A nondestructive testing method and system for carbon fiber reinforced composite materials

Abstract

The present invention discloses a nondestructive testing method and system for carbon fiber reinforced composite materials, including: assembling excitation electrodes at both ends of a carbon fiber reinforced composite material sample; connecting the two excitation electrodes to an alternating current source; measuring the potential distribution on the surface of the sample when an alternating current is passed at a certain height to obtain alternating potential gradient data; and placing the alternating potential gradient data into an image high-resolution reconstruction neural network model to obtain a potential distribution image. Under the excitation of a current signal, the change in conductivity is manifested as a sudden change in the surface potential gradient. The surface potential is tested using a potential sensor, and the potential distribution characteristics are analyzed to achieve quantitative detection of defects.

Description

Technical Field

[0001] The present invention relates to a detection technology, and in particular to a nondestructive detection method and system for carbon fiber reinforced composite materials. Background Art

[0002] Carbon fiber reinforced polymer (CFRP) is a composite material made of carbon fiber as reinforcement material and resin as matrix through winding, molding or pultrusion molding processes. It has the characteristics of light weight, high strength and high modulus, fatigue resistance, corrosion resistance, easy processing and molding, etc. It is widely used in aerospace, national defense, rail transportation, energy industry and other fields. It has played a significant role in both military and civilian fields, profoundly affecting national security, social economy, science and technology and human life. It has become an indispensable key material for the development of high-tech, defense cutting-edge weapons and equipment and other strategic industries in various countries around the world. However, due to the influence of many uncertain factors such as manufacturing process, working environment and load impact, various types of defects or damage will occur on the surface or inside of the material, such as pores, rich and poor glue, foreign inclusions, matrix cracks, fiber breakage and debonding and delamination. This directly affects the performance indicators of CFRP, threatens the reliability and safety of its structural parts, and may even lead to catastrophic consequences.

[0003] Based on conventional nondestructive testing technology, domestic and foreign scholars have conducted extensive research on the detection methods of CFRP defects. However, due to the characteristics of CFRP such as non-magnetic conductivity, low conductivity and anisotropy, and an insulating layer attached to the surface, conventional technologies have shown certain limitations. For example, the eddy current induced by eddy current detection technology inside CFRP is weak, making it more difficult to pick up the magnetic field disturbance signal at the defect. Ultrasonic detection technology requires the replacement of probes according to the size of the object and the use of coupling agents to increase the transmission efficiency of ultrasonic waves, which is not conducive to the detection of large structural parts. The imaging contrast of X-ray detection technology for carbon elements with lower atomic numbers is not high, the equipment is complex and there is radiation pollution, and the detection efficiency is low. Infrared thermal imaging detection technology is easily affected by the ambient temperature, and it is difficult to conduct on-site detection of CFRP structural parts in a high temperature environment. Electrical impedance detection technology is restricted by the anisotropy of CFRP conductivity, and the spatial resolution of the detection results is low. Summary of the invention

[0004] In order to solve the above problems, the present invention provides a nondestructive testing method for carbon fiber reinforced composite materials.

[0005] To achieve the above purpose, the technical solution adopted by the present invention is as follows:

[0006] A nondestructive testing method for carbon fiber reinforced composite materials, comprising the following steps:

[0007] Assembling excitation electrodes at both ends of the carbon fiber reinforced composite material sample;

[0008] Connecting the two excitation electrodes to an alternating current source;

[0009] Measuring the potential distribution on the surface of the sample when an alternating current is passed at a certain height to obtain alternating current potential gradient data;

[0010] The AC potential gradient data is put into the image high-resolution reconstruction neural network model to obtain a potential distribution image.

[0011] As a limitation: after the step of assembling excitation electrodes at both ends of the carbon fiber reinforced composite material sample, the method further includes: measuring the potential distribution of the electrostatic field when the surface of the sample is in electrostatic equilibrium at a certain height to obtain DC potential gradient data;

[0012] After measuring the potential distribution on the surface of the sample when an alternating current is passed through it at a certain height to obtain the AC potential gradient data, the method further includes: subtracting the AC potential gradient data from the DC potential gradient data, and then putting the calculated data into an image high-resolution reconstruction neural network to obtain a potential distribution image.

