Method for identifying acoustic emission damage mode of composite material under thermal shock

Abstract

The invention discloses an acoustic emission damage mode recognition method of a composite material under thermal shock. The method comprises the following steps: 1, connecting a composite material sample with an acoustic emission sensor; 2, connecting the acoustic emission sensor with an acoustic emission instrument, and starting a signal acquisition display of the acoustic emission instrument; 3, a lead breaking experiment is conducted on the surface of the composite material sample, and stable collection of acoustic emission signals in the fuel gas thermal shock service environment is achieved; 4, setting experimental parameters of fuel gas thermal shock on the simulation device; 5, carrying out a fuel gas thermal shock experiment, and displaying the parameters of the acoustic emission signal on a signal acquisition display; step 6, carrying out damage mode identification on the collected acoustic emission signals; 7, k-means clustering is conducted on the collected acoustic emission signals, wavelet packet analysis is conducted on the clustered acoustic emission signals, and the damage mode of the composite material sample under fuel gas thermal shock is recognized. According to the invention, the damage mode identification of the composite material in the gas thermal shock environment is realized.

Description

Technical Field

[0001] The invention belongs to the technical field of nondestructive testing of composite materials, and in particular relates to an acoustic emission damage pattern recognition method for composite materials under thermal shock. Background Art

[0002] At present, most studies still analyze the failure mechanism of composite materials through the evolution of their microstructure. There is a lack of real-time detection methods to study the key parameters in the failure process of gas thermal shock, and then derive the failure mechanism of composite materials. General non-destructive testing methods are difficult to study the damage evolution and failure mode of composite materials, and are not sensitive to defect detection. In addition, some studies focus on studying the damage mechanism of composite materials at room temperature or under static loads through acoustic emission detection technology, and find that there are damage modes. However, there is still a huge gap between this and the actual service environment of composite materials. In the gas thermal shock environment, the failure behavior of composite materials is relatively unclear.

[0003] Ultrasonic testing, X-ray testing, infrared testing and acoustic emission testing can all achieve non-destructive testing of composite materials. Ultrasonic testing can detect defects inside materials, but ultrasonic signals are severely attenuated in composite materials, and the sensitivity to small-sized defects is not high; XCT technology is used to detect changes in the porosity inside the material, but it is insensitive to delamination defects in composite materials and has high costs; infrared detection technology is mainly used to measure surface and internal defects, but its ability to detect defects in composite materials above 2mm is limited, and it is easily affected by the surface state of the material, interfering with the identification of defects. Acoustic emission detection technology can collect the generation of crack signals in the process of material failure in real time. It has the advantages of online real-time dynamic detection, high sensitivity to defects, no need for external excitation, and a wide range of applications. The source of the acoustic emission signal is the material itself. Acoustic emission signals are only generated when the material undergoes structural or state changes.

[0004] Patent CN116990119A discloses a composite material damage pattern recognition method based on acoustic emission, which analyzes the damage mechanism of composite materials by monitoring and analyzing the acoustic emission signals during the tensile process of composite materials. However, there is still a huge gap between this and the actual service environment of composite materials. In the gas thermal shock environment, the failure form of composite materials is relatively unclear. Summary of the invention

[0005] In order to overcome the shortcomings of the above-mentioned prior art, the purpose of the present invention is to provide an acoustic emission damage pattern recognition method for composite materials under thermal shock, that is, during the failure process of the composite material in a gas thermal shock environment, an acoustic emission detection method is used to set a threshold value, sampling frequency, preamplifier and other means for the system to collect damage signals, extract specified acoustic emission signal characteristic parameters, perform cluster analysis and wavelet packet decomposition on the acoustic signal, and realize damage pattern recognition of the composite material in a gas thermal shock environment.

