A wind power tower drum concave detection method based on phased array technology

Abstract

The present application relates to a kind of wind power tower drum recess detection method based on phased array technique, the indirect quantification of recess is realized by detecting convex region;Customized calibration test block containing step convex and paint layer thickness gradient, the mapping relationship of geometric parameter and ultrasonic response is established, the mathematical model of paint layer thickness (g) and gain compensation value (i) is established, the linear equation of fitting convex angle tangent value (tanθ) and gain value (h) is fitted, the horizontal size and length of defect are positioned by wave amplitude disappearance boundary, the recess depth is inverted by simultaneously solving paint layer compensation gain and angle gain equation, the size and position of defect are obtained by three-dimensional quantitative calculation, detection precision is high, error is controlled within 1%, while achieving free scaffold operation, reduce equipment downtime and scaffold rental cost, single detection cost is greatly reduced, it is convenient to use, and the effect is good, it is the innovation on wind power tower drum recess detection method.

Description

Technical Field

[0001] This invention relates to the field of nondestructive testing technology, and in particular to a method for detecting dents in wind turbine towers based on phased array technology. Background Technology

[0002] Against the backdrop of the global energy structure's accelerated transition to renewable energy, wind power, as one of the mainstays, has seen its unit capacity continuously increase, leading to a significant increase in the weight of core components (such as nacelles, hubs, and blades) and the height of the tower. As a critical load-bearing structure supporting the entire wind turbine, the tower's structural integrity and safety directly impact the operational safety and economic benefits of wind farms. With the increasing complexity of the tower's service environment and its age, localized indentations in the tower wall are frequently observed in the industry, posing a significant threat to the structural safety of the tower.

[0003] However, the industry currently lacks effective means for efficient and accurate quantitative detection of tower dent damage. Existing mainstream methods rely on manual measurement using simple rulers and compasses, which has significant limitations: First, large-scale scaffolding must be erected before measurement, consuming substantial manpower, resources, and time, and posing safety risks due to working at heights; second, manual ruler and compasses methods suffer from poor positioning accuracy and low repeatability in defining the boundary of dent damage and measuring dent depth, making it difficult to obtain accurate and reliable quantitative data; most importantly, the feasibility of this method is severely restricted, or even completely impossible, when there are auxiliary ladders, cables, platforms, or other obstructions on the tower surface.

[0004] Therefore, in the face of increasingly larger and taller modern wind turbine towers, developing an efficient and safe detection technology and method that does not require scaffolding, can overcome on-site facility obstructions, and can accurately locate the depression boundary and quantify the depression depth has become an urgent need and an inevitable trend to ensure the safe operation of wind farms, improve operation and maintenance efficiency, and promote technological progress in the industry. Summary of the Invention

[0005] In view of the above situation and to overcome the defects of the prior art, the purpose of this invention is to provide a wind turbine tower dent detection method based on phased array technology, which can effectively solve the problem of wind turbine tower dent detection.

[0006] The technical solution solved by this invention is:

[0007] A method for detecting dents in wind turbine towers based on phased array technology includes the following steps:

[0008] Step 1: Prepare control test blocks

[0009] Comparison block A:

[0010] Prepare a control block A with a protrusion. Prepare at least 10 control blocks A. Each control block 1 has a protrusion at its center. Define the direction along the length of the block as the Y direction, the direction along the width of the block as the X direction, and the vertical upward direction as the Z direction. The angle θ is defined as the inclination angle of the tower protrusion. The 10 control blocks 1 are different in the inclination angle θ of the protrusion.

[0011] Comparison with test block B:

[0012] Multiple comparative test blocks B are prepared. Each comparative test block B has a cuboid structure and its surface is coated with a paint layer of different thicknesses. The paint layer thicknesses of the test blocks are g1, g2, g3, g4, g5, etc., and there are at least 3 different paint layer thicknesses.

[0013] Step 2: Probe Selection

[0014] The probe is a phased array probe, specifically a 10L64-0.5×10 linear array probe, where 10 represents the probe's center frequency of 10MHz, L represents a linear array probe, 64 represents the number of array elements of the probe is 64, 0.5×10 represents the center-to-center spacing of the array elements is 0.5mm and the spacing between adjacent array elements is 0.1mm, and 10 represents the length of the array element is 10mm.

[0015] It is equipped with N63-90S surface wave wedge block, and a displacement sensor is installed inside the probe to record the probe movement distance;

[0016] Step 3: Calibration Test

[0017] 3.1 Paint layer calibration test

[0018] The compensation for surface wave gain value with different paint layer thicknesses was calculated. Comparison test block B was selected, and surface waves were emitted on test block 2. The bottom wave amplitude was adjusted to 80% to obtain the compensation gain for different paint layer thicknesses.

