A Big Data-Based Method and System for Optimizing Heat Treatment Process Parameters of Mold Steel

Abstract

This invention belongs to the field of heat treatment process control technology, specifically involving a method and system for optimizing heat treatment process parameters of mold steel based on big data. The method includes: obtaining the actual percentage content of multiple key elements in the mold steel raw material, and calculating the material quenching sensitivity index based on the weighting coefficients of each element; obtaining the total furnace loading mass, total external surface area of ​​the workpiece, and the envelope volume after stacking, and calculating the effective thermal damping coefficient of the furnace loading; detecting the measured value of the specific saturation magnetization of a standard sample after cooling or initial tempering, and determining the target specific saturation magnetization; and calculating the tempering temperature correction value using a compensation model based on the above indices, coefficients, and magnetic deviation ratio, thereby obtaining the optimized tempering set temperature. This invention solves the technical problems of poor process adaptability caused by batch fluctuations in raw material composition, inability to quantitatively assess furnace loading thermal inertia, and the inability to adjust the process in a closed-loop manner due to the lack of non-destructive testing methods.

Description

Technical Field

[0001] This invention relates to the field of heat treatment process control technology. More specifically, this invention relates to a method and system for optimizing heat treatment process parameters of mold steel based on big data. Background Technology

[0002] High-end mold steels, such as H13, Dievar, and LD, are crucial materials in industrial production, and their performance directly determines the service life and processing quality of molds. Vacuum heat treatment is a key process for improving the performance of mold steels. By precisely controlling the heating temperature, holding time, and cooling rate, an ideal phase transformation occurs within the material's internal structure. Currently, the setting of heat treatment process parameters for mold steels is mainly based on the standard process curves corresponding to the material grade. Technicians typically rely on personal experience, making simple minor adjustments to the process parameters according to the workpiece dimensions.

[0003] However, existing heat treatment process control methods have significant limitations. Even though the content of key alloying elements such as carbon, chromium, molybdenum, vanadium, manganese, and silicon in the raw materials of the same grade of mold steel from different heat batches meets national standards, there are still objective batch-to-batch fluctuations. These minute compositional deviations are enough to cause the continuous cooling transformation curve to shift left or right, and traditional fixed-parameter processes cannot adapt to these fluctuations, easily leading to uneven workpiece hardness or cracking. At the same time, the cooling rate in a vacuum furnace is not only affected by gas pressure but also heavily dependent on the physical state of the furnace load. Existing technologies lack quantitative calculation models for loading mass, stacking density, and porosity, resulting in huge differences in core cooling rates between fully loaded and half-loaded workpieces, easily causing soft spots or incomplete quenching.

[0004] Furthermore, traditional methods for controlling retained austenite mainly rely on metallographic microscopy or X-ray diffraction, which involve long inspection cycles and destructive sampling, making it impossible to obtain data in real time during heat treatment. Due to the lack of rapid non-destructive testing methods based on physical quantities, existing technologies cannot infer the adequacy of tempering based on the actual microstructure transformation, and process parameters cannot be adjusted in a closed loop, thus affecting the final dimensional stability and performance of the mold steel. Summary of the Invention

[0005] To address the technical problems of poor adaptability of heat treatment processes due to batch fluctuations in raw material composition, uncontrollable cooling rates due to the inability to quantitatively assess furnace loading thermal inertia, and the inability to achieve closed-loop process adjustment due to the lack of non-destructive testing methods, this invention provides solutions in the following aspects.

[0006] In a first aspect, the present invention provides a method for optimizing heat treatment process parameters of mold steel based on big data, including: obtaining the actual percentage content of multiple key elements in the mold steel raw material, and calculating the material quenching sensitivity index based on the deviation between the actual percentage content and the standard composition center value, combined with the weight coefficient of each element.

[0007] The total loading mass of the currently loaded workpiece, the total outer surface area of ​​the workpiece, and the enveloping volume after stacking are obtained. Based on the total loading mass, the total outer surface area of ​​the workpiece, the enveloping volume after stacking, and the loading void ratio, the effective thermal damping coefficient of the loading is calculated.

[0008] The specific saturation magnetization of the standard sample after cooling or initial tempering was measured using a magnetic analysis device, and the target specific saturation magnetization corresponding to the mold steel raw material was determined.

[0009] Based on the material quenching sensitivity index, the effective thermal damping coefficient of the furnace, and the deviation ratio between the measured value of the specific saturation magnetization and the target specific saturation magnetization, the correction value of the tempering temperature is calculated using a compensation model to obtain the optimized tempering set temperature.