[0006] In order to achieve the above object, the technical solution adopted by the present invention is:

[0007] A method for identifying acoustic emission damage patterns of composite materials under thermal shock comprises the following steps;

[0008] Step 1: Connect the composite material sample to the acoustic emission sensor;

[0009] Step 2: connecting the acoustic emission sensor to the acoustic emission instrument through a signal transmission line, turning on the signal acquisition display of the acoustic emission instrument, and setting the acoustic emission acquisition parameters of the composite material sample under the gas thermal shock service environment on the display;

[0010] Step 3: After setting the acoustic emission acquisition parameters, conduct a lead breaking experiment on the surface of the composite material sample, check the attenuation of the transmission signal on the signal acquisition display, and achieve stable acquisition of the acoustic emission signal under the gas thermal shock service environment;

[0011] Step 4: After completing the acoustic emission signal collection preparation work of steps 1 to 3, set the experimental parameters of the gas thermal shock on the gas thermal shock service environment simulation device;

[0012] Step 5: Turn on the ignition button on the simulation device to conduct a gas thermal shock experiment. When the gas impacts the surface of the composite material sample, damage will occur inside the composite material sample. Then, the acoustic emission signal is collected through steps 1 to 3, and the amplitude, peak frequency, duration, rise time, count and energy parameters of the acoustic emission signal are displayed on the signal collection display;

[0013] Step 6:

[0014] The acoustic emission signals collected in step five are subjected to damage pattern recognition, and tensile tests are carried out on the substrate material, fiber bundle and composite material samples at room temperature on a high-temperature universal mechanical testing machine, and acoustic emission detection is carried out using an acoustic emission instrument. The specific process repeats steps two to three to obtain the amplitude, peak frequency, duration, rise time, count and energy parameters of the three samples of the substrate material, fiber bundle and composite material, and then these parameters are subjected to k-means cluster analysis and wavelet packet analysis to obtain the frequency characteristic ranges of the four damage modes of matrix cracking, interface slip, interface debonding and fiber breakage;

[0015] Step 7: Perform k-means clustering on the acoustic emission signals collected in step 5, perform wavelet packet analysis on the clustered acoustic emission signals, and compare them with the results of step 6 to identify the damage mode of the composite material sample under gas thermal shock.

[0016] The step 1 is specifically as follows:

[0017] The round composite material sample is fixed on the thermal shock fixture, the top surface on the side facing the spray gun is used for the gas thermal shock test, and the bottom surface on the other side facing away from the spray gun is mechanically or welded to one end of the high-temperature waveguide wire of the nickel-chromium alloy material, and the other end of the high-temperature waveguide wire is connected to the acoustic emission sensor by means of a rubber plug fixed with butter coupling;

[0018] The step 2 is specifically as follows:

[0019] The parameters set are specifically preamplifier 20 / 40 / 60dB, sampling threshold 0-100dB, sampling frequency 2MHz, frequency range 0.1-1MHz and impact length 2k;

[0020] When a composite material sample is damaged, a stress wave, namely an acoustic emission signal, is generated, which is then transmitted to the sensor and finally collected on the signal display.

[0021] The step three is specifically as follows:

[0022] Lead is used to break the composite material sample, and the acoustic emission signal emitted is collected to observe the amplitude attenuation. The higher the amplitude, the better the coupling with the sensor, and finally the stable collection of acoustic emission signals in the gas thermal shock service environment is achieved;

[0023] The lead breaking experiment is to slowly press a pencil at an angle of 30 degrees on the surface of the composite material sample 30 mm away from the contact point of the high-temperature waveguide wire to break the lead core, collect the emitted acoustic emission signal, and observe the amplitude attenuation. The higher the amplitude, the better the coupling with the sensor.

[0024] The step 4 is specifically as follows:

[0025] The parameters are specifically gas shock temperature, oxygen outlet pressure, oxygen flow rate, kerosene outlet pressure, kerosene flow rate, cooling gas outlet pressure and cooling gas flow rate;

[0026] Among them, the gas impact temperature is 0-1300℃, the oxygen outlet pressure is above 1.3MPa, the oxygen flow rate is 230-500L / min, the kerosene outlet pressure is above 0.5MPa, the kerosene flow rate is 5-15L / h, the cooling gas outlet pressure is about 0.7MPa, and the cooling gas flow rate is about 60L / min.

