Method and apparatus for producing a hot crack resistant steel roll

Abstract

The application relates to the technical field of metallurgical engineering, in particular to a production method and device of a thermal crack-resistant steel roller. The method comprises the following steps: in the process of pouring outer-layer molten metal into a mold, a spindle speed setting value is adjusted based on multivariate sensing data at a first adjustment frequency until the pouring of the outer-layer molten metal is completed; after the pouring of the outer-layer molten metal is completed, whether a second pouring condition is established is determined based on the multivariate sensing data at a second adjustment frequency; in the process of pouring inner-layer molten metal, a pouring speed setting value is adjusted based on the multivariate sensing data at a third adjustment frequency until the pouring of the inner-layer molten metal is completed; after the pouring of the inner-layer molten metal is completed, whether a centrifuge stop condition is established is determined based on the multivariate sensing data at a fourth adjustment frequency. It can be seen that the method can set corresponding parameter control strategies for different process stages such as outer-layer pouring, inner-layer pouring and stopping, and the forming quality and production stability of the thermal crack-resistant steel roller are significantly improved.

Description

Technical Field

[0001] This application relates to the field of metallurgical engineering technology, and in particular to a method and apparatus for producing heat-crack resistant steel rolls. Background Technology

[0002] Steel rolls are core components in steel rolling production, primarily used for rolling and shaping metal sheets and profiles. Their performance directly determines the quality of the rolled material and production efficiency. Currently, high-performance composite steel rolls are mainly produced using centrifugal casting, a process that relies on a steel roll centrifuge. The basic process involves pouring molten alloy steel into a high-speed rotating mold cavity. Under strong centrifugal force, the molten metal is thrown against the inner wall of the mold and adheres to the mold, subsequently solidifying sequentially from the outside in, ultimately forming a composite structure with an outer layer of high-hardness, wear-resistant material and an inner layer of high-toughness supporting material.

[0003] However, existing production processes generally use a preset constant speed for casting and solidification, which cannot be adjusted in real time according to dynamic changes such as the fluidity of the molten metal, temperature distribution, and solidification shrinkage. This easily leads to defects such as uneven wall thickness, microstructure segregation, poor interlayer bonding, and internal shrinkage porosity in the rolls, which seriously affect the quality and service life of steel roll products and make it difficult to meet the production needs of high-performance rolled steel. Summary of the Invention

[0004] This application provides a method and apparatus for producing heat-crack resistant steel rolls to solve the problem of low product quality of steel rolls produced by traditional methods.

[0005] In a first aspect, embodiments of this application provide a method for producing a heat-crack-resistant steel roll. The heat-crack-resistant steel roll is a two-layer composite structure formed by two centrifugal castings during centrifugal casting of a mold driven by a centrifuge, comprising an outer layer and an inner layer. The method includes: acquiring multi-dimensional sensing data at a preset sampling frequency during centrifugal operation; wherein the multi-dimensional sensing data includes one or more of the following: mold radial runout, instantaneous vibration acceleration of the spindle, current casting speed, current spindle speed, and temperature data; during the casting of the outer layer of molten metal into the mold, adjusting the spindle speed setting value based on the multi-dimensional sensing data at a first adjustment frequency until the outer layer of molten metal is cast. End; After the outer layer of molten metal is poured, the second pouring condition is determined based on multi-sensor data at the second adjustment frequency. If the second pouring condition is met, a second pouring command is generated to inject the inner layer of molten metal into the mold. During the pouring of the inner layer of molten metal, the pouring speed setting is adjusted based on multi-sensor data at the third adjustment frequency until the pouring of the inner layer of molten metal is completed. After the pouring of the inner layer of molten metal is completed, the centrifuge stop condition is determined based on multi-sensor data at the fourth adjustment frequency. If the centrifuge stop condition is met, a centrifuge stop command is generated to control the centrifuge spindle to stop.

[0006] In one possible implementation, during the process of pouring the outer layer of molten metal into the mold, the step of adjusting the spindle speed setting value based on multi-sensor data at a first adjustment frequency includes: performing a Fourier transform on the time-domain data corresponding to the radial runout of the mold within a first time window at the first adjustment frequency to obtain a spectral feature vector of the radial runout of the mold; wherein, the first adjustment frequency is less than a preset sampling frequency, and the first time window has the same period length as the first adjustment frequency; the time-domain data is a continuous data sequence with time as the abscissa and the radial runout of the mold as the ordinate; and the condition is that the amplitude of the fundamental frequency in the spectral feature vector exceeds a first frequency threshold, and the amplitude of the second harmonic in the spectral feature vector does not exceed a second frequency threshold. In this case, the axial temperature difference of the mold within the first time window is calculated based on the temperature data; wherein, the axial temperature difference of the mold is equal to the difference between the maximum and minimum values ​​of the average temperature of the inner wall of the top, middle, and bottom of the mold; if the axial temperature difference of the mold does not exceed the first temperature threshold, the spindle speed setting is adjusted to Z times the current spindle speed, 0 < Z < 1; if the axial temperature difference of the mold exceeds the first temperature threshold, the spindle speed setting is adjusted to W times the current spindle speed, 0 < W < Z; if the amplitude of the second harmonic in the spectral feature vector exceeds the second frequency threshold, the spindle speed setting is adjusted to Y times the current spindle speed, 0 < Y < W.

[0007] In one possible implementation, after performing a Fourier transform on the time-domain data corresponding to the radial runout of the mold within a first time window at a first adjustment frequency to obtain the spectral feature vector of the radial runout of the mold, the method further includes: if the amplitude of the fundamental frequency in the spectral feature vector does not exceed a first frequency threshold and the amplitude of the second harmonic in the spectral feature vector does not exceed a second frequency threshold, calculating the peak-to-peak value of the radial runout of the mold within the first time window and the root mean square value of the instantaneous vibration acceleration of the spindle; if the peak-to-peak value is not within a first preset range and / or, if the root mean square value is not within a second preset range, adjusting the spindle speed setting to P times or O times the current spindle speed, where Z≤P<1<O.

[0008] In one possible implementation, the temperature data includes a mold thermal imaging image, which is generated by collecting infrared radiation energy along the side of the mold. The step of determining whether the second pouring conditions are met based on multi-sensor data at a second adjustment frequency, and generating a second pouring command if the second pouring conditions are met, includes: calculating the outer inner wall temperature based on the mold thermal imaging image and the outer pouring temperature within a second time window at the second adjustment frequency; wherein the second adjustment frequency is less than a preset sampling frequency, the second time window is equal to the period length corresponding to the second adjustment frequency, and the outer pouring temperature refers to the pouring temperature setting value of the outer metal liquid during pouring; determining that the second pouring conditions are met if the outer inner wall temperature is less than or equal to a second temperature threshold; generating a second pouring command if the second pouring conditions are met, to set the centrifuge spindle speed setting value to a first speed and to inject the inner metal liquid at a first pouring speed; wherein the first speed and the first pouring speed are determined based on a preset mapping relationship between the spindle speed, the metal liquid density, and the pouring speed.

[0009] In one possible implementation, the step of calculating the outer inner wall temperature based on the mold thermal imaging image and the outer pouring temperature within a second time window at a second adjustment frequency includes: extracting the outer wall temperature value for each mold thermal imaging image within the second time window at the second adjustment frequency; wherein the outer wall temperature value is equal to the average of the temperature values ​​of the top region, middle region, and bottom region along the mold axis; the temperature value of each region is the arithmetic mean of the temperatures of all thermal imaging pixels in the corresponding region, and the pixel temperature is calculated based on the standard emissivity of the mold material or the real-time acquired emissivity; calculating the interfacial heat transfer coefficient between the outer molten metal and the mold based on the outer pouring temperature; and calculating the outer inner wall temperature using a finite difference inverse iterative algorithm based on the outer wall temperature value, the interfacial heat transfer coefficient, and the outer pouring temperature.

[0010] In one possible implementation, the steps for calculating the inner wall temperature of the outer layer using a finite difference inverse iterative algorithm based on the outer wall temperature, interfacial heat transfer coefficient, and outer layer casting temperature include: discretizing the mold wall thickness radially into k nodes; where the first node corresponds to the position of the inner wall surface of the mold, and the kth node corresponds to the position of the outer wall surface of the mold; cyclically executing the following steps S1-S5 based on the heat conduction equation and the heat flow balance equation, where the heat flow balance equation is established based on the interfacial heat transfer coefficient and the outer layer casting temperature; S1: setting the initial value of the iteration number m to 1, and extracting the imaging time. t m The corresponding outer wall temperature value is assigned to the k-th node at the imaging time. t m node temperature And based on the k-th node at the imaging time t m node temperature For nodes 1 to (k-1) at the imaging time t m The node temperature is assigned a value; where the imaging time... t m S1: Indicates the imaging time of the m-th thermal image of the mold within the second time window; S2: Increment the iteration number m by 1 and extract the imaging time. t m The corresponding outer wall temperature value is assigned to the k-th node at the imaging time. t m node temperature And each node at the previous imaging time t m-1 node temperature Substituting into the heat conduction equation, we obtain the values ​​of the 2nd to (k-1)th nodes at the imaging time. t m node temperature ,2≤ j ≤k-1; S3: Place the second node at the imaging time t m node temperature Substituting into the heat flow balance equation, we obtain the first node at the imaging time. t m node temperature If m < n, proceed to step S2; if m = n, proceed to step S4; where n is the total number of thermal images of the mold within the second time window; S4: Based on the relative error formula, calculate the value of the first node at the imaging time. t m node temperature Rather than in the previous imaging moment t m-1 node temperature The relative error; S5: If the relative error is less than the error threshold, assign the temperature of the outer inner wall to the value of the first node at the imaging time. t m node temperature If the relative error is greater than or equal to the error threshold, the temperature of the outer inner wall is assigned to a preset value and the loop ends; wherein the preset value is greater than the second temperature threshold.

[0011] In one possible implementation, during the pouring of the inner layer of molten metal, the step of adjusting the pouring speed setting value based on multi-sensor data at a third adjustment frequency includes: calculating the axial temperature difference and the circumferential temperature difference of the mold based on temperature data within a third time window at the third adjustment frequency; wherein, the third adjustment frequency is less than the preset sampling frequency, the third time window has the same period length as the third adjustment frequency, the axial temperature difference of the mold is equal to the difference between the maximum and minimum values ​​of the average temperature of the inner wall at the top, middle, and bottom of the mold, and the circumferential temperature difference of the mold is equal to the difference between the maximum and minimum values ​​of the inner wall temperature at different circumferential positions at the same height section of the mold; when the axial temperature difference of the mold is greater than a third temperature threshold, the pouring speed setting value is adjusted to Q times the current pouring speed, 0 < Q < 1; when the axial temperature difference of the mold is less than or equal to the third temperature threshold and the circumferential temperature difference of the mold is greater than a fourth temperature threshold, the pouring speed setting value is adjusted to L times the current pouring speed, 0 < L < Q.

