A method for on-line non-destructive testing of a large rolling mill drive shaft

Abstract

This invention discloses a method for online non-destructive testing of large rolling mill drive shafts, belonging to the field of drive shaft testing technology. The steps of this invention are as follows: S1: Preparation and operational condition linkage deployment before online testing; S2: Customized online testing of different parts of the roll mill without stopping the mill; S3: Real-time acquisition, anti-interference transmission, and rapid processing of test data; S4: Online qualitative and quantitative analysis of defects and real-time assessment of health status; S5: Output of test results and guidance on targeted maintenance strategies. This invention provides an online non-destructive testing method for large rolling mill drive shafts, focusing on the need for continuous mill operation and no disassembly of the drive shaft. It achieves full-coverage, real-time, and accurate detection of surface, near-surface, and internal defects of the drive shaft, accurately identifying various defects such as fatigue cracks and incomplete penetration. It enables real-time acquisition, transmission, processing, and imaging of test data, rapidly outputting defect results and drive shaft health status for early warning, reducing maintenance costs, and improving equipment operational reliability.

Description

Technical Field

[0001] This invention relates to the field of transmission shaft inspection technology, and more specifically, to a method for online non-destructive inspection of transmission shafts in large rolling mills. Background Technology

[0002] Drive shafts are core components for power transmission in large rolling mills, and their operational status directly determines the continuous production capacity, safety, and economic benefits of the rolling mill production line. With the continuous increase in steel industry production and the advancement of domestic production of drive shafts, such as... Figure 1 As shown, domestic rolling mill production lines frequently experience failures such as crosshead fracture of drive shaft universal joint (a), shear fracture of middle butt weld (b), brittle fracture of support bearing installation part (c), and internal crack propagation of shaft head (d). These failures can cause unplanned production line interruptions or even large-scale shutdowns, resulting in huge economic losses. They also pose safety hazards such as equipment damage and personnel casualties.

[0003] Current methods for inspecting large rolling mill drive shafts primarily involve offline disassembly and inspection, with mainstream techniques including fluorescent magnetic particle testing, penetrant testing, and conventional ultrasonic testing. These methods require disassembling the drive shaft and critical components such as crossheads and bearings only after the rolling mill has been completely shut down, presenting numerous technical and engineering bottlenecks.

[0004] 1. Testing and production are highly conflicting: Offline disassembly and testing takes up several days of production time, which seriously disrupts the production plan and causes a large loss of production capacity;

[0005] 2. Insufficient detection effectiveness: Fluorescent magnetic particle and penetrant testing can only detect surface or near-surface (1-2mm) defects and cannot identify internally initiated fatigue cracks; conventional ultrasonic testing has a low signal-to-noise ratio and poor accuracy in defect localization and quantification, which easily leads to missed detections and misjudgments.

[0006] 3. Lack of early warning capability: Offline detection has a fixed cycle, and micro-cracks inside the drive shaft can easily expand rapidly within the detection interval, causing sudden failures;

[0007] 4. Existing online monitoring technologies for a limited number of drive shafts only indirectly monitor operating parameters such as torque, vibration, and temperature. They cannot directly identify the location, size, or type of defects, resulting in extremely poor detection specificity and accuracy, which is insufficient to meet actual production needs.

[0008] Currently, advanced non-destructive testing (NDT) technologies such as ultrasonic phased array (PAUT), arrayed eddy current (ECA), synthetic aperture focusing (SAFT), full matrix acquisition-full focusing (FMC-TFM), and plane wave imaging-full focusing (PWI-TFM) have been initially applied in the industrial field. However, these technologies are mostly limited to offline component inspection, and a dedicated online NDT solution for in-service drive shafts of large rolling mills has not yet been developed. Significant technical shortcomings exist in areas such as probe condition adaptation, non-contact signal acquisition, anti-interference in complex environments, and real-time data processing, failing to meet the core inspection requirements of rolling mills for continuous operation, non-disassembly, high precision, and real-time testing. Therefore, developing an online NDT method for large rolling mill drive shafts adapted to the complex operating conditions of rolling mills has become an urgent technical problem to be solved in the industry.

[0009] In existing technologies, patent CN118883097B uses an electromagnetic coil array to collect magnetic field data and combines it with a quantum convolutional neural network. This method only detects the geometric parameters of small, low-load automotive driveshafts and can only identify morphological changes, but cannot detect internal cracks or welding defects. The electromagnetic excitation has limited penetration and is difficult to adapt to the complex working conditions of thick-walled, high-load driveshafts in large rolling mills. Patent CN120720338A relies on temperature and acceleration sensors to monitor vibration and temperature. It can only provide early warnings of macroscopic operational faults in the driveshaft and cannot identify early defects such as microcracks and internal inclusions. The sensor signals are easily affected by environmental interference, resulting in insufficient detection accuracy. This method is not suitable for the high-temperature and high-vibration working conditions of rolling mills. This solution uses a fusion detection of ultrasonic phased array and electromagnetic sensors, which can accurately quantify defects. Patent CN119880416B relies on magnetorheological fluid to adjust stiffness and conduct accelerated fatigue tests in the laboratory. This is a destructive test and can only assess the overall service life, but cannot locate defects. It cannot be used for in-service component testing, while this invention is suitable for on-site non-destructive testing of rolling mill driveshafts. Summary of the Invention

[0010] 1. The technical problem that the invention aims to solve

[0011] To address the shortcomings and deficiencies of existing technologies, this invention provides a method for online non-destructive testing of large rolling mill drive shafts. This invention enables direct online non-destructive testing of large rolling mill drive shafts without shutting down or disassembling the machine, completely overcoming the technical limitations of traditional offline disassembly testing. It is suitable for complex working conditions on the rolling mill site, and the testing accuracy, efficiency, and stability all meet the actual needs of industrial production, achieving real-time, accurate, and comprehensive detection of drive shaft defects.