[0027] The step six specifically comprises: performing a substrate tensile test at room temperature based on an acoustic emission device to obtain a characteristic range of acoustic emission frequency for substrate cracking;

[0028] Based on the acoustic emission device, the fiber bundle tensile test was carried out at room temperature to obtain the characteristic range of the acoustic emission frequency of fiber breakage;

[0029] Based on the acoustic emission device, tensile tests of composite materials were carried out at room temperature, and failure modes mainly including matrix cracking, interface slip, interface debonding, and fiber fracture damage were obtained. K-means clustering analysis was performed on the obtained acoustic emission characteristic parameters. Based on the acoustic emission frequency characteristic ranges of matrix cracking and fiber fracture obtained above, the frequency characteristic ranges of interface slip and interface debonding were obtained.

[0030] The step seven is specifically as follows:

[0031] Amplitude, peak frequency, duration, rise time, count and energy parameters are selected as characteristic parameters for cluster analysis, and the mean variance normalization method is used to make the selected characteristic parameters dimensionless, so that the measured data are defined between [0,1], and the expression is as shown in formula (1);

[0032]

[0033] In the formula, A i is the original signal parameter, A μ is the average value of the original signal parameter, is the signal parameter after standardization, A σ is the standard deviation of the original signal parameter;

[0034] Secondly, the contour coefficient value is calculated for the dimensionless characteristic parameters, as shown in formula (2), to determine the optimal clustering number k;

[0035]

[0036] Where a(i) represents the average distance between the i-th vector and other vectors in the same category, and b(i,k) represents the average distance between the i-th vector and other vectors in different categories.

[0037] Finally, perform k-means clustering analysis on the processed feature parameters, select k cluster centers, calculate the distance from the initial cluster centroid to the input vector, and then classify the input vector into the category of the nearest cluster centroid; calculate the position of the cluster center according to the shortest distance principle; repeat the above steps until the position of the initial cluster centroid no longer changes.

[0038] Perform j-layer wavelet packet transform on an acoustic emission signal f(t) to obtain the sum of a series of decomposed parts, as follows:

[0039]

[0040] f(t) refers to an acoustic emission signal, is the decomposed signal at frequency band i and scale j. The energy of each part after decomposition is Defined as:

[0041]

[0042] R i The wavelet packet energy spectrum coefficient is the ratio of the energy of each node to the total energy of all nodes after decomposition;

[0043]

[0044] Beneficial effects of the present invention:

[0045] Based on the damage mechanism of composite materials, most of the research is carried out under normal temperature environment. Using the technology in the present invention, a damage pattern recognition method of composite materials under gas thermal shock environment based on acoustic emission technology is established. The present invention not only realizes the collection of composite material damage acoustic emission signals under gas thermal shock environment, but also provides a new solution for the identification of failure modes of composite materials under extreme service conditions. It has broad and good development prospects and high practical value.

[0046] 1. The present invention realizes real-time detection of composite materials under high temperature conditions by introducing a high-temperature waveguide rod and connecting the waveguide rod, sensor and sample through a mechanical device, and sets the acquisition parameters of acoustic emission, and finally realizes the stable acquisition of acoustic emission signals of composite materials under a gas thermal shock environment, thus solving the technical problem that composite materials cannot perform acoustic emission signal detection under a high temperature environment.

[0047] 2. The present invention identifies the damage mode of composite materials based on a single load. By performing tensile tests on the matrix, fiber, and composite materials respectively, the acoustic emission characteristic parameter ranges of matrix cracking, fiber breakage, interface debonding, and interface slip are obtained, solving the difficult problem of identifying the damage mode of composite materials.

[0048] 3. The present invention realizes damage pattern recognition of composite materials in a gas thermal shock environment by performing k-means clustering and wavelet packet analysis on the acoustic emission signals collected in a gas thermal shock environment, realizes qualitative and quantitative evaluation of composite material damage, provides rich data for the diagnosis of composite material failure mechanism and failure process, and also provides a direct basis for the life prediction of composite materials. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 It is a schematic diagram of the process of the present invention.

[0050] Figure 2 Schematic diagram of gas thermal shock experiment based on acoustic emission test equipment.

[0051] Figure 3 Schematic diagram of the Sk curve of the AE signal of the composite material under gas thermal shock.

