Dynamic regulation and control device and method for aluminum alloy electron beam welding residual stress

Abstract

The invention discloses a dynamic regulation and control device and method for aluminum alloy electron beam welding residual stress, and belongs to the technical field of electron beam machining, and the device integrates a multi-sensing information acquisition system, a vibration system and a control system. The sensing information acquisition system measures the surface residual stress of a welding seam and a heat affected zone in real time and in situ in the welding process through a vacuum compatible X-ray probe and a detector, compares the predicted stress with a target value, and generates a regulation and control instruction of electron beam welding parameters through the control system; the electron beam welding behavior is adjusted in real time to dynamically control stress, and the vibration system is used for applying high-frequency micro-amplitude vibration to a local high-stress area of a weldment to promote residual stress release. The device integrates monitoring, regulation and stress release functions in a vacuum environment, can remarkably reduce welding residual stress and improve welding seam quality, and is particularly suitable for electron beam welding of large thin-wall aluminum alloy structural parts.

Description

Technical fields:

[0001] This invention relates to a device and method for dynamic control of residual stress during electron beam welding of aluminum alloys, belonging to the field of electron beam processing technology. Background technology:

[0002] With the increasing demands for material performance in high-end manufacturing fields such as aerospace and nuclear energy, the demand for welding special alloys, composite materials, and dissimilar materials is growing. Electron beam welding (EBW) is widely used in the manufacturing of high-performance structural components due to its advantages such as high energy density, low heat input, and large weld depth-to-width ratio.

[0003] However, traditional electron beam welding technology suffers from problems such as residual stress concentration and difficulty in controlling welding deformation during the welding of large thin-walled structures, dissimilar metal joints, and high-precision thin-walled box-shaped parts. Residual tensile stress can easily induce fatigue cracks, while welding deformation leads to assembly errors, seriously affecting the service performance and lifespan of structural components. For some materials (such as precipitation-hardening stainless steel, TiAl intermetallic compounds, and aluminum-lithium alloys), due to metallurgical properties or structural size limitations, residual stress cannot be eliminated using traditional heat treatment or overall vibration aging methods, becoming a bottleneck restricting the further promotion of electron beam welding technology.

[0004] Therefore, there is an urgent need for a technical means to monitor and dynamically control residual stress in real time during the welding process in order to improve welding quality and structural reliability. Summary of the Invention:

[0005] To address the problems existing in the prior art, this invention provides a dynamic control device and method for residual stress in electron beam welding of aluminum alloys. The device and method use a monitoring and feedback system to monitor the welding process in real time to obtain stress signals and further feed them back to the computer control system to determine whether the electron beam parameters need to be adjusted. This achieves control over residual stress in electron beam welding, thereby avoiding problems such as excessive stress during the welding process and improving the quality of electron beam welded joints.

[0006] This invention is achieved through the following technical solution:

[0007] This invention provides a device and method for dynamically controlling residual stress in electron beam welding of aluminum alloys, characterized by comprising the following steps:

[0008] (1) System initialization: Load the pre-trained residual stress prediction proxy model, and set the target residual stress threshold and control parameters;

[0009] (2) Synchronous real-time data acquisition: During the electron beam welding process, temperature field data from the infrared thermal imager, data from the detector, and process parameter data from the electron beam machine are acquired synchronously and in real time.

[0010] (3) Real-time prediction of residual stress: Input the real-time data collected in step (2) into the residual stress prediction proxy model to calculate the predicted post-weld residual stress value σ under the current state. vm ;

[0011] (4) Intelligent decision-making and control parameter generation: The predicted residual stress value σ obtained in step (3) is generated. vm Compare with the preset target stress threshold σ; if σ vm If the value is greater than σ, the adjustment amount of the electron beam welding parameters is calculated by the intelligent control algorithm, the electron beam incident point jumps to the next position, and the updated parameters are written into the control system storage module for use in the next cycle.

[0012] (5) Dynamic execution control: The welding parameter adjustment command generated in step (4) is sent to the electron beam machine, and the welding parameters are smoothly transitioned within the specified normal range through the control algorithm;

[0013] (6) Adjusted micro-vibration scheme settings: If there is still a high-stress area after welding, the control system sends the coordinates and dimensions (25mm×25mm) of the area to the vibration system, drives the composite slider to drive the ruby ​​ball to apply high-frequency micro-vibration to the area, and detects it through an X-ray detector;

[0014] (7) Closed-loop feedback: Repeat steps (2) to (6) until the welding process is completed, forming a closed-loop control system for continuous sensing, decision-making and regulation of residual stress.