[0012] 2. Technical Solution

[0013] To achieve the above objectives, the technical solution provided by the present invention is as follows:

[0014] The present invention provides a method for online non-destructive testing of a large rolling mill drive shaft, comprising the following steps:

[0015] S1: Preparations and Deployment in Sync with Online Testing

[0016] S1.1: Simple cleaning of the inspection area: During the normal operation interval of the rolling mill, use a high-pressure air gun to perform dry simple cleaning of the key inspection parts, only blowing away surface dust, oil stains and iron filings. No grinding or disassembly is required. The original protective layer and assembly structure of the drive shaft are retained. Production is resumed immediately after cleaning.

[0017] S1.2: Deployment of online inspection equipment: Based on the on-site installation layout of the rolling mill drive shaft, a movable industrial-grade protective bracket is fixed around the drive shaft. The customized online inspection probe and multi-channel high-speed data acquisition module are installed on the bracket. The wireless data transmission module and on-site real-time data processing terminal are deployed in the rolling mill control room to ensure physical isolation between the inspection equipment and the rolling mill site.

[0018] S1.3: Real-time linkage of operating parameters: The online detection system is connected to the rolling mill PLC control system to synchronize the in-service operating parameters of the drive shaft in real time, including speed, torque, rolling load, running speed, etc. The detection system can automatically adjust the probe scanning frequency and data acquisition frame rate according to the operating parameters to achieve dynamic adaptation between the detection and the in-service operating conditions of the drive shaft.

[0019] S2: Customized online inspection of specific parts of the rolling mill while it is not shut down:

[0020] Based on the structural characteristics, defect types, and in-service rotational state of different parts of the drive shaft, advanced non-destructive testing technologies such as arrayed eddy current (ECA), ultrasonic phased array (PAUT), SAFT (synthetic aperture focusing technology), FMC-TFM (full matrix acquisition-full focusing technology), and PWI-TFM (plane wave imaging-focusing technology) are integrated to design an online dedicated testing probe and dynamic scanning strategy, which utilizes the natural rotation of the drive shaft to achieve 360° full circumferential scanning;

[0021] S3: Real-time acquisition, interference-resistant transmission, and rapid processing of detection data

[0022] S3.1: Real-time high-speed data acquisition: Through a multi-channel industrial-grade high-speed data acquisition module, the detection signals of array eddy currents and ultrasonic phased arrays of various parts are acquired synchronously;

[0023] S3.2: Wireless anti-interference transmission: Using LoRa+5G dual wireless transmission protocols, the collected detection data is transmitted to the field data processing terminal in the control room in real time.

[0024] S3.3: Real-time data processing and imaging: The on-site data processing terminal has a built-in lightweight real-time processing algorithm. First, the rainflow method is used to reduce signal noise and extract defect features, eliminating interference signals from drive shaft rotation, mill vibration, and the on-site environment. Then, a multi-source data fusion algorithm is used to integrate the detection data from different detection technologies, and generate two-dimensional and three-dimensional imaging maps of defects in various parts in real time, intuitively displaying the location, shape, size and key parameters of the defects. The detection data processing delay is ≤5s.

[0025] S4: Online qualitative and quantitative analysis of defects and real-time assessment of health status:

[0026] By combining online synchronous in-service operating parameters and typical damage modes of the drive shaft, online qualitative and quantitative analysis of defects in real-time imaging atlases is performed, and real-time assessment of the health status of the drive shaft is completed.

[0027] S5: Output of test results and guidance on targeted maintenance strategies:

[0028] After the online inspection is completed, the system automatically generates a standardized online inspection report. The report includes detailed information on defects in various parts of the drive shaft, defect imaging maps, health status assessment results, working condition correlation analysis, and defect development trend prediction. At the same time, it provides targeted maintenance strategy suggestions in real time based on the defect type, size, and development trend.

[0029] Furthermore, in S1.1, the key inspection parts include, but are not limited to, the universal joint crosshead of the drive shaft, the shaft body, the diameter change point, the weld, and the shaft end.

[0030] Furthermore, in S1.2, the distance between the probe and the detection part of the transmission shaft is adjusted to ≤5mm to achieve non-contact, near-contact precise alignment.

[0031] Furthermore, S2 specifically includes the following detection steps:

[0032] S2.1: Online Inspection of Universal Joint Crosshead (Part a): Using ultrasonic phased array (PAUT) combined with FMC-TFM technology, the shape and size of the crosshead at the universal joint of different specifications of drive shafts are analyzed to determine the ultrasonic phased array transducer arrangement method. Industrial-grade anti-vibration ultrasonic phased array probes are arranged on the end faces of the four crossheads, and the end faces are inspected using an S-shaped scanning method to detect linear damage defects such as delamination, holes, and cracks inside the crossheads. Crosshead samples with defects are prepared, and simulations and experiments are conducted to determine the time-delay focusing law of the phased array on complex curved surfaces. A linear dynamic multidimensional imaging method was established using full-focus imaging to achieve full-coverage detection of crosshead defects. Based on the time-delay focusing law of complex curved surface phased array, combined with SAFT algorithm to optimize the echo signal, and using the rotation of the drive shaft to achieve dynamic circumferential scanning, full-focus imaging detection of delamination, holes, and cracks within a depth range of 500mm inside the crosshead was achieved. By combining physical parallel ultrasonic phased array focusing method with synthetic sequential focusing method, the amplitude and phase nonlinear changes caused by microcracks were captured, achieving online early warning of microcracks with a depth ≥0.2mm.

[0033] S2.2: Online inspection of drive shaft (part b): The "array eddy current + ultrasonic phased array" dual technology fusion scheme is adopted to achieve full coverage inspection of surface / near surface and internal defects;

[0034] S2.3: Online Inspection of Shaft Diameter Change and Welds: The shaft diameter change area is a stress concentration zone. A customized flexible array electromagnetic online probe is used, which is arranged in close contact with the curved surface structure of the diameter change area. The rotation of the drive shaft is used to achieve circumferential online scanning, and the surface and near-surface cracks at the diameter change area are detected in real time, so as to visualize the inspection results. For the weld and heat-affected zone, an ultrasonic phased array oblique-entry online probe is used, which is arranged on the outside of the weld. The full coverage of the weld area is achieved by dynamically adjusting the sound beam angle. Combined with PWI-TFM technology, the online inspection efficiency is improved, and welding defects such as incomplete fusion, incomplete penetration, inclusions, and porosity are identified in real time.

