WO2023032370A1 - 金属材料の引張圧縮反転負荷挙動予測方法 - Google Patents
金属材料の引張圧縮反転負荷挙動予測方法 Download PDFInfo
- Publication number
- WO2023032370A1 WO2023032370A1 PCT/JP2022/021879 JP2022021879W WO2023032370A1 WO 2023032370 A1 WO2023032370 A1 WO 2023032370A1 JP 2022021879 W JP2022021879 W JP 2022021879W WO 2023032370 A1 WO2023032370 A1 WO 2023032370A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- model
- behavior
- tension
- compression
- metal material
- Prior art date
Links
- 239000007769 metal material Substances 0.000 title claims abstract description 89
- 238000007906 compression Methods 0.000 title claims abstract description 75
- 238000000034 method Methods 0.000 title claims abstract description 32
- 239000000463 material Substances 0.000 claims abstract description 51
- 238000012360 testing method Methods 0.000 claims abstract description 26
- 230000006835 compression Effects 0.000 claims description 27
- 238000009864 tensile test Methods 0.000 claims description 26
- 229910000831 Steel Inorganic materials 0.000 claims description 21
- 239000010959 steel Substances 0.000 claims description 21
- 238000012669 compression test Methods 0.000 claims description 19
- 230000008859 change Effects 0.000 claims description 16
- 239000002436 steel type Substances 0.000 claims description 15
- 230000006399 behavior Effects 0.000 abstract description 98
- 238000010801 machine learning Methods 0.000 abstract description 26
- 238000013528 artificial neural network Methods 0.000 description 20
- 230000000052 comparative effect Effects 0.000 description 19
- 238000010586 diagram Methods 0.000 description 13
- 230000000694 effects Effects 0.000 description 13
- 125000004122 cyclic group Chemical group 0.000 description 8
- 230000007246 mechanism Effects 0.000 description 5
- 238000005482 strain hardening Methods 0.000 description 5
- 238000005728 strengthening Methods 0.000 description 5
- 229910000922 High-strength low-alloy steel Inorganic materials 0.000 description 4
- 230000006870 function Effects 0.000 description 3
- 230000008569 process Effects 0.000 description 3
- 238000010998 test method Methods 0.000 description 3
- 238000012549 training Methods 0.000 description 3
- 238000002474 experimental method Methods 0.000 description 2
- 230000006872 improvement Effects 0.000 description 2
- 229910000734 martensite Inorganic materials 0.000 description 2
- 230000002265 prevention Effects 0.000 description 2
- 239000013077 target material Substances 0.000 description 2
- PXFBZOLANLWPMH-UHFFFAOYSA-N 16-Epiaffinine Natural products C1C(C2=CC=CC=C2N2)=C2C(=O)CC2C(=CC)CN(C)C1C2CO PXFBZOLANLWPMH-UHFFFAOYSA-N 0.000 description 1
- 229910000885 Dual-phase steel Inorganic materials 0.000 description 1
- 229910001035 Soft ferrite Inorganic materials 0.000 description 1
- 230000004913 activation Effects 0.000 description 1
- 229910045601 alloy Inorganic materials 0.000 description 1
- 239000000956 alloy Substances 0.000 description 1
- 238000005275 alloying Methods 0.000 description 1
- 238000000137 annealing Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000005489 elastic deformation Effects 0.000 description 1
- 238000005098 hot rolling Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 239000002184 metal Substances 0.000 description 1
- 229910052751 metal Inorganic materials 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
Images
Classifications
-
- 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
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N3/00—Investigating strength properties of solid materials by application of mechanical stress
- G01N3/08—Investigating strength properties of solid materials by application of mechanical stress by applying steady tensile or compressive forces
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B21—MECHANICAL METAL-WORKING WITHOUT ESSENTIALLY REMOVING MATERIAL; PUNCHING METAL
- B21D—WORKING OR PROCESSING OF SHEET METAL OR METAL TUBES, RODS OR PROFILES WITHOUT ESSENTIALLY REMOVING MATERIAL; PUNCHING METAL
- B21D22/00—Shaping without cutting, by stamping, spinning, or deep-drawing
-
- 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
- G06F30/23—Design optimisation, verification or simulation using finite element methods [FEM] or finite difference methods [FDM]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N7/00—Computing arrangements based on specific mathematical models
