Aluminum alloy thin-walled pipe pulse TIG welding quality control method based on monitoring of droplet transfer

Abstract

The application belongs to the technical field of welding, and discloses an aluminum alloy thin-walled pipe pulse TIG welding quality control method based on droplet transfer monitoring. Key welding process parameters are screened through experiments; the screened parameters are used as optimization variables, expert visual inspection scores, tensile strength, fusion width, fusion depth and pore occurrence rate are used as response targets, a response surface approximate model is established through experiments, an entropy weight-TOPSIS method is combined to determine comprehensive weights, and static optimal welding process parameters are solved; in the welding experiment, droplet transfer characteristic data are extracted through a droplet monitoring device, and the data are used as optimal droplet transfer characteristic parameters; in actual welding, the static optimal process parameters are used as a benchmark, the optimal droplet characteristics are used as a reference, droplet behavior is monitored in real time, a fuzzy-PID control method is used to dynamically adjust welding current, wire feeding speed or welding gun height, so that the droplet transfer is stably kept within the optimal range, and whole-process closed-loop control of welding quality is realized.

Description

Technical Field

[0001] This application belongs to the field of welding technology, specifically relating to a quality control method for pulse TIG welding of thin-walled aluminum alloy tubes based on droplet transfer monitoring. Background Technology

[0002] Aluminum alloys are widely used in critical components such as fuel lines and hydraulic systems of aero engines due to their excellent specific strength and corrosion resistance. Among them, 5B02 aluminum alloy, as a typical aluminum-magnesium alloy, has good formability and weldability and is often used to manufacture aircraft duct components. However, aerospace ducts are mostly thin-walled tubular structures with complex spatial arrangements, which places extremely high demands on the quality and consistency of welded joints.

[0003] Currently, manual tungsten inert gas (TIG) welding is still widely used in aerospace duct welding. While this method can adapt to welding requirements in complex spatial locations, it is significantly affected by human factors in actual operation, easily leading to problems such as unstable welding heat input, fluctuating molten pool behavior, and poor wire feeding uniformity. These factors collectively result in low weld uniformity, especially in the arc initiation and arc termination areas of circumferential welds, where defects such as undercut, weld beads, and porosity are prone to occur, seriously affecting the service safety and lifespan of duct components.

[0004] To improve welding quality and efficiency, automated TIG welding systems have been gradually introduced into the aerospace manufacturing field. While automated welding overcomes the instability of manual operation to some extent, it still faces many challenges in welding thin-walled aluminum alloy tubes: First, the welding process parameter system is complex, encompassing multiple variables such as current parameters, wire feeding parameters, pulse parameters, and shielding gas parameters, with significant interactions between these parameters; second, process windows for specific materials and joint types have not yet been established, and existing research focuses mainly on flat plate welding, lacking systematic research on thin-walled circumferential welds, especially in key areas such as arc initiation and arc termination; third, droplet transfer behavior during welding has a significant impact on weld formation, but current research on the monitoring and control of droplet transfer in TIG welding is insufficient, making it difficult to establish a quantitative correlation between droplet morphology, transfer frequency, and weld quality.

[0005] In existing technologies, welding process parameter optimization often relies on trial-and-error methods or single-objective optimization strategies, which are not only inefficient but also difficult to achieve synergistic optimization among multiple quality indicators such as weld morphology, mechanical properties, and defect rate. Furthermore, disturbances during the welding process can lead to unstable droplet transfer, further exacerbating fluctuations in weld quality.

