Infrared thermal imaging nondestructive testing method for carbon fiber composite pultrusion plate

Abstract

The invention provides an infrared thermal imaging nondestructive testing method for a carbon fiber composite pultrusion plate, which comprises the following steps: step 1, feeding a carbon fiber tow at a yarn outlet device into a pultrusion process workbench through a roller type guider, infiltrating the carbon fiber tow with resin after the carbon fiber tow passes through a tensioner, and then curing the carbon fiber tow, cutting to obtain a carbon fiber composite pultrusion plate; 2, the central control processing system collects infrared images of the carbon fiber composite pultrusion plate in real time through an infrared thermal imager, and temperature distribution of the carbon fiber composite pultrusion plate in the heating process is recorded; step 3, the central control processing system processes the infrared image, identifies an abnormal area in temperature distribution, and extracts key features from the infrared image; 4, identifying product defects or abnormities, and generating a detailed detection report; and step 5, after the detection of the product is completed, waiting for the arrival of the next carbon fiber composite pultrusion plate, and repeating the operations from the step 2 to the step 3, so that the damage detection of the carbon fiber composite pultrusion plate is efficiently realized.

Description

Technical Field

[0001] The present invention relates to a method for non-destructive testing of pultruded plates of carbon fiber composite materials by infrared thermal imaging. Background Art

[0002] With the development of modern industry, with the rapid development of modern industry, pultrusion technology has been widely used in the fields of aerospace, automotive manufacturing, medical devices, etc. Pultrusion is a common composite material manufacturing process, and its core feature is to utilize the stretching characteristics of resin and fiber to closely combine the two, thereby producing products with high strength, high hardness, and excellent corrosion resistance. However, due to the complex structure and diverse materials of pultruded products, non-destructive testing of them remains a challenging task.

[0003] Currently, common non-destructive testing methods on the market include X-ray testing, ultrasonic testing, and CT scanning, etc. X-ray testing is suitable for detecting internal defects of metal materials, but its detection effect on composite materials is relatively limited. Ultrasonic testing can effectively detect internal defects of composite materials, but it requires special equipment, is complex to operate, and has a high cost. Although CT scanning can provide detailed internal structure information of composite material products, the equipment is expensive and the detection speed is relatively slow, making it difficult to meet the requirements of efficient detection.

[0004] In view of the limitations of existing detection technologies, infrared thermal imaging non-destructive testing has become an effective solution due to its simple operation and low cost. This method can not only achieve high-precision detection and identification of surface defects of pultruded products, but also has the advantages of automation and non-contact detection. The application of this technology has greatly overcome the deficiencies of existing detection methods and improved the detection efficiency and accuracy. Summary of the Invention

[0005] In view of this, the purpose of the present invention is to provide a method for non-destructive testing of pultruded plates of carbon fiber composite materials by infrared thermal imaging, which uses a gantry-type mobile vehicle and an infrared thermal imaging device, combined with advanced algorithms, to efficiently achieve damage detection of pultruded plates of carbon fiber composite materials.

[0006] To solve the above technical problems, the technical solution of the present invention is: A method for non-destructive testing of pultruded plates of carbon fiber composite materials by infrared thermal imaging, comprising the following steps:

[0007] Step 1, the carbon fiber tow at the yarn outlet device is fed into the pultrusion process workbench through a roller-type guide, and after passing through a tensioner, the carbon fiber tow is infiltrated with resin and then cured, and the pultruded plate of carbon fiber composite material is obtained by cutting;

[0008] Step 2: Gradually increase the temperature of the pultruded carbon fiber composite plate using a laser heating device. The central control processing system collects infrared images of the pultruded carbon fiber composite plate in real time through an infrared thermal imager, and records the temperature distribution of the pultruded carbon fiber composite plate during the heating process;

[0009] Step 3: The central control processing system processes the infrared images, identifies abnormal areas in the temperature distribution, and extracts key features from the infrared images;

[0010] Step 4: Use the extracted features for anomaly detection, compare the features in the normal state, identify product defects or anomalies, and generate a detailed inspection report;

[0011] Step 5: After completing the inspection of the product, wait for the arrival of the next pultruded carbon fiber composite plate, and repeat the operations in Step 2 to Step 3.

[0012] As a preferred embodiment of the present invention, in Step 2, the infrared thermal imaging image acquisition method is as follows:

[0013] Step 2-1: Turn on the infrared thermal imager;

[0014] Step 2-2: The infrared thermal imager starts recording thermal images at a certain fixed frame rate. After a delay period, the thermal images are saved.

