Pyrolytic carbon coating crack initiation on-line monitoring method based on acoustic emission signals

Abstract

The application provides an online pyrolytic carbon coating crack initiation monitoring method based on acoustic emission signals. The method obtains the current temperature field distribution and hierarchical structure parameters of the multi-layer pyrolytic carbon coating structure, determines the equivalent transmission gain of the multi-layer pyrolytic carbon coating structure in each frequency band based on the temperature field distribution and hierarchical structure parameters, then, according to the equivalent transmission gain, the monitoring frequency band and the front-end gain of the acoustic emission monitoring device are adaptively configured, the configured monitoring frequency band is aligned to the frequency band with the equivalent transmission gain higher than the preset threshold, and finally, on the configured monitoring frequency band, the crack initiation feature is extracted and the crack initiation recognition is performed, so that the detectability of the crack initiation acoustic emission signal under high-temperature working conditions is improved.

Description

Technical Field

[0001] This application relates to crack monitoring technology, and more particularly to an online monitoring method for crack initiation in pyrolytic carbon coatings based on acoustic emission signals. Background Technology

[0002] Pyrolytic carbon coatings are widely used in high-temperature protective components to withstand ablation, thermal shock, and mechanical impact loads in ultra-high temperature environments.

[0003] Acoustic emission technology is commonly used in engineering to achieve online monitoring of early microcracks and interface debonding within such protective coatings. However, under operating conditions exceeding 1800℃, pyrolytic carbon coatings typically exhibit a multi-layered, non-uniform structure, with material parameters changing significantly with temperature. As a result, sound waves experience strong attenuation, mode scattering, and frequency distortion during propagation, causing the acoustic emission signals generated during the crack initiation stage to be severely attenuated before reaching the sensor.

[0004] Existing acoustic emission monitoring systems are typically designed with monitoring frequency bands and amplifier gain parameters based on transmission assumptions under normal or medium temperature conditions. This makes them unsuitable for the frequency-selective attenuation characteristics of high-temperature multilayer structures, resulting in a lack of reliable detection capability for crack initiation stages under actual service conditions. Summary of the Invention

[0005] This application provides an online monitoring method for crack initiation in pyrolytic carbon coatings based on acoustic emission signals, which improves the detectability of acoustic emission signals for crack initiation in multilayer pyrolytic carbon coating structures under high-temperature conditions.

[0006] This application provides an online monitoring method for crack initiation in pyrolytic carbon coatings based on acoustic emission signals, including: Obtain the current temperature field distribution and hierarchical structure parameters of the multilayer pyrolytic carbon coating structure; Based on the temperature field distribution and hierarchical structure parameters, the equivalent transmission gain of the multilayer pyrolytic carbon coating structure in each frequency band is determined. Based on the equivalent transmission gain, the monitoring frequency band and front-end gain of the acoustic emission monitoring device are adaptively configured so that the configured monitoring frequency band is aligned with the frequency band where the equivalent transmission gain is higher than a preset threshold. On the configured monitoring frequency band, crack initiation features are extracted and crack initiation identification is performed.

[0007] The online monitoring method for crack initiation in pyrolytic carbon coatings based on acoustic emission signals provided in this application obtains the current temperature field distribution and hierarchical structure parameters of the multilayer pyrolytic carbon coating structure. Based on the temperature field distribution and hierarchical structure parameters, the equivalent transmission gain of the multilayer pyrolytic carbon coating structure in each frequency band is determined. Then, according to the equivalent transmission gain, the monitoring frequency band and front-end gain of the acoustic emission monitoring device are adaptively configured so that the configured monitoring frequency band is aligned with the frequency band where the equivalent transmission gain is higher than a preset threshold. Finally, crack initiation features are extracted and crack initiation identification is performed on the configured monitoring frequency band to improve the detectability of acoustic emission signals for crack initiation under high-temperature conditions. Attached Figure Description

[0008] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0009] Figure 1 This is a schematic flowchart illustrating an online monitoring method for crack initiation in pyrolytic carbon coatings based on acoustic emission signals, according to an example embodiment of this application. Figure 2 This is a schematic diagram of the active waveguide excitation process shown in another example embodiment of this application; Figure 3 This is a schematic diagram of the path difference compensation process according to another example embodiment of this application.

[0010] The accompanying drawings illustrate specific embodiments of this application, which will be described in more detail below. These drawings and descriptions are not intended to limit the scope of the concept in any way, but rather to illustrate the concept of this application to those skilled in the art through reference to particular embodiments. Detailed Implementation

[0011] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims.

[0012] In the actual service environment of ultra-high temperature protective components, the outer multi-layer pyrolytic carbon coating needs to withstand extreme thermal loads exceeding 1800℃. When applying acoustic emission technology to such components to identify early crack initiation, the multi-layer structure causes severe frequency selective attenuation under high temperature conditions. Traditional acoustic emission systems with fixed monitoring frequencies often cannot receive effective signals, making it difficult to identify the crack initiation state in a timely manner.

[0013] Therefore, this application needs to solve the problems of uneven frequency band attenuation during the propagation of acoustic emission signals in high-temperature multilayer pyrolytic carbon coating structures, which makes it difficult to determine the effective transmittable frequency band, the inability of fixed monitoring frequency bands to adapt to real working conditions, and the easy omission of crack initiation signals, so as to achieve reliable detection of crack initiation acoustic emission events in extreme environments.

[0014] It is worth noting that the aforementioned technical problems arise primarily from the nonlinear changes in material parameters of the multilayer pyrolytic carbon coating under high-temperature conditions, the enhanced reflection and scattering caused by abrupt changes in multi-interface impedance, and the significant dissipation of high-frequency energy, resulting in a marked frequency-dependent attenuation of sound waves during propagation. Particularly at temperatures exceeding 1600℃, mid-to-high frequency signals are often completely untransmittable, while the low-frequency band is relatively preserved, making it impossible to match the fixed monitoring frequency band with the actual transmittable frequency band.

[0015] This application aims to construct a propagation model of a multi-layer structure in different frequency bands using the transmission matrix method, and to calculate the equivalent transmission gain of each frequency band under the current temperature field constraint. Then, based on this transmission gain, an effective frequency band is automatically selected as the monitoring frequency band, and the gain is adjusted through a programmable front end to ensure that the system's sensitive bandwidth matches the actual transmittable frequency band. By extracting low-frequency features related to crack initiation within the matched frequency band, a stable detection capability for crack initiation can be maintained in a severely attenuated high-temperature propagation environment.

