Low-Frequency Electromagnetic Detection Method For Large-Scale Damage Of Ferromagnetic Materials Based On Broadband Excitation

Abstract

The invention discloses a low-frequency electromagnetic detection method for large-scale damage of ferromagnetic materials based on broadband excitation. Detection direction of the magnetic field signal of low-frequency electromagnetic sensor is determined according to the size of ferromagnetic member detection defect; the reference signal and detection signal acquisition position are selected, fix the distance between sensor and tested part, excite a Chirp signal as a broadband excitation signal to perform broadband excitation low-frequency electromagnetic detection; the computer processes collected broadband detection signal; use the difference of Euclidean distance between reference signal and defect detection signal as a defect characterization parameter to obtain the Euclidean distance curve of different depth defects on the upper and lower surfaces of ferromagnetic components with the detection position. Through the analysis and processing of the low-frequency electromagnetic broadband detection signal, the Euclidean response signal and reference signal under broadband excitation are used to characterize the change of material damage degree, which can effectively reduce the influence of magnetic field skin effect, and is beneficial to the effective characterization of the upper and lower material surface defects of at different depths.

Claims

1. A low-frequency electromagnetic detection method for large-scale damage of ferromagnetic materials based on broadband excitation, comprising the steps of: 1) a piece of ferromagnetic steel plate is selected as a test piece; wherein the thickness of the ferromagnetic steel plate is between 12 mm-16 mm; wherein there are crack defects of different depths on the surface of the ferromagnetic steel plate; wherein each crack defect has the same dimensions except for the depth; and wherein the maximum depth is less than the thickness of the ferromagnetic steel plate; 2) adjust the signal pickup direction of a magnetic sensor inside a low-frequency electromagnetic sensor to make the detection direction parallel to a surface of the test piece; wherein the detection result is the tangential leakage magnetic field strength on the surface of the test piece; wherein magnetic field strength in this direction is sensitive to the depth of defect; 3) place the low-frequency electromagnetic sensor on the surface of the ferromagnetic steel plate so that the low-frequency electromagnetic sensor is in a defect-free reference area of the ferromagnetic steel plate, and adjust the lift-off distance between the low-frequency electromagnetic sensor and the ferromagnetic steel plate to be less than 1 mm; 4) adjust a function generator to generate a Chirp signal with fixed output voltage and bandwidth for excitation, and activate a power amplifier; using a digital oscilloscope, simultaneously display an excitation signal and an electromagnetic detection signal of the low-frequency electromagnetic sensor in the defect-free reference area of the ferromagnetic steel plate; assign detected signal as a reference signal when there is no defect and collect the no-defect reference signal; 5) place the low-frequency electromagnetic sensor at one side of a crack, and under the same excitation conditions, manually control the moving direction of the low-frequency electromagnetic sensor so that the moving direction is perpendicular to the longitude of the crack; activate a power amplifier; whenever the low-frequency electromagnetic sensor is placed at a detection point, simultaneously display on the digital oscilloscope the excitation signal and electromagnetic detection signal of the low-frequency electromagnetic sensor at the detection point, collect the broadband detection signals of all detection points as defect detection signals; 6) process the collected reference signals and defect detection signals via a computer, by the steps of: 1, perform frequency domain analysis on reference signals and each defect detection signal to obtain the reference signal spectrum and the defect detection signal spectrum at different positions; 2, Euclidean distance calculation formula is used to obtain the frequency domain Euclidean distance between detection the signal and the reference signal at each detection point; 3, using the frequency domain Euclidean distance as a defect characteristic parameter to draw a curve of Euclidean distance versus detection position; 7) flip the tested piece and repeat the steps 3) to 6) on other surface of the test piece; obtain a curve of Euclidean distance versus detection position when defect appears below the other surface of the tested piece; and 8) based on the Euclidean distance curves, the surface defects of ferromagnetic steel plates at different depths are quantitatively characterized.