[0019] For the compensation gain corresponding to different paint layer thicknesses, a graph showing the relationship between paint layer thickness g and compensation gain i was plotted using Origin software. A formula conforming to the g-i relationship was fitted using the formula fitting method in Origin, requiring a Pearson correlation coefficient above 0.97 and an R-squared goodness of fit above 0.97. This is the paint layer compensation gain equation. 3.2 Tilting Angle Calibration Test

[0020] Place the probe at the upper left edge of the comparison block A and emit surface waves. The upper left corner of the comparison block A is defined as the origin of the coordinate system, i.e., point (0,0). The scanning path is as follows: the probe first starts from point (0,0) and scans the comparison block A along the positive Y-axis until it reaches the edge of the protruding area of ​​the block. Then, the probe scans along the negative Y-axis from the edge of the protruding area of ​​the comparison block A with 40-50% overlap with the previous scanning path in the Y-axis direction until it reaches the left edge of the comparison block A. This cycle is repeated until the probe reaches the lower edge of the comparison block A and the scanning is completed. The scanning speed is no more than 150 mm / s. When scanning within the protruding area of ​​the surface, there will be reflected waves at the protrusions. The intensity of the reflected waves is related to the protrusion angle θ. Adjust the amplitude of the reflected waves to 80% to obtain the minimum gain corresponding to different protrusion angles and 80% amplitude.

[0021] Using Origin software, plot the relationship between the convexity angle tan(θ) and the gain value h corresponding to 80% amplitude. Fit a formula that conforms to the relationship between tan(θ) and h. The formula should have a Pearson correlation coefficient of 0.97 or higher and an R-squared goodness of fit of 0.97 or higher. This is the gain equation for the convexity angle tan(θ) and the gain value h corresponding to 80% amplitude.

[0022] Step 4: On-site testing

[0023] The X-direction is defined as the horizontal direction of the tower, the Y-direction as the height direction of the tower, and the Z-direction as the direction of the tower's bulge. The tower is composed of different sections, which are connected by welds. The scanning method is to first scan the bottom of the tower to weld H1, then scan welds H1-H2, welds H2-H3, and so on until weld Hn to the top of the tower, thus completing the scanning.

[0024] When scanning areas without protrusions or depressions, the surface wave reflects the bottom wave without defects. If a surface protrusion is found, a reflected wave will be generated at the protrusion's starting point. Fix the amplitude at this point to 80%. After finding the point with the minimum gain, fix the gain value at this point to hxdB. Record the probe position at this point as (1,1).

[0025] 4.1 Defect X-direction dimension measurement

[0026] The probe moves horizontally along the positive X direction from the position of (1,1) and the amplitude gradually decreases. When there is no reflected surface wave, the probe moves a distance Xa along the positive X direction. The probe moves along the negative X direction from the position of (1,1) and the amplitude gradually decreases. When there is no reflected surface wave, the probe moves a distance Xb along the negative X direction. Xa + Xb is the size of the defect in the X direction.

[0027] 4.2 Defect Y-direction dimension measurement

[0028] Starting from the position (1, 1), move horizontally along the positive X direction until the wave amplitude disappears. Return to the position (1, 1) and move horizontally along the negative X direction until the wave amplitude disappears again. Record the minimum x-coordinate of the reflected wave during this process, which is the minimum forward propagation distance of the surface wave, denoted as Y1. min ;

[0029] Place the probe on the opposite side of the convex area along the tower height direction, fix the defect wave amplitude at 80%, find the point of minimum gain, fix the gain value at this moment, and record the probe position at this moment; denoted as position (2,2); move horizontally along the positive X direction from position (2,2) until the wave amplitude disappears, return to position (2,2), and move horizontally along the negative X direction until the wave amplitude disappears. Record the minimum abscissa of the reflected wave during this process, that is, the minimum forward propagation distance of the surface wave, denoted as Y2. min ;

[0030] Measure the Y-direction distance between the two probes along the height of the tower, and denot it as Y3;

[0031] The dimension in the Y direction is calculated as follows: Y = Y3 - Y1 minn -Y2 min ;

[0032] 4.3 Defect Z-direction dimension measurement

[0033] Measure the paint layer thickness of the tower to determine the paint layer compensation gain value. Measure the paint layer thickness g and substitute it into the paint layer compensation gain equation obtained in step 3.1 to obtain the paint layer compensation gain value i for thickness g.

[0034] With the paint layer compensation idB added, the total gain is h = hx + i;

[0035] Substituting the total gain value h into step 3.2, we obtain the gain equation for the gain value h corresponding to the convex angle tan(θ) and the amplitude 80%, thus obtaining tan(θ);

[0036] tan(θ) = Z / (Y / 2)

[0037] The defect height, i.e., the Z value, is obtained.

[0038] This allows us to obtain the location and size of the tower protrusion defect.