[0010] This invention constructs three independent quantitative characterization parameters: material quenching sensitivity index, effective thermal damping coefficient of furnace loading, and deviation ratio based on magnetic feedback. It simultaneously incorporates raw material composition detection data, loading physical state, and degree of microstructure transformation into the compensation calculation of tempering temperature, thereby achieving coupled correction of multi-dimensional process factors and ensuring the consistency of heat treatment quality for different batches and under different furnace loading conditions.

[0011] Preferably, the material quenching sensitivity index satisfies the following relationship:

[0012] ;

[0013] In the formula, Indicates the quenching sensitivity index of the material; This indicates the total number of key alloying elements involved in the calculation; Indicates the first The actual percentage content of each element; This indicates the corresponding standard component center value; Indicates the first The weight coefficient of each element.

[0014] This invention calculates the material quenching sensitivity index by weighted superposition of the deviations between the actual content of each key alloying element and the center value of the standard composition. This allows the index to retain the directionality and magnitude of the deviations of multiple elements, thereby accurately reflecting the comprehensive change in the hardenability of the current batch of material relative to the standard material. This provides a quantitative basis at the material level for targeted correction of tempering temperature.

[0015] Preferably, the weighting coefficient is the hardenability influence weighting coefficient, which is determined as follows: A historical process database containing the alloy element content of historical batches of materials and the corresponding mapping relationship between quenching cooling rate and hardness gradient is established; a multiple regression analysis is performed with the content of each alloy element as the independent variable and the hardenability index as the dependent variable to obtain the contribution rate of each element to hardenability; and the hardenability influence weighting coefficient of each element is determined based on the contribution rate, wherein the hardenability influence weighting coefficient corresponding to elements that improve hardenability is positive, and the hardenability influence weighting coefficient corresponding to elements that reduce hardenability is negative.

[0016] Preferably, the effective thermal damping coefficient of the furnace loading satisfies the following relationship:

[0017] ;

[0018] In the formula, Indicates the effective thermal damping coefficient of the furnace; Indicates the total mass of the furnace charge; Indicates the baseline full load mass; Indicates the volume of the stacked envelope; Indicates the volume of the effective heating zone of the furnace; This represents the total external surface area of ​​the workpiece; Represents the reference surface area constant; Indicates the loading void ratio; Represented by natural constant An exponential function with base 0.

[0019] This invention constructs an effective thermal damping coefficient for furnace loading, encompassing three dimensions: mass, space occupancy, and gas convection channels. This is achieved by multiplying the ratio of the total furnace loading mass to the baseline full-load mass, the ratio of the stacked envelope volume to the effective heating zone volume of the furnace, and an exponential term determined by the total outer surface area of ​​the workpiece and the loading void ratio. The exponential term is highly sensitive to changes in void ratio, accurately capturing the hindering effect of stacking density on gas convection heat transfer. This solves the problem that traditional methods cannot quantitatively characterize the impact of heat conduction under complex furnace loading conditions.

[0020] Preferably, the loading void ratio is calculated as follows: the void volume is obtained by subtracting the workpiece volume from the stacked envelope volume, and the void volume is then divided by the stacked envelope volume to obtain the loading void ratio, and the loading void ratio must meet a preset minimum air duct limit.

[0021] Preferably, the optimized tempering set temperature satisfies the following relationship:

[0022] ;

[0023] In the formula, This indicates the optimized tempering set temperature; Indicates the base tempering temperature; Indicates the material composition compensation coefficient; Represent the natural logarithm function; Indicates the quenching sensitivity index of the material; Indicates the furnace load compensation coefficient; Indicates the effective thermal damping coefficient of the furnace; This represents the gain coefficient adjusted by magnetic feedback. Indicates the target specific saturation magnetization; This represents the measured value of the specific saturation magnetization.

[0024] This invention addresses material composition deviation, furnace loading thermal damping deviation, and magnetic feedback deviation by adding the base tempering temperature to three independent compensation terms. The material composition compensation term uses the logarithm of the material quenching sensitivity index to smoothly adjust the compensation amount as the hardenability of the material increases or decreases. The furnace loading load compensation term takes the excess of the effective furnace loading thermal damping coefficient relative to the reference value as input, directly responding to the absolute increment of loading thermal damping. The magnetic feedback compensation term takes the deviation ratio between the measured specific saturation magnetization and the target specific saturation magnetization as input, converting the non-destructive magnetic detection results into a temperature correction amount, forming a real-time closed-loop feedback on the actual degree of microstructure transformation.

[0025] Preferably, the process further includes adjusting the tempering temperature parameters of the heat treatment equipment to the optimized tempering temperature after calculating the optimized tempering set temperature; performing the tempering process; and detecting the measured value of the specific saturation magnetization again after tempering. If the measured value of the specific saturation magnetization still does not reach the target specific saturation magnetization, the secondary correction value of the tempering temperature is calculated iteratively again based on the new deviation ratio.

