A method for optimizing process parameters of coaxial powder feeding laser cladding

Abstract

This invention discloses a method for optimizing process parameters in coaxial powder-feed laser cladding, belonging to the field of laser cladding technology. The method includes the following steps: First, using laser cladding process parameters as independent variables and cladding layer performance evaluation indicators as target response values, relevant data are collected using an orthogonal experimental design. Then, using range analysis and variance analysis, the range analysis table, variance analysis table, and mean response diagram of the target response values ​​for the process parameters to the cladding layer performance evaluation indicators are calculated. Significantly influential process parameters are selected as optimization variables, and regression prediction models are established for each target response value and the optimization variables. A multi-objective optimization model for laser cladding process parameters is constructed. Finally, a fast non-dominated genetic algorithm is used to solve the Pareto front solution set of the multi-objective optimization series to obtain the optimal combination of laser cladding process parameters. This invention features a simple calculation process and high versatility; the composite material cladding layer obtained using this method has a smooth, continuous, and highly flat surface.

Description

Technical Field

[0001] This invention belongs to the field of laser cladding technology, specifically a method for optimizing process parameters of coaxial powder feeding laser cladding. Background Technology

[0002] 27SiMn steel is a hypoeutectoid alloy structural steel characterized by high strength, high toughness, good wear resistance, weldability, and machinability. Hydraulic supports are core mining equipment in coal mine exploration, and 27SiMn steel columns, as key load-bearing components, directly impact the production efficiency and safety of underground coal mine operations. Due to their long-term service in the complex and humid underground environment, the surface of these columns is highly susceptible to wear and corrosion, potentially leading to major safety accidents in coal mines. To improve the service life of these columns in harsh environments, electroplating is a traditional method for surface repair. However, the poor dispersion and coverage of electroplating solutions result in uneven coating thickness. The presence of microcracks and even through-cracks exacerbates corrosion, ultimately causing the entire hydraulic column to fail. Furthermore, the electroplating process generates waste that is harmful to human health and pollutes the environment.

[0003] Laser cladding offers advantages such as a dense and uniform cladding layer, minimal thermal impact on the substrate, controllable dilution rate, fast forming speed, and low pollution, making it a widely used remanufacturing technology in additive manufacturing. This method primarily uses a high-energy-density laser beam to rapidly melt and solidify powder and substrate materials, forming a high-quality cladding layer with excellent metallurgical bonding, effectively improving the wear resistance and corrosion resistance of the substrate surface. It can not only form complex precision parts but also be used for repairing parts damaged by manufacturing defects or misprocessing, as well as for surface modification. Due to its excellent low-carbon and energy-saving effects and significant economic benefits, it is currently widely used in petrochemical, marine engineering, transportation, and machinery manufacturing industries.

[0004] Laser cladding is the result of the coupled interaction of multiple parameters, and the performance and quality of the cladding layer depend on multiple evaluation indicators. This leads to a complex nonlinear relationship between process parameters and cladding layer quality, making it difficult to establish regression prediction models or combine them with machine algorithms for further optimization. Inappropriate process parameter matching can result in defects of varying degrees in the cladding layer, such as cracks, porosity, and surface unevenness, making it difficult for metal powder to form on the substrate surface, and even rendering the entire metal workpiece unusable. Therefore, optimizing laser cladding process parameters is of great significance for improving the quality of the cladding layer and enhancing economic efficiency.