[0006] Therefore, achieving precise monitoring and control of droplet transfer behavior under the coupling effect of multiple process parameters, and establishing its mapping relationship with weld formation and joint performance, thereby achieving stable control of welding quality, has become a problem that needs to be solved in the automatic TIG welding of aluminum alloy thin-walled tubes. Summary of the Invention

[0007] To address the problems of uncontrollable droplet transfer behavior and poor weld formation consistency and unstable welding quality caused by the coupling effect of multiple process parameters, this application aims to provide a quality control method for pulsed TIG welding of thin-walled aluminum alloy tubes based on droplet transfer monitoring. By establishing a mapping relationship between droplet transfer morphology and weld quality, and constructing a parameter decision-making and real-time control system based on multi-objective optimization, the method realizes the transformation of welding quality from experience-based judgment to precise control.

[0008] To achieve the above technical objectives, this application specifically adopts the following technical solution: In one aspect of this application, a method for quality control of pulsed TIG welding of thin-walled aluminum alloy tubes based on droplet transfer monitoring is provided, comprising the following steps: S1. The Plackett-Burman experimental design method was used to screen welding process parameters and obtain key welding process parameters related to droplet transfer mode and welding quality. S2. Using the selected key welding process parameters as optimization variables, and expert visual inspection scores, tensile strength, weld width, weld depth, and porosity as response targets, an approximate response surface model between the welding process parameters and the welding quality targets is established through Box-Behnken experiments; and the entropy weight-TOPSIS method is used to determine the comprehensive weight of each response target. The static optimal welding process parameters are obtained by solving the approximate model. S3. During the welding test, a droplet monitoring device is used to collect droplet transition video in real time, and the video is processed to extract droplet transition characteristic data such as droplet size, shape, transition frequency and transition speed. The droplet transition characteristic data corresponding to the static optimal welding process parameters are used as the optimal droplet transition characteristic parameters. S4. In the actual welding process, the static optimal welding process parameters are used as the benchmark setting values, and the optimal droplet transition characteristic parameters are used as the reference standard. The actual droplet transition behavior is monitored in real time using a droplet monitoring device, and the actual droplet characteristic parameters are compared with the reference standard. When a deviation occurs, the welding current, wire feed speed or welding torch height is dynamically adjusted using a fuzzy-PID control method to stabilize the actual droplet transition behavior within the range of the optimal droplet transition characteristic parameters, thereby achieving closed-loop control of the entire welding quality process.

[0009] In one implementation, step S1 involves screening welding process parameters, which include current parameters, welding wire parameters, shielding gas parameters, pulse parameters, arc initiation parameters, and arc termination parameters.

[0010] In one implementation, the Plackett-Burman experimental design method screens out the significance levels of N-1 factors through N trials, where N is a multiple of 4, and determines key welding process parameters based on analysis of variance.

[0011] In one implementation, the entropy weight-TOPSIS method includes: Calculate the entropy weight of each response objective, and determine the proportion of each response objective based on the entropy weight; The TOPSIS method is used to calculate the closeness of each welding process parameter scheme to the ideal solution, thereby determining the optimal welding process parameters.

[0012] In one implementation, the response surface approximation model is established using Design-Expert or Isight software, and the type of optimization variables is determined to be extremely large, extremely small, or intermediate; wherein, the expert visual inspection score and tensile strength are extremely large, the melt width and melt depth are intermediate, and the porosity is extremely small.

[0013] In one implementation, the fuzzy-PID control method is applied to the arc initiation stage, the stable welding stage, and the arc termination stage of the welding process; During the arc initiation stage, if unstable arc initiation is detected, adjust the welding current or welding torch height. During the stable welding phase, when changes in the droplet transfer pattern or welding trajectory are detected, the welding parameters are adjusted. During the arc-ending phase, reduce the welding current and welding speed to prevent arc craters or incomplete fusion defects.

[0014] In one embodiment, the molten droplet monitoring device includes an information acquisition system, an information processing system, and an information export system; The information acquisition system includes an industrial CMOS camera, which is configured to suppress arc light interference and capture video of droplet transfer during the welding process. The information processing system includes image processing software configured to filter and denoise the droplet transition video, and decompose the video into multiple images to extract droplet transition feature data; The information export system is configured to output droplet transition characteristic data as a text data file and establish the correspondence between droplet change patterns and welding process parameters.

[0015] In one embodiment, the aluminum alloy thin-walled tube is a 5B02 aluminum alloy thin-walled tube with a wall thickness of 1 mm and a diameter of 63 mm.

[0016] The beneficial effects of this application are as follows: This application achieves multi-dimensional improvement in welding quality by constructing a control system that combines droplet transition monitoring with multi-objective parameter optimization. High-speed vision-based droplet behavior analysis establishes a quantitative relationship between process parameters and weld formation, breaking through the traditional experience-based parameter setting mode. Plackett-Burman experimental design is used to rapidly screen key process parameters, and a multi-objective optimization system is constructed by combining response surface methodology and entropy weight-TOPSIS decision-making, significantly improving parameter optimization efficiency. A fuzzy-PID controller is introduced to achieve real-time control of the droplet transition process, effectively suppressing quality fluctuations during welding. This method places quality control upstream in the welding process, forming a closed-loop control of "monitoring-decision-execution," significantly improving the formation consistency of the arc initiation and termination areas of the circumferential weld, and providing reliable technical support for automated welding of thin-walled aluminum alloy tubes. Attached Figure Description

[0017] Figure 1 This is a flowchart illustrating the quality control method for pulse TIG welding of thin-walled aluminum alloy tubes according to an embodiment of this application. Detailed Implementation

[0018] The technical solution of this application will be clearly and completely described below with reference to specific embodiments. However, those skilled in the art will understand that the embodiments described below are only some embodiments of this application, not all embodiments, and are only used to illustrate this application, and should not be regarded as limiting the scope of this application. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0019] This application constructs a closed-loop control method between droplet transition characteristics and welding quality. Droplet images during the welding process are captured in real time using a CMOS high-speed camera, and key characteristic parameters such as droplet size, morphology, and transition rate are obtained after processing. These parameters are compared with the optimal process benchmark obtained through a multi-objective hybrid optimization strategy (integrating Plackett-Burman experiments, response surface methodology, and entropy-weighted TOPSIS). The system uses a fuzzy-PID controller to dynamically adjust parameters such as pulse current and wire feed speed to ensure that the droplet is always in an optimal transition state. This application's method particularly focuses on the stability during the arc initiation and termination stages, aiming to achieve high-precision and high-efficiency control of the circumferential weld formation quality of thin-walled aluminum alloy tubes, effectively reducing welding defects.

[0020] In one specific implementation method, refer to Figure 1As shown, a method for quality control of pulsed TIG welding of thin-walled aluminum alloy tubes based on droplet transfer monitoring is provided, including the following steps: S1. The Plackett-Burman experimental design method was used to screen welding process parameters and obtain key welding process parameters related to droplet transfer mode and welding quality.

[0021] In some embodiments, the welding process parameters cover six major categories: current parameters, welding wire parameters, shielding gas parameters, pulse parameters, arc initiation parameters, and arc termination parameters.

[0022] The current parameter is controlled by dividing the total welding time into multiple time sectors. Each sector corresponds to a specific welding period. The welding current value of each sector is set independently according to the heat accumulation characteristics of that stage, taking into account the special thermal cycle requirements of the arc initiation and arc termination processes, thereby achieving precise segmented control of the welding heat input.

[0023] The core of welding wire parameters is wire feed speed control. This parameter directly affects the amount of molten metal deposited in the weld, and plays a decisive role in the weld reinforcement, penetration depth, and surface finish. During the arc initiation and termination stages, the proper setting of wire feed parameters is particularly important for suppressing end defects and ensuring weld continuity.

[0024] The shielding gas parameters mainly include gas flow rate and composition ratio. The shielding gas forms a barrier that isolates the weld from the atmosphere during welding, and its parameter settings directly affect arc stability and the effectiveness of molten pool protection. Improper parameters can lead to defects such as porosity and oxidation throughout the weld area.

[0025] The pulse parameters include pulse frequency, peak current, base current, and duration. The pulse effect promotes the formation of the fish-scale pattern in the weld and improves the quality of critical areas through heat input adjustment. During the arc initiation and termination phases, adjusting the pulse parameters can effectively reduce localized heat input and prevent burn-through of thin-walled pipes.

[0026] The arc initiation and arc termination parameters, as independent control units, work in conjunction with the pulse parameters and welding wire parameters. The arc initiation parameters ensure reliable arc ignition and initial molten pool formation; the arc termination parameters, through current attenuation and wire feed adjustment, achieve crater filling and crack prevention, jointly ensuring the weld end forming quality and service performance.

[0027] It is understandable that other parameters not categorized above, which have been proven through testing to be related to weld formation quality, are also included in the process considerations and together constitute a complete welding process parameter system.

[0028] In some embodiments, the Plackett-Burman experimental design method is used to construct the experimental protocol. The advantage of this method is that it can effectively identify key factors that significantly affect the output response with a relatively small number of trials. Specifically, when N trials are arranged, this method can evaluate the significance level of up to N-1 factors, where the number of trials N is a multiple of 4.

[0029] During implementation, welding tests were arranged and executed according to the PB experimental design matrix. For each test, relevant quality response indicators were recorded. Based on all experimental data, a first-order mathematical model for each response was established, with the following expression:

[0030] in, y It is a response; It is a constant; It is the first i The coefficient of each variable, which reflects the direction and magnitude of the parameter's influence on the response; It is the first i One variable.

[0031] To scientifically determine the significance of each process parameter, an analysis of variance (ANOVA) is required. The F-statistic is calculated to assess the significance of each experimental factor's influence on the response. The formula for calculating the F-statistic is:

[0032] in, N It is the number of sample points; It is the first i The true value of each sample point; It is the average value of the sample points; It is the first i Predicted values ​​for each sample point; is the number of variables; n is the total number of samples.

[0033] After calculating the F-statistic, perform a hypothesis test on it and obtain the corresponding... p Value. According to statistical standards, it is generally when p When the value is less than or equal to 0.05, the factor is considered to have a significant effect on the response; if... p A value less than or equal to 0.01 indicates that the factor's influence is highly significant. Ultimately, based on this significance criterion, key welding process parameters that decisively influence droplet transfer patterns and welding quality are identified and screened from numerous welding process parameters, providing clear targets for subsequent process optimization.

[0034] S2. Using the selected key welding process parameters as optimization variables, and expert visual inspection scores, tensile strength, weld width, weld depth, and porosity as response targets, an approximate response surface model between the welding process parameters and the welding quality target is established through Box-Behnken experiments; and the entropy weight-TOPSIS method is used to determine the comprehensive weight of each response target. The static optimal welding process parameters are obtained by solving the approximate model.

[0035] In some embodiments, the Box-Behnken experimental design method is used to arrange welding experiments. This method is suitable for optimization scenarios with three or more factors. Its advantage is that it does not include polar axis endpoints, which can effectively reduce the interference of random factors on the optimization results.

[0036] After conducting welding tests according to the experimental design matrix, quality data for each specimen were collected. In some embodiments, expert visual inspection was performed by experienced inspectors; tensile strength testing strictly followed the national standard GB / T 228.1-2010; weld width and weld depth were obtained through cross-sectional metallographic measurements; and porosity was quantitatively analyzed using X-ray detection.

[0037] In some embodiments, based on experimental data, an approximate response surface model between welding process parameters and various response targets is established using Design-Expert or Isight software. During the modeling process, the optimization type of each response target is clearly defined: expert visual inspection score and tensile strength are extremely large indicators, meaning the larger the value, the better the quality; weld width and weld depth are intermediate indicators, which need to be controlled within an ideal range; porosity is an extremely small indicator, requiring its value to be minimized.

[0038] In some embodiments, the entropy-weighted TOPSIS method is used for multi-objective comprehensive optimization. First, the objective weight of each response objective is calculated using the entropy-weighted method: based on the dispersion of each indicator data, its information entropy is calculated, thereby determining the proportion of each indicator in the comprehensive evaluation. Then, the TOPSIS method is used to evaluate the merits of each combination of process parameters: the Euclidean distance between each experimental scheme and the ideal solution (composed of the optimal values ​​of each indicator) and the negative ideal solution (composed of the worst values ​​of each indicator) is calculated, and finally, the scheme that is closest to the ideal solution and furthest from the negative ideal solution is identified as the optimal one.

[0039] S3. During the welding test, a droplet monitoring device is used to collect droplet transition video in real time, and the video is processed to extract droplet transition characteristic data such as droplet size, shape, transition frequency and transition speed. The droplet transition characteristic data corresponding to the static optimal welding process parameters are used as the optimal droplet transition characteristic parameters.

[0040] In some embodiments, the molten droplet monitoring device consists of three parts: an information acquisition system, an information processing system, and an information export system.

[0041] The information acquisition system includes an industrial CMOS camera mounted near the automated welding torch. Its optical parameters are adjusted to ensure the focus remains on the welding wire, tungsten electrode, and the upper surface of the molten pool. The camera remains relatively stationary with the welding torch and features arc suppression capabilities, ensuring clear video sequences of the droplet transfer process even in strong arc light environments.

[0042] The information processing system integrates image processing software to receive video data from the information acquisition system. First, the video signal is filtered and denoised to eliminate arc interference and noise. Then, the video is decomposed into multiple still images in time sequence, with a sampling frequency of no less than 500 Hz. Using image recognition algorithms, feature parameters such as droplet size, geometry, transition frequency, and transition speed are extracted from each image.

[0043] The information export system converts the processed droplet characteristic data into a text-format data file and establishes a correspondence between the droplet variation pattern and the current welding process parameters.

[0044] In some embodiments, after obtaining circumferential weld joints under multiple sets of welding process parameters, the forming quality (e.g., weld width, weld depth) and performance reliability (e.g., tensile strength, porosity) of each set of joints are analyzed. Simultaneously, the droplet transfer mode and characteristic data corresponding to each set of process parameters are recorded. Through comparative analysis, the correlation between droplet transfer characteristics and weld forming and performance is clarified. The droplet transfer characteristic data (including size, shape, frequency, and speed) corresponding to the static optimal welding process parameters are determined as the optimal droplet transfer characteristic parameters. This parameter serves as the benchmark for subsequent optimization of welding parameters for other pipe diameters and wall thicknesses, ensuring the consistency and repeatability of welding quality control.

[0045] S4. In the actual welding process, the static optimal welding process parameters are used as the benchmark setting values, and the optimal droplet transition characteristic parameters are used as the reference standard. The actual droplet transition behavior is monitored in real time using a droplet monitoring device, and the actual droplet characteristic parameters are compared with the reference standard. When a deviation occurs, the welding current, wire feed speed or welding torch height is dynamically adjusted using a fuzzy-PID control method to stabilize the actual droplet transition behavior within the range of the optimal droplet transition characteristic parameters, thereby achieving closed-loop control of the entire welding quality process.

[0046] In the actual welding process, the control system uses the pre-determined static optimal welding process parameters as initial settings, and simultaneously uses the determined optimal droplet transition characteristic parameters as a reference standard for real-time comparison. The droplet monitoring device continuously collects image information of the welding area, and after processing, extracts the actual characteristic parameters such as the size, shape, transition frequency, and speed of the current droplet.

[0047] The control system compares and analyzes the actual acquired droplet characteristic parameters with the preset optimal reference standard in real time. When the monitoring data shows that the actual droplet transition behavior deviates from the optimal parameter range, the system automatically activates the fuzzy-PID control algorithm to adjust the parameters. This control method uses fuzzy logic to process nonlinear characteristics and combines the precise adjustment characteristics of PID control to dynamically adjust key operating variables such as welding current, wire feed speed, or welding torch height, so that the actual droplet transition characteristics gradually converge and stabilize within the optimal parameter range.

[0048] This closed-loop control strategy is applied throughout all stages of the welding process. During the arc ignition stage, if unstable arc ignition or abnormal initial droplet formation is detected, the system prioritizes adjusting the welding current or welding torch height to ensure a smooth arc ignition process. Once the welding process stabilizes, if changes in the droplet transfer pattern or deviations in the welding trajectory are detected, the control system corrects the relevant welding parameters in real time according to preset rules. During the arc termination stage, the system gradually reduces the welding current and welding speed according to a predetermined program, while continuously monitoring the droplet transfer rate. Precise parameter control prevents the formation of defects such as craters and lack of fusion.

[0049] Example S1: System Overview and Equipment Configuration S11: Based on droplet transition monitoring, and combined with droplet transition control and a multi-objective hybrid optimization strategy, this system achieves high-precision and high-efficiency control of the automatic pulse TIG welding quality of thin-walled aluminum alloy tubes. The system mainly consists of welding equipment, a droplet monitoring device, a data processing device, and a feedback control system.

[0050] S12: Welding equipment: A high-precision automatic pulse TIG welding machine is selected. This welding machine has multiple welding process parameter adjustment functions, supports precise control of current parameters, wire feeding parameters, pulse parameters, etc., and can interact with the droplet monitoring and control system in real time.

[0051] S13: Molten Droplet Monitoring Device: Includes an information acquisition system (such as an industrial CMOS camera), an information processing system, and an information export system, used to acquire real-time video of molten droplet transition during the welding process and extract characteristic data such as droplet size, shape, transition frequency, and transition speed.

[0052] S14: Data processing device: integrates image processing software to filter and denoise the acquired droplet transition video, extract key indicators of droplet transition characteristics, and communicate with the welding machine system to provide feedback on optimization parameters, thereby achieving dynamic closed-loop control of the welding process.

[0053] S15: Feedback Control System: Employs a fuzzy-PID control method to dynamically adjust the welding current, wire feed speed, or welding torch height based on real-time monitored droplet transition characteristics data, ensuring stable weld joint quality.

[0054] S2: Preparations before welding S21: Ensure that the automatic TIG welding machine, droplet monitoring device (including high-speed CMOS camera), data processing software and control system have been correctly installed and debugged.

[0055] S22: Prepare aluminum alloy tubing, using 5B02 aluminum alloy thin-walled tubing with a wall thickness of 1 mm and a diameter of 63 mm as the welding target. Cut 100 mm long tubing to be welded using a CNC machine tool, clean the surface to be welded, remove oxide film and grease with sandpaper or wire brush, and wipe with acetone or anhydrous ethanol. Fix the tubing to be welded using a tack welding method.

[0056] S23: Check that the welding torch's contact tip, tungsten electrode, and shielding gas nozzle are properly installed and ensure sufficient welding wire allowance. Operate the automatic welding robotic arm, use the clamping device to fix the workpiece, and adjust the relative position of the welding torch and the workpiece to ensure that the rotation of the pipe and the movement of the welding torch are coordinated during the welding process.

[0057] S25: Set the initial welding parameters. The selection of the initial parameters is based on literature review and manual TIG welding experience, and serves as the benchmark for subsequent optimization experiments.

[0058] S3: A multi-objective hybrid optimization strategy was adopted to optimize the circumferential weld process of 5B02 aluminum alloy thin-walled tubes. The initial welding parameter range was defined as shown in Table 1. Table 1 Initial range of welding process parameters

[0059] S31: Based on the parameter range determined in S3, the welding test is arranged using the Plackett-Burman experimental design method.

[0060] S32: After completing the welding test, the key welding process parameters that have a significant impact on welding quality are screened based on the results of the analysis of variance. Usually, 5-6 parameters are selected as optimization variables. A multi-objective optimization system is constructed with expert visual inspection score, tensile strength, weld width, weld depth, and porosity as response targets.

[0061] S33: The welded circumferential weld specimens are visually inspected and scored by experienced welding inspectors; the weld width and depth are measured by cross-sectional metallography; and the porosity is quantitatively analyzed using X-ray inspection. After completing the above inspections, the circumferential weld specimens are cut, and tensile tests are performed according to GB / T 228.1-2010 standard to obtain the tensile strength data of the joint.

[0062] S4: Based on the clearly defined optimization variables and response objectives, use Design-Expert 14.0 software to arrange the welding test plan according to the Box-Behnken experimental design method, and execute the welding test according to the steps described in S3.

[0063] S41: Based on experimental data, establish an approximate response surface model between welding process parameters and each response objective, and combine the entropy weight-TOPSIS method to perform multi-objective integrated decision-making to complete the optimization process of welding process parameters.

[0064] S42: Welding verification tests are conducted using optimized welding process parameters. The results of response surface model predictions are quantitatively compared with the actual welding test results (including parameters such as tensile strength, weld width, and weld depth) to determine the final optimal combination of welding process parameters.

[0065] S5: During the above welding test, while simultaneously acquiring data on the circumferential weld joint, an industrial CMOS camera equipped with arc suppression function was used, combined with image filtering and noise reduction technology, to obtain a video sequence of droplet transfer during the welding process.

[0066] S51: The information processing system in the droplet monitoring device extracts characteristic data during the droplet transition process, including droplet size, shape, transition frequency, and transition speed. A mapping relationship is established between droplet transition characteristics and welding quality response targets, and the variation patterns are summarized to determine the optimal droplet transition characteristic parameters.

[0067] S6: In the actual welding process, considering the influence of factors such as welding process parameter disturbance and differences in the initial formability of the pipe, even on the basis of setting the optimal process parameters, it is still necessary to monitor the welding process in real time through the droplet monitoring device and dynamically adjust the parameters according to the actual droplet transition behavior in order to achieve stable control of welding quality.

[0068] S61: During the arc initiation stage, if unstable arc initiation or droplet transition deviates from the optimal range, the system automatically adjusts the welding current or welding torch height based on the fuzzy-PID control method to ensure a smooth arc initiation process and stable droplet transition.

[0069] S62: During the stable welding stage, when a deviation in the droplet transfer pattern or welding trajectory is detected, the system corrects the welding parameters in real time through a fuzzy-PID controller, so that the droplet transfer behavior is restored to the range of optimal characteristic parameters.

[0070] S63: During the arc termination phase, the system gradually reduces the welding current and welding speed according to a preset program, continuously monitors the droplet transfer rate, and dynamically adjusts the current and wire feed speed through a fuzzy-PID control method to effectively prevent defects such as arc craters and incomplete fusion.

[0071] S64: After the welding process is completed, the welding machine and fixture automatically return to the initial position, completing this welding cycle.

[0072] S7: New data collected during each welding process is used to verify and improve the control model. The actual welding parameters are still executed based on the determined optimal welding process parameters to ensure the consistency and repeatability of the process.

[0073] Although the embodiments of this application have been described above in conjunction with the accompanying drawings, this application is not limited to the specific embodiments and application fields described above. The specific embodiments described above are merely illustrative and instructive, not restrictive. Those skilled in the art can make many other forms based on the guidance of this specification and without departing from the scope of protection of the claims of this application, and these are all within the scope of protection of this application.

Claims

1. A method for quality control of pulsed TIG welding of thin-walled aluminum alloy tubes based on droplet transfer monitoring, characterized in that, Includes the following steps: S1. The Plackett-Burman experimental design method was used to screen welding process parameters and obtain key welding process parameters related to droplet transfer mode and welding quality. S2. Using the selected key welding process parameters as optimization variables, and expert visual inspection scores, tensile strength, weld width, weld depth, and porosity as response targets, an approximate response surface model between the welding process parameters and the welding quality target is established through Box-Behnken experiments; and the entropy weight-TOPSIS method is used to determine the comprehensive weight of each response target. The static optimal welding process parameters are obtained by solving the approximate model. The response surface approximation model was established using Design-Expert or Isight software, and the optimization variable type was determined to be extremely large, extremely small, or intermediate; among them, expert visual inspection score and tensile strength were extremely large, melt width and melt depth were intermediate, and porosity was extremely small. S3. During the welding test, a droplet monitoring device is used to collect droplet transition video in real time, and the video is processed to extract droplet transition characteristic data such as droplet size, shape, transition frequency and transition speed. The droplet transition characteristic data corresponding to the static optimal welding process parameters are used as the optimal droplet transition characteristic parameters. S4. In the actual welding process, the static optimal welding process parameters are used as the benchmark setting values, and the optimal droplet transition characteristic parameters are used as the reference standard. The actual droplet transition behavior is monitored in real time using a droplet monitoring device, and the obtained actual droplet characteristic parameters are compared with the reference standard. When a deviation occurs, the welding current, wire feed speed or welding torch height is dynamically adjusted using a fuzzy-PID control method to stabilize the actual droplet transition behavior within the range of the optimal droplet transition characteristic parameters, thereby achieving closed-loop control of the entire welding quality process. The fuzzy-PID control method is applied to the arc initiation stage, the stable welding stage, and the arc termination stage of the welding process. In the arc initiation stage, when arc instability is detected, the welding current or welding torch height is adjusted. In the stable welding stage, when changes in the droplet transfer pattern or welding trajectory are detected, the welding parameters are adjusted. In the arc termination stage, the welding current and welding speed are reduced to prevent arc craters or incomplete fusion defects.

2. The method for quality control of pulse TIG welding of thin-walled aluminum alloy tubing according to claim 1, characterized in that, The welding process parameters include current parameters, welding wire parameters, shielding gas parameters, pulse parameters, arc initiation parameters, and arc termination parameters.

3. The method for quality control of pulse TIG welding of thin-walled aluminum alloy tubes according to claim 1, characterized in that, The Plackett-Burman experimental design method screens out the significance levels of N-1 factors through N trials, where N is a multiple of 4, and determines the key welding process parameters based on analysis of variance.

4. The method for quality control of pulse TIG welding of thin-walled aluminum alloy tubes according to claim 1, characterized in that, The entropy weight-TOPSIS method includes: Calculate the entropy weight of each response objective, and determine the proportion of each response objective based on the entropy weight; The TOPSIS method is used to calculate the closeness of each welding process parameter scheme to the ideal solution, thereby determining the optimal welding process parameters.

5. The method for quality control of pulse TIG welding of thin-walled aluminum alloy tubing according to claim 1, characterized in that, The molten droplet monitoring device includes an information acquisition system, an information processing system, and an information export system; The information acquisition system includes an industrial CMOS camera, which is configured to suppress arc light interference and capture video of droplet transfer during the welding process. The information processing system includes image processing software configured to filter and denoise the droplet transition video, and decompose the video into multiple images to extract droplet transition feature data; The information export system is configured to output droplet transition characteristic data as a text data file and establish the correspondence between droplet change patterns and welding process parameters.

6. The method for quality control of pulse TIG welding of thin-walled aluminum alloy tubes according to claim 1, characterized in that, The aluminum alloy thin-walled tube is a 5B02 aluminum alloy thin-walled tube with a wall thickness of 1 mm and a diameter of 63 mm.

Images