Component determination method of martensitic steel and preparation method of martensitic steel

Abstract

The invention discloses a component determination method and a preparation method of martensitic steel. The component determination method comprises the following steps that the adjustment range and the step length of Ni and Ti are obtained; obtaining a plurality of groups of components according to the adjustment range and step length of Ni and Ti; simulating and generating a plurality of phase structures based on each group of components, and calculating the molar fractions of a BCC phase, an FCC phase, a Ni3Ti phase and an LEAVES phase; and inputting the mole fractions corresponding to each group of components into a model to obtain a plurality of performances corresponding to each group of components, so as to determine an optimal component according to the plurality of performances. According to the scheme provided by the invention, a brand-new correlation model among components, tissues and performance is constructed, the influence of the components on the phase tissues is quickly predicted through thermodynamic calculation, and a prediction model of the tissues on the performance is constructed in combination with a machine learning algorithm, so that quick feedback from the components to the tissues to the performance is realized; and the existing martensitic steel components are effectively improved.

Description

Technical Field

[0001] The present invention relates to the field of alloys, and in particular to a method for determining the composition of martensitic steel and a method for preparing martensitic steel. Background Art

[0002] The traditional method of designing martensitic steel compositions involves repeated testing to determine the compositional combinations that meet different performance requirements. This method requires a large number of tests and comprehensive analysis of the test results to roughly obtain the optimal combination that meets the corresponding performance requirements. The accuracy of obtaining the optimal combination is not high, the entire process is time-consuming and costly, and the efficiency is low, making it difficult to meet the needs of modern industry for martensitic steel. Summary of the Invention

[0003] In view of this, in order to overcome at least one aspect of the above-mentioned problems, an embodiment of the present invention provides a method for determining the composition of martensitic steel, comprising the following steps: Get the adjustment range and step size of Ni and Ti; Multiple groups of components are obtained according to the adjustment range and step size of Ni and Ti; Multiple phase structures are generated based on the simulation of each component, and the mole fractions of BCC phase, FCC phase, Ni3Ti phase and LEAVES phase are calculated; The mole fraction corresponding to each component is input into the model to obtain multiple properties corresponding to each component, so as to determine the optimal composition according to the multiple properties.

[0004] In some embodiments, obtaining the adjustment range and step size of Ni and Ti further includes: The adjustment range and step size of Ni and Ti are set as follows: the adjustment range of Ni is 14%~19%, the step size is 0.05%, and the adjustment range of Ti is 0~2%, the step size is 0.02%. In some embodiments, before inputting the mole fraction corresponding to each component group into the model to obtain the performance corresponding to each component group, the method further includes: Acquire multiple sets of actual data of the martensitic steel, each set of the actual data including the composition and properties of the martensitic steel, and obtain the mole fractions of the corresponding BCC phase, FCC phase, Ni3Ti phase and LEAVES phase based on the composition of the martensitic steel; The mole fractions of BCC phase, FCC phase, Ni3Ti phase and LEAVES phase corresponding to each component are used as samples, and the performance is used as a label to train the model.

[0005] In some embodiments, the mole fractions of the BCC phase, FCC phase, Ni3Ti phase, and LEAVES phase corresponding to each group of components are used as samples, and the performance is used as a label to train the model, further comprising: Divide multiple sets of real data into training sets, validation sets, and test sets; The model is trained using the mole fractions of BCC phase, FCC phase, Ni3Ti phase and LEAVES phase corresponding to each component in the training set as samples and the performance as labels; The mole fractions of BCC phase, FCC phase, Ni3Ti phase and LEAVES phase corresponding to each component in the validation set were used as samples and the performance was used as labels to validate the model. The mole fractions of BCC phase, FCC phase, Ni3Ti phase and LEAVES phase corresponding to each component in the test set are used as samples and the performance is used as labels to test the model.

[0006] In some embodiments, obtaining multiple sets of data of the martensitic steel further includes: Each set of data is processed with missing values, unified data format, and vectorized.

