A Method for Determining the Hydrogen Damage State of Materials Based on Ultrasonic Online Monitoring

Abstract

A method for determining the hydrogen damage state of materials based on online ultrasonic monitoring includes establishing a hydrogen damage state prediction model and real-time monitoring of the hydrogen damage state of the test sample based on the prediction model. Establishing the hydrogen damage state prediction model involves first conducting a hydrogen charging test on a calibrated sample, plotting a calibration curve of the relative nonlinear optimal estimation coefficient versus hydrogen charging time, and a reference curve of the nth derivative of the calibration curve. Then, the relationship between hydrogen charging time and hydrogen damage state is determined, and a hydrogen damage state prediction model based on the change of the relative nonlinear optimal estimation coefficient is established. Real-time monitoring of the hydrogen damage state of the test sample includes real-time acquisition of the nonlinear ultrasonic signal of the test sample, processing it to obtain a real-time curve of the relative nonlinear optimal estimation coefficient of the test sample, and determining the hydrogen damage state of the test sample based on the above prediction model. This method can conveniently, quickly, accurately, and intuitively measure hydrogen damage in materials.

Description

Technical Field

[0001] This invention relates to a method for determining the hydrogen damage state of materials based on online ultrasonic monitoring, belonging to the field of ultrasonic nondestructive testing technology. Background Technology

[0002] Hydrogen damage to metals refers to the phenomenon where hydrogen interacts with materials, causing changes in their mechanical properties. Whether endogenous hydrogen is generated during the production process or exogenous hydrogen is introduced from the external environment, it can enter the material, reducing its toughness, plasticity, and mechanical properties, leading to a decrease in its resistance to stress corrosion. Common metals and alloys such as ferritic steel, austenitic stainless steel, aluminum alloys, and titanium alloys are highly sensitive to hydrogen. In environments containing H2 / H2S, humid air, or water, these metals and alloys are extremely prone to hydrogen damage, causing cracking or other damage, severely impacting the safe service life of the structure.

[0003] Currently, methods for detecting hydrogen damage in metallic materials are mainly divided into destructive testing methods and non-destructive testing methods. In the field of industrial inspection, non-destructive testing methods that do not damage, alter, or affect the performance of in-service equipment are irreplaceable means of ensuring product quality. Ultrasonic non-destructive testing technology is widely used in industrial inspection due to its numerous advantages, such as speed, portability, ease of operation, and applicability to in-service facilities. Existing ultrasonic methods for detecting the degree of hydrogen damage in materials are all in the linear ultrasonic domain, that is, using indicators such as ultrasonic velocity and sound attenuation to achieve detection and characterization. However, ultrasonic velocity is easily affected by ambient temperature, sound attenuation is easily affected by surface condition, and the detection resolution is limited by the wavelength of the sound wave. It is not sensitive to the detection of defects such as microcracks much smaller than the wavelength and the degradation of material mechanical properties, and cannot detect early hydrogen damage. Moreover, existing technologies for measuring hydrogen damage in materials using linear ultrasound mainly use surface acoustic waves. Longitudinal linear ultrasonic signals are not sensitive to hydrogen damage, and surface acoustic wave measurements have drawbacks such as not being able to use a self-transmitting mode and being unable to measure deep hydrogen damage. Nonlinear ultrasonic testing technology can overcome the above limitations. Nonlinear ultrasonic testing technology utilizes the nonlinear effects generated by the interaction between finite-amplitude sound waves and micro-defects as they propagate through a specimen to detect micro-defects and evaluate material properties. The basic principle of nonlinear ultrasonic technology is that the inherent nonlinearity of the material and the nonlinear interaction caused by damage defects distort the detection signal as it propagates in the solid, generating second-order and higher-order harmonic frequency components. Utilizing this characteristic, the nonlinear ultrasonic time-domain information of the tested material is converted into frequency-domain information, from which higher-order harmonic signals are extracted, thereby obtaining the damage information of the tested material. Based on this technology, it is possible to achieve online ultrasonic monitoring and determination of the hydrogen damage state of materials. Summary of the Invention

[0004] The purpose of this invention is to propose a method for determining the hydrogen damage state of materials based on online ultrasonic monitoring. This method can conveniently, quickly, accurately, and intuitively measure hydrogen damage in materials. Compared with traditional linear ultrasound, nonlinear ultrasound can detect hydrogen damage of smaller size.

