Device and method for nondestructive testing of grain size of hot rolled strip

Through the integrated device of iron oxide removal module and laser ultrasonic detection module, the problem of iron oxide interference and low detection efficiency is solved, and efficient and accurate detection of the grain size of hot-rolled strips is achieved. It is suitable for production sites and improves detection efficiency and precision, and promote to other sheet fields.

CN118961602APending Publication Date: 2024-11-15ANSTEEL BEIJING RES INST CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411055833.8
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-02
Publication Date
2024-11-15

AI Technical Summary

Technical Problem

The prior art has interference with iron oxide sheets, low detection efficiency, low accuracy, and difficult detection of large-sized steel plates in the detection of grain size of hot-rolled plates. The traditional methods have limited applicability to production sites.

Method used

A device with integrated iron oxide strip removal module, surface gloss detector, laser ultrasonic detection module and control system was designed to remove iron oxide strips through electric grinding wheels and electromagnets. The frequency domain characteristics are calculated using laser ultrasonic detection and time-frequency transformation, combined with Mechanical model and compensation function fit to achieve contactless and real-time detection.

Benefits of technology

It improves detection efficiency and accuracy, realizes real-time inspection of any position of hot-rolled strip, has a compact and flexible structure, is suitable for production sites, and can be promoted to other plate fields, reducing production costs and improving production quality.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118961602A_ABST
    Figure CN118961602A_ABST
Patent Text Reader

Abstract

The invention relates to a device and method for nondestructive detection of grain size of a hot rolled strip, and the device comprises an oxide scale removal module, a surface glossiness detector, a control system, a laser ultrasonic detection module, an installation platform and a notebook computer. The oxide scale removal module, the surface gloss detection module, the control system and the laser ultrasonic detection module are arranged on the mounting platform, and movable baffles are arranged outside the oxide scale removal module, the surface gloss detection module and the laser ultrasonic detection module; the oxide scale removal module, the surface gloss detection module and the laser ultrasonic detection module are connected to a control system, and the control system is connected with a notebook computer; according to the device, oxide scale removal and laser ultrasonic detection are realized, the detection efficiency and the detection precision are greatly improved, and in addition, the device has the characteristics of compact structure, convenience, flexibility, non-contact measurement, high detection speed and the like.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of nondestructive testing, and in particular to a device and method for nondestructive testing of grain size of hot-rolled strips. Background Art

[0002] The commonly used method for hot-rolled strip grain size detection is to perform wire cutting sampling on a specified part, then obtain the microstructure image of the detection part through metallographic method and other methods, and then count the grain size in the image. This method is time-consuming and has a delayed feedback on the production process. The ultrasonic method can realize in-situ detection of grain size, but the hot-rolled strip at the production site has obvious iron oxide scale, which has a great impact on the acquisition of ultrasonic signals.

[0003] The patent application with application number 202111419601.2 discloses an ultrasonic nondestructive testing method for detecting the grain size distribution of aero-engine turbine disks. The ultrasonic detection system and the water-immersion ultrasonic focusing probe are used to scan and collect ultrasonic signals from the nickel-based high-temperature alloy test block of the aero-engine turbine disk. Based on the calculation results of the backscatter coefficient, a two-dimensional distribution image of the grain size is obtained; however, this method is based on a water-immersion ultrasonic system and requires the test sample to be placed in water. For large-sized and large-sized steel plates at the production site, the detection is difficult and the detection efficiency is low.

[0004] Patent application number 202311468359.7 discloses a measurement method for characterizing the ferrite grain size and phase fraction of duplex stainless steel. The method obtains the grain size information of the tested sample by constructing a response database between the change in single ferrite grain size, the change in ferrite phase fraction in duplex steel and the electromagnetic signal. However, this method is based on electromagnetic characterization technology and is more suitable for the measurement of smaller samples. It has limitations for use in production sites. The residual magnetism of the steel plate has a certain interference with the detection results. In addition, the frequency inductance curve is not the most sensitive feature of the microstructure, and the detection resolution is limited.

