A GPU-based approach for 5-axis flute grinding of end-mills with complex grinding wheel

Abstract

End-mills are used widely in industry, in which efficient chip evacuation and enhanced tool performance are directly influenced by the flute shapes. Generally, complex grinding wheels are employed to create intricate flute shapes with CNC grinding. In the kinematic of flute grinding, several transcendental equations with high non-linearity is derived and required to be solved. In some cases, the explicit expression for flute cannot be derived and the numerical method are generally applied, which requires various derivation or numerical program for different flute shapes. To address these challenges, this paper proposed a GPU-based approach for 5-axis flute grinding of end-mills with complex grinding wheel. In this method, a mesh model of the grinding wheel was established to transform the above complex equations into finding a set of point cloud which satisfied the envelope condition. To accelerate the computation time for searching those point cloud, a generalized GPU parallel processing algorithm were used to execute multiple compute threads. The validity of the approach was verified through a series of experiments. It demonstrated that this method achieved remarkable precision and broad applicability, fulfilling the diverse requirements of various grinding wheels and flute shapes. Furthermore, the high efficiency and versatility of this approach make it have great potential in the application of flute-grinding with various complex wheel in real-time path planning.