[0035] S2.4: Online inspection of shaft head: A focused ultrasonic phased array online probe is used, which is placed in close contact with the shaft surface on the outside of the shaft sleeve. By controlling the excitation time of the probe, the focus point of the sound wave is dynamically adjusted to accurately match the rotation state of the drive shaft. The internal structure of the shaft head is dynamically scanned online, and defects such as hidden cracks and fractures inside the shaft head are identified in real time, solving the problem of difficult detection of internal defects of the shaft head.

[0036] Furthermore, S2.2 specifically includes the following detection steps:

[0037] S2.2.1: Surface and Near-Surface Defect Detection: Considering the structural characteristics of shaft wall thickness and curvature, as well as the current situation of surface oil and rust with the basic protective layer, after screening and optimization of detection methods such as magnetic flux leakage, eddy current, and magnetic disturbance, a differential fitting-based electromagnetic disturbance detection probe is used in combination with a circumferential array eddy current online detection probe to perform shaft detection. During detection, the circumferential array electromagnetic probe is non-contactly wrapped around the outside of the shaft. Combined with the axial scanning strategy of the array probe, 360° full circumferential online scanning is achieved by taking advantage of the rotation of the drive shaft itself. Relying on the differential fitting electromagnetic disturbance detection technology, the influence of surface rust, impurities, and probe lifting caused by drive shaft rotation can be effectively offset, reducing signal interference and improving the consistency and reliability of detection signals. Electromagnetic signals of surface and near-surface cracks, local corrosion, and other defects can be collected in real time without grinding the shaft surface. After analysis and processing, the collected data completes the two-dimensional imaging detection of shaft defects.

[0038] S2.2.2: Internal Defect Detection: A high-speed scanning ultrasonic phased array probe is used in conjunction with PWI-TFM technology. The rotation of the drive shaft is used to achieve dynamic scanning of the shaft's interior. Taking advantage of the rapid plane wave imaging of PWI-TFM, the rotational speed of the drive shaft is matched to achieve high-precision real-time imaging detection of internal delamination, voids, and buried cracks. Finally, the detection data of the array eddy current and ultrasonic phased array are fused to form an online full-coverage evaluation result of shaft defects.

[0039] Furthermore, in S3.1, the data acquisition frame rate is linked to the rotational speed of the drive shaft; the higher the rotational speed, the higher the acquisition frame rate, ensuring the integrity and continuity of the detection data during the rotation of the drive shaft.

[0040] Furthermore, S4 specifically includes the following steps:

[0041] S4.1: Defect qualitative judgment: Based on the defect imaging characteristics and signal amplitude changes, combined with the service conditions of the drive shaft, determine the defect type and its cause;

[0042] S4.2: Quantitative Defect Judgment: Based on the quantitative advantages of SAFT and FMC-TFM technologies, the depth, length, and width of defects are accurately measured, ensuring that the positioning error is ≤ ±0.3mm and the quantitative error is ≤ ±10%.

[0043] S4.3: Health Status Assessment: Based on the defect detection results and combined with the design requirements and service standards of the mill drive shaft, the health status of the drive shaft is determined in real time and divided into four levels: qualified, requiring shutdown for maintenance, and requiring emergency replacement. The assessment results are pushed to the mill control room terminal in real time and displayed visually.

[0044] Furthermore, S5 specifically includes the following steps:

[0045] S5.1: For drive shafts in "qualified" condition, record the test data, include them in the equipment health record, and conduct online re-inspection according to the regular cycle;

[0046] S5.2: For drive shafts in the "to be observed" state, set up online continuous monitoring reminders, and the system will automatically start re-inspection periodically to track defect changes in real time;

[0047] S5.3: For drive shafts that require "shutdown for maintenance", provide suggestions on the optimal maintenance window period in conjunction with the rolling mill production plan to reduce the impact of maintenance on production;

[0048] S5.4: For drive shafts in "emergency replacement" status, the system will immediately issue an audible and visual alarm to remind on-site operators to take emergency measures to prevent safety accidents from occurring.

[0049] 3. Beneficial effects

[0050] Compared with the prior art, the technical solution provided by this invention has the following advantages:

[0051] The innovative point of this invention is:

[0052] 1. Pioneering online non-destructive testing solution for large rolling mill drive shafts without stopping or disassembling the machine: This solution completely breaks through the technical limitations of traditional offline disassembly testing. It eliminates the need to disassemble any parts or stop production, enabling testing and production to proceed simultaneously. This fundamentally solves the core industry pain point of conflict between testing and production, and significantly improves the engineering practicality of testing.

[0053] 2. Customized R&D of online detection probes and scanning strategies for rolling mills: Design of vibration-proof, oil-proof, and dust-proof non-contact / near-contact probes (flexible array electromagnetic probes, vibration-proof ultrasonic phased array probes), combined with the in-service rotation state of the drive shaft, to design a dynamic circumferential scanning strategy, utilizing the natural rotation of the drive shaft to achieve 360° full coverage detection, perfectly adapting to the complex on-site working conditions of rolling mills with high vibration, strong electromagnetic fields, and heavy oil contamination;

[0054] 3. Achieve online fusion and real-time data processing of multiple advanced non-destructive testing technologies: Integrate array eddy current, ultrasonic phased array, SAFT, FMC-TFM, and PWI-TFM technologies, develop lightweight online real-time data processing algorithms, realize synchronous acquisition of test data, wireless anti-interference transmission, real-time imaging and analysis, and solve the technical problems of data processing lag and multiple interference signals in online testing;

[0055] 4. Establish an online health status assessment system that links detection with operating conditions: Link online detection data with the operating condition parameters of the drive shaft in the rolling mill PLC system in real time, and combine the in-service damage mode of the drive shaft to conduct qualitative and quantitative analysis of defects and graded assessment of health status, so as to realize the integration of "detection-operating condition-assessment" and provide accurate basis for the rolling mill production line to formulate preventive maintenance strategies based on equipment status.

