CN113870957B - Eutectic high-entropy alloy composition design method and device based on machine learning - Google Patents
Eutectic high-entropy alloy composition design method and device based on machine learning Download PDFInfo
- Publication number
- CN113870957B CN113870957B CN202111232545.1A CN202111232545A CN113870957B CN 113870957 B CN113870957 B CN 113870957B CN 202111232545 A CN202111232545 A CN 202111232545A CN 113870957 B CN113870957 B CN 113870957B
- Authority
- CN
- China
- Prior art keywords
- alloy
- entropy alloy
- eutectic
- components
- eutectic high
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 229910045601 alloy Inorganic materials 0.000 title claims abstract description 248
- 239000000956 alloy Substances 0.000 title claims abstract description 248
- 230000005496 eutectics Effects 0.000 title claims abstract description 116
- 238000000034 method Methods 0.000 title claims abstract description 66
- 238000010801 machine learning Methods 0.000 title claims abstract description 42
- 239000000203 mixture Substances 0.000 title claims abstract description 39
- 238000013461 design Methods 0.000 title claims abstract description 19
- 238000012549 training Methods 0.000 claims abstract description 56
- 230000015572 biosynthetic process Effects 0.000 claims abstract description 13
- 238000007619 statistical method Methods 0.000 claims abstract description 12
- 238000012795 verification Methods 0.000 claims description 15
- 238000004364 calculation method Methods 0.000 claims description 14
- 238000007711 solidification Methods 0.000 claims description 14
- 230000008023 solidification Effects 0.000 claims description 14
- 230000008569 process Effects 0.000 claims description 9
- 238000004458 analytical method Methods 0.000 claims description 8
- 201000004569 Blindness Diseases 0.000 abstract description 5
- 238000003723 Smelting Methods 0.000 description 8
- 238000002844 melting Methods 0.000 description 7
- 230000008018 melting Effects 0.000 description 7
- XKRFYHLGVUSROY-UHFFFAOYSA-N Argon Chemical compound [Ar] XKRFYHLGVUSROY-UHFFFAOYSA-N 0.000 description 6
- 239000002994 raw material Substances 0.000 description 6
- 238000005266 casting Methods 0.000 description 5
- 238000010587 phase diagram Methods 0.000 description 5
- 238000003860 storage Methods 0.000 description 5
- 238000012512 characterization method Methods 0.000 description 4
- 230000006870 function Effects 0.000 description 4
- 229910052782 aluminium Inorganic materials 0.000 description 3
- 229910052786 argon Inorganic materials 0.000 description 3
- 238000004590 computer program Methods 0.000 description 3
- 239000000470 constituent Substances 0.000 description 3
- 210000001787 dendrite Anatomy 0.000 description 3
- 238000002441 X-ray diffraction Methods 0.000 description 2
- 230000005540 biological transmission Effects 0.000 description 2
- 238000004140 cleaning Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000011049 filling Methods 0.000 description 2
- 239000000463 material Substances 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 239000006104 solid solution Substances 0.000 description 2
- 238000003756 stirring Methods 0.000 description 2
- 239000010936 titanium Substances 0.000 description 2
- LFQSCWFLJHTTHZ-UHFFFAOYSA-N Ethanol Chemical compound CCO LFQSCWFLJHTTHZ-UHFFFAOYSA-N 0.000 description 1
- RTAQQCXQSZGOHL-UHFFFAOYSA-N Titanium Chemical compound [Ti] RTAQQCXQSZGOHL-UHFFFAOYSA-N 0.000 description 1
- 238000005275 alloying Methods 0.000 description 1
- XAGFODPZIPBFFR-UHFFFAOYSA-N aluminium Chemical compound [Al] XAGFODPZIPBFFR-UHFFFAOYSA-N 0.000 description 1
- 238000010314 arc-melting process Methods 0.000 description 1
- QVGXLLKOCUKJST-UHFFFAOYSA-N atomic oxygen Chemical compound [O] QVGXLLKOCUKJST-UHFFFAOYSA-N 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 229910002056 binary alloy Inorganic materials 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000000576 coating method Methods 0.000 description 1
- 230000006835 compression Effects 0.000 description 1
- 238000007906 compression Methods 0.000 description 1
- 230000007797 corrosion Effects 0.000 description 1
- 238000005260 corrosion Methods 0.000 description 1
- 238000007418 data mining Methods 0.000 description 1
- 230000007423 decrease Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000001035 drying Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 239000006023 eutectic alloy Substances 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 239000012535 impurity Substances 0.000 description 1
- 239000007788 liquid Substances 0.000 description 1
- 239000008204 material by function Substances 0.000 description 1
- 230000010355 oscillation Effects 0.000 description 1
- 239000001301 oxygen Substances 0.000 description 1
- 229910052760 oxygen Inorganic materials 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 229910002059 quaternary alloy Inorganic materials 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 238000005204 segregation Methods 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 229910052719 titanium Inorganic materials 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
- 238000010200 validation analysis Methods 0.000 description 1
- 238000005303 weighing Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16C—COMPUTATIONAL CHEMISTRY; CHEMOINFORMATICS; COMPUTATIONAL MATERIALS SCIENCE
- G16C20/00—Chemoinformatics, i.e. ICT specially adapted for the handling of physicochemical or structural data of chemical particles, elements, compounds or mixtures
- G16C20/70—Machine learning, data mining or chemometrics
-
- C—CHEMISTRY; METALLURGY
- C22—METALLURGY; FERROUS OR NON-FERROUS ALLOYS; TREATMENT OF ALLOYS OR NON-FERROUS METALS
- C22C—ALLOYS
- C22C19/00—Alloys based on nickel or cobalt
- C22C19/03—Alloys based on nickel or cobalt based on nickel
- C22C19/05—Alloys based on nickel or cobalt based on nickel with chromium
- C22C19/058—Alloys based on nickel or cobalt based on nickel with chromium without Mo and W
-
- C—CHEMISTRY; METALLURGY
- C22—METALLURGY; FERROUS OR NON-FERROUS ALLOYS; TREATMENT OF ALLOYS OR NON-FERROUS METALS
- C22C—ALLOYS
- C22C30/00—Alloys containing less than 50% by weight of each constituent
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
