CN116908304B - Grain size evaluation method of polycrystalline materials based on ultrasonic wake average power attenuation - Google Patents

Abstract

The invention provides a polycrystalline material grain size assessment method based on ultrasonic wake average power attenuation, which comprises the following steps: preparing a plurality of samples with the same thickness as the polycrystalline material to be detected, determining the mode and parameters of an excitation signal, and constructing a polycrystalline material average grain size detection system; respectively exciting and receiving ultrasonic signals of samples with different average grain sizes, and carrying out band-pass filtering pretreatment; selecting signals of a wake wave time window and a reference time window, respectively extracting average power, and calculating wake wave average power attenuation according to a time difference; establishing a logarithmic linear model between the average grain size of the sample and the average power attenuation of the wake wave based on the equivalent attenuation coefficient model; and collecting signals of the polycrystalline material with unknown grain size to be detected, filtering, extracting corresponding wake wave average power attenuation, and determining the average grain size of the polycrystalline material according to the logarithmic linear model. The invention has the advantages of high detection efficiency, relatively easy operation, strong practicability and accurate detection result.

Description

Grain size evaluation method of polycrystalline materials based on ultrasonic wake average power attenuation

Technical field

The invention belongs to the technical field of non-destructive testing of polycrystalline material grain size, and in particular relates to a polycrystalline material grain size evaluation method based on the average power attenuation of ultrasonic wake waves.

Background technique

As a representative key material for manufacturing core parts such as turbine disks of modern advanced aerospace engines, nickel-based high-temperature alloys are widely used in some advanced aerospace engines. Due to processing technology limitations, the current large-sized high-temperature alloy blanks used for turbine disks are prone to uneven grain structure during processing. Some grains are coarse, resulting in reduced tensile strength and fatigue resistance, and dispersion of mechanical properties. If serious, the product will fail and be scrapped, seriously affecting the integrity and reliability of the equipment. Therefore, developing an efficient and accurate grain size evaluation method for alloy blanks will help achieve stable quality of high-temperature alloy turbine disks under mass production conditions, and has important engineering practical value.

Currently, commonly used grain size testing methods are mainly divided into two categories: destructive testing methods and non-destructive testing methods. Destructive testing techniques, such as optical metallography and electron backscattered diffraction, cut small test pieces from the sample to be inspected and then photograph the grain structure characteristics of the surface of the test piece through an optical or electron microscope to evaluate the average grain size. Although the destructive testing method can characterize the grain microstructure with high accuracy, cutting samples is often not allowed for the actual structure to be tested in engineering, and the testing cost is high, which limits its application scope. Therefore, non-destructive testing technology has been highly developed, such as X-ray testing methods, body wave ultrasonic testing methods, etc. The X-ray detection method uses the attenuation or scattering characteristics of the interaction between X-rays and materials when propagating in the object to be tested to observe the internal grain microstructure of the object. However, the equipment is expensive and requires long-term measurement, which is inefficient. The detection method based on body wave ultrasound generates ultrasonic waves in the material and analyzes the characteristics of the waves after interacting with the internal microstructure of the material to evaluate the material properties. The selected body wave ultrasonic waveform characteristics mainly include wave velocity characteristics, backscattering characteristics and attenuation characteristics. The principle of wave speed characteristics is to calculate the sound speed based on the thickness of the material being measured and the time delay between two echoes, and then evaluate the average grain size. However, the relationship between wave speed and grain size is not monotonic under a certain grain size range. It is also less sensitive to some metal materials and requires an accurate measurement system for detection. The backscattering feature focuses on the large number of scattered waves generated by the interaction between ultrasonic waves and the grain interface of the internal acoustic impedance mismatch in the material. In the time domain signal, it appears as a noise-like disturbance signal between the main wave and the echo signal, but the backscattered signal It needs to be extracted through multiple spatial sampling averages, which is inefficient. The attenuation characteristic means that during the propagation of ultrasonic waves inside the material, the energy carried will weaken as the propagation distance increases. Based on this, the sequential decrease in the amplitude of the surface echo and each bottom wave in the pulse echo method can be observed to evaluate the grains. Size and stability are good, but due to the limitations of body wave ultrasound, it can only detect the average grain size of a single path. Large structures require spatially dense sampling, which is less efficient and may lead to the missed detection of local grain abnormalities. .