[0005] Laser cladding technology has developed rapidly, with various equipment and technologies reaching a high level. However, the selection and optimization of laser cladding process parameters lack unified industry standards. Methods such as Taguchi method, response surface methodology, Kriging method, and full factorial analysis are widely used to establish models relating process parameters to cladding layer evaluation indicators. However, predictive models for multi-factor nonlinear data are often difficult to establish, and models built using a large amount of experimental data often have significant errors, failing to meet the accuracy requirements of actual production. Furthermore, the application of intelligent algorithms such as simulated annealing, neural networks, particle swarm optimization, and support vector machines to optimize process parameters can lead to getting stuck in local optima or insufficient data density, complicating the computational process. Summary of the Invention

[0006] To address the aforementioned technical problems, the purpose of this invention is to provide a method for optimizing process parameters in coaxial powder-feeding laser cladding. This method establishes a high-precision regression prediction model between process parameters and response indicators, and uses the Non-Dominated Genetic Algorithm (NSGA-II) to solve a series of Pareto solutions for multi-objective optimization, thereby obtaining a combination of laser cladding process parameters that meets the requirements of actual working conditions.

[0007] To achieve the above objectives, the present invention adopts the following technical solution:

[0008] A method for optimizing process parameters of coaxial powder-feeding laser cladding includes the following steps:

[0009] S1: Determine n laser cladding process parameters as independent variables and m cladding layer performance evaluation indicators as target response values. Select the range of each process parameter and p levels, design an n-factor p-level orthogonal experimental scheme, and conduct laser single-pass cladding test to complete relevant data collection.

[0010] S2: Using range analysis and variance analysis, calculate the range analysis table, variance analysis table, and mean response diagram of the target response values ​​for the process parameters on the performance evaluation index of the cladding layer; determine the weight ranking of the influence of process parameters on the performance evaluation index of the cladding layer; and judge the significance of the influence of process parameters on the response index of the cladding layer, and select significant process parameters.

[0011] S3: Using the significant process parameters selected in step S2 as optimization variables and the cladding layer performance evaluation index determined in step S1 as target response values, establish a regression prediction model between each target response value and the optimization variables, and verify whether the accuracy of the model meets the requirements of subsequent multi-objective optimization of process parameters.

[0012] S4: Using the range of significant process parameters selected in step S2 as constraints, determine the maximum / minimum requirements of the expected response values ​​of each objective, optimize each process parameter using the minimum value solution method, and construct a multi-objective optimization model for laser cladding process parameters;

[0013] S5: Optimize process parameters through a multi-objective optimization model, stratify individuals in the population based on their non-dominated solution level, sort the solutions in the non-dominated solution set by rank in ascending order, and select the better individuals in the population.

[0014] S6: After quicksort, solutions with the same rank are sorted in descending order of cluster distance, and individuals in the quasi-Pareto domain are extended to the entire Pareto domain.

[0015] S7: Use the fast non-dominated sorting genetic algorithm (NSGA-II) to solve for the Pareto front solution set of the optimization variables and the target response value. Combined with the actual production requirements, select the optimization variables corresponding to the target response value that meets the operating conditions, which is the optimal combination of process parameters.

[0016] Furthermore, in step S1, the selection of the range of each process parameter and p levels, and the design of an n-factor p-level orthogonal experimental scheme, are determined based on the equipment usage experience and actual production requirements. A corresponding blank column is set to determine the sum of squares of deviations caused by random errors during variance analysis.

[0017] Furthermore, in step S1, the surface of the workpiece needs to be polished before the cladding test, and the surface oil stains need to be cleaned with anhydrous ethanol and acetone; at the same time, the alloy powder is placed in an oven for heating and heat preservation.

[0018] Furthermore, in step S1, after the laser single-pass cladding test is performed and relevant data is collected, the sample should be cooled to room temperature after the laser cladding test. The cross-section of the cladding layer should be obtained by wire cutting perpendicular to the laser cladding direction. After grinding, polishing and etching, the sample is prepared. The width and height of the cladding layer and the molten pool, as well as the area of ​​the cladding layer and the molten pool, are measured, and the response value of the cladding layer performance evaluation index of each group of samples is calculated.

[0019] Furthermore, in step S2, a range analysis table of the cladding layer performance evaluation index is calculated using the range analysis method, where the range R = max{K1,…K i ,…,K p}-min{K1,…K i ,…,K p}, K i This represents the mean of the target response when the level is i (i=1,2,3…p) in the column containing each factor; when using the analysis of variance method, the F value of each process parameter needs to be calculated and compared with the critical F distribution table, while requiring the P value to be less than 0.05.