[0052] Figure 4 Schematic diagram of the clustering results of AE signals of composite materials under gas thermal shock.

[0053] Figure 5 These are three typical wavelet packet energy spectra of composite materials during gas thermal shock; among them, (a) substrate cracking; (b) interface slip; (c) interface debonding.

[0055] 1. Acoustic emission sensor; 2. Waveguide rod (nickel-chromium alloy wire); 3. Clamping fixture; 4. Composite material sample. DETAILED DESCRIPTION

[0056] The present invention will be further described in detail below in conjunction with the accompanying drawings.

[0057] like Figure 1 , Figure 2 As shown, a method for identifying acoustic emission damage patterns of composite materials under thermal shock comprises the following steps;

[0058] Step 1, connect the composite material sample 4 and the acoustic emission sensor 1. Since the composite material is in service under a gas thermal shock environment, and the operating temperature of the acoustic emission sensor is generally room temperature, in order to realize the application of acoustic emission in the composite material service environment simulation experiment, a high-temperature waveguide rod 2 is introduced to realize the real-time detection of the composite material sample 4 under high temperature conditions. Nickel-chromium alloy wire is selected as the high-temperature waveguide rod 2 through the lead breaking experiment. One end of the waveguide rod 2 is mechanically connected to the back of the composite material sample 4 and fixed by a clamping fixture 3, and the other end adopts the holding device of the acoustic emission probe, which is tightly connected to the sensor under the action of the butter coupling agent, ensuring the tight connection between the acoustic emission sensor 1 and the waveguide rod 2, i.e., the nickel-chromium wire.

[0059] Step 2: Connect the acoustic emission system and set the acoustic emission acquisition parameters suitable for composite materials serving in a gas thermal shock environment, which are preamplifier, sampling threshold, sampling frequency, frequency range and impact length.

[0060] Step 3: Check whether the sensor of the acoustic emission system is well coupled with the composite material, and check the attenuation of the signal by performing a lead breaking experiment on the surface of the composite material sample. Finally, the stable acquisition of the acoustic emission signal of the composite material sample under the gas thermal shock environment is achieved.

[0061] Step 4: setting the main parameters of the gas thermal shock experiment on the gas thermal shock service environment simulation device, the parameters are specifically the gas shock temperature, oxygen outlet pressure, oxygen flow rate, kerosene outlet pressure, kerosene flow rate, cooling gas outlet pressure and cooling gas flow rate,

[0062] Step 5: After setting the gas thermal shock test parameters, conduct a gas thermal shock test on the composite material sample, and use an acoustic emission instrument to perform acoustic emission detection, and collect parameters such as amplitude, peak frequency, duration, rise time, count and energy of the acoustic emission signal on the signal acquisition display.

[0063] Step 6. The frequency characteristics of the acoustic emission signals of different damage modes and the energy distribution of the signals in each frequency band are different, and the frequency characteristics and energy distribution of the signals are almost independent of the size of the damage and the size of the load. Therefore, acoustic emission detection of the tensile failure process of the base material, fiber bundle and composite material was carried out respectively, and a large number of known acoustic emission signals of damage modes (matrix cracking, interface slip, interface debonding, fiber breakage) were obtained. The measured acoustic emission signals were subjected to k-means clustering analysis and wavelet packet analysis to establish the acoustic emission signal characteristics of the four damage modes. Finally, the frequency range of matrix cracking damage, fiber breakage damage, interface slip damage and interface debonding damage was determined.

[0064] Step 7, based on the composite material damage pattern recognition method of acoustic emission, the received acoustic emission signal is subjected to k-means clustering analysis and wavelet packet analysis. Before performing the k-means clustering analysis, the contour coefficient value is first used to determine the number of classifications k. The characteristic parameter characteristics of each type of acoustic emission signal after clustering are used to correspond the obtained clustering analysis results to the damage pattern. The "db5" wavelet basis is selected to perform three-layer wavelet packet decomposition on the acoustic emission signal to obtain the frequency band of each damage pattern, and then obtain the wavelet energy spectrum coefficient distribution diagram. Compare it with the characteristic energy spectrum coefficient distribution diagram of the damage pattern that has been determined in step 5 to determine the damage pattern of the composite material sample damage signal.

