CN118604110A - Composite excitation detection method and detection probe for shallow cracks on wheel rim tread - Google Patents

Abstract

The present invention belongs to the technical field of nondestructive testing, and in particular, relates to a composite excitation detection method and detection probe for shallow cracks on a rim tread. The detection method and detection probe can realize efficient detection and differentiation of surface and near-surface cracks on the rim tread, thereby satisfying the accurate quantitative evaluation of rim surface cracks by technical personnel. The present invention provides a composite excitation detection method and detection probe for shallow cracks on a rim tread, wherein the composite excitation detection probe for shallow cracks on a rim tread includes: a probe housing, a Lemo joint, a permanent magnet magnetization module, an AC excitation module, a sensor array, and a signal conditioning circuit.

Description

Composite excitation detection method and detection probe for shallow cracks on wheel rim tread

Technical Field

The invention belongs to the technical field of nondestructive testing, and in particular relates to a composite excitation detection method for shallow cracks on a rim tread and a detection probe.

Background Art

During the operation of the train wheel rim, it is subjected to complex vertical and lateral forces at the same time. Therefore, after long-term service, rolling contact fatigue (RCF) cracks will be generated on the wheel rim tread, and this type of rolling contact fatigue crack accounts for the highest proportion of wheel rim damage. After research, it was found that the RCF cracks on the wheel rim tread are usually distributed in a direction of 15 to 60 degrees (mostly around 45 degrees) to the direction of the train's travel, and are located within 3mm of the wheel rim surface, thus posing a severe challenge to the safety, economy, comfort and efficient operation of the train.

In order to ensure the high-speed, smooth and safe running of trains, it is particularly necessary for technicians to regularly inspect and maintain important components such as wheel rims. As an existing common non-destructive testing method, the magnetic flux leakage (MFL) detection method is fast and does not require coupling agents. It can achieve fast and reliable detection without destroying the geometric characteristics and performance of the workpiece. Therefore, it has a wide range of application prospects in the fields of railways, petrochemicals, aerospace, etc. The detection principle can be specifically described as follows: when there are defects in the material that cut the magnetic lines of force, the flow direction of the magnetic induction lines inside the material will change accordingly: in addition to part of the magnetic flux directly passing through the defect or the inside of the material to bypass the defect, part of the magnetic flux will leak to the surface of the material, bypass the defect through the air and then enter the material; thus, a leakage magnetic field is formed on the surface of the material. When there are no defects in the material, no leakage magnetic field will be generated. Therefore, the detection of wheel rim tread cracks can be achieved through leakage magnetic field detection.

However, after further research, the inventors found that in order to overcome the phenomenon of insufficient sensitivity of leakage magnetic field detection and obtain better leakage magnetic field detection effect, it is usually necessary to saturate magnetize the test piece to be tested. However, the rim size is huge, and it is difficult to saturate magnetize it; and during the detection process, saturation magnetization will make it extremely difficult to accurately control the test piece to be tested, and additional demagnetization operations are required after the detection, which is time-consuming and laborious. In addition, affected by the skin effect, the AC excitation signal only exists within 1mm of the surface, so it is difficult to meet the full detection of defects within the depth range of the rim tread, and it is impossible to achieve the technical requirements for precise maintenance. Therefore, it is urgent for technical personnel in this field to provide a detection method and detection probe that can effectively and fully meet the requirements for shallow crack detection on the rim tread, so as to solve the detection bottleneck of defects on the surface (within 3mm) of the rim tread in the prior art, realize accurate quantification of RCF cracks, and provide data support for accurate maintenance of the tread.

Summary of the invention

The present invention provides a composite excitation detection method and detection probe for shallow cracks on a rim tread. The detection method and detection probe can realize efficient detection and differentiation of surface and near-surface cracks on the rim tread, thereby satisfying the technical personnel's accurate quantitative evaluation of rim surface cracks.