[0056] 5. Achieve online early warning and dynamic tracking monitoring of microcracks: By combining physical parallel focusing and synthetic sequential focusing methods, the nonlinear acoustic response characteristics of microcracks are captured. The SAFT algorithm is integrated to improve quantification accuracy. It can identify early microcracks with a depth ≥0.2mm online and support online continuous dynamic tracking monitoring of defects, realizing the transformation of equipment operation and maintenance mode from "passive maintenance" to "proactive prevention". Attached Figure Description

[0057] Figure 1 This is a structural diagram of the rolling mill drive shaft and its failure location according to the present invention;

[0058] Figure 2 This is a schematic diagram of the crosshead detection method of the present invention;

[0059] Figure 3 This is a diagram illustrating the implementation effect of the array eddy current detection method of the present invention.

[0060] Figure 4 This is a schematic diagram illustrating the ultrasonic testing principle of the weld seam at the diameter change point of the present invention.

[0061] Figure 5 This is a schematic diagram of the internal fracture detection of the shaft head according to the present invention. Detailed Implementation

[0062] The present invention will be further described below with reference to the accompanying drawings and embodiments:

[0063] Example 1

[0064] from Figure 1 As can be seen, the steps of the online non-destructive testing method for a large rolling mill drive shaft in this embodiment are as follows:

[0065] S1: Preparations and Deployment in Sync with Online Testing

[0066] S1.1: Simple cleaning of inspection area: During the normal operation interval of the rolling mill, use a high-pressure air gun to perform dry simple cleaning of key inspection parts. Key inspection parts include, but are not limited to, universal joint crosshead of drive shaft, shaft body, diameter change, weld, shaft head, etc. Only blow away surface dust, oil stains, and iron filings. No grinding or disassembly is required. The original protective layer and assembly structure of the drive shaft are retained. Production is resumed immediately after cleaning without occupying effective production time.

[0067] S1.2: Deployment of online inspection equipment: Based on the on-site installation layout of the rolling mill drive shaft, a movable industrial-grade protective bracket is fixed around the drive shaft. The distance between the probe and the detection part of the drive shaft is adjusted to ≤5mm to achieve non-contact or near-contact precise alignment. The customized online inspection probe and multi-channel high-speed data acquisition module are installed on the bracket. The wireless data transmission module and on-site real-time data processing terminal are deployed in the rolling mill control room to ensure physical isolation between the inspection equipment and the rolling mill site and avoid interference from on-site working conditions.

[0068] S1.3: Real-time linkage of operating parameters: The online detection system is connected to the rolling mill PLC control system to synchronize the in-service operating parameters of the drive shaft in real time, including speed, torque, rolling load, running speed, etc. The detection system can automatically adjust the probe scanning frequency and data acquisition frame rate according to the operating parameters to achieve dynamic adaptation between the detection and the in-service operating conditions of the drive shaft, and ensure the integrity of the detection data under different operating conditions.

[0069] S2: Customized online inspection of specific parts of the rolling mill while it is not shut down:

[0070] Based on the structural characteristics, defect types, and in-service rotational state of different parts of the drive shaft, advanced non-destructive testing technologies such as arrayed eddy current (ECA), ultrasonic phased array (PAUT), SAFT (synthetic aperture focusing technology), FMC-TFM (full matrix acquisition-full focusing technology), and PWI-TFM (plane wave imaging-focusing technology) are integrated to design an online dedicated inspection probe and dynamic scanning strategy. The drive shaft's natural rotation is used to achieve 360° full circumferential scanning, and the inspection is completed during the normal operation of the rolling mill.

[0071] like Figure 2 As shown, S2 specifically includes the following detection steps:

[0072] S2.1: Online Inspection of Universal Joint Crosshead (Part a): Using ultrasonic phased array (PAUT) combined with FMC-TFM technology, the shape and size of the crosshead at the universal joint of different specifications of drive shafts are analyzed to determine the ultrasonic phased array transducer arrangement method. Industrial-grade anti-vibration ultrasonic phased array probes are arranged on the end faces of the four crossheads, and the end faces are inspected using an S-shaped scanning method to detect linear damage defects such as delamination, holes, and cracks inside the crossheads. Crosshead samples with defects are prepared, and simulations and experiments are conducted to determine the time-delay focusing law of the phased array on complex curved surfaces. A linear dynamic multidimensional imaging method was established using full-focus imaging to achieve full-coverage detection of crosshead defects. Based on the time-delay focusing law of complex curved surface phased array, combined with SAFT algorithm to optimize the echo signal, and using the rotation of the drive shaft to achieve dynamic circumferential scanning, full-focus imaging detection of delamination, holes, and cracks within a depth range of 500mm inside the crosshead was achieved. By combining physical parallel ultrasonic phased array focusing method with synthetic sequential focusing method, the amplitude and phase nonlinear changes caused by microcracks were captured, achieving online early warning of microcracks with a depth ≥0.2mm.