[0012] In one possible implementation, after the inner layer of molten metal is poured, the step of determining whether the centrifuge shutdown condition is met based on multivariate sensing data at a fourth adjustment frequency includes: generating a mold temperature change curve based on temperature data within a fourth time window at the fourth adjustment frequency; wherein the fourth adjustment frequency is less than a preset sampling frequency, and the fourth time window has the same period length as the fourth adjustment frequency; inputting the mold temperature change curve into a temperature estimation model to determine the internal temperature; and determining that the centrifuge shutdown condition is met when the internal temperature is below the solidus line; the solidus line is determined by differential thermal analysis experiments based on the material composition of the inner layer of molten metal.

[0013] In one possible implementation, before adjusting the spindle speed setting value based on multi-sensor data at a first adjustment frequency during the pouring of the outer layer of molten metal into the mold, the method further includes: when the spindle speed setting value of the centrifuge spindle is a second speed, calculating the peak-to-peak value of the mold radial runout and the root mean square value of the instantaneous vibration acceleration of the spindle within a fifth time window at a fifth adjustment frequency; wherein the fifth adjustment frequency is less than a preset sampling frequency, and the fifth time window has the same period length as the fifth adjustment frequency; when the peak-to-peak value is less than a first runout threshold and the root mean square value is less than a first acceleration threshold, generating a first pouring command to inject the outer layer of molten metal at a second pouring speed; wherein the second speed and the second pouring speed are determined based on a preset mapping relationship between the spindle speed, molten metal density, and pouring speed.

[0014] Secondly, embodiments of this application provide a production apparatus for heat-crack resistant steel rolls. The heat-crack resistant steel rolls are two-layer composite structures formed by two centrifugal casting processes performed while a centrifuge drives a mold to rotate. The rolls include an outer layer and an inner layer. The apparatus includes: a data acquisition module configured to acquire multi-dimensional sensing data at a preset sampling frequency during centrifuge operation; wherein the multi-dimensional sensing data includes one or more of the following: mold radial runout, instantaneous spindle vibration acceleration, current casting speed, current spindle speed, and temperature data; a first adjustment module configured to adjust the spindle speed setting value based on the multi-dimensional sensing data at a first adjustment frequency during the casting of the outer layer of molten metal into the mold, until the casting of the outer layer of molten metal is completed; and a first judgment module. The interruption module is configured to: after the outer layer of molten metal is poured, determine whether the conditions for a second pouring are met based on multi-sensor data at a second adjustment frequency, and generate a second pouring command if the conditions for a second pouring are met, to inject the inner layer of molten metal into the mold; the second adjustment module is configured to: during the pouring of the inner layer of molten metal, adjust the pouring speed setting value based on multi-sensor data at a third adjustment frequency until the pouring of the inner layer of molten metal is completed; the second judgment module is configured to: after the inner layer of molten metal is poured, determine whether the centrifuge stop condition is met based on multi-sensor data at a fourth adjustment frequency, and generate a centrifuge stop command if the centrifuge stop condition is met, to control the centrifuge spindle to stop.

[0015] As can be seen from the above, the embodiments of this application provide a method and apparatus for producing heat-crack resistant steel rolls. The method includes: acquiring multi-dimensional sensing data at a preset sampling frequency during the operation of a centrifuge; wherein, the multi-dimensional sensing data includes one or more of the following: mold radial runout, instantaneous vibration acceleration of the spindle, current pouring speed, current spindle speed, and temperature data.

[0016] During the pouring of the outer layer of molten metal into the mold, the spindle speed setting is adjusted at a first adjustment frequency based on multi-sensor data until the pouring of the outer layer of molten metal is completed. After the pouring of the outer layer of molten metal is completed, the second adjustment frequency is used to determine whether the conditions for the second pouring are met based on multi-sensor data. If the conditions for the second pouring are met, a second pouring command is generated to inject the inner layer of molten metal into the mold. During the pouring of the inner layer of molten metal, the pouring speed setting is adjusted at a third adjustment frequency based on multi-sensor data until the pouring of the inner layer of molten metal is completed. After the pouring of the inner layer of molten metal is completed, the fourth adjustment frequency is used to determine whether the centrifuge stop condition is met based on multi-sensor data. If the centrifuge stop condition is met, a centrifuge stop command is generated to control the centrifuge spindle to stop. It is evident that this method employs a phased, differentiated adjustment approach, setting corresponding parameter control strategies for different process stages such as outer layer casting, inner layer casting, and shutdown. This ensures the uniform spreading and forming of the outer layer molten metal under centrifugal force, while precisely controlling the casting process of the inner layer molten metal to achieve a good metallurgical bond with the outer layer. Simultaneously, it avoids internal stress cracks in the rolls caused by sudden centrifuge shutdown, significantly improving the forming quality and production stability of the heat-crack resistant steel rolls. Attached Figure Description

[0017] Figure 1 A schematic diagram of the first process flow for the production method of heat-crack-resistant steel rolls provided in this application embodiment;

[0018] Figure 2 This is a flowchart illustrating the process of adjusting the spindle speed setting during the outer layer molten metal pouring stage, as provided in an embodiment of this application.

[0019] Figure 3 A flowchart illustrating the process of determining whether the conditions for a second pouring are met, provided for an embodiment of this application;

[0020] Figure 4 A flowchart illustrating the process of adjusting the pouring speed setting value during the inner layer molten metal pouring stage, as provided in an embodiment of this application.

[0021] Figure 5 A flowchart illustrating the process of determining whether the conditions for centrifuge to stop are met, provided in an embodiment of this application;

[0022] Figure 6 Metallographic diagram of the heat-crack-resistant steel roll provided in the embodiments of this application;

[0023] Figure 7 This is a schematic diagram of the production apparatus for heat-resistant crack-resistant steel rolls provided in the embodiments of this application.

[0024] Among them, 1001-data acquisition module, 1002-first adjustment module, 1003-first judgment module, 1004-second adjustment module, 1005-second judgment module, and 1006-third judgment module. Detailed Implementation

[0025] To enable those skilled in the art to better understand the technical solutions in this application, the technical solutions in the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of this application.

[0026] To address the issue of poor product quality and service life of steel rolls, this application provides a method and apparatus for producing heat-crack resistant steel rolls. This method uses thermal imaging technology to collect temperature data in real time during the roll forming process, accurately calculates the internal thermal state of the steel roll, dynamically adjusts key parameters of the heat treatment process, achieves precise monitoring of steel roll production, and effectively suppresses the generation of heat cracks.

[0027] Specifically, the production method provided in this application embodiment can be used to produce heat-crack resistant steel rolls. The heat-crack resistant steel rolls can be a two-layer composite structure formed by two centrifugal castings when the centrifuge drives the mold to rotate. The rolls include an outer layer and an inner layer. The outer layer of molten metal can be composed of metal alloy elements, with a specific content of at least 5%-10% chromium (Cr), 9% tungsten (W), 9% vanadium (V), and 9% molybdenum (Mo). The inner layer of molten metal can be composed of high-strength ductile iron.

[0028] Understandably, in the actual production process, the centrifuge can drive the mold to rotate stably according to the set speed value. Then, the prepared outer layer molten metal and inner layer molten metal can be injected into the high-speed rotating mold cavity (or mold cylinder) through the gate in sequence. Under the action of centrifugal force, the molten metal will spread evenly along the inner wall of the mold and solidify layer by layer to finally produce a hollow or solid multi-layer structure steel roll.

[0029] In some implementations, the dimensions (product specifications) of the heat-crack resistant steel roll can be: Φ490mm×850mm.

[0030] Figure 1 This is a first process diagram of a method for producing heat-crack-resistant steel rolls provided in an embodiment of this application.

[0031] like Figure 1 As shown, the production method of the heat-crack resistant steel roll provided in this application includes the following steps S100-S500.

[0032] S100: During the operation of the centrifuge, multi-dimensional sensing data is acquired at a preset sampling frequency; the multi-dimensional sensing data includes one or more of the following: mold radial runout, spindle instantaneous vibration acceleration, current pouring speed, current spindle speed, and temperature data.

[0033] In this embodiment, the preset sampling frequency can be 20Hz (i.e., 20 times per second), 30Hz (i.e., 30 times per second), or 50Hz (i.e., 50 times per second). The specific frequency can be set according to the rolling mill production process stage to meet the monitoring requirements of different stages (e.g., outer layer casting stage, outer layer solidification stage, inner layer casting stage, inner layer solidification stage).

[0034] Furthermore, the radial runout of the mold can be collected by a laser displacement sensor. The collection points are three evenly distributed test points on the outer wall of the mold (e.g., the top test point, the middle test point, and the bottom test point). The average value of the three test points is taken as the valid data to determine the coaxiality of the mold operation.

[0035] The instantaneous vibration acceleration of the spindle can be acquired by a triaxial piezoelectric accelerometer, which can be installed in the bearing housing on both the drive and non-drive ends of the spindle. The unit of instantaneous vibration acceleration of the spindle is m / s². 2 (meters per second squared) is used for real-time monitoring of mechanical shocks and resonances.

[0036] The current pouring speed refers to the mass of molten metal injected into the mold per unit time, which is measured in real time by an electromagnetic flowmeter and is measured in kg / s (kilograms per second). It is used to precisely control the injection of molten metal.

[0037] The current spindle speed refers to the rotational speed of the centrifuge spindle, which is acquired by an incremental photoelectric encoder or Hall effect speed sensor, and is measured in r / min (revolutions per minute). It is used for closed-loop control of the centrifugal force field intensity.

[0038] Furthermore, the temperature data can include the mold inner wall temperature and a mold thermal image. The mold inner wall temperature can be collected by S-type or K-type armored thermocouples. The thermocouples are installed at three axial positions on the mold inner wall: top, middle, and bottom. Three measuring points are evenly distributed along the circumference (same cross-section) at each position, for a total of nine measuring points, used to monitor the axial temperature. The mold thermal image is generated by collecting infrared radiation energy along the side of the mold. Specifically, an industrial-grade short-wave infrared thermal imager (temperature range 0~2000℃) can be fixed perpendicular to the mold axis, and the imager's imaging range covers the entire length of the mold, used for non-contact monitoring of temperature distribution and determination of the solidification process.

[0039] S200: During the process of pouring the outer layer of molten metal into the mold, the spindle speed setting is adjusted based on multi-sensor data at the first adjustment frequency until the pouring of the outer layer of molten metal is completed.

[0040] The first adjustment frequency is lower than the preset sampling frequency, which can be 5Hz, 10Hz, or 15Hz. This allows for rapid detection and response to changes in the outer layer casting process based on multi-sensor data, thereby adjusting the rotation speed setting and reducing the risk of hot cracking.

[0041] It should also be noted that the centrifuge spindle is driven by a variable frequency motor. In this embodiment, the speed setting value can be converted into a corresponding motor frequency command. By adjusting the output frequency through the frequency converter, the operating speed of the motor is changed, thereby driving the spindle to rotate stably according to the set value. If the speed needs to be increased, the output frequency of the frequency converter is increased; if the speed needs to be decreased, the output frequency is decreased, thereby precisely matching the speed control requirements of the outer layer casting stage.