[0039] Compared with the prior art, the present invention

[0040] Compared with existing technologies, the present invention is simple and detects concave-to-convex changes, breaking through the limitations of traditional direct measurement of concave. It proposes a physical model that "tower concave is equivalent to reverse convexity," and indirectly quantifies concave by detecting convex areas. A calibration block containing stepped convexity (comparison block A) and paint thickness gradient (comparison block B) is customized to establish a mapping relationship between geometric parameters and ultrasonic response. A mathematical model of paint thickness (g) and gain compensation value (i) is established, and a linear equation is fitted between the tangent of the convex tilt angle (tanθ) and the gain value (h). The horizontal dimension (Xa+Xb) and length (Y=Y3-Y1min-Y2min) of the defect are located through the amplitude vanishing boundary. The paint compensation gain and angle gain equations are solved simultaneously to invert the concave depth. Through three-dimensional quantitative calculation, the size and location of the defect are obtained. X-axis dimension: the probe is moved horizontally to the amplitude vanishing point, and the positive and negative movement distances are accumulated. (Xa+Xb); Y-axis dimension: Measure the distance Y3 between the two probes, and subtract the minimum propagation distances Y1min and Y2min for amplitude disappearance; Z-axis depth: Combine the paint layer compensation gain and angle gain equations, solve tanθ and derive the Z value; thereby realizing the detection of the tower's dent area and dent depth. Compared with the existing technology, this invention has the following advantages: eliminates paint layer interference through mathematical modeling, improves the accuracy of depth measurement, has high detection precision, and controls the error within 1%; at the same time, it achieves scaffold-free operation, ground scanning replaces high-altitude manual measurement, avoids erection costs and safety risks, has strong anti-interference ability, overcomes the limitations of on-site facilities, realizes full-area tower coverage detection (especially the outer wall suspended area), and achieves low-cost maintenance, reduces equipment downtime and scaffold rental costs, significantly reduces the cost of a single detection, is easy to use, and has good results. It is an innovation in wind power tower dent detection methods. Attached Figure Description

[0041] Figure 1 This is a top view of the comparative test block A of the present invention.

[0042] Figure 2 This is a cross-sectional view of the protrusion of the comparative test block A of the present invention.

[0043] Figure 3 This is a diagram showing the relationship between paint layer thickness and compensation gain in an application example of the present invention.

[0044] Figure 4 This is a graph showing the relationship between the convex angle tan(θ) and the gain value corresponding to 80% amplitude in an application example of the present invention.

[0045] Figure 5 This is a real-world scene image of the detection site, illustrating an application example of the present invention. Detailed Implementation

[0046] The specific embodiments of the present invention will be further described in detail below with reference to the accompanying drawings.

[0047] like Figure 1-2 As shown, the present invention provides a method for detecting dents in wind turbine towers based on phased array technology, comprising the following steps:

[0048] Step 1: Prepare control test blocks

[0049] Comparison block A:

[0050] Since tower dents mainly occur on the outer wall of the tower, which is suspended and cannot be measured, dent detection can be performed by inspecting the convex detection area in the opposite direction. A control block A containing the convex area should be prepared, with at least 10 control blocks A prepared. A schematic diagram of control block A is shown below. Figure 1 , 2 As shown, each control block A can be equipped with a nameplate, which is placed in the lower right corner for identification. Control block 1 measures 1000×1000×15mm, and each control block 1 has a protrusion in the center, such as... Figure 1 , 2 As shown, the Y direction is defined along the length of the test block, the X direction is defined along the width of the test block, the Z direction is defined as the vertical upward direction, and the angle θ is defined as the inclination angle of the tower protrusion. The difference between the 10 comparative test blocks 1 is that the inclination angle θ of the protrusion is different.

[0051] The dimensions of the protrusions are shown in Table 1:

[0052] Table 1. Protrusion Dimensions

[0053]

[0054]

[0055] In the table, X represents the width of the protrusion and Y represents the length of the protrusion. All protrusions have the same shape and size. The top view of the protrusion can be circular or elliptical, and all protrusions are located at the center of the control block A.

[0056] Comparison with test block B:

[0057] Multiple control test blocks B are prepared. Each control test block B has a cuboid structure and dimensions of d×e×f, where d is the length of the test block (at least 500 mm), e is the width of the test block (greater than 100 mm), and f is the height of the test block (greater than 10 mm). The surfaces of the control test blocks are coated with paint layers of different thicknesses, namely g1, g2, g3, g4, g5, etc. The thicknesses can be increased or decreased as needed, but at least three different paint layer thicknesses are included (i.e., at least three control test blocks B are prepared).

[0058] Step 2: Probe Selection

[0059] The scanning method in this application is to inspect along the height of the tower. Since the tower diameter is relatively large, it can be regarded as a planar structure. The probe surface does not need to be ground to a curved surface; a flat surface is sufficient.

[0060] A phased array probe is selected. Ultrasonic phased array equipment often uses 5L32-0.5×10 linear array probes for detection. However, in order to improve the directivity of the ultrasonic beam, make the focusing effect more obvious, and more accurately control the deflection angle and focusing position of the beam, so as to more accurately locate and evaluate defects at different depths and angles, achieve more refined beam scanning and dynamic focusing, cover a larger detection area, and reduce detection blind spots, a 10L64-0.5×10 linear array probe is used. Here, 10 represents the center frequency of the probe, 10MHz; L represents the linear array probe; 64 represents the number of array elements, which is 64; 0.5×10 represents the center-to-center spacing of the array elements, which is 0.5mm and the spacing between adjacent array elements, which is 0.1mm; and 10 represents the length of the array element, which is 10mm.