2. A low-frequency electromagnetic detection method for large-scale damage of ferromagnetic materials based on broadband excitation according to claim 1, further comprising: establish a theoretical calculation model for low-frequency electromagnetic detection magnetic fields at different frequencies based on the theory of magnetic dipoles as follows: with defects such as holes, cracks; and pits in the magnetized ferromagnetic material in a magnetization field H and the area density of magnetic charge is Q, theoretical formula of defect two-dimensional magnetic dipole model is: $\begin{array}{cc}{H}_x\left(x,y\right)=\frac{Q}{4{\mathrm{\pi \mu }}_0}\left\{\underset{-h_2}{\overset{-h_1}{\int }}\frac{\left(x+w\right)d{y}^{\prime }}{{\left[\begin{array}{c}{\left(x+w\right)}^2+\\ {\left(y-y^{\prime }\right)}^2\end{array}\right]}^{3/2}}-\underset{-h_2}{\overset{-h_1}{\int }}\frac{\left(x-w\right)d{y}^{\prime }}{{\left[\begin{array}{c}{\left(x-w\right)}^2+\\ {\left(y-y^{\prime }\right)}^2\end{array}\right]}^{3/2}}\right\}& \left(1\right)\\ H_y\left(x,y\right)=\frac{Q}{4{\mathrm{\pi \mu }}_0}\left\{\underset{-h_2}{\overset{-h_1}{\int }}\frac{\left(y-y^{\prime }\right)d{y}^{\prime }}{{\left[\begin{array}{c}{\left(x+w\right)}^2+\\ {\left(y-y^{\prime }\right)}^2\end{array}\right]}^{3/2}}-\underset{-h_2}{\overset{-h_1}{\int }}\frac{\left(y-y^{\prime }\right)d{y}^{\prime }}{{\left[\begin{array}{c}{\left(x-w\right)}^2+\\ {\left(y-y^{\prime }\right)}^2\end{array}\right]}^{3/2}}\right\}& \left(2\right)\end{array}$ wherein, Hx represents the tangential magnetic field intensity, Hy represents the normal magnetic field intensity, x represents the abscissa of the observation point P, y represents the ordinate of the observation point P, μ.sub.0 represents the permeability of the vacuum permeability, and h.sub.1 represents the upper surface of the defect buried depth, h.sub.2 represents the buried depth of the bottom surface, w represents half of the defect width, and y′ represents the distance from left and right groove wall micro-line elements dy to the material surface; wherein relationship between magnetic charge density at depth h below the surface of the test piece and its surface magnetic charge density can be approximately expressed as: $\begin{array}{cc}\frac{Q_h}{Q_0}=e^{-h\sqrt{\pi f\mu \sigma }}& \left(3\right)\end{array}$ wherein, Q.sub.h represents the magnetic charge density at distance h from the component surface, Q.sub.0 represents the magnetic charge density at the component surface, a represents the permeability of ferromagnetic material, u represents the conductivity of ferromagnetic material, and f represents the excitation signal frequency; wherein equivalent magnetic dipole model for low-frequency electromagnetic detection is: $\begin{array}{cc}{H}_x\left(x,y,f\right)=\frac{Q_0}{4{\mathrm{\pi \mu }}_0}\left\{\underset{-h_2}{\overset{-h_1}{\int }}\frac{e^{-y^{\prime }\sqrt{\pi f\mathrm{\mu \sigma }}}\left(x+w\right)d{y}^{\prime }}{{\left[\begin{array}{c}{\left(x+w\right)}^2+\\ {\left(y-y^{\prime }\right)}^2\end{array}\right]}^{3/2}}-\underset{-h_2}{\overset{-h_1}{\int }}\frac{e^{-y^{\prime }\sqrt{\pi f\mathrm{\mu \sigma }}}\left(x-w\right)d{y}^{\prime }}{{\left[\begin{array}{c}{\left(x-w\right)}^2+\\ {\left(y-y^{\prime }\right)}^2\end{array}\right]}^{3/2}}\right\}& \left(4\right)\\ H_y\left(x,y,f\right)=\frac{Q_0}{4{\mathrm{\pi \mu }}_0}\left\{\underset{-h_2}{\overset{-h_1}{\int }}\frac{e^{-y^{\prime }\sqrt{\pi f\mathrm{\mu \sigma }}}\left(y-y^{\prime }\right)d{y}^{\prime }}{{\left[\begin{array}{c}{\left(x+w\right)}^2+\\ {\left(y-y^{\prime }\right)}^2\end{array}\right]}^{3/2}}-\underset{-h_2}{\overset{-h_1}{\int }}\frac{e^{-y^{\prime }\sqrt{\pi f\mathrm{\mu \sigma }}}\left(y-y^{\prime }\right)d{y}^{\prime }}{{\left[\begin{array}{c}{\left(x-w\right)}^2+\\ {\left(y-y^{\prime }\right)}^2\end{array}\right]}^{3/2}}\right\}& \left(5\right)\end{array}$ such that the magnetic charge density of low-frequency electromagnetic detection is no longer uniformly distributed with the increase of defect depth, but with an exponential decay trend; wherein the Euclidean distance characterization parameter E(H.sub.d, H.sub.n) can be calculated by formula (6):
E(H.sub.d,H.sub.n)=√{square root over ((Σ(H.sub.di−H.sub.ni).sup.2))}  (6) wherein, H.sub.d represents magnetic field strength of the area to be detected, H.sub.n represents the reference magnetic field strength of the reference area, H.sub.di represents the magnetic field strength of the area to be detected at each frequency, and H.sub.ni represents the reference magnetic field strength of the reference area at each frequency.