[0015] Furthermore, in step (5), the execution of the welding parameter adjustment command is achieved through a parameter smooth transition control algorithm, so that the welding speed and power parameters continuously change to the target value within a time window of 50-200ms.

[0016] Furthermore, in step (3), the proxy model dataset is derived from the combination of finite element simulation and XRD measured data. The network structure is a three-layer feedforward neural network with seven features (temperature, process parameters, and diffraction data) in the input layer and the output is the predicted residual stress.

[0017] Furthermore, in step (4), the adjustment of the scanning parameters includes the adjustment of welding speed, current and power, wherein the welding speed is reduced when the predicted stress is higher than the target value;

[0018] Furthermore, the residual stress prediction proxy model is established through the following steps: obtaining a training dataset through numerical simulation and experiment, and using the dataset to train a neural network model to establish the mapping relationship between process parameters and post-weld residual stress;

[0019] Furthermore, the stress threshold σ is based on the material yield strength σ. s The safety factor k is set, where 0.4≤k≤0.7, and temperature decay correction is performed based on the molten pool temperature T. The corrected threshold is used to determine whether to start the stress adjustment subroutine.

[0020] Furthermore, when the electron beam incident point jumps, the distance between the jump position and the original incident point is sufficient to avoid stress concentration caused by heat accumulation. The jump position can be reasonably determined based on parameters such as material thermophysical properties and welding power.

[0021] Furthermore, when the electron beam incident point changes, the miniaturized X-ray stress meter is installed on the industrial robot as an end effector. The robot moves in coordination with the electron beam welding head to ensure the real-time monitoring system's information collection during the welding process.

[0022] Furthermore, when the residual stress after intermittent welding is still higher than the threshold, the vibration system immediately uses a ruby ​​ball to perform small-scale zone vibration on the center of the high-stress sub-region until the stress reaches the standard.

[0023] Furthermore, a dynamic control device for residual stress in electron beam welding of aluminum alloys is characterized by comprising: an electron beam welding multi-sensor information acquisition system, a vibration system, and a control system.

[0024] The multi-sensor information acquisition system includes: a miniaturized X-ray stress meter, i.e., an X-ray probe or detector, which is installed at the end of the robot and moves with the dual electron beam probes; an XRD data analysis module, which calculates the interplanar spacing in real time based on the acquired diffraction data; a temperature field monitoring unit, i.e., an infrared thermal imager, installed on the side of the electron beam probe; and a process parameter acquisition unit, which is used to acquire electron beam current, accelerating voltage, welding speed, and focusing current in real time from the electron beam machine controller. The electrical wiring and signal transmission of this multi-sensor information acquisition system, the vibration system, and the robot motion mechanism all adopt vacuum-compatible design and electromagnetic shielding treatment to ensure that each subsystem operates stably, reliably, and without interference in the electron beam welding environment with vacuum and strong electromagnetic interference.

[0025] The vibration system uses a composite slider and a ruby ​​ball to locally apply high-frequency micro-vibration to the welded parts to release residual stress. It includes an elastic support, a U-shaped vibration platform, a composite slider, a titanium alloy microrod, a mover coil, a stator permanent magnet ring, a piezoelectric stack, a ruby ​​ball, and a high-voltage driving power supply. The U-shaped vibration platform is provided with an elastic support below it, and a cross track is provided on the lower surface of the platform. The composite slider can move on the track, and the ruby ​​ball is driven by the piezoelectric stack to achieve high-frequency micro-vibration in the Z-axis direction.

[0026] The control system is connected to the multi-sensor information acquisition system and is used to predict post-weld residual stress in real time based on the acquired data and generate control commands. It includes: a residual stress prediction unit with a built-in machine learning proxy model; the model's inputs are real-time acquired temperature field characteristics, process parameters, and stress values, and its output is the predicted post-weld residual stress; and an intelligent control unit, used to compare the predicted residual stress with a preset threshold and calculate the adjustment amount of the electron beam parameters based on a control algorithm, and then adjust the value at σ. vm When the stress exceeds the preset stress threshold σ, the electron beam incident point is controlled to jump, and the position information is fed back to the multi-sensor information acquisition system to realize real-time information acquisition, and to drive the vibration system to perform high-frequency micro-vibration on the high-stress area to release residual stress.