[0061] Probe incident angle determination: When a longitudinal wave is incident obliquely at the interface between two media, a wave mode conversion will occur, generating transverse waves, etc. As the longitudinal wave incident angle increases, the transverse wave refraction angle in the second medium will also increase. When the transverse wave refraction angle reaches 90°, there is no transverse wave propagation in the second medium. At this time, the ultrasonic wave will propagate along the surface of the medium, forming a surface wave. The longitudinal wave incident angle that makes the transverse wave refraction angle reach 90° is the angle at which the surface wave is generated.

[0062] Finally, it was determined that when the longitudinal wave is incident on the steel surface at an angle of approximately 63° in the plexiglass, the angle of refraction in the steel is 90°; therefore, an N63-90S surface wave wedge was used, and a displacement sensor was installed inside the probe to record the probe's movement distance; Step 3: Calibration Test

[0063] 3.1 Paint layer calibration test

[0064] The surface wave gain compensation for different paint layer thicknesses was calculated. Comparison block B was selected, and the paint layer thicknesses were g1, g2, g3, g4, and g5, respectively. Surface waves were emitted on block B, and the bottom wave amplitude was adjusted to 80% to obtain the compensation gain for different paint layer thicknesses, as shown in Table 2 below.

[0065] Table 2 Compensation gain values ​​for different paint layer thicknesses

[0066] Paint layer thickness Compensation gain (dB) g1 i1 g2 i2 g3 i3 g4 i4 g5 i5

[0067] For the compensation gain corresponding to different paint layer thicknesses (as shown in the table above), the relationship between paint layer thickness g and compensation gain i is plotted using Origin software. The formula fitting method in Origin is used to fit a formula that conforms to the relationship between g and i. The formula is required to have a Pearson (correlation coefficient) of 0.97 or higher and an R-Square (fitness) of 0.97 or higher. This is the paint layer compensation gain equation.

[0068] 3.2 Inclination Angle Calibration Test

[0069] Place the probe at the upper left edge of the control block A (e.g., ...). Figure 1 At the origin O in the upper left corner of the test block A, a surface wave is emitted. The upper left corner of the test block A is defined as the origin, i.e., point (0,0). The scanning path is as follows: the probe first starts from point (0,0) and scans along the positive Y-axis (right) towards the edge of the raised area of ​​the test block A until it reaches the edge of the raised area of ​​the test block (e.g., ...). Figure 1 (As shown at L in the middle), the probe then scans along the negative Y-axis from the edge of the raised area of ​​the comparison block A with 40-50% overlap with the previous scan path in the Y-axis direction until the left edge of the comparison block A. This process is repeated until the probe reaches the lower edge of the comparison block A and the scan is completed. The scanning speed is no more than 150 mm / s. When scanning within the raised area of ​​the surface, there will be reflected waves at the raised area. The intensity of the reflected wave is related to the raised angle θ. The amplitude of the reflected wave is adjusted to 80% to obtain the minimum gain corresponding to different raised angles and 80% amplitude, as shown in Table 3 below.

[0070] Table 3 Gain values ​​corresponding to different protrusion angles

[0071]

[0072]

[0073] Using Origin software, plot the relationship between the convexity angle tan(θ) and the gain value h corresponding to 80% amplitude. Fit a formula that conforms to the relationship between tan(θ) and h. The formula should have a Pearson correlation coefficient of 0.97 or higher and an R-squared goodness of fit of 0.97 or higher. This is the gain equation for the convexity angle tan(θ) and the gain value h corresponding to 80% amplitude.

[0074] Step 4: On-site testing

[0075] The X-direction is defined as the horizontal direction of the tower, the Y-direction as the height direction of the tower, and the Z-direction as the direction of the tower's bulge. The tower is composed of different sections, which are connected by welds. The scanning method is to first scan the bottom of the tower to weld H1, then scan welds H1-H2, welds H2-H3, and so on until weld Hn to the top of the tower, thus completing the scanning.

[0076] Taking the scanning between welds H1 and H2 as an example, place the probe at any position on the upper edge of weld H1, scan upwards along the Y direction, and stop at the lower edge of weld H2. Then reverse the probe so that it overlaps with the previous scanning path by 40-50%, and scan downwards along the Y direction until the upper edge of weld H1 is reached. Continue scanning until the initial position is reached.

[0077] When scanning areas without protrusions or depressions, the surface wave reflects the bottom wave without defects. If a surface protrusion is found, a reflected wave will be generated at the protrusion's starting point. Fix the amplitude at this point to 80%. After finding the point with the minimum gain, fix the gain value at this point to hxdB. Record the probe position at this point as (1,1).

