Method for detecting internal defect depth of composite material based on frequency modulation type infrared thermal imaging

Abstract

The invention discloses a method for detecting the depth of an internal defect of a composite material based on frequency modulation type infrared thermal imaging. The method comprises the following steps: step 1, preparing a calibration test piece; step 2, obtaining a surface temperature response alternating-current component of the calibration test piece; 3, determining a change curve of the phase difference with respect to the frequency according to the thermal image sequence; 4, establishing a relation between the sensitive frequency and the defect depth; and step 5, evaluating the defect depth according to the sensitive frequency measurement value. According to the method, a material surface temperature image is obtained through an infrared thermal wave radar technology, and a temperature phase is extracted, so that the influence of non-uniform material surface heating and non-uniform emissivity on a test result is avoided, and a sensitive frequency is further extracted from the temperature phase to detect the defect depth. Compared with the traditional blind frequency, the sensitive frequency for detecting the defect depth is easier and more accurate to obtain, the detection range is larger than the delay time and the cross-correlation phase, and the method has the characteristics of high measurement precision and wide detection range.

Description

Technical Field

[0001] The present invention belongs to the technical field of nondestructive testing, and relates to a defect depth detection method based on frequency modulation infrared thermal imaging, and specifically to a method for evaluating the defect depth by extracting sensitive frequencies in a phase difference curve. Background Art

[0002] Infrared thermal wave radar (TWR) is a new type of infrared thermal imaging nondestructive testing technology. This technology heats the material under test by applying frequency modulated thermal excitation, and uses an infrared thermal imager to measure the surface temperature of the material to achieve nondestructive detection of internal defects in the material. Traditional infrared thermal imaging methods mainly include pulse method and phase lock method. The pulse method uses high-energy light pulses to excite the material under test and analyze the cooling curve to detect defects. The phase lock method uses periodic sinusoidal wave excitation and extracts the phase information of the thermal image for defect detection. TWR uses frequency modulation to generate a wide-band thermal excitation signal, which overcomes the limitations of traditional technologies in the distinguishability and detection range of defects at different depths.

[0003] At present, the depth of internal defects of materials is detected based on infrared thermal imaging technology. The conventional method is to extract characteristic parameters from the infrared thermal image sequence, establish the corresponding relationship between characteristic parameters and depth, and identify the defect depth by measuring characteristic parameters. At present, it is more common to extract characteristic time from the temperature change curve or blind frequency from the phase difference change curve as characteristic parameters. However, the surface temperature is easily affected by factors such as non-uniform heating of the material surface and uneven distribution of emissivity. On the other hand, extracting characteristic time from the temperature curve requires calculating derivatives, which leads to the greater influence of noise on the defect depth detection. The phase in the frequency domain has higher anti-interference ability than the temperature in the time domain. This is because the phase is less sensitive to non-uniform heating and uneven distribution of surface emissivity. However, under the influence of measurement noise, it is also difficult to extract blind frequency from the phase difference change curve, which ultimately leads to reduced detection accuracy of defect depth. For TWR, the defect depth can also be detected by cross-correlating the excitation signal and the response signal and extracting the delay time and cross-correlation phase. However, the correlation between the excitation and the response gradually weakens with the increase of the diffusion depth of the thermal wave, resulting in the delay time and cross-correlation phase can only detect the defect depth within a limited range, which has great limitations. Summary of the invention

[0004] In view of the above problems existing in the prior art, the present invention provides a method for detecting the depth of internal defects of composite materials based on frequency-modulated infrared thermal imaging. The method obtains the material surface temperature image through infrared thermal wave radar technology and extracts the temperature phase, thereby avoiding the influence of uneven heating and uneven emissivity of the material surface on the test results, and further extracts the sensitive frequency from the temperature phase to detect the defect depth. The sensitive frequency used by the present invention to detect the depth of defects is easier to obtain accurately than the traditional blind frequency, and the detection range is also larger than the delay time and the cross-correlation phase. It has the characteristics of high measurement accuracy and wide detection range, and is suitable for the detection of the depth of internal stratified defects in composite materials.