In order to solve the above technical problems, the present invention adopts the following technical solutions:

The composite excitation detection method for shallow cracks on a wheel rim tread comprises the following steps:

Step 1: Build a wheel rim tread shallow crack composite excitation detection probe required by the wheel rim tread shallow crack composite excitation detection method, and complete the detection debugging initialization between it and the test piece under test;

Step 2: Use composite excitation to obtain multi-channel sensor data and return its background value to zero;

Step 3: Sort the sensor data according to the time when the sensor signals corresponding to the sensor data arrive at the same position, and calculate the hysteresis value of the sensor data of other paths relative to the sensor data of the first path, so as to eliminate the hysteresis between the sensor data of multiple paths;

Step 4: Set the minimum threshold X and the maximum threshold Y;

Step 5: If there is a data value less than the minimum threshold value X in the sensor data of the road before the middle road, and there is a data value greater than the maximum threshold value Y in the sensor data of the road after the middle road, then the minimum value of the data values less than the minimum threshold value X in the sensor data of the road before the middle road is recorded as A1, and the driving distance corresponding to the data A1 is recorded as B1; the maximum value of the data values greater than the maximum threshold value Y in the sensor data of the road after the middle road is recorded as A2, and the driving distance corresponding to the data A2 is recorded as B2; at the same time, the driving distance corresponding to the minimum value data in the sensor data of the middle road is recorded as B3, and the driving distance corresponding to the maximum value data in the sensor data of the middle road is recorded as B4;

Then the predicted value of the horizontal cracking angle is calculated to satisfy: ;

in, , , a1=（A3-A1）*w, a2=（A2-A3）*w, b1=（B3-B1）, b2=（B2-B4）; w is the spacing between adjacent sensor units in the sensor array;

Step 6: If there is no data value less than the minimum threshold value X in the sensor data of the road before the middle road, the maximum value data in the sensor data of the road before the middle road is recorded as A'1, and the driving distance corresponding to the data A'1 is recorded as B'1; and the maximum value data in the sensor data of the road after the middle road is recorded as A'2, and the driving distance corresponding to the data A'2 is recorded as B'2;

Then the predicted value of the horizontal cracking angle is calculated to satisfy: ;

Where A=（A'2-A'1）*w，B=（B'2-B'1）；w is the spacing between adjacent sensor units in the sensor array;

Step 7: Obtain each magnetic field characteristic signal under composite excitation;

If the magnetic field characteristic signal of the lower excitation frequency of the two excitation frequencies of the composite excitation is not stronger than the preset reference under composite excitation, it means that there are no surface and near-surface cracks in the tested specimen;

If the magnetic field characteristic signal of the lower excitation frequency of the two excitation frequencies of the compound excitation is stronger than the preset reference under compound excitation, it indicates that there is a crack in the test piece;

In the case of judging whether there is a crack in the test piece, if the magnetic field characteristic signal of the higher excitation frequency of the two excitation frequencies of the composite excitation is not stronger than the preset reference under composite excitation, it indicates that the crack is a near-surface crack, otherwise it indicates that the crack is a surface crack.

More preferably, the lower excitation frequency of the two excitation frequencies of the composite excitation is selected as a 2KHz excitation frequency; and the higher excitation frequency of the two excitation frequencies of the composite excitation is selected as a 20KHz excitation frequency.

Preferably, the method further comprises the following steps:

Step 8: When the crack is detected as a surface crack, the predicted value of the horizontal cracking angle of the crack is calculated. As known physical information, it is integrated into the convolutional neural network;

The convolutional neural network is used to realize the crack depth after training. and vertical crack angle Accurate quantitative assessment.

On the other hand, the present invention also provides a composite excitation detection probe for shallow cracks on a rim tread, comprising: a probe housing, a Lemo connector, a permanent magnet magnetization module, an AC excitation module, a sensor array, and a signal conditioning circuit;

Among them, the Lemo connector is installed on one side of the probe housing to achieve signal transmission;

The permanent magnet magnetization module consists of two permanent magnets and a permanent magnet yoke; the two permanent magnets are symmetrically distributed on both sides of the sensor array to realize magnetization of the test piece;

The AC excitation module includes an AC excitation coil and an AC excitation U-shaped magnetic core; the AC excitation coil is evenly wound around the middle of the AC excitation U-shaped magnetic core to generate an AC excitation magnetic field signal;

The sensor array is composed of multiple groups of sensor units, which are arranged at the middle position of the lower leg of the AC excitation U-shaped magnetic core and are evenly and equidistantly arranged along the crossbeam direction of the AC excitation U-shaped magnetic core; the sensor units are TMR magnetic sensors with a Z-axis sensitive direction;

The signal conditioning circuit is used to amplify and filter the signal output by the sensor array.