[0015] As a preferred embodiment of the present invention, in Step 3, the abnormal area of the product is determined by feature extraction, and the method is as follows:

[0016] Step 3-1: Remove salt-and-pepper noise through median filtering and Gaussian filtering, smooth the image, and reduce the influence of noise; use histogram equalization to improve the contrast of the image;

[0017] Step 3-2: Set a threshold according to the gray value to divide the image into foreground and background;

[0018]

[0019] where g(x, y) is the binary image after segmentation, f(x, y) is the original image, and T is the threshold;

[0020] Step 3-3: Use an edge detection algorithm to find the edges in the image;

[0021] Step 3-4: Extract the shape features and texture features of the image

[0022]

[0023] where M pq is the (p, q) -th moment of the image.

[0024]

[0025] Among them, P(i, j) is the gray-level co-occurrence matrix.

[0026] Step 3-5: Classify and identify the extracted features.

[0027]

[0028] Among them, ω is the weight vector, is the mapping of the input feature vector, and b is the bias term.

[0029]

[0030] Among them, I is the input image and K is the convolution kernel.

[0031] Step 4: Use the extracted features for anomaly detection, compare with the features in the normal state, identify product defects or anomalies, and generate a detailed detection report;

[0032] Step 4-1: Extract the statistical features of the image and compare with the reference image

[0033]

[0034] If μ test and significantly deviate from μ baseline and then it is determined as abnormal;

[0035] Step 5: If there is no anomaly in the product, it is transported to the next process by the pulling device.

[0036] In summary, the present invention has the following beneficial effects:

[0037] 1. High-precision detection: The method and device can achieve high-precision detection and identification of surface defects of pultruded plates of carbon fiber composite materials, and the detection accuracy can meet the requirements of industrial production. Traditional detection methods often require manual visual inspection or the use of contact detection equipment, which are prone to problems such as missed detection and misdetection. However, the present invention adopts advanced image processing technology and deep learning algorithms, which can achieve high-precision detection and identification of defects, improving the reliability and accuracy of detection;

[0038] 2. Automatic detection: The method and device can achieve automatic detection of pultruded carbon fiber composite plates, improving the detection efficiency and reducing the detection cost. Traditional detection methods often require manual visual inspection or the use of contact detection equipment, with low detection efficiency and high cost. However, the present invention adopts automatic detection equipment and processes, enabling rapid and accurate detection of products, improving the detection efficiency and reducing the detection cost;

[0039] 3. Non-contact detection: The method and device adopt a non-contact detection method and will not damage the pultruded products. Traditional detection methods often require the use of contact detection equipment, which is likely to damage the products. Moreover, for some fragile and easily deformed products, the use of contact detection equipment may cause product damage. The present invention adopts a non-contact detection method, which will not damage the products and can better protect the quality and safety of the products. Description of the Drawings

[0040] Figure 1 is the overall effect diagram of the production line;

[0041] Figure 2 is the structural schematic diagram of the movable gantry;

[0042] Figure 3 is Figure 2 the effect diagram of the infrared thermal imager in

[0043] Reference numerals: 1. Yarn outlet device; 2. Forward roller type guide; 3. Reverse roller type guide; 4. Production workbench; 5. Resin bath area; 6. Heating device; 7. Cutting machine; 8. Pultruded carbon fiber composite plate; 9. Infrared detection device; 10. Pull-out device; 11. First laser heating device; 12. Central control processing system; 13. Infrared thermal imager; 14. Lens; 15. Second laser heating device; 16. Movable gantry. Detailed Embodiments

[0044] The following further details the specific embodiments of the present invention in conjunction with the drawings, so that the technical solutions of the present invention are easier to understand and master.

[0045] An infrared thermal imaging non-destructive detection method for pultruded carbon fiber composite plates is composed of a first laser heating device 11, a second laser heating device 15, a central control processing system 12, an infrared thermal imager 13, a lens 14, and a movable gantry 16.

[0046] Among them, the movable gantry 16 is a gantry of the prior art. The first laser heating device 11, the second laser heating device 15, the central control processing system 12, the infrared thermal imager 13 and the lens 14 are respectively installed on the movable gantry 16. Among them, the first laser heating device 11 and the second laser heating device 15 are respectively fixed at the corners at both ends of the movable gantry 16. The central control processing system 12, the infrared thermal imager 13 and the lens 14 are fixed in the middle section of the movable gantry 16.