[0013] As a limitation: obtaining the electrostatic balance of the sample includes: loading a DC voltage on one side of the excitation electrode so that the surface of the sample becomes an equipotential surface.

[0014] As a limitation: obtaining the potential gradient data and the reference potential gradient data includes measuring the sample using a non-contact electrostatic potential sensor.

[0015] As a limitation: wherein the non-contact electrostatic potential sensors include a plurality of sensors forming a sensor array.

[0016] As a limitation: the image high-resolution reconstruction neural network includes a lateral inhibition network layer, feature extraction, nonlinear transformation and feature restoration, the digital neuron array of the lateral inhibition network layer corresponds to the sensor array, and in the lateral inhibition network layer, the input signal of the neuron in the xth row and yth column is I0, and the output signal is R, then:

[0017]

[0018] Among them, α ij is the side inhibition weight, M and N are the side inhibition radius.

[0019] A nondestructive testing system for carbon fiber reinforced composite materials, comprising:

[0020] Carbon fiber reinforced composite material samples;

[0021] A signal acquisition system, used to collect electric potential data on the surface of the carbon fiber reinforced composite material sample;

[0022] Excitation electrodes, mounted on both ends of the carbon fiber reinforced composite material sample;

[0023] A signal source, comprising a DC voltage source connected to one of the excitation electrodes and an AC current source connected to the excitation electrode;

[0024] A current limiting resistor is connected in series at both ends of the excitation electrode.

[0025] As a limitation: the signal acquisition system includes a plurality of non-contact electrostatic potential sensors fixed above the carbon fiber reinforced composite material sample, and the non-contact electrostatic potential sensors form a sensor array above the carbon fiber reinforced composite material sample.

[0026] As a limitation: each of the non-contact electrostatic potential sensors includes a sensing electrode, a conditioning amplifier circuit and a shielding tube, the sensing electrode is arranged in the shielding tube, the sensing electrode is connected to the conditioning amplifier circuit, and the conditioning amplifier circuit sends the potential data of the surface of the carbon fiber reinforced composite material sample detected by the sensing electrode to the host computer.

[0027] As a limitation: the conditioning amplifier circuit includes a resistor R in , capacitor C in and operational amplifier A, the resistor R in One end of the capacitor C in One end of the capacitor C in The other end of the resistor R in The other end is connected to the input end of the operational amplifier A, and the output end of the operational amplifier A is connected to the host computer.

[0028] Compared with the prior art, the present invention has the following technical advances:

[0029] The present invention discloses a nondestructive testing method for carbon fiber reinforced composite materials. Defects of carbon fiber reinforced composite materials will cause deformation of carbon fibers, thereby leading to changes in the spatial distribution of material conductivity. Under the excitation of a current signal, the change in conductivity is manifested as a sudden change in surface potential gradient. The surface potential is tested using a potential sensor, and quantitative detection of defects can be achieved by analyzing the potential distribution characteristics. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] The accompanying drawings are used to provide further understanding of the present invention and constitute a part of the specification. They are used to explain the present invention together with the embodiments of the present invention and do not constitute a limitation of the present invention.

[0031] In the attached picture:

[0032] Figure 1 It is a schematic diagram of the principle of the present invention.

[0033] Figure 2 This is a flowchart of the detection method provided in the first embodiment of the present invention.

[0034] Figure 3 Schematic diagram of a neural network for high-resolution reconstruction of images.

[0035] Figure 4 This is a flowchart of a detection method provided in the second embodiment of the present invention.

[0036] Figure 5 It is a schematic diagram of the nondestructive testing device of carbon fiber reinforced composite material of the present invention.

[0037] Figure 6 This is a measurement principle diagram of the non-contact electrostatic potential sensor of the present invention.

[0038] Figure 7 This is the equivalent circuit model of the non-contact electrostatic potential sensor of the present invention.

