GLCSI generates a NURBS-based baseline global smoothing curve and applies the switching mask to deterministically replace only WCS/MCS-critical portions with locally regenerated geometry, while retaining noncritical portions as global blends. Experimentally, the GLCSI approach effectively addressed the limitations of conventional methods, including over-smoothing during purely global interpolation and excessive deceleration when purely local interpolation was in play. The geometric accuracy of GLCSI was comparable to that of local interpolation and the machining efficiency approached that of global interpolation. GLCSI significantly reduced the contour and orientation errors in critical regions and improved the final surface quality of a machined part. In our 1‑kHz controller implementation, GLCSI adapts the interpolation strategy based on WCS and MCS evaluations, enabling stable and precise five-axis motion in real time. Under the tested conditions, the experimental results suggest a practical accuracy-efficiency trade-off for industrially relevant high-speed, high-precision five-axis machining, provided that kinematic mapping and axis-limit parameters are available.