[0005] The objective of the present invention is achieved through the following technical solutions:

[0006] A method for detecting the depth of internal defects of composite materials based on frequency modulation infrared thermal imaging comprises the following steps:

[0007] Step 1: Prepare the calibration specimen;

[0008] Step 2: Obtain the AC component of the surface temperature response of the calibration specimen;

[0009] Step 3: Determine a curve of phase difference change with respect to frequency according to the thermal image sequence;

[0010] Step 4: Establish the relationship between sensitive frequency and defect depth;

[0011] Step 5: Estimate the defect depth based on the sensitive frequency measurements.

[0012] Compared with the prior art, the present invention has the following advantages:

[0013] 1. Compared with the commonly used characteristic time method, there is no need to derive the original temperature data, which reduces the impact of measurement noise on the detection results; by extracting phase information to identify the defect depth, it is not easily affected by non-uniform heating of the specimen surface and uneven emissivity distribution.

[0014] 2. Compared with the commonly used method of detecting defect depth based on infrared thermal wave radar technology, it has a larger detection range, and there is a linear relationship between the sensitive frequency and the defect depth, which is easy to calibrate; it takes into account the influence of defect size on the sensitive frequency, and corrects the detection error caused by the inconsistency between the preset defect size of the calibration specimen and the defect size of the specimen to be tested. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 This is a schematic diagram of the infrared thermal wave radar nondestructive testing experiment;

[0016] Figure 2 Schematic diagram of the experimental specimen (unit: mm);

[0017] Figure 3 It is the phase difference change curve before and after smoothing;

[0018] Figure 4 is the curve of phase difference changing with frequency;

[0019] Figure 5 Verify the feasibility of defect depth detection. DETAILED DESCRIPTION

[0020] The technical solution of the present invention is further described below in conjunction with the accompanying drawings, but is not limited thereto. Any modification or equivalent replacement of the technical solution of the present invention without departing from the spirit and scope of the technical solution of the present invention should be included in the protection scope of the present invention.

[0021] The present invention provides a method for detecting the depth of internal defects of composite materials based on frequency modulation infrared thermal imaging. The method measures the depth of internal defects of materials based on infrared thermal wave radar technology. The main feature of the method is that the phase information of the defect is obtained by measuring the phase change of the surface temperature of the specimen and using image processing technology to determine the sensitive frequency. There is a one-to-one correspondence between the sensitive frequency and the defect depth. The specific steps for detecting the defect depth of the target specimen are as follows:

[0022] Step 1: Prepare the calibration specimen.

[0023] Step 11: Use infrared thermal wave radar technology to pre-detect the test piece, identify the location and shape of the defect based on the collected thermal image, and use a general method to identify the defect size.

[0024] Step 12: The calibration specimen is the same as the test specimen and contains multiple flat-bottomed blind holes of different sizes and depths. The shape of the blind holes is similar to the defects of the test specimen (not required to be completely consistent), and sufficient distance is maintained between adjacent blind holes to ensure that they do not interfere with each other during the measurement process.

[0025] Step 13: Use appropriate black spray paint to evenly spray the surface of the calibration specimen.

[0026] Step 2: Obtain the AC component of the surface temperature response of the calibration specimen.

[0027] Step 21: Apply frequency modulated thermal excitation to the surface of the calibration specimen, as described in the following formula:

[0028] q(t)=A(1+sin(2πft)) (1)

[0029] In the formula, q is the heat flux value of the excitation, A is the heat flux amplitude, and f is the frequency, which satisfies the following form:

[0030]

[0031] Where f1 and f2 are the starting frequency and ending frequency respectively, T s Increasing the sampling frequency of the thermal image sequence (usually higher than 30 Hz) and extending the heating time (usually more than 10 seconds) can help improve detection accuracy.

