CN113061767A - Research method for high-temperature deformation behavior of tungsten-rhenium-hafnium carbide alloy - Google Patents

Research method for high-temperature deformation behavior of tungsten-rhenium-hafnium carbide alloy Download PDF

Info

Publication number
CN113061767A
CN113061767A CN202110295972.8A CN202110295972A CN113061767A CN 113061767 A CN113061767 A CN 113061767A CN 202110295972 A CN202110295972 A CN 202110295972A CN 113061767 A CN113061767 A CN 113061767A
Authority
CN
China
Prior art keywords
tungsten
rhenium
strain
temperature
hafnium carbide
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.)
Pending
Application number
CN202110295972.8A
Other languages
Chinese (zh)
Inventor
李延超
张文
林小辉
梁静
薛建嵘
张新
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Northwest Institute for Non Ferrous Metal Research
Original Assignee
Northwest Institute for Non Ferrous Metal Research
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Northwest Institute for Non Ferrous Metal Research filed Critical Northwest Institute for Non Ferrous Metal Research
Priority to CN202110295972.8A priority Critical patent/CN113061767A/en
Publication of CN113061767A publication Critical patent/CN113061767A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • C—CHEMISTRY; METALLURGY
    • C22—METALLURGY; FERROUS OR NON-FERROUS ALLOYS; TREATMENT OF ALLOYS OR NON-FERROUS METALS
    • C22C—ALLOYS
    • C22C1/00—Making non-ferrous alloys
    • C22C1/04—Making non-ferrous alloys by powder metallurgy
    • C22C1/05—Mixtures of metal powder with non-metallic powder
    • C—CHEMISTRY; METALLURGY
    • C22—METALLURGY; FERROUS OR NON-FERROUS ALLOYS; TREATMENT OF ALLOYS OR NON-FERROUS METALS
    • C22C—ALLOYS
    • C22C27/00—Alloys based on rhenium or a refractory metal not mentioned in groups C22C14/00 or C22C16/00
    • C22C27/04—Alloys based on tungsten or molybdenum
    • C—CHEMISTRY; METALLURGY
    • C22—METALLURGY; FERROUS OR NON-FERROUS ALLOYS; TREATMENT OF ALLOYS OR NON-FERROUS METALS
    • C22C—ALLOYS
    • C22C32/00—Non-ferrous alloys containing at least 5% by weight but less than 50% by weight of oxides, carbides, borides, nitrides, silicides or other metal compounds, e.g. oxynitrides, sulfides, whether added as such or formed in situ
    • C22C32/0047—Non-ferrous alloys containing at least 5% by weight but less than 50% by weight of oxides, carbides, borides, nitrides, silicides or other metal compounds, e.g. oxynitrides, sulfides, whether added as such or formed in situ with carbides, nitrides, borides or silicides as the main non-metallic constituents
    • C22C32/0052—Non-ferrous alloys containing at least 5% by weight but less than 50% by weight of oxides, carbides, borides, nitrides, silicides or other metal compounds, e.g. oxynitrides, sulfides, whether added as such or formed in situ with carbides, nitrides, borides or silicides as the main non-metallic constituents only carbides
    • C—CHEMISTRY; METALLURGY
    • C22—METALLURGY; FERROUS OR NON-FERROUS ALLOYS; TREATMENT OF ALLOYS OR NON-FERROUS METALS
    • C22F—CHANGING THE PHYSICAL STRUCTURE OF NON-FERROUS METALS AND NON-FERROUS ALLOYS
    • C22F1/00—Changing the physical structure of non-ferrous metals or alloys by heat treatment or by hot or cold working
    • C22F1/16—Changing the physical structure of non-ferrous metals or alloys by heat treatment or by hot or cold working of other metals or alloys based thereon
    • C22F1/18—High-melting or refractory metals or alloys based thereon
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/20—Design optimisation, verification or simulation
    • G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/044—Recurrent networks, e.g. Hopfield networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • G06N3/084—Backpropagation, e.g. using gradient descent
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2111/00—Details relating to CAD techniques
    • G06F2111/10—Numerical modelling
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2113/00—Details relating to the application field
    • G06F2113/26—Composites
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
    • G06F2119/08—Thermal analysis or thermal optimisation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
    • G06F2119/14—Force analysis or force optimisation, e.g. static or dynamic forces

Landscapes

  • Engineering & Computer Science (AREA)
  • Chemical & Material Sciences (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Materials Engineering (AREA)
  • Mechanical Engineering (AREA)
  • Metallurgy (AREA)
  • Evolutionary Computation (AREA)
  • Organic Chemistry (AREA)
  • General Engineering & Computer Science (AREA)
  • Artificial Intelligence (AREA)
  • Software Systems (AREA)
  • General Physics & Mathematics (AREA)
  • Computational Linguistics (AREA)
  • Health & Medical Sciences (AREA)
  • Mathematical Physics (AREA)
  • Molecular Biology (AREA)
  • General Health & Medical Sciences (AREA)
  • Computing Systems (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Biomedical Technology (AREA)
  • Biophysics (AREA)
  • Data Mining & Analysis (AREA)
  • Geometry (AREA)
  • Computer Hardware Design (AREA)
  • Medical Informatics (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Thermal Sciences (AREA)
  • Crystallography & Structural Chemistry (AREA)
  • Investigating Strength Of Materials By Application Of Mechanical Stress (AREA)

Abstract

The invention discloses a method for researching high-temperature deformation behavior of a tungsten-rhenium-hafnium carbide alloy, which comprises the following steps: weighing W powder, Re powder and HfC powder, performing high-energy ball milling, and performing hot-pressing sintering to obtain a tungsten-rhenium-hafnium carbide alloy; secondly, stress and strain data under different conditions are obtained through a thermal compression test, and then a constitutive relation model and an artificial neural network training model are established; judging the prediction accuracy of the two models and selecting a model with high accuracy for predicting the high-temperature deformation behavior of the tungsten-rhenium-hafnium carbide alloy; and fourthly, researching the structure and texture evolution of the sample in the hot compression test, and combining the prediction result to obtain the high-temperature deformation behavior of the tungsten-rhenium-hafnium carbide alloy. According to the invention, a stress-strain curve obtained by a tungsten-rhenium-hafnium carbide alloy hot compression test is combined with a constitutive equation, an artificial neural network and a microstructure, so that the relation between the deformation behavior of the tungsten-rhenium-hafnium carbide alloy and the dynamic recovery and dynamic recrystallization of the microstructure is disclosed, and the tungsten-rhenium-hafnium carbide alloy deformation behavior is more predictable.