[0005] The patent application with application number 202222462321.6 discloses a laser ultrasonic non-contact non-destructive testing device, including a testing table, a laser testing mechanism and an ultrasonic testing mechanism. This method requires the steel plate to be placed on the testing table. From the perspective of the testing space, it is not very flexible, that is, it has requirements on the shape and volume of the workpiece to be tested. In addition, the device is designed with a cleaning mechanism that can automatically clean the surface oil, but it still has certain limitations for parts with iron oxide on the surface.

[0006] The paper titled "Laser Ultrasonic Evaluation of Al-Cu-Mg Aluminum Alloy Based on Frequency Domain Attenuation Coefficient" fits the formula of grain size and frequency domain attenuation coefficient of Al-Cu-Mg aluminum alloy through a cubic polynomial. Among them, the thickness L of the sample in the frequency domain attenuation coefficient calculation formula needs to be obtained through measurement. Due to problems such as the plate shape of the steel plate at the production site, the thickness of each inspection position may not be exactly the same. Therefore, repeated measurement of thickness is undoubtedly a large workload, which will affect the detection efficiency. If the thickness of the steel plate is not considered, it will affect the detection accuracy. Summary of the invention

[0007] The present invention provides a device and method for nondestructive testing of hot-rolled strip grain size, which realizes the removal of iron oxide scale and laser ultrasonic testing, greatly improving the testing efficiency and accuracy. In addition, the device has the characteristics of compact structure, convenience and flexibility, non-contact measurement, and fast testing speed.

[0008] In order to achieve the above object, the present invention adopts the following technical solutions:

[0009] A device for nondestructive testing of grain size of hot-rolled plates and strips comprises an iron oxide scale removal module, a surface gloss detector, a control system, a laser ultrasonic detection module, a mounting platform and a laptop computer. The iron oxide scale removal module, the surface gloss detector, the control system and the laser ultrasonic detection module are all arranged on the mounting platform, and movable baffles are arranged outside the iron oxide scale removal module, the surface gloss detector and the laser ultrasonic detection module. The iron oxide scale removal module, the surface gloss detector and the laser ultrasonic detection module are connected to the control system, and the control system is connected to the laptop computer.

[0010] Furthermore, the iron oxide scale removal module includes an electric grinding wheel, a guide rail, an iron oxide scale sucker and a waste collection bucket. The electric grinding wheel is an electric driven grinding wheel that can move up and down. The iron oxide scale sucker includes a spring and an electromagnet. The electromagnet is fixedly connected to the spring, and the spring is connected to the guide rail. The iron oxide scale sucker can move forward and backward along the guide rail. The waste collection bucket is arranged below the iron oxide scale sucker at a set discharge point, and the electrical signals of the electric grinding wheel and the electromagnet are connected to a control system.

[0011] Furthermore, the laser ultrasonic detection module includes a pulse laser, a pulse laser mounting platform, a pulse laser fiber probe, a continuous laser, a dual-wave hybrid interferometer, a mounting frame, a signal light fiber probe, a high-speed signal acquisition system and an acquisition system mounting platform. The pulse laser is arranged on the pulse laser mounting platform, the continuous laser and the dual-wave hybrid interferometer are arranged on the mounting frame, the high-speed signal acquisition system is arranged on the acquisition system mounting platform, the signal line of the pulse laser fiber probe is connected to the pulse laser, and the signal light fiber probe is connected to the dual-wave hybrid interferometer.

[0012] Furthermore, the control system includes a portable power supply, a controller, a mobile device, a limit device, a high-speed signal acquisition result display system and a signal analysis and processing system, and the iron oxide removal module, the surface gloss detector, the laser ultrasonic detection module, the mobile device and the limit device are all connected to the controller.

[0013] Furthermore, the pulse laser installation platform, the installation frame and the acquisition system installation platform adopt a hard aluminum alloy fixing plate specially used for optical instruments.