[0032] Step 22: The measured temperature generally contains both DC and AC components, but the DC component will interfere with phase extraction. In order to remove the DC component, after the calibration specimen is stimulated as shown in equation (1), it is cooled to the initial temperature and then loaded with the following thermal excitation waveform with opposite phase:

[0033] q p (t) = A(1 + sin(2πft + π)) (3)

[0034] Step 23: If the temperatures at a point on the surface of the calibration specimen are T1 and T2 respectively under the thermal excitations shown in equations (1) and (3), then the AC component at the point is:

[0035] T ac =(T1-T2) / 2 (4)

[0036] When applying the thermal excitations shown in equations (1) and (3), it is necessary to ensure that the infrared thermal imager synchronously acquires thermal images to ensure phase consistency.

[0037] Step 3: Determine the change curve of the phase difference with respect to the frequency according to the thermal image sequence.

[0038] Step 31: Substitute the temperature AC component calculated in step 2 into equations (5) and (6) to obtain the specific frequency f t The in-phase correlation function and orthogonal correlation function are as follows:

[0039]

[0040]

[0041] Where N is the total number of thermal image frames, f t ∈[f1,f2], R(n) and Q(n) represent the in-phase reference signal and the orthogonal reference signal respectively, satisfying the following form:

[0042]

[0043]

[0044] In the formula, f samp is the sampling frequency.

[0045] Step 32: Calculate the phase of the AC component using the in-phase correlation function and the orthogonal correlation function:

[0046]

[0047] Step 33: Extract the temperature AC components of several pixels in the defect area and the reference area respectively, calculate the corresponding phase values ​​by equations (5)-(9), and calculate the phase difference:

[0048] ΔΦ=Φ s -Φ d (10)

[0049] Where ΔΦ is the phase difference, Φ d and Φ s are the phase values ​​calculated based on the pixels in the defect area and the reference area, respectively.

[0050] Step 34: Process the temperature data of each defect area on the surface of the calibration specimen according to steps 31 to 33 to obtain the phase difference ΔΦ corresponding to defects of different sizes and depths with respect to the frequency f t The change curve.

[0051] Step 35: The phase difference frequency variation curve has an oscillation phenomenon, so the Loess formula is used to smooth the curve. The Loess formula is a common non-parametric regression smoothing method. Its principle is to weight each data point in the curve and use nearby data for local fitting to obtain a smooth curve.

[0052] Step 4: Establish the relationship between sensitive frequency and defect depth.

[0053] Step 41: Define the frequency corresponding to the maximum phase difference ΔΦ as the sensitive frequency, denoted as f s , sensitive frequency f s The following relationship is satisfied with the defect depth:

[0054]

[0055] Where C1 and C2 are unknown coefficients, D is the defect size, and H is the defect depth (the distance between the defect plane and the measured surface of the specimen).

[0056] Step 42: Calculate the corresponding sensitive frequency according to the phase difference variation curve of each defect area on the surface of the calibration specimen relative to the reference area.

[0057] Step 43: According to the sensitive frequency f of the defect s , size D and depth H, draw f s and 1 / (D·H 2 ) and determine the coefficients C1 and C2.

[0058] Step 5: Estimate the defect depth based on the sensitive frequency measurements.

[0059] Step 51: applying the same thermal excitation as the calibration specimen to the test piece, extracting the temperature AC components of the defective area of ​​the test piece according to step 2, and obtaining the phase difference with respect to the frequency change curve according to the temperature AC components according to step 3, and determining the sensitive frequency of the defective area of ​​the test piece;

[0060] Step 52: Substitute the sensitive frequency and defect size into equation (11) to calculate the defect depth.

[0061] Example:

[0062] This embodiment uses infrared thermal wave radar technology to pre-test the test piece, such as Figure 1 As shown, the experimental device mainly includes an infrared thermal image acquisition system and a light excitation system. A signal generator or a data acquisition card is used to control the light excitation system and the infrared thermal image acquisition system so that the light excitation signal output and the thermal image acquisition are synchronized to ensure the phase consistency between the excitation and the response in multiple experiments. Before applying the method described in the present invention, the test piece should be pre-tested to determine the location, shape and size of the internal defects of the test piece. According to the size and shape of the defect to be tested, a calibration test piece is designed. Taking the internal defect depth detection of alumina ceramics as an example, the specimen size is 150mm×40mm×5mm, and a circular flat-bottomed blind hole is used to approximate the internal defects. There are 14 flat-bottomed blind holes in the specimen, of which the diameter of defects #1 to #7 is 10mm, the diameter of defects #8 to #14 is 8mm, and the defect depth is as follows Figure 2 As shown in Figure 1, 10 flat-bottomed blind holes were randomly selected to calibrate the relationship between the sensitive frequency and the defect depth.