Description

Research method for high-temperature deformation behavior of tungsten-rhenium-hafnium carbide alloy
Technical Field
The invention belongs to the technical field of metal materials, and particularly relates to a method for researching high-temperature deformation behavior of a tungsten-rhenium-hafnium carbide alloy.
Background
Tungsten (W) has unique mechanical properties including high melting point (3410 ℃), high modulus, excellent heat resistance and ablative properties, and is a desirable material for high temperature applications, ranging from turbines, fusion reactors and kinetic energy penetrators. But the service strength of pure tungsten at high temperature is greatly reduced. Alloying is an effective way to improve the strength of the material, and the addition of rhenium (Re) has been shown to improve the room temperature ductility and high temperature strength of tungsten (W). However, rhenium cannot be added in large quantities to enhance the strength of the metal due to the small reserves and high price. The combination of refractory metals as the matrix and high temperature ceramics as the reinforcement creates a series of composite materials with unique chemical, thermal and mechanical properties. Among high temperature ceramic reinforcements, hafnium carbide (HfC), which has the highest melting point and superior mechanical properties, is considered to be the most effective second phase particle for high temperature strengthening W. Dongju Lee et al report that mixed carbides may be beneficial to the reinforcement of HfC-W composite materials, form strong interface bonds, and effectively transfer loads to HfC particles; meanwhile, the W-3Re-5HfC alloy has good high-temperature mechanical property, and the W-3Re-5HfC alloy can also be applied to a reducing atmosphere or an inert atmosphere and has good compatibility with various components in application occasions, such as hydrazine, nitrogen, hydrogen, ammonia and the like in a propeller.
In recent years, many researchers have studied the deformation process of tungsten-based alloys and have established constitutive equation models. The thermal deformation behavior of the W-10 wt% Cu composite material at high temperature is researched by a thermo-compression test in the eastern Lin nations and the like, and a stress-strain curve of pure tungsten at room temperature and at a low strain rate is obtained; the Zhou Li establishes a relationship between the rheological stress, the strain rate, and the strain temperature based on the Arrhenius model. Furthermore, studies suggest that dynamic recrystallization of tungsten alloys occurs only at high temperatures and low strain rates, with stress increasing with decreasing strain temperature or increasing strain rate.
In order to meet the performance requirements of different alloy applications, it is necessary to obtain the required texture and mechanical properties by hot deformation. Therefore, optimizing the heat treatment process parameters is the most effective method for controlling the structure and the performance of the tungsten-rhenium-hafnium carbide alloy. However, the tungsten-rhenium-hafnium carbide alloy is a material difficult to deform, the structure and the performance of the tungsten-rhenium-hafnium carbide alloy are more sensitive to deformation process parameters, and the high-temperature deformation behavior of the tungsten-rhenium-hafnium carbide alloy is more complex. The constitutive equation of the material provides a theoretical basis for reducing deformation defects, and the process design time is saved, so that the material with excellent organization and performance is obtained. But reports on the research on the HfC doped tungsten-rhenium alloy high-temperature deformation mechanism and the constitutive equation model are not found. Therefore, designing a W-3Re-5HfC alloy high-temperature deformation behavior research method, and establishing a high-precision constitutive model or an Artificial Neural Network (ANN) model is very necessary for further researching and understanding the high-temperature deformation behavior of the tungsten-rhenium-hafnium carbide alloy.
Disclosure of Invention
The invention aims to solve the technical problem of providing a method for researching the high-temperature deformation behavior of the tungsten-rhenium-hafnium carbide alloy aiming at the defects of the prior art. According to the method, the relation between the deformation behavior of the tungsten-rhenium-hafnium carbide alloy and the dynamic recovery and dynamic recrystallization of the microstructure is better disclosed through a research method of combining a stress-strain curve obtained by a hot compression test of the tungsten-rhenium-hafnium carbide alloy with a constitutive equation, an artificial neural network and a microstructure, so that the deformation behavior of the tungsten-rhenium-hafnium carbide alloy is better predicted.
In order to solve the technical problems, the technical scheme adopted by the invention is as follows: a method for researching high-temperature deformation behavior of a tungsten-rhenium-hafnium carbide alloy is characterized by comprising the following steps of:
respectively weighing W powder, Re powder and HfC powder as raw material powder, performing mechanical alloying treatment by adopting high-energy ball milling to obtain mixed alloy powder, and then performing hot-pressing sintering to obtain a tungsten-rhenium-hafnium carbide alloy;
secondly, taking a sample from the tungsten-rhenium-hafnium carbide alloy obtained in the first step, carrying out a hot compression test to obtain stress and strain data under the conditions of different strain temperatures and different strain rates, and then respectively establishing a constitutive relation model and training an artificial neural network model;
thirdly, judging the prediction accuracy of the constitutive relation model established in the second step and the trained artificial neural network model, and selecting a model with high accuracy for predicting the high-temperature deformation behavior of the tungsten-rhenium-hafnium carbide alloy;
and step four, studying the structure and texture evolution of the sample under the conditions of different strain temperatures and different strain rates in the thermal compression test in the step two, and obtaining the high-temperature deformation behavior of the tungsten-rhenium-hafnium carbide alloy by combining the prediction result of the model selected in the step three on the high-temperature deformation behavior of the tungsten-rhenium-hafnium carbide alloy.
The method disclosed by the invention is characterized in that a method combining mechanical alloying and hot-pressing sintering is adopted to prepare the tungsten-rhenium-hafnium carbide alloy, a hot-pressing test is carried out to obtain stress and strain data, then a constitutive relation model and an artificial neural network training model are respectively established to predict the high-temperature deformation behavior of the tungsten-rhenium-hafnium carbide alloy, a more appropriate prediction method is determined through comparison, the high-temperature deformation behavior of the tungsten-rhenium-hafnium carbide alloy is obtained by combining the research on the tissue and texture evolution of a hot-pressing test sample, and the relation between the deformation behavior of the tungsten-rhenium-hafnium carbide alloy and the dynamic recovery and dynamic recrystallization of the microstructure is better disclosed through a research method combining a stress-strain curve with a constitutive equation, an artificial neural network and a microstructure, so that the tungsten-rhenium-hafnium carbide alloy has better prediction on the deformation behavior.
The method for researching the high-temperature deformation behavior of the tungsten-rhenium-hafnium carbide alloy is characterized in that in the step one, the high-energy ball milling is carried out by adopting a planetary high-energy ball mill, the rotating speed of the high-energy ball milling is 350-450 rpm, the time is 35-45 h, and the ball-to-material ratio is 10-15: 1; and after high-energy ball milling, putting the mixed alloy powder into a graphite die for hot-pressing sintering at 2100 ℃ for 30min, and cooling along with the furnace to obtain the tungsten-rhenium-hafnium carbide composite material.