[0026] Preferably, the key alloying elements include carbon, chromium, molybdenum, vanadium, manganese, and silicon, and the actual percentage content is obtained by multi-point spark excitation detection of the current batch of mold steel raw materials using a photoelectric direct-reading spectrometer.

[0027] This invention utilizes a photoelectric direct-reading spectrometer for multi-point spark excitation detection, enabling rapid acquisition of the precise content of six key alloying elements—carbon, chromium, molybdenum, vanadium, manganese, and silicon—in the current batch of raw materials. This covers the element combinations that have the most significant impact on the hardenability of mold steel, ensuring the comprehensiveness and reliability of the material quenching sensitivity index calculation results.

[0028] Preferably, the target specific saturation magnetization is determined based on the theoretical specific saturation magnetization when the residual austenite content corresponding to the grade of the mold steel raw material is lower than a preset upper limit; the measured value of the specific saturation magnetization is obtained by non-destructive testing of the furnace-fed standard sample using a magnetic analyzer.

[0029] This invention establishes a strict quantitative correspondence between magnetic physical quantities and microstructure transformation indicators by using the theoretical specific saturation magnetization when the residual austenite content is lower than a preset upper limit as the target specific saturation magnetization. By adopting non-destructive testing with furnace-fed standard samples, the invention achieves real-time inversion of the degree of martensite transformation, replacing traditional destructive metallographic testing, and providing timely and accurate data support for closed-loop iterative adjustment of process parameters.

[0030] Secondly, the present invention provides a big data-based optimization system for heat treatment process parameters of mold steel, including a processor and a memory. The memory stores computer program instructions, and when the computer program instructions are executed by the processor, the above-mentioned big data-based optimization method for heat treatment process parameters of mold steel is implemented.

[0031] By adopting the above technical solution, the above-mentioned method for optimizing the heat treatment process parameters of mold steel based on big data is generated into a computer program and stored in a memory so that it can be loaded and executed by a processor. In this way, a terminal device can be made based on the memory and the processor for convenient use.

[0032] The beneficial effects of this invention are as follows:

[0033] This invention transforms raw material composition detection data into a material quenching sensitivity index, characterizes the loading physical state as an effective thermal damping coefficient for furnace loading, and uses the residual austenite transformation deviation obtained from magnetic non-destructive testing as real-time feedback. This enables the tempering temperature compensation model to synchronously couple disturbances in three dimensions: material, environment, and microstructure transformation. Thus, it dynamically generates optimal temperature parameters under multi-source process fluctuations, achieving refined and intelligent control of the mold steel heat treatment process and ensuring the continuous achievement of batch-to-batch hardness uniformity and microstructure stability standards. Attached Figure Description

[0034] Figure 1 This is a flowchart illustrating the method for optimizing process parameters of mold steel heat treatment based on big data according to the present invention;

[0035] Figure 2 This is a schematic diagram illustrating the comparison and analysis of the cooling rate response of the present invention and the prior art under different furnace loads;

[0036] Figure 3 This is a schematic diagram illustrating the distribution of raw material composition fluctuations and quenching sensitivity in this invention;

[0037] Figure 4 This is a schematic diagram illustrating the correlation between specific saturation magnetization and residual austenite content in this invention. Detailed Implementation

[0038] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are some embodiments of the present invention, but not all embodiments.

[0039] This invention discloses a method for optimizing heat treatment process parameters of mold steel based on big data, referring to... Figure 1 This includes steps S1-S4:

[0040] S1. Obtain the actual percentage content of various key elements in the mold steel raw material, and calculate the material quenching sensitivity index, which characterizes the hardenability of the material, based on the deviation between the actual percentage content and the standard composition center value, combined with the weight coefficient of each element.

[0041] It should be noted that, although the content of trace alloying elements such as carbon, chromium, molybdenum, vanadium, manganese, and silicon in the raw materials of the same grade of mold steel in different heats meets national standards, there are still objective batch-to-batch fluctuations. These minor compositional deviations are sufficient to cause the continuous cooling transformation curve to shift to the left or right, thereby affecting hardenability and final hardness distribution. Therefore, this embodiment constructs a material quenching sensitivity index to evaluate the quenching difficulty of the current batch of material relative to the standard material, providing data basis for subsequent process parameter adjustments.

