Brazing process state monitoring and regulation method based on infrared thermal field characteristic identification

Abstract

The present application relates to the technical field of brazing, in particular to a brazing process state monitoring and regulation method based on infrared thermal field feature recognition, comprising the steps of: collecting thermal field information in the brazing process in real time by using infrared thermal imaging technology, analyzing and extracting thermal field evolution characteristics, establishing the corresponding relationship between thermal field characteristics and brazing process state, identifying the real-time state of the whole process of brazing filler melting, liquid spreading, capillary filling and filling completion, and dynamically adjusting the process parameters according to the identification result, characterized in that it specifically comprises the following steps: S1, infrared thermal field collection, S2, thermal field image preprocessing, S3, thermal field feature extraction, S4, brazing process state recognition, and S5, process regulation. The present application can master the thermal field change law of the workpiece in real time during the brazing process, discover abnormal state in time, and dynamically adjust the process parameters such as heating power, temperature rising rate and holding time according to the identification result, so as to improve the monitorability, controllability and precision of process regulation of the brazing process.

Description

Technical Field

[0001] This invention relates to the field of brazing technology, specifically a method for monitoring and controlling the state of the brazing process based on infrared thermal field feature recognition. Background Technology

[0002] Brazing, as a classic modern material joining technology, has a heating temperature significantly lower than the melting point of the base material. It has natural advantages in avoiding the deterioration of the microstructure of the base material, controlling the thermal deformation of the workpiece, and reducing residual stress. At the same time, through the capillary filling and interfacial diffusion of liquid brazing filler metal, metallurgical joints with high density and excellent shear strength can be obtained, which has been widely used in aerospace, precision electronic components, and intelligent manufacturing.

[0003] In existing industrial production, quality control of brazing processes relies heavily on "preset process curve control modes." This means that before welding, technicians pre-set the heating power, heating rate, stepped holding temperature, holding time, and cooling regime based on experience or prior process experiments. However, in actual mass production processes, numerous uncontrollable disturbances are prevalent, including the geometrical tolerances of the workpieces to be welded, microscopic differences in assembly gaps, the degree of micro-oxidation on the base material and brazing filler metal surfaces, surface roughness, uniformity of brazing filler metal coating distribution, and the unevenness of the thermal field distribution within the brazing furnace. This means that even if the equipment strictly executes identical external heating process parameters, significant deviations in microscopic local thermodynamic conditions can still occur, leading to substantial fluctuations in the final brazed joint connection quality.

[0004] Throughout the dynamic process of brazing heating, the melting chamber or protective gas chamber is in a highly sealed state. Both the operator and the control system are in a "blind box welding" state, making it impossible to obtain real-time, direct, and quantitative information about the internal microscopic evolution. Specifically, it is impossible to know when and where the solid brazing filler metal begins to undergo phase change melting, when it reaches a complete liquefaction state, whether the dynamic wetting and spreading of the liquid brazing filler metal on the base material surface is sufficient, whether it smoothly completes the capillary filling of micro-gaps, and whether key process states such as substrate burns or localized overheating are caused by localized heat accumulation.

[0005] Given this lack of dynamic information, the industry currently relies solely on post-weld X-ray, ultrasonic, or metallographic analysis to evaluate brazed joint quality and conduct post-weld quality inspections. Because detection lags behind the welding process, if abnormalities occur during brazing heating, such as filler metal loss, insufficient capillary filling, or substrate overheating, the control system cannot promptly identify and intervene online, leaving the process defects to fester. Furthermore, discovering defects after the fact often means scrapping the entire batch of workpieces or requiring extremely difficult-to-repair secondary repairs. This results in significant economic losses, particularly in aerospace and other fields using expensive substrates or precious silver-based brazing filler metals.

[0006] A search revealed that patent CN122252727A provides a method and system for online monitoring of the quality of laser brazing of aluminum and steel. This method simultaneously acquires high-speed visual images of the welding area, infrared signals of the temperature field, stress acoustic emission signals, and welding process parameter signals. It then performs spatiotemporal correlation and feature-level fusion of the multi-physics data (e.g., extracting the phase difference between the visual area and the infrared temperature oscillation). Subsequently, the fused features are input into a pre-trained CNN-BiLSTM hybrid neural network model to identify welding defects such as porosity and cracks and predict trends. Finally, the correlation model outputs commands to adjust parameters such as laser power and speed in a closed loop.

