3D defect detection method with magnetic flux leakage testing

Abstract

The present invention discloses a 3D defect detection method with magnetic flux leakage testing (MFLT). It has advantages of higher accuracy of 3D detection of defect and simpler testing device relative to the prior MFLT art. This method includes the following steps: S1: artificially magnetizing a to-be-tested structure, and measuring its MFLT signals {B}; S2: inverting magnetic charge distribution of the interior of the to-be-tested structure by using a magnetic charge distribution reconstruction algorithm to obtain the magnetic charge density of a non-defective region of the to-be-tested structure; and S3: using the magnetic charge density of the non-defective region of the to-be-tested structure as a known constant, and conducting inverse iteration to reconstruct defect depth of the defective region to obtain a 3D image of the defective region of the to-be-tested structure.

Claims

1. A 3D imaging method for defect using a reconstruction algorithm with the signals detected by a magnetic flux leakage testing (MFLT) device, wherein the method comprises the following steps: S1: applying magnetic field to a to-be-tested structure and acquiring associated MFLT signals {B} via the MFLT device; S2: reconstructing its distribution of magnetic charge density inside the to-be-tested structure by using an algorithm, and extracting the magnetic charge density of a non-defective region near the boundary of the defective region; S3: intercepting the MFLT signals over a defective region, and reconstructing a depth field of the defect by applying an iteration algorithm by assuming the magnetic charge density over the entire structure to be a constant value, which is obtained from S2 above, the depth field forming a 3D image of the defect, where S3 includes: S31: according to a magnetic charge distribution theory, and on the premise of knowing the magnetic charge density of the to-be-tested structure, the normal components of the MFLT signals at field point r.sub.j outside a plate, {B.sub.z(r.sub.j)}, can be expressed as $\begin{array}{cc}{B}_z\left(r_j\right)=\frac{{\mu }_0\rho v}{4\pi }\underset{i=1}{\overset{n}{.Math.}}\frac{d_i+h}{{\left(\sqrt{{\left(x_i-x_j\right)}^2+{\left(y_i-y_j\right)}^2+{\left(d_i+h\right)}^2}\right)}^3}& \left(1\right)\end{array}$ where μ.sub.0 is the air permeability, v is the finite element volume of a source point r′.sub.i, ρ is the magnetic charge density of the to-be-tested structure, h is the lift-off for the magnetic sensor, which is equal to the distance in the z-axis direction between the field point r.sub.j and the top surface of the plate, x.sub.j, x.sub.i, y.sub.j, y.sub.i respectively are coordinates of the field point r.sub.j and the source point r′.sub.i, and d.sub.i is the associated depth of a source point r.sub.i′ on the defect surface, i is the number of source point, j is the number of field point, and n is the total number of source point, S32: all variables d.sub.i are collected as the defect depth field {d.sub.i}, and h is used as an initial value of {d.sub.i}, $\begin{array}{cc}{B}_z\left(r_j\right)=\frac{{\mu }_0\rho v}{4\pi }\underset{i=1}{\overset{n}{.Math.}}\frac{d_i}{{\left(\sqrt{{\left(x_i-x_j\right)}^2+{\left(y_i-y_j\right)}^2+d_i^2}\right)}^3}& \left(2\right)\end{array}$ S33: according to step S32, a correlation equation of the MFLT signal {B.sub.z(r.sub.j)} and the defect depth field {d.sub.i} and a coefficient matrix k.sub.d(i,j) is constructed, that is, $\begin{array}{cc}\left\{B_z\left(r_j\right)\right\}=\left[k_d\left(i,j\right)\right]\left\{d_j\right\}\left\{B_z\left(j\right)\right\}=\left[k_d\left(i,j\right)\right]\left\{d_i\right\},\mathrm{and}& \left(3\right)\\ k_d\left(i,j\right)=\frac{\rho {\mu }_{0}v}{4\pi }\left(\frac{1}{{\left(\sqrt{{\left(x_i-x_j\right)}^2+{\left(y_i-y_j\right)}^2+d_i^2}\right)}^3}-\frac{3d_i^2}{{\left(\sqrt{{\left(x_i-x_j\right)}^2+{\left(y_i-y_j\right)}^2+d_i^2}\right)}^5}\right);& \left(4\right)\end{array}$ S34: {B.sub.z(r.sub.j)}, {d.sub.i}, and [k.sub.d(i,j)] are abbreviated as {B}, {d} and [k.sub.d], respectively; and according to the correlation equation of the MFLT signal {B}, the defect depth field {d} and the coefficient matrix [k.sub.d], inverse iteration is conducted to obtain the defect depth field {d} of the defective region.