[0063] The specimen was installed at a distance of 25 cm from the infrared thermal imager, and the test surface of the specimen was kept perpendicular to the direction of the infrared thermal imager lens, and the specimen was located at the center of the infrared thermal imager field of view. Laser excitation was applied to the specimen, the laser wavelength was 808 nm, the angle between the laser irradiation direction and the perpendicular direction of the specimen surface was about 15°, and the heating time was 100 s. First, the optical excitation system was controlled to generate a frequency modulation excitation signal as shown in formula (1), where the frequency modulation starting frequency f1 and the end frequency f2 were 0.001 Hz and 1 Hz respectively, and the amplitude A was 17.18 A (the laser energy was controlled by adjusting the current), and the infrared thermal image sequence of the specimen surface was recorded. After the specimen was cooled to room temperature, the anti-phase excitation signal as shown in formula (3) was applied again, and the infrared thermal image sequence of the specimen surface was recorded. The thermal image was first subjected to background subtraction, that is, the initial temperature was uniformly subtracted, and then the defect area and reference area in the thermal image were determined. The temperature data of the pixels in the defect area and the reference area were extracted respectively, and the temperature average was calculated. The thermal image sequences obtained under the two previous excitations are processed in the same way, and then the temperature AC component is calculated using formula (4).

[0064] After calculating the temperature AC components of each defect area (a total of 10) and the reference area, the phase values ​​of each AC component in the frequency range [0.001, 0.1] are calculated using equations (5)-(9). The phase difference between each defect area and the reference area is calculated according to equation (10), and the curve is smoothed using the Loess formula. The original phase difference curve and the smoothed phase difference curve are shown in Figure 1. Figure 3 shown.

[0065] like Figure 3 As shown, the phase difference curve after smoothing has a trend of increasing first and then decreasing, and there is an extreme point, and the frequency value corresponding to this point is the sensitive frequency. Figure 4 The phase difference curves corresponding to defects of different depths are given. Figure 4 It can be seen that the defect size and depth can change the sensitive frequency. For defects of the same size, the phase difference extreme value and the sensitive frequency gradually decrease with the increase of the defect depth. The sensitive frequencies f of defects #1, #4, #5, #6, #7, #8, #9, #12, #13 and #14 were extracted in the experiment. s , and plotted f according to the size and depth of the defect s About 1 / (D·H 2 ) is a linear fitting curve, such as Figure 5 As shown. The R2 value obtained by linear fitting is 0.96209, indicating that the fitting curve has a high degree of consistency with the actual data. According to the linear fitting curve, the unknown coefficients in formula (11) can be obtained.

[0066] In addition to the defects used for calibration, two defects (#2 and #10) were selected for actual measurement. Figure 5 The temperature AC component of each defect area is extracted respectively, and the phase difference curve with respect to frequency is obtained in turn according to the temperature AC component, and the sensitive frequency corresponding to each defect is determined according to the extreme point. Figure 5 The defect depth is calculated from the sensitive frequency and defect size. The actual defect depth measurement results are also plotted in Figure 5 The results show that the measured values ​​of the defect depths are very close to the true values. The true depths of defects #2 and #10 are 1.1 mm and 1.2 mm, respectively. The measured values ​​according to the method of the present invention are 1.142 mm and 1.23 mm, respectively. The relative errors are 3.8% and 2.5%, respectively, and the detection accuracy is high.