[0007] This patent has the following drawbacks: 1. System hardware redundancy and limited deployment environment: This solution heavily relies on the synchronous fusion of multiple sensors, including vision, infrared, and acoustic emission sensors. Acoustic emission sensors need to be attached to the side of the component for contact measurement, while high-speed vision cameras are easily hampered by issues such as lack of light source and obstructed field of view in vacuum brazing furnaces or sealed protective atmosphere furnaces. The multi-sensor array and stringent millisecond-level hard synchronization requirements make it difficult to extend to brazing scenarios in extreme environments such as vacuum or sealed furnaces.

[0008] 2. The algorithm is too complex and relies on massive amounts of sample training: This technology uses deep learning (CNN-BiLSTM neural network) for state recognition. This black-box model not only has extremely high hardware computing power requirements and suffers from signal fusion and inference delays, but also heavily relies on massive amounts of pre-collected multiphysics training data with state labels. Once the solder grade or workpiece structure is changed, the model often needs to be retrained, which is time-consuming, resulting in weak process adaptability and generalization ability.

[0009] 3. Insufficient exploration of the multidimensional potential of infrared thermal fields: In this existing technology, infrared thermographic signals are mainly used as an auxiliary temperature measurement method for fusion (e.g., only the oscillation curve of the highest temperature is extracted and compared with the visual area for phase comparison). It completely ignores the high-dimensional spatiotemporal evolution laws carried by the infrared thermal field itself (such as heat diffusion rate, spatial temperature gradient, heat flow direction, etc.), and fails to achieve low-cost, high-efficiency control that can independently and in a closed loop complete the identification of the entire brazing process status based solely on a single infrared thermal field feature.

[0010] Patent CN122237791A discloses a welding inspection method for multilayer heat dissipation structures used in integrated circuits. This method employs a temperature sensor to collect the temperature value of the vapor chamber layer in the packaged state in real time during the brazing process, and a pressure sensor to collect the contact pressure between the forced cooling head and the workpiece in real time. The control unit calculates the temperature change rate dT / dt and the pressure change rate dP / dt in real time. When the temperature change rate is lower than a preset first threshold and the duration exceeds a first preset time, and simultaneously the absolute value of the pressure change rate exceeds a preset second threshold and the duration exceeds a second preset time, it is determined that the workpiece has experienced high-temperature cracking, thereby triggering protective responses such as stopping heating, alarm, and marking as unqualified.

[0011] This patent has the following drawbacks: 1. Limited Detection Target, Inability to Cover Core Brazing Quality Processes: This solution only detects the extreme abnormal failure state of "high-temperature cracking," completely ignoring the monitoring of normal core brazing processes (such as the initial melting of the brazing filler metal, liquid spreading, and capillary filling). It can only determine "whether the workpiece is cracked," but cannot determine "whether the weld is fully filled" or "whether the brazing filler metal is evenly spread," thus failing to achieve comprehensive monitoring and evaluation of the internal connection quality of the brazed joint.

[0012] 2. Reliance on specific contact-type tooling severely limits application scenarios: This technology heavily relies on the physical contact between the "pressure sensor" and the "forced cooling head" to collect pressure signals, which requires the brazing equipment to have specific pressurization and cooling structures. For pressureless brazing, complex irregularly shaped parts, or conventional vacuum brazing and protective atmosphere brazing scenarios where external mechanical pressure cannot be applied, this solution is completely ineffective because it cannot obtain pressure signals, resulting in extremely low versatility and promotional value.

[0013] 3. This is an open-loop control system that focuses on "post-incident damage control" and lacks dynamic process adjustment capabilities: After a crack is detected, this system can only perform passive protective actions such as "stop heating, alarm, and scrap." It cannot intervene in the early stages of defect formation (such as when there is uneven local temperature or poor spreading), nor can it salvage the workpiece or optimize the process by dynamically adjusting the heating power or holding time. Essentially, it is still an "advanced alarm" that lacks closed-loop control capabilities and cannot improve the yield of qualified products.

[0014] 4. Low information dimensionality and lack of global thermal field spatial evolution characteristics: This technology only uses the time change rate (one-dimensional time series signal) of single-point (or local) temperature and single-point pressure for judgment, and completely fails to obtain global two-dimensional thermal field spatial distribution information of the workpiece surface. This makes it unable to capture physical phenomena such as the release of latent heat of solder melting and local thermal field distortion caused by liquid metal flow, making it difficult to achieve refined and multi-dimensional state recognition of complex brazing processes.

