Simulation And Experiment Study Of The Precision Grinding Of Silicon Nitride Ceramic Based On Single Grain Cutting

Engineering ceramics have been widely used in industry, aerospace, nationaldefense and other important areas, due to its excellent performance on high strength,high hardness, wear resistance, high temperature resistance, chemical resistance, et al.Grinding with superhard fine abrasives is the most commonly method used tomachining engineering ceramics and other difficult-to-machine materials. Theessential of grinding is the material removal process by the interaction of abrasivegrains and material to be machined, while the abrasive grains are discrete distributionon the surface of grinding wheel. Every abrasive grain is carried out micro-grindingon the material to be machined. Therefore the research on mechanism of singleabrasive grain cutting has become an important means of understanding complexgrinding mechanism; the characteristic of grinding wheel surface topography hasbecome an indispensable prerequisite for modeling, simulation and optimization ofthe grinding process. As the forming mechanism of grinding is complex, the grindingprocess optimization has become a problem that plagued to manufacturing companies.In order to deepen and promote the application of engineering ceramics, it isnecessary to carry out deeply study on engineering ceramics precision grinding.In this paper, the precision grinding of silicon nitride ceramic is taken as theresearch subject. Based on the detecting of grinding wheel surface topography, singlegrain cutting experiments and grinding experiments, the modeling of grinding wheelsurface topography, single grain cutting simulation and grinding simulation arecarried out using numerical simulation technology. Furthermore, the grinding processparameters optimization is done with the improved genetic neural network algorithm,to achieve efficient precision grinding of silicon nitride ceramic. The main researchwork includes:1) Considering the distribution of grain size, position, angle, protrusion height, andthe effect of dressing, the grinding wheel surface topography model was established.With the management and counting for the digitization image about surfacetopography of diamond grinding wheel, the parameters of abrasive grains on grindingwheel were characterized by the statistical law of normal distribution, which werediameter, size distribution, equivalent diameter, volume density, area density andaverage space. The grinding wheel surface topography model was established with the selection of truncated octahedron as the abrasive grain shape. The models before andafter dressing were verified correct, that can be effectively used for grindingsimulation.2) The experiment and simulation system about single abrasive grain cutting wasconstructed. The single grain cutting tool that including wheel disk, single abrasivegrain block, balanced block was designed and prepared, and it was used for theorthogonal experiment of single abrasive grain cutting. Using JH-2materialconstitutive model to simulate the mechanical behavior of silicon nitride ceramic, andtruncated octahedron to simulation single diamond abrasive grain, the single abrasivegrain cutting simulation was put up. It revealed the influence rule of grinding wheelspeed, workpiece speed and cutting depth to grinding force and surface topography.The simulation system was verified feasible by experiment.3) The finite element simulation was carried out for the silicon nitride ceramicgrinding with diamond grinding wheel. With the diamond grinding wheel surfacemodel transformed from plane to round, the blocks sizes of grinding wheel andworkpiece dynamically defined according to the grinding depth and contact arc length,the simulation of diamond grinding wheel block grinding silicon nitride ceramic wasbuilt, to analyze the changes of stress, grinding force and sub-surface damage depthduring grinding, and discuss the influence rule of granularity and grinding wheelspeed, workpiece speed, grinding depth to grinding force and sub-surface damagedepth. The concluded results were similar to experiment. It showed that the finiteelement simulation of silicon nitride ceramic grinding with diamond grinding wheelwas correct and reasonable.4) The sub-surface damage depth forecasting model was established based on thegrinding wheel surface topography model and single abrasive grain cutting simulation.Through the finite element simulation of silicon nitride ceramic grinding withdiamond grinding wheel block, the relationship between stress and sub-surfacedamage depth was revealed. Fixing the formula about undeformed chip thickness ofdynamic effective abrasive, and combining the grinding wheel surface topographymodel and single abrasive grain cutting simulation, the relation model was establishedamong grinding wheel, grinding process parameters and sub-surface damage depth.5) The optimization and forecasting model was established between grindingprocess parameters and processing results. Through the experimental, it revealed theinfluence rule of maximum undeformed chip thickness and equivalent grindingthickness to grinding force, surface roughness and sub-surface damage depth, and established the empirical formula of equivalent grinding thickness and sub-surfacedamage depth. Using the improved genetic neural network, the optimization modelwas established for the grinding process parameters, to achieve the prediction andoptimization of silicon nitride ceramic grinding with diamond grinding wheel.6) The optimization system of spherical surface precision grinding on siliconnitride ceramic was established. The movement patterns, contact area and contact arclength were analyzed for spherical surface precision grinding on silicon nitrideceramic with grinding wheel normal tracking. It revealed the influence rule ofgranularity and radius of grinding wheel, grinding wheel speed, feeding speed andgrinding depth to surface roughness, sub-surface damage depth and processing timewith orthogonal experiment. The grinding simulation with grinding wheel surfacetopography model and single abrasive grain, and the improved genetic neural networkoptimization were applied in silicon nitride ceramic spherical grinding, to predict theworkpiece surface roughness, sub-surface damage depth and processing time, and tooptimize the process parameters.