[0016] In one exemplary application scenario, during ground-based hot commissioning of an ultra-high temperature protective component, the outer surface of the component is covered with multiple layers of pyrolytic carbon coating, and the service temperature can reach above 1800°C. To monitor the early crack initiation state of the coating, the test engineers installed an acoustic emission sensor in the low-temperature zone of the component's outer wall, hoping to capture acoustic emission events of crack initiation in the high-temperature region. However, in the initial testing phase, the engineers found that the fixed monitoring frequency band of 150kHz to 300kHz used by traditional acoustic emission systems could hardly receive effective event waveforms under high-temperature conditions. Even when a known microcrack load was artificially applied, the sensor could only obtain a waveform dominated by noise. After on-site diagnosis, it was found that when the coating temperature exceeded 1600°C, the attenuation of the mid-to-high frequency acoustic emission signal propagating in the coating increased sharply, making this frequency band almost completely untransmittable.

[0017] To address this issue, an infrared thermography array was used to obtain the temperature field distribution of the multilayer coating in the region before the experiment. The thickness, density, and elastic constants of each material layer under high-temperature conditions were also obtained from the design data. Based on this, the sound wave propagation path from the potential crack initiation location to the sensor location was modeled using the transmission matrix method, yielding the equivalent transmission gain in the discrete frequency range of 20kHz to 500kHz. The results show that the transmission gain above 150kHz is significantly lower than -40dB, while the low-frequency band of 40kHz to 80kHz still possesses an acceptable transmission gain of -10dB to -15dB, which is within the effective penetration range of high-temperature multilayer structures.

[0018] Based on the aforementioned transmission gain distribution, the monitoring system automatically switched the main monitoring frequency band to the 40kHz–80kHz range before the formal test, and simultaneously increased the low-frequency gain of the programmable front-end amplifier to 40dB, thereby enhancing the transmittable low-frequency band. During the subsequent hot-run test, the system successfully captured multiple crack initiation events under the above configuration conditions. These events exhibited obvious low-frequency envelope abrupt changes and wave group morphology variations.

[0019] Figure 1 This is a schematic flowchart illustrating an online monitoring method for crack initiation in pyrolytic carbon coatings based on acoustic emission signals, according to an example embodiment of this application. Figure 1 As shown, the method provided in this embodiment includes: S100: Obtain the current temperature field distribution and hierarchical structure parameters of the multilayer pyrolytic carbon coating structure.

[0020] In this step, an infrared thermal imager with high-temperature radiation calibration function can be used to acquire radiation images of the surface of the multilayer pyrolytic carbon coating. Then, based on the current surface emissivity and combined with the high-temperature thermophysical parameters of each material layer of the coating, a radiation inversion algorithm is used to determine the temperature of the coating surface and the near-surface region. The temperature of the deep material is then calculated based on the heat conduction model of the multilayer structure to obtain the current temperature field distribution.

[0021] The hierarchical structure parameters can be obtained by reading the thickness, density, and elastic constants of each material layer from the process design document before the experiment. The elastic constants are then corrected for temperature based on the temperature dependence model of the pyrolytic carbon material to obtain the hierarchical structure parameters.

[0022] Alternatively, in another possible implementation, the thickness distribution of each material layer can be obtained by performing X-ray tomography on the multilayer coating before the component enters service. The density and temperature-related elastic parameters of each material layer can then be retrieved from a material database of the component's manufacturing batch to form the hierarchical structure parameters.

[0023] S200. Based on temperature field distribution and hierarchical structure parameters, the equivalent transmission gain of the multilayer pyrolytic carbon coating structure in each frequency band is determined.

[0024] In this step, the wave impedance and propagation constant can be determined for each material layer, and an equivalent transmission matrix can be constructed from the acoustic emission source position to the sensor position using the transmission matrix method. Based on the equivalent transmission matrix, the transmission gain at each discrete frequency point can be determined to obtain the equivalent transmission gain.

[0025] Specifically, the material density, elastic modulus, and Poisson's ratio at the corresponding temperature can be determined based on the current temperature field distribution. Then, the longitudinal and transverse wave velocities are determined based on the density and elastic modulus of each material layer at the current temperature, and the wave impedance is determined based on the wave velocity and density. Finally, the propagation constant is determined based on the wave velocity and the material layer thickness to construct the equivalent transmission matrix.

[0026] Specifically, determining the propagation constant based on wave velocity and material layer thickness can be achieved by calculating the corresponding phase constant at each discrete frequency point within a preset frequency range, based on the frequency at that frequency point and the P-wave velocity and / or S-wave velocity. The phase constant is then multiplied by the corresponding material layer thickness to obtain the cumulative phase of the material layer at that frequency point. When considering intrinsic material losses, an attenuation term is determined based on the product of a preset frequency-related attenuation coefficient and the material layer thickness. The cumulative phase and the attenuation term are then used together to characterize the propagation constant of the material layer at the corresponding frequency point.

[0027] Optionally, the phase constant can be calculated by multiplying 2π by the ratio of frequency to wave speed, the attenuation term can be calculated by multiplying the attenuation coefficient by the thickness of the material layer, and the propagation constant can be introduced in exponential form when constructing a single-layer transmission matrix.

[0028] Next, single-layer transmission matrices are established for each material layer in the order of the acoustic emission propagation path. All single-layer transmission matrices are then multiplied in stacked order to obtain the equivalent transmission matrix from the acoustic emission source location to the sensor location.

[0029] Then, by setting several discrete frequency points within a preset frequency range, the frequency response of the equivalent transmission matrix is ​​solved frequency by frequency. The magnitude of the frequency response is used as the transmission gain at each discrete frequency point, and this is used to construct the equivalent transmission gain. The preset frequency range can be from 20kHz to 500kHz, and discrete frequency points are set at logarithmic intervals within this range to improve the accuracy of the transmission gain solution in the high-frequency region.