[0015] Therefore, further improvements are still needed in brazing process status monitoring and brazing process adjustment to enhance the monitorability, controllability, and consistency of joint quality in the brazing process. Summary of the Invention

[0016] This invention addresses the problem that existing brazing processes rely heavily on preset process parameters for control, making it impossible to acquire real-time information on key process states such as filler metal melting, liquid spreading, capillary filling, and completion of filling. This results in poor weld quality consistency, delayed process control, and difficulty in timely defect detection. This invention provides a brazing process state monitoring and control method based on infrared thermal field feature recognition. This method utilizes infrared thermography to acquire thermal field information in real time during the brazing process. By analyzing the evolution characteristics of the thermal field, it extracts feature parameters that reflect changes in the brazing state, establishes a correspondence between thermal field features and key brazing process states, and achieves real-time identification of the entire process state, including filler metal melting, liquid spreading, capillary filling, and completion of filling. Based on the identification results, it dynamically adjusts process parameters such as heating power, heating rate, and holding time. This improves the monitorability of the brazing process, the accuracy of process control, and the consistency and reliability of brazed joint quality, reducing product rework rates and production costs. It provides a new technical solution for intelligent monitoring and process optimization of the brazing process, effectively solving the problems mentioned in the background.

[0017] To address the aforementioned problems, the technical solution adopted by this invention is: a brazing process state monitoring and control method based on infrared thermal field feature recognition, comprising: real-time acquisition of thermal field information during the brazing process using infrared thermography; analysis of thermal field evolution characteristics to extract characteristic parameters reflecting changes in the brazing state; establishment of a correspondence between thermal field characteristics and brazing process states; real-time identification of the entire process states of filler metal melting, liquid spreading, capillary filling, and completion of filling; and dynamic adjustment of process parameters based on the identification results. The method is characterized by the following steps: S1. Infrared thermal field acquisition: An infrared detection device is installed on the brazing equipment to continuously acquire the thermal field of the brazing process through the infrared observation window on the side of the brazing furnace cavity, and to obtain the infrared thermal image sequence of the workpiece surface in real time at a preset sampling frequency. S2. Thermal image preprocessing: Preprocess the acquired infrared thermal image to eliminate the influence of environmental radiation and equipment noise. S3. Thermal field feature extraction: Extract thermal field feature parameters from the processed thermal field image and establish the evolution curve of each feature parameter over time. S4. Brazing process status identification: Based on the extracted thermal field characteristic parameters and their variation patterns, the brazing process is identified in real time; the thermal field evolution process includes the solid heating stage, the brazing filler metal melting stage, the liquid spreading and capillary filling stage, and the filling completion and cooling stage. The identification results include insufficient solder melting, incomplete spreading and capillary filling of liquid solder, complete spreading and capillary filling of solder, local overheating, thermal field discontinuity, and abnormal changes in hot spot area; S5. Process Control: During brazing, the liquidus temperature T of the brazing filler metal is used as the control factor. L Set the target brazing temperature as a reference, the target brazing temperature is T. L + (20~80℃), heating rate controlled at 5~20℃ / min, vacuum degree is 10. -3 Pa, after reaching the target brazing temperature, hold for 1 to 30 minutes; When the solder is found to be insufficiently melted, the control and analysis system will increase the heating power by 2% to 10% based on the current power, or extend the holding time by 30 to 180 seconds. After one adjustment cycle, the infrared thermal field image will be reacquired and the solder melting state will be judged until the preset melting state is reached. When the liquid brazing filler metal has not yet completed its spreading and capillary filling, maintain the current brazing temperature, control the temperature fluctuation to not exceed ±5~10℃, re-acquire the infrared image of the brazing area every 10~60s, and continue to judge the spreading and filling status of the liquid brazing filler metal. When the brazing filler metal spreading and capillary filling are completed, the heat preservation is ended and the cooling stage begins; When local overheating is detected, reduce the heating power by 5% to 15%, or reduce the heating rate by 1 to 5°C / min. If the local abnormality is not eliminated after 2 to 3 consecutive adjustment cycles, stop heating or terminate the brazing process.

[0018] Preferably, the infrared thermal image preprocessing in step S2 includes emissivity correction, background temperature compensation, image denoising, and region of interest extraction.

[0019] Preferably, in step S3, the thermal field characteristic parameters include the highest temperature, average temperature, temperature change rate, temperature gradient, hot spot area, hot spot area change rate, thermal diffusion range, and thermal field uniformity. As the brazing process proceeds, the rate of temperature change sequentially goes through the solid heating stage, the brazing filler metal melting stage, the liquid spreading and capillary filling stage, and the filling completion and cooling stage. When the curve shows an endothermic inflection point, it is identified as entering the solder melting stage; subsequently, the temperature change rate enters a relatively stable plateau region, which is identified as the liquid solder spreading and capillary filling stage; when the temperature change rate gradually approaches zero and remains stable, it is identified as the filling is complete, and the equipment can be controlled to end the heat preservation and enter the cooling stage.