[0042] S300: After the outer layer of molten metal is poured, the second pouring condition is determined based on multi-sensor data using the second adjustment frequency. If the second pouring condition is met, a second pouring command is generated to inject the inner layer of molten metal into the mold.

[0043] It is understandable that the outer layer of molten metal can be injected in a quantitative manner, and the pouring of the outer layer of molten metal can be considered complete when the cumulative pouring volume reaches the preset weight of the outer layer. After the pouring of the outer layer of molten metal is completed, the outer layer solidification stage can begin. The second adjustment frequency is lower than the preset sampling frequency, and can be 1Hz or 5Hz.

[0044] S400: During the pouring of the inner layer molten metal, the pouring speed setting is adjusted based on multi-sensor data at the third adjustment frequency until the pouring of the inner layer molten metal is completed.

[0045] The third adjustment frequency is lower than the preset sampling frequency, and can be 5Hz, 10Hz, or 15Hz. In this way, changes in the inner layer pouring process can be quickly detected and responded to based on multi-sensor data, thereby adjusting the pouring speed setting value and reducing the risk of hot cracking.

[0046] It should also be noted that, once the pouring speed setting is determined, the pouring speed in this embodiment can be adjusted by regulating the opening of the flow control valve. For example, by adjusting the opening of the flow control valve, the cross-sectional area of ​​the molten metal flowing out of the gate can be controlled, thereby precisely regulating the amount of molten metal poured per unit time to match the pouring speed setting.

[0047] S500: After the inner layer of molten metal is poured, the centrifuge stops at the fourth adjustment frequency based on multi-sensor data to determine whether the centrifuge stops. If the centrifuge stops, a centrifuge stops command is generated to control the centrifuge spindle to stop.

[0048] It is understood that the inner layer molten metal can be injected quantitatively, and the pouring of the inner layer molten metal can be considered complete when the cumulative pouring volume reaches the preset weight of the inner layer. After the inner layer molten metal pouring is completed, the inner layer solidification stage can begin. The fourth adjustment frequency is less than the preset sampling frequency, and can be 1Hz or 5Hz. In practical applications, this embodiment can control the spindle speed to gradually decrease in a 50r / min gradient until it finally stops.

[0049] It should also be noted that in the actual production process, the total amount of liquid metal injected into the outer layer is determined based on the preset outer layer thickness and roll size, while the total amount of liquid metal injected into the inner layer is determined based on the preset core diameter and material density. The quantitative injection can be achieved through an electronic weighing ladle system or an electromagnetic flowmeter.

[0050] As can be seen from the above, this application provides a method for producing heat-crack-resistant steel rolls. The method includes: acquiring multi-dimensional sensing data at a preset sampling frequency during centrifuge operation; wherein the multi-dimensional sensing data includes one or more of the following: mold radial runout, instantaneous spindle vibration acceleration, current pouring speed, current spindle speed, and temperature data; adjusting the spindle speed setting value based on the multi-dimensional sensing data at a first adjustment frequency during the pouring of the outer layer of molten metal into the mold until the pouring of the outer layer of molten metal is completed; determining whether the conditions for a second pouring are met based on the multi-dimensional sensing data at a second adjustment frequency, and generating a second pouring command to inject the inner layer of molten metal into the mold if the conditions for the second pouring are met; adjusting the pouring speed setting value based on the multi-dimensional sensing data at a third adjustment frequency during the pouring of the inner layer of molten metal until the pouring of the inner layer of molten metal is completed; determining whether the centrifuge stop conditions are met based on the multi-dimensional sensing data at a fourth adjustment frequency, and generating a centrifuge stop command if the centrifuge stop conditions are met to control the centrifuge spindle to stop. It is evident that this method employs a phased, differentiated adjustment approach, setting corresponding parameter control strategies for different process stages such as outer layer casting, inner layer casting, and shutdown. This ensures the uniform spreading and forming of the outer layer molten metal under centrifugal force, while precisely controlling the casting process of the inner layer molten metal to achieve a good metallurgical bond with the outer layer. Simultaneously, it avoids internal stress cracks in the rolls caused by sudden centrifuge shutdown, significantly improving the forming quality and production stability of the heat-crack resistant steel rolls.

[0051] Furthermore, prior to step S200, the embodiments of this application may also include the following steps S601-S602.

[0052] S601: With the centrifuge spindle speed set to the second speed, calculate the peak-to-peak value of the mold radial runout and the root mean square value of the instantaneous vibration acceleration of the spindle within the fifth time window using the fifth adjustment frequency; wherein, the fifth adjustment frequency is less than the preset sampling frequency, and the fifth time window is equal to the period length corresponding to the fifth adjustment frequency.

[0053] The second rotational speed can be referred to as the outer layer pouring reference speed, which can be set based on the characteristics of the outer layer molten metal (such as the density of the molten metal). In this embodiment, the centrifuge spindle can be set to rotate at the second rotational speed first, while the mold idles, i.e., the outer layer molten metal is not poured in until the first pouring conditions are met. Specifically, in this embodiment, the first pouring condition can be determined by calculating the peak-to-peak value of the mold's radial runout and the root mean square value of the instantaneous vibration acceleration of the spindle.

[0054] Furthermore, the fifth adjustment frequency can be less than the preset sampling frequency, specifically 5Hz-20Hz, but this application embodiment does not specifically limit this.

[0055] S602: When the peak-to-peak value is less than the first runout threshold and the root mean square value is less than the first acceleration threshold, a first pouring command is generated to inject the outer layer of molten metal at a second pouring speed; wherein, the second rotation speed and the second pouring speed are determined based on a preset mapping relationship between the spindle rotation speed, molten metal density and pouring speed.

[0056] In this embodiment, the first jump threshold can be 0.12mm, 0.15mm, etc., and this embodiment does not specifically limit it. The first acceleration threshold can be 2.0 m / s². 2 2.5 m / s 2 or 3.0 m / s 2 This application does not impose specific limitations on this aspect.

[0057] For example, assuming a second rotational speed of 800 r / min, a preset sampling frequency of 20Hz, and a fifth adjustment frequency of 10Hz, the centrifuge is initially run unloaded after startup. This embodiment of the application can continuously collect the radial runout of the mold at three test points and the vibration acceleration of two bearing seats, calculating the peak-to-peak value and root mean square value every 200ms. If the peak-to-peak value ≥ 0.12mm and / or the root mean square value ≥ 2.5 m / s², the centrifuge will be considered successful. 2 If the peak value is less than 0.12 mm and the root mean square value is less than 2.5 m / s, then the conditions for the first pouring are not met. 2 If so, then the conditions for the first pouring are met.

[0058] Furthermore, the first pouring command can include a second pouring speed. This second pouring speed refers to the initial pouring speed of the outer layer of molten metal (the set pouring speed value). The second rotational speed and the second pouring speed are determined based on a preset mapping relationship between the spindle speed, molten metal density, and pouring speed. Specifically, the process of establishing this preset mapping relationship includes:

[0059] First, based on the theory of centrifugal casting, a theoretical pouring speed model is established with the inner diameter of the mold cylinder (i.e., the roller diameter), the spindle speed, and the density of the molten metal as core parameters. The core of this model is to ensure that the dynamic pressure head generated by the molten metal during injection matches the centrifugal pressure generated by the rotation of the mold cylinder, ensuring that the molten metal can spread smoothly and rapidly to the inner wall of the mold cylinder, avoiding splashing or dripping. The theoretical model can be expressed as follows: the pouring speed is directly proportional to the product of the square of the rotational speed, the inner diameter of the mold cylinder, and the density of the molten metal, where the inner diameter of the mold cylinder is the fundamental geometric parameter determining the spatial distribution of the centrifugal pressure field and the area of ​​molten metal spreading.

[0060] Secondly, the theoretical model was calibrated and corrected through preliminary process experiments. For roll products with different diameters, multiple sets of trial pouring at different pouring speeds were conducted at an initial rotation speed determined according to the diameter, and data on the uniformity of the initial solidified layer under corresponding conditions were collected. By analyzing the data, the pouring speed range corresponding to the optimal initial solidified layer (i.e., the most uniform thickness and the fewest defects) was determined under a specific combination of diameter, rotation speed, and molten metal density.

[0061] Finally, a preset mapping relationship is formed through digital mapping (e.g., a three-dimensional data table or fitting formula indexed by diameter and material). This mapping relationship takes the target roll's diameter and material (which determines the density of the molten metal) as joint inputs and can directly output the second rotational speed optimized for that diameter and its corresponding optimal second pouring speed, thereby providing precise process parameter settings for the first pouring and ensuring the initial stability of the outer layer pouring quality.

[0062] For example, when it is necessary to produce a roll with a body diameter of 500mm and an outer layer of nickel-chromium infinite chilled cast iron (density of about 7.2g / cm³), the embodiments of this application can directly match the second rotation speed of 650 r / min and the second pouring speed of 48 kg / s from the preset mapping relationship according to the input target diameter and material, thereby automatically setting the pouring parameters to ensure the accuracy and repeatability of the process.

[0063] Figure 2 This is a schematic diagram of the process for adjusting the spindle speed setting value during the outer layer molten metal pouring stage, as provided in an embodiment of this application.

[0064] like Figure 2As shown, the process for adjusting the spindle speed setting value during the outer layer molten metal pouring stage provided in this application embodiment includes the following steps S201-S205.

[0065] S201: Using the first adjustment frequency, perform Fourier transform on the time-domain data corresponding to the radial runout of the mold within the first time window to obtain the spectral feature vector of the radial runout of the mold.

[0066] It is understandable that the first time window is equal to the period length corresponding to the first adjustment frequency. For example, if the first adjustment frequency is 10Hz, then the corresponding first time window is 0.1s.

[0067] During the pouring of the outer layer of molten metal into the mold, a laser displacement sensor collects the radial runout at three uniformly measured points on the outer wall of the mold at a preset sampling frequency (e.g., 100Hz), forming a continuous data sequence of time-radial runout. Then, in this embodiment, time-domain data within each 0.1s first time window can be extracted at a first adjustment frequency. The time-domain signal is converted to a frequency-domain signal using a Fast Fourier Transform (FFT), and features such as frequency, amplitude, and phase are extracted. A spectral feature vector of the mold's radial runout, matching the dimensions and frequency points, is constructed, realizing the transformation from "time-domain fluctuation description" to "frequency-domain fault feature identification," accurately locating the frequency cause of the radial runout. It can be understood that the time-domain data is a continuous data sequence with time as the horizontal axis and the mold's radial runout as the vertical axis.

[0068] Furthermore, the spectral feature vector is the set of frequency domain features obtained after Fourier transforming the time-domain data of the mold's radial runout. It is a multi-dimensional vector indexed by frequency and with the vibration amplitude at the corresponding frequency as its element, which can be represented as follows: ,in, These are the discrete frequency points after the Fourier transform. This represents the radial vibration amplitude at the corresponding frequency point.

[0069] The specific steps for constructing the spectral feature vector are as follows: extract the frequency values ​​and corresponding amplitudes of all effective frequency points (excluding noise frequencies) after Fourier transform, and construct the spectral feature vector by taking the amplitudes as vector elements in ascending order of frequency.

[0070] In some implementations, the steps for constructing the spectral feature vector can also be as follows: extract the frequency value, corresponding amplitude, and phase of all effective frequency points after Fourier transform, and use these three parameters as vector elements in ascending order of frequency to construct the spectral feature vector.

[0071] Furthermore, the fundamental frequency is the core inherent vibration frequency that drives the mold to produce radial runout when the centrifuge spindle rotates stably at its current speed. It is the dominant frequency with the highest energy proportion in the radial runout frequency domain signal of the mold and best reflects the normal rotational vibration characteristics of the spindle. Its frequency value has a fixed linear correlation with the actual spindle speed and is the core frequency domain characteristic point that distinguishes normal rotational vibration from abnormal fault vibration of the equipment. In other words, the fundamental frequency corresponds to the vibration frequency caused by the number of revolutions per second of the spindle.

[0072] The second harmonic is twice the fundamental frequency of the centrifuge spindle rotation. It is the core characteristic frequency reflecting mechanical abnormalities in the frequency domain signal of the mold radial runout. Its energy ratio change directly reflects the degree of abnormality in the spindle dynamic balance, mold coaxiality, or installation status. It is a key frequency domain characteristic point for determining whether the equipment has a severe vibration fault.

[0073] S202: If the amplitude of the fundamental frequency in the spectral feature vector exceeds the first frequency threshold and the amplitude of the second harmonic in the spectral feature vector does not exceed the second frequency threshold, calculate the axial temperature difference of the mold within the first time window based on the temperature data.

[0074] In this embodiment, the first frequency threshold is a critical value for determining whether the fundamental frequency band vibration amplitude exceeds the standard, and can be 0.08 mm. The second frequency threshold is a critical value for determining whether the second harmonic band vibration amplitude exceeds the standard, and can be 0.03 mm. This embodiment does not specifically limit this value.

[0075] In this embodiment, when the fundamental frequency amplitude exceeds the first frequency threshold and the second harmonic amplitude does not exceed the second frequency threshold, the radial runout of the mold can be determined to be a slight eccentric fluctuation dominated by the fundamental frequency, rather than a violent fluctuation caused by resonance or mechanical failure. In this case, the rotation speed adjustment strategy needs to be further determined in conjunction with the temperature field distribution to avoid misjudgment of adjustment caused by a single frequency index.

[0076] Furthermore, the axial temperature difference of the mold can be equal to the difference between the maximum and minimum values ​​of the average temperatures of the inner walls at the top, middle, and bottom of the mold. Specifically, in this embodiment, temperature data collected within a first time window by thermocouples embedded in the inner walls at the top, middle, and bottom of the mold can be used to calculate the average inner wall temperature at each of the three locations. Then, the difference between the maximum and minimum values ​​of these three averages is calculated to obtain the axial temperature difference of the mold. This parameter directly reflects the temperature uniformity of the axial spreading and solidification of the molten metal in the mold and is the core temperature basis for speed adjustment.

[0077] S203: If the axial temperature difference of the mold does not exceed the first temperature threshold, adjust the spindle speed setting to Z times the current spindle speed, where 0 < Z < 1.

[0078] For example, the first temperature threshold is the critical value for determining whether the axial temperature field of the mold is uniform, specifically 20℃. The value of Z can be 0.9, that is, when the amplitude of the fundamental frequency in the spectral feature vector exceeds 0.08mm, the amplitude of the second harmonic in the spectral feature vector does not exceed 0.03mm, and the axial temperature difference of the mold does not exceed 20℃, the spindle speed setting is reduced to 0.9 times the current spindle speed. In this way, when the axial temperature difference of the mold is ≤20℃, it indicates that the molten metal is evenly spread in the mold axis and there is no obvious segregation in the temperature field. Only a slight eccentricity exists due to the excessive amplitude of the spindle fundamental frequency vibration. The speed is reduced to 0.9 times the current speed. This slightly reduces the centrifugal force and suppresses the radial runout caused by the slight eccentricity of the mold, while avoiding a large drop in speed that would affect the uniform spreading and forming of the outer layer of molten metal, thus balancing the stability of mold operation and the uniformity of molten metal forming.

[0079] It is worth noting that the current spindle speed can be the "current spindle speed" collected last within the first time window.

[0080] S204: When the axial temperature difference of the mold exceeds the first temperature threshold, adjust the spindle speed setting to W times the current spindle speed, where 0 < W < Z.

[0081] For example, the value of W can be 0.7, which means that the spindle speed setting is reduced to 0.7 times the current spindle speed.

[0082] Thus, when the axial temperature difference of the mold is greater than 20°C, it indicates that the axial temperature field of the mold is uneven, and the molten metal has problems such as "too fast solidification at the top or bottom and insufficient spreading in the middle". The rotation speed is reduced to 0.7 times the current speed (0.7 < 0.9). By further reducing the spindle speed, the radial thrashing effect of centrifugal force on the molten metal is reduced, allowing time for heat conduction and homogenization of the axial temperature field of the mold. At the same time, the eccentricity of the mold is further suppressed, achieving the dual goals of temperature field homogenization and mold vibration suppression.

[0083] S205: If the amplitude of the second harmonic in the spectral feature vector exceeds the second frequency threshold, adjust the spindle speed setting to Y times the current spindle speed, where 0 < Y < W.

[0084] For example, the value of Y can be 0.5, which means that the spindle speed setting is reduced to 0.5 times the current spindle speed.

[0085] When the second harmonic amplitude exceeds the second frequency threshold, it indicates that the radial runout of the mold is a severe fluctuation dominated by the second harmonic. This type of fluctuation is mainly caused by mechanical problems such as failure of the centrifuge spindle dynamic balance and severe misalignment between the mold and the spindle. If the high speed is maintained, it will lead to defects such as severely uneven metal spreading and cracks in the outer layer. In this embodiment, the speed can be reduced to 0.5 times the current speed (0.5 < 0.7). By significantly reducing the spindle speed, the high-frequency severe vibration caused by mechanical failure is quickly weakened, avoiding the vibration from being transmitted to the metal in the mold and causing molding defects. At the same time, it prevents mechanical damage to the centrifuge spindle or mold caused by the increased vibration, ensuring the dual safety of production equipment and product molding.

[0086] Furthermore, step S201 may be followed by steps S206-S207.

[0087] S206: If the amplitude of the fundamental frequency in the spectral feature vector does not exceed the first frequency threshold and the amplitude of the second harmonic in the spectral feature vector does not exceed the second frequency threshold, calculate the peak-to-peak value of the radial runout of the mold and the root mean square value of the instantaneous vibration acceleration of the spindle within the first time window.

[0088] This step is an auxiliary determination trigger condition for the spindle speed during the outer layer casting stage. It is only executed when the fundamental frequency amplitude is less than or equal to the first frequency threshold and the second harmonic amplitude is less than or equal to the second frequency threshold. At this time, the frequency domain characteristics of the mold radial runout are normal (no eccentricity, no mechanical fault-related vibration). The embodiment of this application can further verify the overall stability of the mold operation through the statistical characteristics of the original time domain data, avoid the control omissions caused by the single frequency domain determination, realize the dual stability determination of frequency domain characteristics and time domain statistics, and ensure the equipment operation foundation for the outer layer molten metal casting.

[0089] Furthermore, the peak-to-peak value of the mold radial runout refers to the difference between the maximum and minimum values ​​in the original time-domain data of the mold radial runout collected by the laser displacement sensor within a specified time window (such as the first time window). The calculation formula is as follows:

[0090] X pp = X max - X min ;

[0091] in, X pp This represents the peak-to-peak value of the radial runout of the mold. X max This represents the maximum radial runout of the mold within the first time window. X min This represents the minimum radial runout of the mold within the first time window.

[0092] The root mean square (RMS) value of the instantaneous vibration acceleration of the spindle is a statistical value obtained by squaring, averaging, and taking the square root of the original instantaneous vibration acceleration data of the spindle. It can reflect the overall intensity and energy level of the spindle vibration over a certain period of time. The calculation formula is:

[0093] ;

[0094] in, A rms This represents the root mean square value of the instantaneous vibration acceleration of the spindle. f This represents the number of sampling points for the instantaneous acceleration of the spindle vibration within the first time window. a i For the first i The instantaneous vibration acceleration of the spindle acquired in this second acquisition, 0 < i ≤ f .

[0095] S207: If the peak-to-peak value is not within the first preset range, and / or if the root mean square value is not within the second preset range, adjust the spindle speed setting to P times or O times the current spindle speed, Z≤P<1<O.

[0096] The first preset range can be 0.01mm-0.05mm, indicating that the mold runout is stable and meets the casting requirements. The second preset range can be 0.5 m / s. 2 -2.0 m / s 2 This range indicates low spindle vibration energy and no significant vibration transmission to the mold. Therefore, when the peak-to-peak value is within the first preset range and the root mean square value is within the second preset range, it indicates stable operation and no adjustment is required. When the peak-to-peak value is not within the first preset range, and / or when the root mean square value is not within the second preset range, this embodiment of the application can lower the spindle speed setting to P times or O times the current spindle speed, Z≤P<1<O, where P can be equal to 0.95 and O can be equal to 1.05. This embodiment of the application does not specifically limit this.

[0097] In some implementations, 1-P=O-1, and the specific values ​​of P and O can be determined based on the actual situation. This application does not impose specific limitations on this.

[0098] Figure 3 This is a flowchart illustrating the process for determining whether the conditions for a second pouring are met, as provided in an embodiment of this application.

[0099] like Figure 3 As shown in the embodiments of this application, step S300 may include the following steps S301-S303.

[0100] S301: Calculate the inner wall temperature of the outer layer based on the mold thermal imaging map and the outer layer pouring temperature within the second time window using the second adjustment frequency.

[0101] The second adjustment frequency is less than the preset sampling frequency, and the second time window has the same period length as the second adjustment frequency. For example, when the second adjustment frequency is 5Hz, the duration of the second time window is 0.2 seconds.

[0102] Furthermore, the outer layer pouring temperature refers to the set pouring temperature of the outer layer molten metal during pouring, with an example value of 1520℃. The outer layer pouring temperature is a fixed value preset by the process and can be stored in memory as a basic parameter for temperature calculation.

[0103] Furthermore, the outer layer inner wall temperature refers to the temperature of the inner wall of the outer layer molten metal after it has been formed, and is used to determine the timing of the inner layer pouring.

[0104] S302: If the temperature of the outer inner wall is less than or equal to the second temperature threshold, the conditions for the second pouring are determined to be met.

[0105] The second temperature threshold is the semi-solidification critical temperature of the outer metal layer, determined by process testing based on the alloy composition, phase transformation characteristics, and metallurgical bonding requirements of the outer metal molten metal. When the inner wall temperature of the outer layer is less than or equal to the second temperature threshold, it indicates that the outer metal layer has formed a solidified layer of a certain thickness, possessing sufficient structural strength to prevent erosion and mixing during subsequent injection of the inner metal molten metal. Simultaneously, the outer metal layer maintains a suitable high temperature, enabling metallurgical bonding with the subsequently injected inner metal molten metal, avoiding cold shuts and peeling defects between layers. This represents the optimal critical temperature for inner layer casting. For example, the second temperature threshold can be 1250℃, but this application does not specifically limit it.