Coda waves are formed by multiple scatterings of ultrasonic waves in inhomogeneous media, and appear as the tail behind the direct waves. Because wake waves play a role in repeatedly sampling space on a longer time scale than body waves, they are more sensitive to microscopic grain structure characteristics and can achieve a wide range of measurements with a single measurement at a single sensor location. Grain size assessment. Therefore, it is very urgent and extremely important to combine the advantages of ultrasonic attenuation characteristics and wake detection to find a polycrystalline material grain size assessment method based on the average power attenuation of ultrasonic wake waves to efficiently achieve average grain size assessment.

Contents of the invention

In view of the above-mentioned defects in the prior art, the present invention proposes a polycrystalline material grain size evaluation method based on the average power attenuation of ultrasonic wake waves. It first prepares several samples with the same thickness as the polycrystalline material to be detected, determines the method and parameters of the excitation signal, and then builds a polycrystalline material average grain size detection system; respectively, the samples with different average grain sizes are Excite and receive ultrasonic signals, and perform band-pass filtering preprocessing; select the signals in the coda time window and the reference time window, extract the average power respectively, and calculate the average coda power attenuation based on the time difference; based on the equivalent attenuation coefficient model, establish a test The logarithmic linear model between the average grain size of the sample and the average power attenuation of the coda wave is used; the signal of the polycrystalline material with unknown grain size to be detected is collected, the corresponding average power attenuation of the coda wave is extracted after filtering, and the polycrystalline material is determined based on the logarithmic linear model. The average grain size of a crystalline material. Compared with the traditional body wave method, the present invention has high detection efficiency, is relatively easy, has strong practicability, and has more accurate detection results.

Specifically, the present invention provides a polycrystalline material grain size evaluation method based on ultrasonic wake average power attenuation, which includes the following steps:

S1. Prepare several samples with the same thickness as the polycrystalline material to be detected, determine the method and parameters of the excitation signal, and build a polycrystalline material average grain size detection system;

S2. Excite and receive ultrasonic signals for samples with different average grain sizes respectively, and perform band-pass filtering preprocessing on the collected ultrasonic signals;

S3. Select the ultrasonic signals of the coda time window [t, t + T] and the reference time window [t', t' + T ₀ ], extract the average power respectively, and calculate the average coda power attenuation based on the time difference between the two. The steps include the following sub-steps:

S31. According to the time window starting point t and time window width T, extract the signal voltage amplitude V(τ) at different moments τ in the selected coda time window [t, t+T], and calculate the average coda wave at time t Power P(t):

The average coda power P(t) represents the energy state of the coda part of the ultrasonic signal at time t;

S32. According to the time window starting point t' and the time window width T ₀ , extract the signal voltage amplitude V(τ) at different moments τ within the selected coda time window []t', t'+T ₀ ]. Within the selected reference time window [t',t'+T ₀ ], calculate the reference average power at time t ₀ =t'+T ₀ /2 as P(t ₀ ):

The reference average power P(t ₀ ) represents the energy state of the main wave part of the ultrasonic signal at time t ₀ ;

S33. Calculate the average coda power attenuation based on the average coda power P(t) and the reference average power P(t ₀ ) of the ultrasonic signal at time t and t ₀ and the time difference Δt=tt ₀ for:

Coda average power attenuation Represents the coda power dissipation at the interval Δt due to grain scattering and other factors;

S4. Based on the equivalent attenuation coefficient model, establish the average grain size d of the sample and the average power attenuation of the coda wave. Log-linear model between:

Among them, θ ₁ and θ ₂ respectively represent the first fitting parameter of the model and the second fitting parameter of the model, which are evaluated by the least squares method;

S5. Use the polycrystalline material average grain size detection system to collect the ultrasonic signal of the polycrystalline material with unknown grain size to be detected, filter it, and extract the coda average power attenuation of the polycrystalline material according to step S3. And determine the average grain size of the polycrystalline material based on the logarithmic linear model described in step S4/> The calculation method is as follows:

Preferably, step S1 specifically includes the following sub-steps:

S11. According to the material composition information of the polycrystalline material to be detected, prepare several samples with the same chemical composition and different average grain sizes, and cut them into plate-shaped samples with the same thickness as the polycrystalline material to be detected;

S12. The excitation source used to excite the ultrasonic signal is a Hanning window modulated sine wave pulse with a single center frequency. The frequency of the ultrasonic signal is selected based on the elastic properties and thickness information of the polycrystalline material to be detected;

S13. Use an arbitrary function generator to output ultrasonic waves in a single channel, and form two output terminals through a tee joint. The first output terminal is connected to a high-voltage power amplifier, and is amplified and then transmitted to the excitation ultrasonic transducer chip. The second output terminal is directly Trigger channel passed to mixed domain oscilloscope;

S14. For each sample, place a pair of excitation transducer chips and receiving transducer chips with the same center frequency closely together and fix them at the center of the sample surface;

S15. After each excitation, according to the trigger signal received by the trigger channel, the ultrasonic signal propagating in the sample is collected through the receiving transducer chip and transmitted to the acquisition channel of the mixed domain oscilloscope.

Preferably, the step S2 specifically includes the following sub-steps:

S21. Randomly cut and sample each sample and obtain the actual grain size;

S22. Use the pulse echo method to detect each sample. The excitation transducer chip generates an excitation ultrasonic signal in the sample, and the receiving transducer chip receives the ultrasonic signal transmitted in the sample;

S23. Repeat steps S21 and S22 until ultrasonic signals corresponding to samples with different average grain sizes are obtained, and band-pass filtering preprocessing is performed on the ultrasonic signals.

Preferably, the signal received by the receiving transducer chip in step S2 is the average of 128 consecutive signal acquisitions by the mixed domain oscilloscope.

Preferably, in step S2, pulse echo is used to detect the ultrasonic signals in each sample.

Preferably, the coda wave time window and the reference time window extracted in step S3 are respectively located in the coda wave part and the main wave part of the ultrasonic signal.

Preferably, the selection of the coda part takes the arrival time of twice the shortest boundary echo as the starting time, and the coda signal within the selected coda time window has a high signal-to-noise ratio.

Preferably, in step S21, the average grain size d of the sample is quantified by means of electron backscatter diffraction technology.

Preferably, in step S4, the first fitting parameter of the model and the second fitting parameter of the model are evaluated by the least squares method.

Preferably, the polycrystalline material average grain size detection system in step S1 includes an arbitrary function generator, a tee joint, a high-voltage power amplifier, an excitation transducer wafer, a receiving transducer wafer, a mixed domain oscilloscope, and a host controller. ;

The arbitrary function generator outputs ultrasonic waves in a single channel and transmits them to the three-way joint. The three-way joint has two output ends. The first output end is connected to a high-voltage power amplifier and is amplified and then transmitted to the excitation ultrasonic transducer. chip, the second output end is directly transmitted to the trigger channel of the mixed domain oscilloscope; the excitation transducer chip is used to generate an excitation ultrasonic signal, and the receiving transducer is used to receive the transmitted ultrasonic signal and transmit the ultrasonic signal to The upper controller.

Compared with the prior art, the beneficial technical effects of the present invention are:

(1) The polycrystalline material grain size evaluation method proposed by the present invention based on the average power attenuation of ultrasonic wake waves is based on the ultrasonic scattering theory and uses the influence of the microstructure within the polycrystalline material on the ultrasonic properties to intercept the wake waves respectively. Time window and reference time window, extract the average power attenuation of the coda wave corresponding to the average grain size of the material, and construct the average grain size d of the sample and the average power attenuation of the coda wave Logarithmic linear model between the two, and then obtain the average power attenuation of the coda of the polycrystalline material to be detected, and determine the average grain size of the polycrystalline material to be detected based on the above logarithmic linear model; this detection method does not require destroying the polycrystalline material to be detected That is to say, the average grain size of polycrystalline materials can be accurately obtained. The method combines the advantages of ultrasonic attenuation characteristics and wake detection, and proposes a method that uses the average power attenuation of wake waves to quantitatively evaluate the average grain size of polycrystalline materials. It has high detection efficiency and good measurement robustness.

(2) The polycrystalline material grain size evaluation method proposed by the present invention based on the average power attenuation of ultrasonic coda waves. Compared with the traditional ultrasonic method that uses direct wave attenuation, the attenuation of the coda wave part is used to evaluate ultrasonic waves and materials. It can be used in A single sensor position evaluates a wide range of grain size information through a single measurement, which has higher detection efficiency; the average power is used to describe the energy state of ultrasonic waves, which avoids the difficulty of extracting the ultrasonic amplitude after propagation in a highly non-uniform structure. The disadvantage of being susceptible to interference enables a more robust grain size assessment. The true grain size measured by electron backscatter diffraction technology is included within the 99% confidence interval of the obtained grain size prediction, illustrating the reliability of the prediction. .

Description of the drawings

Other features, objects and advantages of the present application will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings.

Figure 1 is a flow chart of the polycrystalline material grain size evaluation method based on ultrasonic wake average power attenuation according to the present invention;

Figure 2 is a schematic diagram of the average grain size detection system and flow chart of polycrystalline materials in an embodiment of the present invention;

Figure 3 is a schematic diagram of sample size and transducer arrangement in an embodiment of the present invention;

Figure 4 is a comparison diagram of waveforms after filtering in the reference time window and the coda time window measured on different samples in the embodiment of the present invention;

Figure 5 is a comparison of the logarithmic linear relationship between the average grain size of the sample and the average power attenuation of the coda wave established using logarithmic linear fitting in the embodiment of the present invention, and the experimental values of the sample and the polycrystalline material to be detected.

Detailed ways

The present application will be further described in detail below in conjunction with the accompanying drawings and examples. It can be understood that the specific embodiments described here are only used to explain the relevant invention, but not to limit the invention. It should also be noted that, for convenience of description, only the parts related to the invention are shown in the drawings.

It should be noted that, as long as there is no conflict, the embodiments and features in the embodiments of this application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and embodiments.

The present invention proposes a method for evaluating the grain size of polycrystalline materials based on the average power attenuation of ultrasonic coda waves. As shown in Figure 1, the method includes the following steps:

S1. Prepare several samples with the same thickness as the polycrystalline material to be detected, determine the method and parameters of the excitation signal, and build a polycrystalline material average grain size detection system.

The polycrystalline material average grain size detection system includes an arbitrary function generator, BNC tee joint, high-voltage power amplifier, excitation transducer chip, receiving transducer chip, mixed domain oscilloscope and upper controller. The arbitrary function generator outputs ultrasonic waves in a single channel and transmits them to the three-way joint. The three-way joint has two output terminals. The first output terminal is connected to the high-voltage power amplifier and is amplified and then transmitted to the excitation ultrasonic transducer chip. The second output terminal Directly passed to the trigger channel of the mixed domain oscilloscope; the excitation transducer chip is used to generate the excitation ultrasonic signal, and the receiving transducer is used to receive the transmitted ultrasonic signal and transmit the ultrasonic signal to the upper controller. In this embodiment, the upper controller is a computer.

This step specifically includes the following sub-steps:

S11. According to the material composition information of the polycrystalline material to be detected (which can also be called polycrystalline metal material or polycrystalline metal structure), prepare several samples with the same chemical composition and different average grain sizes by adjusting the heat treatment process parameters. , and cut into plate specimens with the same thickness as the structure to be tested.

S12. The excitation source used to excite the ultrasonic signal is a Hanning window modulated sine wave pulse with a single center frequency. Based on the pre-understood elastic properties and thickness information of the polycrystalline material to be detected, the ultrasonic signal frequency is selected to balance the grain size measurement. sensitivity and resolution.

S13. Use an arbitrary function generator to output ultrasonic waves in a single channel, and form two output terminals through a BNC tee connector. The first output terminal is connected to a high-voltage power amplifier, and is amplified and then transmitted to the excitation ultrasonic transducer chip. The second output terminal Directly passed to the trigger channel of a mixed domain oscilloscope.

S14. For each sample, prepare a pair of excitation transducer chips and receiving transducer chips with the same center frequency, place them closely together, and fix them at the center of the sample surface with glue.

S15. After each excitation, according to the trigger signal received by the trigger channel, the ultrasonic signal propagating in the sample is collected through the receiving transducer chip, and transmitted to the acquisition channel of the mixed domain oscilloscope for subsequent data processing.

S2. Excite and receive ultrasonic signals for samples with different average grain sizes respectively, and perform band-pass filtering preprocessing on the collected ultrasonic signals. This step specifically includes the following sub-steps:

S21. Randomly cut and sample each sample, and use electron backscatter diffraction technology to quantify its true grain size as a reference.

S22. Use the pulse echo method to detect each sample. The excitation transducer chip generates an excitation ultrasonic signal in the sample, and the receiving transducer receives the ultrasonic signal transmitted in the sample. The ultrasonic signal received by the receiving transducer is the average of 128 consecutive signal acquisitions by the mixed domain oscilloscope.

S23. Repeat steps S21 and S22 until ultrasonic signals corresponding to samples with different average grain sizes are obtained, and perform band-pass filtering preprocessing on the ultrasonic signals to remove frequency components irrelevant to grain size evaluation.

S3. Select the ultrasonic signals of the coda time window [t, t + T] and the reference time window [t', t' + T ₀ ], respectively extract the average power of the two, and calculate the average coda power attenuation based on the time difference. This step specifically includes the following sub-steps:

S31. According to the starting point t of the time window and the width T of the time window, extract the signal voltage amplitude V(τ) at different times τ within the selected coda time window [t, t+T], and calculate the average power of the coda at time t. P(t) is:

The average coda power P(t) represents the energy state of the coda part of the ultrasonic signal at time t. The selection of the coda part is preferably based on the arrival of twice the shortest boundary echo as the starting time, and the coda signal within the selected time window must have a high signal-to-noise ratio.

S32. According to the time window starting point t' and the time window width T ₀ , extract the signal voltage amplitude V(τ) at different moments τ within the selected coda time window [t', t'+T ₀ ]. Within a certain reference time window [t',t'+T ₀ ], calculate the reference average power at time t ₀ =t'+T ₀ /2 as P(t ₀ ):

The reference average power P(t ₀ ) represents the energy state of the main wave part of the ultrasonic signal at time t ₀ .

S33. Calculate the average coda power attenuation based on the average coda power P(t) and the reference average power P(t ₀ ) of the ultrasonic signal at time t and t ₀ and the time difference Δt=tt ₀ for:

Coda average power attenuation Represents the coda power dissipation at the interval Δt due to grain scattering and other factors.

S4. Based on the equivalent attenuation coefficient model, establish the average grain size d of the sample and the average power attenuation of the coda wave. Log-linear model between:

Among them, θ ₁ and θ ₂ respectively represent the first fitting parameter of the model and the second fitting parameter of the model, both of which are evaluated by the least squares method.

S5. Collect the ultrasonic signal of the polycrystalline material with unknown grain size to be detected, and extract the average power attenuation of the tail wave of the polycrystalline material with unknown grain size to be detected according to step S3. and determine the average grain size of polycrystalline materials based on the log-linear model described in S4/> for:

The present invention will be further described in detail below with reference to a specific polycrystalline material grain size detection case. The overall process is shown in Figure 2.

S1. Prepare five samples with the same composition as the nickel-based high-temperature alloy sample with the brand name GH742 to be tested (#1, #2, #3, #4, #5 respectively), the sample and the sample to be tested The thickness of the samples is 5mm, the length is 190mm, and the width is 100mm, as shown in Figure 3. The specific process is as follows:

S11. The five samples used different solid solution temperatures and times in the heat treatment process, resulting in them having different average grain sizes.

S12. The excitation signal is a 3.5-period Hanning window modulated sinusoidal pulse with a center frequency of 5MHz.

S13. Use an arbitrary function generator (Tektronix, AFG 31022) to generate an excitation signal and form two outputs through a BNC tee connector. The first output end is connected to a high-voltage power amplifier (Aigtek, ATA-4012). After amplification, the peak-to-peak value is 80V is passed to the excitation transducer chip (Siansonic, center frequency is 5MHz), and the second output is directly passed to the trigger channel of the mixed domain oscilloscope (Tektronix, MDO3104).

S14. The circular excitation transducer chip with a diameter of 10mm and the receiving transducer chip (Siansonic, center frequency is 5MHz) are placed closely together and fixed at the center of the sample surface through 502 quick-drying glue, as shown in Figure 3 .

S15. After each excitation, according to the trigger signal received by the trigger channel, the ultrasonic signal propagating in the sample is collected through the receiving transducer chip, and transmitted to the acquisition channel of the mixed domain oscilloscope.

S2. Randomly cut and sample each sample, use electron backscattered diffraction technology to take photos of the microstructure of the sample, obtain the logarithmic normal distribution experimental value of the grain size through the equivalent diameter method, and take the logarithmic normal distribution experimental value of the grain size. The mean value of the normal distribution is used as a reference for its true grain size. The ultrasonic signals are excited and received respectively for samples with different average grain sizes. The collected signals are preliminarily averaged in the time domain of 128 consecutive acquisition signals inside the mixed-domain oscilloscope. After being input to the upper controller, a wavelet filter is used for passing the signals. Bandpass filter preprocessing with [0.1MHz, 10MHz].

S3 and Figure 4 show the ultrasonic signals in the coda time window [66.22μs, 77.22μs] and reference time window [0.5μs, 40μs] selected for the five samples. A feature that is sensitive to grain size is the attenuation of the coda part of the received signal data, and different initial contact states will ultimately change the energy transmitted into the structure, which affects both the coda part of the signal and the reference window part of the signal. . Since the attenuation coefficient is a relative value, the impact of the contact state on the signal can be minimized. In this case, the variation in attenuation coefficient between different specimens is largely attributed to different grain sizes. The average power is extracted from the time window of the corresponding signal of each sample, and the average power attenuation of the coda wave is calculated based on the time difference according to formula (3).

S4. As shown in formula (4), the average grain size d of the sample and the average power attenuation of the coda wave are established through logarithmic linear fitting of the experimental results. The logarithmic linear relationship model is shown in Figure 5. Among them, the solid circle mark represents the relationship between the reference average grain size and the coda average power attenuation quantified by electron backscatter diffraction technology for each sample. The solid line plots the log-linear least squares fitting result. It is obtained that the first fitting parameter of the model θ ₁ =5.5343, and the second fitting parameter of the model θ ₂ =0.8732. When the average grain size increases, the average coda power attenuation increases with a logarithmic linear trend.

S5. Collect the signal of the polycrystalline material with unknown grain size to be detected. After filtering, obtain the corresponding tail wave average power attenuation of the polycrystalline material to be detected through step S3. Determine the average grain size of the polycrystalline material according to the logarithmic linear model. Then The electron backscatter diffraction technique was used to evaluate the true average grain size for verification. The verification results are shown as solid square marks in Figure 5. The results show that the true grain size measured by the electron backscatter diffraction technique is included within the 99% confidence interval of the model drawn by the black dotted line, indicating that the proposed model can be used to evaluate the average grain size of the polycrystalline material under test size, the prediction results are very accurate.

The polycrystalline material grain size evaluation method proposed by the present invention based on the average power attenuation of ultrasonic coda wave utilizes the influence of the microstructure within the polycrystalline material on the ultrasonic properties, selects the ultrasonic coda wave as the measurement signal, and intercepts the coda wave time window and Refer to the time window, extract the average power attenuation of the coda wave corresponding to the average grain size of the material, and construct the average grain size d of the sample and the average power attenuation of the coda wave Through the logarithmic linear model between the polycrystalline materials to be detected, the average power attenuation of the coda wave of the polycrystalline material to be detected can be obtained through experiments, so that the average grain size of the polycrystalline material can be efficiently detected; this detection method combines the characteristics of ultrasonic attenuation and coda wave detection. Advantages: It can quantitatively evaluate the average grain size of polycrystalline materials by using the average power attenuation of the coda wave. It has high detection efficiency, good measurement robustness and accurate detection results.

Finally, it should be noted that the above embodiments are only for illustrating and not limiting the technical solution of the present invention. Although the present invention has been described in detail with reference to the above embodiments, those of ordinary skill in the art should understand that the present invention can still be modified or modified. Equivalent substitutions, and any modifications or partial substitutions that do not depart from the spirit and scope of the present invention shall be included in the scope of the claims of the present invention.

Claims (6)

1. A polycrystalline material grain size assessment method based on ultrasonic wake average power attenuation, which is characterized by comprising the following steps:

s1, preparing a plurality of samples with the same thickness as the polycrystalline material to be detected, determining the mode and parameters of an excitation signal, and constructing a polycrystalline material average grain size detection system;

the step S1 specifically comprises the following steps:

s11, preparing a plurality of samples with the same chemical composition and different average grain sizes according to the material composition information of the polycrystalline material to be detected, and cutting the samples into plate-shaped samples with the thickness consistent with that of the polycrystalline material to be detected;

s12, an excitation source for exciting an ultrasonic signal is a Hanning window modulated sine wave pulse with a single center frequency, and the ultrasonic signal frequency is selected according to the elastic property and thickness information of the polycrystalline material to be detected;

s13, outputting ultrasonic waves by using a single channel of an arbitrary function generator, and forming two output ends through a three-way connector, wherein the first output end is connected to a high-voltage power amplifier, amplified and then transmitted to an excitation ultrasonic transducer chip, and the second output end is directly transmitted to a trigger channel of a mixed domain oscilloscope;

s14, for each sample, a pair of excitation transducer wafers and receiving transducer wafers with the same center frequency are closely placed and fixed at the center position of the sample surface;

s15, after each excitation, according to the trigger signal received by the trigger channel, acquiring an ultrasonic signal transmitted in a sample through a receiving transducer wafer and transmitting the ultrasonic signal to an acquisition channel of the mixed domain oscilloscope;

s2, respectively exciting and receiving ultrasonic signals of samples with different average grain sizes, and carrying out band-pass filtering pretreatment on the collected ultrasonic signals;

s3, selecting a wake time window [ T, t+T ]]Reference time window [ T ', T' +T ] ₀ ]Respectively extracting average power, and calculating wake average power attenuation according to time difference, wherein the method specifically comprises the following sub-steps;

s31, extracting signal voltage amplitude values V (tau) of different moments tau in a selected wake time window [ T, t+T ] according to a time window starting point T and a time window width T, and calculating wake average power P (T) at the moment T as follows:

wherein the wake average power P (t) represents the energy state of the wake part of the ultrasonic signal at the time t;