2. The 3D defect reconstruction method with MFLT signals according to claim 1, wherein in step S34, an inverse iteration process of the defect depth field {d} and the depth coefficient matrix [k.sub.d] is as follows: St0: setting an initial value of {d}, iteration terminating condition ε.sub.end, calculating [k.sub.d] with given initial {d} according to Eq. (4) St1: according to the formula {B}=[k.sub.d]{d}, using the given {d} and [k.sub.d] to calculate corresponding {B}, and obtaining a standard difference ε between the calculated MFLT signal value {B} and an intercepted MFLT signal value {B.sub.tar} over a to-be-tested defective area as shown in S3; St2: comparing ε with ε.sub.end, if c is greater than ε.sub.end, updating {d} and [k.sub.d], repeating step St1 till ε is less than ε.sub.end, and iteration termination to obtain the defect depth field {d}.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) The accompanying drawings are provided for further understanding of the present invention, constitute a part of the specification, are intended to explain the present invention with the embodiments of the present invention, but do not constitute limitations to the present invention. In the accompanying drawings:

(2) FIG. 1 is a schematic diagram showing steps of the present invention.

(3) FIG. 2 is an inverse iteration process of a defect depth field and a depth coefficient matrix in the present invention.

(4) FIG. 3 is a schematic structural diagram of a V-shaped defect of a to-be-tested structure.

(5) FIG. 4 shows MFLT signals of the to-be-tested structure shown in FIG. 3.

(6) FIG. 5 is a reconstructed magnetic charge distribution diagram of the to-be-tested structure.

(7) FIG. 6 is a reconstructed depth field distribution diagram of a defective region of the to-be-tested structure.

DETAILED DESCRIPTION

(8) As shown in FIG. 1, embodiments of the present invention provide a 3D defect detection method with MFLT, which includes the following steps:

(9) S1: conduct artificial magnetization on the to-be-tested structure, and measure its MFLT signals, as shown in FIG. 4;

(10) S2: invert magnetic charge distribution of the interior of the to-be-tested structure by using a magnetic charge distribution reconstruction algorithm to obtain a reconstructed magnetic charge distribution of the to-be-tested structure, as shown in FIG. 5; and

(11) S3: utilize a maximum magnetic charge density ρ=5.6 E8 (namely a magnetic charge density of the non-defective region) of the structure connection portion in the reconstructed magnetic charge distribution as a constant and substitute it into Eq. (1-4), and conduct inverse iteration to reconstruct defect depth of the defective region to obtain a 3D image of the defective region of the to-be-tested structure.

(12) Step S3 specifically includes:

(13) S31: according to a magnetic charge distribution theory, and on the premise of knowing the magnetic charge density of the non-defective region of the to-be-tested structure, the normal components of the MFLT signals at field point r.sub.j outside the plate, {B.sub.z(r.sub.j)} can be expressed as

(14) $\begin{array}{cc}{B}_z\left(r_j\right)=\frac{{\mu }_0\rho v}{4\pi }\underset{i=1}{\overset{n}{.Math.}}\frac{d_i+h}{{\left(\sqrt{{\left(x_i-x_j\right)}^2+{\left(y_i-y_j\right)}^2+{\left(d_i+h\right)}^2}\right)}^3}& \left(1\right)\end{array}$