[0030] In one specific embodiment, the aforementioned S200 can be applied to a high-temperature combustion component undergoing a ground ignition test. The component's surface is covered with a multi-layered pyrolytic carbon coating of four different densities and heat treatment processes. During the test, each layer exhibits significant wave velocity differences due to temperature variations. At a certain moment, a microcrack initiation event occurs near the interface between the third coating layer and the substrate. The transient acoustic emission wave must penetrate the four layers of pyrolytic carbon coating and a protective coupling layer to reach the acoustic emission sensor deployed on the outer surface.

[0031] After the monitoring system receives the real-time temperature field distribution, this embodiment first determines the corresponding longitudinal and transverse wave velocities based on the density and elastic modulus of each material layer at its current temperature point, and further obtains the wave impedance of each layer at the current temperature. Subsequently, at multiple discrete frequency points from 50kHz to 700kHz, the propagation constant of each material layer at each frequency point is calculated using the wave velocity and material layer thickness, and a single-layer transmission matrix is ​​constructed accordingly. After constructing the single-layer transmission matrix, all single-layer transmission matrices are multiplied layer by layer according to the stacking order of the actual propagation path of the acoustic emission wave to obtain the overall equivalent transmission matrix from the crack initiation position to the acoustic emission sensor position.

[0032] After obtaining the overall equivalent transmission matrix, this embodiment calculates the frequency response modulus of the matrix for all preset discrete frequency points, and uses it as the equivalent transmission gain of the multilayer pyrolytic carbon coating structure at the corresponding frequency points. Under the current temperature field, the coating's transmission gain for elastic waves in the 200kHz–260kHz range is significantly higher than other frequency bands, reaching −18dB, while the transmission gain drops to below −45dB in frequency bands above 400kHz. Based on this, the monitoring system configures the 200kHz–260kHz range as a high-priority monitoring band, while appropriately increasing the amplifier gain in the adjacent 150kHz–180kHz range, shielding the high-loss frequency band above 400kHz, and extracting and identifying crack initiation characteristics on the configured monitoring band.

[0033] S300: Based on the equivalent transmission gain, adaptively configure the monitoring frequency band and front-end gain of the acoustic emission monitoring equipment.

[0034] In this step, the monitoring frequency band and front-end gain of the acoustic emission monitoring device can be adaptively configured based on the equivalent transmission gain, so that the configured monitoring frequency band is aligned with the frequency band where the equivalent transmission gain is higher than a preset threshold.

[0035] Specifically, frequency bands with equivalent transmission gain higher than the first threshold can be configured as high-priority monitoring bands, frequency bands with equivalent transmission gain between the first and second thresholds can be configured as secondary-priority monitoring bands and assigned higher amplifier gain to the secondary-priority monitoring bands, and frequency bands with equivalent transmission gain lower than the second threshold can be configured as shielded bands.

[0036] The above steps obtain the equivalent transmission gain of the multilayer coating under the current temperature field and hierarchical structure conditions using the transmission matrix method. Based on this gain distribution, discrete frequency points are classified by threshold, with high-gain regions considered as effectively transmittable frequency bands. Subsequently, the monitoring system readjusts the center frequency and bandwidth of the adjustable monitoring channel according to these frequency bands, aligning it with the high-gain frequency band. Simultaneously, the amplifier gain is increased for the medium-gain frequency band, and the low-gain frequency band is shielded. This allows the system to automatically focus on frequency bands with low propagation loss and high crack initiation signal retention under different temperatures and operating conditions, thereby improving the effective bandwidth and signal-to-noise ratio of the signal. This makes crack initiation characteristics such as envelope rising edge and group velocity shift more prominent, ultimately improving the accuracy of crack initiation identification.

[0037] Furthermore, the specific understanding of aligning the configured monitoring frequency band to a frequency band with an equivalent transmission gain higher than a preset threshold refers to performing threshold judgment on discrete frequency points with equivalent transmission gain within a preset monitoring frequency range to obtain a set of frequency points with equivalent transmission gain higher than the preset threshold. Then, the center frequency of the adjustable monitoring channel of the acoustic emission monitoring device is configured to be within the frequency range of the set of frequency points, and the bandwidth of the adjustable monitoring channel is set to cover at least a portion of the set of frequency points.

[0038] Furthermore, the above-mentioned configuration of the center frequency of the adjustable monitoring channel to be within the frequency range of the frequency point set can be achieved by identifying continuous or approximately continuous frequency bands in the frequency point set and setting the center frequency of the adjustable monitoring channel to point to the frequency center of the continuous or approximately continuous frequency band.

[0039] S400: On the configured monitoring frequency band, extract crack initiation features and perform crack initiation identification.

[0040] Specifically, after completing the adaptive configuration of the monitoring frequency band and front-end gain, real-time feature extraction can be performed on the acoustic emission signals located within the monitoring frequency band.

[0041] First, bandpass filtering is performed on the acoustic emission signal to retain only the effective frequency components configured as high-priority and low-priority monitoring bands.

[0042] Then, the wave group morphology features are extracted by techniques such as short-time Fourier transform, wavelet packet decomposition or envelope demodulation. The steepness of the rising edge of the envelope is obtained by calculating the slope of the envelope curve in the rising phase. The arrival time of different frequency bands is compared by using the group velocity estimation algorithm to extract the group velocity offset. The event time sequence features are determined by combining the time interval pattern between events.

[0043] Subsequently, based on at least one of the aforementioned features, the known feature patterns of crack initiation events are compared. A threshold-based pattern matching method or a statistical learning-based binary classification method, such as a Gaussian model or a linear discriminant algorithm, is used to perform crack initiation identification, thereby determining whether the current acoustic emission event corresponds to crack initiation behavior.

[0044] It is worth noting that the embodiments of this application aim to solve the problems of uneven frequency band attenuation during the propagation of acoustic emission signals in high-temperature multilayer pyrolytic carbon coating structures, which makes it difficult to determine the effective transmittable frequency band, the inability of fixed monitoring frequency bands to adapt to real working conditions, and the easy omission of crack initiation signals. As for the specific model for crack initiation identification, this embodiment does not impose specific limitations, only requiring that it can achieve the identification function.