- G06N20/10—Machine learning using kernel methods, e.g. support vector machines [SVM]
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16C—COMPUTATIONAL CHEMISTRY; CHEMOINFORMATICS; COMPUTATIONAL MATERIALS SCIENCE
- G16C20/00—Chemoinformatics, i.e. ICT specially adapted for the handling of physicochemical or structural data of chemical particles, elements, compounds or mixtures
- G16C20/90—Programming languages; Computing architectures; Database systems; Data warehousing
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16C—COMPUTATIONAL CHEMISTRY; CHEMOINFORMATICS; COMPUTATIONAL MATERIALS SCIENCE
- G16C60/00—Computational materials science, i.e. ICT specially adapted for investigating the physical or chemical properties of materials or phenomena associated with their design, synthesis, processing, characterisation or utilisation
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Chemical & Material Sciences (AREA)
- Computing Systems (AREA)
- Software Systems (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Organic Chemistry (AREA)
- Medical Informatics (AREA)
- Data Mining & Analysis (AREA)
- Crystallography & Structural Chemistry (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Databases & Information Systems (AREA)
- Materials Engineering (AREA)
- Mechanical Engineering (AREA)
- Evolutionary Computation (AREA)
- Metallurgy (AREA)
- Mathematical Physics (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Investigating And Analyzing Materials By Characteristic Methods (AREA)
Abstract
The application discloses a method and a device for designing eutectic high-entropy alloy components based on machine learning, wherein the method comprises the following steps: acquiring training data, wherein the training data comprises input data and output data, the input data is a component of an alloy, and the output data is a phase composition of the alloy; inputting training data into a pre-selected machine learning model for training to obtain a trained model; obtaining a plurality of groups of alloy components with mole fractions of primary phases within a preset range through a model; carrying out statistical analysis on a plurality of groups of alloy components to obtain key elements with important influence on eutectic formation and elements strongly related to the key elements; and adjusting the contents of the key elements and elements strongly related to the key elements, and predicting the eutectic high-entropy alloy to obtain the components of the eutectic high-entropy alloy. The application solves the problem caused by the prediction of the eutectic high-entropy alloy by the existing method, and avoids the blindness of the prediction of the eutectic high-entropy alloy by the existing method, thereby improving the design efficiency of the eutectic high-entropy alloy.
Description
Technical Field
The application relates to the field of alloys, in particular to a method and a device for designing eutectic high-entropy alloy components based on machine learning.
Background
The high-entropy alloy has excellent strength, hardness, corrosion resistance, thermal stability, irradiation resistance and the like, and has great development potential in the fields of structural materials, functional materials, high-temperature coatings, nuclear industry and the like. Such alloys have a "thermodynamic high entropy effect" and thus tend to form solid solution phases such as FCC, BCC and HCP having a simple structure. Studies have shown that: on one hand, the single-phase high-entropy alloy has poor coordination in strength and plasticity, and it is important to obtain the high-entropy alloy with a multiphase structure through reasonable component design so as to improve the strength and plasticity of the high-entropy alloy at the same time; on the other hand, the casting method is a common method for preparing the high-entropy alloy, but the single-phase high-entropy alloy has poor fluidity and castability due to the fact that the single-phase high-entropy alloy contains various high-concentration main elements, so that casting defects such as component segregation, coarse structure, internal shrinkage cavity and the like exist in the casting, and engineering application and development of the high-entropy alloy are greatly limited. In order to solve the problems of single-phase high-entropy alloy, alCoCrFeNi 2.1 eutectic high-entropy alloy is successfully prepared. The eutectic high-entropy alloy has the characteristics of both high-entropy alloy and eutectic alloy, can realize the balance of alloy strength and plasticity, improves the castability of the high-entropy alloy, and has wide prospect in the aspect of obtaining high-entropy alloy castings with uniform components and tissues and excellent mechanical properties.
At present, how to design the composition of eutectic high-entropy alloy effectively is still a great challenge. The traditional trial-and-error method cannot systematically and accurately find out the influence rule of elements on the alloy structure, and is difficult to quickly and accurately position the components of the eutectic high-entropy alloy, and labor and material resources are consumed. The calhad method can be effectively used to design eutectic high entropy alloys, mainly by simulating pseudo-binary phase diagrams and calculating the solidification path of a series of alloys. The partial eutectic high-entropy alloy can be regarded as being composed of a single-phase solid solution and an intermediate phase, and then a simulated phase diagram of the pseudo-binary alloy is calculated by means of a CALPHAD method, so that the positioning of the eutectic component points of the high-entropy alloy is realized. However, eutectic high-entropy alloy systems with composition design through pseudo-binary phase diagrams are more limited. In addition, the eutectic composition of the alloy is obtained by calculating the solidification path of a series of high-entropy alloys and then according to the solidification characteristics of the eutectic high-entropy alloys. However, the computational effort of this method increases with the number of alloying elements, which also greatly increases the time cost of alloy design.