[0073] S2.2: Online inspection of drive shaft (part b): The "array eddy current + ultrasonic phased array" dual technology fusion scheme is adopted to achieve full coverage inspection of surface / near surface and internal defects;

[0074] S2.2 also includes the following specific detection steps:

[0075] S2.2.1: Surface and near-surface defect detection: Considering the structural characteristics of the shaft wall thickness and curvature, as well as the current state of surface oil and rust adhesion and the retention of the basic protective layer, detection methods such as magnetic flux leakage, eddy current, and magnetic disturbance are screened and optimized. Figure 3 As shown, a differential fitting-based electromagnetic disturbance detection probe is used in conjunction with a circumferential array eddy current online detection probe to perform shaft inspection. During inspection, the circumferential array electromagnetic probe is non-contactly placed around the outside of the shaft. Combined with the axial scanning strategy of the array probe, 360° full circumferential online scanning is achieved by utilizing the rotation of the drive shaft itself, which greatly improves the inspection efficiency and significantly shortens the inspection period. Relying on the differential fitting electromagnetic disturbance detection technology, the influence of surface corrosion, impurities, and probe lifting caused by drive shaft rotation can be effectively offset, reducing signal interference and improving the consistency and reliability of the detection signal. Electromagnetic signals of surface and near-surface cracks, local corrosion, and other defects can be collected in real time without grinding the shaft surface. After analysis and processing, the collected data completes the two-dimensional imaging detection of shaft defects.

[0076] S2.2.2: Internal Defect Detection: A high-speed scanning ultrasonic phased array probe is used in conjunction with PWI-TFM technology. The rotation of the drive shaft is used to achieve dynamic scanning of the shaft's interior. Taking advantage of the rapid plane wave imaging of PWI-TFM, the rotational speed of the drive shaft is matched to achieve high-precision real-time imaging detection of internal delamination, voids, and buried cracks. Finally, the detection data of the array eddy current and ultrasonic phased array are fused to form an online full-coverage evaluation result of shaft defects.

[0077] S2.3: Online inspection of shaft diameter change points and welds: Shaft diameter change points are stress concentration areas, such as... Figure 4 As shown, a customized flexible array electromagnetic online probe is used, which is arranged in close contact with the curved surface structure at the diameter change point. The rotation of the drive shaft realizes circumferential online scanning, and the surface and near-surface cracks at the diameter change point are detected in real time, so as to visualize the detection results. The weld and heat-affected zone are equipped with an ultrasonic phased array oblique online probe, which is arranged on the outside of the weld. The full coverage of the weld area is achieved by dynamically adjusting the sound beam angle. Combined with PWI-TFM technology, the online detection efficiency is improved, and welding defects such as incomplete fusion, incomplete penetration, inclusions, and porosity are identified in real time.

[0078] S2.4: Online Inspection of Shaft Head: A focused ultrasonic phased array online probe is used, positioned close to the outer surface of the shaft body. By controlling the probe's excitation time, the focusing point of the sound waves is dynamically adjusted to precisely match the rotational state of the drive shaft, enabling dynamic online scanning of the shaft head's interior. Figure 5 As shown, it can identify internal defects such as hidden cracks and fractures in the shaft head in real time, solving the problem of difficult detection of internal defects in the shaft head.

[0079] S3: Real-time acquisition, interference-resistant transmission, and rapid processing of detection data

[0080] S3.1: Real-time high-speed data acquisition: Through a multi-channel industrial-grade high-speed data acquisition module, the detection signals of array eddy currents and ultrasonic phased arrays of various parts are acquired synchronously; the acquisition frame rate is linked to the speed of the drive shaft, and the higher the speed, the higher the acquisition frame rate, ensuring the integrity and continuity of the detection data during the rotation of the drive shaft;

[0081] S3.2: Wireless anti-interference transmission: The LoRa+5G dual wireless transmission protocol is adopted to transmit the collected detection data to the field data processing terminal in the control room in real time. This transmission scheme has the ability to resist strong electromagnetic and high vibration interference in the rolling mill, ensuring the stability and real-time performance of data transmission. Moreover, the data is transmitted in a closed loop throughout the plant area, effectively ensuring the security of production data.

[0082] S3.3: Real-time data processing and imaging: The on-site data processing terminal has a built-in lightweight real-time processing algorithm. First, the rainflow method is used to reduce signal noise and extract defect features, eliminating interference signals from drive shaft rotation, mill vibration, and the on-site environment. Then, a multi-source data fusion algorithm is used to integrate the detection data from different detection technologies, and generate two-dimensional and three-dimensional imaging maps of defects in various parts in real time, intuitively displaying the location, shape, size and key parameters of the defects. The detection data processing delay is ≤5s.

[0083] S4: Online qualitative and quantitative analysis of defects and real-time assessment of health status:

[0084] By combining online synchronous in-service operating parameters of the drive shaft with typical damage modes (fatigue, wear, overload), online qualitative and quantitative analysis of defects in real-time imaging atlases is performed, and real-time assessment of the health status of the drive shaft is completed.

[0085] S4 specifically includes the following steps:

[0086] S4.1: Defect qualitative judgment: Based on the defect imaging characteristics and signal amplitude changes, combined with the service conditions of the drive shaft, determine the defect type (cracks, inclusions, incomplete penetration, corrosion, porosity, etc.) and its cause (material manufacturing defects, welding process problems, fatigue load, overload impact, etc.).

[0087] S4.2: Quantitative Defect Judgment: Based on the quantitative advantages of SAFT and FMC-TFM technologies, the depth, length, and width of defects are accurately measured, ensuring that the positioning error is ≤ ±0.3mm and the quantitative error is ≤ ±10%.

[0088] S4.3: Health Status Assessment: Based on the defect detection results and combined with the design requirements and service standards of the mill drive shaft, the health status of the drive shaft is determined in real time and divided into four levels: qualified (no defects or defects within the allowable range), under observation (minor defects exist, with no risk of rapid expansion), requiring shutdown for maintenance (defects exceed the allowable range, requiring shutdown for maintenance in conjunction with the production plan), and emergency replacement (serious defects, with a risk of sudden breakage). The assessment results are pushed to the mill control room terminal in real time and displayed visually.

[0089] S5: Output of test results and guidance on targeted maintenance strategies:

[0090] After the online inspection is completed, the system automatically generates a standardized online inspection report. The report includes detailed information on defects in various parts of the drive shaft, defect imaging maps, health status assessment results, working condition correlation analysis, and defect development trend prediction. At the same time, it provides targeted maintenance strategy suggestions in real time based on the defect type, size, and development trend.

[0091] S5 specifically includes the following steps:

[0092] S5.1: For drive shafts in "qualified" condition, record the test data, include them in the equipment health record, and conduct online re-inspection according to the regular cycle;

[0093] S5.2: For drive shafts in the "to be observed" state, set up online continuous monitoring reminders, and the system will automatically start re-inspection periodically to track defect changes in real time;

[0094] S5.3: For drive shafts that require "shutdown for maintenance", provide suggestions on the optimal maintenance window period in conjunction with the rolling mill production plan to reduce the impact of maintenance on production;

[0095] S5.4: For drive shafts in "emergency replacement" status, the system will immediately issue an audible and visual alarm to remind on-site operators to take emergency measures to prevent safety accidents from occurring.

[0096] This invention adopts a core technical approach of "online non-contact and near-contact detection + customized probes for different parts + fusion of multiple advanced technologies + real-time data processing + working condition linkage adaptation". The entire detection process is completed while the rolling mill is running normally and the drive shaft is continuously rotating, without the need for disassembly or production shutdown. The specific steps are as follows:

[0097] All testing processes in this invention strictly follow national standards such as NB / T47013.15-2021 "Non-destructive testing of pressure equipment - Part 15: Phased array ultrasonic testing" and NB / T 47013.6-2015 "Non-destructive testing of pressure equipment - Part 6: Eddy current testing". All testing equipment has IP67 or higher industrial-grade protection capabilities and can withstand the complex working conditions of high temperature, oil, dust, vibration and other factors at the rolling mill site.

[0098] This invention provides a method for online non-destructive testing of large rolling mill drive shafts. The core of this method is to meet the online testing requirements of rolling mill production lines without shutting down the mill and without disassembling the drive shaft, thereby achieving real-time, accurate, and comprehensive testing of drive shaft defects.

[0099] This invention enables direct, non-destructive online testing of large rolling mill drive shafts without shutting down or disassembling the mill. It completely overcomes the limitations of traditional offline disassembly testing, is adaptable to complex rolling mill conditions, and achieves the required accuracy, efficiency, and stability for industrial production. The core implementation results are as follows:

[0100] 1. Testing and production are completely synchronized, with no capacity loss: No need to disassemble any parts of the drive shaft, no need to stop the rolling mill, all position testing can be completed under normal rolling mill operation. The online testing time of a single drive shaft is reduced by more than 90% compared with the traditional offline disassembly testing, completely eliminating the production loss caused by testing and realizing "testing does not affect production".

[0101] 2. Comprehensive defect detection with high accuracy and detection rate: Through the integration of multiple technologies and customized detection solutions for different parts, it achieves comprehensive online detection of surface, near-surface, and internal defects of the drive shaft, accurately identifying various defects such as fatigue cracks, incomplete penetration, inclusions, and corrosion, with a defect detection rate ≥95% and a false alarm rate ≤10%; it can accurately identify internal microcracks with a depth ≥0.2mm, with a positioning error ≤±0.3mm and a quantitative error ≤±10%, which is far superior to traditional detection methods;

[0102] 3. Strong adaptability to working conditions and excellent anti-interference ability: It adopts non-contact and near-contact detection methods, and is equipped with a vibration-proof, oil-proof, and dust-proof online dedicated probe. Combined with LoRa+5G dual wireless transmission protocols, it can effectively resist interference from strong electromagnetic fields, high vibration, oil stains, dust and other factors in the rolling mill. It can stably and reliably detect within the working conditions of 0-500r / min drive shaft speed and 0-500kN rolling load, ensuring the stability and reliability of the detection process.

[0103] 4. Real-time data processing for early defect warning: The entire process of data acquisition, transmission, processing, and imaging has a delay of ≤5s, enabling rapid identification and early warning in the early stages of defect occurrence. This transforms the process from "offline periodic inspection" to "online real-time monitoring" and from "fault repair" to "preventive maintenance," achieving early defect warning and effectively preventing sudden drive shaft breakage caused by rapid defect expansion.

[0104] 5. Simple operation, enabling automated unmanned inspection: After deployment, the inspection equipment can be deeply integrated with the rolling mill PLC system to achieve unmanned, automated online inspection. No professional inspection personnel are required for on-site operation, reducing labor costs and lowering the operational threshold. Simultaneously, inspection data is automatically incorporated into the equipment health record, enabling full lifecycle data management of the drive shaft. This establishes an integrated process of "online inspection - data processing - status assessment - maintenance guidance," providing a scientific basis for developing preventative maintenance strategies based on equipment status for the rolling mill production line, reducing maintenance costs, and improving equipment operational reliability.

[0105] Example 2

[0106] Example 2: Online Inspection of Drive Shafts in Hot Strip Mill Finishing Units

[0107] This embodiment is designed for the operating conditions of hot strip mills, which are characterized by high temperature (above 80°C), oxide scale splashing, and high rotational speed (100~300 rpm).

[0108] Before testing, a high-pressure air gun is used to simply clean the surface dust and oxide scale during the roll changing interval. Water-cooled protective supports are installed on both sides of the mill, and the distance between the high-temperature resistant probe (withstanding 150℃) and the drive shaft is adjusted to 3mm. The testing system is linked with the PLC, automatically adjusting the acquisition frame rate according to the rotation speed, and increasing the sampling rate as the rotation speed increases to ensure data integrity.

[0109] During inspection, the crosshead employs a high-temperature resistant ultrasonic phased array probe combined with FMC-TFM technology to detect internal fatigue cracks up to 500mm deep. The shaft utilizes a dual-technology fusion of "arrayed eddy current + ultrasonic phased array," with eddy current detecting surface thermal fatigue cracks and PWI-TFM ultrasonic imaging achieving rapid imaging of internal defects. Flexible eddy current probes are used to inspect curved surfaces at diameter changes. All data is wirelessly transmitted to the control room, where it is imaged in real-time after noise reduction using the rainflow method. The defect quantification error is ≤±10%, and the health status is classified into four levels with maintenance recommendations provided.