[0045] In this embodiment, the current temperature field distribution and hierarchical structure parameters of the multilayer pyrolytic carbon coating structure are obtained. Based on these parameters, the equivalent transmission gain of the multilayer pyrolytic carbon coating structure in each frequency band is determined. Then, according to the equivalent transmission gain, the monitoring frequency band and front-end gain of the acoustic emission monitoring device are adaptively configured so that the configured monitoring frequency band is aligned with a frequency band where the equivalent transmission gain is higher than a preset threshold. Finally, crack initiation features are extracted and crack initiation identification is performed on the configured monitoring frequency band, thereby achieving dynamic alignment between the monitoring frequency band and the effective transmittable frequency band, thus improving the detectability of crack initiation acoustic emission signals under high-temperature conditions. Specifically, when the multilayer structure causes severe attenuation of mid-to-high frequency signals, a frequency band with higher transmission gain can be actively selected as the monitoring window, and the observability of weak signals can be improved through gain configuration. This allows the system to effectively distinguish the envelope characteristics, wave group morphology, and low-frequency energy changes of the crack initiation stage under real-world conditions. This technology can directly overcome the problem of missed detection caused by fixed frequency band monitoring in traditional methods, thereby achieving stable and reliable monitoring of the crack initiation stage of pyrolytic carbon coatings.

[0046] It is also worth noting that the method provided in this embodiment can be applied to multilayer pyrolytic carbon coating structures with local temperatures exceeding 1800°C.

[0047] To further understand the above embodiments, a specific explanation can be provided in conjunction with the ground-based hot-fire test scenario of the ultra-high temperature protective component. In this scenario, upon startup, the system first utilizes the temperature sensing array on the component surface and the interlayer thickness and material parameters recorded during coating manufacturing to obtain the current temperature field distribution and hierarchical structure parameters of the multilayer pyrolytic carbon coating at the test moment. Based on this, the system performs real-time correction of the material elastic parameters in each temperature range and further constructs a multilayer medium propagation model from the potential crack initiation location to the acoustic emission sensor in the low-temperature region.

[0048] Subsequently, based on the acquired temperature field and hierarchical structure parameters, the system calculates the wave impedance and propagation constant of each layer of material, and constructs the equivalent transmission matrix of the entire coating structure at discrete frequency points using the transmission matrix method. Through this equivalent transmission matrix, the system obtains the equivalent transmission gain of the coating in different frequency bands. This gain characterizes the ability of signals in each frequency band to be successfully transmitted to the sensor when an acoustic emission event occurs in the high-temperature region.

[0049] After obtaining the equivalent transmission gain, the system adaptively configures the monitoring frequency bands according to preset gain thresholds. When the equivalent transmission gain of a frequency band is higher than the first threshold, the band is automatically set as a high-priority monitoring band; when the transmission gain is between the first and second thresholds, the band is classified as a secondary-priority monitoring band and assigned a higher front-end amplifier gain; if the gain is lower than the second threshold, the band is shielded to avoid introducing excessive noise. Through this dynamic configuration process, the available frequency bands of the acoustic emission monitoring equipment are always automatically aligned with the transmittable frequency bands of the structure during the experiment, overcoming the monitoring blind zone problem caused by the rapid change in frequency attenuation under high-temperature conditions.

[0050] After implementing adaptive configuration of monitoring frequency bands, the system extracts crack initiation event features only from confirmed transmittable frequency bands. These features may include at least one of the following: group morphology changes, steepness of the envelope rising edge, group velocity offset, and event timing-based mode features in the transmittable low-frequency bands. The system utilizes these features to identify early crack initiation behavior in high-temperature coatings, enabling online monitoring of the service status of high-temperature protective components.

[0051] In the above embodiments, the elastic modulus, density, interfacial adhesion, and pore structure of the multilayer pyrolytic carbon coating continuously change dynamically under ultra-high temperature conditions, causing the propagation constant and wave impedance of sound waves to exhibit significant time-varying characteristics in different temperature regions. This leads to a time-varying frequency response function of the sound wave propagation path. However, since the energy of passive acoustic emission events is unknown and uncontrollable, when the structure's propagation function changes at high temperatures, it is usually impossible to deduce the true event waveform from the acquired distorted waveform. Therefore, compensation cannot be achieved using a fixed model or fixed filter, resulting in distortion or disappearance of the crack initiation signal at the sensor end.

[0052] In response, Figure 1 Based on the illustrated embodiment, Figure 2 This is a schematic diagram illustrating the active waveguide excitation process according to another exemplary embodiment of this application. For example... Figure 2 As shown, the active guided wave excitation steps provided in this embodiment include: S510. Periodically emit elastic waves of known spectrum on a multilayer pyrolytic carbon coating structure using a waveguide exciter.

[0053] In this step, the waveguide exciter can be fixed to the surface of the multilayer pyrolytic carbon coating structure or to a position in close contact with it. A driving voltage or electromagnetic excitation is applied to the multilayer pyrolytic carbon coating structure within a preset time period, so that the waveguide exciter outputs an elastic wave signal with a known amplitude frequency distribution and phase frequency distribution within a preset frequency range, and the elastic wave propagates in the multilayer pyrolytic carbon coating structure in a predetermined direction.

[0054] The elastic wave with the known spectral shape mentioned above can be a narrowband sinusoidal pulse or a linear sweep frequency signal, and the waveguide exciter can be a piezoelectric element or an electromagnetic acoustic transducer. Furthermore, it is worth noting that in the ground-based hot-testing scenario of the aforementioned ultra-high temperature protective component, the waveguide exciter is located in the low-temperature zone of the component's outer wall.

[0055] S520: Receive the response of guided wave excitation through an acoustic emission sensor, and determine the transfer function of the current structure based on the excitation signal and the response signal.

[0056] In this step, the drive signal of the waveguide exciter and the waveguide response signal received by the acoustic emission sensor are acquired simultaneously. The drive signal and response signal are then subjected to Fourier transform to obtain the frequency domain excitation spectrum and frequency domain response spectrum, respectively. At each discrete frequency point, the frequency domain response spectrum is divided by the frequency domain excitation spectrum to obtain the frequency domain transfer function characterizing the current propagation characteristics of the multilayer pyrolytic carbon coating structure.

[0057] S530. Use the transfer function to perform frequency domain pre-compensation on the signal of the passive acoustic emission event, and perform crack initiation identification on the compensated signal.

