阅读说明：本技术 一种疲劳裂纹识别方法和系统 (Fatigue crack identification method and system ) 是由 陈汉新 李森 刘明明 李梦龙 胡振宇 于 2021-07-02 设计创作，主要内容包括：本发明涉及一种疲劳裂纹识别方法和系统,所述疲劳裂纹识别方法包括：从待测结构的兰姆波检测信号中提取目标信号,获取所述目标信号的特征参数数据；对所述特征参数数据进行序贯概率比检验,得到所述待测结构的疲劳裂纹的识别结果。借助于上述方法,利用特征参数数据反映疲劳裂纹损伤的信息特征,并结合序贯概率比检验实现对待测结构疲劳裂纹的有效识别,降低样本抽取工作量,提高疲劳裂纹识别的效率。(The invention relates to a fatigue crack identification method and a system, wherein the fatigue crack identification method comprises the following steps: extracting a target signal from a lamb wave detection signal of a structure to be detected, and acquiring characteristic parameter data of the target signal; and carrying out sequential probability ratio inspection on the characteristic parameter data to obtain the identification result of the fatigue crack of the structure to be detected. By means of the method, the information characteristics of fatigue crack damage are reflected by the characteristic parameter data, the fatigue crack of the structure to be detected is effectively identified by combining with sequential probability ratio inspection, the sample extraction workload is reduced, and the fatigue crack identification efficiency is improved.)

1. A fatigue crack identification method, comprising:

extracting a target signal from a lamb wave detection signal of a structure to be detected, and acquiring characteristic parameter data of the target signal;

and carrying out sequential probability ratio inspection on the characteristic parameter data to obtain the identification result of the fatigue crack of the structure to be detected.

2. A fatigue crack identification method according to claim 1, wherein the target signal is a second harmonic signal, and the acquisition of the characteristic parameter data comprises:

and sampling the second harmonic signal, grouping the signal value sequences obtained by sampling to obtain a plurality of groups of inspection data, calculating kurtosis values corresponding to each group of inspection data to obtain a kurtosis value sequence, and taking the kurtosis value sequence as the characteristic parameter data.

3. The fatigue crack identification method of claim 2, wherein the kurtosis value calculation process comprises:

according to the formulaCalculating to obtain the kurtosis value k, wherein w is the number of signal values of corresponding inspection data, y_vFor a signal value with sequence number v in the corresponding check data,is the mean of the sequence of signal values obtained by said sampling.

4. A fatigue crack identification method according to claim 1, wherein the characteristic parameter data is a kurtosis value sequence, and the obtaining of the identification result comprises:

and performing sequential probability ratio inspection on the kurtosis value sequence based on a preset fatigue crack state type to obtain a first fatigue crack state type of the structure to be detected, and taking the first fatigue crack state type as the identification result.

5. The fatigue crack identification method according to claim 4, wherein the preset fatigue crack state types include a no fatigue crack damage state and a fatigue crack damage state, and the acquisition process of the first fatigue crack state type includes:

taking the first fatigue crack state type as a non-fatigue crack damage state as a null hypothesis, taking the first fatigue crack state type as a fatigue crack damage state as a preparation hypothesis, sequentially selecting kurtosis values from the kurtosis value sequence, and according to a formulaCalculating to obtain a first likelihood ratio parameter delta corresponding to the currently selected kurtosis value, stopping selecting the kurtosis value until the first likelihood ratio parameter delta does not belong to a preset threshold interval, and comparing the first likelihood ratio parameter delta with a preset threshold to obtain the first fatigue crack state type, wherein P is the fatigue crack state type₁(k₀) And P₀(k₀) Respectively corresponding prior probabilities under alternative and zero hypothesis conditions, P₁(k_e) And P₀(k_e) The independent variable is k under alternative hypothesis and zero hypothesis respectively_eA function value of probability density, r is the serial number of the kurtosis value in the kurtosis value sequence, k_eThe kurtosis value sequence is a kurtosis value with a sequence number of e in the kurtosis value sequence.

6. The fatigue crack identification method of claim 5, wherein the comparing the first likelihood ratio parameter Δ to a preset threshold to obtain the first fatigue crack state type comprises:

when the first likelihood ratio parameter delta satisfies delta & ltb, the first fatigue crack state type is a non-fatigue crack damage state, and when the first likelihood ratio parameter delta satisfies delta & gt a, the first fatigue crack state type is a fatigue crack damage state, wherein a first preset threshold valueSecond preset thresholdAlpha and beta represent the probability of the occurrence of class i errors and the probability of the occurrence of class ii errors, respectively, in the sequential probability ratio test.

7. The fatigue crack identification method according to any one of claims 1 to 6, further comprising:

and transmitting an initial excitation signal to the structure to be detected by adopting an ultrasonic transmitting probe with the center frequency of 2.5Mhz to obtain a lamb wave detection signal of the structure to be detected.

8. A fatigue crack identification system is characterized by comprising a processing module and an identification module;

the processing module is used for extracting a target signal from a lamb wave detection signal of a structure to be detected and acquiring characteristic parameter data of the target signal;

and the identification module is used for carrying out sequential probability ratio inspection on the characteristic parameter data to obtain the identification result of the fatigue crack of the structure to be detected.

9. The fatigue crack identification system of claim 8, wherein the target signal is a second harmonic signal;

the processing module is specifically configured to sample the second harmonic signal, perform grouping processing on a signal value sequence obtained by the sampling to obtain multiple groups of inspection data, calculate a kurtosis value corresponding to each group of inspection data to obtain a kurtosis value sequence, and use the kurtosis value sequence as the feature parameter data.

10. The fatigue crack identification system of claim 9, wherein the processing module comprises a computing module;

the calculation module is used for calculating according to a formulaIs calculated to obtainKurtosis value k, where w is the number of signal values of the corresponding test data, y_vFor a signal value with sequence number v in the corresponding check data,is the mean of the sequence of signal values.

Technical Field

The invention relates to the technical field of nondestructive testing, in particular to a fatigue crack identification method and system.

Background

With the rapid development of society, the requirements of parts in the fields of chemical engineering, machinery and the like are gradually increased, and the problem that the regular reliability maintenance of the equipment structure becomes more and more important in production and life is brought about by applying various nondestructive testing means in consideration of the factors of operation cost and safety. Compared with many other nondestructive detection technologies, the ultrasonic detection technology has the advantages and wide application range, and the principle of the ultrasonic detection technology is that ultrasonic waves propagate in a medium to encounter defects, so that interface reflection is generated or energy attenuation changes are caused to carry out detection. Researches show that the nonlinear ultrasonic detection technology can make up for many defects of the traditional ultrasonic detection technology and effectively detect the early damage of parts.

The nonlinear ultrasonic detection technology is mainly used for detecting early damage and defects of parts, wherein Lamb waves (Lamb waves) are used for detecting micro defects of a thin plate structure due to the outstanding advantages of high propagation speed and long distance. However, Lamb waves also have the defects of frequency dispersion, multi-mode and the like during propagation, and when Lamb waves are used for detecting the micro-defects of the thin plate structure, the sample extraction workload is large and is not representative, and the fatigue cracks of the structure to be detected cannot be effectively identified.

Disclosure of Invention

The invention provides a fatigue crack identification method and system, and aims to solve the problems that when Lamb waves are used for detecting micro defects of a thin plate structure, the sample extraction workload is large and is not representative, and the fatigue crack of the structure to be detected cannot be effectively identified.

In order to solve the above technical problem, the present invention provides a fatigue crack identification method, including:

extracting a target signal from a lamb wave detection signal of a structure to be detected, and acquiring characteristic parameter data of the target signal;

and carrying out sequential probability ratio inspection on the characteristic parameter data to obtain the identification result of the fatigue crack of the structure to be detected.

The invention has the beneficial effects that: the information characteristics of fatigue crack damage are reflected by using the characteristic parameter data, the effective identification of the fatigue crack of the structure to be detected is realized by combining the sequential probability ratio inspection, the sample extraction workload is reduced, and the fatigue crack identification efficiency is improved.

Further, the target signal is a second harmonic signal, and the obtaining process of the characteristic parameter data includes:

and sampling the second harmonic signal, grouping the signal value sequences obtained by sampling to obtain a plurality of groups of inspection data, calculating kurtosis values corresponding to each group of inspection data to obtain a kurtosis value sequence, and taking the kurtosis value sequence as the characteristic parameter data.

The beneficial effect who adopts above-mentioned improvement scheme is: the second harmonic signal is adopted to reduce the noise influence attached to the system, and the kurtosis value is used as a characteristic parameter to carry out sequential probability ratio inspection, so that the accuracy of fatigue crack identification is further improved.

Further, the calculating of the kurtosis value includes:

according to the formulaCalculating to obtain the kurtosis value k, wherein w is the number of signal values of corresponding inspection data, y_vFor a signal value with sequence number v in the corresponding check data,is the mean of the sequence of signal values obtained by said sampling.

The beneficial effect who adopts above-mentioned improvement scheme is: and the kurtosis value sequence corresponding to the multiple groups of test data is convenient to calculate, and the recognition efficiency is ensured.

Further, the feature parameter data is a kurtosis value sequence, and the obtaining process of the recognition result includes:

and performing sequential probability ratio inspection on the kurtosis value sequence based on a preset fatigue crack state type to obtain a first fatigue crack state type of the structure to be detected, and taking the first fatigue crack state type as the identification result.

The beneficial effect who adopts above-mentioned improvement scheme is: and applying sequential probability ratio inspection to Lamb wave nonlinear detection by presetting fatigue crack state types, thereby realizing effective identification of different damage states of the fatigue crack of the structure to be detected.

Further, the preset fatigue crack state types include a no fatigue crack damage state and a fatigue crack damage state, and the obtaining of the first fatigue crack state type includes:

taking the first fatigue crack state type as a non-fatigue crack damage state as a null hypothesis, taking the first fatigue crack state type as a fatigue crack damage state as a preparation hypothesis, sequentially selecting kurtosis values from the kurtosis value sequence, and according to a formulaCalculating to obtain a first likelihood ratio parameter delta corresponding to the currently selected kurtosis value, stopping selecting the kurtosis value until the first likelihood ratio parameter delta does not belong to a preset threshold interval, and comparing the first likelihood ratio parameter delta with a preset threshold to obtain the first fatigue crack state type, wherein P is the fatigue crack state type₁(k₀) And P₀(k₀) Respectively corresponding prior probabilities under alternative and zero hypothesis conditions, P₁(k_e) And P₀(k_e) The independent variable is k under alternative hypothesis and zero hypothesis respectively_eA function value of probability density, r is the serial number of the kurtosis value in the kurtosis value sequence, k_eThe kurtosis value sequence is a kurtosis value with a sequence number of e in the kurtosis value sequence.

The beneficial effect who adopts above-mentioned improvement scheme is: the sequential probability ratio inspection is used for inspecting and identifying the secondary harmonic signals without fatigue damage and with fatigue damage, whether the structure to be detected has early damage and defects can be identified quickly and accurately, and the production efficiency is improved.

Further, the comparing the first likelihood ratio parameter Δ with a preset threshold value to obtain the first fatigue crack state type includes:

when the first likelihood ratio parameter delta satisfies delta & ltb, the first fatigue crack state type is a non-fatigue crack damage state, and when the first likelihood ratio parameter delta satisfies delta & gt a, the first fatigue crack state type is a fatigue crack damage state, wherein a first preset threshold valueSecond preset thresholdAlpha and beta represent the probability of the occurrence of class i errors and the probability of the occurrence of class ii errors, respectively, in the sequential probability ratio test.

The beneficial effect who adopts above-mentioned improvement scheme is: and comparing the first likelihood ratio parameter with the first preset threshold and the second preset threshold so as to identify two fatigue crack damage states, and facilitating calculation and application.

Further, still include:

and transmitting an initial excitation signal to the structure to be detected by adopting an ultrasonic transmitting probe with the center frequency of 2.5Mhz to obtain a lamb wave detection signal of the structure to be detected.

The beneficial effect who adopts above-mentioned improvement scheme is: the frequency dispersion and the multi-mode during lamb wave propagation are avoided, and the accuracy and the efficiency of fatigue crack identification are ensured.

In a second aspect, the invention provides a fatigue crack identification system, comprising a processing module and an identification module;

the processing module is used for extracting a target signal from a lamb wave detection signal of a structure to be detected and acquiring characteristic parameter data of the target signal;

and the identification module is used for carrying out sequential probability ratio inspection on the characteristic parameter data to obtain the identification result of the fatigue crack of the structure to be detected.

Further, the target signal is a second harmonic signal;

the processing module is specifically configured to sample the second harmonic signal, perform grouping processing on a signal value sequence obtained by the sampling to obtain multiple groups of inspection data, calculate a kurtosis value corresponding to each group of inspection data to obtain a kurtosis value sequence, and use the kurtosis value sequence as the feature parameter data.

Further, the processing module comprises a computing module;

the computing module is used for computing according toFormula (II)Calculating to obtain the kurtosis value k, wherein w is the number of signal values of corresponding inspection data, y_vThe signal values with the sequence number v in the corresponding test data are shown, and x is the mean value of the signal value sequence.

Drawings

Fig. 1 is a schematic flow chart of a fatigue crack identification method according to an embodiment of the present invention;

fig. 2 is a schematic diagram of Lamb wave generation according to an embodiment of the present invention;

FIG. 3 is a schematic diagram of a phase velocity dispersion curve according to an embodiment of the present invention;

FIG. 4 is a schematic diagram of a group velocity dispersion curve according to an embodiment of the present invention;

FIG. 5 is a signal flow diagram of a nonlinear ultrasonic inspection system provided by an embodiment of the present invention;

FIG. 6 is a schematic diagram of a second harmonic signal provided by an embodiment of the present invention;

FIG. 7 is a diagram illustrating the results of the tests on the input signals S1 and S2 according to an embodiment of the present invention;

FIG. 8 is a diagram illustrating the results of the tests on the input signals S2 and S3 according to an embodiment of the present invention;

fig. 9 is a schematic structural diagram of a fatigue crack identification system according to an embodiment of the present invention.

Detailed Description

The following examples are further illustrative and supplementary to the present invention and do not limit the present invention in any way.

A fatigue crack identification method according to an embodiment of the present invention is described below with reference to the drawings.

Referring to fig. 1, the present invention provides a fatigue crack identification method, including:

s1, extracting a target signal from the lamb wave detection signal of the structure to be detected, and acquiring characteristic parameter data of the target signal;

and S2, carrying out sequential probability ratio inspection on the characteristic parameter data to obtain the identification result of the fatigue crack of the structure to be detected.

According to the fatigue crack identification method provided by the embodiment, the characteristic parameter data is used for reflecting the information characteristics of the fatigue crack damage, the effective identification of the fatigue crack of the structure to be detected is realized by combining the sequential probability ratio inspection, the sample extraction workload is reduced, and the fatigue crack identification efficiency is improved.

It is understood that the structure to be measured may be a plate-shaped or plate-shaped metal structure, the target signal is a characteristic signal representing a nonlinear effect, and the characteristic parameter data includes a characteristic parameter capable of reflecting characteristic information of the target signal, that is, the characteristic parameter is capable of reflecting an information characteristic of a micro-damage, for example, a characteristic parameter such as a wave arrival time and an amplitude or a kurtosis value may be adopted.

In the embodiment, a nonlinear ultrasonic detection system can be set up to obtain a lamb wave detection signal of a structure to be detected, and a target signal is extracted from the lamb wave detection signal.

Optionally, in an embodiment, the target signal is a second harmonic signal, and the acquiring process of the characteristic parameter data includes:

and sampling the second harmonic signal, grouping the signal value sequences obtained by sampling to obtain a plurality of groups of inspection data, calculating kurtosis values corresponding to each group of inspection data to obtain a kurtosis value sequence, and taking the kurtosis value sequence as the characteristic parameter data.

It is understood that, when the sequential probability ratio test is performed, the characteristic parameters representing the signal characteristics are extracted from the collected data, and the second harmonic signal is preferred as the target signal in consideration of the fact that the higher harmonic signal is weak and is easily covered by noise attached to the system.

Specifically, in this embodiment, noise reduction filtering processing may be performed on the second harmonic signal, then sampling may be performed to obtain a signal value sequence, grouping may be performed on the signal value sequence to obtain a plurality of groups of inspection data, and then feature parameters corresponding to the plurality of groups of inspection data may be obtained to perform sequential probability ratio inspection.

Preferably, the calculating of the kurtosis value includes:

according to the formulaCalculating to obtain the kurtosis value k, wherein w is the number of signal values of corresponding inspection data, y_vAnd x is the mean value of the signal value sequence obtained by sampling, wherein the signal value is the signal value with the serial number v in the corresponding test data.

It will be appreciated that the grouping of the inspection data may be flexibly configured according to the actual situation, for example, in one embodiment, [ x ] may be assumed₁,x₂,x₃,...,x_N]For a sequence of sampled signal values, where N is set to 2000, the above formula for calculating the kurtosis value can be converted to1901 signal values are sequentially selected from the signal value sequence as a group of test data each time, 100 groups of test data can be obtained, and then a kurtosis value sequence [ k ] corresponding to 100 groups of test data is obtained through calculation₁,k₂,k₃,...,k_n]Wherein k is_tIs a sequence of kurtosis values [ k₁,k₂,k₃,...,k_n]The kurtosis value with the sequence number of t is k_tCorresponding to the t-th test data, n is 100, t is 1,2_iSignal values in the t-th group of test data obtained for the grouping, which are associated with a sequence of signal values [ x ]₁,x₂,x₃,...,x_N]The mean value of the sequence of signal values corresponding to the signal value with the middle sequence number iIs composed ofx_gIs a sequence of signal values [ x ]₁,x₂,x₃,...,x_N]The middle sequence number is the signal value of g.

Optionally, in an embodiment, the feature parameter data is a kurtosis value sequence, and the obtaining of the recognition result includes:

and performing sequential probability ratio inspection on the kurtosis value sequence based on a preset fatigue crack state type to obtain a first fatigue crack state type of the structure to be detected, and taking the first fatigue crack state type as the identification result.

It will be appreciated that the material may have inclusions, segregation or defects, or may be of an unreasonable design, or due to the unreasonable processing and manufacturing processes, stress concentration is often generated at certain parts of the structure, cracks are initiated under repeated stress alternation, for example, under the action of cyclic stress and strain, local permanent accumulated damage is gradually generated at one or more positions, cracks are generated or even complete fracture is generated after a certain number of cycles, and fatigue can be divided into high cycle fatigue, low cycle fatigue, subcritical fatigue and the like according to the number of cycles (namely service life) which the material undergoes before being damaged and the stress level of fatigue load, so that the fatigue crack state type can be preset according to the actual requirement, and (4) providing an assumption for the fatigue crack of the structure to be detected, and further obtaining a fatigue crack identification result of the structure to be detected by utilizing sequential probability ratio inspection.

For example, for a metal sheet to be tested, an assumption can be made as to whether the metal sheet has a fatigue crack or the fatigue degree of the fatigue crack, and a preset fatigue crack state type is established, so that whether the metal sheet has a crack or the fatigue degree of the fatigue crack can be identified, and further early damage and defects of the material can be found in time, which is beneficial to subsequent production and manufacturing.

Preferably, the preset fatigue crack state types include a no fatigue crack damage state and a fatigue crack damage state, and the acquiring of the first fatigue crack state type includes:

taking the first fatigue crack state type as a non-fatigue crack damage state as a zero hypothesis, taking the first fatigue crack state type as a fatigue crack damage state as a preparation hypothesis, and taking the kurtosis value as a reference hypothesisSequentially selecting kurtosis values in the sequence and according to a formulaCalculating to obtain a first likelihood ratio parameter delta corresponding to the currently selected kurtosis value, stopping selecting the kurtosis value until the first likelihood ratio parameter delta does not belong to a preset threshold interval, and comparing the first likelihood ratio parameter delta with a preset threshold to obtain the first fatigue crack state type, wherein P is the fatigue crack state type₁(k₀) And P₀(k₀) Respectively corresponding prior probabilities under alternative and zero hypothesis conditions, P₁(k_e) And P₀(k_e) The independent variable is k under alternative hypothesis and zero hypothesis respectively_eA function value of probability density, r is the serial number of the kurtosis value in the kurtosis value sequence, k_eThe kurtosis value sequence is a kurtosis value with a sequence number of e in the kurtosis value sequence.

Preferably, the comparing the first likelihood ratio parameter Δ with a preset threshold value to obtain the first fatigue crack state type includes:

when the first likelihood ratio parameter delta satisfies delta & ltb, the first fatigue crack state type is a non-fatigue crack damage state, and when the first likelihood ratio parameter delta satisfies delta & gt a, the first fatigue crack state type is a fatigue crack damage state, wherein a first preset threshold valueSecond preset thresholdAlpha and beta represent the probability of the occurrence of class i errors and the probability of the occurrence of class ii errors, respectively, in the sequential probability ratio test.

It should be noted that the basic principle of the sequential probability ratio test includes that a group of h independent and equally distributed random variable sequences l can be obtained by observing the random variable l₁,l₂,l₃,......l_hTwo assumptions are made for the total sample: null hypothesis H₀:θ＝θ₀And alternative hypothesis H₁:θ＝θ₁And forming a binary sequential probability ratio test.

The joint probability density of the two can be defined as:

likelihood ratio lambda of sequential probability ratio test_nThe calculation formula of (2) is as follows:

where j is 0,1, f (l/θ) is a conditional probability distribution, and θ is a distribution parameter.

Determining corresponding threshold values according to the probability alpha of the type I error and the probability beta of the type II error in the sequential probability ratio testAnd(where A > B), for example, assume l₁For the data value of the first fatigue damage pattern recognition, the likelihood ratio lambda can be obtained by formula₁(l₁) And compared to a set threshold A, B to identify two fatigue damage patterns. If likelihood ratio lambda₁(l₁) Satisfy lambda₁(l₁) < B, at which point the test is stopped and a null hypothesis H is accepted₀Rejecting alternative hypothesis H₁(ii) a If the likelihood ratio satisfies λ₁(l₁) > A, stop checking, accept alternative hypothesis H₁Reject null hypothesis H₀(ii) a If the likelihood ratio is B ≦ λ₁(l₁) And (4) when the condition occurs, extracting the next group of values to continue the inspection until the requirement of stopping the inspection is met, and finally giving an identification result based on two fatigue damage modes.

From the above formula for calculating the likelihood ratio of the sequential probability ratio test, it can be known that the likelihood ratio is affected by the change of the mean and the standard deviation, and the obtained kurtosis value sequence basically satisfies the normal distribution, so that the kurtosis value can be used as the characteristic parameter of the sequential probability ratio test.

For example, in this embodiment, when it is assumed that the aluminum alloy test piece has no fatigue damage (corresponding to a fatigue crack damage free state), the collected second harmonic signal sequence satisfies the null hypothesis: h₀:μ＝μ₀In the presence of fatigue cracks (corresponding to the fatigue crack damage state), the sequence satisfies the alternative assumption H₁:μ＝μ₁。

The preset fatigue crack state types established at the moment comprise a non-fatigue crack damage state and a fatigue crack damage state, and when the zero hypothesis and the alternative hypothesis are both established, the joint probability density function of the kurtosis value sequence is defined asIn this process, the standard deviation σ is kept constant and only the mean value μ changes, where P is the case when j is 0₀(k_d) Denotes the probability density function under the condition of zero hypothesis, when j is 1, P₁(k_d) Representing the probability density function, k, under alternative assumptions_dRepresents kurtosis values in the kurtosis value sequence, sigma represents standard deviation of the kurtosis value sequence, and mu is when j is 0₀And j is 1 μ₁Respectively representing the mean values of the corresponding sequences satisfying the null hypothesis condition and satisfying the alternative hypothesis condition;

for simplifying calculation and convenient application, the calculation formula of likelihood ratio of sequential probability ratio test can be converted intoWherein λ is_eLikelihood ratio lambda calculated for the e-th test_e，P₁(k₀) And P₀(k₀) Respectively corresponding prior probabilities under a preparation hypothesis condition and a zero hypothesis condition, wherein r is a sequence number value of the currently selected kurtosis value in the kurtosis value sequence, k_eThe kurtosis value sequence is a kurtosis value with a sequence number of e in the kurtosis value sequence.

At this time, set firstPreset threshold valueAnd a second preset thresholdAccording to the sequential probability ratio test theory, when the likelihood ratio delta satisfies delta < b, stopping the test and receiving H₀I.e., the first fatigue crack state type is a no fatigue crack damage state; when Δ > a, stop the test and accept H₁I.e., the first fatigue crack state type is a fatigue crack damage state; if the preset threshold interval b is less than or equal to delta and less than or equal to a, judgment cannot be carried out at the moment, the detection needs to be carried out continuously, the next likelihood ratio is compared with a and b, sampling is stopped until the condition is met, and then judgment is carried out.

It should be noted that the preset threshold interval and the preset threshold may be set based on a sequential probability ratio test theory, or may be reasonably adjusted according to actual requirements.

Further, in one embodiment, the method further comprises:

and transmitting an initial excitation signal to the structure to be detected by adopting an ultrasonic transmitting probe with the center frequency of 2.5Mhz to obtain a lamb wave detection signal of the structure to be detected.

It should be noted that Lamb waves are stress waves as shown in fig. 2. When sound waves propagate in a thin plate material, if the plate thickness is equivalent to the wavelength of Lamb waves and the sound waves are subjected to surface tension which changes alternately, mass points generate vibration in two directions of transverse direction and longitudinal direction, the two vibrations are synthesized into an elliptical track, the formation process of the Lamb waves is the elliptical track, the propagation mode of the Lamb waves is divided into symmetrical and anti-symmetrical modes, the symmetrical mode and the anti-symmetrical mode are further numerically researched through a Rayleigh-Lamb equation, and the following formula can be obtained:

symmetrical mode:

antisymmetric mode:

in the formulae (1) and (2), p is represented byTo obtain q fromAnd (4) calculating. Wave numberAngular frequency ofc_pIs the phase velocity of Lamb wave, λ is the wavelength, c_lAnd c_tRespectively representing the longitudinal wave velocity and the transverse wave velocity, b₁A sheet thickness of 1/2. From the phase velocities c in equations (1) and (2)_pAnd angular frequency ω yields a dispersion curve.

For example, the longitudinal wave velocity selection c_l6441m/s, the transverse wave velocity is taken by c_t3224 m/s. MATLAB is utilized to solve Rayleigh-Lamb equation to obtain Lamb wave phase velocity and group velocity dispersion curves in the aluminum alloy plate, which are respectively shown in FIG. 3 and FIG. 4.