[0007] In some embodiments, obtaining the adjustment range and step size of Ni and Ti further includes: The adjustment range and step size of Ni and Ti are determined according to the composition and performance of each set of actual data.

[0008] In some embodiments, obtaining multiple groups of components according to the adjustment range and step size of Ni and Ti further includes: Grid searches are performed based on the adjustment range and step size corresponding to Ni and Ti to obtain multiple groups of components with different proportions.

[0009] In some embodiments, the mole fraction corresponding to each component is input into a model to obtain the properties corresponding to each component, so as to determine the optimal composition according to the properties, further comprising: The obtained multiple performances are weighted and calculated to obtain the final performance; The optimal composition is determined based on the final properties.

[0010] In some embodiments, simulating and generating multiple phase structures based on each group of components and calculating the mole fractions of the BCC phase, FCC phase, Ni3Ti phase, and LEAVES phase therein further includes: Based on the molar fractions of BCC phase, FCC phase, Ni3Ti phase and LEAVES phase in each component, the influence of Ni and Ti on BCC phase, FCC phase, Ni3Ti phase and LEAVES phase is determined.

[0011] Based on the same inventive concept, according to another aspect of the present invention, an embodiment of the present invention further provides a method for preparing martensitic steel, comprising: Determine the composition ratio of Ni and Ti using the method described in any of the above embodiments; Martensitic steel is prepared according to the composition ratio of Ni and Ti.

[0012] The present invention has the following beneficial technical effects: The proposed solution establishes a new composition-structure-performance correlation model. Using thermodynamic calculations to rapidly predict the influence of composition on phase structure, combined with machine learning algorithms, a predictive model of structure-performance is constructed. This enables rapid feedback from composition to structure and then to performance, effectively improving the composition of existing martensitic steels. This approach not only significantly shortens R&D cycles and testing costs, but also improves the scientific nature and precision of material design. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other embodiments can be obtained based on these drawings without paying any creative work.

[0014] Figure 1 A schematic flow chart of a method for determining the composition of martensitic steel provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0015] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the embodiments of the present invention are further described in detail below in conjunction with specific embodiments and with reference to the accompanying drawings.

[0016] It should be noted that all expressions using "first" and "second" in the embodiments of the present invention are for distinguishing two non-identical entities with the same name or non-identical parameters. It can be seen that "first" and "second" are only for the convenience of expression and should not be understood as limitations on the embodiments of the present invention. Subsequent embodiments will not explain this one by one.

[0017] According to one aspect of the present invention, an embodiment of the present invention provides a method for determining the composition of martensitic steel, such as Figure 1 As shown, it may include the steps of: S1, obtain the adjustment range and step size of Ni and Ti; S2, obtain multiple groups of components according to the adjustment range and step size of Ni and Ti; S3, based on each group of components, multiple phase structures are simulated and the mole fractions of BCC phase, FCC phase, Ni3Ti phase and LEAVES phase are calculated; S4, inputting the mole fraction corresponding to each component into the model to obtain multiple properties corresponding to each component, so as to determine the optimal composition according to the multiple properties.

[0018] The proposed solution establishes a novel composition-structure-performance correlation model. Using thermodynamic calculations to rapidly predict the influence of composition on phase structure, this approach, combined with machine learning algorithms, builds a predictive model of the relationship between structure and performance. This allows for rapid feedback from composition to structure and then to performance, effectively improving the composition of existing martensitic steels. This approach not only significantly shortens R&D cycles and testing costs, but also improves the scientific nature and precision of material design.

[0019] In some embodiments, obtaining the adjustment range and step size of Ni and Ti further includes: The adjustment range and step size of Ni and Ti are set as follows: the adjustment range of Ni is 14%~19%, the step size is 0.05%, and the adjustment range of Ti is 0~2%, the step size is 0.02%.