[0058] In this step, the original time-domain signal of the passive acoustic emission event is subjected to Fourier transform to obtain its frequency-domain signal. The frequency-domain signal is divided by the amplitude and phase frequency components of the transfer function at each discrete frequency point to obtain the compensated frequency-domain signal. The compensated frequency-domain signal is then subjected to inverse Fourier transform to obtain the time-domain acoustic emission signal with waveform distortion corrected. Crack initiation identification is then performed on the compensated time-domain acoustic emission signal.

[0059] In the above embodiments, by periodically emitting narrowband sinusoidal pulses or swept-frequency signals with known spectral shapes onto the coating surface, and by calculating the structure's transfer function in real time based on the relationship between excitation and response, the instantaneous frequency response of the coating under different temperature conditions can be accurately determined. Furthermore, this real-time transfer function can be used for frequency domain pre-compensation of passive acoustic emission events. This effectively restores the frequency-selective attenuation experienced by passive events during transmission within the coating, thereby making the high-loss frequency bands that were originally unable to penetrate the coating identifiable again. This achieves high signal-to-noise ratio and high-fidelity reconstruction of crack initiation signals under ultra-high temperature conditions, effectively solving the problem of acoustic emission signal disappearance caused by high temperatures.

[0060] Furthermore, in the above embodiments, on the one hand, the elastic modulus, density and damping coefficient of each layer of the multilayer coating change significantly with temperature, and the interface bonding quality and residual stress state also evolve with thermal cycling, resulting in the wave impedance distribution and propagation constant of the coating system for elastic waves being highly non-uniform with time and space, making the transmission function of the main sensing path strongly temperature-dependent, strongly frequency-dependent and strongly position-dependent.

[0061] On the other hand, the near-field region of acoustic emission sources is usually located in a high-temperature region. External sensors can only obtain signals by traversing a complex path through multiple coatings and coupling layers, and cannot obtain an approximate intrinsic response as a reference in the near-field region of the source, resulting in a highly uncertain estimation of the main path transfer function.

[0062] Furthermore, it is difficult to set up conventional comparative test specimens or repeatable excitation sources for periodic calibration in high-temperature environments, making it impossible to easily obtain the instantaneous transmission function through active waveguides. Single-path compensation often assumes stable or approximately uniform transmission characteristics, failing to capture path differences caused by local coating thickness variations, debonding, ablation, etc. This results in varying degrees of distortion of crack initiation characteristics at the sensor end, leading to problems such as frequency band mismatch and reversed amplitude relationships.

[0063] In response, Figure 1 Based on the illustrated embodiment, Figure 3 This is a schematic diagram illustrating the path difference compensation process according to another example embodiment of this application. Figure 3 As shown, the path differential compensation steps provided in this embodiment include: S610. Set the main acoustic emission sensing path and the reference acoustic emission sensing path in the same pyrolytic carbon coating area.

[0064] In this step, a main acoustic emission sensing path and a reference acoustic emission sensing path are set in the same pyrolytic carbon coating area. The sensor signal of the main acoustic emission sensing path passes through multiple coating layers and a coupling layer, while the sensor signal of the reference acoustic emission sensing path arrives via a propagation path less affected by the multiple coating layers.

[0065] Optionally, the aforementioned reference acoustic emission sensing path is achieved by a waveguide embedded along the substrate, with one end of the waveguide close to the crack initiation region and the other end connected to the reference sensor in the low-temperature region.

[0066] Specifically, at least one primary acoustic emission sensor can be arranged on the outer surface of the high-temperature region of the multilayer pyrolytic carbon coating structure, and the primary acoustic emission sensor can be coupled to the outer surface of the multilayer pyrolytic carbon coating structure through a coupling layer to form a primary acoustic emission sensing path.

[0067] A waveguide is embedded along a predetermined path inside the substrate of a multilayer pyrolytic carbon coating structure. A waveguide acoustic interface is set at one end of the waveguide near the area where cracks may occur, and a reference acoustic emission sensor is set at the other end of the waveguide away from the high-temperature area to form a reference acoustic emission sensing path.

[0068] The primary acoustic emission sensor is preferably a high-temperature acoustic emission sensor, such as a fiber optic acoustic emission sensor. The reference acoustic emission sensor is preferably located in a low-temperature region where the temperature is below a preset operating upper limit. The waveguide is preferably at least one of a metal rod, a metal tube, or a ceramic waveguide.

[0069] For the sensor signal in the main acoustic emission sensing path, after passing through multiple layers of coating and coupling layer, the acoustic emission signal generated from the crack initiation region can be transmitted sequentially through each layer of the multilayer pyrolytic carbon coating structure, through the interface between the coating and the substrate, and through the coupling layer between the main acoustic emission sensor and the outer surface of the coating to the main acoustic emission sensor, and the main acoustic emission sensor converts the mechanical wave signal after passing through the above propagation path into an electrical signal.

[0070] Furthermore, if the sensor signal of the reference acoustic emission sensing path arrives via a propagation path less affected by the multilayer coating, the acoustic emission signal generated from the crack initiation region can be coupled into one end of the waveguide through the substrate material or a local structure connected to the crack initiation region, propagating along its length in the waveguide, and then received and converted into an electrical signal by the reference acoustic emission sensor at the other end of the waveguide, away from the high-temperature region. This propagation path is less dependent on the multilayer pyrolytic carbon coating and its coupling layer than the main acoustic emission sensing path.

[0071] It is worth noting that in multilayer pyrolytic carbon coating structures used in environments with local temperatures exceeding 1800°C, since it is impossible to directly place the reference acoustic emission sensor in the high-temperature region, the above steps require extracting the reference acoustic emission signal by embedding a waveguide along the substrate. However, under ultra-high temperature conditions, the waveguide still exhibits problems such as interface stiffness fluctuations with temperature, guided wave mode resonance, low-frequency mechanical vibration coupling, and local envelope distortion. This makes it difficult for the reference acoustic emission signal extracted from the waveguide to maintain temporal structure consistency with the main acoustic emission sensing path. Furthermore, differential transfer function calculation requires the reference signal to remain stable. Therefore, the reference path signal in the above steps still suffers from uncleanliness and instability, leading to a significant decrease in the reliability of crack initiation identification.