[0014] A method for nondestructive testing of grain size of hot-rolled strip comprises the following steps:

[0015] S1. Removing the iron oxide scale on the surface of the hot-rolled steel plate at the position to be detected;

[0016] S2. Detect the surface glossiness of multiple parts of the iron oxide scale after removal, and determine whether it is qualified;

[0017] S3. After the glossiness is qualified, the moving device is automatically triggered to move the entire testing equipment to the testing position;

[0018] S4, obtaining the ultrasonic signal of the detection position by exciting and receiving ultrasonic waves on the same side;

[0019] S5, extracting the time of the echo in the ultrasonic signal and calculating the thickness of the tested sample;

[0020] S6. Perform time-frequency transformation on the signal, obtain frequency domain features, and calculate the frequency domain attenuation coefficient;

[0021] S7. Characterize the average grain size at the detection position according to the characteristic parameters of the detected ultrasonic signal.

[0022] Furthermore, the step S1 specifically includes: the baffle corresponding to the iron oxide scale removal module is opened, the electric grinding wheel moves downward to the detection position, and the iron oxide scale begins to be removed; after the iron oxide scale is removed, the electric grinding wheel returns to its original position, and the electromagnet used to absorb the iron oxide scale moves to the working position and then moves downward; after the electromagnet is triggered to be energized, the iron oxide scale on the surface is sucked up, and then the electromagnet moves backward to the position of the waste collection bucket, the electromagnet loses power, and the adsorbed iron oxide scale is placed in the waste bucket.

[0023] Furthermore, the step S5 specifically includes the following steps:

[0024] S5.1. Extract the time corresponding to the mth and nth echo peaks in the ultrasonic signal and record them as t m ,t n ;

[0025] S5.2, according to the pre-calibrated ultrasonic longitudinal wave propagation speed v in this material, echo time interval Δt = t n -t m , calculate the thickness H of the sample at the test point:

[0026]

[0027] Among them, m and n are the mth echo and the nth echo respectively.

[0028] Furthermore, the step S6 specifically includes the following steps:

[0029] S6.1. performing time-frequency transformation on the ultrasonic signal to obtain frequency domain information of the ultrasonic signal;

[0030] S6.2. Extract the frequency domain amplitudes F1 and F2 of the first and second echoes, and calculate the frequency domain attenuation coefficient α:

[0031]

[0032] S6.3, extract the center frequencies f1 and f2 of the first and second echoes, and calculate the center frequency offset F s :

[0033] F s =f1-f2 (3)

[0034] Furthermore, the step S7 specifically includes:

[0035] S7.1. The actual detected data are divided into two groups. One group of data is used as the calibration data set of the mechanism model f(α). The statistical results of the grain size D are obtained from the microstructure image obtained by the metallographic experiment. s , calibration constants b1 and c1;

[0036] D s =f(α)=b1·α k +c1 (4)

[0037] Wherein, k is a parameter determined according to the relative relationship between the grain size and the ultrasonic wavelength;

[0038] S7.2, another set of data is used as the compensation function fitting data set, and the frequency domain attenuation coefficient calculated from a set of signals is substituted into the function f(α) fitted in step S7.1 to calculate the calculated value D of the grain size. c , and metallographic analysis results D s Compare and obtain the difference ΔD;

[0039] ΔD=D S -D C (5)

[0040] S7.3. Using different types of functions to analyze the ultrasonic characteristic parameters v, F s The relationship between the difference ΔD is fitted to obtain the compensation function ψ(v,F s );

[0041] ΔD=b2·ψ(v,F s )+c2 (6)

[0042] Among them, b2 and c2 are constants used for model fitting;

[0043] S7.4. Fitting the mechanism model f(α) and the compensation function ψ(v,F) by machine learning algorithm s )) weight coefficients g1, g2, and constant term c, the characterization model of average grain size is obtained:

[0044] D=g1·f(α)+g2·ψ(v,F s )+c (7)

[0045] Where D is the characterization model of the average grain size.

[0046] Compared with the prior art, the present invention has the following beneficial effects:

[0047] 1) The surface iron oxide treatment device and the laser ultrasonic nondestructive testing device are integrated on a small vehicle, so that the grain size of the hot-rolled strip at any position can be detected in real time at the production site without being restricted by the sampling position and the testing location;

[0048] 2) The device can be used to remove iron oxide scale and perform laser ultrasonic testing on hot-rolled strips. It has a compact structure, is convenient and flexible, has non-contact measurement, and has a fast testing speed.

