A method and system for predicting the content of non-metallic inclusions in parts.

Abstract

A method and system for predicting the content of non-metallic inclusions in a part is disclosed, relating to a method and system for detecting inclusion content. The aim is to overcome the problem that existing methods for detecting non-metallic inclusions in bearing rings cause wear on the bearing rings. The method comprises the following steps: Step 1: Obtaining the transverse wave velocity and longitudinal wave velocity of ultrasonic waves passing through the part to be tested; Step 2: Inputting both the transverse wave velocity and longitudinal wave velocity into a non-metallic inclusion content prediction model to predict the predicted value of the non-metallic inclusion content of the part.

Description

Technical Field

[0001] This invention relates to a method and system for detecting impurity content. Background Technology

[0002] Bearings are a crucial component of modern industrial operations, and the bearing manufacturing industry is experiencing rapid growth. Bearings are widely used in various sectors, including aerospace, electromechanical, and heavy industry. Therefore, quality inspection of bearing rings is of paramount importance.

[0003] Potential quality problems with bearing races include internal defects such as cracks, shrinkage cavities, and non-metallic inclusions. Among these, non-metallic inclusions are a significant cause of fatigue cracks within bearing races. Their presence leads to uneven microstructure, disrupts material continuity, reduces the impact toughness and fatigue strength of the bearing races, and seriously jeopardizes bearing safety. Therefore, it is necessary to test the content of non-metallic inclusions inside bearing races to ensure that it remains within a safe range.

[0004] Non-metallic inclusions in steel are mainly classified into five categories: sulfides, alumina, silicates, spherical oxides, and single spherical particles. Currently, the main methods for evaluating the content of non-metallic inclusions in steel include: worst field of view evaluation method, field-by-field evaluation method, and steel cleanliness method. All of these methods involve preparing standard samples and using an optical microscope to detect and evaluate non-metallic inclusions to determine whether the content of non-metallic inclusions inside the material meets the standards.

[0005] However, the above-mentioned metallographic examination of bearing rings to detect non-metallic inclusions is a destructive experiment, which will cause damage to the bearing rings. The selection and preparation of samples also require a lot of time and manpower. Summary of the Invention

[0006] The purpose of this invention is to overcome the problem that existing methods for detecting non-metallic inclusions in bearing races cause wear to the bearing races, and to provide a method and system for predicting the content of non-metallic inclusions in parts.

[0007] This invention provides a method for predicting the content of non-metallic inclusions in a part, comprising the following steps:

[0008] Step 1: Obtain the transverse wave velocity and longitudinal wave velocity of the ultrasonic wave as it passes through the part under test.

[0009] Step 2: Input both the transverse wave velocity and the longitudinal wave velocity into the non-metallic inclusion content prediction model to obtain the predicted value of the non-metallic inclusion content of the part.

[0010] Furthermore, the process of establishing the prediction model for non-metallic inclusion content is as follows:

[0011] Step 2: 1. Collect the actual values ​​of longitudinal wave velocity, transverse wave velocity, and non-metallic inclusion content of multiple parts of the same specifications in advance.

[0012] Step 22: Using the actual value of non-metallic inclusion content as the response variable and the longitudinal wave velocity and transverse wave velocity as explanatory variables, construct a Bayesian regression model with model parameters.

[0013] The model parameters include the longitudinal wave velocity coefficient, the transverse wave velocity coefficient, and the model intercept;

[0014] Steps 2 and 3: Treat the model parameters as a normal distribution and calculate the values ​​of the model parameters when the posterior distribution of the model parameters is maximized;

[0015] Step 24: Substitute the values ​​of the model parameters into the Bayesian regression model to complete the construction of the non-metallic inclusion content prediction model.

[0016] Furthermore, the prediction model for non-metallic inclusion content is as follows:

[0017]

[0018] Where y is the predicted value of non-metallic inclusion content, and c L For the longitudinal wave speed, c S For transverse wave sound speed, β L β is the longitudinal wave velocity coefficient. S α is the transverse wave velocity coefficient, and α is the model intercept.