Disclosure of Invention
The embodiment of the application provides a method and a device for designing eutectic high-entropy alloy components based on machine learning, which at least solve the problem caused by predicting the eutectic high-entropy alloy by the existing method.
According to one aspect of the present application, there is provided a method for designing a eutectic high-entropy alloy composition based on machine learning, comprising:
obtaining training data, wherein the training data comprises input data and output data, the input data is a component of an alloy, the output data is a phase composition of the alloy, and the phase composition is a primary phase mole fraction generated in the solidification process of the alloy; inputting the training data into a pre-selected machine learning model for training to obtain a trained model; obtaining a plurality of groups of alloy components with primary phase mole fractions within a preset range through the model, wherein the alloy with the primary phase mole fractions within the preset range is the eutectic high-entropy alloy; carrying out statistical analysis on the plurality of groups of alloy components to obtain key elements with important influence on eutectic formation and elements strongly related to the key elements; and adjusting the contents of the key elements and elements strongly related to the key elements, and predicting the eutectic high-entropy alloy to obtain the components of the eutectic high-entropy alloy.
Further, acquiring the training data includes: grasping alloy components and primary phase mole fractions in published literature data; and (5) selecting alloy components by self and obtaining thermodynamic calculation results.
Further, inputting the training data into a pre-selected machine learning model for training, and obtaining a trained model comprises: and under the condition that the machine learning model is an SVM model, acquiring a decision coefficient, and determining to finish training of the model when the decision coefficient is greater than or equal to a preset threshold value to obtain a trained model.
Further, predicting the eutectic high-entropy alloy to obtain the eutectic high-entropy alloy comprises the following components: the calhatd method was used to predict eutectic high entropy alloys.
Further, the method further comprises the following steps: and carrying out experimental verification on the predicted components of the eutectic high-entropy alloy and the nearby component alloys with the preset group numbers.
According to another aspect of the present application, there is also provided a machine learning-based eutectic high entropy alloy composition design apparatus, including: the first acquisition module is used for acquiring training data, wherein the training data comprises input data and output data, the input data is a component of an alloy, the output data is a phase composition of the alloy, and the phase composition is a primary phase mole fraction generated in the solidification process of the alloy; the training module is used for inputting the training data into a pre-selected machine learning model for training to obtain a trained model; the second acquisition module is used for acquiring a plurality of groups of alloy components with primary phase mole fractions within a preset range through the model, wherein the alloy with the primary phase mole fractions within the preset range is the eutectic high-entropy alloy; the analysis module is used for carrying out statistical analysis on the plurality of groups of alloy components to obtain key elements with important influence on eutectic formation and elements strongly related to the key elements; and the prediction module is used for adjusting the contents of the key elements and elements strongly related to the key elements and predicting the eutectic high-entropy alloy to obtain the components of the eutectic high-entropy alloy.
Further, the first acquisition module is configured to: grasping alloy components and primary phase mole fractions in published literature data; and (5) selecting alloy components by self and obtaining thermodynamic calculation results.
Further, the training module is configured to: and under the condition that the machine learning model is an SVM model, acquiring a decision coefficient, and determining to finish training of the model when the decision coefficient is greater than or equal to a preset threshold value to obtain a trained model.
Further, the prediction module is configured to: the calhatd method was used to predict eutectic high entropy alloys.
Further, the method further comprises the following steps: and the verification module is used for carrying out experimental verification on the components of the eutectic high-entropy alloy obtained through prediction and the nearby component alloys with the preset group number.
In the embodiment of the application, training data is acquired, wherein the training data comprises input data and output data, the input data is a component of an alloy, the output data is a phase composition of the alloy, and the phase composition is a primary phase mole fraction generated in the solidification process of the alloy; inputting the training data into a pre-selected machine learning model for training to obtain a trained model; obtaining a plurality of groups of alloy components with primary phase mole fractions within a preset range through the model, wherein the alloy with the primary phase mole fractions within the preset range is the eutectic high-entropy alloy; carrying out statistical analysis on the plurality of groups of alloy components to obtain key elements with important influence on eutectic formation and elements strongly related to the key elements; and adjusting the contents of the key elements and elements strongly related to the key elements, and predicting the eutectic high-entropy alloy to obtain the components of the eutectic high-entropy alloy. The application solves the problem caused by the prediction of the eutectic high-entropy alloy by the existing method, thereby avoiding the blindness of the prediction of the eutectic high-entropy alloy by the existing method and improving the design efficiency of the eutectic high-entropy alloy.
Drawings
The accompanying drawings, which are included to provide a further understanding of the application and are incorporated in and constitute a part of this specification, illustrate embodiments of the application and together with the description serve to explain the application. In the drawings:
FIG. 1 is a flow chart of a method of combining machine learning with CALPHAD assisted eutectic high entropy alloy composition design in accordance with an embodiment of the present application;
FIG. 2 is a plot of the mole fraction of the primary phase as a function of Al content during equilibrium solidification and non-equilibrium solidification of a Ni 68-xCo16Cr16Alx high-entropy alloy, according to an example of the application;
FIG. 3 is an XRD pattern of a Ni 68-xCo16Cr16Alx high-entropy alloy according to an embodiment of the application;
FIG. 4 is a microstructure image of a Ni 49.8Co16Cr16Al18.2 hypoeutectic high-entropy alloy according to an embodiment of the present application;
FIG. 5 is a microstructure image of a Ni 49Co16Cr16Al19 eutectic high-entropy alloy according to an embodiment of the present application;
FIG. 6 is a microstructure image of a Ni 48.2Co16Cr16Al19.8 hypereutectic high-entropy alloy according to an embodiment of the application;
FIG. 7 is a microstructure image of a Ni 47.4Co16Cr16Al20.6 hypereutectic high-entropy alloy according to an embodiment of the present application;
FIG. 8 is a flow chart of a method of designing a eutectic high entropy alloy composition based on machine learning in accordance with an embodiment of the present application.
