Method for improving conductivity of aluminum bar with mark of 6063

Abstract

The invention relates to the technical field of aluminum processing, and discloses a method for improving the conductivity of an aluminum bar with a mark of 6063, which comprises the following steps: screening raw materials, screening an aluminum bar with no damage on the surface, pretreating the surface, detecting defects, monitoring whether bad defects exist in the screened aluminum bar or not, and carrying out gradient heating. According to a set program, the temperature gradient in the annealing furnace is kept to rise, constant-temperature treatment, segmented cooling and performance detection are carried out, and the final conductivity change of the aluminum bar is monitored and compared with the initial conductivity parameter. According to the method, the 6063 aluminum bar with the inventory conductivity not reaching the standard is subjected to annealing treatment of heat preservation at 420 DEG C for 1 hour and furnace cooling, the conductivity of the aluminum bar can be improved by 0.81-1.33 MS / m and is improved by 2.51%-4.14%, the aluminum bar originally not reaching the standard meets the design requirement, inventory materials are effectively utilized, the production requirement of aluminum flexible connection is met, the utilization rate of the inventory 6063 aluminum bar is improved, and the production cost is reduced. And the process is simple, low in cost and high in practicability.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of aluminum processing, in particular to a method for improving the conductivity of aluminum bar of 6063 brand. BACKGROUND

[0002] Aluminum flexible connection is a kind of connecting piece widely used in industrial field. The aluminum flexible connection has good flexible conductive effect, good conductive performance, small resistance value, and can bear large current. The aluminum flexible connection can be used in various high-voltage electrical appliances, vacuum electrical appliances, mining explosion-proof, switches and related products of automobiles and locomotives, and has wide application in metallurgical and chemical industry, wheel power transformation engineering and power equipment. The aluminum flexible connection can also be used in transformer installation, high-low voltage switch cabinet, vacuum electrical appliance, closed busbar, generator and busbar, rectifier equipment, connection between rectifier cabinet and disconnecting switch and connection between busbars, etc., which can improve the conductivity, adjust the installation error of equipment, simultaneously play the role of shock absorption and work compensation, facilitate test and equipment maintenance, etc. The aluminum flexible connection is widely used in high-end electric vehicle batteries, can replace copper foil flexible connection for large battery pack and battery module connection. The aluminum flexible connection can be used in boiler and industrial furnace flue gas desulfurization device, ventilation duct of wet and dry method in petrochemical enterprises.

[0003] The aluminum bar is a commonly used profile in aluminum flexible connection. The aluminum bar of 6063 brand has more applications. The 6063 aluminum alloy is a commonly used deformed aluminum alloy material in high-voltage electrical conductor parts. Such parts not only require good conductivity, but also usually require certain hardness and strength. However, with the increase of voltage level, the size of the parts gradually increases, and the hardness or conductivity performance of large-size aluminum alloy parts is unqualified. Sometimes the design drawing has requirements on the conductivity of the aluminum bar. However, in the case that the conductivity of the existing aluminum bar does not meet the requirements of the design drawing, it is often discarded, thereby significantly increasing the scrap rate of the 6063 aluminum bar. In view of this, a method for improving the conductivity of the aluminum bar of 6063 brand is proposed. SUMMARY

[0004] In view of the deficiencies of the prior art, the present application provides a method for improving the conductivity of the aluminum bar of 6063 brand, which solves the problem that the existing aluminum bar is often discarded when the conductivity does not meet the requirements of the design drawing, thereby significantly increasing the scrap rate of the 6063 aluminum bar.

[0005] To achieve the above purpose, the present application is implemented by the following technical scheme: a method for improving the conductivity of the aluminum bar of 6063 brand, comprising the following steps:

[0006] S1: raw material selection

[0007] Select the chemical composition of the stock in line with Si: 0.20% ~ 0.6%, Mg: 0.45% ~ 0.9% of 6063 aluminum, aluminum thickness of 6 mm, cross section size tolerance control within ± 0.1 mm, surface without length more than 5 mm and depth more than 0.3 mm of the scratch, no more than 1 ° of the plane warping deformation;

[0008] S2: surface pretreatment

[0009] The purity of ≥99.5% of alcohol is used to soak the dust-free wiping cloth, and then the surface of the aluminum bar selected in S1 is reciprocally wiped, each surface is wiped for not less than 3 times, the length of single wiping stroke is ≥200 mm, the wiping direction is consistent with the length direction of the aluminum bar, and the overlapping area of adjacent two wiping strokes is ≥50 mm;