[0020] Furthermore, in step S3, a regression prediction model is established between each target response value and the optimization variable. The regression prediction model is as follows:

[0021]

[0022] Where Y represents the target response value; X ij X represents the level value of factor i (j); i Represented as process parameter optimization variables; The index represents the influence of process parameters; b is the model revision coefficient; β0 and β i represents the fitted value of the regression coefficient; p represents the number of levels of the process parameter; t represents the number of process parameter optimization variables.

[0023] Furthermore, in step S3, the accuracy of the model is checked to see if it meets the requirements for subsequent multi-objective optimization of process parameters. Specifically, the correlation square coefficient between each objective response value and the regression prediction model of the optimization variables is calculated. When the correlation square coefficient is greater than 0.9, the accuracy requirements for subsequent multi-objective optimization of process parameters are met.

[0024] Furthermore, in step S4, when optimizing each process parameter using the minimum value solution method, it is necessary to take the opposite mathematical equation for the target response value to be maximized. The constructed multi-objective optimization model for laser cladding process parameters is as follows:

[0025]

[0026] Among them, X i f represents the process parameter optimization variables (1≤i≤t); k Regression prediction model of target response value and optimization variables (1≤k≤m); t represents the number of process parameter optimization variables; m is the number of target response values.

[0027] Furthermore, in step S5, the non-dominated solution set constraints are as follows:

[0028]

[0029] Where P is the non-dominated set; M is the population; Q is the position of the offspring in space; x k It can be any factor.

[0030] Furthermore, in step S6, the solutions with the same rank after quicksort are sorted in descending order of cluster distance as follows:

[0031]

[0032] Where, Li d Let L[i+1] represent the clustering distance of any individual i. m This represents the m-th objective function value for the (i+1)-th individual; and Let m represent the maximum and minimum values ​​of the objective function at the m-th position in the set.

[0033] Compared with the prior art, the present invention has the following beneficial effects:

[0034] The proposed method for optimizing process parameters in coaxial powder-feeding laser cladding combines orthogonal experimental design, regression prediction models of process parameters and target response values, and a fast non-dominated sorting genetic algorithm. This method can conveniently and quickly determine the weight ranking and significance of the influence of process parameters on the performance evaluation indicators of the cladding layer. Based on this, a regression prediction model with high accuracy, significant correlation, and good predictive ability can be established. Furthermore, a multi-objective optimization model for laser cladding process parameters can be constructed, and the optimal combination of process parameters can be obtained quickly and accurately through the fast non-dominated sorting genetic algorithm.

[0035] The calculation process of this invention is simple and highly versatile, avoiding getting stuck in local optima during the optimization process; it can accurately determine the optimal parameters without a large number of experiments, saving a lot of time in the early stages of pre-experimentation and parameter selection based on experience, resulting in low energy consumption and low production costs; the composite material cladding layer produced using the optimized process parameters has a smooth, continuous, and highly flat surface, clear fusion lines between the cladding layer and the substrate, high degree of bonding, stable mechanical properties, and excellent wear resistance and corrosion resistance. Attached Figure Description

[0036] Figure 1 This is a flowchart of the method for optimizing process parameters of coaxial powder feeding laser cladding in an example of the present invention;

[0037] Figure 2 This is a schematic diagram of the cross-sectional geometric features of the cladding layer in an example of the present invention;

[0038] Figure 3 This is a graph showing the response of each process parameter to the average dilution rate of the cladding layer in an example of the present invention.

[0039] Figure 4 This is a graph showing the response of various process parameters to the average aspect ratio of the cladding layer in an example of the present invention.

[0040] Figure 5 This is a response diagram of the average microhardness of the cladding layer to various process parameters in an example of the present invention.

[0041] Figure 6 This is the Pareto front solution set diagram for the process parameter optimization variables in this invention example;

[0042] Figure 7 This is a Pareto front solution set diagram of the target response value of the cladding layer in an example of the present invention;

[0043] Figure 8 These are macroscopic morphology images of the cladding layers in the comparative test group and the optimal process parameter group in the examples of this invention;

[0044] Figure 9 for Figure 8 Microstructure diagram of region a;

[0045] Figure 10 for Figure 8 Microstructure diagram of region b;

[0046] Figure 11 for Figure 8 Microstructure diagram of region c;

[0047] Figure 2 In the diagram, 1—substrate; 2—heat-affected zone; 3—substrate melting zone; 4—cladding zone. Detailed Implementation

[0048] To better explain the process technology solution of the present invention, the present invention will be described in complete and detailed manner below with reference to the accompanying drawings in the embodiments of the present invention and through specific implementation methods.

[0049] Reference Figure 1 A method for optimizing coaxial laser cladding process parameters, specifically including the following steps:

[0050] S1: Using laser power, scanning speed, powder feeding rate, and protective gas flow rate as independent variables, and cladding layer dilution rate, aspect ratio, and microhardness as performance evaluation indicators of the cladding layer, the parameter ranges and levels of laser power, scanning speed, powder feeding rate, and protective gas flow rate are selected to design a 5-factor, 4-level L16 (4 5 An orthogonal experimental design (with one blank column) was used to conduct laser single-pass cladding tests on GS-Fe01 alloy powder to complete relevant data collection. The design results are shown in Tables 1 and 2.

[0051] Table 1: Levels of Experimental Factors

[0052]

[0053] Table 2: 5-factor, 4-level orthogonal experimental design L16 (4 5 )

[0054]

[0055] Specifically, before the cladding test, the workpiece needs to be polished to remove the surface oxide scale and impurities, exposing the fresh metal inside, and the surface oil stains need to be cleaned with anhydrous ethanol and acetone; at the same time, the alloy powder is placed in an oven and heated and kept at a temperature for 3-5 hours to prevent the alloy powder from agglomerating and clogging the powder feeding pipe.

[0056] Specifically, to complete the data acquisition for the single-pass laser cladding test, the sample should be cooled to room temperature after the laser cladding test. A wire cut perpendicular to the laser cladding direction should be used to obtain the cross-section of the cladding layer. The cross-section should then be polished sequentially with sandpaper of increasing grit, and finally polished to a mirror finish using diamond polishing paste on a metallographic sample polishing machine. Finally, the cross-section of the cladding layer should be etched with an etchant to complete the sample preparation. (Reference) Figure 2 The cross-section of the cladding layer was observed using a microscope, and the width, height, and area of ​​the cladding layer and the molten zone between the cladding layer and the substrate were measured. The response values ​​of the cladding layer performance evaluation indicators for each group of samples were calculated. Among these, in... Figure 2 In the diagram, W represents the width of the cladding layer, H represents the height of the cladding layer, W1 represents the width of the molten zone of the substrate, h represents the height of the molten zone of the substrate, A1 represents the area of ​​the cladding layer, and A2 represents the area of ​​the molten zone of the substrate. Using a Vickers microhardness tester, the microhardness is measured sequentially at equal intervals starting from the surface of the cladding layer, and the average value is taken. The roughness of the cladding layer cross-section, the interval distance, and the loading conditions are determined according to the type of cladding powder material and the macroscopic dimensions of the cladding layer, referring to the national standard for Vickers hardness testing.

[0057] The type of laser cladding equipment, the type and particle size of the alloy powder, and the selection of the range of independent variables all affect the setting of non-independent process parameters in the laser cladding process. In this example, based on previous pre-experiment experience, the spot diameter was adjusted to Φ2.5 mm and the powder feeding gas flow rate was 7.5 L / min during the experiment. Meanwhile, the range of independent process parameters and the number of experimental factor levels were determined based on equipment usage experience and actual production requirements; details are shown in Table 1.