[0005] The technical solution adopted by this invention to achieve its objective is as follows: a method for determining the hydrogen damage state of materials based on online ultrasonic monitoring, comprising the following steps:

[0006] S1. Prepare a sample of the same material and processing technology as the sample to be tested as a calibration sample.

[0007] S2. An ultrasonic transducer is placed on the calibration sample. The calibration sample is placed in a hydrogen charging solution for electrolytic hydrogen charging. The nonlinear ultrasonic system is used to collect the nonlinear ultrasonic signal of the calibration sample after the corresponding hydrogen charging time at regular intervals. The collected nonlinear ultrasonic signal is processed to obtain the relative nonlinear coefficients for different hydrogen charging times, forming a set of relative nonlinear coefficients.

[0008] S3. Perform optimal estimation calculation on the set of relative nonlinear coefficients to obtain the set of optimal estimation coefficients for relative nonlinearity, and plot the calibration curve of the optimal estimation coefficients for relative nonlinearity versus hydrogen charging time and the reference curve of the nth derivative of the calibration curve.

[0009] S4. Select reference time points based on the changes in the reference curve of the nth derivative of the calibration curve. Let p be the number of reference time points. Select time points 10-60 min before and after the reference time points as detection time points. Select q detection time points corresponding to each reference time point, for a total of m detection time points, where m = p * q. Select m calibration samples and place them in a hydrogen charging solution for electrolytic hydrogen charging. The hydrogen charging time is the same as the m detection time points. Then, detect the actual hydrogen damage degree of the m calibration samples to determine the relationship between hydrogen charging time and hydrogen damage state.

[0010] S5. Based on the calibration curve of the relative nonlinear optimal estimation coefficient and hydrogen charging time plotted in step S3 and the relationship between hydrogen charging time and hydrogen damage state determined in step S4, establish a hydrogen damage state prediction model based on the change of the relative nonlinear optimal estimation coefficient.

[0011] S6. Determine the hydrogen damage state of the test sample: Arrange an ultrasonic transducer with the same frequency as in S2 on the test sample. Use a nonlinear ultrasonic system to collect the nonlinear ultrasonic signal of the test sample at regular intervals, and perform signal processing on the nonlinear ultrasonic signal to obtain the relative nonlinear coefficient of the nonlinear ultrasonic signal of the test sample at each sampling time point. Calculate the optimal estimation coefficient of the relative nonlinearity at each sampling time point through optimal estimation, and plot the real-time curve of the optimal estimation coefficient of the relative nonlinearity of the test sample with time as the horizontal axis and the optimal estimation coefficient of the relative nonlinearity as the vertical axis. Based on the hydrogen damage state prediction model based on the change of the optimal estimation coefficient of the relative nonlinearity established in step S5 and the real-time curve of the optimal estimation coefficient of the relative nonlinearity of the test sample, determine the hydrogen damage state of the test sample.

[0012] Furthermore, the specific method for selecting the reference time point based on the change of the nth derivative reference curve of the calibration curve according to the present invention is as follows: the hydrogen charging time point where the derivative of the first derivative reference curve of the calibration curve is zero and the derivative of the second derivative reference curve of the calibration curve is not zero is taken as the reference time point.