[0020] Preferably, in step S4, during the solid heating stage, the thermal field distribution on the workpiece surface is uniform, and the isotherms maintain a basically regular distribution; during the brazing filler metal melting stage, as the temperature rises, the brazing filler metal begins to melt, and the phase change endothermics cause local thermal field depressions, resulting in significant distortion of the isotherms; during the liquid spreading and capillary filling stage, after the brazing filler metal is completely melted, it completes spreading and gap filling under capillary action, the hot spot area continues to expand, the heat diffusion range gradually increases, and the thermal field tends to be continuous and uniform; during the filling completion stage, the temperature change rate gradually approaches zero and remains stable.

[0021] Preferably, in step S4, when the control and analysis system identifies that the temperature of the brazing area has approached or reached the liquidus temperature of the brazing filler metal, but the area of ​​the melting area of ​​the brazing filler metal is lower than the preset melting area threshold, or the melting area of ​​the brazing filler metal does not continue to expand within a certain period of time, it is determined that the brazing filler metal is not melting sufficiently. When it is detected that the solder has melted, but the liquid solder spreading area continues to expand, or the solder joint filling area has not yet reached the preset filling ratio, it is determined that the liquid solder has not yet completed spreading and capillary filling. When the rate of change of the liquid brazing filler metal spreading area or filling area is less than 1% to 5% within 20 to 60 seconds, and the filling ratio of the target area of ​​the brazing seam reaches the preset value, it is determined that the brazing filler metal spreading and capillary filling are basically completed. When the highest local temperature in the brazing area is detected to be 20-50°C higher than the set brazing temperature, or when the area of ​​the local hot spot increases rapidly or the thermal gradient changes abruptly, it is determined that there is a risk of local overheating.

[0022] Compared with the prior art, the present invention provides a method for monitoring and controlling the state of brazing process based on infrared thermal field feature recognition, which has the following beneficial effects: This invention achieves real-time identification of key process states such as brazing melt, liquid spreading, capillary filling, and completion of filling by real-time acquisition of infrared thermal field information during the brazing process and analysis of thermal field evolution characteristics. Compared with traditional control methods that rely on preset process parameters and post-weld inspection, this invention can grasp the changes in the workpiece's thermal field in real time during the brazing process, promptly detect abnormal states such as local overheating, incomplete filling, and insufficient filling, and dynamically adjust process parameters such as heating power, heating rate, and holding time based on the identification results, thereby improving the monitorability, controllability, and accuracy of process adjustment in the brazing process. This invention reduces welding defects caused by fluctuations in process parameters, lowers product rework rates and material waste, and improves the consistency, reliability, and production efficiency of brazed joints. It provides a new technical solution for intelligent monitoring and process optimization of the brazing process, and has good engineering application value and promotion prospects. By extracting multiple thermal field characteristic parameters such as temperature change rate, temperature gradient, hot spot area, hot spot area change rate, heat diffusion range, and thermal field uniformity, and comprehensively judging the brazing status, rather than monitoring based on a single temperature parameter, it can achieve full-process status identification and quality control without contacting the workpiece or affecting the brazing process. Attached Figure Description

[0023] Figure 1 This is a schematic diagram of the brazing equipment of the present invention; Figure 2 This is a schematic diagram of the temperature change rate characteristic curve and state identification during the brazing process of the present invention; Figure 3 This is a schematic diagram illustrating the evolution of typical infrared thermal field characteristics during the brazing process of this invention; Figure 4 a is a schematic diagram of the cross-sectional structure of the lap joint weld obtained in Embodiment 1 of the present invention; Figure 4 b is a schematic diagram of the shear strength of the welded joint in Embodiment 1 of the present invention; Figure 5 a is a schematic diagram of the cross-sectional structure of the lap joint weld obtained in Embodiment 2 of the present invention; Figure 5 b is a schematic diagram of the shear strength of the welded joint in Embodiment 2 of the present invention; Figure 6 This is a flowchart of the method of the present invention.

[0024] The components include: 1. Atmosphere control components; 2. Brazing furnace cavity; 3. Heating system; 4. Infrared observation window; 5. Infrared detection equipment; 6. Detector bracket; 7. Cooling water tank; 8. Frame and base; 9. Molecular pump; 10. Mechanical pump; and 11. Control and analysis system. Detailed Implementation

[0025] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0026] Reference Figure 1-6 This invention proposes a method for monitoring and controlling the state of brazing process based on infrared thermal field feature recognition, including brazing equipment for implementing the method, such as... Figure 1 As shown, the brazing equipment includes an atmosphere control component 1, a brazing furnace chamber 2, a heating system 3, an infrared observation window 4, an infrared detection device 5, a detector bracket 6, a cooling water tank 7, a frame and base 8, a molecular pump 9, a mechanical pump 10, and a control and analysis system 11. The mechanical pump 10 and the molecular pump 9 are connected to the brazing furnace chamber 2 to establish and maintain the vacuum environment required for the brazing process. The atmosphere control component 1 is used to introduce a protective atmosphere into the brazing furnace chamber 2 according to process requirements. The heating system 3 is located inside the brazing furnace chamber 2 and is used to heat the components to be brazed. The infrared observation window 4 is located on the side of the brazing furnace cavity 2, corresponding to the area to be brazed. The infrared detection device 5 is located outside the infrared observation window 4 via the detector bracket 6, and is used to collect infrared radiation information and thermal field images of the brazing area in real time. The control and analysis system 11 is connected to the infrared detection device 5 and the heating system 3 respectively, and is used to receive infrared thermal field data, identify the melting and spreading state of the brazing filler metal, and dynamically adjust process parameters such as heating power, heating rate and holding time according to the identification results, thereby forming an online closed-loop control of the brazing process.

[0027] Brazing process status monitoring and control based on infrared thermal field feature recognition includes the following steps: S1, Infrared thermal field acquisition An infrared thermal imaging acquisition device (infrared detection device) is installed on the brazing equipment. The infrared observation window on the side of the furnace body continuously collects the thermal field of the brazing process. The infrared thermal image sequence of the workpiece surface is obtained in real time at a preset sampling frequency, realizing online non-contact monitoring of the entire brazing process.

[0028] S2, Thermal Field Image Preprocessing The acquired infrared thermal images are preprocessed, including emissivity correction, background temperature compensation, image denoising, and region of interest (ROI) extraction, to eliminate the influence of environmental radiation and equipment noise and obtain stable and reliable thermal field images, providing basic data for subsequent feature extraction.

[0029] S3, Thermal Field Feature Extraction The thermal field characteristic parameters that reflect the changes in the state of the brazing process are extracted from the processed thermal field image, including the maximum temperature, average temperature, temperature change rate (dT / dt), temperature gradient, hot spot area, hot spot area change rate, heat diffusion range and thermal field uniformity, etc., and the evolution curves of each characteristic parameter over time are established.

[0030] Among them, the temperature change rate curve is as follows Figure 2 As shown, during the brazing process, the rate of temperature change sequentially goes through the solid heating stage, the solder melting stage, the liquid spreading and capillary filling stage, and the filling completion and cooling stage. When the curve shows a clear endothermic inflection point (characteristic point A), it can be identified that the solder has begun to melt; subsequently, the rate of temperature change enters a relatively stable plateau region (characteristic segment B), corresponding to the liquid solder spreading and capillary filling process; when the rate of temperature change gradually approaches zero and remains stable (characteristic point C), it is determined that the capillary filling is complete, and the equipment can be controlled to end the heat preservation and enter the cooling stage.

[0031] S4. Brazing process status identification Based on the extracted thermal field characteristic parameters and their variation patterns, real-time status identification of the brazing process is performed. A typical thermal field evolution process is as follows: Figure 3 As shown.

[0032] During the solid heating stage, the thermal field distribution on the workpiece surface is uniform, and the isotherms maintain a basically regular distribution. Figure 3 (a)); During the solder melting stage, as the temperature rises, the solder begins to melt, and the phase change endothermics cause a local thermal field depression, resulting in significant distortion of the isotherms. Figure 3 (b)); During the liquid spreading and capillary filling stage, after the brazing filler metal is completely melted, it completes spreading and gap filling under capillary action, the hot spot area continues to expand, the heat diffusion range gradually increases, and the thermal field tends to be continuous and uniform. Figure 3 (c) If local overheating, incomplete filling, or insufficient filling occurs during brazing, local hot spots or thermal field faults will form in the thermal field, manifested as abnormal isotherm distribution and abrupt changes in thermal gradient ( Figure 3 (d) The system automatically identifies the stage and abnormal state of the brazing process based on the above-mentioned thermal field evolution characteristics.

[0033] S5, Process Control The brazing process is dynamically adjusted based on real-time identification results. When insufficient melting of the brazing filler metal is detected, the heating power is increased or the holding time is appropriately extended; when the liquid brazing filler metal is detected as not having fully spread, the current heating state is maintained to continue capillary filling; when filling is detected as complete, the holding time ends and the cooling stage begins; when local hot spots, thermal field discontinuities, or abnormal changes in the area of ​​hot spots are detected, it is determined that there are abnormal states such as local overheating, incomplete filling, or insufficient filling, and the heating power, heating rate, or holding time are automatically adjusted. If necessary, the brazing process is terminated to prevent further formation of defects, thus achieving online closed-loop control of the brazing process.

[0034] The specific process control is as follows: During the brazing process, the liquidus temperature T of the brazing filler metal used is controlled. L Set the target brazing temperature as a reference, preferably T. L + (20~80℃), the heating rate is preferably controlled at 5~20℃ / min, and the vacuum degree is preferably 10- 3 Pa, after reaching the target brazing temperature, hold for 1 to 30 minutes.

[0035] When the control and analysis system detects that the temperature of the brazing area is close to or has reached the liquidus temperature of the brazing filler metal, but the area of ​​the melted area is lower than the preset melting area threshold, or the melted area does not continue to expand within a certain period of time, it determines that the brazing filler metal is not fully melted. At this time, the control system increases the heating power by 2% to 10% based on the current power, or extends the holding time by 30 to 180 seconds. After one adjustment cycle, the infrared thermal field image is reacquired and the melting state of the brazing filler metal is judged again, until the preset melting state is reached. One adjustment cycle refers to the control and analysis system completing one closed-loop control process of "infrared thermal field acquisition—brazing status identification—process parameter adjustment—adjustment effect feedback detection".

[0036] When it is detected that the brazing filler metal has basically melted, but the liquid brazing filler metal spreading area continues to expand, or the brazing seam filling area has not yet reached the preset filling ratio, it is determined that the liquid brazing filler metal has not completed spreading and capillary filling. At this time, the current brazing temperature is maintained, and the temperature fluctuation is preferably controlled to not exceed ±5~10℃. The infrared image of the brazing area is reacquired every 10~60 seconds, and the spreading and filling status of the liquid brazing filler metal is continuously assessed.

[0037] When the rate of change of the liquid brazing filler metal spreading area or filling area is less than 1% to 5% within 20 to 60 seconds, and the filling ratio of the target area of ​​the brazing seam reaches the preset value, it is determined that the brazing filler metal spreading and capillary filling are basically completed, the control system ends the heat preservation stage, and controls the equipment to enter the cooling process.

[0038] When the highest local temperature in the brazing area is detected to be 20–50°C higher than the set brazing temperature, or when the local hot spot area rapidly increases or the thermal gradient shows abnormal abrupt changes, it is determined that there is a risk of local overheating. In this case, the heating power is reduced by 5%–15%, or the heating rate is reduced by 1–5°C / min. If the local abnormality is not eliminated after 2–3 consecutive adjustment cycles, the heating is stopped or the brazing process is terminated. If local low-temperature areas, thermal field discontinuities, or brazing seam target areas are not covered by high-temperature liquid brazing filler metal for a long time in the infrared image, it is determined that there may be incomplete or insufficient filling. The brazing filler metal is further spread and capillary filling is promoted by extending the holding time and adjusting the heating power.

[0039] As a specific embodiment of the present invention: Example 1: Taking vacuum brazing lap joints as an example, an infrared detection device (infrared thermal imager) is installed in the observation window of the vacuum brazing furnace to continuously acquire thermal field data throughout the entire brazing process. The system acquires thermal field images at a preset frequency and performs image preprocessing and extraction of thermal field feature parameters. When a clear inflection point appears on the temperature change rate curve, the melting of the brazing filler metal is detected. Subsequently, the liquid spreading and capillary filling processes are judged based on changes in the hot spot area and the thermal diffusion range. When the hot spot area changes tend to stabilize, the filling is determined to be complete, and the equipment is controlled to enter the cooling stage.

[0040] Effect verification The vacuum brazing lap joint connection was completed using the brazing process status monitoring method based on infrared thermal field characteristics described in this invention. The brazing filler metal composition was Ag. 26 Cu 38 Ge 30 Sn6 (atomic ratio), with pure copper as the base material. The system can accurately identify key process stages such as the start of solder melting, liquid spreading, capillary filling, and completion of filling, and controls the equipment to enter the cooling stage after filling is completed.

[0041] Figure 4 (a) is the cross-sectional microstructure of the lap joint weld obtained in this embodiment. As can be seen from the figure, the brazing filler metal can fully fill the lap gap, the weld is continuous and the microstructure is uniform, the interface is well bonded, and no obvious macroscopic defects such as unfilled joints, cracks and pores are observed. Figure 4 (b) is the shear tensile curve of the joint in this embodiment. The load-bearing capacity of the joint continues to increase during the loading process, and it breaks after reaching the peak value. This shows that the method of the present invention can obtain a brazed joint with stable quality and good mechanical properties.

[0042] Example 2: Taking vacuum brazing lap joints as the object, thermal field monitoring was performed using the same method as in Example 1. When the system detects abnormal hot spots, thermal field discontinuities, or abnormal changes in the area of ​​hot spots in a localized area, it determines that there are abnormal conditions such as localized overheating, incomplete filling, or insufficient filling during the brazing process. Based on the identification results, the heating power, heating rate, and holding time are adjusted. After the thermal field returns to a normal distribution, the brazing process continues.

[0043] For local overheating, reduce the heating power by 10% or reduce the heating rate by 3℃ / min; if the local abnormality is not eliminated after 2 to 3 consecutive adjustment cycles, stop heating or terminate the brazing process. To address incomplete or insufficient filling, when the control and analysis system identifies localized low-temperature zones, reduced thermal continuity, or insufficient filling ratio in the brazing joint area, it determines that the brazed joint is not fully filled. In this case, the control system promotes further melting and flow of the liquid brazing filler metal by extending the holding time, increasing the heating power, or adjusting the heating rate. Specifically, the holding time is extended by 30–180 seconds (potentially 90 seconds), or increased by 2%–8% (potentially 5%) based on the current heating power. If necessary, the heating rate is increased by 1–3 °C / min (potentially 2 °C / min) to maintain the temperature of the brazing area above the liquidus temperature of the brazing filler metal and promote capillary filling along the overlap gap. After 1–3 control cycles, if the infrared thermal field detection shows that the temperature distribution in the brazing joint area has returned to uniformity and the liquid brazing filler metal coverage area has reached the preset filling threshold, the subsequent brazing process continues. If unfilled areas still exist after multiple consecutive control cycles, the joint is determined to have a filling abnormality, and the brazing process is terminated.

[0044] Effect verification The infrared thermal field monitoring method described in this invention is used to monitor the vacuum brazing process in real time. The brazing filler metal is composed of Ag. 26 Cu 38 Ge 30 Sn6 (atomic ratio), with pure copper as the base material. When the system detects an abnormal local thermal field, it promptly adjusts the heating power, heating rate, and holding time to restore the thermal field to a normal state before continuing the brazing process.

[0045] Figure 5 (a) shows the cross-sectional microstructure of the lap joint weld obtained in Example 2. As can be seen from the figure, after dynamic adjustment of process parameters, the brazing filler metal can also fully fill the lap gap, the weld microstructure is uniform and continuous, the interface is dense, and no obvious unfilled or other welding defects were found. Figure 5(b) is the shear tensile curve of the joint in Example 2. It can be seen that the joint has a high load-bearing capacity and the tensile curve changes smoothly, indicating that the brazed joint after dynamic control still maintains good mechanical properties.

[0046] Comparison of two implementation examples Depend on Figure 4 and Figure 5 As shown in the weld microstructure, the joints obtained in Examples 1 and 2 both have continuous and dense weld microstructures, with sufficient filler metal filling, good interface bonding, and no obvious defects such as cracks, voids, or unfilled areas were found. Figure 4 (b) and Figure 5 As shown in (b), the shear tensile curves indicate that the joints obtained in both embodiments have high load-bearing capacity, with peak shear strength reaching approximately 400 MPa or more. The curve variation patterns are basically consistent, demonstrating that the brazing process status monitoring and process control method based on infrared thermal field characteristics proposed in this invention can achieve stable and reliable brazing connections under both normal and abnormal process control conditions, ensuring that the brazed joint has good microstructure and mechanical properties. This verifies the effectiveness, stability, and engineering application value of the technical solution of this invention.

[0047] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for monitoring and controlling the state of brazing process based on infrared thermal field feature recognition, comprising: real-time acquisition of thermal field information during the brazing process using infrared thermography; analysis of thermal field evolution characteristics to extract characteristic parameters reflecting changes in brazing state; establishment of a correspondence between thermal field characteristics and brazing process state; real-time identification of the entire process state of filler metal melting, liquid spreading, capillary filling, and completion of filling; and dynamic adjustment of process parameters based on the identification results. The method is characterized by... Specifically, the following steps are included: S1. Infrared thermal field acquisition: An infrared detection device is installed on the brazing equipment to continuously acquire the thermal field of the brazing process through the infrared observation window on the side of the brazing furnace cavity, and to obtain the infrared thermal image sequence of the workpiece surface in real time at a preset sampling frequency. S2. Thermal image preprocessing: Preprocess the acquired infrared thermal image to eliminate the influence of environmental radiation and equipment noise. S3. Thermal field feature extraction: Extract thermal field feature parameters from the processed thermal field image and establish the evolution curve of each feature parameter over time. S4. Brazing process status identification: Based on the extracted thermal field characteristic parameters and their variation patterns, the brazing process is identified in real time; the thermal field evolution process includes the solid heating stage, the brazing filler metal melting stage, the liquid spreading and capillary filling stage, and the filling completion and cooling stage. The identification results include insufficient solder melting, incomplete spreading and capillary filling of liquid solder, complete spreading and capillary filling of solder, local overheating, thermal field discontinuity, and abnormal changes in hot spot area; S5. Process Control: During brazing, the liquidus temperature T of the brazing filler metal is used as the control factor. L Set the target brazing temperature as a reference, the target brazing temperature is T. L + (20~80℃), heating rate controlled at 5~20℃ / min, vacuum degree is 10. -3 Pa, after reaching the target brazing temperature, hold for 1 to 30 minutes; When the solder is found to be insufficiently melted, the control and analysis system will increase the heating power by 2% to 10% based on the current power, or extend the holding time by 30 to 180 seconds. After one adjustment cycle, the infrared thermal field image will be reacquired and the solder melting state will be judged until the preset melting state is reached. When the liquid brazing filler metal has not yet completed its spreading and capillary filling, maintain the current brazing temperature, control the temperature fluctuation to not exceed ±5~10℃, re-acquire the infrared image of the brazing area every 10~60s, and continue to judge the spreading and filling status of the liquid brazing filler metal. When the brazing filler metal spreading and capillary filling are completed, the heat preservation is ended and the cooling stage begins; When local overheating is detected, reduce the heating power by 5% to 15%, or reduce the heating rate by 1 to 5°C / min. If the local abnormality is not eliminated after 2 to 3 consecutive adjustment cycles, stop heating or terminate the brazing process.

2. The brazing process status monitoring and control method based on infrared thermal field feature recognition according to claim 1, characterized in that, In step S2, the infrared thermal image preprocessing includes emissivity correction, background temperature compensation, image denoising, and region of interest extraction.

3. The brazing process status monitoring and control method based on infrared thermal field feature recognition according to claim 1, characterized in that, In step S3, the thermal field characteristic parameters include the highest temperature, average temperature, temperature change rate, temperature gradient, hot spot area, hot spot area change rate, thermal diffusion range, and thermal field uniformity. As the brazing process proceeds, the rate of temperature change sequentially goes through the solid heating stage, the brazing filler metal melting stage, the liquid spreading and capillary filling stage, and the filling completion and cooling stage. When the curve shows an endothermic inflection point, it is identified as entering the solder melting stage; subsequently, the temperature change rate enters a relatively stable plateau region, which is identified as the liquid solder spreading and capillary filling stage; when the temperature change rate gradually approaches zero and remains stable, it is identified as the filling is complete, and the equipment can be controlled to end the heat preservation and enter the cooling stage.

4. The brazing process status monitoring and control method based on infrared thermal field feature recognition according to claim 1, characterized in that, In step S4, during the solid heating stage, the thermal field distribution on the workpiece surface is uniform, and the isotherms maintain a basically regular distribution. During the brazing filler metal melting stage, as the temperature rises, the brazing filler metal begins to melt, and the phase change endothermics cause local thermal field depressions, resulting in significant distortion of the isotherms. During the liquid spreading and capillary filling stage, after the brazing filler metal is completely melted, it completes spreading and gap filling under capillary action, the hot spot area continues to expand, the heat diffusion range gradually increases, and the thermal field tends to be continuous and uniform. During the filling completion stage, the temperature change rate gradually approaches zero and remains stable.

5. The brazing process status monitoring and control method based on infrared thermal field feature recognition according to claim 4, characterized in that, In step S4, when the control and analysis system identifies that the temperature of the brazing area has approached or reached the liquidus temperature of the brazing filler metal, but the area of ​​the melting area of ​​the brazing filler metal is lower than the preset melting area threshold, or the melting area of ​​the brazing filler metal does not continue to expand within a certain period of time, it is determined that the brazing filler metal is not melting enough. When it is detected that the solder has melted, but the liquid solder spreading area continues to expand, or the solder joint filling area has not yet reached the preset filling ratio, it is determined that the liquid solder has not yet completed spreading and capillary filling. When the rate of change of the liquid brazing filler metal spreading area or filling area is less than 1% to 5% within 20 to 60 seconds, and the filling ratio of the target area of ​​the brazing seam reaches the preset value, it is determined that the brazing filler metal spreading and capillary filling are basically completed. When the highest local temperature in the brazing area is detected to be 20-50°C higher than the set brazing temperature, or when the area of ​​the local hot spot increases rapidly or the thermal gradient changes abruptly, it is determined that there is a risk of local overheating.

Images