[0058] S2: Using range analysis and variance analysis, calculate the range analysis table, variance analysis table, and mean response diagram of the target response value for the effect of process parameters on the performance evaluation index of the cladding layer; determine the weight ranking of the influence of process parameters on the performance evaluation index of the cladding layer; and judge the significance of the influence of process parameters on the response index of the cladding layer, and select significant process parameters.

[0059] Among them, the range R = max{K1,…K} of the cladding layer performance evaluation index is calculated using the range analysis method. i ,…,K p}-min{K1,…K i ,…,K p}, where K i This represents the mean of the target response when the level is i (i=1,2,3…p) in the column containing each factor; the F-value of each process parameter is calculated using analysis of variance and compared with the F-critical distribution table. 0.05 The comparison is (3,3)=9.28, and the p-value is required to be less than 0.05.

[0060] Among them, the range analysis table and variance analysis table of the effect of process parameters on the performance evaluation index of the cladding layer were calculated, and the calculation results are shown in Tables 3 and 4. According to the mean response diagram of the target response value, the influence trend of laser power, scanning speed, powder feeding rate and protective gas flow rate on the dilution rate, aspect ratio and microhardness of the target response value of the cladding layer can be clearly seen. Figures 3 to 5 Analysis revealed that laser power, scanning speed, and powder feeding rate significantly affected the target response value of the cladding layer, while the protective gas flow rate was an insignificant process parameter.

[0061] Table 3: Range Analysis Table

[0062]

[0063] Table 4: Analysis of Variance Table

[0064]

[0065] S3: Using the significant process parameters selected in step S2 as optimization variables and the cladding layer performance evaluation index selected in step S1 as target response values, establish regression prediction models between each target response value and the optimization variables, and verify whether the accuracy of the models meets the requirements for subsequent process parameter optimization. The regression prediction models are as follows:

[0066]

[0067] Where Y represents the target response value; X ij X represents the level value of factor i (j); i Represented as process parameter optimization variables; The index represents the influence of process parameters; b is the model revision coefficient; β0 and β i represents the fitted value of the regression coefficient; p represents the number of levels of the process parameter; t represents the number of process parameter optimization variables.

[0068] Since the regression prediction model involves many parameters, the Levenberg-Marquardt method, a nonlinear least squares approach, is introduced to address the parameter evaluation of the nonlinear model. This avoids the ill-conditioned and non-singular problems that occur with the Gauss-Newton method, providing more stable and accurate increments. The entire solution process can be optimized using the 1stOpt software.

[0069] Dilution rate—objective function of process parameters:

[0070]

[0071] Aspect Ratio—Objective Function for Process Parameters:

[0072]

[0073] Microhardness—Objective function of process parameters:

[0074]

[0075] Where Y1 represents the dilution rate; Y2 represents the aspect ratio; Y3 represents the microhardness; X1 represents the laser power; X2 represents the scanning speed; and X3 represents the powder feeding rate.

[0076] In addition, the correlation square coefficients of the regression prediction models for dilution rate, aspect ratio, microhardness, laser power, scanning speed, and powder feeding rate were calculated to verify the reliability and interpretability of the models in fitting the experimental data. Generally, a correlation square coefficient greater than 0.9 is considered to meet the accuracy requirements for subsequent multi-objective optimization of process parameters. The correlation square coefficients for the dilution rate, aspect ratio, and microhardness models were verified to be 0.94, 0.95, and 0.92, respectively, all meeting the accuracy requirements.

[0077] S4: Using the range of significant process parameters selected in step S2 as constraints, determine the maximum / minimum requirements of each target response value, optimize each process parameter using the minimum value solution method, and construct a multi-objective optimization model for laser cladding process parameters.