[0013] Furthermore, the specific operation of detecting the actual hydrogen damage degree of m calibration samples and determining the relationship between hydrogen charging time and hydrogen damage state as described in this invention is as follows: Detect the actual hydrogen damage degree of m calibration samples, and record the reference time points corresponding to the detection time points with large changes in hydrogen damage degree as hydrogen damage characteristic time points; using the hydrogen damage characteristic time points as dividing points, divide the hydrogen damage state into different hydrogen damage stages, and determine the hydrogen damage stages corresponding to different hydrogen charging times, which is the relationship between hydrogen charging time and hydrogen damage state.

[0014] The criteria for significant changes in hydrogen damage are set by the sample material and the sample's usage environment. For example, for steel materials, the criteria for significant changes in hydrogen damage is to examine the sample cross-section under a microscope at 200x magnification to see if the difference in cracks can be detected. For some materials, the presence or absence of hydrogen embrittlement can also be used as the criteria for significant changes in hydrogen damage. The detection method for hydrogen embrittlement can directly adopt the national standard method, such as the closed bending method or the repeated bending method.

[0015] Furthermore, the different hydrogen damage stages of the hydrogen damage state described in this invention include a reversible incubation stage of hydrogen damage and an irreversible expansion stage of hydrogen damage.

[0016] Furthermore, the test sample in this invention is steel, and there is one hydrogen damage characteristic time point. The different hydrogen damage stages include the reversible incubation stage and the irreversible expansion stage. In the hydrogen damage state prediction model based on the change of the relative nonlinear optimal estimation coefficient, the hydrogen damage stage before the hydrogen damage characteristic time point is the reversible incubation stage, and the hydrogen damage stage after the hydrogen damage characteristic time point is the irreversible expansion stage.

[0017] Furthermore, the algorithm for calculating the optimal estimate described in this invention includes the Kalman filtering method.

[0018] Furthermore, the specific method for determining the hydrogen damage state of the test sample in step S6 of the present invention, based on the real-time curve of the relative nonlinear optimal estimation coefficient of the test sample established in step S5, is as follows: sampling begins when the test sample is first put into use, and the real-time curve of the relative nonlinear optimal estimation coefficient of the test sample is plotted. If the change in the relative nonlinear optimal estimation coefficient is within 5% of the initial value before hydrogen filling, the hydrogen damage to the material can be ignored; when the change in the relative nonlinear optimal estimation coefficient is greater than 5% of the initial value before hydrogen filling, hydrogen damage to the material is determined. The hydrogen damage state of the test sample is determined based on the relative nonlinear optimal estimation coefficient of the sample sampling points and the hydrogen damage state prediction model based on the change in the relative nonlinear optimal estimation coefficient.

[0019] Furthermore, in step S4 of the present invention, the time points 30 minutes before and after the reference time point are selected as the detection time points, and two detection time points are selected for each reference time point.

[0020] Furthermore, in step S2 of this invention, before electrolytically charging the calibration sample in the hydrogen-charging solution, the non-hydrogen-charging surface of the calibration sample is sealed with non-conductive UV adhesive; the hydrogen-charging solution is 0.5 mol / L sulfuric acid + 0.2 g / L thiourea, and the current density used for electrolytic charging is 10-50 mA / cm². 2 During the electrolytic hydrogen charging process, the calibration sample is placed in a hydrogen charging solution. The nonlinear ultrasonic system collects the nonlinear ultrasonic signal of the calibration sample every 10-60 minutes until the hydrogen charging time reaches 1200 minutes or more and the hydrogen damage in the calibration sample reaches saturation.

[0021] Furthermore, the ultrasonic waves used in the nonlinear ultrasonic system of the present invention are ultrasonic longitudinal waves, transverse waves, or surface waves with a frequency of 2MHz-10MHz.