[0020] In some embodiments, before inputting the mole fraction corresponding to each component group into the model to obtain the performance corresponding to each component group, the method further includes: Acquire multiple sets of actual data of the martensitic steel, each set of the actual data including the composition and properties of the martensitic steel, and obtain the mole fractions of the corresponding BCC phase, FCC phase, Ni3Ti phase and LEAVES phase based on the composition of the martensitic steel; The mole fractions of BCC phase, FCC phase, Ni3Ti phase and LEAVES phase corresponding to each component are used as samples, and the performance is used as a label to train the model.

[0021] Specifically, data is collected from relevant papers, factory production records, and laboratory experimental results. For example, the composition and performance information with a tensile strength ≥ 2 GPa can be extracted. Then, the molar fractions of various phases corresponding to different compositions are calculated using thermodynamic calculation methods at 500°C. Through thermodynamic simulation, the distribution law of the microscopic phase structure of the material can be clarified, providing a histological basis for performance prediction. In some embodiments, the molar fractions of the BCC phase, FCC phase, Ni3Ti phase, and LEAVES phase corresponding to each group of components are used as samples, and the performance is used as a label to train the model, further comprising: Divide multiple sets of real data into training sets, validation sets, and test sets; The model is trained using the mole fractions of BCC phase, FCC phase, Ni3Ti phase and LEAVES phase corresponding to each component in the training set as samples and the performance as labels; The mole fractions of BCC phase, FCC phase, Ni3Ti phase and LEAVES phase corresponding to each component in the validation set were used as samples and the performance was used as labels to validate the model. The mole fractions of BCC phase, FCC phase, Ni3Ti phase and LEAVES phase corresponding to each component in the test set are used as samples and the performance is used as labels to test the model.

[0022] Specifically, the model is trained using the mole fractions of the BCC, FCC, Ni3Ti, and Leaves phases corresponding to each component in the training set as samples, with performance as labels. During this process, the model continuously adjusts its parameters to minimize the error between the predicted performance results and the actual performance labels. A loss function (such as mean squared error or cross-entropy loss) is typically used to measure the difference between the predicted and true values. An optimization algorithm (such as gradient descent) is then used to update the model parameters to minimize the loss function.

[0023] The model is validated using the mole fractions of the BCC, FCC, Ni3Ti, and Leaves phases corresponding to each component in the validation set as samples, with performance as labels. The validation set is used to evaluate model performance at each stage of model training or after a certain number of training steps. By analyzing evaluation metrics on the validation set (such as accuracy, recall, and mean squared error), it is possible to determine whether the model is overfitting or underfitting, and to adjust the model's hyperparameters to improve its generalization.

[0024] The model was tested using the mole fractions of the BCC, FCC, Ni3Ti, and Leaves phases corresponding to each component in the test set as samples, with performance as labels. After model training and hyperparameter tuning were completed, the test set was used to evaluate the final model performance. Since the test set data was not used during model training and validation, the evaluation results on the test set more accurately reflect the model's performance in real-world applications.

[0025] In some embodiments, obtaining multiple sets of data of the martensitic steel further includes: Each set of data is processed with missing values, unified data format, and vectorized.

[0026] Specifically, the collected data can be normalized, including missing value processing, format unification, and vectorized representation, to provide a basic data set for subsequent analysis. For example, Python can be used to normalize the above-mentioned component-related data and performance-related data, and complete missing value processing, data format unification, and vectorization processing in sequence. Using Python's pandas library, first conduct a comprehensive inspection of the data to identify the fields or records where missing values are located. Based on the characteristics of the data, appropriate missing value processing strategies are used to delete, fill, and mark missing values. Use Python to unify the format of the data, such as data type conversion, date and time format processing, and text data processing. Finally, vectorize all features and construct a data set based on the processed data.

[0027] In some embodiments, obtaining the adjustment range and step size of Ni and Ti further includes: The adjustment range and step size of Ni and Ti are determined according to the composition and performance of each set of actual data.