[0110] Example 3

[0111] Online inspection of drive shafts for large thick plate rolling mills

[0112] This embodiment is designed for thick plate rolling mills with extremely high rolling force (>100MN), large shaft diameter (>800mm), low rotational speed (10~60rpm), and the main failure mode being internal fatigue cracking.

[0113] Before testing, coarse oxide scale is removed by high-pressure water rinsing and a steel brush. A gantry-type moving support is used to mount the probe, with a 5mm spacing to accommodate large vibrations. The system reads rolling force and torque; at low speeds, the data density is ensured by extending the single acquisition time.

[0114] The focus of the inspection is on internal defects: a low-frequency ultrasonic phased array (2.5MHz) combined with FMC-TFM is used, with a penetration depth of over 300mm, focusing on detecting buried cracks in the core; flexible phased array probes are used for close scanning in stress concentration areas at diameter changes; focused ultrasound is used to detect the risk of fracture at the joint of the shaft head. Data processing emphasizes sound velocity compensation and attenuation correction under large thicknesses, and health assessment is combined with fatigue life curves. For internal cracks ≥5mm, it is recommended to arrange shutdown and maintenance.

[0115] The present invention and its embodiments have been described above illustratively. This description is not restrictive, and the figures shown are only one embodiment of the present invention; the actual structure is not limited thereto. Therefore, if those skilled in the art are inspired by this description and design similar structures and embodiments without departing from the spirit of the present invention, such designs should fall within the protection scope of the present invention.

Claims

1. A method for online non-destructive testing of a large rolling mill drive shaft, characterized in that: The steps are as follows: S1: Preparations and Deployment in Sync with Online Testing S1.1: Simple cleaning of the inspection area: During the normal operation interval of the rolling mill, use a high-pressure air gun to perform dry simple cleaning of the key inspection parts, only blowing away surface dust, oil stains and iron filings. No grinding or disassembly is required. The original protective layer and assembly structure of the drive shaft are retained. Production is resumed immediately after cleaning. S1.2: Deployment of online inspection equipment: Based on the on-site installation layout of the rolling mill drive shaft, a movable industrial-grade protective bracket is fixed around the drive shaft. The customized online inspection probe and multi-channel high-speed data acquisition module are installed on the bracket. The wireless data transmission module and on-site real-time data processing terminal are deployed in the rolling mill control room to ensure physical isolation between the inspection equipment and the rolling mill site. S1.3: Real-time linkage of operating parameters: The online detection system is connected to the rolling mill PLC control system to synchronize the in-service operating parameters of the drive shaft in real time, including speed, torque, rolling load, running speed, etc. The detection system can automatically adjust the probe scanning frequency and data acquisition frame rate according to the operating parameters to achieve dynamic adaptation between the detection and the in-service operating conditions of the drive shaft. S2: Customized online inspection of specific parts of the rolling mill while it is not shut down: Based on the structural characteristics, defect types, and in-service rotational state of different parts of the drive shaft, advanced non-destructive testing technologies such as arrayed eddy current (ECA), ultrasonic phased array (PAUT), SAFT (synthetic aperture focusing technology), FMC-TFM (full matrix acquisition-full focusing technology), and PWI-TFM (plane wave imaging-focusing technology) are integrated to design an online dedicated testing probe and dynamic scanning strategy, which utilizes the natural rotation of the drive shaft to achieve 360° full circumferential scanning; S3: Real-time acquisition, interference-resistant transmission, and rapid processing of detection data S3.1: Real-time high-speed data acquisition: Through a multi-channel industrial-grade high-speed data acquisition module, the detection signals of array eddy currents and ultrasonic phased arrays of various parts are acquired synchronously; S3.2: Wireless anti-interference transmission: Using LoRa+5G dual wireless transmission protocols, the collected detection data is transmitted to the field data processing terminal in the control room in real time. S3.3: Real-time data processing and imaging: The on-site data processing terminal has a built-in lightweight real-time processing algorithm. First, the rainflow method is used to reduce signal noise and extract defect features, eliminating interference signals from drive shaft rotation, mill vibration, and the on-site environment. Then, a multi-source data fusion algorithm is used to integrate the detection data from different detection technologies, and generate two-dimensional and three-dimensional imaging maps of defects in various parts in real time, intuitively displaying the location, shape, size and key parameters of the defects. The detection data processing delay is ≤5s. S4: Online qualitative and quantitative analysis of defects and real-time assessment of health status: By combining online synchronous in-service operating parameters and typical damage modes of the drive shaft, online qualitative and quantitative analysis of defects in real-time imaging atlases is performed, and real-time assessment of the health status of the drive shaft is completed. S5: Output of test results and guidance on targeted maintenance strategies: After the online inspection is completed, the system automatically generates a standardized online inspection report. The report includes detailed information on defects in various parts of the drive shaft, defect imaging maps, health status assessment results, working condition correlation analysis, and defect development trend prediction. At the same time, it provides targeted maintenance strategy suggestions in real time based on the defect type, size, and development trend.

2. The method for online non-destructive testing of a large rolling mill drive shaft according to claim 1, characterized in that: In S1.1, the key inspection areas include, but are not limited to, the universal joint crosshead of the drive shaft, the shaft body, the diameter change area, the weld, and the shaft head.

3. The method for online non-destructive testing of a large rolling mill drive shaft according to claim 1, characterized in that: In S1.2, the distance between the probe and the detection part of the transmission shaft is adjusted to ≤5mm to achieve non-contact or near-contact precise alignment.