The method for researching the high-temperature deformation behavior of the tungsten-rhenium-hafnium carbide alloy is characterized in that in the second step, the strain temperature is 1200-1500 ℃, and the strain rate is 0.001s-1～1s-1。
The method for researching the high-temperature deformation behavior of the tungsten-rhenium-hafnium carbide alloy is characterized in that in the second step, the thermal compression test is carried out by adopting a Gleeble-1500 thermal simulation testing machine, the sample is a cylindrical sample with the diameter of 6mm and the height of 9mm, the strain temperatures are 1200 ℃, 1300 ℃, 1400 ℃ and 1500 ℃, and the strain rates are 0.001s-1、0.01s-1、0.1s-1、1s-1And in the process of the thermal compression test, all samples reach the strain temperature at the heating rate of 10 ℃/s and are kept for 0.5min to 1.5 min.
The method for researching the high-temperature deformation behavior of the tungsten-rhenium-hafnium carbide alloy is characterized in that in the step two, the specific process of taking the sample on the tungsten-rhenium-hafnium carbide alloy to perform the hot compression test is as follows: firstly, the sample is chiseled to have an aperture X depth of
Figure BDA0002984366310000046
Inserting a high-temperature thermocouple wire into the hole, fixing the thermocouple wire by using high-temperature cement, then pasting a tantalum foil with the thickness of 0.1mm on the surface of the sample, reaching the strain temperature at the heating rate of 10 ℃/s, and keeping the strain temperature for 0.5-1.5 min to ensure the temperature homogeneity of the sampleAnd (3) selecting a strain rate, compressing the sample along the axial direction by 50%, and immediately performing water quenching, wherein the delay time is not more than 3s so as to keep the microstructure of the sample after the thermal compression deformation.
The method for researching the high-temperature deformation behavior of the tungsten-rhenium-hafnium carbide alloy is characterized in that the establishing process of the constitutive relation model in the step two comprises the following steps:
(1) a simplified Arrhennius equation is used as a constitutive relation model, namely:
Figure BDA0002984366310000041
wherein A, alpha, beta and n1、n2Q is the strain activation energy in kJ. mol for a material constant independent of the specific strain temperature-1R is a general gas constant of 8.31 J.mol-1T is
The absolute temperature, in units of K,
Figure BDA0002984366310000045
to strain rate, σ is stress; the first formula in equation (1), the power law, applies to α σ<0.8, the second equation, the exponential law, applies to σ>1.2, the third formula, namely the hyperbolic sine law formula is applicable to alpha sigma which is more than or equal to 0.8 and less than or equal to 1.2;
(2) the logarithm was taken simultaneously on both sides of equation (1) and the results are shown below:
Figure BDA0002984366310000042
Figure BDA0002984366310000043
Figure BDA0002984366310000044
according to the formulas (2) and (3)
Figure BDA0002984366310000051
And
Figure BDA0002984366310000052
the mean slope of two straight lines is n1And β, α is defined as n1And beta;
linear fitting according to formula (4) to obtain
Figure BDA0002984366310000053
And ln [ sinh (. alpha. sigma.)]-a fitted line plot of 1/T, the mean slopes of the two fitted lines being n2And k, and the slope k ═ ln [ sinh (α σ)]Q/n2R, obtaining a calculation formula of deformation activation energy Q:
Figure BDA0002984366310000054
(3) selecting peak stress as a stress value to calculate the equation, and selecting Z, namely a Zener-Hollomon parameter as a temperature compensation factor to express the synergistic effect of the strain rate and the deformation temperature, wherein Z is expressed as:
Figure BDA0002984366310000055
substituting a hyperbolic sine function of a hyperbolic sine law formula, which is a third formula in formula (1), into formula (6) to obtain formula (7), wherein n in formula (7) represents a pressure index:
Z＝A[sinh(ασ)]n (7)
taking logarithm of two sides of the formula (7) to obtain a formula (8):
lnZ＝lnA+nln[sinh(ασ)] (8)
according to formula (8) pair lnZ and ln [ sinh (. alpha. sigma.)]Linear fitting was performed to obtain the slope n-2.06709 and intercept lnA-1.62 × 10 of the fitted line15Therefore, equation (1) is simplified as:
Figure BDA0002984366310000056
(4) the stress is rewritten into hyperbolic sine function formula (10) according to formula (9), and the numerical values of the parameters are substituted to obtain formula (11), which is the constitutive equation of the tungsten-rhenium-hafnium carbide alloy, as follows:
Figure BDA0002984366310000057
Figure BDA0002984366310000061
the method for researching the high-temperature deformation behavior of the tungsten-rhenium-hafnium carbide alloy is characterized in that the process of training the artificial neural network model in the step two is as follows:
(1) according to stress and strain data of a thermal compression test under the conditions of different strain temperatures and different strain rates, strain epsilon and strain rate
Figure BDA0002984366310000062
And normalizing the strain temperature T and the stress sigma, wherein the calculation formula of the normalized data X' is shown as the formula (12):
Figure BDA0002984366310000063
wherein X is the original data, X' is the normalized data corresponding to X, XminAnd XmaxThe minimum and maximum values of X, respectively;
(2) constructing a BP neural network model;
(3) the strain epsilon and the strain rate after the normalization treatment in the step (1) are
Figure BDA0002984366310000064
And the strain temperature T is used as an input layer, the stress sigma after normalization treatment is used as an output layer, and the constructed BP neural is inputAnd training the BP neural network model through the network model to obtain the trained BP neural network model.
The method for researching the high-temperature deformation behavior of the tungsten-rhenium-hafnium carbide alloy is characterized in that the sample before and after the hot compression test in the fourth step is sliced along the compression axis, and the sliced sample is processed by 120#～800#Sequentially grinding silicon carbide sand paper, polishing alumina slurry with the granularity of 0.3-5 mu M, electrolytically corroding with 1M NaOH solution for 1-3 s, and observing and analyzing the microstructure by adopting an SEM method.
Compared with the prior art, the invention has the following advantages:
1. according to the invention, through a research method of combining a stress-strain curve obtained by a tungsten-rhenium-hafnium carbide alloy hot compression test with a constitutive equation, an artificial neural network and a microstructure, the relation between the deformation behavior of the tungsten-rhenium-hafnium carbide alloy and the dynamic recovery and dynamic recrystallization of the microstructure is better disclosed, so that the method has better predictability on the deformation behavior of the tungsten-rhenium-hafnium carbide alloy.
2. According to the invention, the high-temperature thermocouple wire is fixed in the sample drill hole in the thermal compression test process, so that the problem of difficult welding of the tungsten-rhenium-hafnium carbide alloy due to high hardness and high strength is solved, the pressure head is prepared by adopting the 5% ZrC material with higher hardness and better high-temperature resistance, the friction between the sample surface and the clamp is reduced by pasting the tantalum foil on the sample surface, the smooth operation of the thermal compression test process is ensured, and the water quenching is carried out immediately after the thermal compression test process so as to ensure the microstructure of the sample after deformation.