[0065] Implementation Cases:

[0066] Prepare a C / SiC composite material sample with a diameter of 30 mm and a thickness of 4 mm. The back of the sample is mechanically connected with a high-temperature waveguide wire made of a nickel-chromium alloy with a diameter of 1 mm and a length of 60 cm. The other end of the high-temperature waveguide wire is connected to the acoustic emission sensor by a rubber plug fixed with butter coupling;

[0067] The 16-channel acquisition system of the PCI-2 acoustic emission instrument produced by PAC Corporation of the United States is used to collect and process the acoustic emission signals. The parameters of the acoustic emission detection system are set as follows: preamplifier 40dB, sampling threshold 40dB, sampling frequency 2MHz, frequency range 0.1-1MHz, and impact length 2k.

[0068] A lead breaking experiment was carried out on the surface of the composite material sample, and the attenuation of the transmission signal was checked on the signal acquisition display. The amplitude of the measured acoustic emission signal was 95, indicating that stable acquisition of the acoustic emission signal was achieved under the gas thermal shock service environment.

[0069] A service simulation device was used to simulate the gas thermal shock environment, and the equipment parameters were set as follows: gas shock temperature of 1200°C, oxygen outlet pressure of 1.3 MPa, oxygen flow rate of 260 L / min, kerosene outlet pressure of 0.6 MPa, kerosene flow rate of 5 L / h, cooling gas outlet pressure of 0.7 MPa, and cooling gas flow rate of 60 L / min.

[0070] While the sample is being impacted by gas, click the acquisition button on the signal acquisition display to collect the acoustic emission signal.

[0071] The experiment conducted acoustic emission detection of the tensile failure process on the base material, fiber bundle and composite material. First, the acoustic emission signal of the base material during tension was decomposed by wavelet packets, and it was determined that the frequency range of matrix cracking damage was 0-150kHz. Secondly, the acoustic emission signal of the fiber bundle material during tension was decomposed by wavelet packets, and it was determined that the frequency range of fiber breakage damage was above 350kHz. The acoustic emission signal of the composite material during tension was subjected to k-means clustering analysis to find out the acoustic emission signals that were different from the previous two damage modes and perform wavelet packet decomposition to determine that the frequency ranges of interface slip and debonding damage correspond to 150-250kHz and 250-350kHz, respectively.

[0072] The acoustic emission signals collected in the gas thermal shock experiment were subjected to k-means cluster analysis and wavelet packet analysis. Before applying the kmeans cluster analysis, the k value should be determined.

[0073] Firstly, amplitude, peak frequency, duration, rise time, count and energy parameters were selected as characteristic parameters for cluster analysis.

[0074] Secondly, the mean variance normalization method is adopted to make the selected characteristic parameters dimensionless, so that the measured data are defined between [0, 1], and its expression is as shown in formula (1).

[0075] Finally, k-means clustering analysis was performed on the processed feature parameters. In order to determine the optimal clustering number k, the silhouette coefficient values ​​under different clustering numbers k were calculated, as shown in Figure 3 As shown in Figure 2. It is known that the larger the silhouette coefficient value, the better the classification effect. Generally speaking, when the silhouette coefficient value is greater than 0.6, the classification effect is already very good. Figure 3As shown in the figure, the contour coefficient value is the largest when the AE signal of C / SiC composite material during the 1200℃ gas impact process is divided into three categories, that is, when k=3, S is 0.64, which means that when performing k-means clustering analysis, the clustering effect is best when three categories are selected.

[0076] Perform k-means cluster analysis on the acoustic emission signals after the above processing, such as Figure 4 shown. Figure 4 It shows the clustering results of the amplitude and peak frequency of the AE signal. From the figure, it can be found that the three types of signals, matrix cracking, interface slip and interface debonding, rarely overlap in peak frequency. The amplitude distribution is between 40-80dB, and the peak frequency bands are: 0-150kHz, 150-250kHz, 250-350kHz. It proves that the peak frequency of the acoustic emission signal can classify the damage signal of the C / SiC composite material. It can be clearly seen that there are obvious differences between the clusters, and there is almost no overlap between the samples.