4. The method for online non-destructive testing of a large rolling mill drive shaft according to claim 1, characterized in that: The S2 process specifically includes the following detection steps: S2.1: Online Inspection of Universal Joint Crosshead (Part a): Using ultrasonic phased array (PAUT) combined with FMC-TFM technology, the shape and size of the crosshead at the universal joint of different specifications of drive shafts are analyzed to determine the ultrasonic phased array transducer arrangement method. Industrial-grade anti-vibration ultrasonic phased array probes are arranged on the end faces of the four crossheads, and the end faces are inspected using an S-shaped scanning method to detect linear damage defects such as delamination, holes, and cracks inside the crossheads. Crosshead samples with defects are prepared, and simulations and experiments are conducted to determine the time-delay focusing law of the phased array on complex curved surfaces. A linear dynamic multidimensional imaging method was established using full-focus imaging to achieve full-coverage detection of crosshead defects. Based on the time-delay focusing law of complex curved surface phased array, combined with SAFT algorithm to optimize the echo signal, and using the rotation of the drive shaft to achieve dynamic circumferential scanning, full-focus imaging detection of delamination, holes, and cracks within a depth range of 500mm inside the crosshead was achieved. By combining physical parallel ultrasonic phased array focusing method with synthetic sequential focusing method, the amplitude and phase nonlinear changes caused by microcracks were captured, achieving online early warning of microcracks with a depth ≥0.2mm. S2.2: Online inspection of drive shaft (part b): The "array eddy current + ultrasonic phased array" dual technology fusion scheme is adopted to achieve full coverage inspection of surface / near surface and internal defects; S2.3: Online Inspection of Shaft Diameter Change and Welds: The shaft diameter change area is a stress concentration zone. A customized flexible array electromagnetic online probe is used, which is arranged in close contact with the curved surface structure of the diameter change area. The rotation of the drive shaft is used to achieve circumferential online scanning, and the surface and near-surface cracks at the diameter change area are detected in real time, so as to visualize the inspection results. For the weld and heat-affected zone, an ultrasonic phased array oblique-entry online probe is used, which is arranged on the outside of the weld. The full coverage of the weld area is achieved by dynamically adjusting the sound beam angle. Combined with PWI-TFM technology, the online inspection efficiency is improved, and welding defects such as incomplete fusion, incomplete penetration, inclusions, and porosity are identified in real time. S2.4: Online inspection of shaft head: A focused ultrasonic phased array online probe is used, which is placed in close contact with the shaft surface on the outside of the shaft sleeve. By controlling the excitation time of the probe, the focus point of the sound wave is dynamically adjusted to accurately match the rotation state of the drive shaft. The internal structure of the shaft head is dynamically scanned online, and defects such as hidden cracks and fractures inside the shaft head are identified in real time, solving the problem of difficult detection of internal defects of the shaft head.

5. The method for online non-destructive testing of a large rolling mill drive shaft according to claim 4, characterized in that: S2.2 further includes the following detection steps: S2.2.1: Surface and Near-Surface Defect Detection: Considering the structural characteristics of shaft wall thickness and curvature, as well as the current situation of surface oil and rust with the basic protective layer, after screening and optimization of detection methods such as magnetic flux leakage, eddy current, and magnetic disturbance, a differential fitting-based electromagnetic disturbance detection probe is used in combination with a circumferential array eddy current online detection probe to perform shaft detection. During detection, the circumferential array electromagnetic probe is non-contactly wrapped around the outside of the shaft. Combined with the axial scanning strategy of the array probe, 360° full circumferential online scanning is achieved by taking advantage of the rotation of the drive shaft itself. Relying on the differential fitting electromagnetic disturbance detection technology, the influence of surface rust, impurities, and probe lifting caused by drive shaft rotation can be effectively offset, reducing signal interference and improving the consistency and reliability of detection signals. Electromagnetic signals of surface and near-surface cracks, local corrosion, and other defects can be collected in real time without grinding the shaft surface. After analysis and processing, the collected data completes the two-dimensional imaging detection of shaft defects. S2.2.2: Internal Defect Detection: A high-speed scanning ultrasonic phased array probe is used in conjunction with PWI-TFM technology. The rotation of the drive shaft is used to achieve dynamic scanning of the shaft's interior. Taking advantage of the rapid plane wave imaging of PWI-TFM, the rotational speed of the drive shaft is matched to achieve high-precision real-time imaging detection of internal delamination, voids, and buried cracks. Finally, the detection data of the array eddy current and ultrasonic phased array are fused to form an online full-coverage evaluation result of shaft defects.

6. The method for online non-destructive testing of a large rolling mill drive shaft according to claim 1, characterized in that: In S3.1, the data acquisition frame rate is linked to the rotational speed of the drive shaft. The higher the rotational speed, the higher the acquisition frame rate, ensuring the integrity and continuity of the detection data during the rotation of the drive shaft.

7. The method for online non-destructive testing of a large rolling mill drive shaft according to claim 1, characterized in that: S4 specifically includes the following steps: S4.1: Defect qualitative judgment: Based on the defect imaging characteristics and signal amplitude changes, combined with the service conditions of the drive shaft, determine the defect type and its cause; S4.2: Quantitative Defect Judgment: Based on the quantitative advantages of SAFT and FMC-TFM technologies, the depth, length, and width of defects are accurately measured, ensuring that the positioning error is ≤ ±0.3mm and the quantitative error is ≤ ±10%. S4.3: Health Status Assessment: Based on the defect detection results and combined with the design requirements and service standards of the mill drive shaft, the health status of the drive shaft is determined in real time and divided into four levels: qualified, pending observation, requiring shutdown for maintenance, and emergency replacement. The assessment results are pushed to the mill control room terminal in real time and displayed visually.

8. The method for online non-destructive testing of a large rolling mill drive shaft according to claim 7, characterized in that: S5 specifically includes the following steps: S5.1: For drive shafts in "qualified" condition, record the test data, include them in the equipment health record, and conduct online re-inspection according to the regular cycle; S5.2: For drive shafts in the "to be observed" state, set up online continuous monitoring reminders, and the system will automatically start re-inspection periodically to track defect changes in real time; S5.3: For drive shafts in the "requiring downtime for maintenance" state, provide suggestions on the optimal maintenance window period in conjunction with the rolling mill production plan to reduce the impact of maintenance on production; S5.4: For drive shafts in "emergency replacement" status, the system will immediately issue an audible and visual alarm to remind on-site operators to take emergency measures to prevent safety accidents from occurring.

Images