Method for Evaluating the Quality of the Internal Structure of Bearing Steel Based on Ultrasonic Waves

Abstract

The present invention discloses a method for evaluating the internal structure quality of bearing steel based on ultrasonic waves, comprising the following steps: S1, constructing a mapping relationship between the amplitude range of multi-level ultrasonic noise signals and the internal structure of bearing steel; S2, performing a credibility evaluation on the mapping relationship between the amplitude range of ultrasonic noise signals at each level and the internal structure of bearing steel, and reconstructing the uncredible mapping relationship; S3, obtaining the amplitude of the noise signals of a batch of bearing steel to be evaluated, and determining the internal structure of this batch of bearing steel based on the credible mapping relationship between the amplitude range of ultrasonic noise signals at each level and the internal structure of bearing steel. The present invention characterizes the internal structure of bearing steel based on ultrasonic noise signals, monitors the internal structure quality in an auxiliary manner during the batch inspection of bearing steel, without adding additional detection processes, and has high efficiency.

Description

Technical Field

[0001] The present invention belongs to the technical field of ultrasonic testing, and more specifically, the present invention relates to a method for evaluating the internal structure quality of bearing steel based on ultrasonic waves. Background Art

[0002] Bearings are key basic components of major equipment and are widely used in key technical fields such as rail transit vehicles, wind power, aeroengines, and national defense heavy equipment. The application environment of bearings is usually relatively complex and harsh, and the performance requirements are high temperature resistance, wear resistance, long life, etc. Therefore, the performance requirements for bearing steel are also high. Research shows that the quantity, size, and morphology of non-metallic inclusions in steel and the metallographic structure have a direct impact on the fatigue and other properties of bearing steel materials.

[0003] For the detection of inclusions, most current bearing steel products need to be subjected to fully automatic water immersion high-frequency (greater than 10 MHz) ultrasonic testing and evaluation (i.e., ultrasonic purity evaluation). The automatic water immersion testing method has good stability, high precision, small blind area, and strong ability to detect small defects at high frequencies. Combined with the ultrasonic C-scan imaging system of the equipment, compared with traditional metallographic inspection of inclusions, the ultrasonic purity detection has a larger volume range of the test specimens, and the ultrasonic imaging shows macroscopic inclusions more intuitively, and the overall evaluation of the internal inclusions of bearing steel is more comprehensive. Currently, the most commonly used ultrasonic purity evaluation standards are the German standard SEP 1927 "Immersion Ultrasonic Testing Method for Macro Inclusions in Forged and Rolled Steel Bars" and the basically equivalent national standard GB / T38683 "Ultrasonic Testing Method for Large Inclusions in Bearing Steel". The standards stipulate relevant sensitivity levels and purity indices for evaluation. The sensitivity is divided into 5 levels, that is, on the basis of the gain of the standard Φ1mm flat-bottomed hole, the gain is increased according to the level (for example, for level 4, an additional 18 dB is added). According to theoretical calculations, if a 10 MHz probe is used for detection and the sensitivity is level 4, non-metallic inclusions with a size of nearly 200 μm can be detected. However, the basic theory of ultrasonic waves also shows that the higher the frequency, the greater the attenuation, and the greater the influence by the internal and external quality of the test specimen. Therefore, ultrasonic purity detection has relatively high requirements for the internal grain size and spheroidized structure of the specimen, which is also the guarantee for ultrasonic purity evaluation to detect inclusions. Therefore, the current research and innovation on ultrasonic purity detection of bearing steel mainly focus on the preparation processes such as heat treatment of bearing steel specimens. However, in the batch production process of heat treatment of bearing steel, abnormalities often occur due to factors such as unreasonable processes and equipment control fluctuations, which have a direct impact on the internal grain size, spheroidization degree, and other structure qualities. How to "discover" and trace the production process in a timely manner during subsequent batch detection is particularly important. Summary of the Invention

[0004] The present invention provides a method for evaluating the internal structure quality of bearing steel based on ultrasonic waves, aiming to improve the above problems.