As can be seen from FIGS. 3 and 4, when the product f.d of the frequency f and the plate thickness d is greater than or equal to 3 MHz.mm, at least three wave modes exist, which is not favorable for identifying fatigue cracks. And when f.d is less than or equal to 2.5 MHz.mm, only A0 and S0 wave modes are available, so as to avoid the problem of multiple modes, and considering that 2.5MHz is a critical value of a single mode and the frequency of the ultrasonic transmitting probe should be large to avoid signal dispersion attenuation, 2.5MHz is selected as the central frequency of the ultrasonic transmitting probe.

Specifically, in this embodiment, as shown in fig. 5, the nonlinear ultrasonic detection system may be composed of a RAM-5000-SNAP system, a computer, an impedance, a step attenuator, a filter, and an amplifier, where a signal flow direction is that a signal generated by the RAM-5000-SNAP system is sent from a transmitting end, passes through a 50 Ω impedance, the step attenuator, and a low-pass filter, is transmitted to a structure to be detected by a transmitting probe, and is then transmitted to a receiving probe to obtain the lamb wave detection signal, where the lamb wave detection signal includes a second harmonic and a higher harmonic that are generated when the signal transmitted by the transmitting probe propagates to a microcrack of the structure to be detected, and in order to collect a second harmonic signal carrying defect information, a single-transmit and double-receive mode is adopted, one of the signals received by the receiving probe is directly transmitted to the receiving end 1 to collect a fundamental wave signal, and the other one of the signals is sequentially subjected to high-pass filtering and signal amplification processing by the filter and the amplifier, and then is transmitted to the receiving end The receiving end 2 is used for collecting a second harmonic signal representing the nonlinear effect, wherein the second harmonic signal can be collected through an oscilloscope based on the frequency relative relationship of the fundamental wave signal and the second harmonic signal. In addition, for better effect of the received second harmonic, the excitation signal sent by the transmitting end adopts a Hanning window modulation signal, and meanwhile, 15 Cycles are selected by the sine pulse train to avoid generating overlapped sound waves.

Further, in this embodiment, a size specification of 300 × 100 × 3 (mm) is used³) The 2A12 aluminum alloy plate is a test piece, a triangular notch is cut at the top end of the same thin plate marked as 1,2 and 3 respectively, so that fatigue is easy to generate, then 0-time, 4000-time and 8000-time fatigue stretching is carried out on three plates, because the second harmonic signal is a main characteristic signal representing nonlinear effect, the second harmonic signal is extracted and correspondingly analyzed, for three groups of extracted second harmonic signals, the signal S1 with 0-time fatigue stretching as a normal state is recorded, and the signal S2 and the signal S3 with 4000-time and 8000-time fatigue stretching as two fatigue damage states are recorded respectively, as shown in figure 6, S1 is the second harmonic signal collected by a test piece without fatigue stretching, and S2 and S3 are the second harmonic signals collected after 4000-time and 8000-time fatigue stretching respectively.

Let the mean value of the corresponding characteristic parameter data of S1 in the fatigue damage free state be μ₀The mean value of the corresponding characteristic parameter data of S2 in the fatigue damage state is μ₁And inspecting the signal to be inspected, as shown in FIG. 7, wherein the dotted line represents the inspection resultThe input signal S1 is used as the test result of the signal to be tested, and the solid line represents the input signal S2 as the test result of the signal to be tested.

In another embodiment, the mean value of the corresponding characteristic parameter data of the signal S2 is used as the distribution parameter μ₀The mean value of the corresponding characteristic parameter data of the signal S3 is taken as the distribution parameter mu₁Inputting S2 and S3 into the likelihood ratio formula for verification can obtain the verification result shown in fig. 8, wherein the dotted line represents the input signal S2 as the verification result of the signal to be tested, and the solid line represents the input signal S3 as the verification result of the signal to be tested.

As can be seen from the test results shown in fig. 7-8, after the input signals to be tested in different fatigue damage states are sequentially tested for a certain number of times, the likelihood ratio values of the different signals corresponding to the number of sequential tests will be significantly different, so that the likelihood ratios corresponding to the signals to be tested can be compared by setting a threshold, and the second harmonic signals in different fatigue damage states can be effectively identified.

The fatigue crack identification method provided in the embodiment is simple, easy to understand and implement, has a good application prospect, and can effectively identify whether the structure to be detected has cracks or the fatigue degree of the cracks on the structure to be detected, so that the reliability maintenance of the subsequent equipment structure is facilitated, the production and manufacturing cost is reduced, the material utilization rate is improved, the production efficiency is improved, and the product can meet the requirements of use performance and economic benefit.