[0010] S3: defect detection

[0011] The surface of the aluminum bar after S2 wiping is detected by using eddy current flaw detector, the detection frequency is set to 15 kHz, the scanning speed is controlled at 30 ~ 40 mm / s, the distance between the flaw detector probe and the surface of the aluminum bar is kept at 0.5 ~ 1.0 mm, and the eddy current signal waveform data is collected and stored synchronously during the detection process;

[0012] S4: gradient heating

[0013] The aluminum bar without defects after detection in S3 is placed horizontally on the ceramic fiber support in the annealing furnace, and the temperature is raised from room temperature to 200℃ at a rate of 5℃ / min, pure nitrogen with purity of ≥99.99% is introduced into the furnace during the heating process, the nitrogen is introduced through 6 evenly distributed air inlets at the bottom of the furnace body, and the flow rate is controlled at 8L / min; after 30 min at 200℃, the temperature is raised to 420℃ at a rate of 3℃ / min, and the nitrogen flow rate is adjusted to 10L / min during this stage;

[0014] S5: constant temperature treatment

[0015] After the gradient heating in S3, when the furnace temperature reaches 420℃, the temperature holding stage is entered, the holding time is 1 hour, 2 variable frequency axial fans are arranged at the top and bottom of the furnace body respectively, the fan diameter is 200 mm, the wind speed is adjusted to 2.5 m / s through the frequency converter, and the fan speed is adjusted in real time through the PLC control system according to the temperature data feedback by the 5 platinum resistance temperature sensors arranged uniformly in the furnace;

[0016] S6: segmented cooling

[0017] After the constant temperature treatment time in S5 is met, the furnace is cooled to 200℃ at a rate of 3℃ / min, and the nitrogen flow rate is maintained at 5L / min during the cooling process; when the furnace temperature drops to 200℃, the forced air cooling system is started, the air cooling device is composed of 4 axial flow fans and a serpentine cooling coil, 18℃ cooling water is introduced into the coil, the cooling rate is controlled at 6℃ / min, until the aluminum bar temperature drops to 25±2℃;

[0018] S7: Performance detection

[0019] The conductivity tester with precision level 0.5 and automatic temperature compensation function is used to detect the two ends and the middle of the aluminum bar cooled in S6 at 15℃±0.5℃ environment.

[0020] Preferably, in the S2 surface pretreatment step, ultrasonic cleaning assisted alcohol wiping is used: the aluminum bar is placed in an alcohol solution at a temperature of 40℃, the ultrasonic cleaning machine frequency is set to 40kHz, the power density is 0.3W / cm 2 , the cleaning time is 8min, and the solution needs to be kept circulating during the cleaning process.

[0021] Preferably, after the S2 surface pretreatment step, a convolutional neural network containing 2 convolutional layers, 2 pooling layers and 1 fully connected layer with a convolution kernel size of 3×3 is used to analyze 5000 groups of eddy current flaw detection signals with labels collected by the eddy current flaw detector, and automatically identify cracks with a length of ≥0.5mm or inclusion defects with a diameter of ≥0.3mm.

[0022] Preferably, in the S4 gradient heating step, the Si content, Mg content and initial hardness value of the aluminum bar are input into a machine learning model to calculate the optimal heating rate of 5-8℃ / min, the model is a three-layer neural network containing 3 input neurons, 10 hidden neurons and 1 output neuron, and uses ReLU activation function.

[0023] Preferably, the annealing furnace used in the S4 gradient heating step adopts a double-layer vacuum insulation structure: the inner layer is a 5mm thick 310S stainless steel plate with a high temperature resistance of ≥1200℃, the outer layer is a 50mm thick nanometer aerogel insulation layer with a thermal conductivity of ≤0.02W / (m·K), and the pressure between the two layers is ≤1Pa, the furnace door sealing uses a silicon rubber sealing ring combined with an inflation structure.

[0024] Preferably, in the 200℃ holding stage of the S4 gradient heating step, the nitrogen pressure in the furnace is maintained at 1.3atm by a pressure regulating valve, the furnace body is provided with a pressure sensor with an accuracy of ±0.05atm, and the pressure change is monitored in real time and the intake valve opening degree is controlled.

[0025] Preferably, in the S5 constant temperature treatment step, an infrared temperature monitoring device is arranged in the furnace, with a measurement range of 0-600 DEG C and an accuracy of plus or minus 1 DEG C, and the surface temperature of the aluminum row is monitored in real time by four infrared probes distributed around the furnace body, and the monitoring data is transmitted to the control system for cross verification synchronously with the platinum resistance sensor data.

[0026] Preferably, in the S6 segmented cooling step, the axial flow fan of the forced air cooling system is arranged in an up-down staggered manner with the cooling coil, the fan outlet is 300 mm away from the surface of the aluminum row, and the cooling water flow in the coil is controlled at 5 m 3 / h by an electromagnetic flowmeter with an accuracy of plus or minus 1%.

[0027] Preferably, in the S7 performance detection step, the conductivity tester uses a standard conductivity test block before testing, the test block is an aluminum block with a purity of greater than or equal to 99.99%, the known value of the conductivity is 35.0 MS / m, three-point calibration is performed, the range is 0%, 50% and 100% respectively, and the calibration deviation is less than or equal to plus or minus 0.3% FS.

[0028] Preferably, in the S7 performance detection step, the contact pressure between the electrode and the surface of the aluminum row is 0.5 MPa, each test point is measured repeatedly for 3 times, the arithmetic mean value is taken as the detection result of the point, and finally the average value of the three test points is taken as the aluminum row conductivity data.

[0029] The application provides a method for improving the conductivity of a 6063 aluminum row.

[0030] 1. The application can improve the conductivity of the 6063 aluminum row by 0.81-1.33 MS / m through annealing treatment of the 6063 aluminum row with a stock conductivity that does not meet the standard at 420 DEG C for 1 hour and furnace cooling, thereby improving the conductivity by 2.51%-4.14%, meeting the design requirements of the aluminum row that does not meet the standard, effectively utilizing the stock material, meeting the production needs of the aluminum soft connection, improving the utilization rate of the stock 6063 aluminum row, and being simple in process, low in cost and high in practicability.

[0031] 2. The application can identify cracks greater than or equal to 0.5 mm and inclusions greater than or equal to 0.3 mm in the 6063 aluminum row eddy current signal through wavelet transform noise reduction, CNN network construction and training, Adam optimization and other steps, with an accuracy of more than 97.5%, a positioning error of plus or minus 0.2 mm, a detection time of less than 40 seconds, an automatic screening of defective aluminum rows, an optimization of the annealing process, an improvement of the conductivity qualified rate to 96%, an annual cost saving of 1.25 million yuan, and high precision, high efficiency and high economy.

[0032] 3, The application constructs a 3-10-1 neural network model by learning 1000 groups of historical data to dynamically calculate the 6063 aluminum bar gradient temperature rising rate, so that the conductivity standard deviation is reduced by 46%, the inclusion and crack defects are reduced by 37% and 62% respectively, the energy consumption is reduced by 12%, the single batch processing time is shortened by 18%, the high Si inventory value of 3.5 million yuan is activated, the R&D cycle is compressed by 50%, the number of operating personnel is reduced by 58% after embedding the automatic system, the accident rate is reduced by 83%, the quality, efficiency and cost are triple optimized, and the intelligent upgrading of aluminum processing is promoted. BRIEF DESCRIPTION OF DRAWINGS

[0033] Figure 1 The method for improving the conductivity of the 6063 aluminum bar is a flowchart;

[0034] Figure 2 The surface defect recognition algorithm flowchart of the application is a flowchart;

[0035] Figure 3 The annealing furnace internal temperature control algorithm flowchart of the application is a flowchart. DETAILED DESCRIPTION

[0036] The technical solutions in the embodiments of the application will be described clearly and completely in combination with the drawings in the specification of the application. Obviously, the described embodiments are only part of the embodiments of the application, not all the embodiments. Based on the embodiments in the application, all other embodiments obtained by those skilled in the art without creative labor are within the protection scope of the application.

[0037] Embodiment:

[0038] Please refer to the drawings in the specification of the application Figure 1 - the drawings in the specification of the application Figure 3 The embodiment of the application provides a method for improving the conductivity of a 6063 aluminum bar, which comprises the following steps:

[0039] S1: raw material screening

[0040] Select 6063 aluminum bars in the inventory with chemical components of Si: 0.20% to 0.6% and Mg: 0.45% to 0.9%, the thickness of the aluminum bars is 6 mm, the cross-sectional size tolerance is controlled within ±0.1 mm, the surface is free of scratches with a length exceeding 5 mm and a depth exceeding 0.3 mm, and the surface is free of plane warping deformation exceeding 1°;

[0041] S2: surface pretreatment

[0042] The surfaces of the aluminum bars screened out in S1 are reciprocally wiped after the dust-free wiping cloth is infiltrated with alcohol with a purity of ≥99.5%, each surface is wiped for no less than 3 times, the length of each wiping stroke is ≥200 mm, the wiping direction is consistent with the length direction of the aluminum bar, and the overlapping area of adjacent two wiping strokes is ≥50 mm;

[0043] S3: Defect detection

[0044] The surfaces of the aluminum bars wiped in S2 are detected in a circumferential full coverage manner using an eddy current detector, the detection frequency is set to 15 kHz, the scanning speed is controlled to be 30-40 mm / s, the distance between the probe of the detector and the surface of the aluminum bar is kept to be 0.5-1.0 mm, and the eddy current signal waveform data are synchronously collected and stored during the detection process;

[0045] S4: Gradient heating

[0046] The aluminum bars without defects detected in S3 are horizontally placed on the ceramic fiber support in the annealing furnace, and are heated from room temperature to 200℃ at a rate of 5℃ / min, pure nitrogen with a purity of ≥99.99% is introduced into the furnace during the heating process, the nitrogen is introduced through 6 air inlets uniformly distributed at the bottom of the furnace body, and the flow rate is controlled to be 8 L / min; after being kept at 200℃ for 30 min, the temperature is increased to 420℃ at a rate of 3℃ / min, and the flow rate of the nitrogen is adjusted to 10 L / min during this stage;

[0047] S5: Constant temperature treatment

[0048] After the gradient heating in S3, the furnace temperature reaches 420℃, and then the constant temperature treatment is carried out for 1 hour, 2 variable frequency axial fans with a diameter of 200 mm are arranged at the top and the bottom of the furnace body respectively, the wind speed is adjusted to 2.5 m / s through the frequency converter, and the rotating speed of the fans is adjusted in real time through the PLC control system according to the temperature data fed back by the 5 platinum resistance temperature sensors uniformly arranged in the furnace;

[0049] S6: Stepwise cooling

[0050] After the constant temperature treatment in S5 is completed, the furnace is cooled to 200℃ at a rate of 3℃ / min, the flow rate of the nitrogen is maintained at 5 L / min during the cooling process; when the furnace temperature drops to 200℃, the forced air cooling system is started, the air cooling device is composed of 4 axial fans and a serpentine cooling coil, 18℃ cooling water is introduced into the coil, the cooling rate is controlled to be 6℃ / min, and the temperature of the aluminum bar is reduced to 25±2℃;

[0051] S7: Performance detection

[0052] The conductivity tester with a precision level of 0.5 and an automatic temperature compensation function is used to detect the two ends and the middle of the aluminum bar cooled in S6.

[0053] The S2 surface pretreatment step adopts ultrasonic cleaning assisted alcohol wiping: the aluminum row is placed in an alcohol solution at a temperature of 40°C, the frequency of the ultrasonic cleaning machine is set to 40 kHz, the power density is 0.3 W / cm 2 , the cleaning time is 8 min, and the solution needs to be kept circulating during the cleaning process.

[0054] After the S2 surface pretreatment step, a convolutional neural network containing 2 convolutional layers, 2 pooling layers and 1 fully connected layer is used to analyze 5000 groups of signals collected by the eddy current flaw detector with labels, automatically identifying cracks with a length of ≥0.5 mm or inclusions with a diameter of ≥0.3 mm. The following defect recognition algorithm is established:

[0055] I. Signal preprocessing stage

[0056] 1. Wavelet transform denoising

[0057] Discrete wavelet transform is performed on the original signal collected by the eddy current flaw detector to decompose the signal into different frequency components. Taking three-layer Daubechies wavelet (db4) decomposition as an example, the formula is:

[0058]

[0059] Where: the detail coefficient d j,k corresponds to high-frequency noise, and the approximation coefficient c 3,k corresponds to the main part of the signal. The detail coefficient is processed using a soft threshold function:

[0060]

[0061] Threshold σ is the noise standard deviation, and N is the signal length. The reconstructed signal after processing eliminates interference

[0062] II. CNN network construction and training

[0063] 2. First layer convolution operation

[0064] A 3x3 convolution kernel is used to extract features from the denoised signal, and the calculation formula of the output feature map Y1 is:

[0065]

[0066] Where X is the input signal matrix, K1 is the first layer convolution kernel, and b1 is the bias. The activation function uses ReLU:

[0067] ReLU(x) = max(0, x)

[0068] This layer is used to extract local mutation features in the signal (such as eddy current anomalies caused by cracks).

[0069] 3. First layer pooling operation

[0070] 2x2 max-pooling is performed on the convolution output, with the formula:

[0071]

[0072] After pooling, the feature map size is halved, retaining peak features while reducing dimensions, such as high-frequency peaks of crack signals that are preferentially retained.

[0073] 4. Second layer convolution and pooling

[0074] The second layer convolution kernel K2 is also 3x3, with a similar operation formula as the first layer:

[0075]

[0076] After ReLU activation again, 2x2 max-pooling is performed:

[0077]

[0078] This layer is used to extract more abstract defect features, such as eddy current energy distribution anomalies caused by inclusions

[0079] 5. Fully connected layer classification

[0080] The pooled feature map is unfolded into a one-dimensional vector, and classified by a fully connected layer: Y5 = σ(W3 · Y4 + b3)

[0081] where W3 is the weight matrix, b3 is the bias, and σ is the Softmax function:

[0082] Output 3 class probability values (no defect / crack / inclusion), and determine the defect type by threshold 0.5.

[0083] III. Network training optimization

[0084] 6. Loss function calculation

[0085] The cross-entropy loss function is used to measure the difference between the prediction and the true label, with the formula:

[0086] where N = 5000 is the sample size, y i,c is the true label (0 or 1) of the cth class of the ith sample, is the predicted probability.

[0087] 7. Adam optimization algorithm

[0088] When updating the network parameters, the first and second moments are calculated: t = 0.9m t-1 + 0.1g t ,

[0089] After correcting the bias, update the weights:

[0090] where g t is the gradient,

[0091] Four, defect positioning and classification

[0092] 8. Physical position conversion

[0093] The defect position is deduced from the feature map coordinates, considering the step length accumulation of convolution and pooling:

[0094] Sampling point position = (feature map coordinates x 2 x 2) + convolution kernel offset

[0095] Convert to actual physical position:

[0096] (Assuming a scanning speed of 50 mm / s and a sampling frequency of 1000 Hz).

[0097] 9. Defect type discrimination

[0098] For the three-class probability values output by the fully connected layer, take the maximum value corresponding to the class:

[0099] When the crack probability is ≥0.5 and the signal gradient at the peak position of the corresponding feature map is ≥15 mV / μs, it is determined to be a crack with a length ≥0.5 mm; when the inclusion probability is ≥0.5 and the signal energy spectrum accounts for 230% in the 10-20 kHz frequency band, it is determined to be an inclusion with a diameter ≥0.3 mm.

[0100] Five, model evaluation and application

[0101] 10. Accuracy verification

[0102] Using 1000 groups of test set in 5000 groups of labeled data for verification, the defect recognition accuracy calculation formula is:

[0103]

[0104] When the recognition accuracy of the model for cracks and inclusions reaches 98.2% and 97.5% respectively, it is put into practical application to screen the surface defects of 6063 aluminum bars before annealing treatment.

[0105] In the S4 gradient heating step, the optimal heating rate of 5-8 ℃ / min is calculated by inputting the Si content, Mg content and initial hardness value of the aluminum bar into the machine learning model, the model is a three-layer neural network containing 3 input neurons, 10 hidden neurons and 1 output neuron, using ReLU activation function, here the following temperature control algorithm is established:

[0106] Machine learning model implementation step of S4 gradient heating step

[0107] 1. Data preprocessing

[0108] Collect 1000 sets of historical production data, each set containing Si content x1, Mg content x2, initial hardness value x3 of the aluminum bar and the corresponding optimal heating rate t.

[0109] Z-score standardization is performed on the input features:

[0110] Where μ i and σ i are the mean and standard deviation of the i-th feature, for example, μ1=0.38% (mean of Si content), σ1=0.08%. The standardized input vector

[0111] 2. Model initialization

[0112] Construct a neural network with a structure of 3-10-1, randomly initialize the weight matrix W∈R 3×10 and Bias vector b∈R 10 and c∈R. For example, the elements w ij of W are sampled from the uniform distribution U(-0.1, 0.1).

[0113] 3. Forward propagation calculation

[0114] For the input sample X, calculate the hidden layer neuron output h j :

[0115] The output layer calculates the predicted heating rate

[0116] For example, when a set of standardized features X=[0.5,-0.3,1.2] is input, the calculation gives

[0117] 4. Loss function calculation

[0118] The difference between the predicted value and the true value t is evaluated using the mean squared error (MSE):

[0119] where N = 1000, and for example, L = 0.12 in an iteration, indicating that the average prediction error is about

[0120] 5. Backpropagation to compute gradients

[0121] Compute the output layer error term:

[0122] Hidden layer error term: δ j = δ y · v j · 1(z j > 0)

[0123] where 1(·) is the indicator function. Compute the gradients:

[0124] 6. Parameter update

[0125] Update the weights and biases using stochastic gradient descent (SGD) with a learning rate η = 0.01:

[0126] For example, v3 is updated from 0.21 to 0.21 - 0.01 x 0.05 = 0.2095 in an iteration

[0127] 7. Batch training and optimization

[0128] Divide the 1000 samples into 50 batches, each with 20 samples, and train for 500 rounds. Calculate the MSE on the validation set (200 samples) after each round, and stop early when the validation error changes by <0.001 for 10 consecutive rounds. For example, the validation error converges to 0.085 at the 320th round of training.

[0129] 8. Prediction and application

[0130] For a new aluminum sample, input the standardized features X new , and compute the optimal heating rate y through forward propagation, and limit the result to 5-8 ℃ / min:

[0131] For example, if the model output is then the actual heating rate y final = 8 ℃ / min.

[0132] 9. Model evaluation

[0133] The mean absolute error (MAE) is calculated on the test set (300 samples):

[0134] The final model MAE is about 0.25℃ / min, which meets the industrial precision requirements.

[0135] The annealing furnace used in the S4 gradient temperature rising step adopts a double-layer vacuum insulation structure: the inner layer is a 5mm thick 310S stainless steel plate with a high temperature resistance of ≥1200℃, the outer layer is a 50mm thick nano aerogel insulation layer with a thermal conductivity of ≤0.02W / (m·K), and the pressure between the two layers is ≤1Pa, and the furnace door seal uses a silicon rubber sealing ring with an inflatable structure.

[0136] In the 200℃ holding stage of the S4 gradient temperature rising step, the nitrogen pressure in the furnace is maintained at 1.3 atmospheres by a pressure regulating valve, and the furnace body is provided with a pressure sensor with an accuracy of ±0.05 atmospheres to monitor the pressure change and control the intake valve opening degree in real time.

[0137] In the S5 constant temperature treatment step, an infrared temperature monitoring device is provided in the furnace, with a measurement range of 0-600℃ and an accuracy of ±1℃, which monitors the surface temperature of the aluminum bar in real time through four infrared probes distributed around the furnace body, and the monitoring data and platinum resistance sensor data are transmitted to the control system for cross verification.

[0138] In the S6 segmented cooling step, the axial flow fan of the forced air cooling system is arranged in an up-down staggered manner with the cooling coil, the fan outlet is 300mm away from the surface of the aluminum bar, and the cooling water flow in the coil is controlled at 5m 3 / h by an electromagnetic flowmeter with an accuracy of ±1%.

[0139] In the S7 performance detection step, the conductivity tester uses a standard conductivity test block before testing, the test block is an aluminum block with a purity of ≥99.99%, the known conductivity value is 35.0MS / m, and the three-point calibration is performed with a range of 0%, 50%, and 100%, and the calibration deviation is ≤±0.3%FS.

[0140] In the S7 performance detection, the contact pressure between the electrode and the surface of the aluminum bar is 0.5MPa, each test point is measured 3 times, the arithmetic mean value is taken as the detection result of the point, and the average value of the three test points is taken as the conductivity data of the aluminum bar. According to the above content, the corresponding 6063 aluminum bar performance test is established,

[0141] (I) Experimental process

[0142] 1. Clean the surface of 4 aluminum bars to remove surface oil, dust and other impurities.

[0143] 2. Without annealing, use a conductivity meter to test the conductivity of the four aluminum busbars.

[0144] 3. Place the four aluminum pieces into the annealing furnace for annealing treatment. Set the annealing temperature to 420℃ and the holding time to 1 hour. Then cool them in the furnace.

[0145] 4. After the annealing process is completed, the conductivity of the four aluminum busbars is tested again using a conductivity meter.

[0146] 5. Statistical and comparative analysis of the conductivity test results before and after annealing.

[0147]

[0148] (II) Data Processing and Analysis

[0149] Calculation of conductivity improvement: Subtract the conductivity before annealing from the conductivity after annealing to obtain the conductivity improvement value of each aluminum busbar.

[0150] #1 Aluminum busbar: 33.29 - 32.03 = 1.26 (MS·m) -1 )

[0151] #2 aluminum busbar: 33.36 - 32.13 = 1.23 (MS·m) -1 )

[0152] #3 aluminum busbar: 33.46 - 32.13 = 1.33 (MS·m) -1 )

[0153] #4 aluminum busbar: 33.05 - 32.24 = 0.81 (MS·m) -1 )

[0154] Calculation of the increase in conductivity: Divide the conductivity increase value by the conductivity before annealing, and then multiply by 100% to obtain the increase in conductivity of each aluminum busbar.

[0155] #1 Aluminum busbar: (1.26 ÷ 32.03) × 100% ≈ 3.93%

[0156] #2 aluminum busbar: (1.23 ÷ 32.13) × 100% ≈ 3.82%

[0157] #3 aluminum busbar: (1.33 ÷ 32.13) × 100% ≈ 4.14%

[0158] #4 aluminum busbar: (0.81 ÷ 32.24) × 100% ≈ 2.51%

[0159] (III) Experimental Results

[0160] From the experimental data, it can be seen that after annealing treatment, the electrical conductivity of the four 6063 aluminum bars is improved to different degrees. The conductivity improvement value is between 0.81MS·m -1 and 1.33MS·m -1 , and the improvement amplitude is between 2.51% and 4.14%. Among them, the conductivity improvement value and the improvement amplitude of the 3# aluminum bar are the largest, and the conductivity improvement value and the improvement amplitude of the 4# aluminum bar are relatively small

[0161] (Four) Experimental results

[0162] From the experimental data, it can be seen that after annealing treatment, the electrical conductivity of the four 6063 aluminum bars is improved to different degrees. The conductivity improvement value is between 0.81MS·m -1 and 1.33MS·m -1 , and the improvement amplitude is between 2.51% and 4.14%. Among them, the conductivity improvement value and the improvement amplitude of the 3# aluminum bar are the largest, and the conductivity improvement value and the improvement amplitude of the 4# aluminum bar are relatively small.

[0163] (Five) Discussion of results

[0164] The effect of annealing treatment on electrical conductivity: Annealing treatment can improve the electrical conductivity of 6063 aluminum bars, because during annealing, the grains inside the aluminum bar will recrystallize, the grain size increases, and the number of grain boundaries decreases, thereby reducing the scattering of electrons in the crystal and improving the electrical conductivity. At the same time, annealing treatment can also eliminate the residual stress inside the aluminum bar and reduce the lattice distortion, further improving the electrical conductivity.

[0165] (Six) Experimental standards

[0166]

[0167] Although embodiments of the present application have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made therein without departing from the principles and spirit of the application, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for improving the electrical conductivity of an aluminum alloy of the 6063 series, characterized in that, The method comprises the following steps: S1: raw material screening Selecting 6063 aluminum bars in stock with chemical composition of Si: 0.20% to 0.6% and Mg: 0.45% to 0.9%, the thickness of the aluminum bars is 6 mm, the cross-sectional size tolerance is controlled within ±0.1 mm, the surface is free of scratches with length exceeding 5 mm and depth exceeding 0.3 mm, and free of planar warping deformation exceeding 1°; S2: surface pretreatment After the surface of the aluminum bar selected in S1 is reciprocally wiped with a dust-free wiping cloth soaked with alcohol with purity of ≥99.5%, each surface is wiped for no less than 3 times, the length of each wiping stroke is ≥200 mm, the wiping direction is consistent with the length direction of the aluminum bar, and the overlapping area of adjacent two wiping strokes is ≥50 mm; S3: defect detection The surface of the aluminum bar after wiping in S2 is detected by using an eddy current flaw detector, the detection frequency is set to 15 kHz, the scanning speed is controlled to be 30-40 mm / s, the distance between the probe of the flaw detector and the surface of the aluminum bar is kept to be 0.5-1.0 mm, and the eddy current signal waveform data is synchronously collected and stored during the detection process; S4: gradient heating The aluminum bar without defects after detection in S3 is horizontally placed on a ceramic fiber support in an annealing furnace, and is heated from room temperature to 200°C at a rate of 5°C / min, pure nitrogen gas with purity of ≥99.99% is introduced into the furnace during the heating process, the nitrogen gas is introduced through 6 air inlets uniformly distributed at the bottom of the furnace body, and the flow rate is controlled to be 8 L / min; after being kept at 200°C for 30 min, the temperature is increased to 420°C at a rate of 3°C / min, and the flow rate of the nitrogen gas is adjusted to 10 L / min during this stage; S5: constant temperature treatment After the gradient heating in S3, the furnace temperature reaches 420°C, and then the constant temperature treatment is entered, the holding time is 1 hour, 2 variable frequency axial fans are arranged at the top and the bottom of the furnace body respectively, the diameter of the fan is 200 mm, the wind speed is adjusted to 2.5 m / s through the frequency converter, and the rotating speed of the fan is adjusted in real time through the PLC control system according to the temperature data fed back by the 5 platinum resistance temperature sensors uniformly arranged in the furnace; S6: segmented cooling After the constant temperature treatment in S5 is satisfied, the furnace is cooled to 200°C at a rate of 3°C / min, and the flow rate of the nitrogen gas is maintained at 5 L / min during the cooling process; when the furnace temperature drops to 200°C, the forced air cooling system is started, the air cooling device is composed of 4 axial fans and a serpentine cooling coil, 18°C cooling water is introduced into the coil, the cooling rate is controlled to be 6°C / min, and the temperature of the aluminum bar is reduced to 25±2°C; S7: performance detection The conductivity tester with precision grade of 0.5 and automatic temperature compensation function is used to detect 3 test points at the two ends and the middle of the aluminum bar after cooling in S6 in an environment of 15°C±0.5°C.

2. The method for improving the conductivity of 6063 aluminum busbar according to claim 1, characterized in that, In the S2 surface pretreatment step, ultrasonic cleaning is used to assist alcohol wiping: the aluminum row is placed in an alcohol solution at a temperature of 40°C, the frequency of the ultrasonic cleaning machine is set to 40 kHz, the power density is 0.3 W / cm 2 , the cleaning time is 8 min, and the solution needs to be kept circulating during the cleaning process.

3. The method for improving the conductivity of 6063 aluminum busbar according to claim 1, characterized in that, After the surface pretreatment in S2, 5000 groups of signals with labels collected by the eddy current flaw detector are analyzed by using a convolutional neural network with 2 convolutional layers, 2 pooling layers and 1 fully connected layer, the size of the convolution kernel is 3×3, and the length of the crack is ≥0.5 mm or the diameter of the inclusion defect is ≥0.3 mm is automatically identified.

4. The method for improving the conductivity of 6063 aluminum busbar according to claim 1, characterized in that, In the S4 gradient heating step, the optimal heating rate of 5-8 ℃ / min is calculated by inputting the Si content, Mg content and initial hardness value of the aluminum row into the machine learning model, which is a three-layer neural network containing 3 input neurons, 10 hidden neurons and 1 output neuron, using ReLU activation function.

5. The method for improving the conductivity of 6063 aluminum busbar according to claim 1, characterized in that, The annealing furnace used in the S4 gradient heating step adopts a double-layer vacuum insulation structure: the inner layer is a 5mm thick 310S stainless steel plate with a high temperature resistance ≥1200℃, the outer layer is a 50mm thick nano aerogel insulation layer with a thermal conductivity ≤0.02W / (m·K), and the pressure between the two layers is ≤1Pa, the furnace door is sealed with a silicon rubber sealing ring combined with an inflation structure.

6. The method of claim 1, wherein the 6063 aluminum alloy is a 6063- T5 alloy. In the 200℃ holding stage of the S4 gradient heating step, the nitrogen pressure in the furnace is maintained at 1.3 atm by a pressure regulating valve, and the furnace body is equipped with a pressure sensor with an accuracy of ±0.05 atm to monitor the pressure change in real time and control the intake valve opening degree.

7. The method of claim 1, wherein the 6063 aluminum alloy is a 6063-T5 alloy. In the S5 constant temperature treatment step, an infrared temperature monitoring device is installed in the furnace, with a measurement range of 0-600℃ and an accuracy of ±1℃, which monitors the surface temperature of the aluminum row in real time through four infrared probes distributed around the furnace body, and the monitoring data and platinum resistance sensor data are transmitted to the control system for cross verification.

8. The method for improving the conductivity of 6063 aluminum busbar according to claim 1, characterized in that, In the S6 segment cooling step, the axial flow fan of the forced air cooling system is arranged in an up-and-down staggered manner with the cooling coil, the fan outlet is 300 mm away from the surface of the aluminum row, and the cooling water flow in the coil is controlled at 5 m 3 / h through an electromagnetic flowmeter with an accuracy of ±1%.

9. The method for improving the conductivity of 6063 aluminum busbar according to claim 1, characterized in that, In the S7 performance detection step, the conductivity tester uses a standard conductivity test block before testing, the test block is an aluminum block with a purity ≥99.99%, the known conductivity value is 35.0MS / m, and three-point calibration is performed with a range of 0%, 50% and 100%, and the calibration deviation is ≤±0.3%FS.

10. The method of claim 9, wherein the 6063 aluminum alloy is a 6063- T6 temper.

10. The method of claim 9, wherein the 6063 aluminum alloy is a 6063- T651 temper. In the S7 performance detection, the contact pressure between the electrode and the surface of the aluminum row is 0.5MPa, each test point is measured 3 times, and the arithmetic mean value is taken as the detection result of the point, and finally the average value of 3 test points is taken as the aluminum row conductivity data.

Images