[0078] 4.1 Defect X-direction dimension measurement

[0079] The probe moves horizontally along the positive X direction from the position of (1,1) and the amplitude gradually decreases. When there is no reflected surface wave, the probe moves a distance Xa along the positive X direction. The probe moves along the negative X direction from the position of (1,1) and the amplitude gradually decreases. When there is no reflected surface wave, the probe moves a distance Xb along the negative X direction. Xa + Xb is the size of the defect in the X direction.

[0080] 4.2 Defect Y-direction dimension measurement

[0081] Starting from the position (1, 1), move horizontally along the positive X direction until the wave amplitude disappears. Return to the position (1, 1) and move horizontally along the negative X direction until the wave amplitude disappears again. Record the minimum x-coordinate of the reflected wave during this process, which is the minimum forward propagation distance of the surface wave, denoted as Y1. min ;

[0082] Place the probe on the opposite side of the convex area along the tower height direction, fix the defect wave amplitude at 80%, find the point of minimum gain, fix the gain value at this moment, and record the probe position at this moment; denoted as position (2,2); move horizontally along the positive X direction from position (2,2) until the wave amplitude disappears, return to position (2,2), and move horizontally along the negative X direction until the wave amplitude disappears. Record the minimum abscissa of the reflected wave during this process, that is, the minimum forward propagation distance of the surface wave, denoted as Y2. min ;

[0083] Measure the Y-direction distance between the two probes along the height of the tower, and denot it as Y3;

[0084] The dimension in the Y direction is calculated as follows: Y = Y3 - Y1 minn -Y2 min ;

[0085] 4.3 Defect Z-direction dimension measurement

[0086] Measure the paint layer thickness of the tower to determine the paint layer compensation gain value. Measure the paint layer thickness g and substitute it into the paint layer compensation gain equation obtained in step 3.1 to obtain the paint layer compensation gain value i for thickness g.

[0087] With the paint layer compensation idB added, the total gain is h = hx + i;

[0088] Substituting the total gain value h into step 3.2, we obtain the gain equation for the gain value h corresponding to the convex angle tan(θ) and the amplitude 80%, thus obtaining tan(θ);

[0089] tan(θ) = Z / (Y / 2)

[0090] The defect height, i.e., the Z value, is obtained.

[0091] This allows us to obtain the location and size of the tower protrusion defect.

[0092] The present invention has achieved good technical results in practical applications, and the application examples are as follows:

[0093] Step 1: Prepare control test blocks

[0094] Comparison block A:

[0095] Since tower dents mainly occur on the outer wall of the tower, which is suspended and cannot be measured, dent detection can be performed by checking the bulge detection area in the opposite direction. A bulge test block was designed with dimensions of 1000×1000×15mm. The bulge dimensions are shown in Table 4. A schematic diagram of the test block is shown below. Figure 1 , 2 As shown:

[0096] Table 4 Specimen Dimensions

[0097] test block X(mm) Y(mm) Z(mm) tan(θ) Test Block 1 250 250 1 0.008 Test Block 2 250 250 2 0.016 Test Block 3 250 250 3 0.024 Test Block 4 250 250 4 0.032 Test Block 5 250 250 5 0.04 Test Block 6 250 250 6 0.048 Test Block 7 250 250 7 0.056 Test Block 8 250 250 8 0.064 Test Block 9 250 250 9 0.072 Test Block 10 250 250 10 0.08

[0098] Comparison with test block B:

[0099] The test blocks were 500×200×15mm in size and coated with paint layers of different thicknesses, namely 50μm, 100μm, 150μm and 200μm.

[0100] Step 2: Probe Selection

[0101] The probe selected is a 10L64-0.5×10 linear array probe, matched with an N63-90S surface wave wedge block. A displacement sensor is installed inside the probe to record the probe's movement distance.

[0102] Step 3: Calibration Test

[0103] 3.1 Paint layer calibration test

[0104] Using control block B, the surface wave gain compensation for different paint layer thicknesses was calculated. Surface waves were emitted onto control block B, and the bottom wave amplitude was adjusted to 80% to obtain the gain compensation for different paint layer thicknesses. See Table 5 below.

[0105] Table 5 Compensation gain values ​​for different paint layer thicknesses

[0106] Paint layer thickness Compensation gain (dB) 50 0.21 100 0.40 150 0.59 200 0.81 250 1.02

[0107] Using the values ​​in the table, plot the relationship between paint layer thickness g and compensation gain i using Origin software (e.g., ...). Figure 3 As shown), by using the formula fitting method in Origin, a formula that conforms to the relationship between g and i is fitted, such as... Figure 4 As shown; the relation is:

[0108] i = 0.00405g - 0.00143 (4-1) The Pearson correlation coefficient is 0.9997 and the R-Square fit is 0.9994, therefore the formula meets the requirements.