Claims

1. A method for detecting the depth of internal defects of composite materials based on frequency modulation infrared thermal imaging, characterized in that The method comprises the following steps: Step 1: Prepare the calibration specimen; Step 2: Obtain the AC component of the calibration specimen surface temperature response: Step 21: Apply frequency modulated thermal excitation to the surface of the calibration specimen: q(t)=A(1+sin(2πft)) (1) In the formula, q is the heat flux value of the excitation, A is the heat flux amplitude, and f is the frequency; Step 22: After the calibration specimen is stimulated as shown in formula (1), it is cooled to the initial temperature and then loaded with the following thermal stimulation waveform with opposite phase: q p (t)=A(1+sin(2πft+π)) (2) Step 23: If the temperatures at a point on the surface of the calibration specimen are T1 and T2 respectively under the thermal excitations shown in equations (1) and (2), then the AC component at the point is: T ac =(T1-T2) / 2 (3) Step 3: Determine the phase difference curve with respect to frequency based on the thermal image sequence: Step 31: Substitute the temperature AC component calculated in step 2 into equations (4) and (5) to obtain the specific frequency f t The in-phase correlation function and orthogonal correlation function are as follows: Where N is the total number of thermal image frames, f t ∈[f1,f2], R(n) and Q(n) represent the in-phase reference signal and the quadrature reference signal respectively; Step 32: Calculate the phase of the AC component using the in-phase correlation function and the orthogonal correlation function: Step 33: Extract the temperature AC components of several pixels in the defect area and the reference area respectively, calculate the corresponding phase values ​​according to steps 31 to 32, and calculate the phase difference: ΔΦ=Φ s -F d (7) Where ΔΦ is the phase difference, Φ d and Φ s are the phase values ​​calculated based on the pixels in the defect area and the reference area respectively; Step 34: Process the temperature data of each defect area on the surface of the calibration specimen according to steps 31 to 33 to obtain the phase difference ΔΦ corresponding to defects of different sizes and depths with respect to the frequency f t The change curve of Step 35: Use the Loess formula to smooth the curve of phase difference with respect to frequency; Step 4: Establish the relationship between sensitive frequency and defect depth: Step 41: Define the frequency corresponding to the maximum phase difference ΔΦ as the sensitive frequency, denoted as f s , sensitive frequency f s The following relationship is satisfied with the defect depth: In the formula, C1 and C2 are unknown coefficients, D is the defect size, and H is the defect depth; Step 42: Calculate the corresponding sensitive frequency according to the phase difference variation curve of each defect area on the surface of the calibration specimen relative to the reference area; Step 43: According to the sensitive frequency f of the defect s , size D and depth H, draw f s and 1 / (D·H 2 ) fitting curve, determination coefficients C1 and C2; Step 5: Evaluate defect depth based on sensitive frequency measurements: Step 51: applying the same thermal excitation as the calibration specimen to the test piece, extracting the temperature AC components of the defective area of ​​the test piece according to step 2, and obtaining the phase difference with respect to the frequency change curve according to the temperature AC components according to step 3, and determining the sensitive frequency of the defective area of ​​the test piece; Step 52: Substitute the sensitive frequency and defect size into equation (8) to calculate the defect depth.

2. The method for detecting the depth of internal defects of composite materials based on frequency modulation infrared thermal imaging according to claim 1 is characterized in that The specific steps of step 1 are as follows: Step 11: Use infrared thermal wave radar technology to pre-detect the test piece, identify the location and shape of the defect based on the collected thermal image, and use a general method to identify the defect size; Step 12: The calibration specimen is the same as the specimen to be tested and contains a plurality of flat-bottomed blind holes of different sizes and depths; Step 13: Use black spray paint to evenly spray the surface of the calibration specimen.

3. The method for detecting the depth of internal defects of composite materials based on frequency modulation infrared thermal imaging according to claim 2 is characterized in that The shape of the blind hole is similar to the defect of the tested piece, and a sufficient distance is maintained between adjacent blind holes to ensure that they do not interfere with each other during the measurement process.

4. The method for detecting the depth of internal defects of composite materials based on frequency modulation infrared thermal imaging according to claim 1, characterized in that The f satisfies the following form: Where f1 and f2 are the starting frequency and ending frequency respectively, T s The loading time.

5. The method for detecting the depth of internal defects of composite materials based on frequency modulation infrared thermal imaging according to claim 1, characterized in that The R(n) and Q(n) satisfy the following form: In the formula, f samp is the sampling frequency.

Images