[0047] The infrared thermal imager 13 is connected to the central control processing system 12 and is controlled by the central control processing system 12. That is, all the detection information obtained by the infrared thermal imager 13 is transmitted to the central control processing system 12, processed, and then transmitted by the central control processing system 12 to the computer.

[0048] Since infrared thermal imaging is very sensitive to changes in the surface temperature of an object, this high sensitivity enables the device to detect tiny temperature differences, thereby discovering problems or defects during product forming. No special preparation of the object to be measured is required, such as surface covering or marking, which reduces the complexity of operation, especially suitable for large or complex-structured devices; the infrared thermal image provided is an overall thermal image, not just local information, which helps to comprehensively understand the thermal performance of the object to be measured and is also conducive to predicting potential problems.

[0049] The lower end of the movable gantry 16 is rotatably connected with rollers, so that the movable gantry 16 can move in a plane, enabling real-time detection of products. While greatly improving the efficiency, it can also reduce the loss of its own device.

[0050] An infrared thermal imaging non-destructive testing method for a pultruded carbon fiber composite material plate includes the following steps:

[0051] Step 1, the carbon fiber tow at the yarn feeding device 1 first passes through the forward roller type guide 2, then through the reverse roller type guide 3, and then is fed into the pultrusion process workbench; the carbon fiber tow passes through the tensioner and then forms into a carbon fiber plate. Subsequently, the carbon fiber plate enters the resin bath area 5 filled with resin, and the carbon fiber plate is fully infiltrated in the resin; then it is cured by wire heating through the heating device 6, and finally cut by the cutting machine 7 to obtain the pultruded carbon fiber composite material plate 8. After that, the movable gantry 16 is started and moves to the cutting machine 7 of the pultrusion process workbench.

[0052] Step 2, the central control processing system 12 activates the infrared thermal imager 13, and the first laser heating device 11 and the second laser heating device 15 generate an appropriate heat source to gradually increase the temperature of the pultruded carbon fiber composite material plate, and the infrared image of the area to be measured is collected in real time, and the temperature distribution during the heating process is recorded.

[0053] Specifically, the method for collecting infrared thermal imaging images is as follows:

[0054] Step 2-1: Turn on the infrared thermal imager 13.

[0055] Step 2-2: The infrared thermal imager 13 starts to record thermal imaging at a certain fixed frame rate. After a delay, when the first laser heating device 11 and the second laser heating device 15 are excited, the formed thermal imaging is saved.

[0056] Step 2-3: Process the collected infrared thermal imaging.

[0057] Step 3: The central control processing system 12 processes the infrared image, identifies the abnormal area in the temperature distribution, and extracts key features from the infrared image.

[0058] The damage features in Step 3 adopt the following method:

[0059] Step 3-1: Process the image through a denoising algorithm to remove salt-and-pepper noise, achieve image smoothing, and reduce noise interference; subsequently, through an enhancement method based on pixel distribution, enhance the image contrast to make the details clearer.

[0060] Specifically, for the temperature data in the detection of pultruded carbon fiber composite plates, these data can be regarded as a two-dimensional data matrix. In this matrix, a weight distribution method can be used to perform weighted averaging on each data point to weaken the interference. By adjusting the parameters of the weighted distribution, the noise influence in the data will be suppressed, and the processed data matrix is obtained. Its process can be described in the following way:

[0061]

[0062] Among them, T(x, y) represents the temperature matrix after weighted smoothing processing, T(u, v) represents the element located in the i-th row and j-th column in the original temperature matrix, and σ represents the weight distribution parameter in the smoothing operation.

[0063] Step 3-2: Set a threshold according to the gray value to divide the image into foreground and background.

[0064]

[0065] Among them, g(x, y) is the binary image after segmentation, f(x, y) is the original image, and T is the threshold.

[0066] Step 3-3: Use an edge detection algorithm to find the edges in the image.

[0067] Step 3-4: Extract the shape features (shape matrix) and texture features (gray-level co-occurrence matrix) of the image

[0068]

[0069] where M pq is the (p,q)-order moment of the image.

[0070]

[0071] where P(i,j) is the gray-level co-occurrence matrix.

[0072] Step 3-5: Classify and identify the extracted features.

[0073]

[0074] where ω is the weight vector, is the mapping of the input feature vector, and b is the bias term.

[0075]

[0076] where I is the input image and K is the convolution kernel.

[0077] Step 4: Use the extracted features for anomaly detection, compare with the features in the normal state, identify product defects or anomalies, and generate a detailed detection report.