[0049] 3) The average grain size calculation method adds a precision compensation model of multiple ultrasonic characteristic parameters on the basis of the traditional mechanism model, which improves the characterization accuracy;

[0050] 4) It can be extended to the fields of cold-rolled steel plates, titanium alloy plates, aluminum alloy plates, etc., which has positive significance for reducing production costs, improving production quality, and timely improving production processes. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 It is a structural schematic diagram of the device of the present invention.

[0052] Figure 2 It is a schematic diagram of the position of the baffle of the present invention.

[0053] Figure 3 It is a schematic diagram of the structure of the iron oxide scale removal module of the present invention.

[0054] Figure 4 It is a schematic diagram of the structure of the laser ultrasonic detection module of the present invention.

[0055] Figure 5 It is a flow chart of the method of the present invention.

[0056] Figure 6 It is a flow chart of the grain size calculation method of the present invention.

[0057] Figure 7 Schematic diagram of the laser ultrasonic signal in an embodiment of the present invention.

[0058] Figure 8 It is a schematic diagram of the echo frequency domain information described in an embodiment of the present invention.

[0059] In the figure: 1. Laptop 2. Mounting platform 201. Cable management slot 202. Baffle 3. Iron oxide scale removal module 301. Electric grinding wheel 302. Guide rail 303. Iron oxide scale suction device 304. Waste collection bucket 4. Surface gloss detector 5. Control system 6. Laser ultrasonic detection module 601. Pulse laser 602. Pulse laser mounting platform 603. Pulse laser fiber probe 604. Continuous laser 605. Dual-wave hybrid interferometer 606. Mounting frame 607. Signal light fiber probe 608 Collection system mounting platform 609. High-speed signal acquisition system DETAILED DESCRIPTION

[0060] The specific implementation of the present invention will be further described below in conjunction with the accompanying drawings:

[0061] See Figure 1, is a schematic diagram of the structure of the device of the present invention. The device of the present invention for nondestructive detection of grain size of hot-rolled strip comprises an oxide scale removal module 3, a surface gloss detector 4, a control system 5, a laser ultrasonic detection module 6, an installation platform 2 and a laptop computer 1. The oxide scale removal module 3, the surface gloss detector 4, the control system 5, and the laser ultrasonic detection module 6 are all arranged on the installation platform 2; see Figure 2 The installation platform 2 is provided with a wire management groove 201, and the outside of the iron oxide removal module 3, the surface gloss detector 4 and the laser ultrasonic detection module 6 are all equipped with a controllable opening and closing square baffle 202 to reduce the contact of precision instruments with dust, lubricating oil, etc.; the iron oxide removal module 3, the surface gloss detector 4 and the laser ultrasonic detection module 6 are connected to the control system 5, and the control system 5 is connected to the laptop computer 1; the surface gloss detector 4 automatically judges whether it meets the detection requirements according to the gloss detection result.

[0062] See Figure 3 The iron oxide scale removal module 3 includes an electric grinding wheel 301, a guide rail 302, an iron oxide scale suction device 303 and a waste collection bucket 304. The electric grinding wheel 301 is an electric driven grinding wheel that can move up and down. The iron oxide scale suction device includes a spring and an electromagnet. The electromagnet is fixedly connected to the spring, and the spring is connected to the guide rail 302. The iron oxide scale suction device can move forward and backward along the guide rail 302. The waste collection bucket 304 is arranged below the iron oxide scale suction device 303 at a set discharge point. The electrical signals of the electric grinding wheel 301 and the electromagnet are connected to the control system 5.

[0063] See Figure 4 The laser ultrasonic detection module 6 includes a pulse laser 601, a pulse laser mounting platform 602, a pulse laser fiber probe 603, a continuous laser 604, a dual-wave mixing interferometer 605, a mounting frame 606, a signal light fiber probe 607, a high-speed signal acquisition system 609 and an acquisition system mounting platform 608. The pulse laser 601 is an air-cooled single portable laser, which is arranged on the pulse laser mounting platform 602. The continuous laser 604 and the dual-wave mixing interferometer 605 are arranged on the mounting frame 606. The high-speed signal acquisition system 609 is arranged on the acquisition system mounting platform 608. The signal line of the pulse laser fiber probe 603 is connected to the pulse laser 601, and the signal light fiber probe 607 is connected to the dual-wave mixing interferometer 605. The device adopts an ablation mechanism and a same-side optical path to excite and receive the laser; the pulse laser mounting platform 602, the mounting frame 606 and the acquisition system mounting platform 608 adopt a hard aluminum alloy fixing plate dedicated to optical instruments.