[0077] The above methods were all run using MATLAB software.

[0078]

[0079] In the formula, A i is the original signal parameter, A μ is the average value of the original signal parameter, is the signal parameter after standardization, A σ is the standard deviation of the original signal parameter.

[0080]

[0081] Where a(i) represents the average distance between the i-th vector and other vectors in the same category, and b(i,k) represents the average distance between the i-th vector and other vectors in different categories.

[0082] The collected acoustic emission signals are transformed by wavelet packet, and the "db5" wavelet basis is selected to perform three-layer wavelet packet decomposition on the acoustic emission signals, and the wavelet energy spectrum coefficients are extracted as characteristic parameters for identifying the acoustic emission signals of composite material damage modes, such as Figure 5 shown.

[0083] The wavelet packet energy spectrum coefficient can intuitively, simply and effectively reflect the situation of various types of signals in each frequency band, and then identify the failure mode of AE signals. The maximum values ​​of the wavelet packet energy spectrum coefficients of the three typical failure modes are concentrated in Node (0-0.125MHz), Node (0.125-0.25MHz), Nodes (0.25-0.375MHz). Obviously, they represent substrate cracking, interface slip, and interface debonding, respectively.

Claims

1. A method for identifying acoustic emission damage patterns of composite materials under thermal shock, characterized in that: The steps include: Step 1: Connect the composite material sample to the acoustic emission sensor; Step 2: connecting the acoustic emission sensor to the acoustic emission instrument through a signal transmission line, turning on the signal acquisition display of the acoustic emission instrument, and setting the acoustic emission acquisition parameters of the composite material sample under the gas thermal shock service environment on the display; Step 3: After setting the acoustic emission acquisition parameters, conduct a lead breaking experiment on the surface of the composite material sample, check the attenuation of the transmission signal on the signal acquisition display, and achieve stable acquisition of the acoustic emission signal under the gas thermal shock service environment; Step 4: After completing the acoustic emission signal collection preparation work of steps 1 to 3, set the experimental parameters of the gas thermal shock on the gas thermal shock service environment simulation device; Step 5: Turn on the ignition button on the simulation device to conduct a gas thermal shock experiment. When the gas impacts the surface of the composite material sample, damage will occur inside the composite material sample. Then, the acoustic emission signal is collected through steps 1 to 3, and the amplitude, peak frequency, duration, rise time, count and energy parameters of the acoustic emission signal are displayed on the signal collection display; Step 6: The acoustic emission signals collected in step five are subjected to damage pattern recognition, and tensile tests are carried out on the substrate material, fiber bundle and composite material samples at room temperature on a high-temperature universal mechanical testing machine, and acoustic emission detection is carried out using an acoustic emission instrument. The specific process repeats steps two to three to obtain the amplitude, peak frequency, duration, rise time, count and energy parameters of the three samples of the substrate material, fiber bundle and composite material, and then these parameters are subjected to k-means cluster analysis and wavelet packet analysis to obtain the frequency characteristic ranges of the four damage modes of matrix cracking, interface slip, interface debonding and fiber breakage; Step 7: Perform k-means clustering on the acoustic emission signals collected in step 5, perform wavelet packet analysis on the clustered acoustic emission signals, and compare them with the results of step 6 to identify the damage mode of the composite material sample under gas thermal shock.

2. The method for identifying acoustic emission damage patterns of composite materials under thermal shock according to claim 1, characterized in that: The step 1 is specifically as follows: The round composite material sample is fixed on the thermal shock fixture. The top surface on the side facing the spray gun is used for the gas thermal shock test. The bottom surface on the other side facing away from the spray gun is mechanically or welded to one end of a high-temperature waveguide wire made of a nickel-chromium alloy. The other end of the high-temperature waveguide wire is connected to the acoustic emission sensor by means of a rubber plug fixed with butter coupling.