[0005] The present invention is implemented as follows. A method for evaluating the internal structure quality of bearing steel based on ultrasonic waves, the method comprising the following steps:

[0006] S1. Construct a mapping relationship between the amplitude range of multi-level ultrasonic noise signals and the internal structure of bearing steel;

[0007] S2. Conduct a credibility evaluation on the mapping relationship between the amplitude range of ultrasonic noise signals at each level and the internal structure of bearing steel, and reconstruct the uncredible mapping relationship;

[0008] S3. Obtain the amplitude of the noise signal of the batch of bearing steel to be evaluated, and determine the internal structure of this batch of bearing steel based on the credible mapping relationship between the amplitude range of ultrasonic noise signals at each level and the internal structure of bearing steel.

[0009] Furthermore, the internal structure of bearing steel is grain size and spheroidization degree.

[0010] Furthermore, the construction method of the mapping relationship is specifically as follows:

[0011] S11. Set the water immersion ultrasonic purity detection parameters;

[0012] S12. Specify the specification dimensions of the bearing steel sample and its corresponding surface roughness;

[0013] S13. Use a Φ1mm flat-bottomed hole standard test block, and draw a DAC curve with the defect wave height of the flat-bottomed hole reaching 80% of the full screen as the benchmark;

[0014] S14. Collect bearing steel samples with typical tissue characteristics, and automatically scan them respectively in the ultrasonic depth compensation TCG mode, with the sensitivity set to an increase of 18dB based on the gain of the Φ1mm flat-bottomed hole of the standard test block;

[0015] S15. Establish a multi-level ultrasonic noise signal amplitude range, and then obtain the corresponding internal structure of the bearing steel through dissection to form a mapping relationship between the multi-level ultrasonic noise signal amplitude range and the internal structure of the bearing steel.

[0016] Furthermore, the method for determining the roughness under the specification dimensions of the bearing steel sample is specifically as follows:

[0017] Select bearing steel of the same specification, fix the sensitivity, detect the ultrasonic noise amplitude under different roughness conditions. When there is no obvious change in the ultrasonic noise amplitude caused by the roughness change, the roughness requirement under this sample specification is solidified, that is, less than this roughness value, and the influence of surface roughness change on the ultrasonic noise amplitude can be basically ignored.

[0018] Furthermore, for the round bar specimen, the ovality is required to be ≤ 1% of the specimen diameter. When the diameter of the round bar specimen ≥ 100 mm, the surface roughness ≤ 1.0 μm; when the diameter < 100 mm, the surface roughness < 0.7 μm.

[0019] Furthermore, the amplitude range of the multi-level ultrasonic noise signal is set as follows:

[0020] The amplitude range of the first-level ultrasonic noise signal is less than 10%; the amplitude range of the second-level ultrasonic noise signal is 10% to 20%; the amplitude range of the third-level ultrasonic noise signal is 20% to 30%; the amplitude range of the fourth-level ultrasonic noise signal is 30% to 40%.

[0021] Furthermore, the specific process of the credibility evaluation is as follows:

[0022] Calculate the total normal probability distribution P of the grain size within the amplitude range of the ultrasonic noise signal at each level. Set the mapping relationship corresponding to the amplitude range of the ultrasonic noise signal at each level where the total normal probability distribution P is greater than the probability threshold as credible, otherwise set it as non-credible.

[0023] Furthermore, randomly obtain n specimens within the grain size range corresponding to the amplitude range of the ultrasonic noise signal at each level. The grain size is N, and lgN follows a normal distribution. Let χ = lgN, then the calculation formula for the total normal probability distribution P is as follows:

[0024]

[0025] where P(χ) is the failure probability of the logarithmic grain size χ, σ is the population standard deviation, and μ is the population mean.

[0026] Furthermore, the determination process of the parameters σ and μ is as follows:

[0027] ① Under the same conditions, arrange the grain sizes N of each specimen from small to large, then take the logarithm of N to obtain the order of the logarithmic grain size: lgN1 < lgN2 … lgN i … < lgN n ,;

[0028] ② Determine the cumulative failure probability P(N i ):

[0029] P(N i ) = (i - 0.3) / (n + 0.4)

[0030] where i is the specimen serial number and n is the number of specimens;

[0031] ③ Obtain P(N i) corresponding u values to obtain n pairs of data (u i , lg N i );

[0032] ④ Use the least squares method to determine the parameters σ and μ;

[0033]

[0034]

[0035] Where: X i = u i , Y i = lgN i .