[0022] The principle of this invention is:

[0023] When metallic materials are exposed to hydrogen, hydrogen damage worsens with increasing exposure time. Users cannot definitively determine whether the degree of hydrogen damage affects the safe use of the test sample. This invention establishes a hydrogen damage state prediction model based on the change of the relative nonlinear optimal estimation coefficient to determine different stages of hydrogen damage, enabling online ultrasonic monitoring and stage determination of the hydrogen damage state of the test sample. The invention first involves charging a calibration sample with hydrogen and periodically detecting the relative nonlinear coefficient of the calibration sample using an ultrasonic nonlinear system, plotting a calibration curve of the relative nonlinear optimal estimation coefficient versus hydrogen charging time. The relative nonlinear coefficient is collected from the moment the test sample is first used. Based on the calibration curve, the degree of hydrogen damage is inferred from the change of the relative nonlinear optimal estimation coefficient. A reference time point is then found using the nth derivative of the calibration curve, i.e., the point where a qualitative change in the degree of hydrogen damage occurs based on the degree of change in the calibration curve. Finally, a detection time point is determined, and the degree of hydrogen damage of the calibration sample after the hydrogen charging time at the detection time point is measured, thereby determining different stages of the hydrogen damage state and establishing a hydrogen damage state prediction model based on the change of the relative nonlinear optimal estimation coefficient. Based on the hydrogen damage state prediction model based on the relative nonlinear optimal estimation coefficient change, the ultrasonic signal of the test sample is collected and processed when the test sample is put into use, thereby realizing the ultrasonic online monitoring of the hydrogen damage state of the test sample.

[0024] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0025] I. The detection resolution of traditional linear ultrasonic assessment of hydrogen damage is limited by the wavelength of sound waves, and it is not very sensitive to microcracks much smaller than the wavelength. To address the limitations of traditional linear ultrasonic assessment of hydrogen damage, this invention proposes a method for measuring hydrogen damage based on nonlinear ultrasonics. This method is more sensitive to small hydrogen damage in metals and early hydrogen damage, and has higher detection accuracy.

[0026] Second, this invention acquires the nonlinear signal of the test sample in real time and obtains the relative nonlinear coefficient, calculates the optimal estimation coefficient of the relative nonlinearity, and obtains the hydrogen damage state of the test sample based on the hydrogen damage state prediction model of the change of the optimal estimation coefficient of the relative nonlinearity. It provides an intuitive nonlinear characterization of hydrogen damage in metallic materials and enables monitoring of hydrogen damage state of key components of large equipment.

[0027] Third, the hydrogen damage measurement method of this invention can obtain the hydrogen damage state of the tested metal without the need for destructive assessment of the material. The detection process is low-cost and has the advantages of fast detection, intuitiveness and high accuracy.

[0028] The present invention will be further described in detail below with reference to specific embodiments and accompanying drawings. Attached Figure Description

[0029] Figure 1 The calibration curves of the relative nonlinear optimal estimation coefficient versus hydrogen charging time are plotted for embodiments of the present invention.

[0030] Figure 2 The first derivative reference curve of the calibration curve is plotted for an embodiment of the present invention.

[0031] Figure 3 The image shows the scanning electron microscope (SEM) morphology of the calibration sample at the detection time point T1, as described in this embodiment of the invention.

[0032] Figure 4 The image shows the scanning electron microscope (SEM) morphology of the calibration sample at the detection time point T2 in this embodiment of the invention.

[0033] Figure 5 The present invention establishes a hydrogen damage state prediction model based on the variation of the optimal estimation coefficients with relative nonlinearity. Detailed Implementation

[0034] A method for determining the hydrogen damage state of materials based on online ultrasonic monitoring, comprising the following steps:

[0035] S1. Prepare a sample of the same material and processing technology as the sample to be tested as a calibration sample.

[0036] S2. An ultrasonic transducer is placed on the calibration sample. The calibration sample is placed in a hydrogen charging solution for electrolytic hydrogen charging. The nonlinear ultrasonic system is used to collect the nonlinear ultrasonic signal of the calibration sample after the corresponding hydrogen charging time at regular intervals. The collected nonlinear ultrasonic signal is processed to obtain the relative nonlinear coefficients for different hydrogen charging times, forming a set of relative nonlinear coefficients.