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N2203/00—Investigating strength properties of solid materials by application of mechanical stress
- G01N2203/0014—Type of force applied
- G01N2203/0016—Tensile or compressive
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N2203/00—Investigating strength properties of solid materials by application of mechanical stress
- G01N2203/0058—Kind of property studied
- G01N2203/0069—Fatigue, creep, strain-stress relations or elastic constants
- G01N2203/0075—Strain-stress relations or elastic constants
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N2203/00—Investigating strength properties of solid materials by application of mechanical stress
- G01N2203/02—Details not specific for a particular testing method
- G01N2203/0202—Control of the test
- G01N2203/0212—Theories, calculations
- G01N2203/0216—Finite elements
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N2203/00—Investigating strength properties of solid materials by application of mechanical stress
- G01N2203/02—Details not specific for a particular testing method
- G01N2203/0202—Control of the test
- G01N2203/0212—Theories, calculations
- G01N2203/0218—Calculations based on experimental data
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2113/00—Details relating to the application field
- G06F2113/24—Sheet material
-
- 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
Definitions
- the present invention relates to a method for predicting the tension-compression reverse loading behavior of a metal sheet.
- FEM analysis finite element analysis
- FIG. Figure 16 shows the relationship between stress and strain (stress-strain curve (diagram)) to which metal materials are subjected.
- the magnitude of the stress in the direction negative indicates the magnitude of the stress in the compressive direction.
- the yield stress at the time of reloading is ) becomes almost the same as the stress before unloading.
- the initial load in the tension direction is unloaded and then the load direction is reversed and reloaded in the compression direction, as shown in FIG. It yields at a stress lower than the yield stress of the case (see the dashed circle in the figure). This phenomenon also appears in the case of compression-unloading-tension.
- the target metal material is subjected to a tension-compression applied stress reverse test (hereinafter simply referred to as "tension-compression test”) is performed to obtain a stress-strain curve of the metal material, and a method of identifying a model constant based on the stress-strain curve is common.
- tension-compression test that evaluates the Bauschinger effect as described above, buckling deformation may occur in the material during compression. Since buckling deformation is more likely to occur when the thickness of the steel becomes thinner, there is a problem that high-precision testing is difficult.
- Patent Documents 1 and 2 tensile compression test methods for thin plates have been studied, and examples of such test methods are disclosed in Patent Documents 1 and 2.
- the techniques disclosed in Patent Documents 1 and 2 measure the strain during a cyclic load test using an extensometer fixed to the side surface of a test piece with a specified shape with an elastic body.
- jigs that prevent buckling during compression buckling prevention jig and buckling prevention base jig are used to measure the strain during repeated load tests with high accuracy.
- the present invention has been made to solve the above problems, and its object is to obtain model constants of a material model that expresses the tension-compression reversal load behavior of a metal material without requiring a special test method or jig. It is an object of the present invention to provide a method for predicting the reverse tension-compression load behavior of a metal material, which is determined based on the results of a simple test and thereby predicts the reverse tension-compression load behavior of the metal material.
- FIG. 17 a cyclic stress-strain curve ( diagram)) is shown in FIG.
- positive indicates the magnitude of true stress in the tensile direction
- negative indicates the magnitude of true stress in the compressive direction.
- Fig. 17 after tensile deformation is applied to the target material to yield, the load in the tensile direction is removed (the true stress is reduced to 0), and the load in the compressive direction is reapplied.
- the above formula (1) expresses the relationship between the amount of change in stress after re-yielding ⁇ and the amount of change in plastic strain after re-yielding ⁇ p using three model constants Y, A, and B. is.
- the amount of change ⁇ p in plastic strain is obtained by subtracting the amount of change ⁇ e in elastic strain from the amount of change ⁇ in strain.
- Y is the re-yielding stress after reversing the loading direction
- A is the convergence stress corresponding to the stress change from the re-yielding stress after reversing the loading direction.
- B represents the degree of the Bauschinger effect.
- the regression accuracy of the above formula (1) will be described with reference to FIG. 18 .
- the solid line (experimental values) in FIG. 18 plots the relationship between the amount of change in stress ( ⁇ ) after re-yielding and the amount of change in plastic strain ( ⁇ p ) in the tensile compression test results of FIG. 17 .
- the dashed line (regression value) identifies the model constants Y, A, and B of formula (1) based on the experimental values in FIG. The relationship between ⁇ and ⁇ p is plotted.
- the solid line showing the experimental values and the dashed line showing the regression values of Equation (1) are almost in agreement, and the compression behavior model of Equation (1) is sufficient for the compression behavior after re-yielding. It can be seen that it has a good regression accuracy.
- the Bauschinger effect means that there is anisotropy in the work hardening behavior during pre-deformation, and the material strength after reversing the load direction is in the same direction. This is the phenomenon of stress acting in a direction (herein referred to as "reverse Therefore, the Bauschinger effect is a phenomenon closely related to work hardening.” History and Current Status of Research on Effects, Hachiko). From this, it can be concluded that the behavior of the metal material when it is work-hardened by uniaxial tension affects the behavior of the Bauschinger effect when the same metal material is subjected to tension-compression reverse loading. Presumed.
- Indices that indicate behavior in uniaxial tension include representative mechanical property values such as yield stress (YP), maximum tensile stress (TS), uniform elongation (u-EL), etc. There is a (mechanical property value). These mechanical property values can be obtained from the stress-strain curve obtained as the test result by subjecting the metal material to a uniaxial tensile test.
- the inventors used the model constants Y, A, and B in the compression behavior model of formula (1) described above, and the yield stress (YP), maximum tensile stress (TS), and uniform elongation (u-EL) described above. In order to confirm the relationship between , the following experiments were conducted.
- FIG. 19(a) shows the relationship between the model constant Y and the yield stress (YP) for each high-tensile material. As shown in FIG. 19(a), it was found that there is a strong correlation between Y, which is the re-yield stress after reversal of the load direction, and the yield stress (YP).
- FIG. 19(b) shows the relationship between the model constant A and the maximum tensile stress (TS) for each high-tensile material. As shown in FIG. 19(b), it was found that there is a strong correlation between A, which is the convergent stress after reversing the load direction, and the maximum tensile stress (TS).
- FIG. 19(c) shows the relationship between the model constant B and the uniform elongation (u-EL) in each high-tensile steel. As shown in FIG. 19(c), a certain correlation, although not strong, was observed between B, which defines the degree of the Bauschinger effect, and uniform elongation (u-EL).
- the model constants of the material model that expresses the Bauschinger effect are YS, TS, u- We thought that it could be formulated by a multiple regression model, etc., depending on the relationship with typical mechanical property values such as EL.
- high-tensile steel such as a dual-phase structure in which a soft ferrite phase and a hard martensite phase coexist ) structure-hardened high-tensile steel (DP steel (dual-phase steel)) and precipitation-hardened high-strength low-strength steel with the addition of minute amounts of alloying elements such as Nb, V, and Ti. Due to the difference in microstructure between high strength low alloy steels (HSLA steel) and the same maximum tensile stress (TS grade), material deformation (work hardening) in uniaxial tension behave differently.
- HSLA steel high strength low alloy steel
- TS grade maximum tensile stress
- the steel type has the same maximum tensile stress and the same strengthening mechanism, it can be processed due to changes in the metallographic structure caused by variations in manufacturing conditions (hot-rolling conditions, annealing conditions, etc.). Since the hardening behavior also changes, the multiple regression model with only representative mechanical property values as explanatory variables has a problem in terms of estimation accuracy.
- the inventors provide information on these representative values and the behavior of work hardening closely related to the Bauschinger effect instead of the representative mechanical property values obtained from the uniaxial tensile stress-strain curve of the uniaxial tensile test.
- Uniaxial tensile stress-strain curves of uniaxial tensile tests containing information can be discretized and point sequence data (true strain, true stress) can be used as explanatory variables to solve the above problems.
- point sequence data true strain, true stress
- the model constants could be derived with good accuracy by doing so.
- I came up with the idea of using a neural network (NN) that has no restrictions on the selection of input variables.
- the present invention has been made under the circumstances described above, and specifically has the following configuration.
- a method for predicting tension-compression reverse load behavior of a metallic material is a method for predicting tension-compression reverse load behavior by determining model constants of a material model expressing the tension-compression reverse load behavior of a metal material, Machine learning with metal materials test data including factors related to the uniaxial tensile behavior of metal materials as input variables and the values of the model constants determined based on tension-compression tests of the metal materials as output variables.
- the factors associated with the uniaxial tensile behavior include point sequence data discretized from uniaxial tensile stress-strain curves obtained from uniaxial tensile tests.
- the factors related to the uniaxial tensile behavior may further include a set of mechanical property values including yield stress, maximum tensile stress, and uniform elongation.
- the metal material test data should further include steel grade information.
- the material model may be a steel compression behavior model represented by the following formula (1), and the model constants may be Y, A, and B in the following formula (1).
- ⁇ is the change in stress after re-yielding
- ⁇ p is the change in plastic strain after re-yielding
- the metal material test data including factors related to the uniaxial tensile behavior of the metal material is used as an input variable, and the value of the model constant determined based on the tension-compression test of the metal material is used as an output variable for machine learning.
- the model constant of the metallic material is obtained, and the factors related to uniaxial tensile behavior include uniaxial
- point sequence data discretized uniaxial tensile stress-strain curves obtained from tensile tests it expresses the tension-compression reversal loading behavior of metal materials based on the results of simple uniaxial tensile tests.
- the values of the model constants of the material model can be determined. Then, by substituting the determined values into the material model, it is possible to predict the tension-compression reversal load behavior of the metal material without performing a tension-compression test that requires a special testing machine.
- FIG. 1 is an explanatory diagram of a method for predicting the tension-compression reversal load behavior of a metal material according to Embodiment 1 of the present invention.
- FIG. 2 is a flow diagram illustrating the steps for implementing the (i) learning phase of FIG.
- FIG. 3 is a diagram showing a uniaxial tensile stress-strain curve obtained from a uniaxial tensile test and point sequence data obtained by discretizing the uniaxial tensile stress-strain curve.
- FIG. 4 is a schematic diagram showing the configuration of the trained model (neural network) of FIG.
- FIG. 5 is a flow diagram illustrating the steps for performing the (ii) recognition phase of FIG. FIG.
- FIG. 6 is an explanatory diagram of a method for predicting the tension-compression reversal load behavior of a metal material according to Embodiment 2 of the present invention.
- FIG. 7 is a schematic diagram showing the configuration of the trained model (neural network) of FIG.
- FIG. 8 is an explanatory diagram of a method for predicting tension-compression reversal load behavior of a metal material according to Embodiment 3 of the present invention.
- FIG. 9 is a schematic diagram showing the configuration of the trained model (neural network) of FIG.
- FIG. 10 is a graph for explaining the learning accuracy of the trained model of the invention example in the embodiment.
- FIG. 11 is a graph explaining prediction accuracy of a comparative example in the example.
- FIG. 12 is a graph explaining the prediction accuracy of invention example 1 in the working example.
- FIG. 13 is a graph explaining the prediction accuracy of invention example 2 in the working example.
- FIG. 14 is a graph for explaining the prediction accuracy of Invention Example 3 in the example.
- FIG. 15 is a graph comparing the prediction accuracy of Comparative Example and Invention Examples 1 to 3 in Examples.
- FIG. 16 is an explanatory diagram of the Bauschinger effect.
- FIG. 17 is an example of a cyclic stress-strain curve obtained by a tension-compression test.
- FIG. 18 is a graph for explaining the regression accuracy when the compression behavior after re-yielding in FIG. 17 is regressed using the compression behavior model of formula (1).
- FIG. 19 is a graph showing the correlation between the model constants Y, A, and B of the compression behavior model of formula (1) and the mechanical property values YP, TS, and u-EL obtained from the uniaxial tensile test.
- a method for predicting tension-compression reverse load behavior of a metallic material is a method of predicting tension-compression reverse load behavior by determining model constants of a material model that expresses the tension-compression reverse load behavior of a metal material. is. As shown in FIG. 1, this embodiment includes a learning phase in which a neural network performs machine learning, and a recognition phase in which the neural network learned in the learning phase is used to predict the tension-compression reversal load behavior of a metal material.
- a learning phase in which a neural network performs machine learning
- a recognition phase in which the neural network learned in the learning phase is used to predict the tension-compression reversal load behavior of a metal material.
- the learning phase includes a learning data creation step for creating learning data used for machine learning of the neural network, and machine learning of the neural network using the learning data created in the learning data creation step. It includes a machine learning step that builds a trained model by allowing
- the learning data creation step is a step of creating input data and ground truth data (also called teacher data (labeled training data)) used for machine learning of the neural network.
- input data are created based on the results of uniaxial tensile tests of various metal materials, and correct data are created based on the results of tension-compression tests of the same metal materials. Specifically, it is as follows.
- a uniaxial tensile test is performed on multiple metal materials, and a uniaxial tensile stress-strain curve (see Fig. 3) that shows the uniaxial tensile behavior of the target material is obtained.
- the obtained uniaxial tensile stress-strain curve shall be in a range that indicates uniform elongation until localized necking occurs in the test piece in the uniaxial tensile test.
- the discretization process is a process of discretizing the uniaxial tensile stress-strain curve into predetermined sections N and converting it into (N+1) point sequence data. This point will be specifically described with reference to FIG.
- FIG. 3 is an example of a uniaxial tensile stress-strain curve discretized into 100 segments. From the discretized uniaxial tensile stress-strain curve, the boundary points and the coordinates of the start and end points (true strain ⁇ , true stress ⁇ ) of each section are acquired, and 101 coordinate data ( ⁇ 1 , ⁇ 1 ) ⁇ ( ⁇ 101 , ⁇ 101 ) is generated. These 101 pieces of coordinate data (sequence of points) are used as input data for neural network machine learning. Note that ( ⁇ 1 , ⁇ 1 ), which is the starting point coordinates of the point sequence data, is input 1 of the input variable, and coordinate data ( ⁇ 2 , ⁇ 2 ) adjacent to ( ⁇ 1 , ⁇ 1 ) is input 2. Below, it corresponds to input 1 to input 101 of the input variables in order up to the end point ( ⁇ 101 , ⁇ 101 ).
- the model constants of the material model that expresses the tensile-compressive behavior are identified.
- the compression behavior model of Formula (1) mentioned above shall be used as a material model.
- the cyclic stress-strain curve of each metallic material is regressed using a compression behavior model, and the values of the model constants Y, A, and B identified as a result are used as the correct data for machine learning of the neural network.
- the machine learning step is a step of performing machine learning on the neural network using the input data and the correct answer data created in the learning data creating step to construct a trained model.
- a learned model constructed in the machine learning step will be described with reference to FIG.
- FIG. 4 is a schematic diagram showing the configuration of a trained model.
- a trained model comprises an input layer, one or more intermediate layers (hidden layers), and an output layer, as shown in FIG.
- Each layer is an Affine layer, and the output of the intermediate layer has a suitable activation function, such as a Sigmoid function, a ReLu function, or the like.
- input data sequence of points ( ⁇ 1 , ⁇ 1 ) to ( ⁇ 101 , ⁇ 101 )
- correct data model constants Y, A, B
- the recognition phase is created by an input data creation step for creating input data to be input to the trained model based on the results of a uniaxial tensile test on the metal material to be predicted, and an input data creation step.
- a tension-compression reversal load behavior prediction step of predicting the tension-compression reversal load behavior of the metal material based on the input data is included.
- the input data creation step is a step of creating input data to be input to the learned model based on the results of the uniaxial tensile test for the metallic material to be predicted.
- the creation of input data in this step is also performed in the same manner as the creation of input data used for machine learning in the learning data creation step of the learning phase. Specifically, a uniaxial tensile test is performed on the metal material to be predicted, and a uniaxial tensile stress-strain curve indicating uniaxial tensile behavior is obtained. Then, the resulting uniaxial tensile stress-strain curve is subjected to discretization processing to generate point sequence data ( ⁇ 1 , ⁇ 1 ) to ( ⁇ 101 , ⁇ 101 ) (see FIG. 3).
- the tension-compression reversal load behavior prediction step inputs the input data created in the input data creation step to the learned model, acquires the predicted value of the model constant output by the learned model based on the input, and uses the predicted value This is the step of predicting the tension-compression reversal load behavior by substituting it into the compression behavior model.
- the trained model As an output of , obtain the predicted value Y e of the model constant Y, the predicted value A e of the model constant A, and the predicted value B e of the model constant B.
- the obtained predicted values Y e , A e , and B e are substituted into the model constants Y, A, and B of the compression behavior model in the aforementioned formula (1) (see formula (2) below).
- the compressive behavior after re-yielding of the metal material to be predicted can be predicted from the following formula (2).
- ⁇ is the change in stress after re-yielding
- ⁇ p is the change in plastic strain after re-yielding
- the model constants of the material model (in this example, the compression behavior model after re-yielding) using the learned model based on the results of the uniaxial tensile test for the metal material to be predicted
- the tension-compression reversal load behavior in this example, the compression behavior after re-yielding
- the relational expression of the above formula (2) is set in the FEM analysis, it is possible to perform the press forming FEM analysis considering the tension-compression reversal load behavior in the metal material.
- the present invention uses a neural network to machine-learn the relationship between factors related to uniaxial tensile behavior and factors related to tensile-compressive behavior (model constants of a material model) in a metallic material, so that the metallic material to be predicted It is possible to predict the tension-compression reversal load behavior from the uniaxial tension behavior of .
- point sequence data obtained by discretizing a uniaxial tensile stress-strain curve is used as a factor related to uniaxial tensile behavior.
- the prediction accuracy of the tension-compression reversal load behavior is further improved by using the mechanical property value of the metal material to be predicted as a factor related to the uniaxial tension behavior.
- the method for predicting the tension-compression reversal load behavior of a metal material according to the second embodiment includes a learning phase and a recognition phase similar to those of the first embodiment, as shown in FIG.
- the difference from Embodiment 1 is that the input data in the learning phase and the recognition phase include a set of mechanical property values, which are factors related to the uniaxial tensile behavior of metallic materials. This point will be specifically described below.
- point sequence data is created by discretizing the uniaxial tensile stress-strain curve in the same way as in the first embodiment (see FIG. 3).
- a mechanical property value is also obtained from the uniaxial tensile stress-strain curve.
- the mechanical property values are the aforementioned yield stress (YP), maximum tensile stress (TS), and uniform elongation (u-EL).