[0109] 3.2 Inclination Angle Calibration Test

[0110] Place the nameplate of the comparison block A in the lower right corner. Define the length direction of the block as the Y direction, the width direction as the X direction, and the vertical upward direction as the Z direction. Define the upper left corner of the comparison block A as point (0,0). The probe first starts from point (0,0) and scans the comparison block A along the positive Y-axis until it reaches the edge of the convex area of ​​the block. Then, the probe scans along the negative Y-axis from the edge of the simulated block until it reaches the edge of the simulated block, with 50% overlap with the previous scan path. Repeat the above steps until the scan is complete. The scanning speed is no more than 150 mm / s. When scanning to the surface convexity, there will be a reflected wave at the convexity. The intensity of the reflected wave is related to the convexity angle (θ angle). Adjust the reflected wave amplitude to 80% to obtain the minimum gain corresponding to different convexity angles and 80% amplitude, as shown in Table 6 below.

[0111] Table 6. Gain values ​​corresponding to different protrusion angles.

[0112]

[0113] Using Origin software, plot the relationship between the convexity angle tan(θ) and the gain value h corresponding to 80% amplitude (e.g., ...). Figure 4 As shown), by using the formula fitting method in Origin, a formula that conforms to the relationship between tan(θ) and h is fitted, as follows: Figure 5 As shown; the relation is:

[0114] h = -454.4tan(θ) + 99.54 (4-2) The Pearson correlation coefficient is 0.99752 and the R-Square fit is 0.99504, so the formula meets the requirements.

[0115] Step 4: On-site testing

[0116] The X-direction is defined as the horizontal direction of the tower, the Y-direction as the height direction of the tower, and the Z-direction as the convex direction of the tower. The tower is composed of different sections, which are connected by welds. The scanning method is to first scan the bottom of the tower to weld H1, then scan welds H1-H2, welds H2-H3, and so on until weld Hn to the top of the tower, thus completing the scan. Taking the scan between welds H1 and H2 as an example, place the probe at any position on the upper edge of weld H1 and scan upwards along the Y-direction until the lower edge of weld H2 is reached. Then, reverse the probe so that it overlaps with the previous scan path by 40-50% and scans downwards along the Y-direction until the upper edge of weld H1 is reached. This process continues until the initial position is reached.

[0117] When scanning areas without depressions (protrusions), the surface wave reflects the bottom wave without defects. If a surface protrusion is found, a reflected wave will be generated at the protrusion's starting point. Fix the amplitude at 80% at this point, find the point with the minimum gain, and fix the gain value at this point to be 74.44dB. Record the probe position at this point as position (1,1).

[0118] 4.1 Defect X-direction dimension measurement

[0119] Moving horizontally along the positive X direction from position (1,1), the amplitude gradually decreases. When there is no reflected surface wave, the recording probe moves 220mm along the positive X direction. Moving along the negative X direction from position (1,1), the amplitude gradually decreases. When there is no reflected surface wave, the recording probe moves 230mm along the negative X direction. The dimension of the defect in the X direction is 220 + 230 = 450mm.

[0120] 4.2 Defect Y-direction dimension measurement

[0121] Starting from position (1,1), move horizontally along the positive X direction until the amplitude disappears. Return to position (1,1) and move horizontally along the negative X direction until the amplitude disappears. Record the minimum abscissa of the reflected wave during this process, which is the minimum forward propagation distance of the surface wave, denoted as 350mm.

[0122] Place the probe on the opposite side of the convex area along the height of the tower. Fix the defect wave amplitude at 80%. Find the point of minimum gain and fix the gain value at this moment. Record the probe position at this moment as (2,2). From the recorded position (2,2), move horizontally along the positive X direction until the wave amplitude disappears. Return to the position (2,2) and move horizontally along the negative X direction until the wave amplitude disappears. Record the minimum abscissa of the reflected wave during this process, which is the minimum forward propagation distance of the surface wave, and record it as 350mm.

[0123] The distance between the two probes along the height of the tower in the Y direction is measured to be 1000 mm.

[0124] The Y-direction dimension is calculated as follows: Y = 1000 - 350 - 350 = 300 mm.

[0125] 4.3 Defect Z-direction dimension measurement

[0126] The thickness of the paint layer on the tower was measured to determine the paint layer compensation gain value. Referring to the design drawings, the tower has a double-layer paint coating: an epoxy zinc-rich primer with a thickness of 50 micrometers, and an epoxy intermediate coat with a thickness of 90 micrometers, for a total thickness of 140 micrometers. Substituting this into the paint layer-compensation gain formula (4-1) obtained in 4.3.1, the compensation gain value for a 140-micrometer paint layer is 0.57 dB.

[0127] With the paint layer compensation of 0.57dB, 74.44 + 0.57 = 75.01dB.

[0128] Substituting into 4.3.2, we obtain the relationship between the convexity angle tan(θ) and the gain value h corresponding to 80% amplitude.

[0129] h = -454.4tan(θ) + 99.54

[0130] The tan(θ) is 0.054.

[0131] 0.054 = Z / (300 / 2)

[0132] Z = 8.1 mm

[0133] Therefore, the dimensions of the protrusion are 450×300×8.1mm. Figure 5 As shown.

[0134] V. Test Block Verification

[0135] In the application example, the third test block in Comparison Block A was selected for experimental verification in Step 1. The test block dimensions were 1000×1000×15mm, with protrusions of 250mm in the X direction, 250mm in the Y direction, and 3mm in the Z direction. A custom-designed 10L64-0.5×10 linear array probe and an N63-90S surface wave wedge were used. A displacement sensor was installed inside the probe to record the probe's movement distance.

[0136] Place the test block nameplate in the lower right corner. Define the length of the test block as the Y-axis, the width as the X-axis, and the vertical upward direction as the Z-axis. Define the upper left corner of the simulated test block as point (0,0). The probe first starts from point (0,0) and scans the simulated test block along the positive Y-axis until it reaches the edge of the raised area. Then, the probe scans along the negative Y-axis from the edge of the simulated test block until it reaches the edge of the simulated test block, with 50% overlap with the previous scan path. Repeat the above steps until the scan is complete. The scanning speed should not exceed 150 mm / s.

[0137] When scanning areas without depressions (protrusions), the surface wave reflects the bottom wave without defects. If a surface protrusion is found, a reflected wave will be generated at the protrusion's starting point. Fix the amplitude at 80% at this point, find the point with the minimum gain, and fix the gain value at this point to be 88.7dB. Record the probe position at this point as position (3,3).

[0138] 5.1 Step 1: Defect X-direction dimension measurement

[0139] Moving horizontally along the positive X direction from position (3,3), the amplitude gradually decreases. When there is no reflected surface wave, the recording probe moves 120mm along the positive X direction. Moving along the negative X direction from position (3,3), the amplitude gradually decreases. When there is no reflected surface wave, the recording probe moves 130mm along the negative X direction. The dimension of the defect in the X direction is 120mm + 130mm = 250mm. 5.2 Step Two: Measurement of Defect Dimensions in the Y Direction

[0140] Starting from position (3,3), move horizontally along the positive X direction until the amplitude disappears. Return to position (3,3) and move horizontally along the negative X direction until the amplitude disappears. Record the minimum abscissa of the reflected wave during this process, which is the minimum forward propagation distance of the surface wave, denoted as 300mm.

[0141] Place the probe on the opposite side of the convex area along the Y direction of the test block. Fix the defect wave amplitude at 80%. After finding the point of minimum gain, fix the gain value at this moment and record the probe position at this moment. Record this position as (4,4). From the recorded position (4,4), move horizontally along the positive X direction until the wave amplitude disappears. Return to the position (4,4) and move horizontally along the negative X direction until the wave amplitude disappears. Record the minimum abscissa of the reflected wave during this process, which is the minimum forward propagation distance of the surface wave, and record it as 300mm.

[0142] The distance between the two probes along the height of the tower in the Y direction was measured to be 850 mm.

[0143] The Y-direction dimension is calculated as follows: Y = 850 - 300 - 300 = 250 mm.

[0144] 5.3 Step 3: Defect Z-direction dimension measurement

[0145] With a fixed amplitude of 80%, the minimum gain value was 88.7 dB.

[0146] Substituting into 4.3.2, we obtain the relationship between the convexity angle tan(θ) and the gain value h corresponding to 80% amplitude.

[0147] h = -454.4tan(θ) + 99.54

[0148] The tan(θ) is 0.024.

[0149] 0.024 = Z / (250 / 2)

[0150] Z = 2.99 mm

[0151] Therefore, the dimensions of the protrusion are 250×250×2.99mm, which is basically consistent with the dimensions of test block 3, which are 250×250×3mm.