[0039] In the figure:

[0040] 1-Signal acquisition system, 2-Carbon fiber reinforced composite material sample, 3-Excitation electrode, 4-Signal source, 5-Current limiting resistor, 6-Conductive surface. DETAILED DESCRIPTION

[0041] Embodiment 1

[0042] This embodiment discloses a nondestructive testing system for carbon fiber reinforced composite materials. Defects in carbon fiber reinforced composite materials can cause deformation of carbon fibers, thereby causing changes in the spatial distribution of the material's electrical conductivity. The principle is as follows: Figure 1 As shown in the figure, under the stimulation of the current signal, the change of conductivity is manifested as a sudden change of the surface potential gradient. The surface potential of the carbon fiber reinforced composite material is measured by a non-contact electrostatic potential sensor, and the quantitative detection of defects can be achieved by analyzing the potential distribution characteristics. Figure 2 As shown, this embodiment specifically includes the following steps:

[0043] S100: Assemble excitation electrodes at both ends of the carbon fiber reinforced composite material sample;

[0044] S200: connecting the two excitation electrodes to an alternating current source;

[0045] S300: measuring the potential distribution on the surface of the sample when an alternating current is passed through it at a certain height to obtain alternating current potential gradient data;

[0046] S400: Putting the AC potential gradient data into an image high-resolution reconstruction neural network model to obtain a potential distribution image.

[0047] In this embodiment, excitation electrodes 3 are assembled at both ends of the carbon fiber reinforced composite material sample 2, because a part of the carbon fiber reinforced composite material is a composite material made of carbon fiber as a reinforcing material and a resin as a matrix through a forming process such as winding, molding or pultrusion. The resin as the matrix is ​​generally non-conductive, and the carbon fiber is conductive, so the carbon fiber is used as a conductor to connect to the excitation electrode 3.

[0048] The two excitation electrodes 3 are connected to an alternating current source. Specifically, because the alternating current passes through the two ends of the excitation electrode 3, a self-induced electromotive force appears in the conductor to resist the passage of the current, which results in a larger current density near the surface of the conductor. Since the self-induced electromotive force increases with the increase of frequency, the skin effect is also more significant with the increase of frequency. When a high-frequency current passes through the conductor, it can be considered that the current only flows through a very thin layer on the surface of the conductor, which is equivalent to a reduction in the cross-section of the conductor and an increase in resistance, which greatly reduces the effective utilization rate of the conductor material. Therefore, in this embodiment, the alternating current selects a low-frequency alternating current below 1000Hz, so that the current is as evenly distributed as possible inside the conductor to reduce the impact of the skin effect. Therefore, using a low-frequency alternating current to pass through the carbon fiber reinforced composite material sample 2 can make the current flow through most of the area inside the carbon fiber reinforced composite material sample 2 as much as possible. If there are defects such as bubbles, holes, and foreign objects inside the carbon fiber reinforced composite material sample 2, the potential gradient will suddenly change after the current passes.

[0049] In this embodiment, in order to obtain AC potential gradient data, a non-contact electrostatic potential sensor can be set directly above the carbon fiber reinforced composite material sample 2. For the test of the spatial distribution of electrostatic potential, since the spatial resolution and detection speed of the single sensor scanning method restrict each other, the contradiction is difficult to reconcile. In order to increase the accuracy, this embodiment adopts a multi-sensor parallel method, that is, a sensor array formed by a non-contact electrostatic potential sensor is used, and the detection speed of the sensor array is faster.

[0050] The measured AC potential gradient data is put into the image high-resolution reconstruction neural network model through a non-contact electrostatic potential sensor to obtain a potential distribution image. The image high-resolution reconstruction neural network includes a lateral inhibition network layer, feature extraction, nonlinear transformation and feature restoration. The digital neuron array of the lateral inhibition network layer corresponds to the sensor array. In the lateral inhibition network layer, the input signal of the neuron in the xth row and yth column is I0, and the output signal is R, then:

[0051]

[0052] Among them, α ijis the lateral inhibition weight, M and N are the lateral inhibition radii, and their values ​​are determined by the lateral inhibition function. The typical lateral inhibition function model is shown in the following table. The model describes the spatial distribution of the intensity of regulation of a certain neuron on surrounding neurons. The horizontal axis represents the lateral inhibition range, and the vertical axis represents the lateral inhibition weight. Positive values ​​represent excitatory regulation, and negative values ​​represent inhibitory regulation.