[0072] The aforementioned problems arise because, as a slender elastic structure, the waveguide's inherent guided wave modes resonate at specific frequencies, introducing spikes, envelope inversions, or nonlinear jumps into the reference path. Simultaneously, the waveguide exhibits high coupling efficiency to low-frequency mechanical vibrations of the test bench, resulting in a slowly drifting envelope modulation effect on the reference signal in the time domain. Furthermore, the difference in thermal expansion and contraction at the waveguide-substrate interface causes the incident acoustic impedance to change periodically with temperature, leading to inconsistent transmission characteristics of the reference path across multiple time periods.

[0073] The combined effect of the aforementioned resonance superposition and low-frequency modulation causes the reference path signal to exhibit local peak structure shifts and energy abrupt changes that do not correspond to the main path, affecting the prerequisite for the differential transfer function that the main path and the reference path are of the same origin and comparable.

[0074] To this end, envelope extraction can be performed on the acoustic emission signals of the main acoustic emission sensing path and the reference acoustic emission sensing path. Based on the multi-segment time delay alignment window, the local peak structure consistency detection and energy ratio stability detection are performed on the envelopes of the main path and the reference path. The reference path time windows that are judged to be abnormal are removed or reconstructed to obtain a clean reference envelope for calculating the path differential transfer function.

[0075] Optionally, the above-mentioned local peak structure consistency detection includes: extracting the local peak position sequence from the main path envelope and the reference path envelope respectively within each time delay alignment window, calculating the deviation of the corresponding local peak on the time axis, and when the statistical quantity of the time deviation within the window exceeds the preset consistency threshold, marking the reference path envelope corresponding to the time delay alignment window as a local mode aliasing abnormal window and removing it.

[0076] Optionally, the above energy ratio stability detection includes: calculating the main path envelope energy and the reference path envelope energy in each time delay alignment window to obtain the energy ratio sequence of the reference path and the main path; when the change in the energy ratio of adjacent windows exceeds the preset stability threshold, the corresponding reference path envelope window is marked as an abnormal window affected by low-frequency mechanical vibration or waveguide mode resonance and is removed or reconstructed.

[0077] Furthermore, when reconstructing the reference path envelope window marked as abnormal, a combination of neighborhood linear interpolation and main path envelope proportional compensation can be used for reconstruction. That is, the reference envelope value in the abnormal window is set as a weighted sum of the main path envelope and the linear interpolation result based on the reference envelope of the adjacent normal window. The weight of the main path envelope is a pre-set fixed proportional coefficient, and the reconstructed reference envelope is used to calculate the path differential transfer function.

[0078] Specifically, when reconstructing a reference path envelope window marked as abnormal, the start and end times of the abnormal window are first obtained from the time series of the reference path envelope. The corresponding reference envelope values ​​are then extracted from the preceding and following normal windows, and the interpolated envelope values ​​at the same time within the abnormal window are calculated using a time-position-based linear interpolation method. Simultaneously, the main path envelope value at the corresponding time position is extracted from the envelope of the main acoustic emission sensing path within the abnormal window, and its amplitude is scaled according to a pre-set coefficient. The reference envelope value obtained through linear interpolation is then weighted and summed with the scaled main path envelope value to obtain the reconstructed reference envelope within the abnormal window. In the example, the weight of the main path envelope can be set to a fixed value of 0.2 to 0.4 to ensure that the reconstructed envelope retains the trend of the reference path while having a local peak structure consistent with the main path. After reconstruction, the reconstructed reference envelope is concatenated with the reference envelope of the normal window in chronological order for calculating the subsequent path differential transfer function.

[0079] In the above steps, by performing envelope extraction on the acoustic emission signals of the main path and the reference path, and performing local peak structure consistency detection and energy ratio stability detection based on multiple time delay alignment windows, the reference path time windows identified as mode aliasing or low-frequency modulation anomalies are eliminated or reconstructed, thus obtaining a clean reference envelope that is consistent with the main path in terms of envelope shape and does not contain the influence of waveguide inherent modes and test vibration.

[0080] This clean reference envelope enables the subsequently calculated path differential transfer function to have stable amplitude and phase characteristics, which can effectively improve the accuracy of identifying crack initiation characteristics (such as envelope rise steepness, group velocity shift, low-frequency wave group morphology, etc.), thereby effectively solving the problem that the reference path is destroyed by waveguide interference and affects the reliability of crack identification.

[0081] S620: Synchronously acquire the acoustic emission signals of the main acoustic emission sensing path and the reference acoustic emission sensing path, and calculate the ratio of their spectra in the frequency domain to obtain the path differential transfer function.

[0082] In this step, the main acoustic emission sensor and the reference acoustic emission sensor can be simultaneously connected to the same acoustic emission acquisition system or a clock-synchronized multi-channel acoustic emission acquisition system. A unified sampling frequency is used to digitally acquire the two acoustic emission signals, and hardware trigger signals or high-precision timestamps are used to ensure the synchronization of the two acoustic emission signals on the acquisition time axis. Preferably, the sampling frequency is set to at least twice the upper limit of the transmittable frequency band of the multilayer pyrolytic carbon coating structure, and more preferably at least five times that upper limit frequency.

[0083] Then, the synchronously acquired signals of the main acoustic emission sensing path and the reference acoustic emission sensing path are preprocessed, including at least one of DC removal, bandpass filtering, and time window truncation, to obtain signal segments for spectrum calculation. Fast Fourier Transform is then performed on the preprocessed main path signal and reference path signal to obtain the main path spectrum and reference path spectrum, respectively. The complex ratio of the main path spectrum to the reference path spectrum is calculated at each discrete frequency point. The amplitude and phase of the complex ratio characterize the amplitude-frequency and phase-frequency characteristics of the main path relative to the reference path, respectively, to form the path differential transfer function using the complex ratios at all frequency points. To avoid introducing numerical instability into the reference path spectrum at low amplitude frequencies, interpolation, smoothing, or ignoring of frequencies where the reference path spectrum amplitude is below a preset noise threshold is preferably used.

[0084] S630. The path differential transfer function is used to perform frequency domain compensation on the signal of the main acoustic emission sensing path, and crack initiation identification is performed on the compensated signal.