3. Aiming at the high-temperature deformation behavior of the tungsten-rhenium-hafnium carbide alloy, the invention respectively adopts the constitutive relation model establishment and the artificial neural network model training for prediction, and compares the correlation coefficient R and the average absolute relative error AARE predicted by the two models to obtain better prediction accuracy of the artificial neural network model and better prediction on the high-temperature deformation behavior of the tungsten-rhenium-hafnium carbide alloy.
The technical solution of the present invention is further described in detail by the accompanying drawings and examples.
Drawings
FIG. 1a is a stress-strain curve of W-3Re-5HfC alloy at 1200 ℃ at different strain rates in example 1 of the present invention.
FIG. 1b is a stress-strain curve of the W-3Re-5HfC alloy of example 1 of the present invention at 1300 ℃ and different strain rates.
FIG. 1c is a stress-strain curve of W-3Re-5HfC alloy of example 1 of the present invention at 1400 ℃ and different strain rates.
FIG. 1d is a stress-strain curve of the W-3Re-5HfC alloy of example 1 of the present invention at 1500 ℃ and different strain rates.
Fig. 2 is a BP neural network model in embodiment 1 of the present invention.
Fig. 3a is a graph showing a relationship between a stress experimental value and a predicted value of a constitutive equation model in embodiment 1 of the present invention.
Fig. 3b is a graph showing a relationship between a stress experimental value and a predicted value of a BP neural network model in embodiment 1 of the present invention.
FIG. 4a is a diagram illustrating a relative error distribution of the constitutive equation model prediction in example 1 of the present invention.
Fig. 4b is a relative error distribution diagram of the BP neural network model prediction in embodiment 1 of the present invention.
FIG. 5a is a graph showing the relationship between the experimental stress value of the W-3Re-5HfC alloy at 1200 ℃ and different strain rates, the predicted value of the constitutive equation model, and the predicted value of the BP neural network model in example 1 of the present invention.
FIG. 5b is a graph showing the relationship between the experimental stress value of the W-3Re-5HfC alloy at 1300 ℃ and different strain rates, the predicted value of the constitutive equation model, and the predicted value of the BP neural network model in example 1 of the present invention.
FIG. 5c is a graph showing the relationship between the experimental stress value of the W-3Re-5HfC alloy at 1400 ℃ and different strain rates, the predicted value of the constitutive equation model, and the predicted value of the BP neural network model in example 1 of the present invention.
FIG. 5d is a graph showing the relationship between the experimental stress value of the W-3Re-5HfC alloy at 1500 ℃ and different strain rates, the predicted value of the constitutive equation model, and the predicted value of the BP neural network model in example 1 of the present invention.
FIG. 6a is an initial microstructure view of a sample of the W-3Re-5HfC alloy of example 1 of the present invention.
FIG. 6b shows the W-3Re-5HfC alloy samples of example 1 of the present invention at 1200 ℃ for 1s-1Microstructure after thermo-compression deformation under the conditions.
FIG. 6c shows the W-3Re-5HfC alloy samples of example 1 of the present invention at 1300 ℃ for 1s-1Microstructure after thermo-compression deformation under the conditions.
FIG. 6d shows the W-3Re-5HfC alloy samples of example 1 of the present invention at 1300 ℃ for 0.001s-1Microstructure after thermo-compression deformation under the conditions.
FIG. 6e shows the W-3Re-5HfC alloy samples of example 1 of the present invention at 1400 ℃ for 0.001s-1Microstructure after thermo-compression deformation under the conditions.
FIG. 6f shows the temperature of the W-3Re-5HfC alloy sample at 1500 ℃ for 0.001s in example 1 of the present invention-1Microstructure after thermo-compression deformation under the conditions.
Detailed Description
Example 1
The method for researching the high-temperature deformation behavior of the tungsten-rhenium-hafnium carbide alloy comprises the following steps of:
step one, respectively weighing W powder, Re powder and HfC powder according to the component content of each element in the tungsten-rhenium-hafnium carbide composite material, and carrying out high-energy ball milling under the protection of inert gas for mechanical alloying treatment to obtain mixed alloy powder; the mass purity of the W powder is 99.9%, the average particle size is 5 microns, the mass purity of the Re powder is 99.95%, the average particle size is 2 microns, the mass purity of the HfC powder is 99.9%, and the average particle size is 1 micron;
the high-energy ball milling is carried out by adopting a QM-3SP2 planetary high-energy ball mill, the adopted ball milling tank is an agate tank, the grinding balls are agate balls, the high-energy ball milling adopts the rotation speed of 400rpm, the time is 40h, and the ball-to-material ratio is 12: 1;
adding absolute ethyl alcohol as a control agent in the high-energy ball milling process, wherein the addition amount of the absolute ethyl alcohol of the control agent is 15% of the total mass of the W powder, the Re powder and the HfC powder, cleaning the ball milling tank for 3 times by adopting the absolute ethyl alcohol after the high-energy ball milling is finished, and the mass of the absolute ethyl alcohol adopted for cleaning is 10 times of the total mass of the W powder, the Re powder and the HfC powder;
the high-energy ball-milling powder obtained by high-energy ball milling is sequentially subjected to absolute ethyl alcohol ultrasonic cleaning for 30min and acetone ultrasonic cleaning for 30min, then placed in an oven at 60-70 ℃ for drying for 9h, and ground by an agate mortar to obtain mixed alloy powder;
then laying graphite paper on the inner surface of the graphite mold, filling the mixed alloy powder into the graphite mold, fixing the mixed alloy powder outside the graphite mold by adopting a carbon fiber mold, then carrying out hot-pressing sintering, and cooling along with a furnace to obtain a W-3Re-5HfC alloy; the hot-pressing sintering process comprises the following steps: under the pressure condition of 30-60 MPa, firstly heating to 1200 ℃ at the speed of 20 ℃/min, then heating to 2100 ℃ at the speed of 5 ℃/min, and keeping the temperature for 25-35 min;
the composition of the W-3Re-5HfC alloy is shown in Table 1;
TABLE 1
Figure BDA0002984366310000091
Step two, taking a sample from the W-3Re-5HfC alloy obtained in the step one, and carrying out a thermal compression test by adopting a Gleeble-1500 thermal simulation testing machine, wherein the sample is a cylindrical sample with the diameter of 6mm and the height of 9mm, and the selected strain rates are 0.001s-1、0.01s-1、0.1s-1And 1s-1The strain temperatures are 1200 ℃, 1300 ℃, 1400 ℃ and 1500 ℃ respectively, and the specific process is as follows: firstly, the sample is chiseled to have an aperture X depth of
Figure BDA0002984366310000101
Inserting a high-temperature thermocouple wire into the hole, fixing the thermocouple wire by using high-temperature cement, preparing a pressure head by adopting a 5% ZrC material with higher hardness and better high-temperature resistance, then pasting a tantalum foil with the thickness of 0.1mm on the surface of a sample, reaching the strain temperature at the heating rate of 10 ℃/s, keeping the strain temperature for 1min to ensure the temperature homogenization of the sample, selecting the strain rate to axially compress 50%, immediately performing water quenching, keeping the microstructure of the sample after thermal compression deformation for a delay time of not more than 3s, and connecting the sample by a Gleeble-1500 thermal simulation testing machineThe computer automatically records and obtains stress and strain data under different conditions, and further obtains stress-strain curves of the W-3Re-5HfC alloy under different strain temperatures and different strain rates, as shown in figures 1a to 1 d;