[0053]

[0054] Specifically, as shown in Figure 3, the sensor array corresponds to the lateral inhibition network layer one by one, the output of each sensor corresponds to the input of a neuron of the lateral inhibition neural network, the sensor array obtains AC potential data, and transmits the AC potential data to the lateral inhibition network layer, the data processed by the lateral inhibition network layer is subjected to convolution calculation to extract features to obtain low-resolution n1-dimensional features, the low-resolution n1-dimensional features are subjected to multiple convolution calculations and then subjected to nonlinear transformation to obtain high-resolution n1-dimensional features, the high-resolution n1-dimensional features are subjected to deconvolution calculations and feature restoration to obtain the sample surface potential distribution map. Figure 3 The darker color changes indicate the location of the potential mutation. The greater the color change, the greater the potential mutation and the more serious the defect.

[0055] Embodiment 2

[0056] This embodiment discloses a nondestructive testing system for carbon fiber reinforced composite materials. Defects in carbon fiber reinforced composite materials can cause deformation of carbon fibers, thereby causing changes in the spatial distribution of the material's electrical conductivity. Under the excitation of a current signal, the change in electrical conductivity is manifested as a sudden change in the surface potential gradient. The surface potential is tested using a potential sensor, and the potential distribution characteristics are analyzed to achieve quantitative detection of defects, such as Figure 4 As shown, this embodiment specifically includes the following steps:

[0057] S100: Assemble excitation electrodes at both ends of the carbon fiber reinforced composite material sample;

[0058] S101: connecting one electrode end of the excitation electrode to a direct current voltage source, so that the carbon fiber reinforced composite material sample becomes an equipotential body;

[0059] S102: measuring the potential distribution of the electrostatic field when the surface of the sample is in electrostatic equilibrium at a certain height to obtain DC potential gradient data;

[0060] S200: connecting the two excitation electrodes to an alternating current source;

[0061] S300: measuring the potential distribution on the surface of the sample when an alternating current is passed through it at a certain height to obtain alternating current potential gradient data;

[0062] S301: performing a subtraction operation on the AC potential gradient data and the DC potential gradient data;

[0063] S400: Putting the calculated data into the image high-resolution reconstruction neural network model to obtain the electric potential distribution image.

[0064] In this embodiment, after the step of assembling excitation electrodes at both ends of the carbon fiber reinforced composite material sample at S100, the method further includes: S102, measuring the potential distribution of the electrostatic field when the surface of the sample is in electrostatic equilibrium at a certain height to obtain DC potential gradient data.

[0065] Before step S102, the method further includes: S101, applying a DC voltage on one of the excitation electrodes to make the sample surface an equipotential surface.

[0066] After the step of measuring the potential distribution on the surface of the sample when an alternating current is passed at a certain height to obtain AC potential gradient data in S300, the method further includes: S301, performing a subtraction operation on the AC potential gradient data and the DC potential gradient data.

[0067] Different from implementation 1, the above three steps are added for the following reasons:

[0068] 1. To prevent the carbon fiber reinforced composite material sample 2 from being out of electrostatic equilibrium and affecting the measurement results, one electrode end of the excitation electrode 3 is connected to a DC voltage source, so that the carbon fiber reinforced composite material sample 2 becomes an equipotential body and an equipotential surface is formed on its surface;

[0069] 2. Because carbon fiber reinforced composite materials are mostly woven structures, if the potential distribution data is obtained by directly applying an alternating current to both ends of the carbon fiber reinforced composite material sample 2, the different distances from each sensor probe to the surface of the carbon fiber reinforced composite material sample 2 may cause uneven potential distribution, affecting the measurement results. Therefore, it is necessary to subtract the DC potential gradient data from the AC potential gradient data to avoid the above errors.

[0070] Both AC potential gradient data and DC potential gradient data are potential gradient data, which is a two-dimensional matrix. During DC scanning, the DC potential gradient data is recorded by the host computer software and recorded as V dc After the AC scan, the host computer software records the AC potential gradient data, recorded as V ac , V dc With V ac With the same size, matrix subtraction is used to obtain the result, and then the result is put into the image high-resolution reconstruction neural network model to obtain the potential distribution image.