[0106] S303: When the conditions for the second pouring are met, a second pouring command is generated to set the spindle speed of the centrifuge to the first speed and to inject the inner metal liquid at the first pouring speed; wherein the first speed and the first pouring speed are determined based on a preset mapping relationship between the spindle speed, the metal liquid density and the pouring speed.

[0107] It is understandable that the first rotation speed and the first pouring speed can be determined by querying a preset mapping relationship based on the density of the inner layer of molten metal. The steps for establishing the preset mapping relationship can be referred to the above content, and will not be repeated here.

[0108] Furthermore, step S301 may include the following steps S3031-S3033.

[0109] S3031: Extract the outer wall temperature value from each thermal image of the mold within the second time window using the second adjustment frequency.

[0110] It is understood that the outer wall temperature value referred to here is not the inner wall temperature of the mold obtained by thermocouple measurement at the measuring point, but rather the outer wall temperature value of the mold calculated from the thermal imaging image of the mold. Specifically, the outer wall temperature value is equal to the average of the temperature values ​​of the top, middle, and bottom regions along the mold's axial direction; the temperature value of each region is the arithmetic mean of the temperatures of all thermal imaging pixels within the corresponding region, and the pixel temperature is calculated based on the standard emissivity of the mold material or the real-time acquired emissivity.

[0111] Pixel temperature refers to the temperature of a single pixel in a thermal image of a mold. Each pixel corresponds to a tiny physical area on the outer wall of the mold. For example, when the pixel resolution is 384×288, a single pixel corresponds to an area of ​​approximately 1.5mm×1.5mm on the outer wall of the mold, which is the smallest unit for temperature acquisition.

[0112] The specific steps for determining pixel temperature are as follows:

[0113] ① Acquiring the raw radiation signal: The optical system and infrared detector of the thermal imager focus the infrared radiation emitted from a tiny area (e.g., 1.5mm × 1.5mm) on the outer wall of the mold and convert it into a corresponding raw digital signal value (usually called a "count" or "radiation value"). This value is proportional to the infrared radiation power reaching the detector.

[0114] ② Calculate apparent radiance (uncorrected): The thermal imager's internal processor can convert the raw digital signal from the previous step into an apparent radiance value based on the detector's calibration parameters. At this point, the processor defaults to targeting an ideal blackbody (emissivity ε = 1.0).

[0115] ③ Emissivity Correction: Since the mold steel is not an ideal blackbody, its emissivity ε < 1.0. To obtain the true temperature, the apparent radiance must be corrected using emissivity. The correction formula is as follows:

[0116] True radiance = Apparent radiance / ε;

[0117] Where ε represents emissivity.

[0118] The emissivity ε can be either the standard emissivity of the mold material (ε_std) or the real-time emissivity (ε_real). The standard emissivity of the mold material is a known reference value of the material properties within a specific wavelength and temperature range. The real-time emissivity is an emissivity value that is obtained through auxiliary means (such as using a contact temperature gun to simultaneously measure a calibration point) and is closer to the actual operating conditions.

[0119] In some implementations, ε_real is preferred for computation; if ε_real is unavailable, ε_std is used instead.

[0120] ④ Inversion to Temperature Value: By substituting the corrected true radiance into the inverse formula of Planck's blackbody radiation law, the true surface temperature value of the tiny area on the outer wall of the mold corresponding to that pixel can be calculated. This calculation is usually performed in real time by the software or accompanying algorithm inside the thermal imager.

[0121] In some implementations, embodiments of this application can divide the pixel range of the mold thermal imaging image along the mold axis into a top region, a middle region, and a bottom region according to one-third of the actual total axial length of the mold. That is, the pixel range corresponding to the upper one-third of the length is the top region, the pixel range corresponding to the middle one-third of the length is the middle region, and the pixel range corresponding to the lower one-third of the length is the bottom region.

[0122] S3032: Calculate the interfacial heat transfer coefficient between the outer layer molten metal and the mold based on the outer layer pouring temperature.

[0123] The interfacial heat transfer coefficient (unit: W / (m²·K)) is a physical quantity characterizing the intensity of heat transfer between the molten metal and the inner wall of the mold. It is used to represent the amount of heat transferred per unit time, per unit area, and per unit temperature difference. Specifically, the interfacial heat transfer coefficient h can be calculated using the following empirical formula:

[0124] h = 1500 + 2.5 × T_liquid;

[0125] Where h represents the interfacial heat transfer coefficient and T_liquid represents the outer layer casting temperature.

[0126] For example, when T_liquid=1470℃, h=1500+2.5×1470=5175 W / (m²·K).

[0127] S3033: The outer inner wall temperature is calculated using a finite difference inverse iterative algorithm based on the outer wall temperature, the interfacial heat transfer coefficient, and the outer layer casting temperature.

[0128] This step involves solving a one-dimensional unsteady-state inverse heat conduction problem to infer the measurable outer wall temperature from the unmeasurable inner wall temperature. Specifically, step S3033 may include the following steps S3033a-S3033b.

[0129] S3033a: Discretize the mold wall thickness radially into k nodes; where the first node corresponds to the position of the inner wall surface of the mold, and the kth node corresponds to the position of the outer wall surface of the mold.

[0130] In this embodiment, the Inverse Finite Difference Iteration algorithm is a numerical method for inferring the unmeasurable inner wall temperature from the measurable outer wall temperature. This step involves using the finite difference method to uniformly discretize the mold wall thickness along a radial direction (perpendicular to the outer wall of the mold, which is also the main direction of heat conduction), transforming the continuous mold wall thickness into a discrete sequence of nodes. Each discretized node represents a fixed physical location on the mold wall thickness, and the node temperature is the theoretical temperature at that location.

[0131] For example, the mold wall thickness is 80mm, and the number of discrete nodes k is 10. Then, the first node corresponds to the position on the inner wall surface of the mold, the tenth node corresponds to the position on the outer wall surface of the mold, and the second to ninth nodes are the intermediate nodes inside the mold wall.

[0132] S3033b: Based on the heat conduction equation and the heat flow balance equation, the following steps S1-S5 are executed cyclically. The heat flow balance equation is established based on the interface heat transfer coefficient and the outer layer casting temperature.

[0133] S1: Set the initial value of the iteration number m to 1, and extract the imaging time. t m The corresponding outer wall temperature value is assigned to the k-th node at the imaging time. t m node temperature And based on the k-th node at the imaging time t m node temperature For nodes 1 to (k-1) at the imaging time t m The node temperature is assigned a value; where the imaging time... t m This indicates the imaging time of the m-th thermal image of the mold within the second time window.

[0134] Understandably, within the second time window, the "total number of imaging moments" equals the preset sampling frequency multiplied by the second time window duration, and the "maximum number of iterations" equals the "total number of imaging moments" equals n. For example, with a preset sampling frequency of 100Hz and a second adjustment frequency of 5Hz, n=20. Therefore, the maximum number of iterations is 20, and the iteration process starts at m=1, incrementing by 1 each time until m=n, at which point it terminates.

[0135] The initial value of the iteration number m is 1, which is the first iteration and corresponds to the first imaging time. t 1Specifically, in this embodiment, the outer wall temperature value corresponding to the first thermal image of the mold within the second time window can be determined (assigned) as the k-th node at the first imaging time. t 1 node temperature .

[0136] Furthermore, embodiments of this application can be based on the principle of "linear temperature gradient" from the inner wall to the outer wall of the mold, assuming that the temperature within the mold wall decreases linearly from the first node to the kth node (the inner wall temperature is high, and the outer wall temperature is low). Then, based on the kth node at the first imaging time... t 1 node temperature For nodes 1 to (k-1) at the imaging time t 1 The node temperature is assigned a value.

[0137] For example, with k=10, the node temperature of the 10th node. Given a reference temperature of 950℃, based on the principle of "linear temperature gradient," the initial reference temperature of the first node is first set to 1200℃ (this is just an example value; the actual temperature is determined by the outer pouring temperature, and can be equal to the outer pouring temperature). Then, based on the radially discrete node spacing of the mold wall thickness, the linear temperature decrease step size between the first and tenth nodes is calculated, i.e., the temperature difference between two adjacent nodes is approximately 27.78℃. Subsequently, the initial temperatures of the second to ninth nodes are assigned sequentially according to this decrease step size. The temperature of the second node... The value is 1200 - 27.78 = 1172.22℃, which is the temperature of the third node. The temperature is 1144.44℃, the temperature of the 4th node. The temperature is 1116.66℃, the temperature of the 5th node. The temperature is 1088.88℃, the temperature of the 6th node. The temperature is 1061.10℃, the temperature of the 7th node. The temperature is 1033.32℃, the temperature of the 8th node. The temperature is 1005.54℃, the temperature of the 9th node. The temperature reached 977.76℃, marking the completion of imaging at nodes 1 through 9. t 1 The initial temperature of each node is assigned a value, forming a linear temperature gradient distribution from the inner wall to the outer wall of the mold, providing an initial reference value that conforms to the heat conduction law for subsequent iterative calculations.

[0138] S2: Increment the iteration count m by 1 to extract the imaging time. t m The corresponding outer wall temperature value is assigned to the k-th node at the imaging time.t m node temperature And each node at the previous imaging time t m-1 node temperature Substituting into the heat conduction equation, we obtain the values ​​of the 2nd to (k-1)th nodes at the imaging time. t m node temperature ,2≤ j ≤k-1.

[0139] It is understandable that incrementing the iteration count m by 1 results in the second iteration, at which point the imaging time can be extracted. t 2 The corresponding outer wall temperature value, i.e., the outer wall temperature value corresponding to the second thermal image of the mold within the second time window, is determined (assigned) as the k-th node at the second imaging time. t 2 node temperature Then, the previous imaging time can be... t 1 Node temperature of each node Substituting into the heat conduction equation, we obtain the values ​​of the 2nd to (k-1)th nodes at the imaging time. t 2 node temperature .

[0140] In this embodiment, the heat conduction equation is specifically a one-dimensional unsteady-state heat conduction equation, and the specific construction process is as follows:

[0141] The formula describing the heat transfer over time along the mold wall thickness is as follows:

[0142] ;

[0143] in, ρ For mold density, c p The specific heat capacity of the mold under constant pressure. λ mold The thermal conductivity of the mold, This is the first derivative of temperature with respect to time. Let be the second derivative of temperature with respect to radial position.

[0144] Furthermore, the derivative is approximated using the finite difference method.

[0145] First time derivative (forward difference): , indicating that the r-th node is t +Δ t Temperature difference over time;

[0146] Second derivative in space (intermediate difference): , where represents the temperature gradient between the r-th node and its neighboring nodes.

[0147] Substituting the derivative approximation into Formula 1, we get: ;

[0148] Calculate the coefficient terms based on the example parameters. The simplified final heat conduction equation is: ;

[0149] In this way, you only need to know t time r -1、 r , r +1 Node temperature of three adjacent nodes ( ), then it can be calculated t +Δ t Time of the first r Node temperature of each node .