FIG. 1a is a stress-strain curve of the W-3Re-5HfC alloy at 1200 ℃ and different strain rates in the present embodiment, FIG. 1b is a stress-strain curve of the W-3Re-5HfC alloy at 1300 ℃ and different strain rates in the present embodiment, FIG. 1c is a stress-strain curve of the W-3Re-5HfC alloy at 1400 ℃ and different strain rates in the present embodiment, and FIG. 1d is a stress-strain curve of the W-3Re-5HfC alloy at 1500 ℃ and different strain rates in the present embodiment, and it can be seen from FIGS. 1a to 1d that the stress has two different trends: under the low temperature condition of high strain rate (figure 1 a-figure 1b), the stress-strain curve rises rapidly, then slowly and finally tends to be stable; under the conditions of higher temperature and lower strain rate (fig. 1 c-1 d), the stress-strain curve rapidly reaches the peak value and then slowly decreases to gradually form a balance, namely the balance of work hardening and dynamic softening; at the same time, the stress decreases with increasing deformation temperature at the same strain rate, which indicates that with increasing temperature, the atomic vibration and deformation enhancing lower critical shear stress with stronger activation energy and the slip system also increase, and at the same time, the increase of temperature also facilitates the nucleation process and the provided Dynamic Recrystallization (DRX) process under the thermal activation condition, thereby facilitating the deformation of the W-3Re-5HfC alloy. Furthermore, when the deformation temperature is constant, the rate of increase of the stress increases with increasing strain rate, so that the density and effect of dislocations hindering the interaction of dislocations becomes stronger, which will result in more difficult deformations and the electrical resistance of the large deformation alloy.
In summary, under four temperature variables and four strain rate variables, the real stress-strain curve in the initial stage rapidly increases with the increase of strain, and this stage is the strain hardening stage; at a lower temperature of 1200 ℃ and a high strain rate of 1s-1Under the conditions, the true stress of the W-3Re-5HfC alloy increases with strain, however, 0.001s-1、0.01s-1At a strain rate of 1300 ℃, 1400 ℃ and 1500 ℃, in the initial stageThe force increases with increasing strain, but after a certain value the stress does not increase significantly. The peak stresses of the W-3Re-5HfC alloy in this example at different deformation temperatures and at different deformation rates are shown in Table 2.
TABLE 2
Figure BDA0002984366310000111
1) Establishing constitutive relation model
(1) A simplified Arrhennius equation is used as a constitutive relation model, namely:
Figure BDA0002984366310000112
wherein A, alpha, beta and n1、n2Q is the strain activation energy in kJ. mol for a material constant independent of the specific strain temperature-1R is a general gas constant of 8.31 J.mol-1T is the absolute temperature in K,
Figure BDA0002984366310000113
to strain rate, σ is stress; the first formula in equation (1), the power law, applies to α σ<0.8, the second equation, the exponential law, applies to σ>1.2, the third formula, namely the hyperbolic sine law formula is applicable to alpha sigma which is more than or equal to 0.8 and less than or equal to 1.2;
(2) the logarithm was taken simultaneously on both sides of equation (1) and the results are shown below:
Figure BDA0002984366310000121
Figure BDA0002984366310000122
Figure BDA0002984366310000123
according to the formulas (2) and (3)
Figure BDA0002984366310000124
And
Figure BDA0002984366310000125
the mean slope of two straight lines is n1And β, α is defined as n1And beta;
linear fitting according to formula (4) to obtain
Figure BDA0002984366310000126
And ln [ sinh (. alpha. sigma.)]-a fitted line plot of 1/T, the mean slopes of the two fitted lines being n2And k, and the slope k ═ ln [ sinh (α σ)]Q/n2R, obtaining a calculation formula of deformation activation energy Q:
Figure BDA0002984366310000127
(3) selecting peak stress as a stress value to calculate the equation, and selecting Z, namely a Zener-Hollomon parameter as a temperature compensation factor to express the synergistic effect of the strain rate and the deformation temperature, wherein Z is expressed as:
Figure BDA0002984366310000128
substituting a hyperbolic sine function of a hyperbolic sine law formula, which is a third formula in formula (1), into formula (6) to obtain formula (7), wherein n in formula (7) represents a pressure index:
Z＝A[sinh(ασ)]n (7)
taking logarithm of two sides of the formula (7) to obtain a formula (8):
lnZ＝lnA+nln[sinh(ασ)] (8)
according to formula (8) pair lnZ and ln [ sinh (. alpha. sigma.)]Linear fitting was performed to obtain the slope n-2.06709 and intercept lnA-1.62 × 10 of the fitted line15Thus simplifying the formula (1)Comprises the following steps:
Figure BDA0002984366310000131
(4) the rheological stress is rewritten into hyperbolic sine function formula (10) according to formula (9), and the numerical values of the parameters are substituted to obtain formula (11), which is the constitutive equation of the tungsten-rhenium-hafnium carbide alloy, as follows:
Figure BDA0002984366310000132
Figure BDA0002984366310000133
TABLE 3
Figure BDA0002984366310000134
2) Training an artificial neural network model:
(1) according to stress and strain data of a thermal compression test under the conditions of different strain temperatures and different strain rates, strain epsilon and strain rate
Figure BDA0002984366310000135
And normalizing the strain temperature T and the stress sigma, wherein the calculation formula of the normalized data X' is shown as the formula (12):
Figure BDA0002984366310000136
wherein X is the original data, X' is the normalized data corresponding to X, XminAnd XmaxThe minimum and maximum values of X, respectively;
(2) constructing a BP neural network model, as shown in FIG. 2;
(3) the strain epsilon and the strain rate after the normalization treatment in the step (1) areRate of change
Figure BDA0002984366310000137
And taking the strain temperature T as an input layer, taking the stress sigma after normalization treatment as an output layer, and inputting the constructed BP neural network model for training to obtain the trained BP neural network model.
(3) The strain epsilon and the strain rate after the normalization treatment in the step (1) are
Figure BDA0002984366310000141
And the strain temperature T is used as an input layer, the stress sigma after normalization processing is used as an output layer, the constructed BP neural network model is input, transfer functions from the input layer to the hidden layer and from the hidden layer to the output layer respectively adopt a logarithm s-type function and linear functions namely tan sigmoid and pure linear, a training function is Trainlm, and 416 data sets (including strain epsilon and strain rate) are selected from a real stress-strain curve
Figure BDA0002984366310000142
Corresponding values and corresponding strain temperature T values), and is divided into a training data set and a testing data set, wherein 332 data sets with 0.1-0.6 true strain values and 0.1 interval are selected as training data set training network models, 84 data sets with 0.1-0.6 true strain values and 0.1 interval are used as testing data sets to test the performance of the neural network models, and in order to determine the proper number of neurons in the hidden layer, trial and error are started from two neurons in the hidden layer and then from more neurons. Research finds that a single hidden layer network consisting of 12 hidden neurons has the minimum Mean Square Error (MSE) and is the optimal structure for predicting the rheological stress of the W-3Re-5HfC alloy; after 12000 cycles of training, the BP neural network model reaches a stable state, and the trained BP neural network model is obtained;
step three, judging the prediction accuracy of the constitutive relation model established in the step two and the trained BP neural network model, and selecting a model with high accuracy for predicting the high-temperature deformation behavior of the tungsten-rhenium-hafnium carbide alloy, wherein the specific process is as follows:
the predictability of the trained BP neural network model and the constitutive equation is verified by using standard statistical parameters such as a correlation coefficient (R), an Average Absolute Relative Error (AARE), an average Root Mean Square Error (RMSE) and the like, corresponding calculation formulas are shown in (13) to (15), and the results are shown in FIGS. 3a to 3 b;
Figure BDA0002984366310000151
Figure BDA0002984366310000152
Figure BDA0002984366310000153
wherein EiAnd PiAre respectively an experimental value and a predicted value,
Figure BDA0002984366310000154
and
Figure BDA0002984366310000155
are respectively EiAnd PiN is the total number of data used for the survey.
Fig. 3a is a graph showing a relationship between a stress experimental value and a predicted value of a constitutive equation model in this embodiment, and fig. 3b is a graph showing a relationship between a stress experimental value and a predicted value of a BP neural network model in this embodiment, and as can be seen from fig. 3a and fig. 3b, correlation coefficients of the stress experimental value and the predicted value of the constitutive equation model, and correlation coefficients of the stress experimental value and the predicted value of the BP neural network model are 0.99892 and 0.99299, respectively, which show that both models have better correlation with the experimental values, corresponding average absolute relative errors are 4.09% and 0.99%, corresponding average root mean square errors are 5.60MPa and 1.43MPa, respectively, and show that the stress prediction accuracy based on the neural network model is higher than the calculation accuracy of the constitutive equation model. In addition, the performance of the constitutive equation and the trained neural network model is further researched through the statistical analysis of relative errors. Comparing the predicted results with corresponding experimental data, calculating the Relative Error (RE) as shown in formula (16), and the results are shown in fig. 4a and 4 b;
Figure BDA0002984366310000156
wherein E isiAnd PiThe meaning of (a) is the same as previously mentioned.
FIG. 4a is a relative error distribution diagram predicted by a constitutive equation model in example 1 of the present invention, FIG. 4b is a relative error distribution diagram predicted by a BP neural network model in example 1 of the present invention, Mean in the two diagrams represents Mean, SD represents standard deviation, Max represents maximum, and Min represents minimum, and it can be seen from FIG. 4a and FIG. 4b that for the constitutive equation model, the relative error ranges from 10.09% to-11.41%, and the relative error range of the artificial neural network model ranges from 8.09% to-3.44%; the BP neural network model established in the embodiment has good performance and can be used for understanding and predicting the flow behavior of the W-3Re-5HfC alloy;
FIG. 5a is a graph showing the relationship between the stress experimental value of the W-3Re-5HfC alloy at 1200 ℃ and different strain rates and the predicted value of the constitutive equation model and the predicted value of the BP neural network model in this embodiment, FIG. 5b is a graph showing the relationship between the stress experimental value of the W-3Re-5HfC alloy at 1300 ℃ and different strain rates and the predicted value of the constitutive equation model and the predicted value of the BP neural network model in this embodiment, FIG. 5c is a graph showing the relationship between the stress experimental value of the W-3Re-5HfC alloy at 1400 ℃ and different strain rates and the predicted value of the constitutive equation model and the predicted value of the BP neural network model in this embodiment, FIG. 5d is a graph showing the relationship between the stress experimental value of the W-3Re-5HfC alloy at 1500 ℃ and different strain rates and the predicted value of the constitutive equation model and, as can be seen from fig. 5a to 5d, the stress prediction value based on the BP neural network model can be kept consistent with the experimental value in the whole temperature and stress rate range, which indicates that the BP neural network model has higher precision; at the same time, under certain deformation conditions, e.g. inThe temperature is 1200 ℃ and the strain rate is 0.1s-1And 1s-1The stress predicted by the constitutive equation has deviation from an experimental value, which shows that the constitutive equation can be used for rough estimation, and the BP neural network model has higher precision and can be used for understanding and predicting the flow behavior of the W-3Re-5HfC alloy;
step four, researching the structure and texture evolution of the sample under different strain temperatures and strain rates in the thermal compression test in the step two by adopting an SEM method, wherein the specific process is as follows: slicing the sample before and after the hot compression test in the fourth step along the compression axis, and slicing the sample 120 times#～800#Sequentially grinding silicon carbide sand paper, polishing alumina slurry with the granularity of 0.3-5 mu M, electrolytically corroding with 1M NaOH solution for 1-3 s, and observing and analyzing the microstructure by adopting an SEM method, wherein the result is shown in FIGS. 6 a-6 f;
FIG. 6a is a view showing an initial microstructure of a sample of the W-3Re-5HfC alloy in the present example, and it can be seen from FIG. 6a that the initial structure of the W-3Re-5HfC alloy is composed of fine equiaxed grains having an average grain size of about 1 μm to 2 μm.
FIG. 6b shows the W-3Re-5HfC alloy samples of this example at 1200 ℃ for 1s-1The microstructure after thermo-compression deformation under the conditions, FIG. 6c is a view showing the W-3Re-5HfC alloy sample of this example at 1300 ℃ for 1s-1The microstructure after thermo-compression deformation under the conditions, FIG. 6d is a microstructure of the W-3Re-5HfC alloy sample of this example at 1300 ℃ for 0.001s-1The microstructure after thermo-compression deformation under the conditions, FIG. 6e is a view showing the W-3Re-5HfC alloy sample of this example at 1400 ℃ for 0.001s-1The microstructure after thermo-compression deformation under the conditions, FIG. 6f is a graph showing the W-3Re-5HfC alloy sample of this example at 1500 ℃ for 0.001s-1Microstructure after thermo-compression deformation under the conditions, as can be seen from FIGS. 6b to 6c, at a strain rate of 1s-1At 1200 ℃ and 1300 ℃, the initial structure is compressed to the long strip shape without recrystallization, HfC is spherical or subsphaeroidal (fig. 6c), fig. 6d is 1300 ℃ and 0.001s-1The initial dynamic recovery state at intermediate transition conditions, however, did not reach the critical recrystallization temperature and no recrystallization occurred.
As can be seen from FIG. 6e, HfC particlesThe fine grains nearby replaced a portion of the coarse primary grains, indicating that Dynamic Recrystallization (DRX) occurred under this condition. In addition, the strain rate was 0.001s at a temperature of 1400 ℃-1When the HfC particles had some micro-cracks or pores around them, as shown in black in fig. 6e, indicating severe deformation around the HfC particles, indicating that the particles had some deformation resistance, and as the temperature increased, more recrystallized grains appeared and the grain size increased (fig. 6f), it was concluded that the increased temperature caused new grains to grow coarser and the DRX fraction was higher.
The above description is only for the preferred embodiment of the present invention, and is not intended to limit the present invention in any way. Any simple modification, change and equivalent changes of the above embodiments according to the technical essence of the invention are still within the protection scope of the technical solution of the invention.

Claims (8)

1. A method for researching high-temperature deformation behavior of a tungsten-rhenium-hafnium carbide alloy is characterized by comprising the following steps of:
respectively weighing W powder, Re powder and HfC powder as raw material powder, performing mechanical alloying treatment by adopting high-energy ball milling to obtain mixed alloy powder, and then performing hot-pressing sintering to obtain a tungsten-rhenium-hafnium carbide alloy;
secondly, taking a sample from the tungsten-rhenium-hafnium carbide alloy obtained in the first step, carrying out a hot compression test to obtain stress and strain data under the conditions of different strain temperatures and different strain rates, and then respectively establishing a constitutive relation model and training an artificial neural network model;
thirdly, judging the prediction accuracy of the constitutive relation model established in the second step and the trained artificial neural network model, and selecting a model with high accuracy for predicting the high-temperature deformation behavior of the tungsten-rhenium-hafnium carbide alloy;
and step four, studying the structure and texture evolution of the sample under the conditions of different strain temperatures and different strain rates in the thermal compression test in the step two, and obtaining the high-temperature deformation behavior of the tungsten-rhenium-hafnium carbide alloy by combining the prediction result of the model selected in the step three on the high-temperature deformation behavior of the tungsten-rhenium-hafnium carbide alloy.
2. The method for researching the high-temperature deformation behavior of the tungsten-rhenium-hafnium carbide alloy according to claim 1, wherein in the step one, the high-energy ball milling is carried out by using a planetary high-energy ball mill, the high-energy ball milling is carried out at a rotating speed of 350-450 rpm for 35-45 h, and the ball-to-material ratio is 10-15: 1; and after high-energy ball milling, putting the mixed alloy powder into a graphite die for hot-pressing sintering at 2100 ℃ for 30min, and cooling along with the furnace to obtain the tungsten-rhenium-hafnium carbide composite material.
3. The method for researching high-temperature deformation behavior of the tungsten-rhenium-hafnium carbide alloy as claimed in claim 1, wherein the strain temperature in the second step is 1200-1500 ℃, and the strain rate is 0.001s-1～1s-1。
4. The method for researching the high-temperature deformation behavior of the tungsten-rhenium-hafnium carbide alloy according to claim 1, wherein the thermal compression test in the second step is performed by a Gleeble-1500 thermal simulation testing machine, the sample is a cylindrical sample with a diameter of 6mm and a height of 9mm, the strain temperature is 1200 ℃, 1300 ℃, 1400 ℃ and 1500 ℃, and the strain rate is 0.001s-1、0.01s-1、0.1s-1、1s-1And in the process of the thermal compression test, all samples reach the strain temperature at the heating rate of 10 ℃/s and are kept for 0.5min to 1.5 min.
5. The method for researching the high-temperature deformation behavior of the tungsten-rhenium-hafnium carbide alloy according to claim 1, wherein the specific process of taking the sample on the tungsten-rhenium-hafnium carbide alloy and performing the hot compression test in the step two is as follows: firstly, the sample is chiseled to have an aperture X depth of
Figure FDA0002984366300000025
A high temperature thermocouple wire is inserted into the hole and fixed by high temperature cement, and then the thickness of the thermocouple wire is adhered to the surface of the sample to be 0.1mmThe tantalum foil reaches the strain temperature at the heating rate of 10 ℃/s and is kept for 0.5 min-1.5 min to ensure the homogenization of the temperature of the sample, then the strain rate is selected to be compressed by 50 percent along the axial direction, water quenching is immediately carried out, and the delay time is not more than 3s to keep the microstructure of the sample after the thermal compression deformation.
6. The method for studying the high-temperature deformation behavior of the tungsten-rhenium-hafnium carbide alloy according to claim 1, wherein the step two of establishing the constitutive relation model comprises:
(1) a simplified Arrhennius equation is used as a constitutive relation model, namely:
Figure FDA0002984366300000021
wherein A, alpha, beta and n1、n2Q is the strain activation energy in kJ. mol for a material constant independent of the specific strain temperature-1R is a general gas constant of 8.31 J.mol-1T is the absolute temperature in K,
Figure FDA0002984366300000024
to strain rate, σ is stress; the first formula in equation (1), the power law, applies to α σ<0.8, the second equation, the exponential law, applies to σ>1.2, the third formula, namely the hyperbolic sine law formula is applicable to alpha sigma which is more than or equal to 0.8 and less than or equal to 1.2;
(2) the logarithm was taken simultaneously on both sides of equation (1) and the results are shown below:
Figure FDA0002984366300000022
Figure FDA0002984366300000023
Figure FDA0002984366300000031
according to the formulas (2) and (3)
Figure FDA0002984366300000032
And
Figure FDA0002984366300000033
the mean slope of two straight lines is n1And β, α is defined as n1And beta;
linear fitting according to formula (4) to obtain
Figure FDA0002984366300000034
And ln [ sinh (. alpha. sigma.)]-a fitted line plot of 1/T, the mean slopes of the two fitted lines being n2And k, and the slope k ═ ln [ sinh (α σ)]Q/n2R, obtaining a calculation formula of deformation activation energy Q:
Figure FDA0002984366300000035
(3) selecting peak stress as a stress value to calculate the equation, and selecting Z, namely a Zener-Hollomon parameter as a temperature compensation factor to express the synergistic effect of the strain rate and the deformation temperature, wherein Z is expressed as:
Figure FDA0002984366300000036
substituting a hyperbolic sine function of a hyperbolic sine law formula, which is a third formula in formula (1), into formula (6) to obtain formula (7), wherein n in formula (7) represents a pressure index:
Z＝A[sinh(ασ)]n (7)
taking logarithm of two sides of the formula (7) to obtain a formula (8):
lnZ＝lnA+nln[sinh(ασ)] (8)
according to formula (8) pair lnZ and ln [ sinh (. alpha. sigma.)]Linear fitting was performed to obtain the slope n-2.06709 and intercept lnA-1.62 × 10 of the fitted line15Therefore, equation (1) is simplified as:
Figure FDA0002984366300000037
(4) the rheological stress is rewritten into hyperbolic sine function formula (10) according to formula (9), and the numerical values of the parameters are substituted to obtain formula (11), which is the constitutive equation of the tungsten-rhenium-hafnium carbide alloy, as follows:
Figure FDA0002984366300000041
Figure FDA0002984366300000042
7. the method for researching high-temperature deformation behavior of the tungsten-rhenium-hafnium-carbide alloy according to claim 1, wherein the process for training the artificial neural network model in the second step is as follows:
(1) according to stress and strain data of a thermal compression test under the conditions of different strain temperatures and different strain rates, strain epsilon and strain rate
Figure FDA0002984366300000043
And normalizing the strain temperature T and the stress sigma, wherein the calculation formula of the normalized data X' is shown as the formula (12):
Figure FDA0002984366300000044
wherein X is the original data, X' is the normalized data corresponding to X, XminAnd XmaxThe minimum and maximum values of X, respectively;
(2) constructing a BP neural network model;
(3) the strain epsilon and the strain rate after the normalization treatment in the step (1) are
Figure FDA0002984366300000045
And taking the strain temperature T as an input layer, taking the stress sigma after normalization treatment as an output layer, and inputting the constructed BP neural network model for training to obtain the trained BP neural network model.
8. The method for researching the high-temperature deformation behavior of the tungsten-rhenium-hafnium carbide alloy as claimed in claim 1, wherein the sample before and after the hot compression test in the fourth step is sliced along the compression axis, and the sliced sample is 120#～800#Sequentially grinding silicon carbide sand paper, polishing alumina slurry with the granularity of 0.3-5 mu M, electrolytically corroding with 1M NaOH solution for 1-3 s, and observing and analyzing the microstructure by adopting an SEM method.
CN202110295972.8A 2021-03-19 2021-03-19 Research method for high-temperature deformation behavior of tungsten-rhenium-hafnium carbide alloy Pending CN113061767A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110295972.8A CN113061767A (en) 2021-03-19 2021-03-19 Research method for high-temperature deformation behavior of tungsten-rhenium-hafnium carbide alloy