Detailed Description
It should be noted that, without conflict, the embodiments of the present application and features of the embodiments may be combined with each other. The application will be described in detail below with reference to the drawings in connection with embodiments.
It should be noted that the steps illustrated in the flowcharts of the figures may be performed in a computer system such as a set of computer executable instructions, and that although a logical order is illustrated in the flowcharts, in some cases the steps illustrated or described may be performed in an order other than that illustrated herein.
In this embodiment, a method for designing a eutectic high-entropy alloy composition based on machine learning is provided, and fig. 8 is a flowchart of the method for designing a eutectic high-entropy alloy composition based on machine learning according to an embodiment of the present application, as shown in fig. 8, the flowchart includes the following steps:
Step S802, training data is obtained, wherein the training data comprises input data and output data, the input data is a component of an alloy, the output data is a phase composition of the alloy, and the phase composition is a primary phase mole fraction generated in the solidification process of the alloy;
There are various ways to obtain training data, for example, grabbing alloy components and primary phase mole fractions in published literature data; and (5) selecting alloy components by self and obtaining thermodynamic calculation results. This way of obtaining data is relatively easy.
Step S804, inputting the training data into a pre-selected machine learning model for training to obtain a trained model;
The machine learning model may also select a plurality of existing models, in this embodiment, an SVM model is selected, and in the case where the machine learning model is an SVM model, a decision coefficient is obtained, and training of the model is determined to be completed when the decision coefficient is greater than or equal to a preset threshold value, so as to obtain a trained model.
Step S806, obtaining a plurality of groups of alloy components with primary phase mole fractions within a preset range through the model, wherein the alloy with the primary phase mole fractions within the preset range is the eutectic high-entropy alloy;
step S808, carrying out statistical analysis on the plurality of groups of alloy components to obtain key elements with important influence on eutectic formation and elements strongly related to the key elements;
The key factors having influence on the eutectic generation can be obtained, then the key factors are arranged according to the influence size, and a predetermined number of key elements having great influence are selected from the arranged key factors as key elements having important influence.
And step S810, adjusting the contents of the key elements and elements strongly related to the key elements, and predicting the eutectic high-entropy alloy to obtain the components of the eutectic high-entropy alloy.
There are also various ways of prediction, for example, the calhatd method may be used to predict eutectic high entropy alloys. After prediction, the components of the eutectic high-entropy alloy obtained by prediction and the nearby component alloys with preset groups can be subjected to experimental verification.
The method solves the problem caused by the prediction of the eutectic high-entropy alloy by the existing method, and avoids blindness of the prediction of the eutectic high-entropy alloy by the existing method, thereby improving the design efficiency of the eutectic high-entropy alloy.
An alternative embodiment is described below with reference to the accompanying drawings. In this embodiment, machine learning is considered as a typical data mining analysis method, and has the characteristics of low trial and error cost, short period, strong adaptability and the like, so that the embodiment combines the machine learning and phase diagram calculation (Calculation of PHASE DIAGRAMS, abbreviated as calpha) method to rapidly locate eutectic components in the high-entropy alloy. The embodiment provides a method for designing a co-crystallized high-entropy alloy by combining machine learning and CALPHAD, which establishes a database based on thermodynamic calculation results and literature data, and utilizes the relation between component-phase composition of machine learning unlocking to realize the purpose of rapidly and accurately designing the co-crystallized high-entropy alloy component according to the CALPHAD results.
Fig. 1 is a flowchart of a method for designing a composition of a co-eutectic high-entropy alloy in combination with machine learning and calhatd according to an embodiment of the present application, as shown in fig. 1, the method for designing a co-eutectic high-entropy alloy in combination with machine learning and calhatd according to the embodiment of the present application includes the steps of:
(1) Establishing a database according to literature data and thermodynamic calculation results; wherein, each group of data of the database comprises components and phase compositions of the alloy, and the phase compositions are mole fractions of primary phases generated in the solidification process of the alloy.
(2) Selecting a proper machine learning model, setting input and output parameters according to the database established in the step (1), and training the model.
As an alternative implementation manner, in the step (2), the selected machine learning model may be an SVM model, the alloy component is taken as input, the mole fraction of the primary phase corresponding to the alloy component is taken as output, the model is trained by utilizing a database, the decision coefficient R 2 is adopted to quantitatively evaluate the predicted performance of the model, and the training of the model is completed when R 2 is more than or equal to 0.9.
(3) Predicting by using the model trained in the step (2) to obtain a large amount of near-eutectic high-entropy alloy components, and carrying out statistical analysis on the near-eutectic high-entropy alloy components obtained by prediction to find out key elements with important influence on eutectic phase formation and elements strongly related to the key elements.
Optionally, in the step (3), the primary phase mole fraction of the near-eutectic high-entropy alloy component is less than or equal to 4%, and the obtained near-eutectic high-entropy alloy component is not less than 300 groups to ensure the accuracy of the statistical analysis result.
(4) And (3) adjusting the contents of the key elements and the elements strongly related to the key elements obtained in the step (3), and predicting the eutectic high-entropy alloy by using a CALPHAD method. Can be adopted in the stepSoftware, TTNI thermodynamic database was selected and the calhatd method was used to predict eutectic high entropy alloys.
