Method for optimizing process parameters of electric arc additive manufacturing based on material heterogeneity

Abstract

The present application relates to the field of arc additive manufacturing material performance optimization, and specifically provides an arc additive manufacturing process parameter optimization method based on material non-uniformity, which comprises the following steps: S1, determining the process parameters of arc additive manufacturing material, and preparing an arc additive sample; S2, quantifying the anisotropy and non-uniformity indexes of the arc additive sample by using an ultrasonic detection method; S3, modeling the anisotropy index and non-uniformity index of the arc additive sample; S4, optimizing the process parameters according to the anisotropy index model and the non-uniformity index model. The present application maximally reduces the non-uniformity of arc additive manufacturing material within the effective range of main process control parameters, quantifies the anisotropy and non-uniformity of the material based on the ultrasonic attenuation coefficient, establishes the correlation model of the process parameters and the anisotropy and non-uniformity indexes, and determines the process parameters capable of preparing the material with the minimum anisotropy and non-uniformity by using an optimization strategy.

Description

Technical Field

[0001] This invention belongs to the field of material performance optimization technology in electric arc additive manufacturing, and specifically relates to a method for optimizing process parameters in electric arc additive manufacturing based on material non-uniformity. Background Technology

[0002] Metal additive manufacturing (AM) is an emerging manufacturing technology increasingly used in rapid manufacturing, repair, and remanufacturing. Arc additive manufacturing (WAAM) technology uses an electric arc as a heat source and wire as a raw material to build metal parts by stacking wires layer by layer on a substrate. WAAM parts undergo a complex thermal cycling process, resulting in a non-uniform microstructure. Furthermore, for aluminum alloys, the significant difference in hydrogen solubility between solid and liquid aluminum inevitably leads to porosity during solidification in the arc additive manufacturing process. This unique microstructure and porosity affect the material's properties, posing a significant challenge to the application of WAAM materials in critical fields.

[0003] To utilize WAAM parts as load-bearing structural components in engineering applications, it is necessary to manufacture WAAM parts with performance comparable to or better than conventionally manufactured parts. For additively fabricated parts, improvements in material properties largely depend on the optimization of process parameters. Due to the path dependence and complex thermal history of additive manufacturing, additively manufactured materials have been observed to exhibit varying degrees of anisotropy and inhomogeneity.

[0004] To optimize the anisotropy and inhomogeneity of WAAM material properties in a multidimensional parameter space, two aspects of research are required: quantifying the anisotropy and inhomogeneity of the material, and establishing a model that correlates the quantification index with process parameters. For the former, traditional quantification processes involve metallographic analysis or destructive testing-based mechanical property assessments sampling at one or more locations. To obtain statistically sufficient information about the formed parts, sampling at multiple locations is necessary to assess the spatial homogeneity of material properties, which is very time-consuming and labor-intensive for large parts. For the latter, the lack of sufficient statistical information prevents the implementation of a multi-objective optimization model for minimizing the anisotropy and inhomogeneity of WAAM materials. Therefore, this invention proposes a method based on ultrasonic nondestructive testing that can effectively reduce the anisotropy and inhomogeneity of arc additive manufacturing materials. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a method for optimizing process parameters in arc additive manufacturing based on material inhomogeneity. This invention minimizes the inhomogeneity of materials in arc additive manufacturing within the effective range of key process control parameters. It quantifies the anisotropy and inhomogeneity of materials using parameters based on ultrasonic attenuation coefficients, establishes a correlation model between process parameters and anisotropy and inhomogeneity indices, and uses optimization strategies to determine the process parameters that can prepare materials with minimal anisotropy and inhomogeneity.

[0006] To achieve the above objectives, the present invention discloses the following technical solution:

[0007] A method for optimizing process parameters in arc additive manufacturing based on material inhomogeneity, comprising:

[0008] S1: Determine the process parameters for the arc additive manufacturing material and prepare arc additive test samples:

[0009] S11: Determine the process parameters that affect the anisotropy and inhomogeneity of material properties; the process parameters are welding torch travel speed A, arc current B, and shielding gas flow rate C;

[0010] S12: Prepare metal parts using an electric arc additive manufacturing process based on cold metal transition. By setting n different combinations of process parameters, prepare n electric arc additive samples, where n is the number of process parameter combinations in step S11.

[0011] S2: The anisotropy and inhomogeneity of the arc additive manufacturing sample are quantified using ultrasonic testing methods, which includes the following sub-steps:

[0012] S21: Use the ultrasonic pulse echo method to obtain the ultrasonic data of the arc additive sample in step S12, and calculate the total energy loss of ultrasonic wave propagation, i.e., the attenuation coefficient α.

[0013] S22: The formula for calculating the anisotropy index of the arc additive manufacturing specimen based on the ultrasonic attenuation coefficient is shown below:

[0014]

[0015] Among them, A α An anisotropy index based on the ultrasonic attenuation coefficient; The average ultrasonic attenuation coefficient along the scanning path perpendicular to the Z-axis; The average ultrasonic attenuation coefficient along the scanning path perpendicular to the Y-axis;

[0016] S23: The calculation formula for the non-uniformity index of arc additive manufacturing specimens based on the ultrasonic attenuation coefficient is shown below:

[0017]

[0018] Wherein, CV(α) is the non-uniformity index based on the attenuation coefficient; CD(α) is the standard deviation of the ultrasonic attenuation coefficient along the same scanning path; MV(α) is the average value of the ultrasonic attenuation coefficient along the same scanning path; and α is the attenuation coefficient.

[0019] S3: Model the anisotropy and inhomogeneity indices of the arc additive manufacturing sample, which specifically includes the following sub-steps:

[0020] S31: Obtain the anisotropy index A based on the ultrasonic attenuation coefficient in step S22. α Anisotropic index models based on the ultrasonic attenuation coefficient are established for the process parameters in step S11, as shown below:

[0021] A α =μ1B+μ2C+μ3AB+μ4AC+μ5

[0022] Where A is the welding torch travel speed; B is the arc current; C is the shielding gas flow rate; μ1, μ2, μ3, μ4 and μ5 are the first, second, third, fourth and fifth fitting parameters of the anisotropic index model based on the attenuation coefficient, respectively.

[0023] S32: Obtain the non-uniformity index CV(α) based on the ultrasonic attenuation coefficient from step S23, and establish non-uniformity index models based on the ultrasonic attenuation coefficient with the process parameters from step S11, as shown below:

[0024]

[0025] Wherein, CV(α) Z ) represents the non-uniformity index model based on the ultrasonic attenuation coefficient along the scanning path perpendicular to the Z-axis; CV(α) Y ) represents the non-uniformity index model based on the ultrasonic attenuation coefficient along the scanning path perpendicular to the Y-axis; t1, t2, t3, t4, and t5 are the first, second, third, fourth, and fifth fitting parameters of the non-uniformity index model based on the ultrasonic attenuation coefficient along the scanning path perpendicular to the Z-axis, respectively; h1, h2, h3, h4, h5, and h6 are the first, second, third, fourth, fifth, and sixth fitting parameters of the non-uniformity index model based on the ultrasonic attenuation coefficient along the scanning path perpendicular to the Y-axis, respectively.

[0026] S4: Optimize process parameters based on anisotropic index model and non-uniformity index model;

[0027] Obtain the anisotropic index model based on the ultrasonic attenuation coefficient in step S31, and the non-uniformity index model based on the attenuation coefficient in step S32. Construct the process parameter optimization constraint function as follows:

[0028]

[0029] First, single-objective optimization is performed on the anisotropy and non-uniformity indices. Then, multi-objective optimization is performed on the anisotropy and non-uniformity indices to finally determine the optimal electric arc additive manufacturing process parameters.

[0030] Preferably, the attenuation coefficient α in step S21 is obtained by using an ultrasonic transducer to obtain a primary bottom wave BW1 with a stroke of 2S and a secondary bottom wave BW2 with a stroke of 4S, and then calculating the attenuation coefficient α using the amplitude difference between the two bottom waves.

[0031]

[0032] Where A1 is the amplitude of the primary bottom wave; A2 is the amplitude of the secondary bottom wave; and S is the sample thickness along the wave propagation direction.

[0033] Preferably, in step S22, the anisotropy index of the arc additive manufacturing sample based on the ultrasonic attenuation coefficient is obtained through... To measure the difference in attenuation coefficients of ultrasonic waves propagating in two orthogonal directions; through This is used to measure the average attenuation coefficient of ultrasound in a material.

[0034] Preferably, the modeling of the anisotropy and inhomogeneity indices of the arc additive manufacturing sample in step S3 is based on surrogate modeling using the response surface methodology, where y represents the observed response variable and x is the independent variable. The response variable is approximated using a regular polynomial as follows:

[0035]

[0036] Where y is the observed response variable; x is the independent variable; β is the model parameter vector; ε is the zero-mean Gaussian error variable; n is the number of process parameter combinations; i is the welding torch travel speed number in the process parameters; j is the arc current number in the process parameters; and k is the shielding gas flow rate number in the process parameters.

[0037] Preferably, in step S4, the process parameters are optimized using a single objective, specifically as follows:

[0038] S411: Minimize the single-objective function of the process parameter optimization constraint function in step S4 |A α |, thus obtaining the optimal process parameters for anisotropy;

[0039] S412: The two inhomogeneities CV(α) in two orthogonal directions are combined. Y ) and CV(α) ZThe parameters are combined into a single overall non-uniformity index function CV(α) for parameter optimization. The overall non-uniformity function is adjusted using a weighting coefficient p to achieve optimization with emphasis in both directions. The overall non-uniformity index function CV(α) is shown in the following equation:

[0040] CV(α)=p·|CV(α Z )|+(1-p)·|CV(α Y )|

[0041] Where CV(α) is the overall non-uniformity index function; p is the weight coefficient, p∈[0,1];

[0042] S413: The above formula can be used to obtain the solution of the non-uniformity index function CV(α) under different weight parameters p, thereby obtaining the optimal process parameter value for material non-uniformity.