[0150] Therefore, when calculating the imaging time of the 2nd to (k-1)th nodes... t m node temperature To calculate the final result (i.e., the temperature of the second node at the previous imaging time), simply substitute the node temperature of the associated nodes (e.g., nodes 1-3) at the previous imaging time into the heat conduction equation. t m (node ​​temperature).

[0151] S3: Place the second node at the imaging time t m node temperature Substituting into the heat flow balance equation, we obtain the first node at the imaging time. t m node temperature If m < n, proceed to step S2; if m = n, proceed to step S4; where n is the total number of thermal images of the mold within the second time window.

[0152] Specifically, the heat flow balance equation is:

[0153] ;

[0154] in, Indicates the first node at the imaging time t m The node temperature, Indicates the second node at the imaging time tm The node temperature, a Indicates the first coefficient. b This indicates the second coefficient.

[0155] For example, based on instance parameters, a The value can be 0.7377. b The value can be 393.44.

[0156] Understandably, when m < n, we need to jump to step S2 to continue the iteration; when m = n, we can jump to step S4, that is, end the iteration.

[0157] S4: Based on the relative error formula, calculate the first node at the imaging time. t m node temperature Rather than in the previous imaging moment t m-1 node temperature The relative error.

[0158] The formula for relative error is:

[0159] ;

[0160] in, err Indicates relative error. Indicates the first node at the imaging time t m The node temperature, This indicates that the first node was at the previous imaging time. t m-1 The node temperature.

[0161] S5: If the relative error is less than the error threshold, assign the temperature of the outer inner wall to the value of the first node at the imaging time. t m node temperature If the relative error is greater than or equal to the error threshold, the temperature of the outer inner wall is assigned to a preset value and the loop ends; wherein the preset value is greater than the second temperature threshold.

[0162] For example, the error threshold can be 0.03, but this application does not specifically limit it. If err If the value is less than 0.03, convergence is determined, the loop ends, and the node temperature at the last moment can be output (the temperature of the first node at the imaging moment). t m node temperature () is used as the temperature of the inner wall of the outer layer.

[0163] Furthermore, for example, the second temperature threshold can be 1250℃, then the preset value can be 1300℃, and the specific value can be determined based on the actual situation. This application embodiment does not specifically limit this. If err If the value is ≥0.03, it is considered non-converged, the loop ends, and the preset value of 1300℃ can be output as the outer inner wall temperature.

[0164] This allows the loop to end and the outer inner wall temperature to be calculated. Based on this temperature, it can be determined whether the conditions for the second pouring are met. If the outer inner wall temperature is less than or equal to the second temperature threshold, the conditions for the second pouring can be confirmed.

[0165] Figure 4 This is a schematic diagram illustrating the process of adjusting the pouring speed setting value during the inner layer molten metal pouring stage, as provided in an embodiment of this application.

[0166] like Figure 4 As shown in the embodiments of this application, step S400 may include the following steps S401-S403.

[0167] S401: Calculate the axial temperature difference and circumferential temperature difference of the mold based on the temperature data within the third time window using the third adjustment frequency; wherein, the third adjustment frequency is less than the preset sampling frequency, the period length corresponding to the third time window and the third adjustment frequency is equal, the axial temperature difference of the mold is equal to the difference between the maximum and minimum values ​​of the average temperature of the inner wall of the top, middle, and bottom of the mold, and the circumferential temperature difference of the mold is equal to the difference between the maximum and minimum values ​​of the inner wall temperature at different circumferential positions of the same height section of the mold.

[0168] In this embodiment, the mold inner wall temperature data within a third time window is captured using a third adjustment frequency (lower than the preset sampling frequency) as the period, and the two temperature differences are calculated respectively:

[0169] Axial temperature difference of mold: Calculate the average temperature of the inner walls of the top, middle and bottom of the mold, and take the difference between the maximum and minimum values ​​to reflect the temperature uniformity of the mold along the length direction;

[0170] Circumferential temperature difference of mold: Select a cross section at the same height of the mold and calculate the difference between the maximum and minimum values ​​of the inner wall temperature at different circumferential positions of the cross section. This reflects the temperature uniformity of the mold at the same height along the circumferential direction.

[0171] S402: When the axial temperature difference of the mold is greater than the third temperature threshold, adjust the pouring speed setting to Q times the current pouring speed, where 0 < Q < 1.

[0172] If the calculated axial temperature difference of the mold exceeds the third temperature threshold, it indicates that the temperature deviation of the mold along the length direction is too large. Adjust the current pouring speed setting of the inner layer molten metal to Q times the original speed (current pouring speed, i.e. the last "current pouring speed" in the third time window). By slightly reducing the speed, the molten metal can be fully spread and dissipated, thus improving the axial temperature uniformity.

[0173] For example, the third temperature threshold can be 30°C, and the value of Q can be 0.8. This application embodiment does not specifically limit this.

[0174] S403: When the axial temperature difference of the mold is less than or equal to the third temperature threshold and the circumferential temperature difference of the mold is greater than the fourth temperature threshold, adjust the pouring speed setting to L times the current pouring speed, where 0 < L < Q.

[0175] If the axial temperature difference of the mold is less than or equal to the third temperature threshold (the axial temperature is normal), but the circumferential temperature difference is greater than the fourth temperature threshold (the circumferential temperature deviation is too large), it indicates that the temperature uniformity of the mold in the circumferential direction is poor, and this abnormality has an impact on the molding quality. Adjust the current pouring speed setting to L times the original speed (the last "current pouring speed" in the third time window) (0 < L < Q). By slightly reducing the speed, the molten metal can be fully and evenly spread under the action of centrifugal force, thereby improving the circumferential temperature distribution.

[0176] For example, the fourth temperature threshold can be 15°C, and the value of L can be 0.9. This application embodiment does not specifically limit this.

[0177] Figure 5 This is a flowchart illustrating the process for determining whether the centrifuge stop condition is met, as provided in an embodiment of this application.

[0178] like Figure 5 As shown, step S500 may include the following steps S501-S503.

[0179] S501: Generate a mold temperature change curve based on the temperature data within the fourth time window using the fourth adjustment frequency; wherein, the fourth adjustment frequency is less than the preset sampling frequency, and the period length of the fourth time window is equal to that corresponding to the fourth adjustment frequency.

[0180] In this embodiment, the mold temperature data within the fourth time window is captured using a fourth adjustment frequency (lower than the preset sampling frequency) as the period. Based on this batch of data, a mold temperature change curve is generated, which intuitively reflects the dynamic change trend of the mold temperature within the fourth time window.

[0181] In some implementations, the mold temperature change curve can refer to the temperature change curve of the top inner wall, middle inner wall, or bottom inner wall of the mold, or it can refer to the average change curve of the temperature of the top inner wall, middle inner wall, and bottom inner wall of the mold. This application does not specifically limit this.

[0182] S502: Input the mold temperature change curve into the temperature calculation model to determine the internal temperature.

[0183] In this embodiment, the generated mold temperature change curve can be input into a pre-trained temperature calculation model, which can then deduce and calculate the internal temperature of the roll based on the temperature change pattern outside the mold.

[0184] Furthermore, the temperature estimation model is a hybrid model trained based on "heat conduction mechanism + solidification characteristics + experimental data". It takes the mold temperature change curve as input features and integrates process factors such as mold material, molten metal solidification characteristics, and centrifugal casting heat exchange law. After training and iterative optimization with multiple sets of process test data, it can accurately fit the correlation between the external temperature change of the mold and the internal temperature of the roll. It can quickly deduce the actual internal temperature of the roll, which cannot be directly detected, based on the real-time temperature change trend of the mold, providing accurate temperature data support for determining the stop condition of the centrifuge.

[0185] For example, the training data comes from multiple sets of process test data. The input data includes: the mold temperature change curve collected in real time during the test, and the corresponding process parameters (rotation speed, pouring speed, etc.); the output labels include: the internal temperature of the roll measured by differential thermal analysis, embedded thermocouples, etc. (as the real value for model training); for different roll sizes, molten metal materials, and combinations of process parameters, multiple sets of trial pouring are carried out, and the model parameters are iteratively optimized after data collection to ensure that the mapping relationship between input (mold temperature) and output (internal temperature of the roll) is accurate.

[0186] S503: The centrifuge shutdown condition is determined when the internal temperature is below the solidus line; the solidus line is determined by differential thermal analysis based on the material composition of the inner molten metal.

[0187] In this embodiment, the internal temperature of the roll obtained by deduction can be compared with the solidus line temperature of the inner metal liquid: if the internal temperature drops below the solidus line, it indicates that the metal liquid inside the roll has been fully solidified and has sufficient structural strength. At this time, it is determined that the centrifuge stop condition is met, and the centrifuge can be controlled to gradually reduce the speed until it stops running.

[0188] The solidus line is the critical temperature at which the inner layer of molten metal completely solidifies. It is determined by the specific material composition of the inner layer of molten metal and can be determined through differential thermal analysis experiments.

[0189] It should be added that differential thermal analysis is a commonly used thermal analysis method for determining the critical temperature of phase transition of materials. During the experiment, the sample of the inner metal liquid is heated or cooled simultaneously with a thermally inert reference, and the temperature difference between the two is detected in real time. When the sample undergoes phase transitions such as solidification or melting, it will release or absorb heat, resulting in a temperature difference with the reference. By recording the temperature difference change curve, the critical temperature at which the metal liquid of the material solidifies into a solid state, i.e., the solidus temperature, can be accurately determined.

[0190] Figure 6 Metallographic diagram of the heat-crack resistant steel roll provided in the embodiments of this application.

[0191] like Figure 6 As shown in the metallographic photographs of the heat-crack-resistant steel rolls, the hardness test results corresponding to the metallographic images reveal that, for nine test points selected from the three busbars, the Shore Scleroscope Hardness D-scale (HSD) hardness of the outer layer is as follows: 82, 82, 83; 83, 82, 82; 83, 83, 83. The outer layer hardness uniformity is good, with the maximum hardness difference being less than 2 HSD. The HSD hardness of the ductile iron core is 40 and 39, both meeting the predetermined hardness requirements.

[0192] As can be seen from the above, the embodiments of this application provide a method for producing heat-crack-resistant steel rolls. This method can accurately determine the timing of outer and inner layer pouring through multi-frequency temperature monitoring and algorithmic deduction throughout the centrifugal casting process, and can dynamically adjust the spindle speed setting and pouring speed setting. Simultaneously, the internal temperature of the roll can be deduced using a temperature calculation model to determine the timing of centrifuge shutdown, and the process parameters can be dynamically controlled throughout the process, effectively suppressing the initiation of hot cracks in the rolls.

[0193] Figure 7 This is a schematic diagram of the production apparatus for heat-resistant crack-resistant steel rolls provided in the embodiments of this application.