[0027] Furthermore, the dual electron beam probes are positioned above the workpiece, forming a tandem pattern during welding. They weld the workpiece one after the other. Utilizing the auxiliary beam from the dual electron beam probes, after a jump in the incident point or parameter adjustment, a light modification scan is performed on the jump area to smooth the surface morphology. The dual electron beam probes constitute the core execution carrier for achieving stress-morphology dual-objective optimization in this invention. The main beam is responsible for the energy input required for basic welding and stress control; the auxiliary beam acts as a dynamic compensator, possessing three functions: (a) extending the solidification time of the molten pool to reduce porosity; (b) performing a modification scan on the weld after a jump in the incident point to smooth the morphology; and (c) assisting in the fine control of the thermal gradient by adjusting its power and focusing state. The coordination between the main and auxiliary beams is programmable, and their energy ratio and action sequence are direct outputs of the closed-loop decision algorithm of this invention, which is impossible to achieve with a single electron beam system.

[0028] The beneficial effects of this invention are as follows: This invention breaks through the traditional sequential process mode of "forming first, then relieving stress" in welding, and creatively proposes a parallel and collaborative control mechanism of "shape-following monitoring, real-time prediction, dynamic energy adjustment, and local energy release." This mechanism achieves proactive intervention in stress evolution in the time dimension, precise positioning and energy redistribution of high-stress sub-regions in the spatial dimension, and comprehensively utilizes electron beam thermal energy and vibrational mechanical energy in the energy dimension. These three elements work in a closed-loop linkage to jointly ensure synchronous and optimal control of stress and morphology in complex thermal cycles. On the one hand, the use of X-ray diffraction to directly measure lattice strain ensures the reliability of the control basis. Monitoring technology and machine learning prediction models enable "in-process" prediction of the "final residual stress," gaining valuable time for real-time control and solving the problem of post-weld measurement lag. Dual-electron beam welding transforms a single, concentrated ultra-high energy heat source into a programmable and coordinateable "composite energy field." Through the redistribution of energy in space and time, it achieves precise control of the weld pool thermal history and solidification process. Utilizing the compensation capability of dual-electron beams, it solves the problem of abrupt changes in weld morphology that may occur during dynamic control, achieving dual assurance of residual stress control and weld formation quality. Meanwhile, The intermittent welding strategy of this invention, through a combination of incident point jumps and smooth parameter transitions, allows for precise control of local energy input, cooling rate, and thermal gradient. Furthermore, the ruby ​​ball undergoes local vibration during welding, suppressing stress generation and instantly releasing existing localized high stress, reducing subsequent rework processes. The entire device is compact, with all steps completed within a vacuum chamber, eliminating the need to break the vacuum or transfer the workpiece, thus avoiding secondary contamination and additional clamping errors. It is particularly suitable for electron beam welding of large, thin-walled structural components such as aluminum alloy cabins, panels, and fuel tanks in aerospace applications requiring high cleanliness, precision, and reliability. This significantly extends weld fatigue life, reduces production costs, and has significant engineering application value. Attached image description:

[0029] Figure 1 This is a flowchart of a method for dynamically controlling residual stress in electron beam welding of aluminum alloys according to the present invention.

[0030] Figure 2 This is a schematic diagram of the overall structure of the dynamic control device;

[0031] Figure 3 This is a schematic diagram of intermittent electron beam welding.

[0032] Figure 4 This is a schematic diagram of a vibration system;

[0033] Figure 5 This is a diagram showing the installation of the composite slider.

[0034] In the diagram, 1-Dual electron beam probe; 2-Electron beam machine; 3-X-ray probe; 4-Detector; 5-Elastic support; 6-Control system; 7-Infrared thermal imager; 8-U-shaped vibration platform; 9-Welded component; 10-Composite slider; 11-Cross slide rail; 12-Ruby ball; 13-Titanium alloy microrod; 14-Piezoelectric stack; 15-Motor coil; 16-Spherical seat; 17-Stator permanent magnet ring. Detailed implementation method:

[0035] This invention provides a device and method for dynamically controlling residual stress in electron beam welding of aluminum alloys. To make the objectives, effects, and technical solutions of this invention clearer, the invention will be described in detail with reference to the accompanying drawings and examples. It should be understood that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the invention. For ease of description, the accompanying drawings show only the parts relevant to this invention, not all of the structures.

[0036] The following description, in conjunction with the accompanying drawings, details a device and method for dynamically controlling residual stress in electron beam welding of aluminum alloys provided by the present invention:

[0037] The experiment focused on the dynamic control of residual stress in electron beam welding of aluminum alloys. The substrate was a rectangular 2219 aluminum alloy with a thickness of 5mm and a diameter of 1000mm × 500mm. The welding experiment was conducted on a vibration platform 15. A commercially available vacuum-compatible XRD probe (Bruker D8 VENTURE series) could be used.

[0038] The apparatus of this invention includes an electron beam machine 2, whose dual electron beam probes 1 weld a workpiece 8. An infrared thermal imager 7 is aimed at the molten pool and heat-affected zone, and an X-ray probe 3 is fixed by a robot approximately 50-100 mm behind the weld (in the solidified area). A multi-sensor information acquisition module receives data from the infrared thermal imager 7, the detector 4, and the machine 2. The control system 6 includes data fusion, stress prediction, and intelligent control units. Through finite element simulation or numerous process experiments, post-weld residual stress data corresponding to different welding parameters (beam current, velocity, etc.), temperature fields, and instantaneous stresses are acquired to form a training sample set. This sample set is used to train a deep learning neural network, enabling the network to learn the complex nonlinear mapping relationship from process parameters to final stress. The trained model is deployed in the prediction unit.

[0039] The information is fed back to the control system 6, which determines the magnitude of the equivalent stress σ during the welding process based on the feedback parameters. vm A preset stress threshold σ is set before welding. When σ vm If σ is less than or equal to σ, it indicates that the residual stress after welding meets the expected requirements, and the welding process proceeds normally; when σ vmWhen the stress exceeds σ, it indicates that the stress does not meet the expected requirements, and the welding parameters are deemed unsuitable. The parameters need to be adjusted based on the stress. Upon receiving the judgment result, the control system jumps the electron beam incident point when the stress magnitude does not meet the expected requirements. The workpiece and the real-time monitoring system move together via a composite axis, adjusting the control system subroutine. This adjusts the electron beam power P, electron beam current I, and welding speed v via the electron gun high-voltage power supply. The adjustment of the electron beam parameters is within the specified normal range, and the updated parameters are written to the control system storage module for use in the next cycle. Simultaneously, the new parameter u1 = u0 + Δu* is written back to shared memory for use in the next cycle.

[0040] If a high-stress sub-region still exists after welding, the controller sends the coordinates (X0, Y0) of the upper left corner of the high-stress sub-region and its area of ​​25mm×25mm to the cross track 11. The composite slider 10 drives the ruby ​​ball 12 to perform high-frequency micro-vibration on the 25mm×25mm sub-region. The X-ray probe 3 continuously verifies until the residual stress is lower than the target value, so as to achieve the purpose of real-time control of residual stress in electron beam welding.

[0041] Specifically, in step (5), the execution of the welding parameter adjustment command is achieved through a parameter smooth transition control algorithm, so that the welding speed and power parameters continuously change to the target value within a time window of 50-200ms.

[0042] Specifically, in step (2), the stress data is obtained by measuring the solidified area behind the weld using the X-ray probe 3. Most of the electronic units of the X-ray probe host are placed outside the vacuum chamber, with only the X-ray probe and detector inside.

[0043] Specifically, in step (3), the proxy model dataset is derived from the combination of finite element simulation and XRD measured data. The network structure is a three-layer feedforward neural network with seven features (temperature, process parameters, and diffraction data) in the input layer and the output is the predicted residual stress.

[0044] Specifically, in step (4), the adjustment of the scanning parameters includes the adjustment of welding speed, current and power, wherein the welding speed is reduced when the predicted stress is higher than the target value;

[0045] Specifically, the residual stress prediction proxy model is established through the following steps: obtaining a training dataset through numerical simulation and experiments, and using the dataset to train a neural network model to establish the mapping relationship between process parameters and post-weld residual stress;

[0046] Specifically, the stress threshold σ is based on the material yield strength σ. s The safety factor k is set, where 0.4≤k≤0.7, and temperature decay correction is performed based on the molten pool temperature T. The corrected threshold is used to determine whether to start the stress adjustment subroutine.

[0047] Specifically, when the electron beam incident point jumps, the distance between the jump position and the original incident point is sufficient to avoid stress concentration caused by heat accumulation. The jump position can be reasonably determined based on parameters such as material thermophysical properties and welding power.

[0048] Specifically, when the electron beam incident point changes, the miniaturized X-ray stress meter is installed on the industrial robot as an end effector. The robot moves in coordination with the electron beam welding head to ensure the real-time monitoring system's information collection during the welding process.

[0049] Specifically, when the residual stress after intermittent welding is still higher than the threshold, the vibration system immediately uses a ruby ​​ball to perform small-scale zone vibration on the center of the high-stress sub-region until the stress reaches the standard.

[0050] Specifically, a dynamic control device for residual stress in electron beam welding of aluminum alloys is characterized by comprising: an electron beam welding multi-sensor information acquisition system, a vibration system, and a control system.

[0051] The multi-sensor information acquisition system includes: a miniaturized X-ray stress meter, namely an X-ray probe 3 and a detector 4, which is installed at the end of the robot and moves with the dual electron beam probe 1; an XRD data analysis module, which calculates the interplanar spacing in real time based on the acquired diffraction data; a temperature field monitoring unit, namely an infrared thermal imager 7, installed on the side of the electron beam probe 7; and a process parameter acquisition unit, which is used to obtain the electron beam current, accelerating voltage, welding speed and focusing current in real time from the electron beam machine controller. The electrical wiring and signal transmission of this multi-sensor information acquisition system, the vibration system and the robot motion mechanism are all designed with vacuum compatibility and electromagnetic shielding to ensure that each subsystem works stably, reliably and without interference in the electron beam welding environment with vacuum and strong electromagnetic interference.

[0052] The vibration system uses a composite slider 10 and a ruby ​​ball 12 to locally implement high-frequency micro-vibration on the welded part 9 to release residual stress. It includes an elastic support 5, a U-shaped vibration platform 8, a composite slider 10, a titanium alloy microrod 13, a mover coil 15, a stator permanent magnet ring 17, a piezoelectric stack 14, a ruby ​​ball 12, and a high-voltage drive power supply.

[0053] The control system is connected to the multi-sensor information acquisition system and is used to predict post-weld residual stress in real time based on the acquired data and generate control commands. It includes: a residual stress prediction unit with a built-in machine learning proxy model; the model's inputs are real-time acquired temperature field characteristics, process parameters, and stress values, and its output is the predicted post-weld residual stress; and an intelligent control unit used to compare the predicted residual stress with a preset threshold and calculate the adjustment amount of the electron beam parameters based on a control algorithm; and in σ... vmWhen the stress exceeds the preset stress threshold σ, the electron beam incident point is controlled to jump, and the position information is fed back to the real-time monitoring system to realize real-time information acquisition, and to drive the vibration system to perform high-frequency micro-vibration on the high-stress area to release residual stress.

[0054] Specifically, the X-ray probe 3 is installed at the end of the robot and moves with the electron beam head. The dual electron beam probe 1 is located above the workpiece 9, forming a tandem pattern during welding, welding the workpiece one after the other. Using the auxiliary beam in the dual electron beam probe, after the incident point jumps or parameters are adjusted, parametric modification scanning is performed on the affected weld area to improve the surface morphology. The infrared thermal imager 7 is installed to the side of the electron beam probe. An elastic support 14 is provided below the U-shaped vibration platform 8, and a cross slide rail 11 is provided on the lower surface of the platform. The composite slider 10 can move on the track. The ruby ​​ball 12 is driven by a piezoelectric stack to achieve high-frequency micro-vibration in the Z-axis direction. The frequency range of the high-frequency micro-vibration is 10-50kHz, and the amplitude is 1-10μm. This parameter range is determined through pre-experiment optimization based on the acoustic characteristics of aluminum alloy in the welding temperature range, dislocation motion activation energy, and threshold for avoiding micro-fatigue damage, and can be fine-tuned in actual applications according to the stress release effect.

[0055] Specifically, the dual electron beam probes are positioned above the workpiece, forming a tandem pattern during welding. They are used one after the other to weld the workpiece. Utilizing the auxiliary beam of the dual electron beam probes, after a jump in the incident point or parameter adjustment, a light-touch refining scan of ≤100ms is performed on the transition area to smooth the surface morphology. The dual electron beam probes constitute the core execution carrier for achieving stress-morphology dual-objective optimization in this invention. The main beam is responsible for the energy input required for basic welding and stress control; the auxiliary beam acts as a dynamic compensator, possessing three functions: (a) extending the solidification time of the molten pool to reduce porosity; (b) performing a refining scan on the weld after a jump in the incident point to smooth the morphology; and (c) assisting in the fine control of the thermal gradient by adjusting its power and focusing state. The coordination between the main and auxiliary beams is programmable, and their energy ratio and timing are direct outputs of the closed-loop decision algorithm of this invention, which is impossible to achieve with a single electron beam system.

[0056] The above description is only a preferred embodiment of the present invention. It should be noted that those skilled in the art can make several improvements without departing from the principle of the present invention, and these improvements should also be considered within the scope of protection of the present invention.

Claims

1. A device and method for dynamic regulation of electron beam welding residual stress of aluminum alloy, characterized in that, The real-time monitoring method comprises the following steps: (1) System initialization: load the pre-trained residual stress prediction proxy model, set the target residual stress threshold and control parameters; (2) Synchronous real-time data acquisition: during the electron beam welding process, synchronous real-time acquisition of temperature field data from the infrared thermal imager, data collected from the detector, and process parameter data from the electron beam machine; (3) Real-time residual stress prediction: input the real-time data collected in step (2) into the residual stress prediction surrogate model to calculate the predicted post-weld residual stress value σ vm ; (4) Intelligent decision and control parameter generation: compare the predicted residual stress value σ vm obtained in step (3) with the preset target stress threshold σ; if σ vm > σ, calculate the adjustment amount of the electron beam welding parameters through an intelligent control algorithm, the electron beam incident point jumps to the next position, and the updated parameters are written into the control system storage module for the next cycle. (5) Dynamic execution control: the welding parameter adjustment instruction generated in step (4) is sent to the electron beam machine, and the welding parameters are smoothly transitioned within the specified normal range through the control algorithm; (6) Adjusted micro-vibration scheme setting: if there are still high stress areas after welding, the control system will send the coordinates and size (25mm x 25mm) of the area to the vibration system to drive the ruby ball to apply high-frequency micro-vibration to the area, and the X-ray detector is used for detection; (7) Closed-loop feedback: repeat steps (2) to (6) until the welding process is completed, forming a closed-loop control system for continuous sensing, decision-making and control of residual stress.

2. The method of claim 1, wherein the method comprises: In step (5), the execution of the welding parameter adjustment instruction is realized through a parameter smooth transition control algorithm, which allows the welding speed and power parameters to change continuously to the target value within a time window of 50-200ms.

3. The method of claim 1, wherein the method comprises: In step (3), the proxy model dataset is derived from the combination of finite element simulation and XRD measurement data, the network structure is a three-layer feedforward neural network, the input layer has 7 features (temperature, process parameters, diffraction data), and the output is the predicted residual stress; in step (4), the adjustment of the scanning parameters includes adjustment of the welding speed, current and power, wherein when the predicted stress is higher than the target value, the welding speed is reduced.

4. The method of claim 1, wherein the method comprises: The residual stress prediction proxy model is established by the following steps: obtaining a training dataset through numerical simulation and experiment, training a neural network model using the dataset to establish a mapping relationship between process parameters and post-weld residual stress.

5. The method of claim 1, wherein the method comprises: The stress threshold σ is set in accordance with the yield strength σ s A safety factor k is set, where 0.4 ≤ k ≤ 0.7, and a temperature decay correction is made in accordance with the molten pool temperature T, and the corrected threshold is used to determine whether to start the stress adjustment subroutine.

6. The method of claim 1, wherein the method comprises: When the electron beam incident point jumps, the distance between the jumping position and the original incident point is sufficient to avoid stress concentration caused by heat accumulation, and the jumping distance can be reasonably determined according to the material thermal physical properties, welding power and other parameters.

7. The method of claim 1, wherein the method comprises: When the electron beam incident point jumps, the miniaturized X-ray stress meter is installed as an end effector on the industrial robot, and the robot moves with the electron beam welding head to ensure real-time information collection of the monitoring system during welding.

8. The method of claim 1, wherein the method comprises: When the residual stress after intermittent welding is still higher than the threshold, the vibration system immediately uses a ruby ball to perform small-range partition vibration on the center of the high-stress sub-area until the stress meets the standard.

9. A device for dynamic regulation of electron beam welding residual stress of an aluminum alloy, characterized in that, It comprises: An electron beam welding multi-sensor information acquisition system, a vibration system and a control system; The multi-sensing information acquisition system comprises a miniaturized X-ray stress meter, i.e., an X-ray probe and a detector, which are installed at the end of a robot and move with the movement of an electron beam head; an XRD data analysis module, which calculates the interplanar spacing in real time based on the collected diffraction data; a temperature field monitoring unit, i.e., an infrared thermal imager, which is installed at the side of the electron beam probe; and a process parameter acquisition unit, which is used to acquire the electron beam current, acceleration voltage, welding speed and focusing current from an electron beam machine controller in real time. The multi-sensing information acquisition system, the vibration system and the electrical wiring and signal transmission of the robot motion mechanism all adopt vacuum-compatible design and electromagnetic shielding treatment, so as to ensure that the subsystems work stably, reliably and without interference in the electron beam welding environment of vacuum and strong electromagnetic interference. The vibration system implements high-frequency micro-vibration on the local part of the ruby ball butt joint through a composite slider, releases residual stress, and comprises an elastic support, a U-shaped vibration platform, a composite slider, a titanium alloy micro-rod, a mover coil, a stator permanent magnet ring, a piezoelectric stack, a ruby ball and a high-voltage driving power supply. The U-shaped vibration platform is provided with an elastic support below, the lower surface of the platform is provided with a cross rail, the composite slider can move on the rail, and the ruby ball is driven by the piezoelectric stack to realize high-frequency micro-vibration in the Z-axis direction. The control system is connected with the multi-sensing information acquisition system, and is used for predicting the post-weld residual stress in real time based on the acquired data and generating a regulation instruction, which comprises a residual stress prediction unit, a machine learning agent model is built in the residual stress prediction unit, input of the model is temperature field characteristics, process parameters and stress values acquired in real time, and output is predicted post-weld residual stress; and an intelligent control unit is used for comparing the predicted residual stress with a preset threshold value, calculating an adjustment amount of an electron beam parameter based on a control algorithm, and controlling the electron beam incident point to jump when the residual stress is greater than the preset stress threshold value σ, and feeding back position information to the multi-sensing information acquisition system, so as to realize real-time acquisition of information and drive the vibration system to implement high-frequency micro-vibration on a high-stress area to release residual stress. vm The control system is connected with the multi-sensing information acquisition system, and is used for predicting the post-weld residual stress in real time based on the acquired data and generating a regulation instruction, which comprises a residual stress prediction unit, a machine learning agent model is built in the residual stress prediction unit, input of the model is temperature field characteristics, process parameters and stress values acquired in real time, and output is predicted post-weld residual stress; and an intelligent control unit is used for comparing the predicted residual stress with a preset threshold value, calculating an adjustment amount of an electron beam parameter based on a control algorithm, and controlling the electron beam incident point to jump when the residual stress is greater than the preset stress threshold value σ, and feeding back position information to the multi-sensing information acquisition system, so as to realize real-time acquisition of information and drive the vibration system to implement high-frequency micro-vibration on a high-stress area to release residual stress.

10. The device for dynamic control of residual stress in electron beam welding of aluminum alloys according to claim 9, characterized in that: The double electron beam probes are located above the welded part, form a tandem mode during welding, and weld the welded part in front and back. The auxiliary beam in the double electron beam probes is used to perform light modification scanning on the jump area after the incident point jumps or the parameters are adjusted, so as to smooth the surface topography.

Images