[0043] Preferably, in step S4, multi-objective optimization is performed on the anisotropy and non-uniformity indices, specifically as follows:

[0044] S421: Based on the composite satisfaction function CDF, anisotropic and non-uniformity indicators are integrated into a single indicator. The composite satisfaction function CDF is shown below:

[0045]

[0046] Wherein, CDF is the composite satisfaction function; w i d represents the weight of the i-th process parameter; i (y) represents the individual satisfaction defined by the i-th process parameter; y is the observed response variable; n is the number of process parameter combinations;

[0047] S422: Set the individual satisfaction d defined by the i-th target parameter. i (y), as shown below:

[0048]

[0049] Where U is the upper limit of the parameter; T is the target value of the individual parameter;

[0050] Individual satisfaction d i The value of (y) or the composite satisfaction function CDF varies between 0 and 1, where 1 represents the ideal situation and 0 represents one or more parameters that are not within the acceptable range.

[0051] S423: Obtain the single objective function CDF after integrating the anisotropic index and the overall inhomogeneity index. ACV As shown below:

[0052]

[0053] Among them, CDF ACV The integrated single objective function; q1 is the weighting coefficient of the anisotropy index; q2 is the weighting coefficient of the overall non-uniformity index; Individual satisfaction as an anisotropic indicator; Individual satisfaction as an indicator of overall unevenness;

[0054] CDF of the integrated single objective function ACV The optimal process parameters for co-optimized anisotropy and non-uniformity indices under different weight coefficients q1, q2 and p are obtained by solving the problem.

[0055] Preferably, in step S4, the process parameter optimization constraint function is solved using the fmincon function in the GNU / Octave software package to minimize the single objective function |A. α | and CV(α).

[0056] Compared with the prior art, the present invention has the following beneficial effects:

[0057] (1) This invention uses phased array ultrasonic testing to obtain ultrasonic data of samples manufactured with different process parameters, obtains ultrasonic attenuation coefficient from ultrasonic data, proposes parameters based on ultrasonic attenuation coefficient to characterize the anisotropy and non-uniformity of materials, establishes a correlation model between process parameters and anisotropy and non-uniformity index based on ultrasonic evaluation results, and determines the optimal process parameters based on different optimization objectives such as directional or isotropic optimization.

[0058] (2) In this invention, the anisotropy and inhomogeneity of the material are quantified by ultrasonic nondestructive testing of the entire sample, rather than by sparse local destructive sampling. Therefore, the anisotropy and inhomogeneity of the material can be quantified and optimized without damaging the sample.

[0059] (3) This invention establishes a surrogate model based on ultrasonic parameters, which allows for the reverse identification of optimal process parameters within the effective range of process parameters. The anisotropy and inhomogeneity of the material along the packing direction and its orthogonal direction can be included in the model individually or simultaneously without special processing, which facilitates the flexible preparation of materials with optimal quality that are oriented or isotropic under different applications, and can be well applied in actual production. Attached Figure Description

[0060] Figure 1 This is a flowchart of the method for optimizing process parameters of arc additive manufacturing based on material inhomogeneity according to the present invention;

[0061] Figure 2 This is a schematic diagram of the layer-by-layer construction of the material block according to the present invention;

[0062] Figure 3 This is a schematic diagram of the sample of the present invention;

[0063] Figure 4 This is a schematic diagram of the propagation of ultrasonic waves in a sample according to the present invention;

[0064] Figure 5 This is a schematic diagram of multi-directional ultrasonic testing of a sample according to the present invention;

[0065] Figure 6 The anisotropy index A predicted by this invention α Schematic diagram;

[0066] Figure 7 A of the present invention α A comparison graph of the surrogate model predictions and the actual experimental measurements;

[0067] Figure 8 The non-uniformity index CV(α) of this invention Z ) Schematic diagram;

[0068] Figure 9 For the CV(α) of the present invention Z A comparison chart of the surrogate model predictions and actual experimental measurements;

[0069] Figure 10 The non-uniformity index CV(α) of this invention Y ) Schematic diagram;

[0070] Figure 11 For the CV(α) of the present invention Y A comparison chart of the surrogate model predictions and actual experimental measurements;

[0071] Figure 12 This is a schematic diagram illustrating the optimal CV(α) parameter values ​​for different weight parameters p according to the present invention;

[0072] Figure 13 For different weight parameters p in this invention, the optimal CV(α) corresponds to |CV(α)|. Z A diagram illustrating the parameter values;

[0073] Figure 14 For different weight parameters p in this invention, the optimal CV(α) corresponds to |CV(α)|. Y A diagram illustrating the parameter values;

[0074] Figure 15 This is a schematic diagram showing the welding torch travel speed parameter values ​​corresponding to the optimal CV(α) for different weight parameters p in this invention;

[0075] Figure 16 This is a schematic diagram showing the arc current parameter values ​​corresponding to the optimal CV(α) for different weight parameters p in this invention;

[0076] Figure 17 This is a schematic diagram showing the protective gas flow rate parameter values ​​corresponding to the optimal CV(α) for different weight parameters p in this invention. Detailed Implementation

[0077] Exemplary embodiments, features, and aspects of the present invention will now be described in detail with reference to the accompanying drawings. The same reference numerals in the drawings denote elements that have the same or similar functions. Although various aspects of the embodiments are shown in the drawings, they are not necessarily drawn to scale unless specifically indicated otherwise.