3. A low-frequency electromagnetic detection method for large-scale damage of ferromagnetic materials based on broadband excitation according to claim 1, further comprising: providing a system that includes the function generator, the power amplifier, the digital oscilloscope, a stabilized current power supply and the low-frequency electromagnetic sensor 5; divide an output port of the function generator into two channels, one connected to the second channel of the digital oscilloscope for displaying broadband excitation signals, and the other connected to the input port of the power amplifier, connect an output end of the power amplifier to the input end of the low-frequency electromagnetic sensor, connect an output end of the low-frequency electromagnetic sensor to the first channel of the digital oscilloscope for displaying the electromagnetic signal detected by the low-frequency electromagnetic sensor, and finally, the positive and negative poles of the stabilized current power supply are respectively connected to the two power supply input terminals of low-frequency electromagnetic sensor for supplying power for the magnetic sensor inside the low-frequency electromagnetic sensor 5.

Description

DESCRIPTION OF DRAWINGS

[0033] FIG. 1 is a cross-sectional view of a rectangular slot two-dimensional magnetic dipole model;

[0034] FIG. 2 is a Euclidean distance defect characterization flowchart;

[0035] FIG. 3 is a low-frequency electromagnetic detection system diagram; wherein: 1. Function generator, 2. Power amplifier, 3. Digital oscilloscope, 4. Stabilized current power supply, 5. Low frequency electromagnetic sensor;

[0036] FIG. 4a is a schematic diagram of upper surface defects scanning and detecting;

[0037] FIG. 4b is a schematic diagram of lower surface defects scanning and detecting;

[0038] FIG. 5a is diagram showing characterization results of broadband Euclidean distance on the upper surface defect;

[0039] FIG. 5b is a diagram showing characterization results of broadband Euclidean distance on the lower surface defect.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0040] The present invention will be further described below in conjunction with preferred embodiment:

[0041] The implementation process includes the following steps:

[0042] Construction of the experimental system:

[0043] The system used in the present invention is built based on in FIG. 3, including a function generator 1, a power amplifier 2, a digital oscilloscope 3, a stabilized current power supply 4 and a low-frequency electromagnetic sensor 5. First, divide the output port of the function generator 1 into two channels, one connected to the second channel of digital oscilloscope 3 for displaying broadband excitation signals, and the other connected to the input port of the power amplifier 2 for magnetizing DUT. Then, output end of the power amplifier 2 is connected to the input end of the low-frequency electromagnetic sensor 5, and its output end is connected to the first channel of the digital oscilloscope 3 for displaying the electromagnetic signal detected by the low-frequency electromagnetic sensor 5. Finally, the positive and negative poles of stabilized current power supply 4 are respectively connected to the two power supply input terminals of low-frequency electromagnetic sensor 5 for supplying the power for magnetic sensor inside low-frequency electromagnetic sensor 5.

[0044] Magnetic sensor detection direction selection: the test piece is 20# low carbon steel plate, a 620×400×12 mm commonly used ferromagnetic material. The steel plate has four standard artificial defects with 25 mm long and 4 mm wide, the depth of these four defects are 4.8 mm, 6.0 mm, 7.2 mm and 9.6 mm. Since the width of the defect is equal and depth is different, the detection direction of magnetic sensor is adjusted to be parallel to the tangential direction on the surface of tested steel plate, so that the electromagnetic detection signal is more sensitive to depth.

[0045] Sensor parameter selection: Place the sensor on defect-free side and defective side of the tested steel plate surface, manually control the sensor to maintain a lift-off distance of 1 mm from steel plate surface. Adjust the function generator, set excitation signal to a broadband Chirp signal, control the voltage to 3V, and the frequency bandwidth to 0-150 Hz.