[0085] In this step, the inverse transfer function can be constructed based on the path differential transfer function. Complex multiplication operations are then performed on the spectrum of the main path signal at each frequency point with the corresponding frequency value of the inverse transfer function to obtain the frequency-domain compensated main path spectrum. The frequency-domain compensated main path spectrum is then converted back to the time domain using an inverse fast Fourier transform to obtain the compensated main path acoustic emission time-domain signal.

[0086] Preferably, to limit noise amplification, when constructing the inverse transfer function, amplitude truncation, regularization, or frequency band shielding are applied to frequency points where the amplitude of the path differential transfer function is lower than a preset threshold.

[0087] Finally, at least one crack initiation feature is extracted from the transmittable low-frequency band of the compensated main path acoustic emission signal, including group morphology, envelope rise steepness, group velocity shift, and event timing pattern. The crack initiation feature is compared with a pre-established crack initiation feature criterion or classification model. Based on the comparison result, it is determined whether the current acoustic emission event corresponds to the crack initiation stage of the pyrolytic carbon coating, and the crack initiation identification result is output.

[0088] In the above embodiments, a main acoustic emission sensing path and a reference acoustic emission sensing path are simultaneously set within the same pyrolytic carbon coating region, and the main path signal passes through multiple coating layers and a coupling layer. The reference path signal arrives via a propagation path less affected by the multiple coating layers (preferably via a waveguide embedded along the substrate, one end close to the crack initiation region and the other end connected to the reference sensor in the low-temperature region). The two signals are simultaneously acquired and their frequency domain spectral ratio is calculated to obtain the path differential transfer function of the main path relative to the reference path.

[0089] By using the differential transfer function to perform frequency domain compensation on the main path signal, the frequency-selective attenuation, phase distortion, and some temperature-related transmission effects introduced by the multilayer coating and coupling layer in the main path can be adaptively eliminated or effectively suppressed in the frequency domain. This makes the compensated signal closer to the real response of the crack initiation source in terms of spectral structure, envelope morphology, group velocity characteristics, and event timing patterns.

[0090] This effectively suppresses feature drift caused by high-temperature conditions and coating evolution, effectively improves the comparability and stability of crack initiation features (such as low-frequency transportable wave group morphology, envelope rise steepness, group velocity shift, and event timing patterns), and enhances the sensitivity, accuracy, and generalization ability of online crack initiation identification under different working conditions, thereby solving the problems of misjudgment and missed detection caused by changes in path and environment.

[0091] For ease of understanding, the implementation of the above technical effects is based on the underlying technical principles of differential transmission and frequency domain compensation: First, the reference acoustic emission sensing path is realized by a waveguide embedded along the substrate. One end of the waveguide is positioned close to the crack initiation area to ensure that the reference path has good sensing capability for acoustic emission events in the near-source region. The other end is positioned in the low-temperature region and connected to the reference sensor, so that the reference path is basically unaffected by the multilayer pyrolytic carbon coating and the strong non-uniformity of the high-temperature region during propagation, thus it can be regarded as a reference path with approximately weak transmission distortion.

[0092] Secondly, in signal processing, acoustic emission signals from the main path and the reference path are acquired simultaneously. Fourier transforms are performed on the two signals to obtain their spectra. The complex ratio of the spectrum of the main path to the spectrum of the reference path is calculated to obtain the path differential transfer function H_diff(f), which includes the amplitude frequency characteristics and the phase frequency characteristics.

[0093] Based on this, the inverse function H_diff⁻¹(f) of H_diff(f) is used as a compensation operator to perform dot product compensation on the spectrum of the main path signal, i.e., S_main,comp(f)=S_main(f)·H_diff⁻¹(f), and the time-domain compensation signal is obtained by inverse Fourier transform.

[0094] Since the reference path and the main path share the same acoustic emission source and near-source region structural characteristics, the difference between them is mainly reflected in the additional transmission effect introduced by the multilayer coating and coupling layer. Therefore, H_diff(f) obtained by the spectral ratio can effectively characterize the comprehensive influence of the multilayer coating transmission, remove it from the main path signal, and realize the online separation and compensation of complex transmission effects at high temperature.

[0095] Thus, without the need for absolute transfer function calibration and external excitation, the inherent characteristics of the acoustic emission signal of crack initiation can be recovered, providing a stable signal basis for subsequent reliable identification based on features such as wave group morphology, envelope rising edge steepness, group velocity shift, and event timing patterns.

[0096] To further understand the above embodiments, the above-described ground hot-testing scenario of ultra-high temperature protective components can be considered. Figure 2 and Figure 3 The illustrated embodiment will be described in detail. In the above monitoring scenario, to further improve the reliability of crack initiation event identification under extreme high temperature conditions, the following is introduced: Figure 2 The active guided wave excitation shown is Figure 3 The path differential compensation mechanism is shown.

[0097] During actual testing, the acoustic propagation characteristics of the multilayer pyrolytic carbon coating may change further over time after prolonged heating above 1800℃. The equivalent transmission gain calculated solely based on the initial temperature field may be insufficient to reflect the real-time structural state. Therefore, in this embodiment, a waveguide exciter is deployed on the coating surface to periodically emit elastic waves with a known spectral shape into the coating. The spectral shape can be a narrowband sinusoidal pulse or a linear sweep signal, and the waveguide exciter can be implemented using a piezoelectric element or an electromagnetic acoustic transducer.

[0098] During the operation of the waveguide exciter, acoustic emission sensors positioned in the low-temperature region of the component synchronously receive the excitation response signal. Since the excitation signal possesses known time-frequency characteristics, the system can calculate the transfer function of the current structure in real time based on the correspondence between excitation and response. The resulting transfer function not only includes frequency-domain attenuation information caused by the temperature field but also implicitly contains changes in the local propagation path of the pyrolytic carbon coating due to ablation, micro-damage, or thermal fatigue. The system uses this real-time transfer function to perform frequency-domain pre-compensation on passive acoustic emission events captured during the experiment, thereby recovering a signal closer to the true waveform of the crack initiation source, providing a higher signal-to-noise ratio and more stable feature representation for subsequent identification.

[0099] In addition to active excitation compensation, a dual-path acoustic emission sensing structure can be arranged to further reduce measurement errors caused by complex thermal environments. A main acoustic emission sensing path and a reference acoustic emission sensing path are set within the multi-layer pyrolytic carbon coating region. The main path signal reaches the low-temperature sensor after passing through the multi-layer high-temperature coating and coupling interface, while the reference path reaches another reference sensor located in the low-temperature region via a waveguide embedded along the substrate structure. One end of this waveguide is close to the crack initiation area, and its propagation characteristics are less affected by changes in the multi-layer coating.

[0100] During the experiment, the system synchronously acquired acoustic emission signals from the main path and the reference path, and calculated the ratio of their spectra in the frequency domain to obtain the path differential transfer function. This path differential transfer function reflects the additional frequency-dependent attenuation and phase distortion caused by the ultra-high temperature coating in the main path. Subsequently, the system used this path differential transfer function to perform frequency domain compensation on the real-time acoustic emission signal in the main path, eliminating the transmission distortion introduced by the multilayer coating, and making the compensated signal more reflective of the intrinsic time-frequency characteristics of the crack initiation event.

[0101] By combining the aforementioned active excitation compensation mechanism with the dual-path differential compensation mechanism, this embodiment achieves high-fidelity online acoustic emission monitoring on multilayer pyrolytic carbon coating structures with temperatures exceeding 1800℃. From the compensated signal, the system can stably extract crack initiation features, including the morphology of transportable low-frequency wave groups, the steepness of the envelope rise edge, group velocity shift, and event timing patterns. Based on these features, accurate identification of coating crack initiation can be achieved in extreme thermal environments.

[0102] Other embodiments of this application will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of this application that follow the general principles of this application and include common knowledge or customary techniques in the art not disclosed herein. The specification and examples are to be considered exemplary only, and the true scope and spirit of this application are indicated by the claims.

[0103] It should be understood that this application is not limited to the precise structure described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of this application is limited only by the appended claims.

Claims

1. A method for online monitoring of crack initiation in pyrolytic carbon coatings based on acoustic emission signals, characterized in that, include: Obtain the current temperature field distribution and hierarchical structure parameters of the multilayer pyrolytic carbon coating structure; Based on the temperature field distribution and hierarchical structure parameters, the equivalent transmission gain of the multilayer pyrolytic carbon coating structure in each frequency band is determined. Based on the equivalent transmission gain, the monitoring frequency band and front-end gain of the acoustic emission monitoring device are adaptively configured so that the configured monitoring frequency band is aligned with the frequency band where the equivalent transmission gain is higher than a preset threshold. On the configured monitoring frequency band, crack initiation features are extracted and crack initiation identification is performed.

2. The method according to claim 1, characterized in that, The determination of the equivalent transmission gain of the multilayer pyrolytic carbon coating structure in each frequency band based on the temperature field distribution and hierarchical structure parameters includes: For each material layer, the wave impedance and propagation constant are determined, and an equivalent transmission matrix is ​​constructed from the acoustic emission source location to the sensor location using the transmission matrix method. Based on the equivalent transmission matrix, the transmission gain is determined at each discrete frequency point to obtain the equivalent transmission gain.

3. The method according to claim 1, characterized in that, The adaptive configuration of the monitoring frequency band and front-end gain of the acoustic emission monitoring equipment includes: Frequency bands with equivalent transmission gain higher than the first threshold are configured as high-priority monitoring bands, frequency bands with equivalent transmission gain between the first threshold and the second threshold are configured as secondary-priority monitoring bands and are assigned higher amplifier gain, and frequency bands with equivalent transmission gain lower than the second threshold are configured as shielded bands.

4. The method according to claim 1, characterized in that, The crack initiation characteristics include at least one of the following in the transmittable low-frequency band: group morphology, envelope rise steepness, group velocity shift, and event timing pattern.

5. The method according to claim 1, characterized in that, Also includes: Elastic waves of known spectrum are periodically emitted onto the multilayer pyrolytic carbon coating structure via a waveguide exciter. The response to the guided wave excitation is received by an acoustic emission sensor, and the transfer function of the current structure is determined based on the excitation signal and the response signal. The transfer function is used to perform frequency domain pre-compensation on the signal of the passive acoustic emission event, and the crack initiation identification is performed on the compensated signal.

6. The method according to claim 1, characterized in that, Also includes: A main acoustic emission sensing path and a reference acoustic emission sensing path are set in the same pyrolytic carbon coating area. The sensor signal of the main acoustic emission sensing path passes through multiple coating layers and a coupling layer, while the sensor signal of the reference acoustic emission sensing path arrives via a propagation path that is less affected by the multiple coating layers. The acoustic emission signals of the main acoustic emission sensing path and the reference acoustic emission sensing path are acquired synchronously, and the spectrum ratio between the two is calculated in the frequency domain to obtain the path differential transfer function. The path differential transfer function is used to perform frequency domain compensation on the signal of the main acoustic emission sensing path, and the crack initiation identification is performed on the compensated signal.

7. The method according to claim 6, characterized in that, After setting the main acoustic emission sensing path and the reference acoustic emission sensing path in the same pyrolytic carbon coating area, the method further includes: Envelope extraction is performed on the acoustic emission signals of the main acoustic emission sensing path and the reference acoustic emission sensing path. Based on a multi-segment time delay alignment window, local peak structure consistency detection and energy ratio stability detection are performed on the envelopes of the main path and the reference path. Reference path time windows that are determined to be abnormal are removed or reconstructed.

8. The method according to claim 7, characterized in that, The local peak structure consistency detection includes: extracting the local peak position sequence from the main path envelope and the reference path envelope respectively within each time delay alignment window, calculating the deviation of the corresponding local peak on the time axis, and removing the corresponding reference path envelope when the statistical quantity of the time deviation within the window exceeds the preset consistency threshold.

9. The method according to claim 7, characterized in that, The energy ratio stability detection includes: calculating the main path envelope energy and the reference path envelope energy in each time delay alignment window to obtain the energy ratio sequence of the reference path and the main path; when the change in the energy ratio of adjacent windows exceeds a preset stability threshold, the corresponding reference path envelope window is reconstructed.

10. The method according to claim 9, characterized in that, The step of reconstructing the corresponding reference path envelope window includes: The reconstruction is performed using a combination of neighborhood linear interpolation and main path envelope ratio compensation.

Images