[0042] Specifically, before the mold steel workpieces are heat-treated in the furnace, a photoelectric direct-reading spectrometer is used to perform multi-point spark excitation detection on the current batch of mold steel raw materials to obtain the actual percentage content of key alloying elements such as carbon, chromium, molybdenum, vanadium, manganese, and silicon. The photoelectric direct-reading spectrometer is a well-known component detection device and will not be described in detail here.

[0043] Furthermore, based on the deviation between the actual percentage content and the central value of the standard composition, and combined with the weighting coefficient of the hardenability influence of each element, the quenching sensitivity index of the material is calculated. Material quenching sensitivity index Satisfying the relation:

[0044]

[0045] In the formula, is a dimensionless material quenching sensitivity index, with a baseline value of 1; This indicates the total number of key alloying elements involved in the calculation; The first value obtained by the spectrometer measurement The actual percentage content of each element; This indicates the standard composition center value of the corresponding grade material; Indicates the first The hardenability of each element affects the weighting coefficient.

[0046] In other words, the material quenching sensitivity index Using 1 as a baseline, when the actual content of each element deviates from the standard value, it is weighted and superimposed onto the baseline value through its respective hardenability influence weighting coefficient; when the content of a certain element is high and the hardenability influence weighting coefficient is positive, An increase in hardenability indicates that the hardenability of the corresponding batch of material is enhanced; conversely, a decrease in hardenability indicates that the hardenability is enhanced. A decrease indicates reduced hardenability. Therefore, the material's quenching sensitivity index... It can comprehensively reflect the superimposed effect of fluctuations in multiple elements on hardenability.

[0047] It should be noted that hardenability affects the weighting coefficient. The determination method is as follows: A historical process database is established, containing the alloy element content of historical batches of materials and the corresponding mapping relationship between quenching cooling rate and hardness gradient; a multiple regression analysis is performed with the content of each alloy element as the independent variable and the hardenability index as the dependent variable to obtain the contribution rate of each element to hardenability, and the hardenability influence weight coefficient of each element is determined based on the contribution rate. Among them, for elements that improve hardenability, such as carbon and molybdenum, hardenability affects the weighting coefficient. Take positive values; for elements that reduce hardenability, the hardenability influence weighting coefficient is... Take a negative value. In other embodiments, the hardenability influence weighting coefficient can be set according to the actual application scenario and requirements. Furthermore, hardenability affects the weighting coefficient. The absolute value ranges from [0.1, 2], and the hardenability influences the weighting coefficient. If the absolute value is too small, the material quenching sensitivity index Insufficient response to elemental deviations makes it impossible to distinguish the differences in hardenability between different batches of materials; excessively high elemental deviations result in a high material quenching sensitivity index. Excessive sensitivity to minute fluctuations in composition leads to drastic fluctuations in the tempering temperature correction.

[0048] The following is a calculation example: For H13 steel, the standard composition center value The following settings were used: carbon 0.4%, chromium 5%, and molybdenum 1.3%. The actual percentage content of the raw materials in the current batch was measured using a spectrometer. The components are: carbon 0.42%, chromium 4.9%, and molybdenum 1.35%. Hardenability affects the weighting coefficient. Historical data regression analysis was used to determine the weighting coefficient for the influence of carbon hardenability. The hardenability of chromium affects the weighting coefficient, which is 0.5. The hardenability of molybdenum affects the weighting coefficient, which is 0.4. It is 0.6.

[0049] Substitute the formula for the material quenching sensitivity index into the equation for calculation:

[0050] Carbon's bias contribution: ;

[0051] Chromium's contribution to the deviation: ;

[0052] Molybdenum's bias contribution: ;

[0053] Material quenching sensitivity index .

[0054] Calculation results The value is 1.04, which is greater than the benchmark value of 1. This indicates that the current batch of materials has a higher overall hardenability than the standard value due to the higher carbon and molybdenum content. The continuous cooling transformation curve has shifted to the right, and the tempering temperature needs to be adjusted appropriately in subsequent processes to balance the microstructure transformation.

[0055] Thus, by collecting raw material composition data and calculating the material quenching sensitivity index, the directional influence of each element's deviation is preserved, and the combined effect of multiple element fluctuations on hardenability can be accurately calculated, providing material-level data basis for subsequent process adjustments.

[0056] S2. Obtain the total loading mass of the currently loaded workpiece, the total outer surface area of ​​the workpiece, and the envelope volume after stacking. Based on the total loading mass, the total outer surface area of ​​the workpiece, the envelope volume after stacking, and the loading void ratio, calculate the effective thermal damping coefficient of the loading, which characterizes the degree of gas convection heat transfer resistance.