[0078] This invention provides a method for optimizing process parameters in arc additive manufacturing based on material inhomogeneity, such as... Figure 1 As shown, the process parameters for arc additive manufacturing materials are determined, and arc additive manufacturing samples are prepared. The anisotropy and inhomogeneity indices of the arc additive manufacturing samples are quantified using ultrasonic testing. The anisotropy and inhomogeneity indices of the arc additive manufacturing samples are modeled, and the process parameters are optimized based on the anisotropy and inhomogeneity indices models. This includes:

[0079] Step S1: Determine the process parameters of the arc additive manufacturing material and prepare arc additive manufacturing samples:

[0080] Step S11: Determine the process parameters that affect the anisotropy and inhomogeneity of material properties; the process parameters are welding torch travel speed A, arc current B, and shielding gas flow rate C.

[0081] In practical engineering applications, the three process parameters—torch travel speed A, arc current B, and shielding gas flow rate C (99.99% pure argon)—are considered to be the main variables affecting material properties in the WAAM process. To investigate the influence of these three parameters on the anisotropy and inhomogeneity of the resulting material microstructure, an orthogonal array design matrix L9 for the three process parameters was established, as shown in Table 1.

[0082] Table 1 Orthogonal array design matrix for three process parameters

[0083] A welding torch travel speed mm / s 8 10 12 B Arc current A 100 120 140 C Protect gas flow L / min 15 20 25

[0084] Nine WAAM blocks were manufactured based on the process parameter values ​​corresponding to the nine design points. The actual recorded values ​​of the process parameters associated with the manufactured samples are shown in Table 2. Other minor process parameters remained unchanged. For example, the wire elongation was 12 mm, and the waiting time between subsequent layer depositions was 30 s. Therefore, the differences in anisotropy and inhomogeneity among the nine resulting samples can be entirely attributed to three process parameters.

[0085] like Figure 3The diagram shown is a schematic of an embodiment of the present invention, in which the manufactured material block is further processed to form a clean surface.

[0086] Table 2. Process parameters corresponding to the nine samples.

[0087] 1 8 98 15 2 8 119 20 3 8 139 25 4 10 101 25 5 10 119 15 6 10 139 20 7 12 101 20 8 12 119 25 9 12 146 15

[0088] Step S12: Prepare metal parts using an electric arc additive manufacturing process based on cold metal transition. By setting n different combinations of process parameters, prepare n electric arc additive samples, where n is the number of process parameter combinations in step S11.

[0089] In this embodiment, the pulse-based cold metal transfer (CMT+P) process in WAAM is selected to manufacture aluminum alloy parts. ER2319 aluminum alloy wire with a diameter of 1.2 mm is used, and the chemical composition of the welding wire is shown in Table 3. A six-axis robotic arm drives a welding torch that simultaneously feeds the wire and provides shielding gas to build material blocks layer by layer, with an oscillation strategy between adjacent layers. WAAM blocks are manufactured using nine design points under three process parameter combinations shown in Table 2, and a smooth surface is post-machined. The deposition direction is defined as the Z-axis, and the X and Y axes are two coordinates within the deposited layer.

[0090] Table 3 shows the chemical composition of the welding wire.

[0091]

[0092] Step S2: Quantify the anisotropy and inhomogeneity of the arc additive manufacturing sample using ultrasonic testing, which includes the following sub-steps:

[0093] Step S21: Obtain the ultrasonic data of the arc additive manufacturing sample from step S12 using the ultrasonic pulse-echo method, and calculate the ultrasonic attenuation coefficient α; Figure 4 The diagram shows the propagation of ultrasonic waves in a sample according to the present invention. A primary backwave BW1 with a travel time of 2 seconds and a secondary backwave BW2 with a travel time of 4 seconds are obtained using an ultrasonic transducer. The ultrasonic attenuation coefficient α is calculated based on the amplitude difference between the two backwaves.

[0094]

[0095] Where α is the attenuation coefficient; A1 is the amplitude of the primary bottom wave; A2 is the amplitude of the secondary bottom wave; and S is the sample thickness along the wave propagation direction.

[0096] Step S22: The calculation formula for the anisotropy index of the arc additive manufacturing specimen based on the ultrasonic attenuation coefficient is shown below:

[0097]

[0098] Among them, A αAn anisotropy index based on the ultrasonic attenuation coefficient; The average ultrasonic attenuation coefficient along the scanning path perpendicular to the Z-axis; The average ultrasonic attenuation coefficient along the scanning path perpendicular to the Y-axis.

[0099] like Figure 2 The diagram illustrates the layer-by-layer material block construction of this invention. A six-axis robotic arm drives a welding torch that simultaneously feeds wire and provides shielding gas, constructing the material block layer by layer. Anisotropy characterizes the differences in orientation-related properties within the material. Since WAAM is a path-dependent layer-by-layer deposition process, its microstructure and material properties exhibit orientation-related characteristics. Two orthogonal directions parallel / perpendicular to the construction direction (…) Figure 2 The Z / Y axes shown are considered to be typical directions characterizing the anisotropy of materials.

[0100] Step S23: The formula for calculating the non-uniformity index of the arc additive manufacturing sample based on the ultrasonic attenuation coefficient is shown below:

[0101]

[0102] Wherein, CV(α) is the non-uniformity index based on the attenuation coefficient; SD(α) is the standard deviation of the ultrasonic attenuation coefficient along the same scanning path; and MV(α) is the average value of the ultrasonic attenuation coefficient along the same scanning path.

[0103] Material inhomogeneity is determined by modeling the spatial variation of material properties. Here, the material property variations of the WAAM sample along the Z / Y axes are used to quantify the degree of inhomogeneity of the sample.

[0104] like Figure 5 The diagram illustrates the multi-directional ultrasonic testing of a sample according to the present invention. The ultrasonic system sequentially acquires data along a scanning path parallel to the Z and Y axes. The attenuation coefficient is calculated using data acquired at each scanning position along the path. Subsequently, the average attenuation coefficient MV(α) and standard deviation SD(α) of all scanning positions can be calculated. The standard deviation SD(α) reflects the degree of deviation of all data from the average MV(α).

[0105] Evaluation of the anisotropy and inhomogeneity of the nine prepared aluminum alloy samples:

[0106] Evaluate the ultrasonic attenuation coefficient α for each of the nine aluminum alloy specimens in two orthogonal directions. Calculate the mean and standard deviation of α for each scan path, and calculate the corresponding coefficient of variation.

[0107] The results based on the attenuation coefficient are shown in Table 4. CV(α) Z ) and CV(α) YThe terms are the variation coefficients of the attenuation coefficient related to the scan path perpendicular to the Z and Y axes, respectively. The parameter values ​​based on the attenuation coefficient vary significantly among different samples. Since ultrasonic attenuation is mainly affected by the microstructure characteristics of the material, such as grain boundaries, inclusions, phase interfaces, and porosity, the large differences in attenuation coefficients between samples indicate that the attenuation coefficient is sensitive to microstructure characteristics. A larger A... α This means that the microstructural features of the two orthogonal directions are quite different.

[0108] Table 4 Results based on attenuation coefficient

[0109]

[0110] Step S3: Model the anisotropy and inhomogeneity indices of the arc additive manufacturing sample, which specifically includes the following sub-steps:

[0111] Step S31: Obtain the anisotropy index A based on the ultrasonic attenuation coefficient from step S22. α Anisotropic index models based on the ultrasonic attenuation coefficient are established for the process parameters in step S1, as shown below:

[0112] A α =-2.171B+14.608C+0.254AB-1.653AC+29.681

[0113] Where A is the welding torch travel speed; B is the arc current; and C is the shielding gas flow rate.

[0114] like Figure 6 The figure shows the anisotropy index A predicted by the present invention. α The diagram illustrates the use of the above formula to predict the anisotropy index A. α The magnitude of the value is represented by the intensity of the color. Figure 7 A of the present invention α The graph comparing the surrogate model predictions with the actual experimental measurements shows the relationship between the surrogate model predictions and A. α A comparison of actual experimental measurements revealed a small deviation between the two near the diagonal, demonstrating the effectiveness of the model for predictive purposes.

[0115] Step S32: Obtain the non-uniformity index CV(α) based on the ultrasonic attenuation coefficient from step S23, and establish a non-uniformity index model based on the ultrasonic attenuation coefficient with the process parameters from step S11, as shown below:

[0116]

[0117] Wherein, CV(α) Z) represents the non-uniformity index model based on the ultrasonic attenuation coefficient along the scanning path perpendicular to the Z-axis; VC(α) Y () is a non-uniformity index model based on the ultrasonic attenuation coefficient on the scanning path perpendicular to the Y-axis.

[0118] like Figure 8 The figure shows the non-uniformity index CV(α) of the present invention. Z The diagram illustrates how the above formula is used to predict the heterogeneity index CV(α). Z The value is represented by the intensity of the color. Figure 9 For the CV(α) of the present invention Z The graph compares the predicted values ​​of the surrogate model with the actual experimental measurements, showing the CV(α) values. Z By comparing the predicted values ​​of the surrogate model with the actual experimental measurements, it was found that CV(α) Z The F-statistic of the surrogate model is 9.229, and the adequacy of the model is statistically guaranteed (F > F). 0.05 =9.117). R of the proxy model 2 The value is 0.925.

[0119] like Figure 10 The figure shows the non-uniformity index CV(α) of the present invention. Y A schematic diagram using CV(α) Y The surrogate model predicts the non-uniformity index, using color depth to represent the magnitude of the value; Figure 11 For the CV(α) of the present invention Y The graph compares the predicted values ​​of the surrogate model with the actual experimental measurements, showing the CV(α) values. Y The surrogate model predictions were compared with the actual experimental measurements, and the CV(α) was calculated. Y The F-statistic of the surrogate model is 41.211 (F>F). 0.05 =28.710), R 2 The value is 0.990, indicating that the model is sufficient to correlate the data. Furthermore, the CV(α) of sample 6 in Table 4... Y The measured values ​​were used to verify the accuracy of the model. The CV(α) of sample No. 6 was... Y The predicted value is 22.331%, if Figure 11 As indicated by the diamond marker, the prediction error is only 2.610%, demonstrating that the constructed model can produce accurate prediction results.

[0120] Step S4: Optimize the process parameters based on the anisotropy index model and the non-uniformity index model;

[0121] Obtain the anisotropy index model based on the ultrasonic attenuation coefficient in step S31 and the non-uniformity index model based on the ultrasonic attenuation coefficient in step S32, and construct the process parameter optimization constraint function as shown in the following equation:

[0122]

[0123] Step S41: First, perform single-objective optimization on the anisotropy and inhomogeneity indices, specifically as follows:

[0124] Step S411: Solve for the single objective function |A, which minimizes the process parameter optimization constraint function in step S4. α |、|CV(α Z )| and |CV(α) Y The process parameters that minimize anisotropy and unidirectional inhomogeneity are obtained by solving the process parameter optimization constraint function using the fmincon function. The obtained single-objective optimal process parameters are shown in Table 5.

[0125] Table 5 Optimal Process Parameters for Single Objective

[0126]

[0127] Step S412: CV(α) is the inhomogeneity index in two orthogonal directions. Y ) and CV(α) Z The parameters are combined into a single overall non-uniformity index function CV(α) for parameter optimization. The overall non-uniformity function is adjusted using a weighting coefficient p to achieve optimization with emphasis in both directions. The overall non-uniformity index function CV(α) is shown in the following equation:

[0128] CV(α)=p·|CV(α Z )|+(1-p)·|CV(α Y )|

[0129] Where CV(α) is the overall non-uniformity index function; p is the weight coefficient, p∈[0,1].

[0130] Step S413: The above formula can be used to obtain the solution of the non-uniformity index function CV(α) under different weight parameters p, thereby obtaining the process parameter value with the minimum material non-uniformity.

[0131] like Figure 12 The diagram illustrates the optimal CV(α) parameter values ​​for different weight parameters p according to the present invention. Based on step S412, the solution of the non-uniformity index function CV(α) for different weight parameters p can be obtained, thereby obtaining the optimal three process parameter values. Under the optimal three process parameters, the following parameters are determined: Figure 13The figure shows the |CV(α) corresponding to the optimal CV(α) for different weight parameters p in this invention. Z The parameter value represents the non-uniformity index model value based on the ultrasonic attenuation coefficient along the scanning path perpendicular to the Z-axis; the optimal parameters were determined under the three optimal process parameters. Figure 14 The optimal CV(α) corresponding to different weight parameters p is shown below. Y The parameter value represents the non-uniformity index model value based on the ultrasonic attenuation coefficient along the scanning path perpendicular to the Y-axis. The three process parameter values ​​corresponding to the optimal CV(α) for different weight parameters p are as follows: Figures 15-17 As shown, Figure 15 Let be the welding torch travel speed parameter value corresponding to the optimal CV(α) for different weight parameters p, representing the optimal welding torch travel speed in the optimized process parameters; Figure 16 Let be the arc current parameter value corresponding to the optimal CV(α) for different weight parameters p, representing the optimal arc current among the optimized process parameters; Figure 17 Let be the protective gas flow rate parameter value corresponding to the optimal CV(α) for different weight parameters p, representing the optimal protective gas flow rate among the optimized process parameters. Figures 12-17 The optimization results show that, for any value p, the optimal torch travel speed A and shielding gas flow rate C are less than 9 mm / s and 18 L / min, respectively. The results indicate that a shielding gas flow rate of 18 L / min is sufficient to protect the arc from air contamination, providing quantitative validation for the use of low-velocity shielding gas. To illustrate this more clearly, [the following is a separate, unrelated section:] ...selecting... Figures 12-17 The optimal process parameters determined for the three arbitrary p values ​​shown are listed in Table 6.

[0132] Table 6. Optimal process parameters determined based on p-value.

[0133]

[0134] Step S42: Then, perform multi-objective optimization on the anisotropy and inhomogeneity indices, specifically as follows:

[0135] Step S421: Based on the composite satisfaction function CDF, integrate the anisotropy index and the non-uniformity index into a composite satisfaction function CDF, as shown below:

[0136]

[0137] Wherein, CDF is the composite satisfaction function; w i d represents the weight of the i-th process parameter; i (y) represents the individual satisfaction defined by the i-th process parameter; y is the observed response variable; n is the number of process parameter combinations.

[0138] Step S422: Set the individual satisfaction d defined by the i-th target parameter. i (y), as shown below:

[0139]

[0140] Where U is the upper limit of the parameter; T is the target value of the individual parameter.

[0141] Individual satisfaction d i The value of (y) or the composite satisfaction function CDF varies between 0 and 1, where 1 represents the ideal situation and 0 indicates that one or more parameters are outside the acceptable range. For illustrative purposes, Table 7 shows the target variable, example target, upper limit, and weights. It is assumed here that the overall quality of the material is unsatisfactory once the value of the anisotropy or inhomogeneity index function exceeds 50%.

[0142] Table 7 Material Quality Judgment Criteria

[0143] <![CDATA[|A α |]]> 0 50％ <![CDATA[q1=1,2,3,4,5]]> |CV(α)| 0 50％ <![CDATA[q2=1,2,3,4,5]]>

[0144] Step S423: Obtain the single objective function CDF after integrating the anisotropy index and the overall inhomogeneity index. ACV As shown below:

[0145]

[0146] Among them, CDF ACV The integrated single objective function; q1 is the weighting coefficient of the anisotropy index; q2 is the weighting coefficient of the overall non-uniformity index; Individual satisfaction as an anisotropic indicator; Individual satisfaction is an indicator of overall unevenness.

[0147] CDF of the integrated single objective function ACV The optimal process parameters after co-optimization of anisotropy and inhomogeneity indices are obtained under different weight coefficients q1, q2, and p. In practical engineering applications, once the weights of the quality indices are determined, the optimal solution can be easily determined. Here, several different weights are randomly selected to demonstrate the CDF. ACV The optimization results are shown in Table 8.

[0148] Table 8 CDF ACV Optimization results table

[0149]

[0150]

[0151] The beneficial effects of this invention are as follows: This invention provides a method for optimizing process parameters in arc additive manufacturing (WAAM) based on material inhomogeneity. It employs phased array ultrasonic testing to acquire ultrasonic data of WAAM samples manufactured using different process parameters, obtains the ultrasonic attenuation coefficient from the ultrasonic data, and proposes parameters based on the ultrasonic attenuation coefficient to characterize the anisotropy and inhomogeneity of the material. Based on the ultrasonic evaluation results, a correlation model is established between process parameters and anisotropy and inhomogeneity indices. The optimal parameters are determined according to different optimization objectives, such as directional optimization or isotropic optimization. Verification and analysis of practical cases demonstrate that this method has good application effects and meets practical application requirements.

[0152] The embodiments described above are merely preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Various modifications and improvements made by those skilled in the art to the technical solutions of the present invention without departing from the spirit of the present invention should fall within the protection scope defined by the claims of the present invention.

Claims

1. A method for optimizing process parameters in arc additive manufacturing based on material inhomogeneity, characterized in that, It includes: S1: Determine the process parameters for the arc additive manufacturing material and prepare arc additive test samples: S11: Determine the process parameters that affect the anisotropy and inhomogeneity of material properties; The process parameters are welding torch travel speed A, arc current B, and shielding gas flow rate C; S12: Prepare metal parts using an electric arc additive manufacturing process based on cold metal transition. By setting n different combinations of process parameters, prepare n electric arc additive samples, where n is the number of process parameter combinations in step S11. S2: The anisotropy and inhomogeneity of the arc additive manufacturing sample are quantified using ultrasonic testing methods, which includes the following sub-steps: S21: Use the ultrasonic pulse echo method to obtain the ultrasonic data of the arc additive sample in step S12, and calculate the total energy loss of ultrasonic wave propagation, i.e., the attenuation coefficient α. S22: The formula for calculating the anisotropy index of the arc additive manufacturing specimen based on the ultrasonic attenuation coefficient is shown below: Among them, A α An anisotropy index based on the ultrasonic attenuation coefficient; The average ultrasonic attenuation coefficient along the scanning path perpendicular to the Z-axis; The average ultrasonic attenuation coefficient along the scanning path perpendicular to the Y-axis; S23: The calculation formula for the non-uniformity index of arc additive manufacturing specimens based on the ultrasonic attenuation coefficient is shown below: Wherein, CV(α) is the non-uniformity index based on the attenuation coefficient; SD(α) is the standard deviation of the ultrasonic attenuation coefficient along the same scanning path; MV(α) is the average value of the ultrasonic attenuation coefficient along the same scanning path; and α is the attenuation coefficient. S3: Model the anisotropy and inhomogeneity indices of the arc additive manufacturing sample, which specifically includes the following sub-steps: S31: Obtain the anisotropy index A based on the ultrasonic attenuation coefficient in step S22. α Anisotropic index models based on the ultrasonic attenuation coefficient are established for the process parameters in step S11, as shown below: A α =μ1B+μ2C+μ3AB+μ4AC+μ5 Where A is the welding torch travel speed; B is the arc current; C is the shielding gas flow rate; μ1, μ2, μ3, μ4 and μ5 are the first, second, third, fourth and fifth fitting parameters of the anisotropic index model based on the attenuation coefficient, respectively. S32: Obtain the non-uniformity index CV(α) based on the ultrasonic attenuation coefficient from step S23, and establish non-uniformity index models based on the ultrasonic attenuation coefficient with the process parameters from step S11, as shown below: Wherein, CV(α) Z ) represents the non-uniformity index model based on the ultrasonic attenuation coefficient along the scanning path perpendicular to the Z-axis; CV(α) Y ) represents the non-uniformity index model based on the ultrasonic attenuation coefficient along the scanning path perpendicular to the Y-axis; t1, t2, t3, t4, and t5 are the first, second, third, fourth, and fifth fitting parameters of the non-uniformity index model based on the ultrasonic attenuation coefficient along the scanning path perpendicular to the Z-axis, respectively; h1, h2, h3, h4, h5, and h6 are the first, second, third, fourth, fifth, and sixth fitting parameters of the non-uniformity index model based on the ultrasonic attenuation coefficient along the scanning path perpendicular to the Y-axis, respectively. S4: Optimize process parameters based on anisotropic index model and non-uniformity index model; Obtain the anisotropic index model based on the ultrasonic attenuation coefficient in step S31, and the non-uniformity index model based on the attenuation coefficient in step S32. Construct the process parameter optimization constraint function as follows: First, single-objective optimization is performed on the anisotropy and non-uniformity indices. Then, multi-objective optimization is performed on the anisotropy and non-uniformity indices to finally determine the optimal electric arc additive manufacturing process parameters.

2. The method for optimizing arc additive manufacturing process parameters based on material inhomogeneity according to claim 1, characterized in that: The attenuation coefficient α in step S21 is obtained by using an ultrasonic transducer to obtain a primary bottom wave BW1 with a stroke of 2S and a secondary bottom wave BW2 with a stroke of 4S, and then calculating it using the amplitude difference between the two bottom waves. Where A1 is the amplitude of the primary bottom wave; A2 is the amplitude of the secondary bottom wave; and S is the sample thickness along the wave propagation direction.

3. The method for optimizing arc additive manufacturing process parameters based on material inhomogeneity according to claim 1, characterized in that: In step S22, the anisotropy index of the arc additive manufacturing sample based on the ultrasonic attenuation coefficient is obtained through... To measure the difference in attenuation coefficients of ultrasonic waves propagating in two orthogonal directions; through This is used to measure the average attenuation coefficient of ultrasound in a material.

4. The method for optimizing arc additive manufacturing process parameters based on material inhomogeneity according to claim 1, characterized in that: In step S3, modeling the anisotropy and inhomogeneity indices of the arc additive manufacturing sample is based on surrogate modeling using the response surface methodology. We use y to represent the observed response variable and x as the independent variable. The response variable is approximated using a regular polynomial: Where y is the observed response variable; x is the independent variable; β is the model parameter vector; ε is the zero-mean Gaussian error variable; n is the number of process parameter combinations; i is the welding torch travel speed number in the process parameters; j is the arc current number in the process parameters; and k is the shielding gas flow rate number in the process parameters.

5. The method for optimizing arc additive manufacturing process parameters based on material inhomogeneity according to claim 1, characterized in that: Step S4 involves single-objective optimization of the process parameters, specifically: S411: Minimize the single-objective function of the process parameter optimization constraint function in step S4 |A α |, thus obtaining the optimal process parameters for anisotropy; S412: The two inhomogeneities CV(α) in two orthogonal directions are combined. Y ) and CV(α) Z The parameters are combined into a single overall non-uniformity index function CV(α) for parameter optimization. The overall non-uniformity function is adjusted using a weighting coefficient p to achieve optimization with emphasis in both directions. The overall non-uniformity index function CV(α) is shown in the following equation: CV(α)=p·|CV(α Z )|+(1-p)·|CV(α Y )| Where CV(α) is the overall non-uniformity index function; p is the weight coefficient, p∈[0,1]; S413: The above formula can be used to obtain the solution of the non-uniformity index function CV(α) under different weight parameters p, thereby obtaining the optimal process parameter value for material non-uniformity.

6. The method for optimizing arc additive manufacturing process parameters based on material inhomogeneity according to claim 5, characterized in that: Step S4 involves multi-objective optimization of the anisotropy and inhomogeneity indices, specifically as follows: S421: Based on the composite satisfaction function CDF, anisotropic and non-uniformity indicators are integrated into a single indicator. The composite satisfaction function CDF is shown below: Where CDF is the composite satisfaction function; w i The weight of the i-th process parameter; d i (y) represents the individual satisfaction defined by the i-th process parameter; y is the observed response variable; n is the number of process parameter combinations; S422: Set the individual satisfaction d defined by the i-th target parameter. i (y), as shown below: Where U is the upper limit of the parameter; T is the target value of the individual parameter; Individual satisfaction d i The value of (y) or the composite satisfaction function CDF varies between 0 and 1, where 1 represents the ideal situation and 0 represents one or more parameters that are not within the acceptable range. S423: Obtain the single objective function CDF after integrating the anisotropic index and the overall inhomogeneity index. AcV As shown below: Among them, CDF ACV The integrated single objective function; q1 is the weighting coefficient of the anisotropy index; q2 is the weighting coefficient of the overall non-uniformity index; Individual satisfaction as an anisotropy index; d |CV(α)| Individual satisfaction as an indicator of overall unevenness; CDF of the integrated single objective function ACV The optimal process parameters for co-optimized anisotropy and non-uniformity indices under different weight coefficients q1, q2 and p are obtained by solving the problem.

7. The method for optimizing arc additive manufacturing process parameters based on material inhomogeneity according to claim 1, characterized in that: In step S4, the process parameter optimization constraint function is solved using the fmincon function from the GNU / Octave package to minimize the single objective function |A. α | and CV(α).

Images