[0078] Step 4-1: Extract the statistical features of the image and compare with the reference image

[0079]

[0080] If μ test and significantly deviate from μ baseline and then it is determined as an anomaly.

[0081] Specifically, the spectrogram can be divided into several sub-regions, and the eigenvalue of each sub-region, such as the mean value and standard deviation, can be calculated respectively for comparison. If the eigenvalue of a certain sub-region is significantly higher than that of other sub-regions, it indicates that the temperature fluctuation in this region is relatively significant and there may be an anomaly. The division process of the spectrogram can be expressed as:

[0082]

[0083] where f k represents the lower frequency limit of the k-th frequency band, f max represents the highest frequency point in the whole spectrogram, K represents how many frequency bands the spectrogram is divided into, and k represents the number of the frequency band.

[0084] Within each frequency band, the mean amplitude and standard deviation of all frequency points in that band can be calculated, and then the amplitude differences between different frequency bands can be compared. If the amplitude of a certain frequency band is significantly higher than that of other frequency bands, it indicates that the temperature fluctuation in that band is relatively significant and there may be temperature anomalies.

[0085] The non-destructive infrared thermal imaging detection method for the pultruded plate of carbon fiber composite material of the present invention can also be: the turning on of the laser lamp, the movement of the gantry, and the turning on of the infrared camera are all commanded by the central control processing system.

[0086] Step 5, if there are no any defect problems in the detected product, it is transported to the next step by the transmission roller in the pulling-out device 10.

[0087] Of course, the above are only typical examples of the present invention. In addition, the present invention can also have many other specific implementation manners. Any technical solutions formed by equivalent replacement or equivalent transformation fall within the scope of protection required by the present invention.

Claims

1. A nondestructive testing method for infrared thermal imaging of carbon fiber composite pultruded plates, characterized in that: The steps include: Step 1, the carbon fiber tow at the yarn outlet device is sent to the pultrusion process workbench through a roller guide, the carbon fiber tow is impregnated with resin and then solidified after passing through a tensioner, and a carbon fiber composite material pultrusion plate is obtained by cutting; Step 2, gradually increasing the temperature of the carbon fiber composite material pultruded plate by a laser heating device, and the central control processing system collects infrared images of the carbon fiber composite material pultruded plate in real time by an infrared thermal imager, and records the temperature distribution of the carbon fiber composite material pultruded plate during the heating process; Step 3: The central control processing system processes the infrared image, identifies abnormal areas in the temperature distribution, and extracts key features from the infrared image; Step 4: Use the extracted features to perform anomaly detection, compare the features under normal conditions, identify product defects or anomalies, and generate a detailed test report; Step 5, after completing the inspection of the product, wait for the arrival of the next carbon fiber composite pultrusion plate and repeat the operations from step 2 to step 3.

2. The infrared thermal imaging nondestructive testing method for carbon fiber composite pultruded plates according to claim 1, characterized in that: In step 2, the infrared thermal imaging image acquisition method is as follows: Step 2-1, turn on the infrared thermal imager; Step 2-2, the infrared thermal imager starts to record thermal images at a fixed frame rate, and after a delay of a period of time, the thermal images are saved.

3. The infrared thermal imaging nondestructive testing method for carbon fiber composite pultruded plates according to claim 1, characterized in that: In step 3, the product abnormal area is determined by feature extraction, and the method is as follows: Step 3-1, remove salt and pepper noise by median filtering and Gaussian filtering, and use histogram equalization to improve the contrast of the image; Step 3-2, set a threshold according to the gray value to divide the image into foreground and background; Among them, g(x,y) is the binary image after segmentation, f(x,y) is the original image, and T is the threshold; Step 3-3, using an edge detection algorithm to find the edge in the image; Step 3-4: Extract shape and texture features of the image Among them, M pq is the (p,q)-order moment of the image. Among them, P(i,j) is the gray-level co-occurrence matrix. Step 3-5: classify and identify the extracted features. Where ω is the weight vector, is the mapping of the input feature vector, and b is the bias term. Among them, I is the input image and K is the convolution kernel. Step 4: Use the extracted features to perform anomaly detection, compare the features under normal conditions, identify product defects or anomalies, and generate a detailed test report; Step 4-1 Extract statistical features of the image and compare it with the reference image If μ test and Significant deviation from μ baseline and It is judged as abnormal; Step 5: If the product is normal, it will be transported to the next stage through the pulling device.

Images