Staggered calibration method for ensuring consistency of high-uniformity foamed aluminum material

Abstract

The invention relates to the technical field of material detection, and particularly discloses a staggered calibration method for ensuring the consistency of a high-uniformity foamed aluminum material, and the staggered calibration method comprises the following steps: pretreating raw materials, heating molten aluminum through a mold, cooling and molding, conveying a demolded foamed aluminum ingot to a saw cutting station of a conveying belt, vertical saw blades of sawing machines on the two sides of a conveying belt are used for sawing in the width direction of the short edges of the two sides of a foam aluminum ingot, the internal foam structure section is exposed, and the problem of collection blind areas existing in traditional machine vision calibration is effectively solved. The right side identification system scans along the cross section from bottom to top, the upper area and the lower area of the cross section of the foamed aluminum can be fully covered, aperture information of each layer is completely obtained, information loss caused by single-direction scanning is avoided, and comprehensive and reliable basic data are provided for subsequent uniformity index calculation.

Description

Technical Field

[0001] This invention relates to the field of materials testing technology, and in particular to an interleaved calibration method for ensuring the consistency of high-uniformity aluminum foam materials. Background Technology

[0002] In the field of aluminum foam production, melt foaming is the mainstream preparation process. However, this process is easily affected by gravity and pressure difference during the foaming stage, causing the aluminum liquid to sink and the bubbles to float. The resulting aluminum foam has the problem of larger pore size at the top and smaller pore size at the bottom, and the uniformity is difficult to control.

[0003] Uniformity is a key indicator affecting the compressive strength and energy absorption buffering performance of aluminum foam, and its control is directly related to product quality. However, existing technologies have obvious defects in the calibration of aluminum foam uniformity.

[0004] Traditional manual calibration methods rely on operators' visual observation and experience, which is not only time-consuming and labor-intensive, but also prone to inaccurate calibration results due to subjective errors, making it difficult to meet the requirements of precision and efficiency for large-scale production.

[0005] Even with the introduction of machine vision into some technologies, such as the machine vision-based device and method for measuring the pore size of aluminum foam (publication number CN102135415A), most of them use single-direction scanning to collect data. This cannot fully cover the upper and lower areas of the cross-section of aluminum foam, and blind spots are easily formed. This results in incomplete pore size information acquisition, making it difficult to accurately quantify the difference in uniformity between the upper and lower parts. Consequently, it cannot provide reliable data support for subsequent grading and screening, thus restricting the stable production of high-uniformity aluminum foam materials. Summary of the Invention

[0006] The purpose of this invention is to provide an interleaved calibration method for ensuring the consistency of high-uniformity aluminum foam materials, thereby solving the problems mentioned in the background art.

[0007] To achieve the above objectives, the present invention provides the following technical solution: an interleaved calibration method for ensuring the consistency of high-uniformity aluminum foam materials, wherein the interleaved calibration method includes the following steps: S1: Raw material pretreatment, the molten aluminum liquid is heated and cooled by the mold and the foamed aluminum ingot is demolded and then transported to the conveyor belt sawing station. The vertical saw blades of the sawing machines on both sides of the conveyor belt are used to saw along the short side width of the foamed aluminum ingot to expose the internal foam structure cross-section. S2: Dual-modal scanning. The sawn foam aluminum ingot is conveyed to the identification station by the conveyor belt. The station is equipped with dual identification systems located on both sides of the conveyor belt. The left identification system scans from top to bottom along the cross-section of the foam aluminum ingot, and the right identification system scans from bottom to top along the cross-section. The dual identification systems start synchronously. S3: Layered Quantitative Analysis: The dual recognition system acquires layered images of the cross-section of the foamed aluminum ingot. The acquisition height of each layer is less than or equal to 10mm. After the images are transmitted to the computer, the computer processes the images and extracts the top and bottom pore diameters of the layer. Then, the preset algorithm is used to calculate the pore diameter uniformity index. S4: Dynamic grading control: Grading decisions are made based on the pore size uniformity index of each layer. If the index of a certain layer exceeds the preset qualified threshold, the foam aluminum ingot is determined to be unqualified and transported to the unqualified product area. If the index of a certain layer meets the qualified threshold, the index of subsequent layers continues to be collected and calculated. When multiple layers meet the qualified threshold, it is determined to be qualified and transported to the qualified product area.