[0046] Electromagnetic testing experiment: Place the sensor in non-defective reference area, turn on the power amplifier, and use digital oscilloscope to record and collect the non-defective reference signal. According to FIG. 4a and FIG. 4b, select a 30 mm linear distance perpendicular to the length of defect to make the sensor moves from left to right within the selected linear distance during detect. The detection step length is controlled to 1 mm, and the detection speed changes manually. Use digital oscilloscope to record the excitation signal and electromagnetic detection signal of a single position, and store the defect detection signal of 31 points during movement.

[0047] The computer processes the collected reference signals and the defect detection signals. First, perform frequency domain analysis on the reference signal and each detection signal to obtain reference signal spectrum and detection signal spectrum at different positions. Then Euclidean distance calculation formula (6) is used to obtain the frequency domain Euclidean distance between detection signal and reference signal at each detection point, and the distance is used as the defect characteristic parameter to draw the curve of Euclidean distance of the surface defect at different depths with the detection position (As shown in FIG. 5a) and the curve of the Euclidean distance between the lower surface defects and the detection position (as shown in FIG. 5b).

[0048] Analysis of experimental results: the depth of each defect and the position of the defect is known. It can be seen from FIG. 5a and FIG. 5b that the peaks in two figures can be judged, and the application of broadband excitation low-frequency electromagnetic detection method can effectively distinguish different depth defects on the upper and lower surfaces of steel plate. For four defects with equal width and different depth, the Euclidean distance detection amplitude on the upper surface is 169.2, 193.0, 212.4, 224.3; the Euclidean distance detection amplitude on the lower surface is 65.36, 75.11, 84.38, 89.33. The detection amplitude of each defect on the upper and lower surfaces is quite different, and the defect separation rate is higher.

[0049] While the invention has been particularly shown and described as referenced to the embodiments thereof, those skilled in the art will understand that the foregoing and other changes in form and detail may be made therein without departing from the spirit and scope of the invention.

REFERENCE

[0050] [1] Cunfu H, Yuegang H, Jingpin J, et al. Array of Fundamental Torsional Mode EMATs and Experiment in Thick-wall Pipe with Small Diameter[J]. Journal of Mechanical Engineering, 2015, 51(2):14-20. [0051] [2] Wang Z, Luo Q, Shi Y, et al. Analysis of eddy current coil coupling in ferromagnetic pipe and pipe's parameter detection[J]. Yi Qi Yi Biao Xue Bao/Chinese Journal of Scientific Instrument, 2014, 35(12):2843-2851. [0052] [3] Sharatchandra Singh W, Rao B P C, Vaidyanathan S, et al. Detection of leakage magnetic flux from near-side and far-side defects in carbon steel plates using a giant magneto-resistive sensor[J]. Measurement Science & Technology, 2008, 19(1):1-8. [0053] [4] Sun Y, Feng B, Liu S, et al. A Methodology for Identifying Defects in the Magnetic Flux Leakage Method and Suggestions for Standard Specimens[J]. Journal of Nondestructive Evaluation, 2015, 34(20):1-9. [0054] [5] Ying T, Mengchun P, Feilu L, et al. Detection of Corrosion in Pipeline Using Pulsed Magnetic Flux Leakage Testing[J]. Computer Measurement & Control, 2010, 18(1):38-43. [0055] [6] Gotoh Y, Takahashi N. Study on problems in detecting plural cracks by alternating flux leakage testing using 3D nonlinear eddy current analysis[J]. IEEE Transactions on Magnetics, 2003, 39(3): 1527-1530. [0056] [7] Singh W S, Rao B P C, Thirunavukkarasu S, et al. Flexible GMR sensor array for magnetic flux leakage testing of steel track ropes[J]. Journal of Sensors, 2012(2012): 1-6. [0057] [8] Singh W S, Rao B P C, Jayakumar T, et al. Simultaneous measurement of tangential and normal component of leakage magnetic flux using GMR sensors[J]. Journal of Non-Destructive Testing & Evaluation, 2009, 8(2): 23-28. [0058] [9] Jingpin J, Yu C, Guanghai L, et al. Study on low frequency AC magnetic flux leakage detection for internal and external cracks of ferromagnetic structures[J]. Chinese Journal of Scientific Instrument, 2016, 37(8):1808-1818. [0059] [10] Hosseini S, Lakis A A. Application of time-frequency analysis for automatic hidden corrosion detection in a multilayer aluminum structure using pulsed eddy current[J]. Ndt & E International, 2012, 47(2):70-79.