(5) And (3) selecting the eutectic high-entropy alloy obtained through prediction and the nearby component alloys (3-5 groups) according to the result of the step (4) for experimental verification. In this example, granular or block raw materials with a purity of 99.99% or more may be proportioned, and the button ingot high-entropy alloy sample prepared by arc melting may be subjected to experimental verification. Optionally, in this step, when preparing the alloy by arc melting, vacuum is applied to 1×10 -3 Pa, argon is filled, and simultaneously the melting is repeated for 5 to 6 times in a turning manner and electromagnetic stirring is performed to ensure the uniformity of the alloy, and phase and microstructure analysis is performed by XRD and SEM, respectively.
The present embodiment will be described below with reference to several examples.
Example 1
In this embodiment, the machine learning and calhad method are combined to rapidly and accurately perform eutectic high-entropy alloy design in the Ni xCoyCryAlz high-entropy alloy system, and the following is a detailed description of embodiment 1.
(1) Establishing a database: the database consists of 100 sets of data, all from literature data and by means ofThe software obtains the equilibrium thermodynamic calculation result.
Table 1 summarizes the eutectic high entropy alloys reported in the Ni-Co-Cr-Al system, with the corresponding primary phase mole fraction of 0. Each set of data in the database includes an alloy composition and a primary phase mole fraction corresponding to the alloy composition.
Table 1 Ni eutectic high entropy alloys reported in Co-Cr-Al System
(2) Training a model: taking alloy components as input and the mole fraction of the primary phase corresponding to the alloy components as output; randomly selecting 80 (80%) group data and 20 (20%) group data from the database established in the step (1) as a training set and a verification set respectively for training and verifying the SVM model; the mean square coefficients R 2 of the training set and validation set, calculated by the formula, are 0.9239 and 0.9156 respectively, which values are particularly close to 1, which indicates that the model is very predictable.
(3) Element classification: the purpose of element classification is to find out key elements with important influence on eutectic phase formation and elements strongly related to the key elements in a Ni xCoyCryAlz high-entropy alloy system.
The constituent concentrations of the Ni xCoyCryAlz high-entropy alloy can be set as: x is more than or equal to 16 and less than or equal to 21, y is 15-20, and z=100-x-y. And (3) randomly selecting 5000 components in a set component range by using the trained model in the step (2) to predict, wherein 341 groups of near eutectic components (mole fraction of primary phases is less than 4%) are obtained in the embodiment. Based on predicted 341 groups of near-eutectic compositions, the present example classifies the constituent elements of the alloy and analyzes their correlation, thereby understanding the effect of each constituent element on the eutectic phase formation of the Ni xCoyCryAlz high-entropy alloy. The results show that in the Ni-Co-Cr-Al system, al is a key element in the formation of a eutectic phase, and Ni is an element strongly related to the key element Al.
(4) The calhad method predicts eutectic high entropy alloy: in the step, ni 68-xCo16Cr16Alx alloy in a Ni xCoyCryAlz high-entropy system is selected as a specific research object, and the specific influence of the Al element in the Ni 68-xCo16Cr16Alx high-entropy alloy on the formation of a eutectic phase is known by adjusting the content of the Al element through thermodynamic calculation. FIG. 2 shows the mole fraction of the primary phase of Ni 68- xCo16Cr16Alx high-entropy alloy as a function of Al content during equilibrium solidification. When the Ni 68- xCo16Cr16Alx high-entropy alloy is a eutectic component, its primary phase mole fraction should be 0. Therefore, the embodiment can preliminarily predict the components of the Ni 68-xCo16Cr16Alx eutectic high-entropy alloy according to the simulation result. However, due to the limitation of thermodynamic databases, ni 48.2Co16Cr16Al19.8 alloy and 3-5 groups of nearby high-entropy alloy components need to be selected for the next experimental verification.
(5) And (3) experimental verification: and (3) selecting 4 groups of high-entropy alloy components according to the prediction result of the step (4) for experimental verification, wherein the specific components are shown in a table 2, and then carrying out phase analysis and microstructure characterization on the 4-group alloy.
The arc melting process can be used to prepare 4 groups of high entropy alloys in table 2, with the following operations performed to ensure consistency and uniformity of the alloy composition: selecting granular or massive raw materials with purity more than or equal to 99.99%, and proportioning the raw materials according to a proportion; vacuumizing to 1X 10 -3 Pa and filling argon; and when the alloy is prepared by arc melting, the overturning and repeated melting are carried out for 5 to 6 times.
Table 2 experimentally verified 4-group high-entropy alloy and components thereof
Phase analysis is carried out on the Al18.2, al19, al19.8 and Al20.6 high-entropy alloy samples obtained in the example, specifically, the phase analysis is carried out by adopting X-ray diffraction (XRD, D8 ADVANCEDAVINCI), the scanning angle is 20-100 degrees, the scanning speed is 4 degrees/min, and the characterization result is shown in figure 3. From XRD results, it is clear that the four alloys are composed of FCC phase and BCC phase.