[0008] Preferably, in the S2 dual-modal scanning, the dual system includes three or more CCD cameras arranged in a row. The CCD cameras are respectively mounted on the recognition brackets on both sides of the conveyor belt and correspond to the cross-section of the aluminum foam ingot. The longitudinal spacing of the CCD cameras is 8 to 12 mm, and the overlap rate of the acquisition areas of adjacent CCD cameras is 5% to 10%.

[0009] Preferably, the preset algorithm in the S3 stratified quantification analysis is a relative aperture difference formula, which is as follows: ; in, As a uniformity index, The average aperture at the top. The average aperture at the bottom; The average aperture at the top The arithmetic mean of 5 to 8 apertures randomly selected from the top region of this layer is the bottom average aperture. The arithmetic mean of 5 to 8 apertures were randomly selected from the bottom region of this layer.

[0010] Preferably, the preset qualified threshold is When the thickness is greater than or equal to 100mm, the preset qualified threshold is .

[0011] Preferably, in the S1 raw material pretreatment, the sawing depth along the short side width direction of both sides of the foamed aluminum ingot is 1 / 3 to 1 / 2 of the width of the foamed aluminum ingot, the sawing speed of the vertical saw blade during the sawing process is 20 to 30 mm / s, and the number of teeth of the saw blade is 30 to 40 teeth / inch.

[0012] Preferably, in the S3 hierarchical quantization analysis, when the computer processes the image, an abnormal bubble interference filtering algorithm is used, which is implemented based on OpenCV; First, morphological opening operations are performed using a 3×3 rectangular structural element; Then, a morphological closing operation is performed using a 5×5 rectangular structural element.

[0013] Preferably, in the S2 dual-modal scanning, the conveyor belt speed and the image acquisition frame rate satisfy the following conditions: ; in, This refers to the conveyor belt speed, expressed in mm / s. The image acquisition frequency of the dual recognition system is expressed in Hz. The height of each layer collected in step S3 is in mm.

[0014] Preferably, in the S4 dynamic grading control, when all multiple layers meet the qualified threshold, specifically when the pore size uniformity index of three or more consecutive layers meets the preset qualified threshold.

[0015] Preferably, in the S2 dual-modal scanning, the dual system further includes a laser-assisted positioning module. The laser-assisted positioning module is installed at the acquisition entrance of the dual recognition system. The laser-assisted positioning module emits a linear laser with an accuracy of ±0.1mm, and the overlap between the laser beam and the central axis of the cross-section of the aluminum foam ingot is less than or equal to 0.05mm.

[0016] Preferably, in the S2 dual-modal scanning, the dual recognition system is also equipped with an ambient light compensation device, which automatically adjusts the camera exposure parameters according to the lighting conditions of the recognition station.

[0017] Compared with the prior art, the beneficial effects of the present invention are: 1. This invention effectively solves the problem of blind spots in traditional machine vision calibration. Through a dual-recognition system with synchronous staggered scanning design, the left recognition system scans from top to bottom along the cross-section of the aluminum foam ingot, while the right recognition system scans from bottom to top along the same cross-section. This can fully cover the upper and lower areas of the aluminum foam cross-section, completely acquire the pore size information of each layer, and avoid information loss caused by scanning in a single direction. This provides comprehensive and reliable basic data for subsequent uniformity index calculation, ensuring the comprehensiveness and accuracy of uniformity calibration, and helping to screen out truly high-uniformity aluminum foam materials that meet the requirements.

[0018] 2. This invention automates the entire process, from sawing the aluminum foam ingot to expose its internal cross-section, to the dual-identification system acquiring images layer by layer, the computer processing data and calculating the uniformity index, and then making dynamic grading decisions based on the index. This significantly shortens the calibration time. At the same time, by quantifying the uniformity index through a preset algorithm and combining it with dynamic grading control logic, it can accurately distinguish between qualified and unqualified products, avoiding subjective errors caused by human experience judgment. This provides a standardized and precise technical means for the quality control of aluminum foam materials.

[0019] 3. This invention has good practicality and adaptability, and can be stably applied to the large-scale production of aluminum foam. The sawing pretreatment step can effectively expose the internal structure of aluminum foam, creating favorable conditions for subsequent scanning and acquisition. The layered quantitative analysis step ensures the detail of pore size information acquisition by controlling the acquisition height of each layer. The dynamic grading control step achieves automated screening based on the uniformity index. The entire process is smoothly connected and can be seamlessly integrated with existing conveyor belt systems without the need for large-scale modification of the production line, thus reducing the difficulty of technology implementation. Attached Figure Description

[0020] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0021] Figure 1 This is a flowchart illustrating the staggered calibration method of the present invention. Detailed Implementation

[0022] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0023] Example 1, please refer to Figure 1 As shown, an interleaved calibration method is used to ensure the consistency of highly uniform aluminum foam materials. The interleaved calibration method includes the following steps: S1: Raw material pretreatment. Molten aluminum liquid is heated and cooled in a mold, and the foamed aluminum ingot is then demolded and transported to the conveyor belt sawing station. The vertical saw blades on both sides of the conveyor belt are used to saw along the width of the short side of the foamed aluminum ingot, exposing the cross-section of the internal foam structure. When sawing, the saw blade is aligned with the edge of the short side of the ingot 10-15mm away to avoid damaging the core foam structure. The compressed air blowing direction is at 45° to the sawing surface.

[0024] The molten aluminum is 99.7% industrial pure aluminum, with 0.5% titanium and 1.2% titanium hydride added. The mold is heated to 720℃ and held for 15 minutes, with a cooling rate of 5-8℃ / min. After demolding, it is cooled to room temperature. The conveyor belt is of type DTII with a polytetrafluoroethylene anti-slip layer. The speed is 10-15 mm / s. The saw blade is made of high-speed steel, with a sawing speed of 20-30 mm / s. After sawing, aluminum chips are blown away with 0.2 MPa compressed air. A drawer-type chip collection box is installed under the sawing station and is cleaned once a day to prevent aluminum chips from scattering and contaminating the equipment. The mold is made of heat-resistant steel. When demolding, use a wooden tool to gently tap it to prevent the ingot from being bumped. Position sensors are set every 2 meters on the conveyor belt to ensure accurate delivery of the ingot.

[0025] S2: Dual-modal scanning. The sawn foam aluminum ingot is conveyed to the identification station by the conveyor belt. The station is equipped with dual identification systems located on both sides of the conveyor belt. The left identification system scans from top to bottom along the cross-section of the foam aluminum ingot, and the right identification system scans from bottom to top along the cross-section. The dual identification systems start synchronously. The conveyor belt is 1 to 1.5 meters away from the identification station, with side guide plates on both sides to prevent deviation. The identification bracket is 1.2 to 1.8 meters high. It has a dual system with white backlight LED light source, brightness of 800-1200 lux, and synchronization error ≤10ms. The side guide plate surface is covered with rubber pads to avoid scratching the spindle. The bottom of the identification bracket is equipped with leveling feet to ensure that the bracket is stable and does not wobble. The brightness of the light source can be finely adjusted by the knob to adapt to different ambient light. S3: Layered Quantitative Analysis: The dual recognition system acquires layered images of the cross-section of the foamed aluminum ingot. The acquisition height of each layer is less than or equal to 10mm. After the images are transmitted to the computer, the computer processes the images and extracts the top and bottom pore diameters of the layer. Then, the preset algorithm is used to calculate the pore diameter uniformity index. Images are acquired at intervals of 0.5 to 1 second per layer and transmitted to an industrial computer via Ethernet. The computer processes the images using OpenCV with a threshold of 120-150. 5-8 bubbles are selected and averaged, avoiding the edges by 2-3mm. The Ethernet interface uses shielded cables to reduce signal interference. The computer periodically clears the cache to ensure image processing speed. The threshold can be adjusted slightly according to the image grayscale to ensure clear separation of bubbles from the background. S4: Dynamic grading control: Grading decisions are made based on the pore size uniformity index of each layer. If the index of a certain layer exceeds the preset qualified threshold, the foam aluminum ingot is determined to be unqualified and transported to the unqualified product area. If the index of a certain layer meets the qualified threshold, the index of subsequent layers continues to be collected and calculated. When multiple layers meet the qualified threshold, it is determined to be qualified and transported to the qualified product area. The specific steps of hierarchical decision-making in S4 dynamic hierarchical control are as follows: Step 1: Thickness detection and threshold matching: Use a thickness measuring device to detect the thickness of the foamed aluminum ingot, take the average value of the detection results at the center and both sides of the ingot, and preset the qualified threshold according to the pore size uniformity corresponding to the thickness matching.

[0026] Step 2: Index Comparison and Preliminary Judgment: The computer compares the pore size uniformity index of each layer with the preset qualified threshold. If the index exceeds the threshold, it is judged as unqualified. If it meets the threshold, the subsequent layers are tested.

[0027] Step 3: Continuous layer verification: Continuously detect the indicators of subsequent layers of foamed aluminum ingots that meet the threshold, and record the number of layers that continuously meet the threshold. If the set number is reached, it is judged as qualified. If the indicator exceeds the threshold, the detection is terminated and it is judged as unqualified.

[0028] Step 4: Diversion and Recording: Start the diversion device to send qualified products to the qualified area and unqualified products to the processing area. The computer automatically records the thickness of the foamed aluminum ingot, the indicators of each layer, the judgment result and the destination. The data is stored in the database and backed up regularly.

[0029] Thickness < 100mm threshold ≥100mm The conveyor belt ends with a pneumatic push rod for diversion, rubber pads are laid in the qualified area, computer records information and stores it in a database, the pneumatic push rod is regularly checked for sealing to prevent air leakage from affecting the operation, the rubber pad is 5-8mm thick to avoid the spindle from impact and deformation, and the database is backed up once a week to prevent information loss.

[0030] Example 2, please refer to Figure 1 As shown, in the S2 dual-modal scanning, the dual system includes a row of linearly arranged CCD cameras. The number of cameras is adapted to the width of the foamed aluminum ingot. When the width is <100mm, 2 cameras are configured per row; when the width is ≥100mm, ≥3 cameras are configured per row. The CCD cameras are respectively installed on the recognition brackets on both sides of the conveyor belt and correspond to the cross-section of the foamed aluminum ingot.

[0031] The longitudinal spacing between CCD cameras is 8–12 mm, with a spacing error of ≤0.1 mm. The overlap rate of the acquisition areas of adjacent CCD cameras is 5%–10%, and the data from the overlapping area is averaged.

[0032] CCD camera resolution ≥ 1280×960, frame rate 1~3Hz, the frame rate must be matched with the conveyor belt speed, the matching formula is as follows: V is the conveyor belt speed, h is the acquisition height of each layer of S3, and f is the image acquisition frequency. The support is equipped with guide rails, and the surface of the guide rails is coated with lubricating oil. The camera position can be adjusted along the guide rails. After adjustment, the spacing is checked with calipers to ensure accuracy.

[0033] The default algorithm in S3 stratified quantitative analysis is the relative aperture difference formula, which is as follows: ; in, As a uniformity index, The average aperture at the top. The average aperture at the bottom; Top average aperture The arithmetic mean of 5 to 8 pore sizes were randomly selected from the top region of this layer, and the average pore size at the bottom was calculated. The arithmetic mean of 5 to 8 apertures were randomly selected from the bottom region of this layer.

[0034] The range is 0-1, with the closer to 0 being more uniform. The bubbles must be intact. The aperture is measured 3 times and the average is taken. The result is rounded to four decimal places. When selecting bubbles, prioritize areas with uniform distribution and avoid concentration in the same local area. Use an image measurement tool to measure the aperture and ensure that the starting point and ending point of each measurement are consistent. Rounding the result to four decimal places can accurately reflect the uniformity difference.

[0035] The preset qualified threshold is When the thickness is greater than or equal to 100mm, the preset acceptable threshold is... .

[0036] Thickness is measured using the Keyence LK-G80 with an accuracy of ±0.1mm. The average of three points (center and both sides) is taken. The computer reads the data and automatically matches the threshold, requiring no manual intervention. Before measurement, the thickness gauge needs to be calibrated, and the accuracy is verified using a standard thickness block. The measurement points are marked on the surface of the ingot to ensure that the measurement position is consistent each time. The computer and the thickness gauge are connected via a data cable to achieve real-time data transmission.

[0037] S1 raw material pretreatment, sawing depth along the short side width direction of both sides of the foamed aluminum ingot is 1 / 3-1 / 2 of the width of the foamed aluminum ingot, sawing speed of vertical saw blade is 20-30mm / s during sawing, saw blade teeth number is 30-40 teeth / inch.

[0038] For a width of 80-120mm, cut to a depth of 1 / 3; for 120-180mm, cut to a depth of 1 / 2. For a 30-tooth blade, cut at a speed of 20-25mm / s; for a 40-tooth blade, cut at a speed of 25-30mm / s. Replace the saw blade if its wear exceeds 0.2mm. After replacing the saw blade, perform a test cut to check the quality of the cut surface. If it meets the requirements, proceed with batch cutting. Measure the wear amount with a special measuring tool to ensure timely replacement.

[0039] In S3 hierarchical quantization analysis, when the computer processes images, it uses the abnormal bubble interference filtering algorithm, which is implemented based on OpenCV. First, morphological opening operations are performed using a 3×3 rectangular structural element; Then, a morphological closing operation is performed using a 5×5 rectangular structural element.

[0040] When the computer processes images in the S3 hierarchical quantization analysis, the specific steps are as follows: Define abnormal bubbles as tiny bubbles with a size < 0.5 mm, super-large bubbles with a size > 3 times the average pore diameter of this layer, and残缺 bubbles with broken edges or connected to other bubbles. For the filtering operation, first perform morphological opening operation with a 3×3 rectangular structural element to remove tiny bubble impurities, and then perform morphological closing operation with a 5×5 rectangular structural element to fill the tiny gaps of qualified bubbles. Each operation is iterated 1 time. If it is detected that the proportion of super-large bubbles in the local area > 15%, start secondary filtering, perform closing operation with a 7×7 rectangular structural element, and mark local bubble anomalies in the data record.

[0041] Control the gray value of the processed image within 80 - 120, and the contrast ≥ 30. Verify through the image analysis tool to ensure that the bubbles and the background are clearly separated.

[0042] Example three, please refer to Figure 1 As shown, in the S2 dual-modal scanning, the conveyor belt speed and the image acquisition frame rate satisfy ; Among them, is the conveyor belt speed, with the unit of mm / s, is the image acquisition frequency of the dual-recognition system, with the unit of Hz, is the acquisition height of each layer in step S3, with the unit of mm.

[0043] The actual speed takes 80% - 9, such as h = 10 mm, f = 2 Hz, v ≤ 20 mm / s is set to 16 - 18 mm / s. The variable-frequency speed regulator controls the speed with an accuracy of ±0.1 mm / s. Try to collect 3 - 5 times without missing frames before using. Regularly maintain the variable-frequency speed regulator and clean the dust. Observe the continuity of the image during the test collection. If there is no overlap and no missing, it can be officially run. The speed adjustment is operated through the control panel, which is intuitive and convenient.

[0044] In the S4 dynamic grading control, when all multi-layer levels meet the qualified threshold, specifically when the pore size uniformity indexes of three consecutive layers and more than three layers all meet the preset qualified threshold.

[0045] Verify from the first layer. If the first three layers are qualified, it is judged as qualified. If any layer is unqualified, the collection is terminated and judged as unqualified. The computer interface shows green for qualified and red for unqualified, and generates a report containing the ingot number and indexes. The report can be exported in Excel format for easy statistical analysis. The ingot number corresponds to the foam aluminum ingot one by one for easy traceability.

[0046] In the S2 dual-modal scanning, the dual system also includes a laser-assisted positioning module. The laser-assisted positioning module is installed at the acquisition entrance of the dual recognition system. The laser-assisted positioning module emits a linear laser with an accuracy of ±0.1mm, and the overlap between the laser beam and the central axis of the cross-section of the aluminum foam ingot is less than or equal to 0.05mm.

[0047] In the S2 dual-modal scanning, the dual system also includes a laser-assisted positioning module. The module is a 650nm Class IIIa semiconductor laser with a power of 5-10mW, which complies with the "Laser Safety Protection Specification" GB7247.1. It is installed 50-80mm before the acquisition entrance of the dual identification system, emits a linear laser with an accuracy of ±0.1mm, and the overlap between the laser beam and the central axis of the cross-section of the aluminum foam ingot is ≤0.05mm.

[0048] Laser safety protection measures include: installing a transparent PC board protective cover in the laser emission area, with a thickness of ≥3mm, and affixing laser hazard and direct-view warning signs to the protective cover; The system identifies workstations equipped with human body induction interlock devices. When a person is detected entering the laser radiation range, and the radiation range radius is 1m, the laser power supply is automatically cut off. Operators must wear laser safety glasses with a wavelength of 650nm and an optical density of OD4+. The equipment calibration and abnormal handling procedure is as follows: before starting the production line each day, the laser position is calibrated with a 20x microscope. If the deviation exceeds 0.05mm, the equipment will automatically stop and sound an alarm. After the operator manually adjusts the laser position to the qualified range, the overlap is checked by microscopy. The production line can only be started after confirming that the deviation is ≤0.05mm.

[0049] In the S2 dual-modal scanning system, the dual recognition system is also equipped with an ambient light compensation device, which automatically adjusts the camera exposure parameters according to the lighting conditions at the recognition station.

[0050] The device includes an E3Z-D61 photosensitive sensor and a 15-20W LED light. When the light level is less than 500 lux, the exposure time is extended to 15ms when the light is turned on and to 8ms when the light level is greater than 800 lux. The response time is ≤100ms. The photosensitive sensor is installed on the top of the recognition station to avoid obstruction. The LED light angle is adjustable to ensure uniform illumination. The response time is tested with a timer to ensure rapid adaptation to changes in light.

[0051] The specific steps for automatically adjusting camera exposure parameters are as follows: Step 1: Light Monitoring Start-up: When the dual recognition system starts scanning, the photosensitive sensor on the top of the recognition station is turned on simultaneously to continuously monitor the light at the station. The monitoring frequency is consistent with the image acquisition frequency.

[0052] Step 2: Light Data Transmission and Analysis: The photosensitive sensor transmits light intensity data to the ambient light compensation device control module in real time. The control module analyzes whether the light meets the requirements for clear image acquisition.

[0053] Step 3: Exposure parameter adjustment decision: If the light is below the appropriate range, the control module instructs the camera to increase the exposure time; if the light is above the appropriate range, the camera instructs to shorten the exposure time; if the light is within the appropriate range, the exposure parameters remain unchanged.

[0054] Step 4: Parameter Adjustment and Feedback: The camera adjusts the exposure parameters according to the instructions and then feeds the parameters back to the control module. If the image is still not clear, repeat the light monitoring and parameter adjustment process.

[0055] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. An interleaved calibration method for ensuring the consistency of highly uniform aluminum foam materials, characterized in that: The interleaved calibration method includes the following steps: S1: Raw material pretreatment, the molten aluminum liquid is heated and cooled by the mold and the foamed aluminum ingot is demolded and then transported to the conveyor belt sawing station. The vertical saw blades of the sawing machines on both sides of the conveyor belt are used to saw along the short side width of the foamed aluminum ingot to expose the internal foam structure cross-section. S2: Dual-modal scanning. The sawn foam aluminum ingot is conveyed to the identification station by the conveyor belt. The station is equipped with dual identification systems located on both sides of the conveyor belt. The left identification system scans from top to bottom along the cross-section of the foam aluminum ingot, and the right identification system scans from bottom to top along the cross-section. The dual identification systems start synchronously. S3: Layered Quantitative Analysis: The dual recognition system acquires layered images of the cross-section of the foamed aluminum ingot. The acquisition height of each layer is less than or equal to 10mm. After the images are transmitted to the computer, the computer processes the images and extracts the top and bottom pore diameters of the layer. Then, the preset algorithm is used to calculate the pore diameter uniformity index. S4: Dynamic grading control: Grading decisions are made based on the pore size uniformity index of each layer. If the index of a certain layer exceeds the preset qualified threshold, the foam aluminum ingot is determined to be unqualified and transported to the unqualified product area. If the index of a certain layer meets the qualified threshold, the index of subsequent layers continues to be collected and calculated. When multiple layers meet the qualified threshold, it is determined to be qualified and transported to the qualified product area.

2. The staggered calibration method for ensuring the consistency of high-uniformity aluminum foam materials according to claim 1, characterized in that: In the S2 dual-modal scanning, the dual system includes three or more CCD cameras arranged in a row. The CCD cameras are respectively mounted on the recognition brackets on both sides of the conveyor belt and correspond to the cross-section of the aluminum foam ingot. The longitudinal spacing of the CCD cameras is 8 to 12 mm, and the overlap rate of the acquisition areas of adjacent CCD cameras is 5% to 10%.

3. The staggered calibration method for ensuring the consistency of high-uniformity aluminum foam materials according to claim 1, characterized in that: The preset algorithm in the S3 stratified quantitative analysis is the relative aperture difference formula, which is as follows: ； in, As a uniformity index, The average aperture at the top. The average aperture at the bottom; The average aperture at the top The arithmetic mean of 5 to 8 pore sizes are randomly selected from the top region of this layer, and the bottom average pore size is... The arithmetic mean of 5 to 8 apertures were randomly selected from the bottom region of this layer.

4. The staggered calibration method for ensuring the consistency of high-uniformity aluminum foam materials according to claim 3, characterized in that: The preset qualified threshold is: When the thickness is greater than or equal to 100mm, the preset qualified threshold is .

5. The staggered calibration method for ensuring the consistency of high-uniformity aluminum foam materials according to claim 1, characterized in that: The pretreatment of raw material S1 involves sawing along the short side width of both sides of the aluminum foam ingot to a depth of 1 / 3 to 1 / 2 of the width of the aluminum foam ingot. During the sawing process, the sawing speed of the vertical saw blade is 20 to 30 mm / s, and the number of teeth on the saw blade is 30 to 40 teeth / inch.

6. The staggered calibration method for ensuring the consistency of high-uniformity aluminum foam materials according to claim 1, characterized in that: In the S3 hierarchical quantization analysis, when the computer processes the image, it uses an abnormal bubble interference filtering algorithm, which is implemented based on OpenCV. First, morphological opening operations are performed using a 3×3 rectangular structural element; Then, a morphological closing operation is performed using a 5×5 rectangular structural element.

7. The staggered calibration method for ensuring the consistency of high-uniformity aluminum foam materials according to claim 1, characterized in that: In the S2 dual-modal scanning, the conveyor belt speed and the image acquisition frame rate satisfy the following conditions: ; in, This refers to the conveyor belt speed, expressed in mm / s. The image acquisition frequency of the dual recognition system is expressed in Hz. The height of each layer collected in step S3 is in mm.

8. The staggered calibration method for ensuring the consistency of high-uniformity aluminum foam materials according to claim 1, characterized in that: In the S4 dynamic grading control, when multiple layers meet the qualified threshold, specifically when the pore size uniformity index of three or more consecutive layers meets the preset qualified threshold.

9. The staggered calibration method for ensuring the consistency of high-uniformity aluminum foam materials according to claim 1, characterized in that: In the S2 dual-modal scanning, the dual system also includes a laser-assisted positioning module. The laser-assisted positioning module is installed at the acquisition entrance of the dual recognition system. The laser-assisted positioning module emits a linear laser with an accuracy of ±0.1mm, and the overlap between the laser beam and the central axis of the cross-section of the aluminum foam ingot is less than or equal to 0.05mm.

10. The staggered calibration method for ensuring the consistency of high-uniformity aluminum foam materials according to claim 1, characterized in that: In the S2 dual-modal scanning, the dual recognition system is also equipped with an ambient light compensation device, which automatically adjusts the camera exposure parameters according to the lighting conditions at the recognition station.

Images