[0071] refer to Figure 4, which shows a carbon fiber reinforced composite material nondestructive testing system provided by an embodiment of the utility model, which includes a carbon fiber reinforced composite material sample 2; a signal acquisition system 1, used to collect potential data on the surface of the carbon fiber reinforced composite material sample 2; an excitation electrode 3, assembled at both ends of the carbon fiber reinforced composite material sample 2; a signal source 4, including a DC voltage source connected to one electrode of the excitation electrode 3 and an alternating current source connected to the two electrodes of the excitation electrode 3; a current limiting resistor 5, connected in series at both ends of the excitation electrode 3, and when in use, the system is connected to a conductive surface 6 to form a circuit loop.

[0072] The signal acquisition system 1 includes a plurality of non-contact electrostatic potential sensors fixed above the sample. The non-contact electrostatic potential sensors form a sensor array above the carbon fiber reinforced composite material sample. Each non-contact electrostatic potential sensor includes a sensing electrode, a conditioning amplifier circuit and a shielding tube. The sensing electrode is used for sensing and is arranged in the shielding tube. The sensing electrode is connected to the conditioning amplifier circuit. The conditioning amplifier circuit sends the potential data of the surface of the carbon fiber reinforced composite material sample detected by the sensing electrode to the host computer. The non-contact electrostatic potential sensor is based on the principle of electrostatic induction. The conditioning amplifier circuit specifically includes a resistor R in , capacitor C in and operational amplifier A, resistor R in One end of the capacitor C in One end of the capacitor C in The other end of the resistor R in The other end is connected to the input end of the operational amplifier A, and the output end of the operational amplifier A is connected to the host computer. By measuring the distorted electrostatic field between the carbon fiber reinforced composite material sample 2 and the sensing electrode, the electrostatic potential of the carbon fiber reinforced composite material sample 2 is indirectly obtained.

[0073] like Figure 6 As described above, the sensing electrode of the non-contact electrostatic potential sensor is brought close to the carbon fiber reinforced composite material sample 2, and the equivalent coupling capacitance C between the sensing electrode and the carbon fiber reinforced composite material sample 2 is used. e , the electrostatic potential on the surface of the carbon fiber reinforced composite material sample 2 can be measured. In order to avoid the interference of the stray electrostatic field in space, the sensing electrode of the non-contact electrostatic potential sensor is usually equipped with a grounded shielding tube. Shielding depth S d It should be appropriate. Too deep shielding will cause the electrode sensing signal to attenuate, making it difficult for the conditioning and amplifier circuit to pick up; too shallow shielding will expand the measured area and reduce the spatial resolution of the test results.

[0074] The equivalent circuit of the non-contact electrostatic potential sensor is as follows: Figure 7 As shown, the potential of the surface of the carbon fiber reinforced composite material sample 2 is V s , the equivalent coupling capacitance between the sensing electrode and the test point is Ce , when the measurement distance is constant, C e It can be approximated as a constant. The input resistance of the sensor is R in , the input capacitance is C in , the output voltage is V i , then:

[0075]

[0076] From the above formula, we can see that the test result V i The amplifier input resistance R is usually required to decay exponentially with time t to avoid errors in DC and low-frequency quasi-DC testing. in Not less than 10 14 Ohms, to increase the input capacitance C in For resistor R in The time constant of the discharge.

[0077] Among them, the input coupling capacitance C between the sensing electrode and the carbon fiber reinforced composite material sample 2 is in The calculation formula is (approximately a parallel plate capacitor, ε is the dielectric constant of air, S is the surface area of ​​the sensing electrode, and d is the distance from the sensing electrode to the surface of the carbon fiber reinforced composite material sample) In order to increase the strength of the electrostatic test signal, it is necessary to increase the input capacitance. ε and S are fixed values. C can be increased by reducing d. in When d is below 1 mm, the electrostatic induction signal intensity can reach the mV level, meeting the sensitivity requirements of the subsequent test circuit.