[0078] Constraints:

[0079] st

[0080] Excessive dilution can significantly reduce the performance of the cladding layer, even leading to defects such as cracks; therefore, the dilution rate should be as small as possible. A small aspect ratio will reduce the wettability of the cladding layer; therefore, the aspect ratio should be maximized. Microhardness, as an evaluation index of the mechanical properties of the cladding layer, is desired to be as high as possible. It should be noted that when optimizing the process parameters using the minimum value method in this invention, the aspect ratio and microhardness target models need to be expressed with opposite mathematical equations.

[0081] Multi-objective optimization model for laser cladding process parameters:

[0082]

[0083] S5: Optimize process parameters using a multi-objective optimization model. Based on the non-dominated solution level of each individual, stratify the population. Sort the solutions in the non-dominated solution set by rank in ascending order, and select the better individuals from the population. The fast non-dominated solution set constraints are as follows:

[0084]

[0085] Where P is the non-dominated set; M is the population; Q is the position of the offspring in space; x k It can be any factor.

[0086] S6: After quicksort, solutions with the same rank are then sorted in descending order of cluster distance, expanding individuals in the quasi-Pareto domain to the entire Pareto domain. The descending sorting principle is as follows:

[0087]

[0088] Where, Li d Let L[i+1] represent the clustering distance of any individual i. m This represents the m-th objective function value for the (i+1)-th individual; and Let m represent the maximum and minimum values ​​of the objective function at the m-th position in the set.

[0089] S7: Use the Fast Non-Dominated Sorting Genetic Algorithm (NSGA-II) to solve for the Pareto front solution set of the optimization variables and objective response values, such as... Figure 6 and 7 As shown; combining actual production requirements, select the optimization variables that meet the target response requirements under the working conditions, which is the optimal combination of process parameters.

[0090] In theory, every point in the Pareto front solution set is the optimal solution. However, excessively pursuing superior performance in one aspect may come at the cost of sacrificing other properties of the cladding layer. Considering practical production requirements, this example limits the dilution rate to between 15% and 25% to obtain a cladding layer with good metallurgical bonding and dense microstructure; an aspect ratio greater than 3 is beneficial for improving the surface smoothness of the cladding layer, ensuring smooth subsequent multi-layer overlaps; and a microhardness greater than 700 HV is required. 0.5 This satisfies the wear resistance requirements of the hydraulic support column. Therefore, within the range that meets the actual working conditions, a laser power of 742 W, a scanning speed of 234 mm / min, and a powder feeding rate of 1.2 R / min were selected as the optimal combination of process parameters for multi-objective optimization. The optimized predicted cladding layer dilution rate was 19.8%, the aspect ratio was 3.005, and the microhardness was 708.51 HV. 0.5 .

[0091] Furthermore, verification experiments were conducted under optimal process parameters of 742 W laser power, 234 mm / min scanning speed, and 1.2 R / min powder feeding rate. The measured cladding layer dilution rate was 19.1%, the aspect ratio was 3.142, and the microhardness was 727.64 HV. 0.5 Comparing the predicted and measured values ​​of cladding layer dilution rate, aspect ratio, and microhardness, the two values ​​show good agreement. The prediction deviations are 3.7%, 4.4%, and 2.6%, respectively, which meet the actual process requirements.

[0092] Furthermore, to verify the effect of the optimal combination of process parameters on the performance improvement of the cladding layer, the sixth group (test number 6 in Table 2), which exhibited good bonding, no obvious cracks or pores, and a superior target response value in step S1, was selected as the control test group. The measured values ​​of cladding layer dilution rate, aspect ratio, and microhardness obtained from the verification test were compared with those of the control test group. The results show that the performance evaluation indicators of the cladding layer obtained using the optimization method of this invention are significantly better than those of the control test group, as shown in Table 5. The improvement rates of cladding layer dilution rate, aspect ratio, and microhardness were 33.2%, 24.5%, and 12.3%, respectively.

[0093] Table 5: Comparison of cladding layer performance evaluation indicators after optimization of optimal process parameters

[0094]

[0095] In addition, the macroscopic morphology of the cladding layer in the comparative experimental group and the optimal process parameter group is as follows: Figure 8 As shown, the cladding layer morphology in the comparative test group was a combination of convex and crescent shapes. This was mainly due to the significant difference in energy required for melting the powder material and the substrate during laser cladding. The process parameters in the comparative test group resulted in a higher powder density per unit area, which shielded the laser beam. Furthermore, most of the laser energy was used to melt the powder material, leading to the continuous accumulation of cladding material and the resulting convexity. The cladding layer morphology under the optimal process parameters was a double crescent shape, with a smooth, continuous, and highly flat surface, free of semi-melted adhesion points. The substrate formed an effective molten pool, and the powder material entered the molten pool uniformly and continuously to form an effective cladding layer. Moreover, the fusion line between the cladding layer and the substrate was clear, with a high degree of bonding. The macroscopic morphology quality of the cladding layer was significantly better than that of the comparative test group.

[0096] Microstructure of different regions of the cladding layer cross section, such as Figures 9 to 11 As shown, the cladding layer exhibits a uniform and dense structure, with good adhesion to the substrate, and is free from defects such as burning, cracks, and porosity. Figure 9 It can be seen that the microstructure at the bottom of the comparative test group is mainly composed of coarse dendritic crystals and columnar crystals. This is because the bottom of the cladding layer is in contact with the substrate, heat transfer is faster, and the tissue grows fully. Figure 10 , Figure 11 In the middle and top microstructure of the cladding layer, the solidification rate increases and the temperature gradient decreases, resulting in a significant reduction in the number of dendritic and columnar crystals and the formation of more equiaxed crystals. Furthermore, due to the influence of heat accumulation and changes in heat dissipation, the growth direction of the microstructure becomes disordered. In contrast, the optimal process parameter set, with its reasonable matching of process parameters, accelerates crystal cooling, increases the number of nucleation points, and makes the crystal morphology and size more uniform. This results in fewer dendritic and columnar crystals in the cladding layer microstructure, a higher proportion of equiaxed crystals, finer grain size, a denser microstructure, and better cladding layer forming quality.

[0097] The above description is merely a preferred embodiment of the present invention and does not limit the scope of the present invention's process technology. For those skilled in the art, any modifications, equivalent substitutions, and improvements made based on the present invention's process technology and concept will not affect the optimization effect of this example. Therefore, the patent protection scope of the present invention should be determined by the content of the claims.