(15) where μ.sub.0 is the air permeability, v is finite element volume of a source point, ρ is the magnetic charge density of the non-defective region of the to-be-tested structure, h is the lift-off for the magnetic sensor, which is equal to the distance in the z-axis direction between the field point r.sub.j and the top surface of the plate, x.sub.j, x.sub.i, y.sub.j, y.sub.i respectively are coordinates of the field point r.sub.j and the source point r.sub.i′, and d.sub.i is the associated depth of a source point r.sub.i′ on the defect surface, i is the number of source point, j is the number of field point, and n is the total number of source point;

(16) S32: all variables d.sub.i are collected as defect depth field {d.sub.i}, and h is used as an initial value of {d.sub.i},

(17) $\begin{array}{cc}{B}_z\left(r_j\right)=\frac{{\mu }_0\rho v}{4\pi }\underset{i=1}{\overset{n}{.Math.}}\frac{d_i}{{\left(\sqrt{{\left(x_i-x_j\right)}^2+{\left(y_i-y_j\right)}^2+d_i^2}\right)}^3}& \left(2\right)\end{array}$

(18) S33: according to step S32, an incidence relation of the MFLT signal {B.sub.z(r.sub.j)} and the defect depth field and a coefficient matrix k.sub.d(i,j) are constructed, that is,

(19) $\begin{array}{cc}\left\{B_z\left(j\right)\right\}=\left[k_d\left(i,j\right)\right]\left\{d_i\right\},\mathrm{and}& \left(3\right)\\ k_d\left(i,j\right)=\frac{\rho {\mu }_{0}v}{4\pi }\left(\frac{1}{{\left(\sqrt{{\left(x_i-x_j\right)}^2+{\left(y_i-y_j\right)}^2+d_i^2}\right)}^3}-\frac{3d_i^2}{{\left(\sqrt{{\left(x_i-x_j\right)}^2+{\left(y_i-y_j\right)}^2+d_i^2}\right)}^5}\right);& \left(4\right)\end{array}$

(20) S34: B.sub.z(r.sub.j)}, and [K.sub.d(i,j)] are abbreviated as {B}, {d} and [K.sub.d], respectively. According to the incidence relation of the MFLT signal {B} and the defect depth field {d} and the depth field coefficient matrix [k.sub.d], inverse iteration is conducted to obtain the defect depth {d} of the defective region, namely a 3D image of the defective region, as shown in FIG. 5.

(21) Preferably, as shown in FIG. 2, in step S34, an inverse iteration process of the defect depth field and the depth coefficient matrix is as follows:

(22) St0: set an initial value of {d}, end an iteration condition ε.sub.end, and calculate [k.sub.d] with given initial {d} according to Eq. (4);

(23) St1: according to the formula {B}=[K.sub.d]{d}, utilize the given {d} and [k.sub.d] to calculate corresponding {B}, and obtain a standard difference ε between the calculated MFLT signal value {B} and an intercepted MFLT signal value {B.sub.tar} over a defective area as shown in S3;

(24) St2: compare ε with ε.sub.end, if ε is greater than ε.sub.end, update {d} and [K.sub.d], repeat step St1 till ε is less than ε.sub.end, and end iteration to obtain the defect depth field {d}.

(25) It can be known by comparing FIG. 3 with FIG. 6, the defect depth field of the to-be-tested structure obtained by using the method of the embodiment of the present invention is consistent with the defect sizes of the to-be-tested structure, so the present invention improves the detection accuracy in comparison with the prior art.

(26) In conclusion, the method provided by the present invention does not necessarily need the saturated magnetization, directly measures the MFLT signals of the defective specimen, and conducts inverse iteration on the defect depth field according to the incidence relation of the MFLT signal and the defect depth field to finally obtain a 3D image of the defective region of the to-be-tested structure. Compared with the prior art, the present invention simplifies the MFLT devices and improves the imaging accuracy of the defects.

(27) Obviously, persons skilled in the art can make various modifications and variations to the present invention without departing from the spirit and scope of the present invention. The present invention is intended to cover these modifications and variations provided that they fall within the scope of protection defined by the following claims and their equivalent technologies.