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110295972.8A CN113061767A (en) 2021-03-19 2021-03-19 Research method for high-temperature deformation behavior of tungsten-rhenium-hafnium carbide alloy

Publications (1)

Publication Number Publication Date
CN113061767A true CN113061767A (en) 2021-07-02

Family

ID=76562465

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110295972.8A Pending CN113061767A (en) 2021-03-19 2021-03-19 Research method for high-temperature deformation behavior of tungsten-rhenium-hafnium carbide alloy

Country Status (1)

Country Link
CN (1) CN113061767A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN118398140A (en) * 2024-06-19 2024-07-26 宝鸡锐航精材科技有限公司 Intelligent analysis method and system for realizing high-temperature performance of titanium rod based on deep learning

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US4658362A (en) * 1984-12-24 1987-04-14 Mxdonnell Douglas Corporation Process modeling for superplastic forming of metal sheets
CN105095962A (en) * 2015-07-27 2015-11-25 中国汽车工程研究院股份有限公司 Method for predicting dynamic mechanical property of material based on BP artificial neural network
CN105300837A (en) * 2014-06-25 2016-02-03 北京有色金属研究总院 Method for measuring viscosity of high solid fraction metal semi-solid slurry
CN106971012A (en) * 2016-01-13 2017-07-21 太原科技大学 A kind of method for building up of magnesium alloy materials constitutive relation model
CN112133373A (en) * 2020-09-14 2020-12-25 北京理工大学 Titanium alloy constitutive relation prediction method based on machine learning

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US4658362A (en) * 1984-12-24 1987-04-14 Mxdonnell Douglas Corporation Process modeling for superplastic forming of metal sheets
CN105300837A (en) * 2014-06-25 2016-02-03 北京有色金属研究总院 Method for measuring viscosity of high solid fraction metal semi-solid slurry
CN105095962A (en) * 2015-07-27 2015-11-25 中国汽车工程研究院股份有限公司 Method for predicting dynamic mechanical property of material based on BP artificial neural network
CN106971012A (en) * 2016-01-13 2017-07-21 太原科技大学 A kind of method for building up of magnesium alloy materials constitutive relation model
CN112133373A (en) * 2020-09-14 2020-12-25 北京理工大学 Titanium alloy constitutive relation prediction method based on machine learning

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
GUOLIANG JI等: ""A comparative study on Arrhenius-type constitutive model and artificial neural network model to predict high-temperature deformation behaviour in Aermet100 steel"", 《MATERIALS SCIENCE AND ENGINEERING:A》 *
李延超等: "W-Re-HfC合金的界面反应及高温拉伸性能", 《中国钨业》 *

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN118398140A (en) * 2024-06-19 2024-07-26 宝鸡锐航精材科技有限公司 Intelligent analysis method and system for realizing high-temperature performance of titanium rod based on deep learning
CN118398140B (en) * 2024-06-19 2024-09-03 宝鸡锐航精材科技有限公司 Intelligent analysis method and system for realizing high-temperature performance of titanium rod based on deep learning

Similar Documents

Publication Publication Date Title
Mendiratta et al. Strength and ductile-phase toughening in the two-phase Nb/Nb 5 Si 3 alloys
Xu et al. Evaluating compressive property and hot deformation behavior of molybdenum alloy reinforced by nanoscale zirconia particles
CN106967900A (en) A kind of titanium-based metal glass particle reinforced aluminum matrix composites and preparation method thereof
CN108570572A (en) A kind of nano oxide dispersion strengthens Cu-Ti alloys and preparation method thereof
Hussain et al. Hardness and wear analysis of Cu/Al2O3 composite for application in EDM electrode
Yuan et al. Processing map and hot deformation behavior of Ta-particle reinforced TiAl composite
Rajamuthamilselvan et al. Processing map for hot working of SiCp/7075 Al composites
Zhang et al. The hyperbolic constitutive equations and modified dynamic material model of TiBw/Ti-6.5 Al-2.5 Zr-1Mo-1V-0.5 Si composites
CN113061767A (en) Research method for high-temperature deformation behavior of tungsten-rhenium-hafnium carbide alloy
Zhou et al. Characterization of hot workability for a near alpha titanium alloy by integrating processing maps and constitutive relationship
Wei et al. The “pinning” effect of in-situ TiB whiskers on the hot deformation behavior and dynamic recrystallization of (10 vol%) TiB/Ti–6Al–4V composites
Kumar et al. Study on influence of die angle in cold extrusion on properties of nano sic reinforced 6061 aluminum alloy
Xiao et al. Hot deformation and processing map of 15% SiC p/2009 Al composite
CN117286357A (en) TZM alloy for large-size isothermal forging die, and preparation process and application thereof
CN115138849B (en) Preparation method of binding phase-free hard alloy cutter material
Gupta Controlling factors in aluminum matrix composites fabrication
Shengpu et al. Constitutive Analysis to Predict High-Temperature Flow Stress of 25vol% B4Cp/2009Al Composite
Kumar et al. Optimization of powder metallurgy process parameters to enhance the mechanical properties of AZ91 magnesium alloy.
Joel et al. Effect of alloying elements on the mechanical behavior of microwave synthesized composites
Mohanavel et al. [Retracted] Synthesis and Workability Behavior of Cu‐X wt.% TiC (x= 0, 4, 8, and 12) Powder Metallurgy Composites
Abdellah et al. Hot Extrusion of Reinforced Aluminum Powder Compacts
Liu et al. Characterization of hot workability of 5 vol.% TiBw/TA15 composites of meshed structures based on hot processing maps and microstructure evolution
CN115747550B (en) TiC particle reinforced high-strength high-wear-resistance tungsten-based composite material and preparation method thereof
Subrahmanyam et al. Evaluation of the mechanical properties on aluminium alloy 2024 fly ash metal matrix composite
Bodukuri et al. Investigation of optimum compacting parameters for aluminium, SiCP and B4C composites

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
RJ01 Rejection of invention patent application after publication

Application publication date: 20210702

RJ01 Rejection of invention patent application after publication