[0036] The present invention is based on the shaft ultrasonic noise signal to characterize the internal structure of bearing steel, and assists in monitoring the quality of the internal structure during the batch inspection of bearing steel, without adding additional detection processes, with high efficiency. Description of the Drawings

[0037] Figure 1 is a flowchart of a method for evaluating the quality of the internal structure of bearing steel based on ultrasonic waves provided by an embodiment of the present invention;

[0038] Figure 2 is the quality of the internal structure of bearing steel corresponding to the ultrasonic noise signal amplitude of 30% to 40% of the full screen in Embodiment 1 of the present invention. Among them, (a) is the range of 30% to 40% of the full screen of the noise signal, (b) the grain size of the structure is 7.5 levels, and (c) the spheroidization of the structure is incomplete;

[0039] Figure 3 is the quality of the internal structure of bearing steel corresponding to the ultrasonic noise signal amplitude of 10% to 20% of the full screen in Embodiment 1 of the present invention. Among them, (a) is the range of 10% to 20% of the full screen of the noise signal, (b) the grain size of the structure is 9.5 levels, and (c) the spheroidization of the structure is 3 levels;

[0040] Figure 4 is to judge the difference in grain size based on the difference in ultrasonic noise signals between the center and the outside of a certain sample in Embodiment 2 of the present invention. Among them, (a) is the noise signal, the outside is in the range of 10% to 20% of the full screen, the center is in the range of 20% to 30% of the full screen, (b) the grain size of the center structure is 8.5 levels, and (c) the grain size of the outside is 9.5 levels. Detailed Embodiments

[0041] The following is a more detailed description of the specific embodiments of the present invention by referring to the drawings and describing the embodiments, so as to help those skilled in the art have a more complete, accurate and in-depth understanding of the inventive concept and technical solution of the present invention.

[0042] The present invention utilizes the scattering mechanism of ultrasonic waves in heterogeneous media. Bearing steel can obtain different metallographic structures under different heat treatment processes. For the propagation of ultrasonic waves, most heat treatment products can be regarded as heterogeneous media. The more uneven the metallographic structure state, the more likely it is to cause alternating changes in the acoustic impedance inside the material. When the wavelength of high-frequency ultrasonic waves is closer to the grain size, the scattering becomes more obvious. The intuitive manifestation is the change in the ultrasonic noise signal. Therefore, the technical solution of the present invention uses the ultrasonic noise signal to characterize the metallographic structure state of bearing steel, and while realizing the evaluation of the inclusion purity of bearing steel in large quantities by water immersion ultrasonic testing, it can also monitor the internal tissue quality of bearing steel.

[0043] Figure 1 FIG. 4 is a flowchart of a method for evaluating the internal tissue quality of bearing steel based on ultrasonic waves provided by an embodiment of the present invention. The method specifically includes the following steps:

[0044] S1. Construct a mapping relationship between the amplitude range of multi-level ultrasonic noise signals and the internal structure of bearing steel. The construction process of this mapping relationship is specifically as follows:

[0045] S11. Set the water immersion ultrasonic purity detection parameters according to the standard SEP 1927;

[0046] S12. Determine the specification size of the bearing steel sample and its corresponding surface roughness;

[0047] The characteristics such as ultrasonic attenuation and scattering are directly affected by the sample size, specification, and surface roughness. The larger the radius of curvature and the surface roughness, the more obvious the ultrasonic attenuation and scattering. Therefore, it is necessary to fix the roughness under the specification size of the bearing steel sample. The process is specifically as follows:

[0048] Select bearing steel of the same specification, fix the sensitivity, and detect the ultrasonic noise amplitude under different roughness conditions. When there is no obvious change in the ultrasonic noise amplitude caused by the change in roughness, the roughness requirement under the specification of this bearing steel sample is fixed. Below this roughness value, the influence of surface roughness change on the ultrasonic noise amplitude can be basically ignored.

[0049] The sample is a round bar sample. Among them, the round bar sample requires that the roundness ≤ 1% of the sample diameter. When the diameter of the round bar sample ≥ 100 mm, the surface roughness requirement ≤ 1.0 μm; when the diameter < 100 mm, the surface roughness requirement < 0.7 μm.