[0064] The control system 5 includes a portable power supply, a controller, a mobile device, a limit device, a high-speed signal acquisition result display system and a signal analysis and processing system. The iron oxide removal module 3, the surface gloss detector 4, the laser ultrasonic detection module 6, the mobile device and the limit device are all connected to the controller, and the mobile device can be moved to the specified position set by the limit device.

[0065] See Figure 5 A method for nondestructive testing of grain size of hot rolled strip comprises the following steps:

[0066] S1. Removing the iron oxide scale on the surface of the hot-rolled steel plate at the position to be detected: the baffle 202 corresponding to the iron oxide scale removal module 3 is opened, and the electric grinding wheel 301 moves downward to the detection position to start removing the iron oxide scale; after the iron oxide scale is removed, the electric grinding wheel 301 moves upward to return to its original position, and the electromagnet used to absorb the iron oxide scale moves to the working position and then moves downward; after the electromagnet is triggered to be energized, the iron oxide scale on the surface is sucked up, and then the electromagnet moves backward to the position of the waste collection bucket 304, and the electromagnet loses power to put the absorbed iron oxide scale into the waste bucket.

[0067] S2. Open the baffle 202 corresponding to the surface gloss tester 4, and use the built-in gloss tester to test the gloss of the part where the iron oxide scale is removed; determine whether it is qualified based on the test results of multiple positions.

[0068] S3, judging whether the surface glossiness test result is qualified, if qualified, closing the baffle 202, and moving the trolley to the laser ultrasonic testing position, if unqualified, repeating step S1.

[0069] S4. Acquire the ultrasonic signal at the detection position by means of same-side excitation and reception of ultrasonic waves: open the baffle 202 corresponding to the optical fiber probe, and after the optical fiber probe is aligned with the detection position, start the light sources of the pulse laser 601 and the continuous laser 604; the signal light reflected by the sample passes through the dual-wave hybrid interferometer 605, and the propagation signal of the ultrasonic wave in the strip is acquired through the high-speed signal acquisition system 609.

[0070] See Figure 6 , grain size calculation method flow chart:

[0071] S5. See Figure 7 , is an example diagram of laser ultrasonic signal. The echo time in the ultrasonic signal is extracted to calculate the thickness of the tested sample:

[0072] S5.1. Extract the time corresponding to the first and second echo peaks in the ultrasonic signal, m = 1, n = 2, recorded as t1 and t2 respectively; in this example, t1 = 1.7584 × 10 -6 s,t2=3.4496×10-6 s;

[0073] S5.2, based on the previously calibrated ultrasonic longitudinal wave propagation speed of 5900m / s in steel, combined with the echo time interval Δt=t n -t m , calculate the thickness H of the sample at the detection point.

[0074]

[0075] S6. Perform time-frequency transformation on the signal, obtain frequency domain features, and calculate the frequency domain attenuation coefficient:

[0076] S6.1. performing time-frequency transformation on the ultrasonic signal to obtain frequency domain characteristics of the ultrasonic signal;

[0077] S6.2, see Figure 8 , extract the first and second echo frequency domain amplitudes F1 and F2, and calculate the frequency domain attenuation coefficient α in combination with the steel plate thickness H obtained in S5.2,

[0078]

[0079] S6.3, extract the center frequencies f1 and f2 of the first and second echoes, and calculate the center frequency offset F s ;

[0080] F s =f1-f2 (10)

[0081] S7, characterizing the average grain size at the detection position according to the characteristic parameters of the detected ultrasonic signal;

[0082] S7.1. The actual detected data are divided into two groups. One group of data is used as the calibration data set of the mechanism model f(α). The statistical results of the grain size D are obtained from the microstructure image obtained by the metallographic experiment. s , and the frequency domain attenuation coefficient α calculated in step S6.2, calibration constants b1 and c1;

[0083] D s =f(α)=b1·α k +c1 (11)

[0084] Wherein, k is a parameter determined according to the relative relationship between the grain size and the ultrasonic wavelength;

[0085] S7.2, another set of data is used as the compensation function fitting data set, and the frequency domain attenuation coefficient calculated from a set of signals is substituted into the function f(α) fitted in step S7.1 to calculate the calculated value D of the grain size. c , and metallographic analysis results D s Compare and obtain the difference ΔD;

[0086] ΔD=D S -D C (12)

[0087] S7.3. Use different types of functions, such as linear, multiple, exponential, etc., to calculate the ultrasonic characteristic parameters v, F s The relationship between the difference ΔD is fitted to obtain the compensation function ψ(v,F s );

[0088] ΔD=b2·ψ(v,F s )+c2 (13)

[0089] Among them, b2 and c2 are constants used for model fitting;

[0090] S7.4. Fitting the mechanism model f(α) and the compensation function ψ(v,F) by machine learning algorithm s )) weight coefficients g1, g2, and constant term c, the characterization model of average grain size is obtained:

[0091] D=g1·f(α)+g2·ψ(v,F s )+c (14)

[0092] Where D is the characterization model of the average grain size to improve the characterization accuracy.

[0093] The above embodiments are implemented based on the technical solution of the present invention, and detailed implementation methods and specific operation processes are given, but the protection scope of the present invention is not limited to the above embodiments. The methods used in the above embodiments are conventional methods unless otherwise specified.

Claims

1. A device for nondestructive testing of grain size of hot-rolled strip, characterized in that: The invention comprises an iron oxide scale removal module, a surface gloss detector, a control system, a laser ultrasonic detection module, an installation platform and a laptop computer. The iron oxide scale removal module, the surface gloss detector, the control system and the laser ultrasonic detection module are all arranged on the installation platform, and movable baffles are arranged on the outside of the iron oxide scale removal module, the surface gloss detector and the laser ultrasonic detection module. The iron oxide scale removal module, the surface gloss detector and the laser ultrasonic detection module are connected to the control system, and the control system is connected to the laptop computer.

2. The device for nondestructive testing of grain size of hot-rolled strip according to claim 1, characterized in that: The iron oxide scale removal module includes an electric grinding wheel, a guide rail, an iron oxide scale sucker and a waste collection bucket. The electric grinding wheel is an electric driven grinding wheel that can move up and down. The iron oxide scale sucker includes a spring and an electromagnet. The electromagnet is fixedly connected to the spring, and the spring is connected to the guide rail. The iron oxide scale sucker can move forward and backward along the guide rail. The waste collection bucket is arranged below the iron oxide scale sucker at a set discharge point. The electrical signals of the electric grinding wheel and the electromagnet are connected to a control system.

3. The device for nondestructive testing of grain size of hot-rolled strip according to claim 1, characterized in that: The laser ultrasonic detection module includes a pulse laser, a pulse laser mounting platform, a pulse laser fiber probe, a continuous laser, a dual-wave hybrid interferometer, a mounting frame, a signal light fiber probe, a high-speed signal acquisition system and an acquisition system mounting platform. The pulse laser is arranged on the pulse laser mounting platform, the continuous laser and the dual-wave hybrid interferometer are arranged on the mounting frame, the high-speed signal acquisition system is arranged on the acquisition system mounting platform, the signal line of the pulse laser fiber probe is connected to the pulse laser, and the signal light fiber probe is connected to the dual-wave hybrid interferometer.

4. The device for nondestructive testing of grain size of hot-rolled strip according to claim 1, characterized in that: The control system includes a portable power supply, a controller, a mobile device, a limit device, a high-speed signal acquisition result display system and a signal analysis and processing system. The iron oxide scale removal module, the surface gloss detector, the laser ultrasonic detection module, the mobile device and the limit device are all connected to the controller.

5. The device for nondestructive testing of grain size of hot-rolled strip according to claim 3, characterized in that: The pulse laser installation platform, the installation frame and the collection system installation platform adopt a hard aluminum alloy fixing plate specially used for optical instruments.

6. A method for nondestructive testing of grain size of hot-rolled strip, characterized in that: The steps include: S1. Removing the iron oxide scale on the surface of the hot-rolled steel plate at the position to be detected; S2. Detect the surface glossiness of multiple parts of the iron oxide scale after removal, and determine whether it is qualified; S3. After the glossiness is qualified, the moving device is automatically triggered to move the entire testing equipment to the testing position; S4, obtaining the ultrasonic signal of the detection position by exciting and receiving ultrasonic waves on the same side; S5, extracting the time of the echo in the ultrasonic signal and calculating the thickness of the tested sample; S6, performing time-frequency transformation on the signal, obtaining frequency domain features, and calculating the frequency domain attenuation coefficient; S7. Characterize the average grain size at the detection position according to the characteristic parameters of the detected ultrasonic signal.

7. A method for nondestructive testing of grain size of hot rolled strip according to claim 6, characterized in that: The step S1 specifically includes: the baffle corresponding to the iron oxide scale removal module is opened, the electric grinding wheel moves downward to the detection position, and the iron oxide scale begins to be removed; after the iron oxide scale is removed, the electric grinding wheel returns to its original position, and the electromagnet used to absorb the iron oxide scale moves to the working position and then moves downward; after the electromagnet is triggered to be energized, the iron oxide scale on the surface is sucked up, and then the electromagnet moves backward to the position of the waste collection bucket, the electromagnet loses power, and the absorbed iron oxide scale is placed in the waste bucket.

8. The method for nondestructive testing of grain size of hot-rolled strip according to claim 6, characterized in that: The step S5 specifically includes the following steps: S5.

1. Extract the time corresponding to the mth and nth echo peaks in the ultrasonic signal and record them as t m ,t n ; S5.2, according to the pre-calibrated ultrasonic longitudinal wave propagation speed v in this material, echo time interval Δt = t n -t m , calculate the thickness H of the sample at the test point: Among them, m and n are the mth echo and the nth echo respectively.

9. The method for nondestructive testing of grain size of hot-rolled strip according to claim 6, characterized in that: The step S6 specifically includes the following steps: S6.

1. performing time-frequency transformation on the ultrasonic signal to obtain frequency domain information of the ultrasonic signal; S6.

2. Extract the frequency domain amplitudes F1 and F2 of the first and second echoes, and calculate the frequency domain attenuation coefficient α: S6.3, extract the center frequencies f1 and f2 of the first and second echoes, and calculate the center frequency offset F s : F s =f1-f2 (3)。 10. The method for nondestructive testing of grain size of hot-rolled strip according to claim 6, characterized in that: The step S7 specifically includes: S7.

1. The actual detected data are divided into two groups. One group of data is used as the calibration data set of the mechanism model f(α). The statistical results of the grain size D are obtained from the microstructure image obtained by the metallographic experiment. s , calibration constants b1 and c1; D s =f(α)=b1·α k +c1 (4) Wherein, k is a parameter determined according to the relative relationship between the grain size and the ultrasonic wavelength; S7.2, another set of data is used as the compensation function fitting data set, and the frequency domain attenuation coefficient calculated from a set of signals is substituted into the function f(α) fitted in step S7.1 to calculate the calculated value D of the grain size. c , and metallographic analysis results D s Compare and obtain the difference ΔD; ΔD=D S -D C (5) S7.

3. Using different types of functions to analyze the ultrasonic characteristic parameters v, F s The relationship between the difference ΔD is fitted to obtain the compensation function ψ(v,F s ); ΔD=b2·ψ(v,F s )+c2 (6) Among them, b2 and c2 are constants used for model fitting; S7.

4. Fitting the mechanism model f(α) and the compensation function ψ(v,F) by machine learning algorithm s ))'s weight coefficients g1, g2, and the constant term c, the characterization model of the average grain size is obtained: D=g1·f(α)+g2·ψ(v,F s )+c (7) Where D is the characterization model of the average grain size.