Claims

1. A method for optimizing process parameters of coaxial powder-feeding laser cladding, characterized in that, Includes the following steps: S1: Determine n laser cladding process parameters as independent variables and m cladding layer performance evaluation indicators as target response values. Select the range of each process parameter and p levels, design an n-factor p-level orthogonal experimental scheme, and conduct laser single-pass cladding test to complete relevant data collection. S2: Using range analysis and variance analysis, the range analysis table, variance analysis table, and mean response diagram of the target response values ​​of the process parameters on the performance evaluation index of the cladding layer are calculated. Determine the weight ranking of the influence of process parameters on the performance evaluation index of the cladding layer; and judge the significance of the influence of process parameters on the response index of the cladding layer, and select significant process parameters. S3: Using the significant process parameters selected in step S2 as optimization variables and the cladding layer performance evaluation index determined in step S1 as target response values, establish regression prediction models between each target response value and the optimization variables, and verify whether the accuracy of the models meets the requirements for subsequent multi-objective optimization of process parameters; the regression prediction models are as follows: ； Where Y represents the target response value; X ij X represents the level value of factor i (j); i Represented as process parameter optimization variables; The index represents the influence of process parameters; b is the model revision coefficient; β0 and β i represents the fitted value of the regression coefficient; p represents the number of levels of the process parameter; t represents the number of process parameter optimization variables; S4: Using the range of the significant process parameters selected in step S2 as constraints, determine the maximum / minimum requirement of each target response value, and optimize each process parameter using the minimum value solution method to construct a multi-objective optimization model for laser cladding process parameters; when optimizing each process parameter using the minimum value solution method, it is necessary to take the opposite mathematical equation for the target response value to be maximized. The constructed multi-objective optimization model for laser cladding process parameters is as follows: ； Among them, X i f represents the process parameter optimization variables (1≤i≤t); k Regression prediction model of target response value and optimization variables (1≤k≤m); t represents the number of process parameter optimization variables; m is the number of target response values; S5: Optimize process parameters using a multi-objective optimization model. Based on the non-dominated solution level of each individual, stratify the population. Then, sort the solutions in the non-dominated solution set by rank in ascending order to select the superior individuals from the population. The constraints of the non-dominated solution set are as follows: ； Where P is the non-dominant set; M x For the population; Q is the position of the offspring in space; x n For any factor; S6: After quicksort, solutions of the same rank are then sorted in descending order of cluster distance, expanding individuals in the quasi-Pareto domain to the entire Pareto domain; the principle for sorting solutions of the same rank in descending order of cluster distance after quicksort is as follows: ； Where, Li d Let L[i+1] represent the clustering distance of any individual i. m This represents the m-th objective function value for the (i+1)-th individual; and Let m represent the maximum and minimum values ​​of the objective function at the m-th position in the set. S7: Use the fast non-dominated sorting genetic algorithm to solve for the Pareto front solution set of the optimization variables and the target response value. Combined with the actual production requirements, select the optimization variables that meet the target response value of the working conditions, which is the optimal combination of process parameters.

2. The method for optimizing process parameters of coaxial powder-feeding laser cladding according to claim 1, characterized in that, In step S1, the selection of the range of each process parameter and p levels, and the design of an n-factor p-level orthogonal experimental scheme, are determined based on the equipment usage experience and actual production requirements. Corresponding blank columns are set to determine the sum of squares of deviations caused by random errors during variance analysis.

3. The method for optimizing process parameters of coaxial powder-feeding laser cladding according to claim 1, characterized in that, In step S1, the surface of the workpiece needs to be polished before the cladding test, and the surface oil stains need to be cleaned with anhydrous ethanol and acetone; at the same time, the alloy powder is placed in an oven for heating and heat preservation.

4. The method for optimizing process parameters of coaxial powder-feeding laser cladding according to claim 1, characterized in that, In step S1, a single-pass laser cladding test is conducted to complete the relevant data acquisition. After the laser cladding test, the sample should be cooled to room temperature and the cross-section of the cladding layer should be obtained by wire cutting perpendicular to the laser cladding direction. The sample is then prepared after grinding, polishing, and etching. The width and height of the cladding layer and the molten pool, as well as the area of ​​the cladding layer and the molten pool, are measured, and the response value of the cladding layer performance evaluation index for each group of samples is calculated.

5. The method for optimizing process parameters of coaxial powder-feeding laser cladding according to claim 1, characterized in that, In step S2, a range analysis table of the cladding layer performance evaluation index is calculated using the range analysis method, where the range R = max{K1,…K i ,…,K p }-min{K1,…K i ,…,K p }, K i This represents the mean of the target response when the level is i (i=1,2,3…p) in the column containing each factor; when using the analysis of variance method, the F value of each process parameter needs to be calculated and compared with the critical F distribution table, while requiring the P value to be less than 0.

05.

6. The method for optimizing process parameters of coaxial powder-feeding laser cladding according to claim 1, characterized in that, In step S3, the accuracy of the model is checked to see if it meets the requirements for subsequent multi-objective optimization of process parameters. Specifically, the correlation square coefficient between the response value of each objective and the regression prediction model of the optimization variable is calculated. When the correlation square coefficient is greater than 0.9, the accuracy requirement for subsequent multi-objective optimization of process parameters is met.

Images