Nondestructive testing method for graphite electrode manufacturing process
By combining X-ray CT and phased array ultrasound in a multimodal detection method, the problem of being unable to identify minute defects in graphite electrodes in existing technologies has been solved, achieving high-precision non-destructive testing and production optimization, and reducing the scrap rate.
Patent Information
- Application Number
- CN202511474562.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-10-15
- Publication Date
- 2026-01-20
AI Technical Summary
Existing detection technologies cannot effectively identify minute defects such as 50μm-level microcracks and 100μm-level micropores inside graphite electrodes, and lack synergistic and complementary capabilities, making it difficult to comprehensively and accurately assess the internal quality of graphite electrodes.
A multimodal collaborative nondestructive testing method is adopted, combining X-ray CT and phased array ultrasound to perform three-dimensional scanning imaging and full-area scanning. Defects are identified through data fusion and artificial intelligence, generating inspection reports and optimizing production processes.
It enables accurate identification of macroscopic and microscopic defects inside graphite electrodes, improves detection accuracy, reduces scrap rate, and achieves closed-loop quality control in the production process.
Abstract
Description
TECHNICAL FIELD
[0001] The present application relates to the technical field of graphite electrode detection, in particular to a nondestructive testing method for the manufacturing process of graphite electrodes. BACKGROUND
[0002] As the core conductive material and key consumable in the high-temperature electric furnace smelting and refining process, the internal quality of graphite electrodes directly determines the stability of the smelting process, the energy consumption level and the product qualification rate.
[0003] The existing detection technology has obvious defects in defect recognition accuracy, and can only effectively identify defects above millimeter level. For the common 50μm microcracks and 100μm small pores inside the graphite electrode, the key micro-defects cannot meet the quality control requirements of high-end graphite electrodes. At the same time, the existing detection means is relatively single and lacks synergistic complementary ability: when X-ray detection is used, although the internal macrostructure and density distribution of the electrode can be accurately presented, the detection sensitivity of microcracks and interface defects is low, and missed detection is easy to occur; when ultrasonic detection is used, although some microcracks can be detected, the problem of internal macroscopic density unevenness of the electrode cannot be directly reflected, and the two types of detection technologies cannot form effective synergy, making it difficult to comprehensively and accurately evaluate the internal quality of the graphite electrode. SUMMARY
[0004] The technical problem to be solved by the present application is to overcome the above technical defects and provide a multi-modal synergistic, intelligent and closed-loop controllable nondestructive testing method, which can accurately identify the internal macroscopic defects and microscopic defects of the graphite electrode, and at the same time link the production system to optimize the process and reduce waste.
[0005] To solve the above technical problems, the technical solution provided by the present application is as follows: a nondestructive testing method for the manufacturing process of graphite electrodes, comprising the following steps:
[0006] S1: cleaning the surface of the graphite electrode to be detected and positioning and fixing it on the conveying track of the detection platform;
[0007] S2: three-dimensional scanning imaging of the graphite electrode to obtain original data of the internal macrostructure and density distribution of the electrode;
[0008] S3: using a phased array ultrasonic probe to scan the graphite electrode in the whole area and synchronously collecting ultrasonic echo signals;
[0009] S4: fusing the two types of data to construct an intelligent defect recognition and classification model based on artificial intelligence, analyzing the fused data, identifying the defect type and preliminarily determining the defect severity;
[0010] S5: quantitatively evaluating the size, position, number and severity of the defects and generating a detection report;
[0011] S6: transmitting the detection data to the enterprise production execution system and the programmable logic controller to realize traceability or trigger the unqualified product rejection instruction, and complete the quality closed-loop control of the graphite electrode manufacturing process.
[0012] Preferably, the three-dimensional scanning imaging in S2 includes using a microfocus or small focus X-ray source and a tiled detector, and processing the original data includes image reconstruction algorithms, artifact correction algorithms, and defect enhancement algorithms to obtain preliminary imaging results of macroscopic defects of the graphite electrode.
[0013] Preferably, the positioning and fixing in S1 is specifically achieved by using an adjustable clamping mechanism on the detection platform, so as to ensure that the coaxiality deviation between the electrode axis and the center line of the conveying track is less than or equal to 0.5 mm.
[0014] Preferably, the processing of the echo signal in S3 includes adaptive filtering, beam synthesis optimization, and signal enhancement.
[0015] Preferably, the focus size of the microfocus or small focus X-ray source is less than or equal to 50 μm, and the tiled detector uses 2-4 groups of detector units for tiling, and the pixel size of a single group of detector units is less than or equal to 100 μm.
[0016] Preferably, the method for data fusion in S4 uses a feature layer fusion algorithm, specifically by extracting the density difference features of the data and the acoustic impedance difference features of the phased array ultrasonic data, and using a weighted fusion model to fuse the two types of features.
[0017] Preferably, the quantitative evaluation in S5 specifically includes: for crack defects, quantitatively detecting cracks with a minimum width of greater than or equal to 50 μm and a minimum length of greater than or equal to 200 μm; for pore defects, quantitatively detecting pores with a minimum diameter of greater than or equal to 100 μm; for inclusion defects, quantitatively detecting inclusions with a minimum volume of greater than or equal to 0.1 mm 3 ; and for interface defects, quantitatively detecting regions with an interface separation width of greater than or equal to 30 μm, and the position positioning error of all defects is less than or equal to 1 mm.
[0018] Preferably, the unqualified product rejection in S6 is specifically: when a defect size exceeding a critical value (such as a crack length of greater than or equal to 5 mm) is detected, the conveying track automatically transfers the electrode to the unqualified product storage.
[0019] Preferably, a synchronous acquisition system is used to realize the time and space alignment of the detection data and the phased array ultrasonic detection data.
[0020] Preferably, the programmable logic controller system automatically adjusts the production process parameters according to the defect judgment result, and when the number of detected defects exceeds a preset threshold, the programmable logic controller system increases the pressing pressure of the subsequent electrode by 5%-10% or prolongs the baking holding time by 1-2 h.
[0021] The advantages of the present application compared with the prior art are:
[0022] 1. Through the deep integration of X-ray CT and phased array ultrasound, both macroscopic density defects and microscopic micro-cracks and interface defects can be accurately identified, the smallest detectable crack is 50 pm wide and the smallest detectable pore is 100 pm in diameter, the detection accuracy is high, and the defects are covered comprehensively.
[0023] 2. Automatic scanning and data processing are adopted to realize online full detection, replace the traditional sampling inspection mode, cover the whole production process, and improve the detection efficiency.
[0024] 3. The detection data are deeply linked with the enterprise production execution system and programmable logic controller, so that the quality data can be traced back, the production process can be automatically optimized, unqualified products can be removed, manual intervention can be reduced, and the waste product rate can be reduced.
[0025] 4. The core equipment can use domestic components, the procurement and maintenance costs are greatly reduced, and the production needs of domestic graphite electrode enterprises are met. DETAILED DESCRIPTION
[0026] The present application will be further described in detail below in combination with the content.
[0027] A non-destructive testing method for a graphite electrode manufacturing process, comprising the following steps:
[0028] S1: cleaning the surface of the graphite electrode to be detected, and positioning and fixing it on the conveying track of the detection platform;
[0029] S2: three-dimensional scanning imaging of the graphite electrode to obtain original data of the internal macroscopic structure and density distribution of the electrode;
[0030] S3: using a phased array ultrasonic probe to scan the graphite electrode in the whole area and synchronously collecting ultrasonic echo signals;
[0031] S4: fusing the two types of data to construct an intelligent defect recognition and classification model based on artificial intelligence, analyzing the fused data, identifying the defect type and preliminarily determining the defect severity;
[0032] S5: quantitatively evaluating the size, position, number and severity of the defects, and generating a detection report;
[0033] S6: transmitting the detection data to the enterprise production execution system and programmable logic controller to realize traceability, or triggering unqualified product removal instructions, and completing the quality closed-loop control of the graphite electrode manufacturing process.
[0034] The three-dimensional scanning imaging in S2 includes using a microfocus or small focus X-ray source, and a tiled detector, and processing the original data includes an image reconstruction algorithm, an artifact correction algorithm and a defect enhancement algorithm to obtain a preliminary imaging result of macroscopic defects of the graphite electrode, the focus size of the microfocus or small focus X-ray source is ≤ 50 μm, and the tiled detector adopts 2-4 groups of detector units for tiling, and the pixel size of a single group of detector units is ≤ 100 μm.
[0035] The processing on the echo signal in S3 includes adaptive filtering, beam synthesis optimization and signal enhancement.
[0036] The method for data fusion in S4 adopts a feature layer fusion algorithm, specifically, density difference features of data and acoustic impedance difference features of phased array ultrasonic data are extracted, a weighted fusion model is used to fuse the two types of features, and time and space alignment of detection data and phased array ultrasonic detection data is realized by using a synchronous acquisition system.
[0037] The quantitative evaluation in S5 specifically includes: for crack defects, cracks with a minimum width ≥ 50 μm and a minimum length ≥ 200 μm are quantitatively detected; for pore defects, pores with a minimum diameter ≥ 100 μm are quantitatively detected; for inclusion defects, inclusions with a minimum volume ≥ 0.1 mm 3 are quantitatively detected; and for interface defects, regions with an interface separation width ≥ 30 μm are quantitatively detected, and the position positioning error of all defects is ≤ 1 mm.
[0038] The unqualified product rejection in S6 specifically includes: when a defect size exceeding a critical value (such as a crack length ≥ 5 mm) is detected, the conveying track automatically transfers the electrode to an unqualified product storage, the programmable logic controller system automatically adjusts production process parameters according to a defect judgment result, and when the number of detected defects exceeds a preset threshold, the programmable logic controller system increases the pressing pressure of a subsequent electrode by 5%-10%, or prolongs the baking holding time by 1-2 h.
[0039] In specific implementation of the present application:
[0040] Step one: graphite electrode pretreatment and positioning and fixing
[0041] Compressed air is used to blow off dust on the surface of the electrode, and then a dust-free cloth dipped in anhydrous ethanol is used to wipe the surface several times along the axial direction to remove residual oil stains on the surface, and then the electrode is placed on the conveying track of the detection platform, the hydraulic clamping mechanism on both sides is started, and the coaxiality deviation between the electrode axis and the center line of the track is calibrated by a laser alignment instrument and is ≤ 0.5 mm.
[0042] Step two: high-resolution X-ray CT three-dimensional scanning imaging
[0043] A domestic micro-focus X-ray source and a 3-group spliced detector are adopted, spiral scanning is performed, a three-dimensional image is reconstructed by using a filtered back-projection algorithm after a certain time, scattering artifacts are eliminated by using a scattering correction algorithm, and a gray-scale adaptive enhancement algorithm is used to improve the contrast of the density difference area.
[0044] Step three: phased array ultrasound full-area scanning and signal processing
[0045] A phased array ultrasound probe is used, and a water immersion or coupling agent type ultrasound coupling process is adopted to ensure stable contact between the probe and the electrode surface, and the graphite electrode is scanned in the full area along the conveying track, and the ultrasonic echo signal is synchronously collected.
[0046] Step four: multi-modal data fusion and AI defect identification
[0047] The X-ray CT data of step two and the phased array ultrasound data of step three are time and space aligned by the synchronous acquisition system, ensuring that the two types of data correspond to the same detection area, and then a feature layer fusion algorithm is used to fuse the data, enhance the defect feature recognition, and analyze the fused data based on artificial intelligence defect intelligent identification. The defect type is automatically identified, and the defect severity is preliminarily determined according to the defect feature parameters.
[0048] Step five: defect quantitative evaluation and detection report generation
[0049] According to the crack defect, the minimum width of the crack is ≥50μm, the minimum length of the crack is ≥200μm; for the pore defect, the minimum diameter of the pore is ≥100μm; for the inclusion defect, the minimum volume of the inclusion is ≥0.1mm 3 The interface defect is quantitatively detected, the interface separation width of the region is ≥30μm, and the position positioning error of all defects is ≤1mm. The evaluation standard is generated to generate a standardized detection report.
[0050] Step six: production system data interaction and quality closed-loop control
[0051] The enterprise production execution system stores the detection data and report, associates the database, realizes quality traceability, and the programmable logic controller automatically executes the control instruction according to the defect determination result. If the number of defects exceeds the preset threshold, the pressing pressure of the subsequent electrode is automatically increased by 5%-10%, or the baking holding time is extended by 1-2h, the production process is optimized, and if the defect size exceeds the critical value (such as crack length ≥5mm), the conveying track linkage instruction is triggered, and the unqualified electrode is automatically transported to the unqualified product storage area to avoid flowing into the next process.
[0052] The above describes the present application and its embodiments, which are not limited. In general, if a person skilled in the art is inspired by the above, without departing from the spirit of the present application, similar structures and embodiments can be designed without creativity, and should be within the protection scope of the present application.
Claims
1. A non-destructive testing method of a graphite electrode manufacturing process, characterized by, The method comprises the following steps: S1: surface cleaning of the graphite electrode to be detected, and positioning and fixing the graphite electrode on the conveying track of the detection platform; S2: three-dimensional scanning imaging of the graphite electrode to obtain original data of the internal macrostructure and density distribution of the electrode; S3: full-area scanning of the graphite electrode by using a phased array ultrasonic probe, and synchronous acquisition of ultrasonic echo signals; S4: fusion of the two types of data, construction of an intelligent defect recognition and classification model based on artificial intelligence, analysis of the fused data, recognition of the defect type, and preliminary determination of the defect severity; S5: quantitative evaluation of the size, position, number, and severity of the defects, and generation of a detection report; S6: transmission of the detection data to an enterprise production execution system and a programmable logic controller to realize traceability, or trigger a rejection instruction for unqualified products, and complete the quality closed-loop control of the graphite electrode manufacturing process.
2. A non-destructive testing method of a graphite electrode manufacturing process according to claim 1, characterized by: The three-dimensional scanning imaging in S2 comprises using a microfocus or small-focus X-ray source and a spliced detector, and the processing of the original data comprises an image reconstruction algorithm, an artifact correction algorithm, and a defect enhancement algorithm to obtain preliminary imaging results of the macroscopic defects of the graphite electrode.
3. A non-destructive testing method of a graphite electrode manufacturing process according to claim 1, characterized by: The positioning and fixing in S1 is specifically performed by using an adjustable clamping mechanism on the detection platform to ensure that the coaxiality deviation between the electrode axis and the center line of the conveying track is ≤0.5 mm.
4. A non-destructive testing method of a graphite electrode manufacturing process according to claim 2, characterized by: The processing of the echo signals in S3 comprises adaptive filtering, beam synthesis optimization, and signal enhancement.
5. A non-destructive testing method of a graphite electrode manufacturing process according to claim 2, characterized by: The focus size of the microfocus or small-focus X-ray source is ≤50 μm, and the spliced detector adopts 2-4 groups of detector units for splicing, and the pixel size of a single group of detector units is ≤100 μm.
6. A non-destructive testing method of a graphite electrode manufacturing process according to claim 1, characterized by: The method for data fusion in S4 adopts a feature layer fusion algorithm, specifically extracting the density difference features of the data and the acoustic impedance difference features of the phased array ultrasonic data, and using a weighted fusion model to fuse the two types of features.
7. A non-destructive testing method of a graphite electrode manufacturing process according to claim 1, characterized by: The quantitative evaluation in S5 specifically includes: for crack defects, quantitatively detecting cracks with a minimum width ≥ 50 μm and a minimum length ≥ 200 μm; for pore defects, quantitatively detecting pores with a minimum diameter ≥ 100 μm; for inclusion defects, quantitatively detecting inclusions with a minimum volume ≥ 0.1 mm 3 ; for interface defects, quantitatively detecting areas with an interface separation width ≥ 30 μm, and the position positioning error of all defects ≤ 1 mm.
8. A non-destructive testing method of a graphite electrode manufacturing process according to claim 1, characterized by: The rejection of unqualified products in S6 is specifically as follows: when the size of a defect detected exceeds a critical value (such as a crack length ≥5 mm), the conveying track automatically transfers the electrode to a storage for unqualified products.
9. A non-destructive testing method of a process for manufacturing a graphite electrode according to claim 6, characterized by: The synchronous acquisition system is used to realize the time and space alignment of the detection data and the phased array ultrasonic detection data.
10. A non-destructive testing method of a graphite electrode manufacturing process according to claim 1, characterized by: The programmable logic controller system automatically adjusts the production process parameters according to the defect determination results, and when the number of defects detected exceeds a preset threshold, the programmable logic controller system increases the pressing pressure of the subsequent electrodes by 5%-10%, or prolongs the baking holding time by 1-2 h.