[0057] It should be noted that the cooling rate in a vacuum furnace is not only affected by gas pressure but also heavily dependent on the physical state of the load. Existing technologies lack quantitative calculation models for loading mass, stacking density, and porosity, resulting in significant differences in core cooling rates between fully loaded and half-loaded workpieces, which can easily lead to soft spots or incomplete quenching. Therefore, this embodiment constructs an effective thermal damping coefficient for the load to assess the specific impact of the loading state on heat conduction.

[0058] Specifically, the total mass of the workpieces currently loaded into the vacuum furnace is obtained through a weighing sensor at the bottom of the furnace. Based on the CAD 3D model of the workpiece and the actual furnace loading arrangement, the total outer surface area of ​​the workpiece was calculated. and the envelope volume after stacking The envelope volume refers to the spatial volume occupied by the entire workpiece stack, and the solid volume of the workpiece is denoted as... .

[0059] Furthermore, loading void ratio The calculation method is as follows: through the envelope volume after stacking. Subtract the solid volume of the workpiece Obtain the void volume, then divide the void volume by the stacked envelope volume. Obtain the loading void ratio ,Right now And loading void ratio A preset minimum airflow limit must be met to ensure unobstructed gas convection channels. In other embodiments, the minimum airflow limit can be set according to the actual application scenario and requirements, and the value range of the minimum airflow limit is [0.05, 0.3]. If the minimum airflow limit is set too small, the lower limit of the allowable loading void ratio will be too low. When the stacking is too dense, the gas convection channel will be severely insufficient, resulting in a sharp drop in the cooling rate of the workpiece core, which is very easy to produce soft spots or incomplete quenching. If the minimum airflow limit is set too large, the loading void ratio requirement will be too high, the number of workpieces that can be loaded in a single furnace will be greatly reduced, and the production efficiency will be reduced.

[0060] Based on this, the effective thermal damping coefficient of the furnace can be calculated using the aforementioned physical quantities. The effective thermal damping coefficient of the furnace satisfies the following relationship:

[0061]

[0062] In the formula, Indicates the effective thermal damping coefficient of the furnace; This indicates the total mass of the furnace charge, obtained by actual measurement using a weighing sensor. This represents the baseline full-load mass, which is the equipment's rated constant. Indicates the volume of the stacked envelope; This represents the effective heating zone volume of the furnace, which is a rated constant of the equipment. This represents the total external surface area of ​​the workpiece; The reference surface area constant is used to balance the dimensions of the independent variable of the exponential term. Dimensionless, reference surface area constant The reference surface area constant is determined based on the inner wall area of ​​the effective heating zone of the furnace. Therefore, when the total outer surface area of ​​the workpiece... With loading void ratio When the product equals the area of ​​the furnace inner wall, the independent variable of the exponential term is 1. The effective thermal damping coefficient for the furnace is 2.72. At a moderate level; when the workpieces are densely stacked, resulting in much smaller At this point, the exponential term increases sharply, accurately reflecting the severe impediment of dense stacking to gas convection heat transfer. In other embodiments, the reference surface area constant can be determined based on the actual furnace geometry of the equipment. ; Indicates the loading void ratio; Represented by natural constant An exponential function with base 0.

[0063] In other words, the effective thermal damping coefficient of the furnace. It is composed of the product of the mass ratio, the volume ratio, and the porosity index; when the total mass of the furnace is... As the mass ratio increases, the thermal damping also increases; when the envelope volume after stacking increases... When the volume ratio increases, the loading space occupies more space, and the thermal damping increases; in the porosity index, when the total surface area of ​​the workpiece increases... Reduce or load void ratio When the porosity index decreases, the internal value of the porosity exponent increases, and the exponent value increases sharply. This means that the denser the packing, the more severely gas convection and heat transfer are hindered, and the greater the thermal damping. Therefore, the effective thermal damping coefficient of the furnace charge is... It can comprehensively reflect the degree of influence of loading status on heat transfer.

[0064] The following is a calculation example. The equipment's rated parameters are: baseline full load mass. The effective heating zone volume of the furnace is 1000 kg. 2m 3 Reference surface area constant 10m 2 The reference surface area constant is determined based on the total surface area of ​​the inner wall of the effective heating zone of the vacuum furnace. The current measured loading data is: total loaded mass. The envelope volume after stacking is 800 kg. 1.5m 3 Workpiece solid volume 0.9m 3 Total outer surface area of ​​the workpiece 20m 2 .

[0065] The first step is to calculate the loading void ratio. :

[0066] ;

[0067] Loading void ratio The value is 0.4, which is greater than the minimum air duct limit of 0.1, and meets the requirement for unobstructed gas convection channels.

[0068] The second step is to substitute the effective thermal damping coefficient of the furnace. Relational calculations:

[0069] Quality ratio: ;

[0070] Volume ratio: ;

[0071] Porosity index term internally: ;

[0072] Index value: ;

[0073] .

[0074] Effective thermal damping coefficient of furnace loading The value is 2.094, which is greater than 1, indicating that although the total mass of the furnace is not fully loaded, the heat conduction is severely hindered due to the high stacking density, and process compensation is required.

[0075] Thus, by constructing a multidimensional thermal damping model that includes mass, volume, and porosity, the specific impact of loading conditions on heat conduction can be accurately assessed. In particular, the impact of stacking density on gas convection heat transfer is keenly captured through the exponential term, solving the problem that traditional methods cannot quantitatively calculate furnace loading thermal damping and achieving accurate prediction and compensation of cooling rate.

[0076] S3. Use magnetic analysis equipment to detect the measured value of the specific saturation magnetization of the standard sample after cooling or initial tempering, and determine the target specific saturation magnetization corresponding to the mold steel raw material.

[0077] It should be noted that traditional methods for controlling retained austenite mainly rely on metallographic microscopy or X-ray diffraction, which involve long testing cycles and destructive sampling, making it impossible to obtain data in real time during heat treatment. Martensite in mold steel is strongly magnetic, while retained austenite is paramagnetic, meaning it is nearly non-magnetic. Martensite and retained austenite differ significantly in their magnetic properties. Therefore, this embodiment utilizes magnetic physical quantities for non-destructive testing, replacing traditional destructive testing methods, to quickly infer the content of retained austenite in the microstructure.

[0078] Specifically, after quenching and cooling, a magnetic analyzer was used to perform non-destructive testing on the furnace-fed standard sample to obtain the measured value of the specific saturation magnetization. .

[0079] Furthermore, the target specific saturation magnetization The specific saturation magnetization is determined based on the theoretical value of the mold steel raw material when the retained austenite content is below a preset upper limit. For example, for H13 steel, when the retained austenite content is below 3%, the theoretical specific saturation magnetization is 18A. m 2 / kg, then the target specific saturation magnetization 18A m 2 / kg. In other embodiments, the target specific saturation magnetization can be set according to the actual application scenario and the grade of the mold steel raw material. .

[0080] Assuming that the measured value of the specific saturation magnetization of the current standard sample is... 15A m 2 / kg. Measured value of specific saturation magnetization. Below the target specific saturation magnetization This indicates that the standard sample contains a large amount of non-magnetic retained austenite, which has not been completely transformed into martensite. Therefore, it is necessary to increase the tempering temperature to promote the decomposition of the retained austenite.

[0081] Non-destructive testing of magnetic physical quantities can quickly back-calculate the content of residual austenite in the microstructure, providing real-time feedback data for closed-loop process control.

[0082] S4. Based on the material quenching sensitivity index, the effective thermal damping coefficient of the furnace, and the deviation ratio between the measured value of the specific saturation magnetization and the target specific saturation magnetization, the correction value of the tempering temperature is calculated using the compensation model to obtain the optimized tempering set temperature.

[0083] In this step, the optimized tempering set temperature is calculated based on the collected parameters. The relationship is as follows:

[0084]

[0085] In the formula, This indicates the optimized tempering set temperature, in °C. This indicates the base tempering temperature, which is the tempering temperature recommended in the standard process instruction. Indicates the material composition compensation coefficient; Represent the natural logarithm function; Indicates the quenching sensitivity index of the material; Indicates the furnace load compensation coefficient; Indicates the effective thermal damping coefficient of the furnace; This represents the gain coefficient adjusted by magnetic feedback. Indicates the target specific saturation magnetization; This represents the measured value of the specific saturation magnetization.

[0086] In other words, the optimized tempering set temperature Based on the basic tempering temperature It consists of three compensation items: material composition compensation item. Response Material Quenching Sensitivity Index Changes, when When the value is greater than 1, indicating that the material has relatively high hardenability, A positive value indicates that the compensation term increases the tempering temperature to balance the microstructure transformation; the furnace load compensation term... Response to effective thermal damping coefficient of furnace loading Changes, when When the value is greater than 1, indicating excessive furnace thermal damping, the compensation term increases the tempering temperature to ensure heat penetration; magnetic feedback compensation term. Measured value of response ratio saturation magnetization saturation magnetization relative to the target The deviation ratio, when compared with the measured value of saturation magnetization. Below the target specific saturation magnetization When the deviation ratio is positive, the compensation term increases the tempering temperature to promote the decomposition of residual austenite. Thus, the three compensation terms dynamically correct the tempering temperature from three dimensions: material, environment, and testing feedback.

[0087] It should be noted that each compensation coefficient is set based on a historical process experience database. For example, the material composition compensation coefficient... The furnace load compensation coefficient is 50℃. At 20℃, the gain coefficient is adjusted by magnetic feedback. The temperature is 100°C. In other embodiments, the material composition compensation coefficient can be set according to the actual application scenario and requirements. And the material composition compensation coefficient The value range is [10, 100]℃; the furnace load compensation coefficient can be set according to the actual application scenario and requirements. And the furnace load compensation coefficient The value range is [5, 50]℃; the magnetic feedback adjustment gain coefficient can be set according to the actual application scenario and requirements. And the gain coefficient is adjusted by magnetic feedback. The value range is [30, 200]℃. If the above compensation coefficient is too small, the temperature correction will be insufficient and it will not be able to effectively compensate for the effects of material fluctuations, furnace loading thermal damping, or residual austenite deviations; if it is too large, the temperature correction will be excessive and may cause the workpiece to overheat and cause grain coarsening.

[0088] Material quenching sensitivity index The effective thermal damping coefficient for furnace loading is 1.04. The target specific saturation magnetization is 2.094. 18A m 2 / kg, measured value of specific saturation magnetization 15A m2 / kg, base tempering temperature The temperature is 560℃. Substitute this into the optimized formula for the tempering setpoint temperature for the final calculation:

[0089] Material composition compensation items:

[0090] ℃;

[0091] This means the material has slightly high hardenability, requiring a slight adjustment in temperature.

[0092] Furnace load compensation items:

[0093] ℃;

[0094] The furnace thermal resistance is too high, so the temperature needs to be increased to ensure that the core of the workpiece is heated through.

[0095] Magnetic feedback compensation item:

[0096] ℃;

[0097] This means that the residual austenite content is too high, and the temperature needs to be increased to promote decomposition.

[0098] Total correction and final temperature:

[0099] ℃.

[0100] The final system output command sets the tempering temperature to 600.5℃. In engineering practice, this can be rounded down to 600℃ or 601℃ to ensure that the performance of this batch of workpieces meets the standards.

[0101] Furthermore, the optimized tempering set temperature was calculated. Then, the tempering temperature parameters of the heat treatment equipment were adjusted to the optimized tempering set temperature. Perform a tempering process, and then re-test the measured value of the specific saturation magnetization after tempering. If compared with the measured value of saturation magnetization The target specific saturation magnetization has not yet been achieved. Then, based on the new deviation ratio, the secondary correction value of the tempering temperature is calculated again through iteration.

[0102] Thus, by constructing a multi-parameter coupled inverse decision-making model, it is possible to simultaneously respond to fluctuations in material composition, differences in physical loading environment, and the degree of actual microstructure transformation, dynamically generating optimal process parameters, thereby achieving intelligent and refined control of the heat treatment process.

[0103] Reference Figure 2 The horizontal axis represents the effective thermal damping coefficient of the furnace. The vertical axis represents the effective cooling rate of the workpiece core. This varies with the effective thermal damping coefficient during furnace loading. With increasing load, i.e., heavier or denser loading, the cooling rate at existing measured points decreases rapidly and nonlinearly, with many points falling into the low-speed zone; however, the measured points using the method of this invention, through adaptive compensation, maintain a stable cooling rate within the target process window over a wider load range.

[0104] Reference Figure 3 The horizontal axis represents the batch sequence of raw material furnace numbers, and the vertical axis represents the material quenching sensitivity index. Normal batches within the standard process window cluster around the baseline value of 1, while high-sensitivity batches, i.e., difficult-to-quench batches, exhibit a higher material quenching sensitivity index. High quenching sensitivity index for batches with low sensitivity, i.e., batches prone to cracking. The result is relatively low. This directly reflects the invention's ability to keenly detect minute differences in composition between different batches of materials.

[0105] Reference Figure 4 The horizontal axis represents the tempering temperature, the left vertical axis represents the retained austenite content, and the right vertical axis represents the specific saturation magnetization. As the tempering temperature increases, the retained austenite gradually decomposes and its content decreases; at the same time, the specific saturation magnetization increases synchronously, and the two show a negative correlation. The optimal tempering temperature point is marked in the figure, which verifies the scientific nature of using magnetic feedback for process optimization.

[0106] This invention also discloses a big data-based optimization system for heat treatment process parameters of mold steel, including a processor and a memory. The memory stores computer program instructions, and when the computer program instructions are executed by the processor, the big data-based optimization method for heat treatment process parameters of mold steel according to this invention is implemented.

[0107] The system also includes other components well known to those skilled in the art, such as communication buses and communication interfaces, the settings and functions of which are known in the art and will not be described in detail here.