s32, according to the starting point T' of the time window and the width T of the time window ₀ Extracting a selected reference time window [ T ', T' +T ] ₀ ]The signal voltage amplitude V (tau) at different moments tau within a selected reference time window T ', T' +T ₀ ]In, calculate t ₀ ＝t′+T ₀ Reference average power P (t) at time/2 ₀ ) The method comprises the following steps:

wherein the reference average power P (t ₀ ) Representing the main wave part of the ultrasonic signal at t ₀ The energy state at the moment;

s33, according to the ultrasonic signals, at t and t ₀ Wake average power at time P (t) and reference average power P (t) ₀ ) Time difference Δt=t-t ₀ Constructing a wake average power equivalent attenuation coefficient model, and calculating wake average power attenuation of the sampleThe wake average power equivalent attenuation coefficient model is as follows:

the wake average power decayWake power dissipation, which represents the separation Δt;

the wake time window and the reference time window extracted in the step S3 are respectively positioned at the wake part and the main wave part of the ultrasonic signal; selecting the wake wave part with the arrival time of the double shortest boundary echo as the starting moment, wherein the wake wave signals in the selected wake wave time window have high signal-to-noise ratio;

s4, establishing the average grain size d of the sample and the average power attenuation of the wake wave based on the equivalent attenuation coefficient model of the average power of the wake waveThe log-linear model between is shown below:

wherein θ ₁ 、θ ₂ Respectively representing a first fitting parameter of the model and a second fitting parameter of the model;

s4, evaluating a model first fitting parameter and a model second fitting parameter through a least square method;

s5, collecting ultrasonic signals of the polycrystalline material with unknown grain size to be detected by utilizing a polycrystalline material average grain size detection system, filtering, and extracting wake average power attenuation of the polycrystalline material according to the step S3And determining the average grain size of the polycrystalline material based on the log-linear model described in step S4>The calculation method comprises the following steps:

2. the method for evaluating grain size of polycrystalline material based on ultrasonic wake average power attenuation of claim 1, wherein step S2 specifically comprises the steps of:

s21, randomly dicing and sampling each sample, and obtaining the actual grain size;

s22, detecting each sample in a pulse echo mode, exciting a transducer wafer to generate an excitation ultrasonic signal in the sample, and receiving the ultrasonic signal transmitted in the sample by the transducer wafer;

s23, repeatedly executing the step S21 and the step S22 until ultrasonic signals corresponding to the samples with different average grain sizes are obtained, and carrying out band-pass filtering pretreatment on the ultrasonic signals.

3. The method of claim 1, wherein the signal received by the receiving transducer wafer in step S2 is an average of 128 consecutive signal acquisitions by a mixed domain oscilloscope.

4. The method for evaluating grain size of polycrystalline material based on average power attenuation of ultrasonic wake wave according to claim 1, wherein the ultrasonic signal in each sample is detected by pulse echo in step S2.

5. The method for evaluating grain size of polycrystalline material based on ultrasonic wake average power attenuation according to claim 1, wherein the actual grain size d of the sample in step S21 is quantified by means of electron back scattering diffraction technique.

6. The method for evaluating grain size of polycrystalline material based on ultrasonic wake average power attenuation of claim 1, wherein the system for detecting grain size of polycrystalline material in step S1 comprises an arbitrary function generator, a three-way connector, a high voltage power amplifier, an excitation transducer wafer, a receiving transducer wafer, a mixed domain oscilloscope, and a host controller;

the three-way connector is provided with two output ends, the first output end is connected to a high-voltage power amplifier, amplified and then transmitted to an excitation ultrasonic transducer wafer, and the second output end is directly transmitted to a trigger channel of the mixed domain oscilloscope; the exciting transducer chip is used for generating an exciting ultrasonic signal, and the receiving transducer is used for receiving the transmitted ultrasonic signal and transmitting the ultrasonic signal to the upper controller.