The samples of the high entropy alloys Al18.2, al19, al19.8 and Al20.6 obtained in the examples were subjected to microstructure characterization, specifically, using an optical microscope (OM, leica-DM 4000M) and a scanning electron microscope (SEM, FEI Quanta 650), and the characterization results are shown in FIGS. 3-6. Wherein, fig. 4 is a microstructure image of Ni 49.8Co16Cr16Al18.2 alloy consisting of dendrite (FCC phase) and eutectic structure, which is a hypoeutectic high-entropy alloy; FIG. 5 is a microstructure image of a Ni 49Co16Cr16Al19 alloy consisting entirely of lamellar eutectic structures, a eutectic high-entropy alloy; FIG. 6 is a microstructure image of a Ni 48.2Co16Cr16Al19.8 alloy consisting of dendrites (BCC phases) and eutectic structures, a hypereutectic high-entropy alloy; FIG. 7 is a microstructure image of a Ni 47.4Co16Cr16Al20.6 alloy consisting of dendrites (BCC phases) and eutectic structures, a hypereutectic high-entropy alloy; as the Al content increases (while the Ni content decreases), the alloy microstructure has a hypo-eutectic-hypereutectic transformation law. Thus, the Ni 49Co16Cr16Al19 alloy is a eutectic high-entropy alloy in the Ni xCoyCryAlz quaternary high-entropy alloy system, with the predicted aluminum content of the eutectic point (19.82 at.%) slightly higher than the experimental result (19 at.%), which may be caused by the choice of thermodynamic database, as it goesThe prediction accuracy of the method can be improved continuously due to the continuous perfection of the related thermodynamic database in the software.
1. The components of the eutectic high-entropy alloy of Ni 49Co16Cr16Al19 (at.%) are:
Ni：55.5～56wt.％；Cr：16～16.5wt.％；Co：18～18.5wt.％；Al：9.5～10wt.％
The preparation method of the Ni 49Co16Cr16Al19 (at%) eutectic high-entropy alloy comprises the following steps:
1) Selecting granular or massive Al, co, cr, ni raw materials with purity of more than or equal to 99.99%, using absolute ethyl alcohol as a cleaning liquid to carry out ultrasonic oscillation cleaning, and drying for later use so as to achieve the purpose of removing surface impurities;
2) Weighing the standby raw materials in the step 1) according to the proportion, and respectively bagging for smelting;
3) According to the melting point of each element, sequentially placing weighed Al, co, ni and Cr granular or massive raw materials into a crucible of a high-vacuum arc melting furnace, and closing a furnace door; vacuumizing to 1X 10 -3 Pa and filling argon; smelting, namely smelting a titanium ingot firstly, and smelting the Ni-Co-Cr-Al quaternary alloy for multiple times; after each smelting is completed, after the alloy is cooled into button-shaped ingots, turning the ingots up and down, and then smelting the ingots for the next time; each furnace of alloy is overturned for 5 times, and the alloy is subjected to electromagnetic stirring during the second and third smelting to ensure the uniformity of alloy tissues and components;
4) After smelting, arc breaking is carried out, the alloy ingot is cooled for 15 minutes along with a furnace to obtain eutectic high-entropy alloy ingot casting, and the furnace is opened for sampling.
In the step (3), the Ti block is smelted first to eliminate residual oxygen molecules in the furnace body.
2. Performance of Ni 49Co16Cr16Al19 (at.%) eutectic high-entropy alloy
| Sequence number | Alloy | Compressive Strength | Fracture compression ratio |
| Example 1 | Ni49Co16Cr16Al19 | 3006.9MPa | 45.5％ |
| Example 2 | Ni48.2Co16Cr16Al19.8 | 2971.3MPa | 44.2％ |
Through the embodiment, on the basis of the internal relation between clear elements and phase compositions, the CALPHAD method is adopted to predict the eutectic composition of the high-entropy alloy, and the design concept and the guiding concept of the novel eutectic high-entropy alloy are provided. Compared with the design of the eutectic high-entropy alloy by the prior CALPHAD method, the embodiment avoids the blindness of predicting the eutectic high-entropy alloy by the CALPHAD method by means of machine learning, thereby improving the efficiency of the design of the eutectic high-entropy alloy, and the beneficial effect becomes more obvious along with the increase of the number of alloy elements.
In this embodiment, there is also provided an electronic device comprising a memory in which a computer program is stored, and a processor arranged to run the computer program to perform the method of the above embodiments.
The above-described programs may be run on a processor or may also be stored in memory (or referred to as computer-readable media), including both permanent and non-permanent, removable and non-removable media, and information storage may be implemented by any method or technique. The information may be computer readable instructions, data structures, modules of a program, or other data. Examples of storage media for a computer include, but are not limited to, phase change memory (PRAM), static Random Access Memory (SRAM), dynamic Random Access Memory (DRAM), other types of Random Access Memory (RAM), read Only Memory (ROM), electrically Erasable Programmable Read Only Memory (EEPROM), flash memory or other memory technology, compact disc read only memory (CD-ROM), digital Versatile Discs (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices, or any other non-transmission medium, which can be used to store information that can be accessed by a computing device. Computer-readable media, as defined herein, does not include transitory computer-readable media (transmission media), such as modulated data signals and carrier waves.
These computer programs may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart block or blocks and/or block diagram block or blocks, and corresponding steps may be implemented in different modules.
Such an apparatus or system is provided in this embodiment. The device is called a machine learning-based eutectic high-entropy alloy composition design device and comprises: the first acquisition module is used for acquiring training data, wherein the training data comprises input data and output data, the input data is a component of an alloy, the output data is a phase composition of the alloy, and the phase composition is a primary phase mole fraction generated in the solidification process of the alloy; the training module is used for inputting the training data into a pre-selected machine learning model for training to obtain a trained model; the second acquisition module is used for acquiring a plurality of groups of alloy components with primary phase mole fractions within a preset range through the model, wherein the alloy with the primary phase mole fractions within the preset range is the eutectic high-entropy alloy; the analysis module is used for carrying out statistical analysis on the plurality of groups of alloy components to obtain key elements with important influence on eutectic formation and elements strongly related to the key elements; and the prediction module is used for adjusting the contents of the key elements and elements strongly related to the key elements and predicting the eutectic high-entropy alloy to obtain the components of the eutectic high-entropy alloy.