[0037] S3. Perform optimal estimation calculation on the set of relative nonlinear coefficients to obtain the set of optimal estimation coefficients for relative nonlinearity, and plot the calibration curve of the optimal estimation coefficients for relative nonlinearity versus hydrogen charging time and the reference curve of the nth derivative of the calibration curve.

[0038] S4. Select reference time points based on the changes in the reference curve of the nth derivative of the calibration curve. Let p be the number of reference time points. Select time points 10-60 min before and after the reference time points as detection time points. Select q detection time points corresponding to each reference time point, for a total of m detection time points, where m = p * q. Select m calibration samples and place them in a hydrogen charging solution for electrolytic hydrogen charging. The hydrogen charging time is the same as the m detection time points. Then, detect the actual hydrogen damage degree of the m calibration samples to determine the relationship between hydrogen charging time and hydrogen damage state.

[0039] S5. Based on the calibration curve of the relative nonlinear optimal estimation coefficient and hydrogen charging time plotted in step S3 and the relationship between hydrogen charging time and hydrogen damage state determined in step S4, establish a hydrogen damage state prediction model based on the change of the relative nonlinear optimal estimation coefficient.

[0040] S6. Determine the hydrogen damage state of the test sample: Arrange an ultrasonic transducer with the same frequency as in S2 on the test sample. Use a nonlinear ultrasonic system to collect the nonlinear ultrasonic signal of the test sample at regular intervals, and perform signal processing on the nonlinear ultrasonic signal to obtain the relative nonlinear coefficient of the nonlinear ultrasonic signal of the test sample at each sampling time point. Calculate the optimal estimation coefficient of the relative nonlinearity at each sampling time point through optimal estimation, and plot the real-time curve of the optimal estimation coefficient of the relative nonlinearity of the test sample with time as the horizontal axis and the optimal estimation coefficient of the relative nonlinearity as the vertical axis. Based on the hydrogen damage state prediction model based on the change of the optimal estimation coefficient of the relative nonlinearity established in step S5 and the real-time curve of the optimal estimation coefficient of the relative nonlinearity of the test sample, determine the hydrogen damage state of the test sample.

[0041] The signal processing methods for nonlinear ultrasound signals in steps S2 and S6 include Fourier transform.

[0042] Preferably, the specific method for selecting the reference time point based on the change of the nth derivative reference curve of the calibration curve is as follows: the hydrogen charging time point where the derivative of the first derivative reference curve of the calibration curve is zero and the derivative of the second derivative reference curve of the calibration curve is not zero is taken as the reference time point.

[0043] More preferably, the specific operation of detecting the actual hydrogen damage degree of m calibration samples and determining the relationship between hydrogen charging time and hydrogen damage state is as follows: Detect the actual hydrogen damage degree of m calibration samples, and record the reference time points corresponding to the detection time points with large changes in hydrogen damage degree as hydrogen damage characteristic time points; using the hydrogen damage characteristic time points as dividing points, divide the hydrogen damage state into different hydrogen damage stages, and determine the hydrogen damage stages corresponding to different hydrogen charging times, which is the relationship between hydrogen charging time and hydrogen damage state.

[0044] The criteria for significant changes in hydrogen damage are set by the sample material and the sample's usage environment. For example, for steel materials, the criteria for significant changes in hydrogen damage is to examine the sample cross-section under a microscope at 200x magnification to see if the difference in cracks can be detected. For some materials, the presence or absence of hydrogen embrittlement can also be used as the criteria for significant changes in hydrogen damage. The detection method for hydrogen embrittlement can directly adopt the national standard method, such as the closed bending method or the repeated bending method.

[0045] Preferably, the different hydrogen damage stages of the hydrogen damage state include a reversible incubation stage and an irreversible expansion stage of hydrogen damage.

[0046] Preferably, the algorithm for calculating the optimal estimate includes the Kalman filter method.

[0047] Preferably, the specific method for determining the hydrogen damage state of the test sample in step S6 based on the real-time curve of the relative nonlinear optimal estimation coefficient of the test sample established in step S5 is as follows: starting from the moment the test sample is put into use, a real-time curve of the relative nonlinear optimal estimation coefficient of the test sample is plotted. If the change in the relative nonlinear optimal estimation coefficient is within 5% of the initial value before hydrogen filling, the hydrogen damage to the material can be ignored; when the change in the relative nonlinear optimal estimation coefficient is greater than 5% of the initial value before hydrogen filling, it is determined that the material has hydrogen damage. The hydrogen damage state of the test sample is determined based on the relative nonlinear optimal estimation coefficient of the sample sampling points and the hydrogen damage state prediction model based on the change in the relative nonlinear optimal estimation coefficient.

[0048] Preferably, in step S4, the time points 30 minutes before and after the reference time point are selected as the detection time points, and two detection time points are selected for each reference time point.

[0049] Preferably, in step S2, before placing the calibration sample in the hydrogen charging solution for electrolytic hydrogen charging, the non-hydrogen-charging surface of the calibration sample is sealed with non-conductive UV adhesive; the hydrogen charging solution is 0.5 mol / L sulfuric acid + 0.2 g / L thiourea, and the current density used for electrolytic hydrogen charging is 10-50 mA / cm2; during the process of placing the calibration sample in the hydrogen charging solution for electrolytic hydrogen charging, the nonlinear ultrasonic system collects the nonlinear ultrasonic signal of the calibration sample every 10-60 minutes until the hydrogen charging time reaches 1200 minutes or more, and the hydrogen damage in the calibration sample reaches saturation.

[0050] Preferably, the ultrasonic waves used by the nonlinear ultrasonic system are longitudinal waves, transverse waves, or surface waves with a frequency of 2MHz-10MHz.

[0051] When the test sample is steel, there is one characteristic time point for hydrogen damage. Different hydrogen damage stages include the reversible incubation stage and the irreversible expansion stage. In the hydrogen damage state prediction model based on the change of the relative nonlinear optimal estimation coefficient, the hydrogen damage stage before the characteristic time point is the reversible incubation stage, and the hydrogen damage stage after the characteristic time point is the irreversible expansion stage.

[0052] Example

[0053] Taking low-carbon steel as the test sample as an example, the hydrogen damage state of the low-carbon steel test sample is determined by the above-mentioned method for determining the hydrogen damage state of materials based on ultrasonic online monitoring. Figure 1The calibration curve of the relative nonlinear optimal estimation coefficient versus hydrogen charging time is plotted for step S3 of this embodiment. Figure 2 The calibration curve is a first-order derivative reference curve plotted in step S3 of the embodiment. The point where the derivative of the first-order derivative reference curve is zero is taken as the reference time point, denoted as C in the figure. Then... Figure 1 Find point C, and use points 30 minutes before and after point C as detection time points. Figure 1 The two calibration samples are denoted as T1 and T2. Two calibration samples are placed in a hydrogen-filled solution for electrolytic hydrogen filling. The hydrogen filling times are T1 and T2, respectively. Then, the cross sections of the two calibration samples are magnified 200 times under a microscope to check for the presence of cracks along the grain boundaries. In other words, the standard for a large change in the degree of hydrogen damage in this case is whether the difference in the degree of cracks can be detected by magnifying the sample cross section under a microscope 200 times. Figure 3 and Figure 4 The images show the scanning electron microscope (SEM) morphology of the calibration samples at detection time points T1 and T2, respectively, after hydrogen charging. It can be seen from the images that the calibration sample at detection time point T1 has no cracks yet, belonging to the reversible incubation stage of hydrogen damage; the calibration sample at detection time point T2 has already developed cracks, belonging to the irreversible propagation stage of hydrogen damage. The reference time point C between these two detection time points is denoted as the characteristic time point of hydrogen damage. Figure 5 The hydrogen damage state prediction model established for this embodiment is based on the change of the relative nonlinear optimal estimation coefficient. The change of the relative nonlinear optimal estimation coefficient is within 5% of the initial value when hydrogen is not charged. This may be due to detection error. Even if there is no detection error, the hydrogen damage in this stage can be ignored. The hydrogen damage in stage AB in the figure can be ignored. Stage BC is the incubation stage of reversible hydrogen damage. After point C, it is the expansion stage of irreversible hydrogen damage.

Claims

1. A method for determining the hydrogen damage state of materials based on online ultrasonic monitoring, comprising the following steps: S1. Prepare a sample of the same material and processing technology as the sample to be tested as a calibration sample. S2. Place the ultrasonic transducer on the calibration sample, place the calibration sample in the hydrogen charging solution for electrolytic hydrogen charging, and use a nonlinear ultrasonic system to collect the nonlinear ultrasonic signal of the calibration sample after the corresponding hydrogen charging time at regular intervals. Then, perform signal processing on the collected nonlinear ultrasonic signal to obtain the relative nonlinear coefficients for different hydrogen charging times, and form a set of relative nonlinear coefficients. S3. Perform optimal estimation calculation on the set of relative nonlinear coefficients to obtain the set of optimal estimation coefficients for relative nonlinearity, and plot the calibration curve of the optimal estimation coefficients for relative nonlinearity versus hydrogen charging time and the basis curve of the nth derivative of the calibration curve. S4. Select reference time points based on the changes in the reference curve of the nth derivative of the calibration curve. Let p be the number of reference time points. Select time points 10-60 min before and after the reference time points as detection time points. Select q detection time points corresponding to each reference time point, for a total of m detection time points, where m = p * q. Select m calibration samples and place them in a hydrogen charging solution for electrolytic hydrogen charging. The hydrogen charging time is the same as the m detection time points. Then, detect the actual hydrogen damage degree of the m calibration samples to determine the relationship between hydrogen charging time and hydrogen damage state. S5. Based on the calibration curve of the relative nonlinear optimal estimation coefficient and hydrogen charging time plotted in step S3 and the relationship between hydrogen charging time and hydrogen damage state determined in step S4, establish a hydrogen damage state prediction model based on the change of the relative nonlinear optimal estimation coefficient. S6. Determine the hydrogen damage state of the test sample: Arrange an ultrasonic transducer with the same frequency as in S2 on the test sample. Use a nonlinear ultrasonic system to collect the nonlinear ultrasonic signal of the test sample at regular intervals, and perform signal processing on the nonlinear ultrasonic signal to obtain the relative nonlinear coefficient of the nonlinear ultrasonic signal of the test sample at each sampling time point. Calculate the optimal estimation coefficient of the relative nonlinearity at each sampling time point through optimal estimation, and plot the real-time curve of the optimal estimation coefficient of the relative nonlinearity of the test sample with time as the horizontal axis and the optimal estimation coefficient of the relative nonlinearity as the vertical axis. Based on the hydrogen damage state prediction model based on the change of the optimal estimation coefficient of the relative nonlinearity established in step S5 and the real-time curve of the optimal estimation coefficient of the relative nonlinearity of the test sample, determine the hydrogen damage state of the test sample.

2. The method for determining the hydrogen damage state of materials based on online ultrasonic monitoring according to claim 1, characterized in that: The specific method for selecting the reference time point based on the change of the nth derivative reference curve of the calibration curve is as follows: the hydrogen charging time point where the derivative of the first derivative reference curve of the calibration curve is zero and the derivative of the second derivative reference curve of the calibration curve is not zero is taken as the reference time point.

3. A method for determining the hydrogen damage state of materials based on online ultrasonic monitoring according to claim 1 or 2, characterized in that: The specific operation of detecting the actual hydrogen damage degree of m calibration samples and determining the relationship between hydrogen charging time and hydrogen damage state is as follows: detect the actual hydrogen damage degree of m calibration samples, and record the reference time points corresponding to the detection time points with large changes in hydrogen damage degree as hydrogen damage characteristic time points. Using the characteristic time point of hydrogen damage as the dividing point, the hydrogen damage state is divided into different hydrogen damage stages, and the hydrogen damage stage corresponding to different hydrogen charging time is determined, which is the relationship between hydrogen charging time and hydrogen damage state.

4. The method for determining the hydrogen damage state of materials based on online ultrasonic monitoring according to claim 3, characterized in that: The different hydrogen damage stages of the hydrogen damage state include the reversible incubation stage of hydrogen damage and the irreversible expansion stage of hydrogen damage.

5. The method for determining the hydrogen damage state of materials based on online ultrasonic monitoring according to claim 3, characterized in that: The test sample is steel, and there is one hydrogen damage characteristic time point. The different hydrogen damage stages include the reversible incubation stage and the irreversible expansion stage. In the hydrogen damage state prediction model based on the change of the relative nonlinear optimal estimation coefficient, the hydrogen damage stage before the hydrogen damage characteristic time point is the reversible incubation stage, and the hydrogen damage stage after the hydrogen damage characteristic time point is the irreversible expansion stage.

6. The method for determining the hydrogen damage state of materials based on online ultrasonic monitoring according to claim 1, characterized in that: The algorithm for calculating the optimal estimate includes the Kalman filter method.

7. The method for determining the hydrogen damage state of materials based on online ultrasonic monitoring according to claim 1, characterized in that: In step S6, the specific method for determining the hydrogen damage state of the test sample based on the real-time curve of the relative nonlinear optimal estimation coefficient established in step S5 and the test sample is as follows: starting from the moment the test sample is put into use, the real-time curve of the relative nonlinear optimal estimation coefficient of the test sample is plotted. If the change in the relative nonlinear optimal estimation coefficient is within 5% of the initial value before hydrogen filling, the hydrogen damage to the material can be ignored; when the change in the relative nonlinear optimal estimation coefficient is greater than 5% of the initial value before hydrogen filling, the material is determined to have hydrogen damage. The hydrogen damage state of the test sample is determined based on the relative nonlinear optimal estimation coefficient of the sample sampling points and the hydrogen damage state prediction model based on the change in the relative nonlinear optimal estimation coefficient.

8. The method for determining the hydrogen damage state of materials based on online ultrasonic monitoring according to claim 1, characterized in that: In step S4, the time points 30 minutes before and after the reference time point are selected as the detection time points, and two detection time points are selected for each reference time point.

9. The method for determining the hydrogen damage state of materials based on online ultrasonic monitoring according to claim 1, characterized in that: In step S2, before electrolytically charging the calibration sample in the hydrogen-charging solution, the non-hydrogen-charging surface of the calibration sample is sealed with non-conductive UV adhesive. The hydrogen-charging solution is 0.5 mol / L sulfuric acid + 0.2 g / L thiourea, and the current density used for electrolytic charging is 10-50 mA / cm². 2 During the electrolytic hydrogen charging process, the calibration sample is placed in a hydrogen charging solution. The nonlinear ultrasonic system collects the nonlinear ultrasonic signal of the calibration sample every 10-60 minutes until the hydrogen charging time reaches 1200 minutes or more and the hydrogen damage in the calibration sample reaches saturation.

10. The method for determining the hydrogen damage state of materials based on online ultrasonic monitoring according to claim 1, characterized in that: The nonlinear ultrasonic system uses ultrasonic longitudinal waves, transverse waves, or surface waves with a frequency of 2MHz-10MHz.

Images