The Mechanism And Experimental Research Of Additive/Subtractive Material Composite Molding Of Functionally Graded Structural Parts

Functionally graded composite materials are based on computer-aided material design and use advanced composite technology to make the composition,structure,and other elements of the material change continuously or stepwise in a certain direction during the material preparation process,so that the material properties and functions A heterogeneous composite material that also changes continuously or stepwise.Additive manufacturing technology can realize continuous change of material composition through real-time control of the proportion of different powders entering the molten pool,and then change the composition of some special parts of the part,so that different parts have different properties,which cannot be achieved by traditional casting and forging.Relying on additive technology to achieve material accumulation,relying on subtractive processing to improve surface quality,accurately control gradient structure,remove internal defects,improve stress state,and finally realize the formation of functionally graded parts through the interchange of two technologies.Based on the five-axis machining machine with additive and subtractive materials,the mechanism and experimental research of additive and subtractive material composite forming of functionally graded structural parts are explored as follows:(1)The single-factor experiment was used to analyze the effect of different ratios of 316L/Inconel718 powder on the cross section height H,cross section width W,and surface quality of the laser cladding layer with a gradient step size of 10%under the same process parameters.Influence,carry out single-channel single-layer single-factor experiment and orthogonal experiment,explore the influence of laser power P,powder feed rate Q,and scanning speed V on the width and height of the cladding layer,and determine the range of process parameters according to the results of the single-factor experiment.In the cross experiment,the experimental results are analyzed by the range method,and the orthogonal experiment is regression analysis.The process parameter regression model is established by using Q,V,and P as independent variables and W and H as the dependent variables,and the particle swarm algorithm and the regression model are combined and optimized.Process parameters.(2)Calculate the Z-axis lift based on the formula and the cladding section width and height data.Perform single-layer,multi-layer single-factor experiments and orthogonal experiments to explore the laser power P,powder feed rate Q,scanning speed V,and material reduction.The effect of the number of additive layers N on the density,density,hardness,residual stress,tensile properties,and fracture morphology of a single multi-layer additive sample,the single factor experiment mainly explores the number of additive layers before reducing these.The effect of performance is analyzed by orthogonal analysis of range analysis and stepwise regression analysis of orthogonal experimental data.(3)In the previous single-layer multi-layer single-factor experiment and orthogonal experiment,the density,hardness and residual stress data were used to build a model of the performance of the test piece based on the BP neural network.Four process parameters(graduated laser power,powder feed amount,scanning speed,and number of layers added before each reduction)that have important effects on the gradient structure are used as input layer variables.The input layer has 4 nodes.Density and residual stress are used as a single network output layer.Data parameters are normalized and denormalized.The parameters in the neural network model are set.The target learning accuracy is set.The maximum number of learning steps is the network learning rate.The single-factor experiments and orthogonal experiments done before use the input and output data to train the network.Hidden layer transfer function and the number of hidden layer nodes were used as variables to analyze the BP neural network model of density,hardness and residual stress.(4)Here,the material composition is selected to change linearly along the gradient direction and to transition with a gradient step size of 10%.Design and process longitudinal single-pass multi-layer pure additive 316L/Inconel718 functionally-gradient thin-walled parts and multiple longitudinal single-pass multiples.Multi-layer processing of 316L/Incone1718 functionally graded thin-walled parts,multi-lane single-layer 316L/Incone1718 functionally graded test pieces,and multi-layered multi-functionally graded blocks.Finally,the functionally graded torsion is machined after the forming is increased or decreased.Reasonably select the overlap ratio and measure the residual stress,microhardness,element distribution,tensile properties and tensile fracture analysis of functionally graded thin-walled pieces and blocks.