Claims

1. A method for detecting dents in wind turbine towers based on phased array technology, characterized in that, Includes the following steps: Step 1: Prepare control test blocks Comparison block A: Prepare a control block A with a protrusion. Prepare at least 10 control blocks A, each with a protrusion at its center. Define the length of the block as the Y direction, the width of the block as the X direction, and the vertical upward direction as the Z direction. Define the angle θ as the inclination angle of the tower protrusion. The 10 control blocks A differ in the inclination angle θ of the protrusion. Comparison with test block B: Multiple comparative test blocks B are prepared. Each comparative test block B has a cuboid structure and its surface is coated with a paint layer of different thicknesses. The paint layer thicknesses of the test blocks are g1, g2, g3, g4, g5, etc., and there are at least 5 different paint layer thicknesses. Step 2: Probe Selection The probe is a phased array probe, specifically a 10L64-0.5×10 linear array probe, where 10 represents the probe's center frequency of 10MHz, L represents a linear array probe, 64 represents the number of array elements of the probe is 64, 0.5×10 represents the center-to-center spacing of the array elements is 0.5mm and the spacing between adjacent array elements is 0.1mm, and 10 represents the length of the array element is 10mm. It is equipped with N63-90S surface wave wedge block, and a displacement sensor is installed inside the probe to record the probe movement distance; Step 3: Calibration Test 3.1 Paint layer calibration test The surface wave gain compensation for different paint layer thicknesses was calculated. Comparison test block B was selected, and surface waves were emitted on the comparison test block B. The bottom wave amplitude was adjusted to 80% to obtain the compensation gain for different paint layer thicknesses. For the compensation gain corresponding to different paint layer thicknesses, the relationship between paint layer thickness g and compensation gain i is plotted using Origin software. The formula fitting method in Origin is used to fit a formula that conforms to the relationship between g and i. The formula is required to have a Pearson (correlation coefficient) of 0.97 or higher and an R-Square (fitness) of 0.97 or higher. This is the paint layer compensation gain equation. 3.2 Tilt Angle Calibration Test Place the probe at the upper left edge of the comparison block A and emit surface waves. The upper left corner of the comparison block A is defined as the origin of the coordinate system, i.e., point (0,0). The scanning path is as follows: the probe first starts from point (0,0) and scans the comparison block A along the positive Y-axis until it reaches the edge of the protruding area of ​​the block. Then, the probe scans along the negative Y-axis from the edge of the protruding area of ​​the comparison block A with 40-50% overlap with the previous scanning path in the Y-axis direction until it reaches the left edge of the comparison block A. This cycle is repeated until the probe reaches the lower edge of the comparison block A, and the scanning is completed. The scanning speed is no more than 150 mm / s. When scanning within the protruding area of ​​the surface, there will be reflected waves at the protrusions. The intensity of the reflected waves is related to the tilt angle θ of the protrusions. Adjust the amplitude of the reflected waves to 80% to obtain the minimum gain corresponding to the tilt angle of different protrusions and the amplitude of 80%. Using Origin software, plot the relationship between the tangent of the tilt angle θ (tan(θ)) and the gain value h corresponding to 80% amplitude. Fit a formula that satisfies the relationship between tan(θ) and h. The formula should have a Pearson correlation coefficient of 0.97 or higher and an R-squared goodness of fit of 0.97 or higher. This is the gain equation between the tangent of the tilt angle θ (tan(θ)) and the gain value h corresponding to 80% amplitude. Step 4: On-site testing The X-direction is defined as the horizontal direction of the tower, the Y-direction as the height direction of the tower, and the Z-direction as the direction of the tower's bulge. The tower is composed of different sections, which are connected by welds. The scanning method is to first scan the bottom of the tower to weld H1, then scan welds H1-H2, welds H2-H3, and so on until weld Hn to the top of the tower, thus completing the scanning. When scanning areas without protrusions or depressions, the surface wave reflects the bottom wave without defects. If a surface protrusion is found, a reflected wave will be generated at the protrusion's starting point. Fix the amplitude at this point to 80%, find the point with the minimum gain, and fix the gain value at this point to hxdB. Record the probe position at this point as (1,1). 4.1 Defect X-direction dimension measurement The probe moves horizontally along the positive X direction from the position of (1,1) and the amplitude gradually decreases. When there is no reflected surface wave, the probe moves a distance Xa along the positive X direction. The probe moves along the negative X direction from the position of (1,1) and the amplitude gradually decreases. When there is no reflected surface wave, the probe moves a distance Xb along the negative X direction. Xa + Xb is the size of the defect in the X direction. 4.2 Defect Y-direction dimension measurement Starting from the position (1,1), move horizontally along the positive X direction until the wave amplitude disappears. Return to the position (1,1) and move horizontally along the negative X direction until the wave amplitude disappears again. Record the minimum x-coordinate of the reflected wave during this process, which is the minimum forward propagation distance of the surface wave, denoted as Y1. min ; Place the probe on the opposite side of the convex area along the height of the tower. Fix the defect wave amplitude at 80%. Find the point of minimum gain and fix the gain value at this moment. Record the probe position at this moment as (2,2). Move the probe horizontally along the positive X direction from the recorded position (2,2) until the wave amplitude disappears. Return to the position (2,2) and move it horizontally along the negative X direction until the wave amplitude disappears. Record the minimum abscissa of the reflected wave during this process, which is the minimum forward propagation distance of the surface wave, and record it as Y2. min ; Measure the Y-direction distance between the two probes along the height of the tower, and denot it as Y3; The dimension in the Y direction is calculated as Y = Y3 - Y1. minn -Y2 min ; 4.3 Defect Z-direction dimension measurement Measure the paint layer thickness of the tower to determine the paint layer compensation gain value. Measure the paint layer thickness g and substitute it into the paint layer compensation gain equation obtained in step 3.1 to obtain the paint layer compensation gain value i for thickness g. With the paint layer compensation idB added, the total gain is h = hx + i; Substituting the total gain value h into step 3.2, we obtain the gain equation of the tangent value tan(θ) of the tilt angle θ of the convexity and the gain value h corresponding to 80% of the amplitude, thus obtaining tan(θ); tan(θ) = Z / (Y / 2) The defect height, i.e., the Z value, is obtained. This allows us to obtain the location and size of the tower protrusion defect.

Images