Research Of High Performance CNC Simulation Display Mode

CNC machining simulation is the process of a dynamic simulation by using the computer graphics display system for CNC machining process model of the workpiece, tool model, the tool’s tracking and the workpiece model by cutting and display. At present, there are a lot of CNC simulation research literatures based on traditional CPU implementation, due to the CPU serial nature, computing speed and memory bandwidth of disadvantage, which cannot make some algorithms on the CPU to rapid calculation. Therefore, this essay puts forward using GPU to replace traditional CPU for parallel computing, which satisfies the requirement of real-time display.In this paper, the five-axis CNC simulation of geometric modeling, cutting calculation, the dynamic display module was studied, the main contents are as follows:1.Geometric modeling aspect: study and analysis of several commonly used geometric modeling method, combining with the characteristics of five-axis CNC simulation, this paper improved the Marching Cubes algorithm, and gives general models of cutting tool. The improved Marching Cubes algorithm eliminates the object outside the empty and internal completely real voxel. Avoid redundant operations of the traditional MC algorithm for empty and full entity in iso-surface extraction, which improved the speed of the display of the machining surface.2.Cutting computing aspect: on CUDA platform, cutting calculation implemented in parallel from XOY direction, YOZ direction and XOZ direction of the workpiece and tool and thus cutting point array of data stored in Shared memory.3.Dynamic display aspect: first, implemented in parallel MC algorithm directly by CUDA, but the disadvantage is that iterates through all voxel causing the waste of resources; the boundary of the CUDA voxel rendering algorithms based on improved the Marching Cubes algorithm is proposed, and voxel status flag in each thread determines whether the voxel requires to be rebuilt, and then machining surface render by OpenGL’s interoperability with CUDA.