[0078] Finally, it should be noted that the above description is only a preferred embodiment of the utility model and is not intended to limit the utility model. Although the utility model is described in detail with reference to the aforementioned embodiments, those skilled in the art can still modify the technical solutions described in the aforementioned embodiments or replace some of the technical features therein with equivalents. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the utility model shall be included in the scope of protection of the claims of the utility model.

Claims

1. A nondestructive testing method for carbon fiber reinforced composite materials, characterized in that: The following steps are involved: Assembling excitation electrodes at both ends of the carbon fiber reinforced composite material sample; Connecting the two excitation electrodes to an alternating current source; Measuring the potential distribution on the surface of the sample when an alternating current is passed at a certain height to obtain alternating current potential gradient data; Putting the AC potential gradient data into an image high-resolution reconstruction neural network model to obtain a potential distribution image; After the step of assembling excitation electrodes at both ends of the carbon fiber reinforced composite material sample, the method further includes: measuring the potential distribution of the electrostatic field when the surface of the sample is in electrostatic equilibrium at a certain height to obtain DC potential gradient data; After measuring the potential distribution on the surface of the sample when an alternating current is passed through it at a certain height to obtain the alternating potential gradient data, the method further includes: performing a subtraction operation on the alternating potential gradient data and the direct current potential gradient data, and then putting the calculated data into an image high-resolution reconstruction neural network to obtain a potential distribution image; The image high-resolution reconstruction neural network includes a lateral inhibition network layer, feature extraction, nonlinear transformation and feature restoration. The digital neuron array of the lateral inhibition network layer corresponds to the sensor array. In the lateral inhibition network layer, the input signal of the neuron in the xth row and yth column is I 0, the output signal is R ,but: ； in, α ij is the side inhibition weight, M and N are the side inhibition radius.

2. A nondestructive testing method for carbon fiber reinforced composite materials according to claim 1, characterized in that: The step of obtaining the electrostatic balance on the sample surface includes: applying a DC voltage on one of the excitation electrodes to make the sample surface an equipotential surface.

3. A nondestructive testing method for carbon fiber reinforced composite materials according to claim 2, characterized in that: The obtaining of the AC potential gradient data and the DC potential gradient data includes measuring the sample using a non-contact electrostatic potential sensor.

4. A nondestructive testing method for carbon fiber reinforced composite materials according to claim 3, characterized in that: The non-contact electrostatic potential sensors include a plurality of sensors forming a sensor array.

5. A nondestructive testing system for carbon fiber reinforced composite materials using the method of claim 1, characterized in that: include: Carbon fiber reinforced composite material samples; A signal acquisition system, used to collect electric potential data on the surface of the carbon fiber reinforced composite material sample; Excitation electrodes, mounted on both ends of the carbon fiber reinforced composite material sample; A signal source, comprising a DC voltage source connected to one of the excitation electrodes and an AC current source connected to the excitation electrode; A current limiting resistor is connected in series at both ends of the excitation electrode.

6. A carbon fiber reinforced composite material nondestructive testing system according to claim 5, characterized in that: The signal acquisition system comprises a plurality of non-contact electrostatic potential sensors fixed above the carbon fiber reinforced composite material sample, and the non-contact electrostatic potential sensors form a sensor array above the carbon fiber reinforced composite material sample.

7. A carbon fiber reinforced composite material nondestructive testing system according to claim 6, characterized in that: Each of the non-contact electrostatic potential sensors includes a sensing electrode, a conditioning amplifier circuit and a shielding tube. The sensing electrode is arranged in the shielding tube. The sensing electrode is connected to the conditioning amplifier circuit. The conditioning amplifier circuit sends the potential data of the surface of the carbon fiber reinforced composite material sample detected by the sensing electrode to the host computer.

8. The nondestructive testing system for carbon fiber reinforced composite materials according to claim 7, characterized in that: The conditioning amplifier circuit includes a resistor R in ,capacitance C in and operational amplifiers A， The resistance R in Connect one end of the capacitor C in One end of the capacitor C in The other end of the resistor is grounded. R in The other end of the operational amplifier A The input terminals of the operational amplifier are connected A The output end is connected to the host computer.

Images