Preferably, it also includes: a probe gland and a wheel;

The probe gland and the probe housing are fixed to form an integrated structure by fixing screws;

The wheel is installed on the outside of the probe housing and is used to roll and adjust the composite excitation detection probe for shallow cracks on the rim tread.

Preferably, the permanent magnet magnetization module and the AC excitation module are arranged at -45° along the traveling direction of the composite excitation detection probe for shallow cracks on the rim tread.

The present invention provides a composite excitation detection method and detection probe for shallow cracks on a rim tread, wherein the composite excitation detection probe for shallow cracks on a rim tread includes: a probe housing, a Lemo joint, a permanent magnet magnetization module, an AC excitation module, a sensor array, and a signal conditioning circuit. The composite excitation detection method and detection probe for shallow cracks on a rim tread with the above characteristics increase the penetration depth of the AC excitation field in the rim tread through a composite excitation method, improve the sensitivity of the surface and near-surface crack detection of the rim tread (within 3mm), and achieve the distinction between the surface and near-surface cracks on the rim tread; at the same time, a crack horizontal crack angle quantification algorithm based on multi-channel sensor data achieves an accurate quantitative evaluation of the upper surface crack depth and vertical crack angle of the rim tread, providing important technical support for the detection of rim tread cracks in industry and the determination of the grinding amount.

BRIEF DESCRIPTION OF THE DRAWINGS

The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the following drawings:

FIG1 is a schematic diagram of the explosion structure of a composite excitation detection probe for shallow cracks on a rim tread provided by the present invention;

FIG2 is a schematic diagram of the structure of a probe housing in a composite excitation detection probe for shallow cracks on a rim tread provided by the present invention;

FIG3 is a schematic diagram of the angle of cracking of the rim tread;

FIG4 is a schematic diagram of a process of a composite excitation detection method for shallow cracks on a rim tread provided by the present invention;

FIG5 is a C-scan image obtained by scanning with the detection probe;

Figure 6 shows the Bz characteristic signal picked up by the sensor just above the surface crack under 2KHz excitation;

Figure 7 shows the Bz characteristic signal picked up by the sensor just above the surface crack under 20KHz excitation;

Figure 8 shows the Bz characteristic signal picked up by the sensor just above the near-surface crack under 2KHz excitation;

Figure 9 shows the Bz characteristic signal picked up by the sensor just above the near-surface crack under 20KHz excitation;

Figure 10 shows the Bx Markov image processing result;

Figure 11 shows the Bz Markov image processing result;

Figure numerals: 10, wheel; 20, probe housing; 201, housing side wall; 202, Lemo connector hole; 203, threaded hole; 204, sensor array slot; 205, wheel axle hole; 206, AC excitation U-shaped magnetic core slot; 207, detection direction mark; 208, permanent magnet slot; 209, reinforcing rib; 30, Lemo connector; 40, sensor array; 50, permanent magnet; 60, AC excitation module; 601, AC excitation U-shaped magnetic core; 602, AC excitation coil; 70, permanent magnet yoke; 80, signal conditioning circuit; 90, probe gland; 100, fixing screw.

DETAILED DESCRIPTION

The present invention provides a composite excitation detection method and detection probe for shallow cracks on a rim tread. The detection method and detection probe can realize efficient detection and differentiation of surface and near-surface cracks on the rim tread, thereby satisfying the technical personnel's accurate quantitative evaluation of rim surface cracks.

Specifically, the present invention provides a composite excitation detection method for shallow cracks on a rim tread, as shown in FIG4 , comprising the following steps:

Step 1: Build the wheel rim and tread shallow crack composite excitation detection probe required by the wheel rim and tread shallow crack composite excitation detection method, and complete the detection debugging initialization between it and the test piece under test.