[0194] like Figure 7 As shown in the embodiment of this application, a production apparatus for heat-crack resistant steel rolls is provided. The heat-crack resistant steel rolls are a two-layer composite structure formed by two centrifugal casting processes when a centrifuge drives a mold to rotate. The structure includes an outer layer and an inner layer. The apparatus includes:

[0195] The data acquisition module 1001 is configured to acquire multi-dimensional sensing data at a preset sampling frequency during the operation of the centrifuge; wherein, the multi-dimensional sensing data includes one or more of the following: mold radial runout, instantaneous vibration acceleration of the spindle, current pouring speed, current spindle speed, and temperature data;

[0196] The first adjustment module 1002 is configured to: adjust the spindle speed setting value based on multi-sensor data at a first adjustment frequency during the process of pouring the outer layer of molten metal into the mold, until the pouring of the outer layer of molten metal is completed.

[0197] The first judgment module 1003 is configured to: after the outer layer of molten metal is poured, determine whether the conditions for the second pouring are met based on multi-sensor data at the second adjustment frequency, and if the conditions for the second pouring are met, generate a second pouring command to inject the inner layer of molten metal into the mold.

[0198] The second adjustment module 1004 is configured to: adjust the pouring speed setting value based on multi-sensor data at a third adjustment frequency during the pouring of the inner layer molten metal, until the pouring of the inner layer molten metal is completed.

[0199] The second judgment module 1005 is configured to: after the inner layer of molten metal is poured, determine whether the centrifuge stop condition is met based on multi-sensor data using the fourth adjustment frequency, and generate a centrifuge stop command if the centrifuge stop condition is met, so as to control the centrifuge spindle to stop.

[0200] In one possible implementation, the first adjustment module 1002 is further configured to: perform a Fourier transform on the time-domain data corresponding to the radial runout of the mold within a first time window at a first adjustment frequency to obtain a spectral feature vector of the radial runout of the mold; wherein the first adjustment frequency is less than a preset sampling frequency, and the first time window has the same period length as the first adjustment frequency; the time-domain data is a continuous data sequence with time as the horizontal axis and the radial runout of the mold as the vertical axis; when the amplitude of the fundamental frequency in the spectral feature vector exceeds a first frequency threshold, and the amplitude of the second harmonic in the spectral feature vector does not exceed a second frequency threshold, the first time is calculated based on the temperature data. The axial temperature difference of the mold within the window; wherein, the axial temperature difference of the mold is equal to the difference between the maximum and minimum values ​​of the average temperature of the top inner wall, the average temperature of the middle inner wall, and the average temperature of the bottom inner wall of the mold; when the axial temperature difference of the mold does not exceed the first temperature threshold, the spindle speed setting is adjusted to Z times the current spindle speed, 0 < Z < 1; when the axial temperature difference of the mold exceeds the first temperature threshold, the spindle speed setting is adjusted to W times the current spindle speed, 0 < W < Z; when the amplitude of the second harmonic in the spectral feature vector exceeds the second frequency threshold, the spindle speed setting is adjusted to Y times the current spindle speed, 0 < Y < W.

[0201] In one possible implementation, the first adjustment module 1002 is further configured to: calculate the peak-to-peak value of the mold radial runout and the root mean square value of the instantaneous vibration acceleration of the spindle within the first time window, provided that the amplitude of the fundamental frequency in the spectral feature vector does not exceed the first frequency threshold and the amplitude of the second harmonic in the spectral feature vector does not exceed the second frequency threshold; and adjust the spindle speed setting to P times or O times the current spindle speed, Z≤P<1<O, provided that the peak-to-peak value is not within the first preset range and / or the root mean square value is not within the second preset range.

[0202] In one possible implementation, the temperature data includes a mold thermal image, which is generated by collecting infrared radiation energy along the side of the mold; the first judgment module 1003 is further configured to: calculate the outer inner wall temperature based on the mold thermal image and the outer pouring temperature within a second time window at a second adjustment frequency; wherein the second adjustment frequency is less than a preset sampling frequency, the second time window is equal to the period length corresponding to the second adjustment frequency, and the outer pouring temperature refers to the pouring temperature setting value of the outer metal liquid during pouring; if the outer inner wall temperature is less than or equal to a second temperature threshold, determine that the second pouring condition is met; if the second pouring condition is met, generate a second pouring command to set the centrifuge spindle speed setting value to a first speed and inject the inner metal liquid at a first pouring speed; wherein the first speed and the first pouring speed are determined based on a preset mapping relationship between the spindle speed, the metal liquid density, and the pouring speed.

[0203] In one possible implementation, the first judgment module 1003 is further configured to: extract the outer wall temperature value from each thermal imaging image of the mold within the second time window at a second adjustment frequency; wherein the outer wall temperature value is equal to the average of the temperature values ​​of the top region, the middle region, and the bottom region along the mold axis; the temperature value of each region is the arithmetic mean of the temperatures of all thermal imaging pixels in the corresponding region, and the pixel temperature is calculated based on the standard emissivity of the mold material or the real-time acquired emissivity; calculate the interfacial heat transfer coefficient between the outer layer molten metal and the mold based on the outer layer pouring temperature; and calculate the outer inner wall temperature using a finite difference inverse iterative algorithm based on the outer wall temperature value, the interfacial heat transfer coefficient, and the outer layer pouring temperature.

[0204] In one possible implementation, the first judgment module 1003 is further configured to: discretize the mold wall thickness radially into k nodes; wherein the first node corresponds to the position of the inner wall surface of the mold, and the kth node corresponds to the position of the outer wall surface of the mold; cyclically execute the following steps S1-S5 based on the heat conduction equation and the heat flow balance equation, wherein the heat flow balance equation is established based on the interface heat transfer coefficient and the outer layer casting temperature; S1: set the initial value of the iteration number m to 1, and extract the imaging time. t mThe corresponding outer wall temperature value is assigned to the k-th node at the imaging time. t m node temperature And based on the k-th node at the imaging time t m node temperature For nodes 1 to (k-1) at the imaging time t m The node temperature is assigned a value; where the imaging time... t m S1: Indicates the imaging time of the m-th thermal image of the mold within the second time window; S2: Increment the iteration number m by 1 and extract the imaging time. t m The corresponding outer wall temperature value is assigned to the k-th node at the imaging time. t m node temperature And each node at the previous imaging time t m-1 node temperature Substituting into the heat conduction equation, we obtain the values ​​of the 2nd to (k-1)th nodes at the imaging time. t m node temperature ,2≤ j ≤k-1; S3: Place the second node at the imaging time t m node temperature Substituting into the heat flow balance equation, we obtain the first node at the imaging time. t m node temperature If m < n, proceed to step S2; if m = n, proceed to step S4; where n is the total number of thermal images of the mold within the second time window; S4: Based on the relative error formula, calculate the value of the first node at the imaging time. t m node temperature Rather than in the previous imaging moment t m-1 node temperature The relative error; S5: If the relative error is less than the error threshold, assign the temperature of the outer inner wall to the value of the first node at the imaging time. t m node temperature If the relative error is greater than or equal to the error threshold, the temperature of the outer inner wall is assigned to a preset value and the loop ends; wherein the preset value is greater than the second temperature threshold.

[0205] In one possible implementation, the second adjustment module 1004 is further configured to: calculate the axial temperature difference and circumferential temperature difference of the mold based on temperature data within a third time window at a third adjustment frequency; wherein, the third adjustment frequency is less than the preset sampling frequency, the third time window has the same period length as the third adjustment frequency, the axial temperature difference of the mold is equal to the difference between the maximum and minimum values ​​of the average temperature of the inner wall at the top, middle, and bottom of the mold, and the circumferential temperature difference of the mold is equal to the difference between the maximum and minimum values ​​of the inner wall temperature at different circumferential positions at the same height section of the mold; when the axial temperature difference of the mold is greater than a third temperature threshold, the pouring speed setting is adjusted to Q times the current pouring speed, 0 < Q < 1; when the axial temperature difference of the mold is less than or equal to the third temperature threshold and the circumferential temperature difference of the mold is greater than a fourth temperature threshold, the pouring speed setting is adjusted to L times the current pouring speed, 0 < L < Q.

[0206] In one possible implementation, the second judgment module 1005 is further configured to: generate a mold temperature change curve based on temperature data within a fourth time window at a fourth adjustment frequency; wherein the fourth adjustment frequency is less than a preset sampling frequency, and the fourth time window has the same period length as the fourth adjustment frequency; input the mold temperature change curve into a temperature calculation model to determine the internal temperature; and determine that the centrifuge stop condition is met when the internal temperature is below the solidus line; the solidus line is determined by differential thermal analysis experiments based on the material composition of the inner metal liquid.

[0207] In one possible implementation, the apparatus provided in this application embodiment further includes a third judgment module 1006, configured to: when the spindle speed of the centrifuge spindle is set to a second speed, calculate the peak-to-peak value of the radial runout of the mold and the root mean square value of the instantaneous vibration acceleration of the spindle within a fifth time window at a fifth adjustment frequency; wherein the fifth adjustment frequency is less than a preset sampling frequency, and the fifth time window is equal to the period length corresponding to the fifth adjustment frequency; when the peak-to-peak value is less than a first runout threshold and the root mean square value is less than a first acceleration threshold, generate a first pouring command to inject the outer layer of molten metal at a second pouring speed; wherein the second speed and the second pouring speed are determined based on a preset mapping relationship between the spindle speed, the molten metal density, and the pouring speed.

[0208] In a specific implementation, the present invention also provides a computer storage medium, wherein the computer storage medium may store a program, which, when executed, may include some or all of the steps in the various embodiments of the production method for heat-crack-resistant steel rolls provided by the present invention. The storage medium may be a magnetic disk, optical disk, read-only memory (ROM), or random access memory (RAM), etc.

[0209] It is readily understood that, based on the several embodiments provided in this application, those skilled in the art can combine, split, or reorganize the embodiments of this application to obtain other embodiments, none of which exceed the protection scope of this application.

[0210] The above detailed embodiments further illustrate the purpose, technical solution, and beneficial effects of the embodiments of this application. It should be understood that the above are merely specific embodiments of the embodiments of this application and are not intended to limit the protection scope of the embodiments of this application. Any modifications, equivalent substitutions, improvements, etc., made on the basis of the technical solutions of the embodiments of this application should be included within the protection scope of the embodiments of this application.

Claims

1. A method for producing heat-crack resistant steel rolls, characterized in that, The heat-crack-resistant steel roll is a two-layer composite structure formed by two centrifugal casting processes while the die is rotated in a centrifuge, including an outer layer and an inner layer; the method includes: During the operation of the centrifuge, multi-dimensional sensing data is acquired at a preset sampling frequency; wherein, the multi-dimensional sensing data includes one or more of the following: mold radial runout, instantaneous vibration acceleration of the spindle, current pouring speed, current spindle speed, and temperature data; During the process of pouring the outer layer of molten metal into the mold, the spindle speed setting value is adjusted based on the multi-sensor data at a first adjustment frequency until the pouring of the outer layer of molten metal is completed. After the outer layer of molten metal is poured, the second pouring condition is determined based on the multi-sensor data using the second adjustment frequency. If the second pouring condition is met, a second pouring command is generated to inject the inner layer of molten metal into the mold. During the pouring of the inner layer molten metal, the pouring speed setting value is adjusted based on the multi-sensor data at a third adjustment frequency until the pouring of the inner layer molten metal is completed. After the inner layer of molten metal is poured, the centrifuge stop condition is determined based on the multi-sensor data using the fourth adjustment frequency. If the centrifuge stop condition is met, a centrifuge stop command is generated to control the centrifuge spindle to stop.

