A GPU Kernel Of NC Programming Based On Isogeometric Analysis Model And Its Application Optimization Algorithms

In order to meet the manufacturing requirement of advanced aerospace products,the numerical control programing technology is developing towards the direction of geometryphysics coupling optimization.However,due to the different demands of the geometric calculation and the physical analysis of the existing NC programming on the model,it is difficult to guarantee the stability and accuracy of the optimization process.Furthermore,the related algorithms exist some problems,such as inconsistent forms,low computational efficiency and poor robustness,which seriously restrict the development of the NC programming technology.Around the machining system model with geometry-physics fusion and the issues of NC programming calculation,this study lucubrates the isogeometric modeling method for machining system,the isogeometric kernel algorithm for NC programing and the application optimization algorithm for NC programing such as the tool-workpiece contact state,the tool dimension and the tool profile.The main contributions and innovation points are as follows:(1)Aiming at the issue of the geometry-physic fusion machining system model,a geometric modeling method of tool and workpiece,as well as a geometric analysis method for NC programming are proposed.By comparing the traditional finite element method and the isogeometric analysis method,the isogeometric analysis model is defined,and the isogeometric modeling method of the two key objects,the tool and the workpiece,is given in the machining system.On this basis,an isogeometric oriented to NC programming and specific to the problem of the geometry-physic coupling optimization in NC programming is proposed,which realizes the model unification in the process of the geometric problem calculation and the physical problem analysis.The successful generation of the trajectory in the deformation compensation machining process of thin-walled workpieces demonstrates the feasibility of the proposed modeling and analysis method in the geometry-physic coupling optimization in NC programming.(2)On the issues of low efficiency,poor precision and inconsistent forms of the existing NC programming algorithms,a GPU based isogeometric kernel algorithm for NC programming is presented.The kernel algorithm includes a distance monitoring and an optimized registration.By analyzing the computing method of the tool envelope surface and the tool axis trajectory surface,the distance monitoring establishes an interference model under the global contact of the tool-workpiece on the single point and the cutting line.Besides,the distance monitoring completes the computation of the interference distribution between the tool and the workpiece by employing the GPU parallel architecture,and solves the contradiction between the accuracy and the efficiency of the existing contact state computation.On the basis of the analysis of common tool control mode,the optimized registration builds an optimized registration model of tool position and orientation.Furthermore,by combining the distance monitoring computing results,the optimized registration realizes the adjustment of tool position and orientation by applying the sequence approximation method and the reduction operation in GPU,which avoids the issue of algorithm inconsistency caused by different offset approaches in the present tool position and orientation adjustment.The analysis results of the case study show that the isogeometric kernel algorithm and the GPU parallel technology can significantly improve the efficiency of NC programming.(3)In response to the issues of the interference detection as well as the tool position and orientation adjustment between the tool and the workpiece in NC programming,a fast interference judgment method and a position and orientation adjustment algorithm based on global contact state are put forward.Based on the convex hull of workpiece unit geometric model and the bounding box of the tool capsule,the fast interference judgment method utilizes equivalence transformation principle and the separating axis test to quickly extract potential interference regions on the workpiece in machining process.In addition,combining with three kinds of tool interference and distance monitoring kernel algorithms,the contact state computation and the interference judgement method between the tool and the workpiece are given.By analyzing common tool position and orientation adjustment strategies,the optimization adjustment algorithm establishes the tool position and orientation constraint region and the corresponding bounding box and obtains potential interference regions on the workpiece in tool adjustment process.Moreover,on the basis of the optimization registration kernel algorithm,it realizes the optimized adjustment of the tool position and orientation in potential interference regions.The case study indicates that two algorithms can quickly and efficiently achieve the interference judgement between the tool and the workpiece as well as the optimized adjustment of the tool position and orientation.(4)With respect to the issue of machining region division and the tool selection of boundary grooves on complex surfaces,a machinability chromatogram based division method for machining regions and a collaborative optimization method of tool size,position and orientation on the cutting line are proposed.The division method focuses on the critical parameter computation of boundary grooves such as the reachable region of tool axis and the maximum tool size in reachable region,the machinability critical parameter chromatographic representation based on boundary grooves,and the machining region division based on the chromatogram.The collaborative optimization method achieves the collaborative optimization of the tool size and tool axis vector on the cutting line in two steps.At first,the tool rigidity is express by the ratio of tool length to diameter,and the tool axis smoothness is described by the variation of the tool axis on adjacent tool positions.A tool size optimization model based on the tool rigidity and tool axis smoothness is established,and then the optimized tool size and the corresponding feasible region are gained.On this basis,a tool axis optimization model with the aim of smoothness under given tool sizes is built,which realizes the tool position and orientation optimization of the optimal tool size.The case study displays that there is no interference between the tool and the workpiece when the selected tool arrives at corresponding position and orientation,and there is no saltation between adjacent tool axes.(5)For the issue of the trajectory planning and tool contour designing in multi-axis side milling of free-form surface,a strip-width machining trajectory planning method based on bandwidth vector field,a global optimization method for machining trajectory,and a simultaneous optimization method of tool contour and machining trajectory are presented.Firstly,by constructing the maximum bandwidth vector field including the maximum machining bandwidth of the workpiece surface and the corresponding tool orientation distribution,streamline clusters in the parameter domain are generated,and selection method of the initial machining trajectory is given.The trajectory planning of strip-width side milling combined with the cutting efficiency is achieved.Secondly,through defining the control vertices on tool axis trajectory surface to express the side milling trajectory,the machining error is evaluated by calculating the interference between tool and workpiece on the cutting line.The tool axis smoothness is evaluated by tool axis trajectory energy,and a global optimization model of the multi-axis machining trajectory based on the machining error and tool axis smoothness is established,which realizes the side milling trajectory planning under the co-constraints of the accuracy and surface quality.Thirdly,the tool profile shape is described by control vertices on tool generatrix,and simultaneous optimization models of the tool contour and machining trajectory in single-line and multi-line side milling are established,respectively.These models realize shape design of the generatrix-type tool for free-form surface and the trajectory generation of the corresponding side milling.The case study and machining result manifest that the proposed algorithm can efficiently improve the machining efficiency,the machining accuracy and the surface quality under same conditions.