Specifically, the composite excitation detection probe for shallow cracks on the rim tread can be constructed as follows: wherein the permanent magnet magnetization module is responsible for providing a constant magnetic field in the composite excitation for the composite excitation detection probe for shallow cracks on the rim tread, and realizing the magnetization of the rim tread. Under the action of the signal conditioning circuit, the AC excitation module modulates to form a dual-frequency AC excitation magnetic field signal; under composite excitation, the dual-frequency AC excitation magnetic field signal can realize the aggregation of surface and buried depth induced currents, which is conducive to the distinction between surface and buried depth defects.

Here, it is optional that the lower excitation frequency of the two excitation frequencies of the compound excitation is selected as the 2KHz excitation frequency; and the higher excitation frequency of the two excitation frequencies of the compound excitation is selected as the 20KHz excitation frequency.

In addition, it is necessary to add that, in the process of initializing the detection and debugging of the composite excitation detection probe for shallow cracks on the rim tread and the test piece, on the one hand, the excitation direction of the composite excitation should be optimized (specifically, the permanent magnet magnetization module and the AC excitation module should be set at -45° along the travel direction of the composite excitation detection probe for shallow cracks on the rim tread); on the other hand, the appropriate permanent magnet magnetization intensity should be selected to concentrate the magnetization effect on the near surface within 3mm of the rim tread, thereby realizing surface and near-surface crack detection.

Step 2: Use composite excitation to obtain multi-channel sensor data and return its background value to zero;

Step 3: Sort the sensor data according to the time when the sensor signals corresponding to the sensor data arrive at the same position, and calculate the hysteresis value of the sensor data of other paths relative to the first path of sensor data, so as to eliminate the hysteresis between the multiple sensor data.

After completing step one, further implement steps two and three.

Specifically, in steps 2 and 3, firstly, the multi-channel sensing data of multiple groups of sensor units are obtained by using composite excitation, and the background values thereof are reset to zero. In the process of obtaining the multi-channel sensing data of multiple groups of sensor units, since the different sensing data transmitted by different sensor units arrive at the same position at different times, in order to eliminate the lag between the sensing data of each channel, the lag value of the sensing data of other channels relative to the first sensing data is calculated, so as to eliminate the lag of the sensing data of other channels (compared with the first sensing data).

Step 4: Set the minimum threshold X and the maximum threshold Y;

Step 5: If there is a data value less than the minimum threshold value X in the sensor data of the road before the middle road, and there is a data value greater than the maximum threshold value Y in the sensor data of the road after the middle road, then the minimum value of the data values less than the minimum threshold value X in the sensor data of the road before the middle road is recorded as A1, and the driving distance corresponding to the data A1 is recorded as B1; the maximum value of the data values greater than the maximum threshold value Y in the sensor data of the road after the middle road is recorded as A2, and the driving distance corresponding to the data A2 is recorded as B2; at the same time, the driving distance corresponding to the minimum value data in the sensor data of the middle road is recorded as B3, and the driving distance corresponding to the maximum value data in the sensor data of the middle road is recorded as B4;

Then the predicted value of the horizontal cracking angle is calculated to satisfy: ;

in, , , a1=（A3-A1）*w, a2=（A2-A3）*w, b1=（B3-B1）, b2=（B2-B4）; w is the spacing distance between adjacent sensor units in the sensor array.

On the basis of completing steps 2 and 3, further implement steps 4 and 5. The reason for this setting is that in actual detection, there are usually the following situations: the Bz characteristic signal deviating from the left end point of the crack presents a trough, the Bz characteristic signal deviating from the right end point of the crack presents a peak, and the Bz at the top of the crack presents a standard peak and trough. Based on this, we set the minimum threshold X and the maximum threshold Y, and further derive the quantification algorithm of the horizontal cracking angle of the crack based on multiple sets of sensor data.

Specifically, taking the horizontal cracking angle schematic diagram shown in Figure 3 as an example, the crack parameter selection reference is as follows: length 18mm, width 0.5mm, depth 5mm, horizontal cracking angle 60 degrees, vertical cracking angle 90 degrees. At this time, the C scan obtained by the detection probe scanning is shown in Figure 5. In Figure 5, the approximate angle range of the horizontal cracking of the crack is intuitively reflected. One point that needs to be explained is that the horizontal cracking of the crack is crucial for the assessment of the crack depth, and the accurate assessment of the crack depth can be achieved under the premise of angle quantification.

Step 6: If there is no data value less than the minimum threshold value X in the sensor data of the road before the middle road, the maximum value data in the sensor data of the road before the middle road is recorded as A'1, and the driving distance corresponding to the data A'1 is recorded as B'1; and the maximum value data in the sensor data of the road after the middle road is recorded as A'2, and the driving distance corresponding to the data A'2 is recorded as B'2;

Then the predicted value of the horizontal cracking angle is calculated to satisfy: ;

Wherein, A=（A'2-A'1）*w, B=（B'2-B'1）; w is the spacing distance between adjacent sensor units in the sensor array.

On the basis of completing step 5, further implement step 6. It is worth noting that through the above crack horizontal cracking angle quantification algorithm, the predicted value of the crack horizontal cracking angle can be calculated to be 59.108°, with an absolute error of only 0.892° and a relative error of 1.49%. The quantification accuracy is high and the result is relatively accurate.

Step 7: Obtain each magnetic field characteristic signal under composite excitation;

If the magnetic field characteristic signal of the lower excitation frequency of the two excitation frequencies of the composite excitation is not stronger than the preset reference under composite excitation, it means that there are no surface and near-surface cracks in the tested specimen;

If the magnetic field characteristic signal of the lower excitation frequency of the two excitation frequencies of the composite excitation is stronger than the preset reference under composite excitation, it indicates that there is a crack in the test piece;

In the case of judging whether there is a crack in the test piece, if the magnetic field characteristic signal of the higher excitation frequency of the two excitation frequencies of the composite excitation is not stronger than the preset reference under composite excitation, it indicates that the crack is a near-surface crack, otherwise it indicates that the crack is a surface crack.

On the basis of completing steps five and six, further implement step seven. It should be noted that since the rim tread of the test piece is a ferromagnetic material, its "skin effect" is very obvious. Through simulation and experimental verification, it is found that higher excitation frequencies have poor detection capabilities for near-surface cracks. Therefore, composite excitation can effectively distinguish between surface and near-surface cracks on the rim tread, providing the possibility for classification, identification, and precise quantification of cracks in the thin metal layer (within 3mm) of the rim tread. The characteristic signals of surface and near-surface cracks under different AC excitations are shown in Figures 6 to 9.

As a more preferred embodiment of the present invention, as shown in FIG4 , a composite excitation detection method for shallow cracks on a rim tread also includes the following steps:

Step 8: When the crack is detected as a surface crack, the predicted value of the horizontal cracking angle of the crack is calculated. As known physical information, it is integrated into the convolutional neural network;

The convolutional neural network is used to realize the crack depth after training. and vertical crack angle Accurate quantitative assessment.

It is worth noting that after the crack detection process is completed, further implementation of step eight can achieve quantitative analysis of crack depth and vertical crack angle. In the process of quantitative analysis, when the crack classification and angle are known, the existing technology can significantly improve the quantification accuracy of crack depth, providing technical feasibility for further precise repair of rim tread.

Here, the simulation and experiment of cracks with a surface depth of 1-9 mm, a horizontal crack angle of 0-90°, a vertical crack angle of 10-90°, and a buried depth of 1-3 mm are used as an example for explanation.

Specifically, the process of incorporating convolutional neural networks into training can be described in detail as follows: converting one-dimensional sequence data into two-dimensional (image) form for analysis and processing; using the Markov migration field to convert the data set signal sequence into two-dimensional data; constructing a Markov transfer matrix and developing it into a Markov migration field to achieve effective encoding of image data.

For the time series X=(xt, t=1, 2, ..., T), the image encoding steps are as follows:

1. Divide the time series X(t) into Q bins (labeled 1, 2, ..., Q, with the same amount of data in each bin);

2. Change each data in the time series to the serial number of its corresponding quantile bin;

3. Construct the transfer matrix W (wij represents the frequency of transfer from bin i to bin j): ;

4. Construct the Markov transfer matrix M: ;

The Bx and Bz feature signals are processed by the Markov image coding method. The processing results are shown in Figures 10 and 11. Finally, the database is built, the image is convolved, and the physical information of the horizontal cracking angle is integrated into the convolutional neural network for auxiliary correction.

When evaluating the crack depth and vertical cracking angle, the crack parameters of the embodiment are 18mm long, 0.5mm wide, 5mm deep, 45° horizontal cracking angle, and 70° vertical cracking angle. The network predicts a vertical crack angle of 75.143°, an absolute error of 5.143°, a relative error of 7.347%, a depth of 5.125mm, an absolute error of 0.125mm, and a relative error of 2.5%. Therefore, it is confirmed that the prediction results of the convolutional neural network are more accurate, with an accuracy of more than 97% for depth prediction, which solves the problem of quantifying the horizontal angle, vertical cracking angle, and depth of surface cracks.

At this point, the composite excitation detection method for shallow cracks on the rim tread provided by the present invention has realized the quantitative and qualitative analysis of shallow cracks on the rim tread; in addition, with the help of convolutional neural networks, the technical problems of quantifying the horizontal angle, vertical cracking angle and depth of surface cracks have also been solved, providing strong technical support for the detection and repair of cracks on the rim tread.

On the other hand, a composite excitation detection probe for shallow cracks on a rim tread, as shown in FIGS. 1 and 2 , includes: a probe housing 20 , a Lemo connector 30 , a permanent magnet magnetization module, an AC excitation module 60 , a sensor array 40 , and a signal conditioning circuit 80 .

The Lemo connector 30 is installed on one side of the probe housing 20 (specifically, fixedly installed through the Lemo connector hole 202 on the housing side wall 201 ) to realize signal transmission.

The permanent magnet magnetization module is composed of two permanent magnets 50 and a permanent magnet yoke 70. The two permanent magnets 50 (the permanent magnets 50 are placed in the permanent magnet slots 208 of the probe housing 20) are symmetrically distributed on both sides of the sensor array 40 to magnetize the test piece.

The AC excitation module 60 includes an AC excitation U-shaped magnetic core 601 and an AC excitation coil 602. The AC excitation coil 602 is evenly wound in the middle of the AC excitation U-shaped magnetic core 601 (the AC excitation U-shaped magnetic core 601 is placed in the AC excitation U-shaped magnetic core slot 206 of the probe housing 20), and is used to generate an AC excitation magnetic field signal (the AC excitation signal uses a composite excitation of 2KHz excitation frequency and 20KHz excitation frequency to meet the detection method's requirements for surface and deep defect detection).

It is worth noting that the permanent magnet magnetization module can reduce the magnetic permeability of the rim tread, increase the penetration depth of the AC excitation signal of the AC excitation module, and realize the detection of surface and sub-surface defects of the rim tread.

The sensor array 40 is composed of multiple groups of sensor units (the sensor array 40 is placed in the sensor array slot 204 of the probe housing 20), and the sensor units are arranged in the middle position of the lower leg of the AC excitation U-shaped magnetic core 601, and are evenly and equidistantly arranged along the crossbeam direction of the AC excitation U-shaped magnetic core 601. Specifically, the sensor unit uses a TMR magnetic sensor with a Z-axis sensitive direction.

The signal conditioning circuit 80 is used to amplify and filter the signal output by the sensor array.

In addition, as a more preferred embodiment of the present invention, as shown in FIG. 1 and FIG. 2 , the rim tread shallow crack composite excitation detection probe further includes: a probe gland 90 and a wheel 10 .