2. The method for producing heat-crack resistant steel rolls according to claim 1, characterized in that, The step of adjusting the spindle speed setting value based on the multi-sensor data at a first adjustment frequency during the process of pouring the outer layer of molten metal into the mold includes: Using the first adjustment frequency, a Fourier transform is performed on the time-domain data corresponding to the radial runout of the mold within the first time window to obtain the spectral feature vector of the radial runout of the mold; wherein, the first adjustment frequency is less than the preset sampling frequency, and the first time window is equal to the period length corresponding to the first adjustment frequency; the time-domain data is a continuous data sequence with time as the horizontal axis and the radial runout of the mold as the vertical axis; If the amplitude of the fundamental frequency in the spectral feature vector exceeds a first frequency threshold and the amplitude of the second harmonic in the spectral feature vector does not exceed a second frequency threshold, the axial temperature difference of the mold within the first time window is calculated based on the temperature data; wherein, the axial temperature difference of the mold is equal to the difference between the maximum and minimum values ​​of the average temperature of the inner wall at the top of the mold, the average temperature of the inner wall at the middle, and the average temperature of the inner wall at the bottom of the mold. If the axial temperature difference of the mold does not exceed the first temperature threshold, the spindle speed setting value is adjusted to Z times the current spindle speed, where 0 < Z < 1; If the axial temperature difference of the mold exceeds the first temperature threshold, the spindle speed setting value is adjusted to W times the current spindle speed, where 0 < W < Z; If the amplitude of the second harmonic in the spectral feature vector exceeds the second frequency threshold, the spindle speed setting value is adjusted to Y times the current spindle speed, where 0 < Y < W.

3. The method for producing heat-crack resistant steel rolls according to claim 2, characterized in that, After the step of performing a Fourier transform on the time-domain data corresponding to the radial runout of the mold within the first time window using the first adjustment frequency to obtain the spectral feature vector of the radial runout of the mold, the method further includes: If the amplitude of the fundamental frequency in the spectral feature vector does not exceed the first frequency threshold and the amplitude of the second harmonic in the spectral feature vector does not exceed the second frequency threshold, calculate the peak-to-peak value of the radial runout of the mold and the root mean square value of the instantaneous vibration acceleration of the spindle within the first time window. If the peak value is not within the first preset range, and / or if the root mean square value is not within the second preset range, the spindle speed setting value is adjusted to P times or O times the current spindle speed, where Z≤P<1<O.

4. The method for producing heat-crack resistant steel rolls according to claim 1, characterized in that, The temperature data includes a mold thermal imaging image, which is generated by collecting infrared radiation energy along the side of the mold; the step of determining whether the second pouring conditions are met based on the multi-sensor data using a second adjustment frequency, and generating a second pouring command if the second pouring conditions are met, includes: The outer inner wall temperature is calculated based on the mold thermal imaging image and the outer pouring temperature within the second time window using the second adjustment frequency; wherein, the second adjustment frequency is less than the preset sampling frequency, the second time window is equal to the period length corresponding to the second adjustment frequency, and the outer pouring temperature refers to the pouring temperature setting value of the outer metal liquid during pouring. If the temperature of the outer inner wall is less than or equal to the second temperature threshold, the conditions for the second pouring are determined to be met. When the conditions for the second pouring are met, the second pouring command is generated to set the centrifuge spindle speed setting to the first speed and to inject the inner layer of molten metal at the first pouring speed; wherein the first speed and the first pouring speed are determined based on a preset mapping relationship between the spindle speed, the density of the molten metal and the pouring speed.

5. The method for producing heat-crack resistant steel rolls according to claim 4, characterized in that, The step of calculating the outer inner wall temperature based on the mold thermal imaging image and the outer pouring temperature within the second time window using the second adjustment frequency includes: Using the second adjustment frequency, the outer wall temperature value is extracted from each of the thermal imaging images of the mold within the second time window; wherein, the outer wall temperature value is equal to the average of the temperature values ​​of the top region, the middle region, and the bottom region along the mold axis; the temperature value of each region is the arithmetic mean of the temperatures of all thermal imaging pixels in the corresponding region, and the pixel temperature is calculated based on the standard emissivity of the mold material or the real-time acquired emissivity; The interfacial heat transfer coefficient between the outer layer molten metal and the mold is calculated based on the outer layer pouring temperature. Based on the outer wall temperature, the interface heat transfer coefficient, and the outer layer casting temperature, the outer inner wall temperature is calculated using a finite difference inverse iterative algorithm.

6. The method for producing heat-crack resistant steel rolls according to claim 5, characterized in that, The step of calculating the inner wall temperature of the outer layer using a finite difference inverse iterative algorithm based on the outer wall temperature, the interface heat transfer coefficient, and the outer layer casting temperature includes: The mold wall thickness is discretized radially into k nodes; where the first node corresponds to the position of the inner wall surface of the mold, and the kth node corresponds to the position of the outer wall surface of the mold. The following steps S1-S5 are executed cyclically based on the heat conduction equation and the heat flow balance equation, wherein the heat flow balance equation is established based on the interface heat transfer coefficient and the outer layer casting temperature. S1: Set the initial value of the iteration number m to 1, and extract the imaging time. t m The corresponding outer wall temperature value is assigned to the value of the k-th node at the imaging time. t m node temperature And based on the k-th node at the imaging time t m node temperature For the first node to the (k-1)th node at the imaging time t m The node temperature is assigned a value; where the imaging time... t m This indicates the imaging time of the m-th thermal image of the mold within the second time window; S2: Increment the iteration count m by 1 to extract the imaging time. t m The corresponding outer wall temperature value is assigned to the value of the k-th node at the imaging time. t m node temperature And each node at the previous imaging time t m-1 node temperature Substituting into the heat conduction equation, we obtain the imaging times of the 2nd to (k-1)th nodes. t m node temperature ,2≤ j ≤k-1; S3: Place the second node at the imaging time t m node temperature Substituting into the heat flow balance equation, we obtain the first node at the imaging time. t m node temperature If m < n, proceed to step S2; if m = n, proceed to step S4; where n is the total number of thermal images of the mold within the second time window. S4: Based on the relative error formula, calculate the first node at the imaging time. t m node temperature Rather than in the previous imaging moment t m-1 node temperature The relative error; S5: If the relative error is less than the error threshold, assign the temperature of the outer inner wall to the value of the first node at the imaging time. t m node temperature If the relative error is greater than or equal to the error threshold, the temperature of the outer inner wall is assigned a preset value and the loop ends; wherein the preset value is greater than the second temperature threshold.

7. The method for producing heat-crack resistant steel rolls according to claim 1, characterized in that, The step of adjusting the pouring speed setting value based on the multi-sensor data at a third adjustment frequency during the pouring of the inner layer molten metal includes: Using the third adjustment frequency, the axial temperature difference and circumferential temperature difference of the mold are calculated based on the temperature data within the third time window; wherein, the third adjustment frequency is less than the preset sampling frequency, the third time window has the same period length as the third adjustment frequency, the axial temperature difference of the mold is equal to the difference between the maximum and minimum values ​​of the average temperature of the inner wall of the top, middle, and bottom of the mold, and the circumferential temperature difference of the mold is equal to the difference between the maximum and minimum values ​​of the inner wall temperature at different circumferential positions of the same height section of the mold; If the axial temperature difference of the mold is greater than the third temperature threshold, the pouring speed setting value is adjusted to Q times the current pouring speed, where 0 < Q < 1; If the axial temperature difference of the mold is less than or equal to the third temperature threshold and the circumferential temperature difference of the mold is greater than the fourth temperature threshold, the pouring speed setting value is adjusted to L times the current pouring speed, where 0 < L < Q.

8. The method for producing heat-crack resistant steel rolls according to claim 1, characterized in that, The step of determining whether the centrifuge shutdown condition is met based on the multi-sensor data at a fourth adjustment frequency after the inner layer molten metal is poured includes: A mold temperature change curve is generated based on the temperature data within a fourth time window using the fourth adjustment frequency; wherein the fourth adjustment frequency is less than the preset sampling frequency, and the fourth time window is equal in length to the period corresponding to the fourth adjustment frequency; The temperature change curve of the mold is input into the temperature calculation model to determine the internal temperature; The centrifuge is deemed to have met the stop condition when the internal temperature is below the solidus line; the solidus line is determined by differential thermal analysis based on the material composition of the inner molten metal.

9. The method for producing heat-crack resistant steel rolls according to claim 1, characterized in that, Before the step of adjusting the spindle speed setting value based on the multi-sensor data at a first adjustment frequency during the process of pouring the outer layer of molten metal into the mold, the method further includes: With the centrifuge spindle speed set to the second speed, the peak-to-peak value of the mold radial runout and the root mean square value of the instantaneous vibration acceleration of the spindle are calculated within the fifth time window at the fifth adjustment frequency; wherein, the fifth adjustment frequency is less than the preset sampling frequency, and the fifth time window is equal to the period length corresponding to the fifth adjustment frequency; When the peak-to-peak value is less than the first fluctuation threshold and the root mean square value is less than the first acceleration threshold, a first pouring command is generated to inject the outer layer of molten metal at a second pouring speed; wherein, the second rotation speed and the second pouring speed are determined based on a preset mapping relationship between the spindle rotation speed, molten metal density and pouring speed.

10. A production apparatus for heat-crack resistant steel rolls, characterized in that, The heat-crack-resistant steel roll is a two-layer composite structure formed by two centrifugal casting processes while the die is rotated in a centrifuge, including an outer layer and an inner layer; the device includes: The data acquisition module is configured to acquire multi-dimensional sensing data at a preset sampling frequency during centrifuge operation; wherein the multi-dimensional sensing data includes one or more of the following: mold radial runout, instantaneous vibration acceleration of the spindle, current pouring speed, current spindle speed, and temperature data. The first adjustment module is configured to: adjust the spindle speed setting value at a first adjustment frequency based on the multi-sensor data during the process of pouring the outer layer of molten metal into the mold, until the pouring of the outer layer of molten metal is completed; The first judgment module is configured to: after the outer layer of molten metal is poured, determine whether the conditions for the second pouring are met based on the multi-sensor data at a second adjustment frequency, and if the conditions for the second pouring are met, generate a second pouring command to inject the inner layer of molten metal into the mold. The second adjustment module is configured to: during the process of pouring the inner layer molten metal, adjust the pouring speed setting value based on the multi-sensor data at a third adjustment frequency until the pouring of the inner layer molten metal is completed. The second judgment module is configured to: after the inner layer of molten metal is poured, determine whether the centrifuge stop condition is met based on the multi-sensor data at the fourth adjustment frequency, and generate a centrifuge stop command if the centrifuge stop condition is met, so as to control the centrifuge spindle to stop.

Images