[0019] Furthermore, the specific process of step one is as follows:

[0020] Step 11: Place the part to be tested into the liquid;

[0021] Step 1 and 2: Emit ultrasonic waves into the part to be tested in the liquid, so that the ultrasonic waves are transverse or longitudinal waves when they are incident on the part to be tested, and record the ultrasonic wave emission time at the same time.

[0022] Step 13: Receive transverse or longitudinal waves and record the transverse wave reception time and longitudinal wave reception time respectively; calculate the transverse wave velocity using the propagation path length of the transverse wave in the part under test, the transverse wave reception time, and the ultrasonic wave emission time; calculate the longitudinal wave velocity using the propagation path length of the longitudinal wave in the part under test, the longitudinal wave reception time, and the ultrasonic wave emission time.

[0023] Furthermore, step one also includes:

[0024] Step 14: Rotate the part to be tested and repeat steps 11 to 13 to calculate the transverse wave velocity and longitudinal wave velocity of the part to be tested at different positions, and calculate the average value of the transverse wave velocity and longitudinal wave velocity respectively.

[0025] This invention also provides a system for predicting the content of non-metallic inclusions in parts, including...

[0026] The sound velocity acquisition module is used to acquire the transverse wave velocity and longitudinal wave velocity of ultrasonic waves when they pass through the part under test.

[0027] The non-metallic inclusion content prediction model is used to input both transverse wave velocity and longitudinal wave velocity into the non-metallic inclusion content prediction model to predict the predicted value of non-metallic inclusion content in the part.

[0028] Furthermore, the modules for establishing a prediction model for non-metallic inclusion content include:

[0029] The data acquisition module is used to pre-collect the actual values ​​of longitudinal wave velocity, transverse wave velocity, and non-metallic inclusion content of N parts with the same specifications.

[0030] The regression model building module is used to construct a Bayesian regression model with model parameters, using the actual value of non-metallic inclusion content as the response variable and the longitudinal wave velocity and transverse wave velocity as explanatory variables.

[0031] The model parameters include the longitudinal wave velocity coefficient, the transverse wave velocity coefficient, and the model intercept;

[0032] The model parameter calculation module is used to treat the model parameters as a normal distribution and calculate the values ​​of the model parameters when the posterior distribution of the model parameters is maximized.

[0033] The parameter substitution module is used to substitute the values ​​of model parameters into the Bayesian regression model to complete the construction of the non-metallic inclusion content prediction model.

[0034] Furthermore, the prediction model for non-metallic inclusion content is as follows:

[0035]

[0036] Where y is the predicted value of non-metallic inclusion content, and c L For the longitudinal wave speed, c S For transverse wave sound speed, β L β is the longitudinal wave velocity coefficient. S α is the transverse wave velocity coefficient, and α is the model intercept.

[0037] Furthermore, the sound velocity acquisition module includes:

[0038] A liquid tank is used to immerse the part to be tested in a liquid.

[0039] An ultrasonic probe is used to emit ultrasonic waves into a part under test in a liquid, so that the ultrasonic waves are transverse or longitudinal waves when they are incident on the part under test.

[0040] A receiving probe, used to receive transverse or longitudinal waves;

[0041] The processing unit is used to control the ultrasonic probe to turn on and record the ultrasonic wave emission time; and to receive transverse waves and longitudinal waves through the receiving probe and record the transverse wave reception time and longitudinal wave reception time respectively; to calculate the transverse wave velocity using the propagation path length of the transverse wave in the part under test, the transverse wave reception time, and the ultrasonic wave emission time; and to calculate the longitudinal wave velocity using the propagation path length of the longitudinal wave in the part under test, the longitudinal wave reception time, and the ultrasonic wave emission time.

[0042] Furthermore,

[0043] The sound velocity acquisition module also includes a rotary table, which is used to rotate the part to be measured;

[0044] The processing unit is also used to calculate the transverse wave velocity and longitudinal wave velocity of the part under test at different positions, and to calculate the average value of the transverse wave velocity and longitudinal wave velocity respectively.

[0045] The beneficial effects of this invention are:

[0046] 1) Ultrasonic testing does not damage the performance of bearing rings, reducing material waste. Compared with non-destructive testing methods such as X-rays and eddy current testing, ultrasonic testing does not produce radiation signals that affect the human body, and it overcomes the limitation of eddy current testing being only applicable to metallic materials, making it extremely effective when measuring the internal structure of materials;

[0047] 2) Compared to methods such as the worst-case field-of-view assessment method, the statistical weighted average method, and the steel purity method, this method avoids interference from human factors, reduces labor costs, and improves testing efficiency and automation.

[0048] 3) A predictive model for the non-metallic inclusion content of bearing rings is constructed based on known sample data. When sufficient sample data is available, this model can effectively and accurately detect the content of non-metallic inclusions inside the bearing. This method can not only be used for detecting the non-metallic inclusion content of bearing rings using ultrasonic testing, but can also be extended to other testing fields. Attached Figure Description

[0049] Figure 1 This is a schematic diagram of the sound velocity acquisition module in this invention.

[0050] Figure 2 This is a schematic diagram illustrating the measurement principle of the sound velocity acquisition module in this invention. Detailed Implementation

[0051] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0052] It should be noted that, unless otherwise specified, the embodiments and features described in the present invention can be combined with each other.

[0053] The present invention will be further described below with reference to the accompanying drawings and specific embodiments, but this is not intended to limit the scope of the invention. Specific Implementation Method 1

[0055] The method for predicting the content of non-metallic inclusions in a part according to this embodiment includes the following steps:

[0056] Step 1: Obtain the transverse wave velocity and longitudinal wave velocity of the ultrasonic wave as it passes through the part to be tested (9).

[0057] Step 2: Input both the transverse wave velocity and the longitudinal wave velocity into the non-metallic inclusion content prediction model to obtain the predicted value of the non-metallic inclusion content of the part. Specific Implementation Method Two

[0059] This embodiment is a further explanation of Embodiment 1. In this embodiment, the process of establishing the prediction model for non-metallic inclusion content is as follows:

[0060] Step 2: 1. Collect the actual values ​​of longitudinal wave velocity, transverse wave velocity, and non-metallic inclusion content of multiple parts of the same specifications in advance.

[0061] Step 22: Using the actual value of non-metallic inclusion content as the response variable and the longitudinal wave velocity and transverse wave velocity as explanatory variables, construct a Bayesian regression model with model parameters.

[0062] The model parameters include the longitudinal wave velocity coefficient, the transverse wave velocity coefficient, and the model intercept;

[0063] Steps 2 and 3: Treat the model parameters as a normal distribution and calculate the values ​​of the model parameters when the posterior distribution of the model parameters is maximized;

[0064] Step 24: Substitute the values ​​of the model parameters into the Bayesian regression model to complete the construction of the non-metallic inclusion content prediction model. Specific Implementation Method 3

[0066] This embodiment is a further explanation of embodiment two. In this embodiment, the prediction model for non-metallic inclusion content is as follows:

[0067]

[0068] Where y is the predicted value of non-metallic inclusion content, and c L For the longitudinal wave speed, c S For transverse wave sound speed, β L β is the longitudinal wave velocity coefficient. S α is the transverse wave velocity coefficient, and α is the model intercept.

[0069] The other technical features of this embodiment are exactly the same as those of Embodiment 2.

[0070] Specifically, N standard bearing rings of the same specifications are selected, and N samples are first made for metallographic testing to obtain N sets of non-metallic inclusion contents. Then, ultrasonic testing is performed to obtain N sets of longitudinal and transverse wave velocities of the bearing rings.

[0071] Adjusting the angle of the adjustable wedge controls the incident angle of the ultrasonic wave, ensuring that only longitudinal or transverse waves are generated inside the bearing ring. The detection ultrasonic wave is emitted by an ultrasonic probe and incident perpendicularly into the bearing ring. The transverse wave is obtained by adjusting the ultrasonic incident angle θ to between the first critical angle θ1 and the second critical angle θ2 at the water-steel interface (water and bearing ring). When the incident wave is a longitudinal wave, the longitudinal wave incident angle that achieves a refraction angle of 90° is the first critical angle θ1, and the longitudinal wave incident angle that achieves a refraction angle of 90° is the second critical angle θ2. Therefore, when the longitudinal wave incident angle θ is between θ1 and θ2, only transverse waves exist inside the bearing ring, and no longitudinal waves. The values ​​of θ1 and θ2 are obtained according to Snell's Law: θ1 = arcsin(c L1 / c L2 ),θ2=arcsin(c L1 / c S2 ), where c L1 and c L2 c represents the longitudinal wave propagation speed in water and steel, respectively. S2 denoted as , where is the propagation speed of transverse waves in steel.

[0072] The calculation principle for the longitudinal and transverse wave velocities inside the bearing race is based on the ratio of the ultrasonic wave propagation path to its propagation time. By setting up corresponding receiving probes on the bearing race wall and measuring the propagation distance and time of the longitudinal and transverse waves inside the bearing race, the wave velocities can be obtained. By rotating the bearing race on a rotary table, the propagation path of the sound waves within the bearing race is changed. The above steps are repeated to complete multiple sets of wave velocity measurements. The average value of both the longitudinal and transverse wave velocities is calculated, and this average value, along with the non-metallic content results from the metallographic examination of the corresponding bearing race, forms a three-dimensional data matrix C. ij(i = 1···N, j = 1, 2, 3), including non-metallic inclusion content (metallographic detection), longitudinal wave velocity, and transverse wave velocity, a total of N sets of data, which together form a data matrix C of non-metallic inclusion content-ultrasonic characteristic information. ij enter.

[0073] The model learning process is as follows:

[0074] Based on the above-obtained non-metallic inclusion content-ultrasonic information data matrix C ij A model was constructed to establish the relationship between the content of non-metallic inclusions in the ring and the longitudinal and transverse wave velocities, where the non-metallic inclusion content y is the response variable and the longitudinal wave velocity c is the response variable. L With transverse wave speed of sound c S As an explanatory variable.

[0075] The model is based on the existing linear regression model, and a Bayesian statistical method is introduced on this basis. That is, the content of non-metallic inclusions in N bearing rings is regarded as a Gaussian distribution, denoted as: Y ~ N(μ,σ) 2 If ), then for the i-th bearing ring, we have:

[0076] The parameters that need to be calculated in the model are {β} L β S α}, where β L and β S The longitudinal wave speeds c are respectively Li and transverse wave speed of sound c Si The coefficients, where α is the model intercept.

[0077] Assuming that the model parameters to be estimated are all normally distributed, the likelihood function for the non-metallic inclusion content of the i-th bearing ring is: Where σ0 is the standard deviation of the likelihood function, and the likelihood functions of all bearing rings are obtained by multiplying the previous equations, the posterior distribution of the model parameters is: P(μ(β) L ,β S ,α)|y)=∫P(y|(β L ,β S ,α))P(β L ,β S ,α)d(β L ,β S ,α).

[0078] The model parameters {β} are calculated based on the principle of maximum posterior distribution when the posterior probability is maximized. L β S α}, according to the obtained model parameters A complete model was constructed to establish the relationship between the non-metallic inclusion content of bearing rings and the longitudinal and transverse wave velocities (non-metallic inclusion content prediction model).

[0079] The prediction and estimation process is as follows:

[0080] The bearing rings to be tested are then tested using the method of this embodiment. The longitudinal and transverse wave velocities of the bearing rings are obtained as known explanatory variables. The predicted value of the non-metallic inclusion content of the bearing rings to be tested is used as the response variable to be predicted. The non-metallic inclusion content prediction model constructed above is used to predict and estimate the content of non-metallic inclusions in the bearing rings to be tested. Specific Implementation Method Four

[0082] This embodiment is a further explanation of embodiment one, two, or three. In this embodiment, the specific process of step one is as follows:

[0083] Step 11: Place the part to be tested, 9, into the liquid;

[0084] Step 1 and 2: Emit ultrasonic waves to the part 9 to be tested in the liquid, so that the ultrasonic waves are transverse or longitudinal waves when they are incident on the part 9 to be tested, and record the ultrasonic wave emission time at the same time.

[0085] Step 13: Receive transverse or longitudinal waves and record the transverse wave reception time and longitudinal wave reception time respectively; calculate the transverse wave velocity using the propagation path length of the transverse wave in the part under test 9, the transverse wave reception time, and the ultrasonic wave emission time; calculate the longitudinal wave velocity using the propagation path length of the longitudinal wave in the part under test 9, the longitudinal wave reception time, and the ultrasonic wave emission time.

[0086] The other technical features of this embodiment are exactly the same as those of Embodiment 1, 2 or 3. Detailed Implementation Method Five

[0088] This embodiment is a further explanation of embodiment four. In this embodiment, step one further includes:

[0089] Step 14: Rotate the part to be tested 9 and repeat steps 11 to 13 to calculate the transverse wave velocity and longitudinal wave velocity of the part to be tested 9 at different positions, and calculate the average value of the transverse wave velocity and longitudinal wave velocity respectively.

[0090] The other technical features of this embodiment are exactly the same as those of embodiment four. Specific Implementation Method Six

[0092] This embodiment provides a system for predicting the content of non-metallic inclusions in a part, including...

[0093] The sound velocity acquisition module is used to acquire the transverse wave velocity and longitudinal wave velocity of the ultrasonic wave as it passes through the part under test 9.

[0094] The non-metallic inclusion content prediction model is used to input both transverse wave velocity and longitudinal wave velocity into the non-metallic inclusion content prediction model to predict the predicted value of non-metallic inclusion content in the part. Detailed Implementation Method Seven

[0096] This embodiment is a further explanation of embodiment six. In this embodiment, the module for establishing the prediction model for non-metallic inclusion content includes:

[0097] The data acquisition module is used to pre-collect the actual values ​​of longitudinal wave velocity, transverse wave velocity, and non-metallic inclusion content of N parts with the same specifications.

[0098] The regression model building module is used to construct a Bayesian regression model with model parameters, using the actual value of non-metallic inclusion content as the response variable and the longitudinal wave velocity and transverse wave velocity as explanatory variables.

[0099] The model parameters include the longitudinal wave velocity coefficient, the transverse wave velocity coefficient, and the model intercept;

[0100] The model parameter calculation module is used to treat the model parameters as a normal distribution and calculate the values ​​of the model parameters when the posterior distribution of the model parameters is maximized.

[0101] The parameter substitution module is used to substitute the values ​​of model parameters into the Bayesian regression model to complete the construction of the non-metallic inclusion content prediction model.

[0102] The other technical features of this embodiment are exactly the same as those of Embodiment Six. Detailed Implementation Method Eight

[0104] This embodiment is a further explanation of embodiment seven. In this embodiment, the prediction model for non-metallic inclusion content is as follows:

[0105]

[0106] Where y is the predicted value of non-metallic inclusion content, and c L For the longitudinal wave speed, c S For transverse wave sound speed, β L β is the longitudinal wave velocity coefficient. S α is the transverse wave velocity coefficient, and α is the model intercept.

[0107] The other technical features of this embodiment are exactly the same as those of Embodiment Seven. Detailed Implementation Method Nine

[0109] This embodiment is a further description of embodiment six, seven or eight. In this embodiment, liquid pool 1 is used to place the part to be tested 9 into the liquid.

[0110] Ultrasonic probe 2 is used to emit ultrasonic waves into the part to be tested 9 in the liquid, so that the ultrasonic waves are transverse or longitudinal waves when they are incident on the part to be tested 9.

[0111] Receiver probe 3 is used to receive transverse or longitudinal waves;

[0112] The processing unit 4 is used to control the ultrasonic probe 2 to turn on and record the ultrasonic emission time; and to receive transverse waves and longitudinal waves through the receiving probe 3 and record the transverse wave reception time and longitudinal wave reception time respectively; to calculate the transverse wave velocity by the propagation path length of the transverse wave in the part under test 9, the transverse wave reception time, and the ultrasonic emission time; and to calculate the longitudinal wave velocity by the propagation path length of the longitudinal wave in the part under test 9, the longitudinal wave reception time, and the ultrasonic emission time.

[0113] The other technical features of this embodiment are exactly the same as those of embodiments six, seven, or eight. Detailed Implementation Method Ten

[0115] This embodiment is a further explanation of embodiment nine. In this embodiment, the sound velocity acquisition module also includes a rotary table 5, which is used to rotate the part to be measured 9.

[0116] The processing unit 4 is also used to calculate the transverse wave velocity and longitudinal wave velocity of the part to be tested 9 at different positions, and to calculate the average value of the transverse wave velocity and longitudinal wave velocity respectively.

[0117] The other technical features of this embodiment are exactly the same as those of Embodiment Nine.

[0118] Specifically, such as Figure 1 and Figure 2 As shown, the sound velocity acquisition module includes: a liquid pool 1 (water tank), an ultrasonic probe 2, a receiving probe 3, a processing unit 4 (oscilloscope, computer), a rotary table 5, a pulse generator 6, a clamping device 7, and an adjustable wedge 8. The ultrasonic probe 2 is connected to the pulse generator 6 and fixed in the liquid pool 1 by the adjustable wedge 8 and the clamping device 7. The adjustable wedge 8 allows for free adjustment of the angle of the ultrasonic probe 2. The bearing ring to be tested is placed on the rotary table 5 in the liquid pool 1, allowing for free horizontal rotation. The receiving probe 3 is placed on the ring wall and connected to the oscilloscope to receive the ultrasonic signal generated by the ultrasonic probe 2. The oscilloscope is connected to the computer, which processes the ultrasonic signal transmitted by the oscilloscope, acquires the corresponding ultrasonic information, and completes the detection of the non-metallic inclusion content of the bearing ring based on the corresponding data.

[0119] The measurement process is as follows: the pulse generated by the pulse generator drives the longitudinal wave probe to generate ultrasonic waves. The position of the adjustable wedge is fixed in the clamping device. By adjusting the angle of the adjustable wedge, the ultrasonic probe is controlled to generate only longitudinal or transverse waves inside the bearing ring. At this time, the receiving probe receives the ultrasonic signal, which is displayed on the oscilloscope and transmitted to the computer. Finally, the computer can calculate the longitudinal wave velocity and transverse wave velocity based on the propagation path and time of the sound wave in the bearing ring, and complete the detection calculation of the non-metallic inclusion content of the bearing ring.

[0120] While the invention has been described herein with reference to specific embodiments, it should be understood that these embodiments are merely examples of the principles and applications of the invention. Therefore, it should be understood that many modifications can be made to the exemplary embodiments, and other arrangements can be designed without departing from the spirit and scope of the invention as defined by the appended claims. It should be understood that different dependent claims and features herein can be combined in ways different from those described in the original claims. It is also understood that features described in conjunction with individual embodiments can be used in other embodiments.

Claims

1. A method for predicting the content of non-metallic inclusions in a part, characterized in that, The steps are as follows: Step 1: Obtain the transverse wave velocity and longitudinal wave velocity of the ultrasonic wave as it passes through the part to be tested (9). Step 2: Input both the transverse wave velocity and the longitudinal wave velocity into the non-metallic inclusion content prediction model to obtain the predicted value of the non-metallic inclusion content of the part. The process of establishing the prediction model for non-metallic inclusion content is as follows: Step 2:

1. Collect the actual values ​​of longitudinal wave velocity, transverse wave velocity, and non-metallic inclusion content of multiple parts of the same specifications in advance. Step 22: Using the actual value of non-metallic inclusion content as the response variable and the longitudinal wave velocity and transverse wave velocity as explanatory variables, construct a Bayesian regression model with model parameters. The model parameters include the longitudinal wave velocity coefficient, the transverse wave velocity coefficient, and the model intercept; Steps 2 and 3: Treat the model parameters as a normal distribution and calculate the value of the model parameters when the posterior distribution of the model parameters is maximized; Step 24: Substitute the values ​​of the model parameters into the Bayesian regression model to complete the construction of the non-metallic inclusion content prediction model; The prediction model for non-metallic inclusion content is as follows: ， Where y is the predicted value of non-metallic inclusion content. For longitudinal wave sound speed, For transverse wave sound speed, For the longitudinal wave velocity coefficient, The transverse wave velocity coefficient, This is the model intercept.

2. The method for predicting the content of non-metallic inclusions in a part according to claim 1, characterized in that, The specific process of step one is as follows: Step 11: Place the part to be tested (9) into the liquid; Step 12: Emit ultrasonic waves to the part to be tested (9) in the liquid, so that the ultrasonic waves are transverse or longitudinal waves when they are incident on the part to be tested (9), and record the ultrasonic wave emission time at the same time. Step 13: Receive transverse waves or longitudinal waves and record the transverse wave reception time and longitudinal wave reception time respectively; calculate the transverse wave velocity by the propagation path length of the transverse wave in the part under test (9), the transverse wave reception time, and the ultrasonic wave emission time; calculate the longitudinal wave velocity by the propagation path length of the longitudinal wave in the part under test (9), the longitudinal wave reception time, and the ultrasonic wave emission time.

3. The method for predicting the content of non-metallic inclusions in a part according to claim 2, characterized in that, Step one also includes: Step 14: Rotate the part to be tested (9) and repeat steps 11 to 13 to calculate the transverse wave velocity and longitudinal wave velocity of the part to be tested (9) at different positions, and calculate the average value of the transverse wave velocity and longitudinal wave velocity respectively.

4. A system for predicting the content of non-metallic inclusions in a part, characterized in that, include The sound velocity acquisition module is used to acquire the transverse wave velocity and longitudinal wave velocity of the ultrasonic wave when it passes through the part to be tested (9). The non-metallic inclusion content prediction model is used to input both the transverse wave velocity and the longitudinal wave velocity into the non-metallic inclusion content prediction model to predict the predicted value of the non-metallic inclusion content of the part. The modules for establishing a prediction model for non-metallic inclusion content include: The data acquisition module is used to pre-collect the actual values ​​of longitudinal wave velocity, transverse wave velocity, and non-metallic inclusion content of N parts with the same specifications. The regression model building module is used to construct a Bayesian regression model with model parameters, using the actual value of non-metallic inclusion content as the response variable and the longitudinal wave velocity and transverse wave velocity as explanatory variables. The model parameters include the longitudinal wave velocity coefficient, the transverse wave velocity coefficient, and the model intercept; The model parameter calculation module is used to treat the model parameters as a normal distribution and calculate the value of the model parameters when the posterior distribution of the model parameters is maximized. The parameter substitution module is used to substitute the values ​​of the model parameters into the Bayesian regression model to complete the construction of the non-metallic inclusion content prediction model. The prediction model for non-metallic inclusion content is as follows: ， Where y is the predicted value of non-metallic inclusion content. For longitudinal wave sound speed, For transverse wave sound speed, For the longitudinal wave velocity coefficient, The transverse wave velocity coefficient, This is the model intercept.

5. The system for predicting the content of non-metallic inclusions in a part according to claim 4, characterized in that, The sound velocity acquisition module includes: Liquid pool (1) is used to place the part to be tested (9) into the liquid; An ultrasonic probe (2) is used to emit ultrasonic waves into the part to be tested (9) in the liquid, so that the ultrasonic waves are transverse or longitudinal waves when they are incident on the part to be tested (9). The receiving probe (3) is used to receive transverse or longitudinal waves; The processing unit (4) is used to control the ultrasonic probe (2) to turn on and record the ultrasonic emission time; and to receive transverse waves and longitudinal waves through the receiving probe (3) and record the transverse wave reception time and longitudinal wave reception time respectively; the transverse wave velocity is calculated by the propagation path length of the transverse wave in the part under test (9), the transverse wave reception time and the ultrasonic emission time, and the longitudinal wave velocity is calculated by the propagation path length of the longitudinal wave in the part under test (9), the longitudinal wave reception time and the ultrasonic emission time.

6. The system for predicting the content of non-metallic inclusions in a part according to claim 5, characterized in that, The sound velocity acquisition module also includes a rotary table (5), which is used to rotate the part to be measured (9). The processing unit (4) is also used to calculate the transverse wave velocity and longitudinal wave velocity of the part to be tested (9) at different positions, and to calculate the average value of the transverse wave velocity and longitudinal wave velocity respectively.

Images