Among them, the probe cover 90 is fixed to the probe housing 20 (the probe housing 20 is provided with a threaded hole 203 for positioning the fixing screw 100) by the fixing screw 100 to form an integrated structure; the wheel 10 is installed on the outside of the probe housing 20 (the probe housing 20 is provided with a wheel axle hole 205 for fixing the wheel 10) for rolling adjustment of the composite excitation detection probe for shallow cracks on the rim tread.

As a more preferred embodiment of the present invention, the permanent magnet magnetization module and the AC excitation module are arranged at -45° in the direction of travel of the composite excitation detection probe for shallow cracks on the rim tread. The reason for this arrangement is that, as shown in FIG3 , it has been found through research that the RCF cracks on the rim tread are usually distributed along the direction of 15°-60° in the direction of train travel, especially mostly at 45°. Therefore, by arranging the permanent magnet magnetization module and the AC excitation module at -45° in the direction of travel of the composite excitation detection probe for shallow cracks on the rim tread, the permanent magnet magnetization module and the AC excitation module can obtain as many magnetic field characteristic signals as possible, and ultimately improve the accuracy of the composite excitation detection probe for shallow cracks on the rim tread in detecting rim surface cracks.

One point that needs to be added is that, in order to facilitate the installation of the permanent magnet magnetization module and the AC excitation module by the technicians, the detection direction mark 207 can be arranged in the probe housing 20, so as to help the technicians achieve the technical purpose of setting the permanent magnet magnetization module and the AC excitation module at a -45° azimuth along the travel direction of the rim tread shallow crack composite excitation detection probe. In addition, the probe housing 20 can be further installed with reinforcing ribs 209 to increase the stability of the probe housing 20 structure.

The present invention provides a composite excitation detection method and detection probe for shallow cracks on a rim tread, wherein the composite excitation detection probe for shallow cracks on a rim tread includes: a probe housing, a Lemo joint, a permanent magnet magnetization module, an AC excitation module, a sensor array, and a signal conditioning circuit. The composite excitation detection method and detection probe for shallow cracks on a rim tread with the above characteristics increase the penetration depth of the AC excitation field in the rim tread through a composite excitation method, improve the sensitivity of the surface and near-surface crack detection of the rim tread (within 3mm), and achieve the distinction between the surface and near-surface cracks on the rim tread; at the same time, a crack horizontal crack angle quantification algorithm based on multi-channel sensor data achieves an accurate quantitative evaluation of the upper surface crack depth and vertical crack angle of the rim tread, providing important technical support for the detection of rim tread cracks in industry and the determination of the grinding amount.

The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art who is familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims (6)

1. The rim tread shallow crack composite excitation detection method is characterized by comprising the following steps of:

Step one: building a rim tread shallow crack composite excitation detection probe required by the rim tread shallow crack composite excitation detection method, and finishing detection debugging initialization between the rim tread shallow crack composite excitation detection probe and a detected test piece;

step two: acquiring multi-path sensing data by using composite excitation, and carrying out zero-resetting treatment on a background value of the multi-path sensing data;

Step three: sequencing the sensing signals corresponding to the sensing data according to the time when the sensing signals reach the same position, and calculating the hysteresis values of the sensing data of other paths relative to the first path of sensing data, so that the hysteresis among the multiple paths of sensing data is eliminated;

step four: setting a minimum threshold value X and a maximum threshold value Y;

step five: if the sensing data of the front road of the middle road has a data value smaller than the minimum threshold value X and the sensing data of the rear road of the middle road has a data value larger than the maximum threshold value Y, the minimum value of the data values smaller than the minimum threshold value X in the sensing data of the front road of the middle road is marked as A1, and the driving distance corresponding to the data A1 is marked as B1; the maximum value of the data values which are larger than the maximum threshold Y in the sensing data of the rear road of the middle road is marked as A2, and the driving distance corresponding to the data A2 is marked as B2; meanwhile, the driving distance corresponding to the minimum value data in the middle road sensing data is marked as B3, and the driving distance corresponding to the maximum value data in the middle road sensing data is marked as B4;

Then calculating the predicted value of the horizontal cracking angle of the crack, and meeting the following conditions: ；

Wherein, ，A1= (A3-A1) ×w, a2= (A2-A3) ×w, b1= (B3-B1), b2= (B2-B4); w is the separation distance between adjacent sensor units in the sensor array;

step six: if the sensing data of the front road of the middle road does not have a data value smaller than the minimum threshold value X, the maximum value data in the sensing data of the front road of the middle road is marked as A '1, and the driving distance corresponding to the data A '1 is marked as B '1; the maximum value data in the sensing data of the rear road of the middle road is recorded as A '2, and the driving distance corresponding to the data A '2 is recorded as B '2;

Then calculating the predicted value of the horizontal cracking angle of the crack, and meeting the following conditions: ；

Wherein a= (a '2-a' 1) w, b= (B '2-B' 1); w is the separation distance between adjacent sensor units in the sensor array;

step seven: acquiring characteristic signals of each magnetic field under composite excitation;

If the magnetic field characteristic signal of the lower excitation frequency of the two excitation frequencies of the composite excitation is not stronger than a preset reference under the composite excitation, the condition that no surface and near-surface cracks exist in the tested test piece is indicated;

If the magnetic field characteristic signal of the lower excitation frequency of the two excitation frequencies of the composite excitation is stronger than a preset reference under the composite excitation, the existence of cracks in the tested test piece is indicated;

and under the condition that the crack exists in the tested test piece, if the magnetic field characteristic signal of the higher excitation frequency of the two excitation frequencies of the composite excitation is not stronger than the preset reference under the composite excitation, the crack is indicated to be the crack at the near surface, otherwise, the crack is indicated to be the surface crack.

2. The method for detecting the rim tread shallow crack composite excitation according to claim 1, wherein the lower excitation frequency of two excitation frequencies of the composite excitation is selected to be 2KHz excitation frequency; the higher excitation frequency of the two excitation frequencies of the composite excitation is selected as the excitation frequency of 20 KHz.

3. The rim tread shallow crack composite excitation detection method according to claim 1, further comprising the steps of:

Step eight: when detecting that the crack exists as the surface crack, calculating the predicted value of the horizontal crack angle of the crack As known physical information, merging into a convolutional neural network;

The convolutional neural network is used for realizing crack depth after training And vertical cracking angleIs used for accurate quantitative evaluation of (a).

4. Rim tread shallow crack composite excitation detection probe, which is based on the rim tread shallow crack composite excitation detection method according to any one of claims 1-3, and is characterized by comprising: the sensor comprises a probe shell, a Ramo joint, a permanent magnet magnetization module, an alternating current excitation module, a sensor array and a signal conditioning circuit;

Wherein, lei Mo joint is installed at one side of the probe shell, which is used for realizing signal transmission;

The permanent magnet magnetizing module consists of two permanent magnets and a permanent magnet yoke; the two permanent magnets are symmetrically distributed on two sides of the sensor array and are used for magnetizing a tested test piece;

The alternating current excitation module comprises an alternating current excitation coil and an alternating current excitation U-shaped magnetic core; the alternating current excitation coil is uniformly wound in the middle of the alternating current excitation U-shaped magnetic core and is used for generating an alternating current excitation magnetic field signal;

the sensor array is composed of a plurality of groups of sensor units, and the sensor units are arranged at the middle position of the lower supporting leg of the alternating current excitation U-shaped magnetic core and uniformly and equidistantly arranged along the direction of the cross beam of the alternating current excitation U-shaped magnetic core; the sensor unit selects TMR magnetic sensor with Z-axis sensitive direction;

And the signal conditioning circuit is used for amplifying and filtering the signals output by the sensor array.

5. The rim tread shallow crack composite excitation detection probe as in claim 4, further comprising: probe gland and wheel;

the probe gland and the probe shell are fixed into an integrated structure through the fixing screw;

the wheel is arranged outside the probe shell and is used for rolling and adjusting the rim tread shallow crack composite excitation detection probe.

6. The rim tread shallow crack composite excitation detection probe according to claim 4, wherein the permanent magnet magnetization module and the alternating current excitation module are arranged along the-45-degree direction of the travelling direction of the rim tread shallow crack composite excitation detection probe.