The system or the device is used for realizing the functions of the method in the above embodiment, and each module in the system or the device corresponds to each step in the method, which has been described in the method, and will not be described herein.
For example, the first acquisition module is configured to: grasping alloy components and primary phase mole fractions in published literature data; and (5) selecting alloy components by self and obtaining thermodynamic calculation results.
For another example, the training module is to: and under the condition that the machine learning model is an SVM model, acquiring a decision coefficient, and determining to finish training of the model when the decision coefficient is greater than or equal to a preset threshold value to obtain a trained model.
For another example, the prediction module is configured to: the calhatd method was used to predict eutectic high entropy alloys. Optionally, the method further comprises: and the verification module is used for carrying out experimental verification on the components of the eutectic high-entropy alloy obtained through prediction and the nearby component alloys with the preset group number.
The problem caused by the fact that the eutectic high-entropy alloy is predicted by the existing method is solved through the embodiment, so that blindness of the eutectic high-entropy alloy predicted by the existing method is avoided, and the efficiency of the eutectic high-entropy alloy design is improved.
The foregoing is merely exemplary of the present application and is not intended to limit the present application. Various modifications and variations of the present application will be apparent to those skilled in the art. Any modification, equivalent replacement, improvement, etc. which come within the spirit and principles of the application are to be included in the scope of the claims of the present application.
Claims (6)
1. The method for designing the eutectic high-entropy alloy composition based on machine learning is characterized by comprising the following steps of:
Obtaining training data, wherein the training data comprises input data and output data, the input data is a component of an alloy, the output data is a phase composition of the alloy, and the phase composition is a primary phase mole fraction generated in the solidification process of the alloy;
Inputting the training data into a pre-selected machine learning model for training to obtain a trained model; the selected machine learning model is an SVM model, alloy components are used as input, primary phase mole fractions corresponding to the alloy components are used as output, the machine learning model is trained by utilizing a database, the predicted performance of the model is quantitatively evaluated by adopting a decision coefficient R2, and the training of the model is completed when R2 is more than or equal to 0.9;
Obtaining a plurality of groups of alloy components with primary phase mole fractions within a preset range through the model, wherein the alloy with the primary phase mole fractions within the preset range is the eutectic high-entropy alloy;
Carrying out statistical analysis on the plurality of groups of alloy components to obtain key elements with important influence on eutectic formation and elements strongly related to the key elements; wherein the primary phase mole fraction of the near-eutectic high-entropy alloy component is less than or equal to 4%, and the obtained near-eutectic high-entropy alloy component is not less than 300 groups to ensure the accuracy of statistical analysis results;
Adjusting the contents of the key elements and elements strongly related to the key elements, and predicting the eutectic high-entropy alloy to obtain the components of the eutectic high-entropy alloy; wherein, the CALPHAD method is used for predicting the eutectic high-entropy alloy.
2. The method of claim 1, wherein obtaining the training data comprises:
grabbing alloy components in the published literature data and obtaining thermodynamic calculation results;
And storing the grabbed alloy components and the thermodynamic calculation result in a database as the training data.
3. The method according to any one of claims 1 to 2, further comprising:
and carrying out experimental verification on the predicted components of the eutectic high-entropy alloy and the nearby component alloys with the preset group numbers.
4. A machine learning-based eutectic high-entropy alloy composition design apparatus, comprising:
The first acquisition module is used for acquiring training data, wherein the training data comprises input data and output data, the input data is a component of an alloy, the output data is a phase composition of the alloy, and the phase composition is a primary phase mole fraction generated in the solidification process of the alloy;
The training module is used for inputting the training data into a pre-selected machine learning model for training to obtain a trained model; the selected machine learning model is an SVM model, alloy components are used as input, primary phase mole fractions corresponding to the alloy components are used as output, the machine learning model is trained by utilizing a database, the predicted performance of the model is quantitatively evaluated by adopting a decision coefficient R2, and the training of the model is completed when R2 is more than or equal to 0.9;
The second acquisition module is used for acquiring a plurality of groups of alloy components with primary phase mole fractions within a preset range through the model, wherein the alloy with the primary phase mole fractions within the preset range is the eutectic high-entropy alloy;
The analysis module is used for carrying out statistical analysis on the plurality of groups of alloy components to obtain key elements with important influence on eutectic formation and elements strongly related to the key elements; wherein the primary phase mole fraction of the near-eutectic high-entropy alloy component is less than or equal to 4%, and the obtained near-eutectic high-entropy alloy component is not less than 300 groups to ensure the accuracy of statistical analysis results;
the prediction module is used for adjusting the contents of the key elements and elements strongly related to the key elements and predicting the eutectic high-entropy alloy to obtain the components of the eutectic high-entropy alloy; wherein, the CALPHAD method is used for predicting the eutectic high-entropy alloy.
5. The apparatus of claim 4, wherein the first acquisition module is configured to:
grabbing alloy components in the published literature data and obtaining thermodynamic calculation results;
And storing the grabbed alloy components and the thermodynamic calculation result in a database as the training data.