[0050] S13. Use a Φ1mm flat-bottom hole standard test block (the diameter difference from the sample to be inspected ≤ 10%), and draw a DAC curve with the flat-bottom hole defect wave height reaching 80% of the full screen as the benchmark, that is, the ultrasonic "distance - amplitude" relationship curve.

[0051] S14. Collect bearing steel specimens with typical tissue characteristics (different grain sizes and spheroidization degrees), and automatically scan them respectively in the ultrasonic depth compensation TCG mode with the sensitivity set to 18 dB higher than the gain of the Φ1 mm flat-bottom hole in the standard test block.

[0052] S15. Establish a multi-level ultrasonic noise signal amplitude evaluation range, and then form a mapping relationship between the ultrasonic noise signal amplitude range and the internal structure of the bearing steel by anatomically analyzing the corresponding internal structure of the bearing steel, i.e., the grain size level and the spheroidization degree.

[0053] In the ultrasonic purity detection, if the amplitude of the noise signal exceeds 40% of the full screen, it will directly affect the appearance of inclusion defects, not meet the detection signal-to-noise ratio requirements, and not meet the ultrasonic inclusion purity evaluation. Therefore, the upper limit of the noise signal amplitude rating is set at 40%. Based on this, the multi-level ultrasonic noise signal amplitude range is set as follows: The first-level ultrasonic noise signal amplitude range: less than 10%; The second-level ultrasonic noise signal amplitude range: 10% to 20%, i.e., [10%, 20%]; The third-level ultrasonic noise signal amplitude range: 20% to 30%, i.e., (20%, 30%]; The fourth-level ultrasonic noise signal amplitude range: 30% to 40%, i.e., (30%, 40%].

[0054] In S13, the Φ1 mm equivalent is selected as the detection sensitivity because if a value greater than the Φ1 mm equivalent (such as Φ2 mm) is selected, the detection sensitivity will decrease, the noise signal will also decrease, and the "sensitivity" to the tissue change will also decrease, making it impossible to assist in the discrimination of the tissue; if a value less than the Φ1 mm equivalent (such as Φ0.5 mm) is selected, conversely, the noise signal will become higher. On the one hand, it will affect the appearance of the defects in the purity evaluation, and on the other hand, being "too sensitive" will also affect the "discrimination degree" of its mapping with the tissue. Therefore, it is most appropriate to select the Φ1 mm equivalent as the detection sensitivity, which neither affects the detection of purity defects nor can implement the present invention.

[0055] S2. Obtain the total distribution probability of the grain size corresponding to each level of the ultrasonic noise signal amplitude range, set the mapping relationship corresponding to each level of the ultrasonic noise signal amplitude range greater than the probability threshold as credible, otherwise set it as non-credible. For the non-credible mapping relationship, it needs to be re-determined, and the mapping relationship between each level of the ultrasonic noise signal amplitude range and the internal structure of the bearing steel at this level is re-constructed by the method recorded in step S1 until it is credible.

[0056] Collect n specimens of grain sizes corresponding to the amplitude range of the fourth-level ultrasonic noise signals, anatomically analyze the specimens, measure and grade the grain sizes in accordance with relevant standards (such as GB / T 6394 "Methods for Determining Average Grain Size of Metals"), and establish the corresponding relationship between the ultrasonic noise signals and the grain sizes using the normal distribution. Set the grain size as the parameter N, where lgN follows the normal distribution. Let χ = lgN, and then the logarithmic grain size statistical analysis can be carried out using the normal distribution theory. The normal probability distribution function is:

[0057]

[0058] In the formula, P(χ) is the failure probability of the logarithmic grain size χ / %, σ is the population standard deviation, and μ is the population mean.

[0059] Let u = (χ - μ) / σ, then we have:

[0060] χ = -μ + uσ (2)

[0061] Substitute Equation (2) into Equation (1), and the standard normal distribution function with a mean of 0 and a standard deviation of 1 can be obtained:

[0062]

[0063] In Equation (1), P(χ) and χ are obviously non-linearly related, but in Equation (2), χ and u are linearly related. Therefore, the regression equation of χ and u in Equation (2) can be fitted by the least squares method to determine the normal distribution parameters σ and μ. The determination process of the parameters σ and μ is as follows:

[0064] ① Under the same conditions, arrange the grain sizes N of each specimen from small to large, and then take the logarithm of N to obtain the order statistics of the logarithmic grain sizes: lgN1 < lgN2 … lgN i … < lgN n ,;

[0065] ② The smaller the size, the lower the detection probability, and the larger the size, the higher the detection probability. Determine the cumulative failure probability P(N i ) through the median rank method:

[0066] P(N i ) = (i - 0.3) / (n + 0.4) (4)

[0067] where i is the specimen serial number and n is the number of specimens.

[0068] ③ Obtain the corresponding u value of P(N i ) by looking up the standard normal distribution function table with a mean of 0 and a standard deviation of 1, so as to obtain n groups of data pairs (u i , lg Ni );

[0069] ④ Determine σ and μ using the least squares method;

[0070]

[0071]

[0072] Where: X i = u i , Y i = lgN i , n is the number of data pairs. Substitute the obtained standard deviation σ and mean μ into Equation (1), which is the grain size probability distribution function statistically obtained based on the normal distribution. The total distribution probability P of the grain size within the amplitude range of the ultrasonic noise signals at each level can be obtained. The mapping relationship reaching the probability threshold (70%) is a credible mapping relationship. For the uncredible mapping relationship, it needs to be reconstructed to improve the accuracy of the mapping relationship between the amplitude range of the ultrasonic noise signals and the internal structure of the bearing steel.

[0073] S3. Obtain the amplitude of the noise signal of the batch of bearing steel to be evaluated, and then determine the internal structure of the bearing steel.

[0074] The present invention will be described in detail below in combination with Example 1 and the embodiments for the above method for evaluating the quality of the internal structure of bearing steel based on ultrasonic waves.

[0075] Example 1: A batch of GCr15 bearing steel with a specification of Φ60mm that needs to be detected in batches.

[0076] Step 1: Set the parameters for water immersion ultrasonic purity detection in accordance with the standard SEP 1927.

[0077] Step 2: The surface roughness of the GCr15 bearing steel sample with a specification of Φ60mm is 0.623μm.

[0078] Step 3: Use a Φ1mm flat-bottom hole standard test block (Φ60mm) to draw a DAC curve based on the condition that the wave height of the flat-bottom hole defect reaches 80% of the full screen.

[0079] Step 4: Collect bearing steel samples with typical tissue characteristics (different grain sizes and spheroidization degrees). Turn on the TCG mode of the equipment to automatically scan these samples with typical tissue characteristics, and increase the sensitivity by 18dB based on the gain of the Φ1mm flat-bottom hole in the standard test block.

[0080] Step 5: Through anatomical analysis of the corresponding grain size level and spheroidization degree of the bearing steel, form and verify the correspondence between the ultrasonic noise signal and the quality of the internal structure, as shown in Table 1.

[0081] Step 6: After batch verification, conduct mass internal structure quality monitoring for GCr15 bearing steel of this specification, as shown in Figure 1 , Figure 2 , and conduct sample noise signal characterization and dissection verification.

[0082] Table 1 Establishment of characterization correspondence in the embodiment

[0083]

[0084] Embodiment 2:

[0085] When batch detecting GCr15 bearing steel of the same specification as in Embodiment 1, it is found that Figure 3 there are obvious differences in the ultrasonic noise signals at the core and the outer side of a certain batch of samples. It is judged that there are differences in the internal and external structures, and then tracing the heat treatment production of this batch, it is found that there are fluctuations in the control of the cooling rate at a certain period of the on-site process, resulting in differences in the internal and external structures.

[0086] The present invention has been described by way of example. Obviously, the specific implementation of the present invention is not limited by the above methods. As long as various non-substantive improvements are made by adopting the method concept and technical solution of the present invention, or the concept and technical solution of the present invention are directly applied to other occasions without improvement, they are all within the protection scope of the present invention.

Claims

1. An ultrasonic-based method for evaluating the internal structure quality of bearing steel, characterized in that, The method includes the following steps: S1. Establish the mapping relationship between the amplitude range of multi-level ultrasonic noise signals and the internal structure of bearing steel; S2. Conduct a credibility evaluation on the mapping relationship between the amplitude range of ultrasonic noise signals at each level and the internal structure of bearing steel, and reconstruct the untrustworthy mapping relationship; S3. Obtain the amplitude of the noise signal of the batch of bearing steel to be evaluated, and determine the internal structure of this batch of bearing steel based on the credible mapping relationship between the amplitude range of ultrasonic noise signals at each level and the internal structure of bearing steel. The internal structure of bearing steel is grain size and spheroidization degree; The specific method for establishing the mapping relationship is as follows: S11. Set the parameters for water immersion ultrasonic purity detection; S12. Specify the specification dimensions of the bearing steel sample and its corresponding surface roughness; S13. Use a Φ1mm flat-bottomed hole standard test block, and draw a DAC curve with the defect wave height of the flat-bottomed hole reaching 80% of the full screen as the benchmark; S14. Collect bearing steel samples with typical tissue characteristics, and automatically scan them respectively in the ultrasonic depth compensation TCG mode, with the sensitivity set to an increase of 18 dB based on the gain of the Φ1mm flat-bottomed hole of the standard test block; S15. Establish the amplitude range of multi-level ultrasonic noise signals, and then obtain the corresponding internal structure of the bearing steel by dissection to form the mapping relationship between the amplitude range of multi-level ultrasonic noise signals and the internal structure of bearing steel.

2. The method for evaluating the internal structure quality of bearing steel based on ultrasonic waves according to claim 1, characterized in that The specific method for determining the roughness under the specification dimensions of the bearing steel sample is as follows: Select bearing steel of the same specification, fix the sensitivity, detect the ultrasonic noise amplitude under different roughness conditions. When there is no obvious change in the ultrasonic noise amplitude caused by the roughness change, the roughness requirement under this sample specification is solidified, that is, less than this roughness value, and the influence of surface roughness change on the ultrasonic noise amplitude can be basically ignored.

3. The method for evaluating the internal structure quality of bearing steel based on ultrasonic waves according to claim 2, wherein For round bar-shaped samples, the roundness is required to be ≤1% of the sample diameter. When the diameter of the round bar-shaped sample ≥100mm, the surface roughness ≤1.0μm; when the diameter <100mm, the surface roughness <0.7μm.

4. The method for evaluating the internal structure quality of bearing steel based on ultrasonic waves according to claim 1, characterized in that The amplitude range of multi-level ultrasonic noise signals is set as follows: The amplitude range of the first-level ultrasonic noise signal is less than 10%; the amplitude range of the second-level ultrasonic noise signal is 10% to 20%; the amplitude range of the third-level ultrasonic noise signal is 20% to 30%; the amplitude range of the fourth-level ultrasonic noise signal is 30% to 40%.

5. The method for evaluating the internal structure quality of bearing steel based on ultrasonic waves according to claim 1, characterized in that, The specific process of credibility evaluation is as follows: Calculate the total normal probability distribution P of the grain size within the amplitude range of the ultrasonic noise signal at each level. Set the mapping relationship corresponding to the amplitude range of the ultrasonic noise signal at each level where the total normal probability distribution P is greater than the probability threshold as credible, otherwise set it as uncredible.

6. The method for evaluating the internal structure quality of bearing steel based on ultrasonic waves according to claim 5, wherein Randomly obtain n samples within the grain size range corresponding to the amplitude range of the ultrasonic noise signal at each level. The grain size is N, and lgN follows a normal distribution. Let χ = lgN, then the calculation formula for the total normal probability distribution P is as follows: Among them, P(χ) is the failure probability of the logarithmic grain size χ, σ is the population standard deviation, and μ is the population mean.

7. The method for evaluating the internal structure quality of bearing steel based on ultrasonic waves according to claim 6, characterized in that, The specific process for determining the parameters σ and μ is as follows: ① Under the same conditions, arrange the grain sizes N of each specimen in ascending order from smallest to largest, then take the logarithm of N to obtain the order of logarithmic grain sizes: lgN1 < lgN2 … lgN i … < lgN n ; ② Determine the cumulative failure probability P(N i ): P(N i ) = (i - 0.3)(n + 0.4) Among them, i is the sample serial number, and n is the number of samples; ③ Obtain the u value corresponding to P(N i ) by looking up the standard normal distribution function table with a mean of 0 and a standard deviation of 1, so as to obtain n pairs of data (u i , lgN i ); ④ Use the least squares method to determine the parameters σ and μ; Where: X i = u i , Y i = lgN i .

Images