3. The method for identifying acoustic emission damage patterns of composite materials under thermal shock according to claim 1, characterized in that: The step 2 is specifically as follows: The parameters set are specifically preamplifier 20 / 40 / 60dB, sampling threshold 0-100dB, sampling frequency 2MHz, frequency range 0.1-1MHz and impact length 2k; When a composite material sample is damaged, a stress wave, namely an acoustic emission signal, is generated, which is then transmitted to the sensor and finally collected on the signal display.

4. The method for identifying acoustic emission damage patterns of composite materials under thermal shock according to claim 1, characterized in that: The step three is specifically as follows: Lead is used to break the composite material sample, and the acoustic emission signal emitted is collected to observe the amplitude attenuation. The higher the amplitude, the better the coupling with the sensor, and finally the stable collection of acoustic emission signals in the gas thermal shock service environment is achieved; The lead breaking experiment is to slowly press a pencil at an angle of 30 degrees on the surface of the composite material sample 30 mm away from the contact point of the high-temperature waveguide wire to break the lead core, collect the emitted acoustic emission signal, and observe the amplitude attenuation. The higher the amplitude, the better the coupling with the sensor.

5. The method for identifying acoustic emission damage patterns of composite materials under thermal shock according to claim 1, characterized in that: The step 4 is specifically as follows: The parameters are specifically gas shock temperature, oxygen outlet pressure, oxygen flow rate, kerosene outlet pressure, kerosene flow rate, cooling gas outlet pressure and cooling gas flow rate; Among them, the gas impact temperature is 0-1300℃, the oxygen outlet pressure is above 1.3MPa, the oxygen flow rate is 230-500L / min, the kerosene outlet pressure is above 0.5MPa, the kerosene flow rate is 5-15L / h, the cooling gas outlet pressure is 0.7MPa, and the cooling gas flow rate is 60L / min.

6. The method for identifying acoustic emission damage patterns of composite materials under thermal shock according to claim 1, characterized in that: The step six specifically comprises: performing a substrate tensile test at room temperature based on an acoustic emission device to obtain a characteristic range of acoustic emission frequency for substrate cracking; Based on the acoustic emission device, the fiber bundle tensile test was carried out at room temperature to obtain the characteristic range of the acoustic emission frequency of fiber breakage; Based on the acoustic emission device, tensile tests of composite materials were carried out at room temperature, and failure modes mainly including matrix cracking, interface slip, interface debonding and fiber fracture damage were obtained. The obtained acoustic emission characteristic parameters were subjected to k-means cluster analysis. According to the acoustic emission frequency characteristic range of matrix cracking and fiber fracture obtained above, the frequency characteristic range of interface slip and interface debonding was obtained.

7. The method for identifying acoustic emission damage patterns of composite materials under thermal shock according to claim 1, characterized in that: The step seven is specifically as follows: Amplitude, peak frequency, duration, rise time, count and energy parameters are selected as characteristic parameters for cluster analysis, and the mean variance normalization method is used to make the selected characteristic parameters dimensionless, so that the measured data are defined between [0,1], and the expression is as shown in formula (1); In the formula, A i is the original signal parameter, A μ is the average value of the original signal parameter, is the signal parameter after standardization, A σ is the standard deviation of the original signal parameter; Secondly, the contour coefficient value is calculated for the dimensionless characteristic parameters, as shown in formula (2), to determine the optimal clustering number k; Where a(i) represents the average distance between the i-th vector and other vectors in the same category, and b(i,k) represents the average distance between the i-th vector and other vectors in different categories. Finally, perform k-means clustering analysis on the processed feature parameters, select k cluster centers, calculate the distance from the initial cluster centroid to the input vector, and then classify the input vector into the category of the nearest cluster centroid; calculate the position of the cluster center according to the shortest distance principle; repeat the above steps until the position of the initial cluster centroid no longer changes. Perform j-layer wavelet packet transform on an acoustic emission signal f(t) to obtain the sum of a series of decomposed parts, as follows: f(t) refers to an acoustic emission signal, is the decomposed signal at frequency band i and scale j. The energy of each part after decomposition is Defined as: The calculation formula of total energy E(t) is as follows: R i The wavelet packet energy spectrum coefficient is the ratio of the energy of each node to the total energy of all nodes after decomposition;

Images