6. The apparatus according to any one of claims 4 to 5, further comprising:
and the verification module is used for carrying out experimental verification on the components of the eutectic high-entropy alloy obtained through prediction and the nearby component alloys with the preset group number.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111232545.1A CN113870957B (en) | 2021-10-22 | 2021-10-22 | Eutectic high-entropy alloy composition design method and device based on machine learning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111232545.1A CN113870957B (en) | 2021-10-22 | 2021-10-22 | Eutectic high-entropy alloy composition design method and device based on machine learning |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113870957A CN113870957A (en) | 2021-12-31 |
| CN113870957B true CN113870957B (en) | 2024-09-03 |
Family
ID=78997128
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111232545.1A Active CN113870957B (en) | 2021-10-22 | 2021-10-22 | Eutectic high-entropy alloy composition design method and device based on machine learning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113870957B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116721724B (en) * | 2023-06-20 | 2024-07-30 | 小米汽车科技有限公司 | Alloy performance prediction method and device, storage medium, electronic equipment and chip |
Family Cites Families (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107739958B (en) * | 2017-10-24 | 2019-10-11 | 辽宁工业大学 | A kind of high-entropy alloy and preparation method thereof containing eutectic structure |
| US11359266B2 (en) * | 2018-11-20 | 2022-06-14 | City University Of Hong Kong | High entropy alloy structure and a method of preparing the same |
| KR102198677B1 (en) * | 2018-12-12 | 2021-01-05 | 포항공과대학교 산학협력단 | Design method for high entropy alloys using computer simulation method |
| CN110010210B (en) * | 2019-03-29 | 2021-03-19 | 北京科技大学 | Multi-component alloy component design method based on machine learning and oriented to performance requirements |
| CN112216356A (en) * | 2020-10-22 | 2021-01-12 | 哈尔滨理工大学 | High-entropy alloy hardness prediction method based on machine learning |
| CN112951343B (en) * | 2021-01-20 | 2022-08-30 | 桂林电子科技大学 | Machine learning-based iron-based amorphous nano soft magnetic alloy design method |
-
2021
- 2021-10-22 CN CN202111232545.1A patent/CN113870957B/en active Active
Non-Patent Citations (1)
| Title |
|---|
| Design of NiCoCrAl eutectic high entropy alloys by combining machine learning with CALPHAD method;Liu Feng 等;Materials Today Communications;20220120(第30期);第103172页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113870957A (en) | 2021-12-31 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Li et al. | Unraveling the dislocation–precipitate interactions in high-entropy alloys | |
| Stryzhyboroda et al. | Phase equilibria in the Al–Co–Cr–Fe–Ni high entropy alloy system: thermodynamic description and experimental study | |
| Nagase et al. | Alloy design and fabrication of ingots in Cu-Zn-Mn-Ni-Sn high-entropy and Cu-Zn-Mn-Ni medium-entropy brasses | |
| Sato et al. | Roles of microalloying elements on the cluster formation in the initial stage of phase decomposition of Al-based alloys | |
| Yeh et al. | On the solidification and phase stability of a Co-Cr-Fe-Ni-Ti high-entropy alloy | |
| Shafiei et al. | A cobalt-rich eutectic high-entropy alloy in the system Al–Co–Cr–Fe–Ni | |
| Dorin et al. | Micro-segregation and precipitates in as-solidified Al-Sc-Zr-(Mg)-(Si)-(Cu) alloys | |
| Li et al. | Correlative characterization of primary Al3 (Sc, Zr) phase in an Al–Zn–Mg based alloy | |
| Gupta et al. | Theoretical study of the influence of microalloying on sensitization of AA5083 and moderation of sensitization of a model Al-Mg-Mn alloy via Sr additions | |
| Liu et al. | Design of NiCoCrAl eutectic high entropy alloys by combining machine learning with CALPHAD method | |
| Jain et al. | Artificial neural network approach for microhardness prediction of eight component FeCoNiCrMnVAlNb eutectic high entropy alloys | |
| Zhang et al. | Morphology, thermal stability, electronic structure and mechanical properties of α-AlFeMnSi phases with varying Mn/Fe atomic ratios: Experimental studies and DFT calculations | |
| CN113870957B (en) | Eutectic high-entropy alloy composition design method and device based on machine learning | |
| Yan et al. | Nucleation and growth of Al3Sc precipitates during isothermal aging of Al-0.55 wt% Sc alloy | |
| Churyumov et al. | Effect of Nb addition on microstructure and thermal and mechanical properties of Fe-Co-Ni-Cu-Cr multiprincipal-element (high-entropy) alloys in as-cast and heat-treated state | |
| Lekakh | Engineering nucleation kinetics of graphite nodules in inoculated cast iron for reducing porosity | |
| Singh et al. | Effects of Cr alloying on the evolution of solidification microstructure and phase transformations of high-Nb containing γ-TiAl based alloys | |
| Lekakh | Searching for graphite nodule nuclei using automated SEM/EDX analysis | |
| Yang et al. | Dissolution behavior of dizinc magnesium phase in Al–Zn–Mg–Cu alloy | |
| Wang et al. | Uncovering the effects of Ce and superheat temperature on Fe-rich intermetallic and microporosity formation in aluminum alloy | |
| Jamaly et al. | Microstructure, macrosegregation, and thermal analysis of direct chill cast aa5182 aluminum alloy | |
| Li et al. | Characterization of the convoluted 3d intermetallic phases in a recycled al alloy by synchrotron x-ray tomography and machine learning | |
| Chen et al. | Phase field simulation of microstructure evolution and process optimization during homogenization of additively manufactured Inconel 718 alloy | |
| Lekakh | Effect of nonmetallic inclusions on solidification of inoculated spheroidal graphite iron | |
| Agrahari et al. | Effect of cooling rate on microstructures and mechanical property of Al 1230 alloy in a sand casting process |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |