Brazing forming process of light high-reliability heat dissipation cold plate

Abstract

The invention belongs to the technical field of quality detection, and provides a light high-reliability heat dissipation cold plate brazing forming process which comprises the steps that through multi-power testing, the surface temperature conditions of a heat dissipation cold plate finished product under different powers are obtained, a local overheating area is screened, the diameter of a welding spot air hole in the heat dissipation cold plate finished product is obtained, and a suspected welding spot defect area is obtained; carrying out coincidence analysis on the local overheating area and the suspected welding spot defect area to obtain a co-occurrence sub-area, carrying out synchronism analysis on the temperature and the pore diameter in the co-occurrence sub-area under overload power, judging whether relevance exists or not, if yes, obtaining a local temperature-welding spot defect relevance model, and if not, obtaining a local temperature-welding spot defect relevance model; and whether a high error exists in the current air hole diameter detection result of the heat dissipation cold plate finished product or not is judged based on the correlation model, if yes, correction is carried out, heat dissipation cold plate defect classification is carried out according to the air hole diameter of the heat dissipation cold plate finished product after correction, and the production quality and use reliability of the heat dissipation cold plate can be improved.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of quality detection, and particularly relates to a brazing forming process of a light high-reliability heat dissipation cold plate. BACKGROUND

[0002] In the manufacturing technology of the heat dissipation cold plate, the heat dissipation cold plate is adapted to the development demand of the continuous improvement of the power density of electronic equipment, and the heat dissipation performance and structural reliability of the heat dissipation cold plate as a core thermal management component directly determine the equipment operation stability and service life. At present, the quality detection link of the heat dissipation cold plate still has many technical limitations, and it is difficult to meet the reliability control demand under full working conditions.

[0003] In the prior art, the traditional detection is mostly focused on the temperature distribution under the rated power of the heat dissipation cold plate, and the overload working condition that may occur in actual application is not fully simulated, so that only the overheating area under the conventional operation can be identified, and the local overheating area that appears only at the overload is missed. Secondly, the existing detection means mostly relies on surface temperature monitoring equipment such as an infrared thermal imager, and can only capture surface temperature abnormalities, ignoring the hidden influence of internal solder joint defects on the heat dissipation performance, hindering the heat conduction path, causing heat to accumulate locally on the surface to form overheating, and not analyzing the associated influence of surface overheating and internal structural defects, so that the core problem source of overheating cannot be accurately located. Finally, the detection data lacks a historical calibration mechanism, and the detection error is not corrected by modeling the correlation between the historical temperature and the average diameter of the pores, so that the repair accuracy is not accurate.

[0004] Therefore, the application provides a brazing forming process of a light high-reliability heat dissipation cold plate. SUMMARY

[0005] In order to make up for the deficiencies of the prior art and solve at least one technical problem proposed in the background art.

[0006] The technical scheme adopted by the application to solve the technical problems is: a brazing forming process of a light high-reliability heat dissipation cold plate, comprising:

[0007] During internal quality detection of the heat dissipation cold plate product, the surface temperature of the heat dissipation cold plate product under different powers is obtained through multi-power testing, and a local overheating area is screened out;

[0008] The diameter of the solder joint pores in the heat dissipation cold plate product is detected by using the micro-CT technology, and a suspected solder joint defect area is obtained;

[0009] The local overheating area and the suspected solder joint defect area are subjected to coincidence analysis, and a co-occurrence sub-area is obtained. The temperature and the pore diameter in the co-occurrence sub-area under the overload power are subjected to synchronous analysis, and whether there is a correlation between the local overheating and the suspected solder joint defect is determined.

[0010] Based on the existence of correlation, the historical detection data is analyzed to obtain a local temperature-welding defect correlation model, and based on the correlation model, it is judged whether the current heat dissipation cold plate product pore diameter detection result exists high error, if exists, it is corrected and after correction, the heat dissipation cold plate defect classification is carried out according to the heat dissipation cold plate product pore diameter.

[0011] Further, the surface temperature of the heat dissipation cold plate product under different power is obtained as follows:

[0012] Access the rated power, use the infrared thermal imager to shoot the heat dissipation cold plate for full coverage, record the temperature distribution of the heat dissipation cold plate surface under the rated power, and obtain the rated infrared distribution map;

[0013] The rated infrared distribution map is divided into several sub-regions with equal area to obtain the rated sub-region, and the temperature threshold is set, and the rated sub-region with temperature exceeding the temperature threshold is marked as an abnormal rated sub-region;

[0014] Access the overload power, use the infrared thermal imager to shoot the heat dissipation cold plate for full coverage, record the temperature distribution of the heat dissipation cold plate surface under the overload power, and obtain the overload infrared distribution map;

[0015] Compare the temperature change of each overload sub-region with the corresponding rated sub-region to obtain the temperature rise amplitude value;

[0016] If the temperature rise amplitude value of the overload sub-region is greater than or equal to the temperature rise amplitude threshold, the corresponding overload sub-region is marked as an abnormal overload sub-region.

[0017] Further, the process of screening the local overheating area is as follows:

[0018] The abnormal overload sub-region and the abnormal rated sub-region are compared, and the non-overlapping part of the abnormal overload sub-region and the abnormal rated sub-region is marked as a local overheating area.

[0019] Further, the process of obtaining the suspected welding defect area is as follows:

[0020] Use the micro-CT device to emit X-rays to scan the heat dissipation cold plate in all directions, and the X-rays penetrate the heat dissipation cold plate and are received by the micro-CT device receiver to obtain three-dimensional signal data of the heat dissipation cold plate and generate a three-dimensional image of the internal structure of the heat dissipation cold plate;

[0021] In the three-dimensional image, the welding point area is identified and located by using image analysis software to obtain the pore diameter, and the pore diameter is compared with the pore diameter threshold;

[0022] If the pore diameter is greater than or equal to the pore diameter threshold, the corresponding sub-region is marked as a suspected welding defect area.

[0023] Further, the local overheating area and the suspected solder defect area are subjected to coincidence analysis to obtain a co-occurrence sub-area, and the process is as follows:

[0024] The sub-area that is both a local overheating area and a suspected solder defect area is marked as a co-occurrence sub-area.

[0025] The number of co-occurrence sub-areas, the total number of heat dissipation cold plate sub-areas, the total number of local overheating sub-areas, and the total number of suspected solder defect sub-areas are counted.

[0026] The total number of local overheating sub-areas and the total number of suspected solder defect sub-areas are subjected to difference processing to obtain a deviation area number, and the deviation area number and the total number of heat dissipation cold plate sub-areas are subjected to ratio processing to obtain a region difference value.

[0027] The number of co-occurrence sub-areas and the total number of local overheating sub-areas are subjected to ratio processing to obtain a local heat proportion value.

[0028] The local heat proportion value and the region difference proportion value are subjected to difference processing to obtain a coincidence coefficient.

[0029] Further, the temperature and the pore diameter in the co-occurrence sub-area under overload power are subjected to synchronization analysis, and the process is as follows:

[0030] Based on the co-occurrence sub-area, the temperature and the pore diameter in each co-occurrence sub-area under overload power are obtained and integrated respectively to obtain a co-occurrence temperature sequence and a co-occurrence pore sequence.

[0031] The co-occurrence sub-area positions in the co-occurrence temperature sequence and the co-occurrence pore sequence are arranged in the same way.

[0032] The Pearson correlation coefficient between the co-occurrence temperature sequence and the co-occurrence pore sequence is calculated, and the absolute value is taken to obtain a synchronization coefficient about the temperature and the pore diameter.

[0033] Further, it is determined whether there is a correlation between the local overheating and the suspected solder defect, and the process is as follows:

[0034] The coincidence coefficient and the synchronization coefficient are summed to obtain a local heat defect correlation value.

[0035] If the local heat defect correlation value is greater than or equal to a local heat defect correlation threshold value, it indicates that there is a correlation between the local overheating and the solder defect (pore diameter).

[0036] Further, the historical detection data is analyzed to obtain a local temperature-solder defect correlation model, and the process is as follows:

[0037] Based on the co-occurrence temperature sequence and the co-occurrence pore sequence obtained from the co-occurrence sub-region, the co-occurrence temperature sequence and the co-occurrence pore sequence are fitted by using the least square method, and a local temperature-welding defect correlation model is obtained.

[0038] Further, the process of judging whether the current heat dissipation cold plate product pore diameter detection result exists high error is as follows:

[0039] Based on the suspected welding defect sub-region, the temperature value in the suspected welding defect sub-region is input into the local temperature-welding defect correlation model, the actual pore diameter of the suspected welding defect sub-region is output, and compared with the pore detection diameter of the suspected welding defect sub-region;

[0040] If the error between the actual pore diameter and the pore detection diameter exceeds the allowable error range, the suspected welding defect sub-region is marked as a high error sub-region;

[0041] The area number proportion of the high error sub-region in the suspected welding defect sub-region contained in the current heat dissipation cold plate product is counted to obtain a high error proportion value;

[0042] The pore diameter error of the high error sub-region is calculated by absolute difference proportion with the nearest allowable error range endpoint value to obtain an error out-of-limit value of the high error sub-region;

[0043] The error out-of-limit values of all high error sub-regions are processed by averaging to obtain a high error degree value;

[0044] The high error proportion value and the high error degree value are summed to obtain a high error performance value;

[0045] If the high error performance value is greater than or equal to the high error performance threshold value, it indicates that the current heat dissipation cold plate product pore diameter detection result exists high error.

[0046] Further, the process of classifying the heat dissipation cold plate defect according to the heat dissipation cold plate product pore diameter is as follows:

[0047] If there is a pore diameter greater than or equal to 0.5mm in the internal region of the heat dissipation cold plate product, the heat dissipation cold plate product is marked as a severe defect heat dissipation cold plate;

[0048] If there is no pore diameter greater than or equal to 0.5mm in the internal region of the heat dissipation cold plate product, the heat dissipation cold plate product is marked as a slight defect heat dissipation cold plate.

[0049] The beneficial effects of the present application are as follows:

[0050] The heat dissipation cold plate is tested at multiple powers, temperature detection is performed at the rated power and the overload power respectively, the non-overlapping area is marked as a local overheating area through area overlapping comparison, the weld point pore diameter in the heat dissipation cold plate product is detected through the micro-CT technology, the area where the pore diameter is greater than a pore diameter threshold is marked as a suspected weld point defect area, overlapping analysis is performed based on the local overheating area and the suspected weld point defect area, a co-occurrence sub-area is obtained, Pearson correlation coefficient is used for synchronism analysis, whether there is a higher possibility of correlation is judged, if there is correlation, a local temperature-weld point defect correlation model is obtained, whether there is a high error in the detection result is judged, the pore detection diameter is corrected according to the pore diameter error, and the heat dissipation cold plate defect classification is performed according to the pore diameter, which is beneficial to improve the production quality and use reliability of the heat dissipation cold plate, reduce the equipment failure rate caused by heat dissipation problems, accurately identify and repair the local overheating and weld point defect problems, effectively prolong the service life of the heat dissipation cold plate, and meanwhile, the predictive maintenance strategy based on the correlation model can discover potential quality risks in advance, and provide sufficient preparation time for operation and maintenance personnel. BRIEF DESCRIPTION OF DRAWINGS

[0051] The application will be further described below with reference to the drawings.

[0052] Figure 1 It is a step flow chart of the brazing forming process of the light-weight high-reliability heat dissipation cold plate described in the embodiments of the application.

[0053] Figure 2 It is a logic judgment diagram of the brazing forming process of the light-weight high-reliability heat dissipation cold plate described in the embodiments of the application. DETAILED DESCRIPTION

[0054] In order to make the technical means, creative features, purposes and effects realized by the application easy to understand, the application will be further described below with reference to the specific embodiments.

[0055] Embodiment 1: The brazing forming process of the light-weight high-reliability heat dissipation cold plate described in the embodiments of the application, comprising:

[0056] S1. Material pretreatment and assembly positioning; the light-weight metal plate material meeting the design requirements is used as a cold plate base material, and the cold plate base material is pretreated, wherein the pretreatment includes but is not limited to surface oil removal, oxide film removal and drying treatment, etc., for example:

[0057] Surface oil removal: ethanol, acetone or trichloroethylene is used as a cleaning agent, and the oil is removed through ultrasonic cleaning, and the cleaning agent is immediately washed away with deionized water after cleaning to avoid residual cleaning agent;

[0058] Oxide film removal: soak the cold plate substrate in a 10% to 30% nitric acid solution or sodium hydroxide solution, the soaking time is adjusted according to the thickness of the oxide film, generally 5 to 20 minutes, and then rinse with deionized water;

[0059] Drying treatment: place the cleaned cold plate substrate in a drying oven at a temperature of 80 to 120 degrees Celsius for 30 to 60 minutes to ensure that the substrate surface is free of moisture;

[0060] Assembly positioning: according to the design drawing, the pretreated cold plate substrate is accurately assembled with the heat dissipation cold plate, water inlet and outlet pipe and other components, and is fixed with positioning clamps to ensure the relative position between the components is accurate and error-free, and to prepare for the subsequent brazing process;

[0061] S2. Core brazing forming; brazing filler metal and flux selection: according to the characteristics of the cold plate substrate (light metal plate such as aluminum alloy), aluminum-silicon brazing filler metal is selected, which can be in the form of brazing filler metal foil or brazing filler metal paste, and corrosion-free aluminum brazing flux is selected to avoid subsequent corrosion of the substrate.

[0062] Pre-brazing assembly review: the cold plate substrate, heat dissipation cold plate, water inlet and outlet pipe assembly fixed by the positioning clamps are reviewed to ensure that the cold plate spacing deviation is ≤0.1mm, and the water pipe and substrate interface coaxiality is ≤0.05mm, to avoid assembly deviation affecting the brazing quality;

[0063] Preheating treatment: place the assembled components into the brazing furnace and use stepwise preheating to raise the temperature from room temperature to 200°C and then to 350°C to reduce the thermal stress between the substrate and the components and prevent deformation;

[0064] Brazing heating and holding: after preheating, the temperature is raised to 580-620°C and held for 15-30 minutes; during this period, nitrogen gas is introduced as a protective atmosphere, with a gas pressure of 0.12-0.15MPa, to isolate air and prevent oxidation of the substrate, ensuring that the brazing filler metal flows fully and fills the gaps between the welds;

[0065] Cooling control: after brazing holding is completed, first reduce to 300°C, and then switch to natural cooling to room temperature to avoid rapid cooling leading to weld point embrittlement or component cracking;

[0066] S3. Post-processing and quality detection; including, flux residue cleaning: place the brazing formed heat dissipation cold plate into a 5% to 10% concentration sodium carbonate solution at a temperature of 40-60°C for 20-30 minutes of ultrasonic cleaning to remove the flux residue on the surface and in the gaps, and then rinse with deionized water three times to ensure no chemical residue;

[0067] Surface drying and finishing: after cleaning, the cold plate is placed in a drying oven at a temperature of 80-100 DEG C for 40-50 min to ensure that the surface is free of moisture; after drying, the burrs at the welding site are lightly sanded with sandpaper to ensure that the cold plate is flat and free of sharp protrusions;

[0068] Appearance quality detection: visual detection is used, and the surface of the welding point is required to be smooth and continuous, free of cracks, incomplete fusion, excess or missing filler material, etc., the heat dissipation cold plate is free of deformation and lodging, the inlet and outlet pipe interfaces are free of skew, etc., to ensure that the quality requirements are met;

[0069] Sealing detection: the internal flow channel of the cold plate is subjected to water pressure sealing test: 0.3-0.5 MPa deionized water is introduced into the flow channel, and the pressure is maintained for 30 min, during which the interface and the welding site are wiped with a dry cloth, and there is no water leakage and sweating phenomenon, which is qualified, and at the same time, a helium mass spectrometer leak detector is used for detection, and the leakage rate needs to be ≤1x10⁻ 9 Pa·m³ / s, to ensure that the flow channel is free of small leaks;

[0070] Mechanical properties and internal quality detection: 10% of the finished products are randomly selected for tensile test to test the tensile strength of the welding point, internal welding points are detected by micro-CT technology, and there are no internal defects such as pores and incomplete fusion with a diameter of ≥0.5 mm, and the heat dissipation performance is sampled and tested to ensure that the heat dissipation design requirements are met;

[0071] Example 2: please refer to Figure 1 - Figure 2 The brazing forming process of the light high-reliability heat dissipation cold plate, as shown in the present application, comprises:

[0072] Step one: during the internal quality detection of the finished heat dissipation cold plate, the surface temperature of the finished heat dissipation cold plate under different powers is obtained through multi-power test, and the local overheating area is screened out;

[0073] The finished heat dissipation cold plate is obtained, and multi-power detection is performed to determine whether there is a local overheating area, and the specific process is as follows:

[0074] The heat dissipation cold plate is connected to a temperature intelligent control system, and the rated power is set as the power value when the heat dissipation cold plate normally works (for example, 250 W), and the overload power is 1.2 to 1.5 times of the rated power (for example, 360 W), and the purpose of setting the rated power and the overload power is to simulate the running state of the heat dissipation cold plate under different working loads;

[0075] It should be noted that the temperature intelligent control system includes but is not limited to an infrared thermal imager and an adjustable power supply;

[0076] During the rated power test stage, the rated power is connected, the infrared thermal imager is used to shoot the whole heat dissipation cold plate, the temperature distribution of the whole heat dissipation cold plate surface under the rated power is recorded, and the rated infrared distribution map is obtained;

[0077] It should be noted that the infrared thermal imager can capture and display the temperature values of each part of the heat dissipation cold plate under rated power in real time, form an intuitive temperature distribution map, and obtain a rated infrared distribution map;

[0078] It should be further noted that the shooting mode is vertical shooting of the surface of the heat dissipation cold plate;

[0079] The rated infrared distribution map is analyzed and processed, the rated infrared distribution map is divided into a plurality of sub-regions with equal areas to obtain rated sub-regions, a temperature threshold is set, the rated sub-regions with a temperature exceeding the temperature threshold in the rated infrared distribution map are marked by using the infrared thermal imager, and the rated sub-regions where the marked regions are located are marked as abnormal rated sub-regions;

[0080] It should be noted that the setting of the temperature threshold needs to refer to the heat resistance limit of the cold plate substrate and the surface coating to avoid material thermal deformation or aging caused by excessively high temperature, and at the same time, in combination with the safe working temperature range of the cold plate adapted to the electronic components, a basic range is determined, combined with pre-experiment verification, a plurality of heat dissipation cold plates that have passed factory detection are selected, and the temperature distribution of the whole domain is recorded by using the infrared thermal imager under the condition that the heat dissipation cold plates are continuously operated under rated power until the temperature is stable, the maximum temperature and the upper limit of the 95% confidence interval of all qualified cold plates are counted, and the upper limit of the confidence interval + 5% safety margin is taken as the temperature threshold;

[0081] In the overload power test stage, the overload power is connected, the infrared thermal imager is used to shoot the whole heat dissipation cold plate, the temperature distribution of the whole heat dissipation cold plate surface under the overload power is recorded, and an overload infrared distribution map is obtained;

[0082] The overload infrared distribution map is analyzed and processed, and the overload infrared distribution map is also divided into sub-regions with equal areas and consistent with the number of the rated infrared distribution map to obtain overload sub-regions;

[0083] Through image analysis software, the temperature change of each overload sub-region and the corresponding rated sub-region is compared, the temperature rise amplitude value is obtained, and compared with the temperature rise amplitude threshold;

[0084] If the temperature rise amplitude value of the overload sub-region is greater than or equal to the temperature rise amplitude threshold, the corresponding overload sub-region is marked as an abnormal overload sub-region;

[0085] The abnormal overload sub-region and the abnormal rated sub-region are compared, the overlapping part of the abnormal overload sub-region and the abnormal rated sub-region is marked as an overlapping region, and the non-overlapping part of the abnormal overload sub-region and the abnormal rated sub-region is marked as a local overheating region;

[0086] It should be noted that the setting of the temperature rise threshold needs to consider the material properties of the heat dissipation cold plate, the safety threshold of electronic components and historical detection data. First, based on the thermal conductivity and thermal expansion coefficient of the cold plate substrate, combined with the heat resistance limit of the surface coating, the material safety temperature range is determined. Second, referring to the maximum allowable operating temperature of the adapted electronic components, a safety margin of 10% to 15% is reserved to prevent component overheating and damage. Finally, through the historical detection database, the temperature variation curve of the same type of cold plate under rated and overload conditions is extracted. The upper limit of the 95% confidence interval of the temperature rise is calculated using statistical methods and used as the initial threshold. The temperature rise threshold needs to be verified by actual working conditions and dynamically adjusted according to the material and structural characteristics of different types of cold plates to ensure detection accuracy and reliability.

[0087] The image analysis software includes but is not limited to FLIR Tools+ and Fluke SmartView.

[0088] It should also be noted that the non-overlapping part is marked as a local overheating area because of the accurate capture of performance differences under different workloads. Specifically, the rated power simulation of the heat dissipation cold plate reflects the daily routine running state. The abnormal rated sub-area under this state reflects the inherent overheating risk that exists under normal load. This risk may be due to design defects (such as blocked heat dissipation channels) or insufficient material intrinsic thermal performance. Such problems will continue to be exposed. The overload power simulates the peak operation of the device and extreme high load and other special scenarios. The abnormal overload sub-area is the area that overheats and has a temperature rise above the threshold under high load. The overlapping part of the abnormal rated sub-area means that this area has an overheating risk under both normal and extreme conditions, which is a "global inherent problem". In contrast, the non-overlapping part is an "overheating area that only appears under overload conditions". This area can meet the demand for heat dissipation under normal load, but when the power is increased and the heat generation rate exceeds the local heat dissipation limit, abnormal heating occurs, indicating insufficient adaptability of the local heat dissipation structure. If this non-overlapping area is not marked as a local overheating area, the unique risks under overload conditions will be hidden and the "no abnormality under normal conditions, problem under extreme conditions" hidden danger cannot be accurately identified.

[0089] After marking separately, both inherent defects (overlapping area) and load-dependent defects (non-overlapping area) can be clearly distinguished, and data support can be provided for subsequent analysis of non-overlapping areas.

[0090] It can be understood that the purpose of obtaining the local overheating area is to accurately locate the unique overheating area of the heat dissipation cold plate under different power operating states. In the actual working process of the heat dissipation cold plate, the rated power represents the conventional operating state, and the overload power simulates an extreme or high-load operating scenario. The overlapping area means that this part has an overheating risk under both conventional and extreme states, which may be due to inherent defects in design or material itself, while the non-overlapping area reflects the performance difference of the heat dissipation cold plate under different load conditions, for example, some parts only abnormally heat up under overload, which may be related to unreasonable local heat dissipation structure, uneven material thermal conductivity, or brazing quality fluctuation. By distinguishing the overlapping and non-overlapping areas, the heat dissipation design can be optimized, such as improving the cold plate layout, adjusting the brazing filler process, or selecting materials with better thermal conductivity, thereby improving the reliability of the heat dissipation cold plate in the full operating range. In addition, the analysis of non-overlapping areas can also provide a basis for fault prediction, such as identifying parts prone to failure under overload in advance, and guiding the development of a more scientific maintenance cycle.

[0091] Step two: Use micro-CT technology to detect the diameter of the weld point pores in the finished heat dissipation cold plate to obtain the suspected weld defect area;

[0092] Place the heat dissipation cold plate that has completed the multi-power test and marked the local overheating area on the detection table of the micro-CT equipment;

[0093] Use the micro-CT equipment to emit X-rays for all-around scanning of the heat dissipation cold plate. After the X-rays penetrate the heat dissipation cold plate, they are received by the micro-CT equipment receiver to obtain three-dimensional signal data of the heat dissipation cold plate and generate a three-dimensional image of the internal structure of the heat dissipation cold plate;

[0094] In the three-dimensional image, use image analysis software to identify and locate the weld point area to obtain the diameter of the pores and compare it with the pore diameter threshold;

[0095] If the pore diameter is greater than or equal to the pore diameter threshold, the corresponding sub-area is marked as a suspected weld defect area;

[0096] If the pore diameter is less than the pore diameter threshold, the corresponding sub-area is marked as a non-suspected weld defect area;

[0097] It should be noted that the setting of the pore diameter threshold needs to be combined with the use requirements of the heat dissipation cold plate and the brazing quality standard. For example, the pore diameter threshold can be 0.5 mm;

[0098] Example 3: Please refer to Figure 1 - Figure 2 As shown in the figure, the brazing forming process of the lightweight and high-reliability heat dissipation cold plate according to the embodiments of the present application comprises:

[0099] Step three: perform coincidence analysis on the local overheating area and the suspected solder defect area to obtain a co-occurrence sub-area, and perform synchronism analysis on the temperature and the pore diameter in the co-occurrence sub-area under the overload power to determine whether there is a correlation between the local overheating and the suspected solder defect;

[0100] The heat dissipation cold plate is divided into a plurality of sub-areas with equal areas;

[0101] It should be noted that the division manner is the same as that of the above-mentioned rated infrared distribution map and overload infrared distribution map;

[0102] The sub-area that is both a local overheating area and a suspected solder defect area is marked as a co-occurrence sub-area;

[0103] The number of co-occurrence sub-areas, the total number of heat dissipation cold plate sub-areas, the total number of local overheating sub-areas, and the total number of suspected solder defect sub-areas are counted;

[0104] The total number of local overheating sub-areas and the total number of suspected solder defect sub-areas are subjected to difference processing to obtain a deviation area number, and the deviation area number and the total number of heat dissipation cold plate sub-areas are subjected to ratio processing to obtain a region difference value;

[0105] The number of co-occurrence sub-areas and the total number of local overheating sub-areas are subjected to ratio processing to obtain a local heat proportion value;

[0106] The local heat proportion value and the region difference proportion value are subjected to difference processing to obtain a coincidence coefficient;

[0107] Based on the co-occurrence sub-area, the temperature and the pore diameter in each co-occurrence sub-area under the overload power are obtained and integrated respectively to obtain a co-occurrence temperature sequence and a co-occurrence pore sequence;

[0108] The co-occurrence sub-area positions in the co-occurrence temperature sequence and the co-occurrence pore sequence are arranged in the same way;

[0109] The Pearson correlation coefficient between the co-occurrence temperature sequence and the co-occurrence pore sequence is calculated, and the absolute value is taken to obtain a synchronism coefficient about the temperature and the pore diameter;

[0110] The coincidence coefficient and the synchronism coefficient are summed to obtain a local heat defect correlation value;

[0111] In some embodiments, the local heat defect correlation value is compared with a local heat defect correlation threshold value;

[0112] If the local heat defect correlation value is greater than or equal to the local heat defect correlation threshold value, it indicates that there is a correlation between the local overheating and the solder defect (pore diameter);

[0113] If the local heat defect correlation value is less than the local heat defect correlation threshold, it indicates that there is no correlation between the local overheating and the weld defect (pore diameter);

[0114] It should be noted that the setting of the local heat defect correlation threshold needs to be based on the positive samples of verified correlation and the negative samples of verified no correlation in the historical detection database, the initial threshold is determined through ROC curve analysis, the initial threshold is adjusted combined with the reliability requirements of application scenarios, the structure characteristics of cold plate and weld points, cold plate samples covering different defect states are selected for pre-experiment, and the local heat defect correlation threshold is obtained;

[0115] It can be understood that the physical meaning of the local heat defect correlation value is that the local heat defect correlation value is calculated by summing the coincidence coefficient and the synchronization coefficient, wherein the coincidence coefficient reflects the coincidence degree of the local overheating sub-region and the suspected weld defect region in the position and quantity dimensions, the higher the coincidence degree is, the higher the possibility of the correlation between the weld defect (pore diameter) and the local overheating of the sub-region, and the synchronization coefficient reflects the synchronization correlation between the temperature in the sub-region and the weld defect degree on the basis of the coincidence of the local overheating sub-region and the suspected weld defect region, for example, the higher the weld defect degree is, the more serious the degree or phenomenon of local overheating in the sub-region is, therefore, the higher the synchronization correlation is, the higher the possibility of the correlation between the weld defect (pore diameter) and the local overheating of the sub-region is;

[0116] For example, the process of calculating the Pearson correlation coefficient between the co-occurrence temperature sequence and the co-occurrence pore sequence can be specifically:

[0117] The co-occurrence sub-regions are numbered as B, B= (1, 2, 3, …, i), wherein i represents the serial number of the heat dissipation cold plate sub-region;

[0118] The temperatures in the co-occurrence sub-regions are obtained in sequence according to the numbering order, and are integrated into a co-occurrence temperature sequence according to the numbering order, and the pore diameters in the co-occurrence sub-regions are obtained in sequence according to the numbering order, and are integrated into a co-occurrence pore sequence according to the numbering order;

[0119] The Pearson correlation coefficient is calculated, and the specific calculation formula is: and the absolute value is taken to obtain the synchronization coefficient, wherein the temperature mean value, the pore diameter mean value, n represents the number of co-occurrence sub-regions, represents the temperature value of the i-th co-occurrence sub-region under the overload power, and di represents the weld pore diameter in the i-th co-occurrence sub-region;

[0120] It also needs to be explained that the purpose of judging the existence of relevance and synchronization is to accurately verify the problem root, provide basis for subsequent quality control, and at the same time improve the accuracy and efficiency of the fault diagnosis of the heat dissipation cold plate. In addition, the judgment is also the premise of subsequent accurate repair and data verification. Only by confirming the existence of relevance and synchronization can the data basis for establishing an associated impact model be provided. This judgment can provide direction for the production end to trace the problem (such as improper brazing temperature and insufficient brazing purity leading to porosity), and can provide basis for predictive maintenance (early warning of overheating risk caused by welding defects) for the operation and maintenance end, to ensure the reliability of the heat dissipation cold plate under normal and overload working conditions.

[0121] Step four: based on the existence of relevance, the historical detection data is analyzed to obtain a local temperature-welding defect correlation model. Based on the correlation model, it is judged whether there is a high error in the current heat dissipation cold plate product porosity diameter detection result. If there is, it is corrected and then classified according to the heat dissipation cold plate product porosity diameter after correction.

[0122] In step four, the way to obtain the local temperature-welding defect correlation model is:

[0123] Based on the co-occurrence temperature sequence and the co-occurrence porosity sequence obtained from the co-occurrence sub-region, the least squares method is used to fit the co-occurrence temperature sequence and the co-occurrence porosity sequence to obtain the local temperature-welding defect correlation model.

[0124] In step four, the process of judging whether there is a high error in the current heat dissipation cold plate product porosity diameter detection result is:

[0125] Based on the suspected welding defect sub-region, the temperature value in the suspected welding defect sub-region is input into the local temperature-welding defect correlation model, and the actual diameter of the porosity in the suspected welding defect sub-region is output and compared with the detected diameter of the porosity in the suspected welding defect sub-region (detected by the micro-CT technology in step two):

[0126] If the error between the actual diameter of the porosity and the detected diameter of the porosity exceeds the allowable error range, the suspected welding defect sub-region is marked as a high error sub-region.

[0127] If the error between the actual diameter of the porosity and the detected diameter of the porosity does not exceed the allowable error range, the suspected welding defect sub-region is marked as a low error sub-region.

[0128] The proportion of the number of high error sub-regions in the suspected welding defect sub-regions contained in the current heat dissipation cold plate product is calculated to obtain the high error proportion value.

[0129] The error out-of-range value of the high-error sub-region is calculated by absolute difference proportion calculation of the pore diameter error of the high-error sub-region and the nearest end value of the allowable error range, and the error out-of-range value of the high-error sub-region is obtained;

[0130] The pore diameter error is the error between the actual pore diameter and the detected pore diameter.

[0131] For example, the absolute difference proportion calculation is as follows:

[0132] For example, the pore diameter error is 0.3 mm, and the allowable error range is (-0.4 mm, 0.4 mm). The absolute difference proportion calculation is as follows: | (0.3-0.4) / 0.4|;

[0133] The error out-of-range values of all high-error sub-regions are averaged to obtain a high-error degree value.

[0134] The high-error proportion value and the high-error degree value are summed to obtain a high-error performance value.

[0135] In some embodiments, the high-error performance value is compared with a high-error performance threshold value.

[0136] If the high-error performance value is greater than or equal to the high-error performance threshold value, it indicates that the current heat dissipation cold plate product has a high-error pore diameter detection result.

[0137] If the high-error performance value is less than the high-error performance threshold value, it indicates that the current heat dissipation cold plate product does not have a high-error pore diameter detection result.

[0138] It should be noted that the high-error performance threshold value is determined in combination with the production precision requirement and the detection standard of the heat dissipation cold plate. For example, according to the distribution of the high-error performance values of the qualified products and the unqualified products in the historical detection data, a value that can effectively distinguish the two is selected as the threshold value.

[0139] In step four, the pore detection diameter is corrected according to the pore diameter error.

[0140] In step four, the process of classifying the heat dissipation cold plate defects according to the pore diameter of the heat dissipation cold plate product is as follows:

[0141] If there is a pore diameter greater than or equal to 0.5 mm in the internal region of the heat dissipation cold plate product, the heat dissipation cold plate product is marked as a severely defective heat dissipation cold plate.

[0142] If there is no pore diameter greater than or equal to 0.5 mm in the internal region of the heat dissipation cold plate product, the heat dissipation cold plate product is marked as a slightly defective heat dissipation cold plate.

[0143] For example, the slightly defective heat dissipation cold plate and the slightly defective heat dissipation cold plate are repaired.

[0144] For the slightly defective area, local polishing and re-soldering can be used to repair, to eliminate pores and improve the quality of the solder joint;

[0145] For the severely defective area, the polishing range needs to be expanded, and after thoroughly removing the defective part, the solder is refilled and soldered to ensure the strength of the solder joint, and if it cannot be eliminated, the heat dissipation cold plate material in this area needs to be replaced and the soldering forming process is re-performed to ensure the overall heat dissipation performance and reliability;

[0146] It can be understood that the purpose of obtaining the temperature and pore diameter average correlation influence model is to establish a quantitative correlation model between temperature and internal solder joint pore diameter average through deep mining of historical data, which can not only accurately identify the possible system error or human operation deviation in the detection process, ensure the authenticity and reliability of the detection data, but also provide scientific basis for the quality control of the heat dissipation cold plate, for example, in actual production, if it is found that the detection data and the model predicted value exist significant deviation, the production link can be traced back immediately to check whether it is caused by improper soldering temperature control, insufficient solder purity or process parameter setting error, etc., so as to adjust the production parameters in time, avoid the occurrence of batch quality problems, at the same time, the model can also provide data support for the design optimization of the heat dissipation cold plate, through analyzing the change rule of the pore diameter under different temperatures, guiding the designer to reasonably adjust the heat dissipation structure and optimize the soldering process, so as to reduce the thermal stress concentration and improve the stability and life of the heat dissipation cold plate under long-term high temperature operation, in addition, the establishment of the correlation influence model is also helpful to realize the predictive maintenance of the heat dissipation cold plate quality, through real-time monitoring of the temperature and pore diameter change in operation, the potential quality risk is warned in advance, scientific basis is provided for maintenance decision, non-planned downtime risk is reduced, and the overall operation efficiency is improved;

[0147] The working principle of the present application is:

[0148] By carrying out temperature detection under rated power and overload power through multi-power testing on the heat dissipation cold plate, the non-overlapping area is marked as a local overheating area through area coincidence comparison, the internal welding point pore diameter of the heat dissipation cold plate product is detected through micro-CT technology, and the area where the pore diameter is greater than the pore diameter threshold is marked as a suspected welding point defect area, coincidence analysis is carried out based on the local overheating area and the suspected welding point defect area, and a co-occurrence sub-area is obtained, synchronicity analysis is carried out through the Pearson correlation coefficient, whether there is a higher possibility of correlation is judged, if there is correlation, a local temperature-welding point defect correlation model is obtained, whether the detection result has a high error is judged, the pore detection diameter is corrected according to the pore diameter error, and the heat dissipation cold plate defect is classified according to the pore diameter, which is beneficial to improve the production quality and use reliability of the heat dissipation cold plate, reduce the equipment failure rate caused by heat dissipation problems, accurately identify and repair the local overheating and welding point defect problems, effectively prolong the service life of the heat dissipation cold plate, and meanwhile, the predictive maintenance strategy based on the correlation model can discover potential quality risks in advance, and provide sufficient preparation time for operation and maintenance personnel;

[0149] The above shows and describes the basic principles, main features and advantages of the present application. Those skilled in the art should understand that the present application is not limited to the above-mentioned embodiments, and the above-mentioned embodiments and descriptions in the specification are only to illustrate the principles of the present application. Without departing from the spirit and scope of the present application, various changes and improvements can be made to the present application, and these changes and improvements all fall within the scope of the present application. The scope of protection of the present application is defined by the appended claims and their equivalents.

Claims

1. A brazing forming process for a lightweight, high-reliability heat dissipation cold plate, characterized in that: Includes the following steps: During the internal quality inspection of the finished heat dissipation plate, multi-power testing is conducted to obtain the surface temperature of the finished heat dissipation plate under different power levels, and to screen for localized overheating areas. Micro-CT technology was used to detect the diameter of pores in the solder joints inside the finished heat dissipation plate to identify areas with suspected solder joint defects. Overlap analysis was performed on the local overheated area and the suspected solder joint defect area to obtain the co-occurrence sub-region. The temperature and pore diameter in the co-occurrence sub-region under overload power were analyzed to determine whether there is a correlation between local overheating and suspected solder joint defects. Based on the correlation, historical detection data is analyzed to obtain a local temperature-solder joint defect correlation model. Based on the correlation model, it is determined whether the current detection result of the pore diameter of the finished heat dissipation plate has a high error. If so, it is corrected and the heat dissipation plate defect is classified according to the pore diameter of the finished heat dissipation plate after correction.

2. The brazing forming process of the lightweight, high-reliability heat dissipation cold plate according to claim 1, characterized in that: The process for obtaining the surface temperature of the finished heat dissipation plate under different power conditions is as follows: Connect to the rated power and use an infrared thermal imager to take full-area coverage pictures of the heat dissipation plate, record the temperature distribution on the surface of the heat dissipation plate under the rated power, and obtain the rated infrared distribution map; The rated infrared distribution map is divided into several sub-regions of equal area to obtain the rated sub-regions. A temperature threshold is set, and the rated sub-regions whose temperature exceeds the temperature threshold are marked as abnormal rated sub-regions. By applying overload power, an infrared thermal imager is used to capture full-area images of the heat dissipation plate, recording the temperature distribution on the surface of the heat dissipation plate under overload power, and obtaining an overload infrared distribution map. By comparing the temperature changes of each overload sub-region with the corresponding rated sub-region, the temperature rise value is obtained; If the temperature rise of the overload sub-region is greater than or equal to the temperature rise threshold, the corresponding overload sub-region will be marked as an abnormal overload sub-region.

3. The brazing forming process of the lightweight, high-reliability heat dissipation cold plate according to claim 1, characterized in that: The process for screening locally overheated areas is as follows: The abnormal overload sub-region is compared with the abnormal rated sub-region, and the non-overlapping part of the abnormal overload sub-region and the abnormal rated sub-region is marked as a local overheating region.

4. The brazing forming process of the lightweight, high-reliability heat dissipation plate according to claim 1, characterized in that: The process for identifying suspected solder joint defect areas is as follows: The heat sink plate is scanned from all directions by emitting X-rays using a micro-CT device. After the X-rays penetrate the heat sink plate, they are received by the receiver of the micro-CT device to obtain three-dimensional signal data of the heat sink plate and generate a three-dimensional image of the internal structure of the heat sink plate. In the 3D image, image analysis software is used to identify and locate the solder joint area, obtain the pore diameter, and compare it with the pore diameter threshold. If the pore diameter is greater than or equal to the pore diameter threshold, the corresponding sub-region is marked as a suspected solder joint defect region.

5. The brazing forming process of the lightweight, high-reliability heat dissipation plate according to claim 1, characterized in that: The process of performing overlap analysis on local overheated areas and suspected solder joint defect areas to obtain co-occurring sub-regions is as follows: Sub-regions that are both localized overheating areas and suspected solder joint defects are marked as co-occurrence sub-regions; The total number of sub-regions with co-occurrence, the total number of sub-regions with heat dissipation and cold plate, the total number of sub-regions with local overheating, and the total number of sub-regions with suspected solder joint defects are counted. The difference between the total number of local overheated sub-regions and the total number of suspected solder joint defect sub-regions is calculated to obtain the number of deviation regions. The ratio of the number of deviation regions to the total number of heat dissipation cold plate sub-regions is then calculated to obtain the region difference value. The ratio of the number of co-occurring sub-regions to the total number of locally overheated sub-regions is used to obtain the local overheating ratio. The overlap coefficient is obtained by performing a difference analysis on the local heat ratio and the regional difference ratio.

6. The brazing forming process of the lightweight, high-reliability heat dissipation plate according to claim 1, characterized in that: The process of synchronously analyzing the temperature and pore diameter within the same sub-region under overload power is as follows: Based on the co-occurrence sub-region, the temperature and pore diameter of each co-occurrence sub-region under overload power are obtained and integrated to obtain the co-occurrence temperature sequence and the co-occurrence pore sequence. Among them, the positions of the co-occurring sub-regions in the co-occurring temperature sequence and the co-occurring stoma sequence are the same; Calculate the Pearson correlation coefficient between the co-occurring temperature sequence and the co-occurring stomatal sequence, and take the absolute value to obtain the synchronization coefficient with respect to temperature and stomatal diameter.

7. The brazing forming process of the lightweight, high-reliability heat dissipation cold plate according to claim 1, characterized in that: The process for determining whether there is a correlation between localized overheating and suspected solder joint defects is as follows: The overlap coefficient and synchronization coefficient are summed to obtain the correlation value of local thermal defects; If the correlation value of localized thermal defects is greater than or equal to the correlation threshold of localized thermal defects, it indicates that there is a correlation between localized overheating and solder joint defects.

8. The brazing forming process of the lightweight, high-reliability heat dissipation cold plate according to claim 1, characterized in that: The analysis of historical detection data yields a local temperature-weld joint defect correlation model, as follows: Based on the co-occurrence temperature sequence and co-occurrence porosity sequence obtained from the co-occurrence sub-region, the least squares method is used to fit the co-occurrence temperature sequence and co-occurrence porosity sequence to obtain a local temperature-solder joint defect correlation model.

9. The brazing forming process of the lightweight, high-reliability heat dissipation plate according to claim 1, characterized in that: The process for determining whether there is a high error in the detection result of the pore diameter of the current finished heat dissipation plate is as follows: Based on the suspected solder joint defect sub-region, the temperature value within the suspected solder joint defect sub-region is input into the local temperature-solder joint defect correlation model, and the actual diameter of the pores in the suspected solder joint defect sub-region is output and compared with the detected diameter of the pores in the suspected solder joint defect sub-region. If the error between the actual diameter of the pore and the detected diameter of the pore exceeds the allowable error range, the suspected weld defect sub-region will be marked as a high error sub-region. The proportion of high-error sub-regions in the current finished heat dissipation cold plate is obtained by statistically analyzing the number of sub-regions with suspected solder joint defects. The error overshoot value of the high error sub-region is obtained by calculating the ratio of the absolute difference between the pore diameter error of the high error sub-region and the nearest allowable error range endpoint value. The error out-of-bounds values ​​of all high-error sub-regions are averaged to obtain the high-error severity value; The high error percentage value is summed with the high error severity value to obtain the high error performance value; If the high error performance value is greater than or equal to the high error performance threshold, it indicates that the current pore diameter detection result of the heat dissipation plate product has a high error.

10. The brazing forming process of the lightweight, high-reliability heat dissipation cold plate according to claim 1, characterized in that: The process of classifying defects in the heat dissipation cold plate based on the diameter of the pores in the finished product is as follows: If there are pores with a diameter of 0.5 mm or more in the internal area of ​​the finished heat dissipation plate, the finished heat dissipation plate shall be marked as a severely defective heat dissipation plate. If there are no pores with a diameter greater than or equal to 0.5 mm in the internal area of ​​the finished heat dissipation plate, the finished heat dissipation plate will be marked as a slightly defective heat dissipation plate.

Images