[0028] Specifically, you can look at the content range of Ni and Ti in all actual data. Find the minimum and maximum values of Ni and Ti respectively, which will preliminarily determine their possible adjustment range. For example, if the Ni content in the actual data is between 10%-30% and the Ti content is between 5%-15%, then the initial adjustment range can be set to be slightly larger than this range, such as Ni is 5%-35% and Ti is 0%-20%, to include possible changes. You can also analyze the sensitivity of performance to changes in Ni and Ti content. If the performance is sensitive to changes in Ni and Ti content, that is, a small change in content will lead to a large change in performance, then the step size should be selected as a smaller value, such as 0.05% or 0.1%, in order to more accurately capture the relationship between performance changes and composition.

[0029] In some embodiments, obtaining multiple groups of components according to the adjustment range and step size of Ni and Ti further includes: Grid searches are performed based on the adjustment range and step size corresponding to Ni and Ti to obtain multiple groups of components with different proportions.

[0030] Specifically, after determining the adjustment range and step size corresponding to Ni and Ti, a grid search can be used to obtain multiple groups of components with different proportions.

[0031] In some embodiments, the mole fraction corresponding to each component is input into a model to obtain the properties corresponding to each component, so as to determine the optimal composition according to the properties, further comprising: The obtained multiple performances are weighted and calculated to obtain the final performance; The optimal composition is determined based on the final properties.

[0032] In some embodiments, simulating and generating multiple phase structures based on each group of components and calculating the mole fractions of the BCC phase, FCC phase, Ni3Ti phase, and LEAVES phase therein further includes: Based on the molar fractions of BCC phase, FCC phase, Ni3Ti phase and LEAVES phase in each component, the influence of Ni and Ti on BCC phase, FCC phase, Ni3Ti phase and LEAVES phase is determined.

[0033] Specifically, focusing on key phases (such as BCC, FCC, Ni3Ti, and Leaves) as optimization targets, we adjust compositions with properties close to the target values (e.g., strength >= 2.5 GPa, elongation >= 5%). By simulating the changes in the proportions of key alloying elements, we can understand the influence of the mole fractions of key phases. By focusing on analyzing the changing trends of key phases and identifying extreme points, we can ultimately select the optimal composition combination that meets the project's performance requirements.

[0034] For example, when the ratio of Ni to Ti is 14.25:1.18, the tensile strength (MPa) can reach 2336, when the ratio of Ni to Ti is 17.75:1.5, the tensile strength (MPa) can reach 2531, and when the ratio of Ni to Ti is 18.25:1.42, the tensile strength (MPa) can reach 2396.

[0035] The proposed solution establishes a novel composition-structure-performance correlation model. Using thermodynamic calculations to rapidly predict the influence of composition on phase structure, this approach, combined with machine learning algorithms, builds a predictive model of the relationship between structure and performance. This allows for rapid feedback from composition to structure and then to performance, effectively improving the composition of existing martensitic steels. This approach not only significantly shortens R&D cycles and testing costs, but also improves the scientific nature and precision of material design.

[0036] Based on the same inventive concept, according to another aspect of the present invention, an embodiment of the present invention further provides a method for preparing martensitic steel, comprising: Determine the composition ratio of Ni and Ti using the method described in any of the above embodiments; Martensitic steel is prepared according to the composition ratio of Ni and Ti.

[0037] The above are exemplary embodiments disclosed in the present invention, but it should be noted that various changes and modifications may be made without departing from the scope of the embodiments disclosed in the claims. The functions, steps and / or actions of the method claims according to the disclosed embodiments described herein do not need to be performed in any particular order. In addition, although the elements disclosed in the embodiments of the present invention may be described or required in individual form, they may also be understood as multiple unless expressly limited to the singular.

[0038] It should be understood that, as used herein, the singular forms "a" and "an" are intended to include the plural forms as well, unless the context clearly supports an exception. It should also be understood that, as used herein, "and / or" is intended to include any and all possible combinations of one or more of the associated listed items.

[0039] The serial numbers of the embodiments disclosed in the above embodiments of the present invention are only for description and do not represent the advantages or disadvantages of the embodiments.