In the above embodiments, although the steps are numbered as S1, S2, etc., but only the specific embodiments are given in this application, a person skilled in the art may adjust the execution sequence of S1, S2, etc. according to the actual situation, and this is within the scope of the present invention, and it is understood that some embodiments may include some or all of the above embodiments.

As shown in fig. 9, an embodiment of the invention provides a fatigue crack identification system 10, which includes a processing module 20 and an identification module 30;

the processing module 20 is configured to extract a target signal from a lamb wave detection signal of a structure to be detected, and acquire characteristic parameter data of the target signal;

the identification module 30 is configured to perform sequential probability ratio inspection on the characteristic parameter data to obtain an identification result of the fatigue crack of the structure to be detected.

Optionally, in an embodiment, the target signal is a second harmonic signal;

the processing module 20 is specifically configured to sample the second harmonic signal, perform grouping processing on a sequence of signal values obtained by the sampling to obtain multiple groups of inspection data, calculate a kurtosis value corresponding to each group of inspection data to obtain a sequence of kurtosis values, and use the sequence of kurtosis values as the feature parameter data.

Preferably, the processing module 20 comprises a computing module;

the calculation module is used for calculating according to a formulaCalculating to obtain the kurtosis value k, wherein w is the number of signal values of corresponding inspection data, y_vFor a signal value with sequence number v in the corresponding check data,is the mean of the sequence of signal values.

Optionally, in an embodiment, the characteristic parameter data is a sequence of kurtosis values;

the identification module 30 is specifically configured to perform sequential probability ratio inspection on the kurtosis value sequence based on a preset fatigue crack state type to obtain a first fatigue crack state type of the structure to be detected, and use the first fatigue crack state type as the identification result.

Preferably, the preset fatigue crack state types include a no fatigue crack damage state and a fatigue crack damage state, and the identification module 30 includes an inspection module;

the inspection module is used for taking the first fatigue crack state type as a non-fatigue crack damage state as a zero hypothesis and taking the first fatigue crack state type as a zero hypothesisSelecting kurtosis values from the kurtosis value sequence in sequence according to a formulaCalculating to obtain a first likelihood ratio parameter delta corresponding to the currently selected kurtosis value, stopping selecting the kurtosis value until the first likelihood ratio parameter delta does not belong to a preset threshold interval, and comparing the first likelihood ratio parameter delta with a preset threshold to obtain the first fatigue crack state type, wherein P is the fatigue crack state type₁(k₀) And P₀(k₀) Respectively corresponding prior probabilities under alternative and zero hypothesis conditions, P₁(k_e) And P₀(k_e) The independent variable is k under alternative hypothesis and zero hypothesis respectively_eA function value of probability density, r is the serial number of the kurtosis value in the kurtosis value sequence, k_eThe kurtosis value sequence is a kurtosis value with a sequence number of e in the kurtosis value sequence.

Preferably, the checking module is specifically configured to determine that the first fatigue crack state type is a no fatigue crack damage state when the first likelihood ratio parameter Δ satisfies Δ < b, and determine that the first fatigue crack state type is a fatigue crack damage state when the first likelihood ratio parameter Δ satisfies Δ > a, where a first preset threshold is setSecond preset thresholdAlpha and beta represent the probability of the occurrence of class i errors and the probability of the occurrence of class ii errors, respectively, in the sequential probability ratio test.

Optionally, in an embodiment, the apparatus further includes a signal acquisition module;

the signal acquisition module is used for transmitting an initial excitation signal to the structure to be detected by adopting an ultrasonic transmitting probe with the center frequency of 2.5Mhz to obtain a lamb wave detection signal of the structure to be detected.

As will be appreciated by one skilled in the art, the present invention may be embodied as a system, method or computer program product. Accordingly, the present disclosure may be embodied in the form of: may be embodied entirely in hardware, entirely in software (including firmware, resident software, micro-code, etc.) or in a combination of hardware and software, and may be referred to herein generally as a "circuit," module "or" system. Furthermore, in some embodiments, the invention may also be embodied in the form of a computer program product in one or more computer-readable media having computer-readable program code embodied in the medium.

In the description herein, references to the description of the term "one embodiment," "some embodiments," "an example," "a specific example," or "some examples," etc., mean that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the invention. In this specification, the schematic representations of the terms used above are not necessarily intended to refer to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples. Furthermore, various embodiments or examples and features of different embodiments or examples described in this specification can be combined and combined by one skilled in the art without contradiction.

Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention, and that variations, modifications, substitutions and alterations can be made to the above embodiments by those of ordinary skill in the art within the scope of the present invention.