[0108] In the description of this specification, "multiple" or "several" means at least two, such as two, three or more, unless otherwise expressly and specifically defined.

Claims

1. A method for optimizing heat treatment process parameters of die steel based on big data, characterized in that, include: The actual percentage content of several key elements in the mold steel raw material is obtained. Based on the deviation between the actual percentage content and the standard composition center value, and combined with the weighting coefficient of each element, the material quenching sensitivity index is calculated, satisfying the following relationship: ； In the formula, Indicates the quenching sensitivity index of the material; This indicates the total number of key alloying elements involved in the calculation; Indicates the first The actual percentage content of each element; This indicates the corresponding standard component center value; Indicates the first The weight coefficients of the elements; The weighting coefficient is the hardenability influence weighting coefficient, which is determined as follows: A historical process database containing the alloy element content of historical batches of materials and the corresponding mapping relationship between quenching cooling rate and hardness gradient is established; a multiple regression analysis is performed with the content of each alloy element as the independent variable and the hardenability index as the dependent variable to obtain the contribution rate of each element to hardenability, and the hardenability influence weighting coefficient of each element is determined according to the contribution rate. The hardenability influence weighting coefficient of elements that improve hardenability is taken as a positive value, and the hardenability influence weighting coefficient of elements that reduce hardenability is taken as a negative value. Obtain the total loading mass of the currently loaded workpieces, the total outer surface area of ​​the workpieces, and the enveloping volume after stacking. Based on the total loading mass, the total outer surface area of ​​the workpieces, the enveloping volume after stacking, and the loading void ratio, calculate the effective thermal damping coefficient of the loading, satisfying the following relationship: ； In the formula, Indicates the effective thermal damping coefficient of the furnace; Indicates the total mass of the furnace charge; Indicates the baseline full load mass; Indicates the volume of the stacked envelope; Indicates the volume of the effective heating zone of the furnace; This represents the total external surface area of ​​the workpiece; Represents the reference surface area constant; Indicates the loading void ratio; Represented by natural constant An exponential function with base 0; The specific saturation magnetization of standard samples after cooling or initial tempering was measured using magnetic analysis equipment, and the target specific saturation magnetization corresponding to the mold steel raw material was determined. Based on the material quenching sensitivity index, the effective thermal damping coefficient of the furnace, and the deviation ratio between the measured value and the target specific saturation magnetization, a compensation model is used to calculate the correction value of the tempering temperature, resulting in the optimized tempering set temperature, which satisfies the following relationship: ； In the formula, This indicates the optimized tempering set temperature; Indicates the base tempering temperature; Indicates the material composition compensation coefficient; Represents the natural logarithm function; Indicates the furnace load compensation coefficient; This represents the gain coefficient adjusted by magnetic feedback. Indicates the target specific saturation magnetization; This represents the measured value of the specific saturation magnetization.

2. The method for optimizing heat treatment process parameters of mold steel based on big data according to claim 1, characterized in that, The loading void ratio is calculated as follows: the void volume is obtained by subtracting the workpiece volume from the stacked envelope volume, and then the void volume is divided by the stacked envelope volume to obtain the loading void ratio. The loading void ratio must meet the preset minimum air duct limit.

3. The method for optimizing heat treatment process parameters of mold steel based on big data according to claim 1, characterized in that, It also includes adjusting the tempering temperature parameters of the heat treatment equipment to the optimized tempering set temperature after calculating the optimized tempering set temperature; performing the tempering process; and detecting the measured value of the specific saturation magnetization again after tempering. If the measured value of the specific saturation magnetization still does not reach the target specific saturation magnetization, the secondary correction value of the tempering temperature is calculated again based on the new deviation ratio.

4. The method for optimizing heat treatment process parameters of mold steel based on big data according to claim 1, characterized in that, The key alloying elements include carbon, chromium, molybdenum, vanadium, manganese, and silicon. The actual percentage content is obtained by multi-point spark excitation detection of the current batch of mold steel raw materials using a photoelectric direct-reading spectrometer.

5. The method for optimizing heat treatment process parameters of mold steel based on big data according to claim 1, characterized in that, The target specific saturation magnetization is determined based on the theoretical specific saturation magnetization when the residual austenite content corresponding to the grade of the mold steel raw material is lower than a preset upper limit; the measured value of the specific saturation magnetization is obtained by non-destructive testing of the furnace-fed standard sample using a magnetic analyzer.

6. A big data-based optimization system for heat treatment process parameters of mold steel, characterized in that, include: A processor and a memory, wherein the memory stores computer program instructions that, when executed by the processor, implement the method for optimizing heat treatment process parameters of mold steel based on big data according to any one of claims 1-5.

Images