[0040] Those skilled in the art should understand that the discussion of any of the above embodiments is merely illustrative and is not intended to limit the scope of the disclosure of the present invention (including the claims) to these examples. Within the spirit of the present invention, the technical features of the above embodiments or different embodiments may be combined, and many other variations exist in different aspects of the above embodiments, which are not provided in detail for the sake of clarity. Therefore, any omissions, modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for determining the composition of martensitic steel, characterized in that: The following steps are involved: Get the adjustment range and step size of Ni and Ti; Multiple groups of components are obtained according to the adjustment range and step size of Ni and Ti; Multiple phase structures are generated based on the simulation of each component, and the mole fractions of BCC phase, FCC phase, Ni3Ti phase and LEAVES phase are calculated; The mole fraction corresponding to each component is input into the model to obtain multiple properties corresponding to each component, so as to determine the optimal composition according to the multiple properties.

2. The method according to claim 1, wherein Obtaining the adjustment range and step size of Ni and Ti further includes: The adjustment range and step size of Ni and Ti are set as follows: the adjustment range of Ni is 14%~19%, the step size is 0.05%, and the adjustment range of Ti is 0~2%, the step size is 0.02%.

3. The method according to claim 1, wherein Before inputting the mole fraction corresponding to each component into the model to obtain the performance corresponding to each component, the method further includes: Acquire multiple sets of actual data of the martensitic steel, each set of the actual data including the composition and properties of the martensitic steel, and obtain the mole fractions of the corresponding BCC phase, FCC phase, Ni3Ti phase and LEAVES phase based on the composition of the martensitic steel; The mole fractions of BCC phase, FCC phase, Ni3Ti phase and LEAVES phase corresponding to each component are used as samples, and the performance is used as a label to train the model.

4. The method according to claim 3, wherein The model is trained using the mole fractions of the BCC phase, FCC phase, Ni3Ti phase, and LEAVES phase corresponding to each component as samples and the performance as labels, further comprising: Divide multiple sets of real data into training sets, validation sets, and test sets; The model is trained using the mole fractions of BCC phase, FCC phase, Ni3Ti phase and LEAVES phase corresponding to each component in the training set as samples and the performance as labels; The mole fractions of BCC phase, FCC phase, Ni3Ti phase and LEAVES phase corresponding to each component in the validation set were used as samples and the performance was used as labels to validate the model. The mole fractions of BCC phase, FCC phase, Ni3Ti phase and LEAVES phase corresponding to each component in the test set are used as samples and the performance is used as labels to test the model.

5. The method according to claim 3, wherein Acquiring multiple sets of data of the martensitic steel further includes: Each set of data is processed with missing values, unified data format, and vectorized.

6. The method according to claim 3, wherein Obtaining the adjustment range and step size of Ni and Ti further includes: The adjustment range and step size of Ni and Ti are determined according to the composition and performance of each set of actual data.

7. The method according to claim 1, wherein According to the adjustment range and step size of Ni and Ti, multiple groups of components are obtained, further including: Grid searches are performed based on the adjustment range and step size corresponding to Ni and Ti to obtain multiple groups of components with different proportions.

8. The method according to claim 1, wherein Inputting the mole fraction corresponding to each component into the model to obtain the properties corresponding to each component, so as to determine the optimal composition according to the properties, further comprising: The obtained multiple performances are weighted and calculated to obtain the final performance; The optimal composition is determined based on the final properties.

9. The method according to claim 1, wherein Based on each component group, multiple phase structures are simulated and the mole fractions of BCC phase, FCC phase, Ni3Ti phase and Leaves phase are calculated, further including: Based on the molar fractions of BCC phase, FCC phase, Ni3Ti phase and LEAVES phase in each component, the influence of Ni and Ti on BCC phase, FCC phase, Ni3Ti phase and LEAVES phase is determined.

10. A method for preparing martensitic steel, characterized in that: include: Determining the composition ratio of Ni and Ti using the method according to any one of claims 1 to 9; Martensitic steel is prepared according to the composition ratio of Ni and Ti.

Images