- the uniaxial tensile stress-strain curve obtained from the uniaxial tensile test is discretized point sequence data and a set of mechanical property values (YP, TS, u-EL) as input data, and the tensile compression test Using the model constants Y, A, and B identified based on the cyclic stress-strain curve obtained from the model as the correct data, machine learning of the neural network is performed.
- FIG. 7 shows an example of a trained model constructed by the above machine learning.
- the mechanical property values are obtained.
- point sequence data input 1 to input 101
- YP TS, u-EL
- Predicted values Y e , A e , and B e of the model constants Y, A, and B can be obtained by inputting to the trained model of 7.
- the mechanical property values of metal materials are also included as factors related to uniaxial tensile behavior, so further improvement in prediction accuracy of tension-compression reversal load behavior can be expected.
- the input data in the learning phase and the recognition phase is the uniaxial tensile stress-strain curve. can. Further, as described in Embodiment 2, by adding the mechanical property value, which is a factor related to uniaxial tensile behavior, to the input data, an improvement in prediction accuracy can be expected. Furthermore, the input data may include test data (for example, steel type information) related to uniaxial tensile tests in addition to factors related to uniaxial tensile behavior. Examples of such are described below.
- the method for predicting the tension-compression reversal load behavior of a metal material according to Embodiment 3 includes a learning phase and a recognition phase similar to Embodiments 1 and 2.
- the difference from Embodiments 1 and 2 is that the input data in the learning phase and the recognition phase includes the steel type information of the metal material. This point will be specifically described below.
- the uniaxial tensile stress described in the first embodiment - point sequence data obtained by discretizing the strain curve, described in the second embodiment
- steel type information is used.
- the steel type information is information indicating the steel type of the metal material used in the uniaxial tensile test.
- a character string combining a TS grade (590 to 1470 MPa grade) representing the maximum tensile stress and an abbreviation (DP, HSLA, FM, etc.) representing the strengthening mechanism can be used.
- the character string "980DP” is used for DP steel of 980 MPa class
- the character string "590HSLA” is used for HSLA steel of 590 MPa class as steel type information.
- FIG. 9 shows an example of a trained model constructed by performing machine learning of a neural network using the values of the model constants Y, A, and B identified based on the stress-strain curve as correct data.
- the steel type information is included in the input data in addition to the factors related to the uniaxial tension behavior, so that the prediction accuracy of the tension-compression reversal load behavior can be expected to be further improved.
- the input data includes a set of mechanical property values (YP, TS, u-EL)
- only the point sequence data and the steel type information may be used as the input data.
- the compression behavior model of formula (1) was used as the material model expressing the tension-compression reversal load behavior of the metal material, but in the present invention, the material model is limited. It goes without saying that a material model other than the above can be used instead, and the model constants in the material model may be determined.
- Non-Patent Document 1 An example of a material model is the conventionally known kinematic hardening model. More specifically, the Y-U model described in Non-Patent Document 1, which is an example of the kinematic curing model, can be mentioned. In this case, Y, C, B, Rsat, b, m, and h in the document Seven parameters can be treated as model constants of the present invention.
- a material model is the elastoplastic constitutive equation used in the "stress-strain relationship simulation method" disclosed in Patent No. 5582211, which evolved the Y-U model.
- the parameters of the formula specifically the 10 parameters of Y, Rsat, b, m, h, Cc, A1, A2, n1, and n2, can be treated as model constants of the present invention.
- High-strength materials for learning include those with TS grades of 590 to 1470 MPa, and those with strengthening mechanisms such as DP steel, HSLA steel and FM (ferrite-martensite) steel. ing. Then, three kinds of prediction high-tensile steels of 980 MPa grade, 1180 MPa grade and 1470 MPa grade, which are not included in the above 29 samples, were prepared as targets for predicting the reverse tension-compression load behavior.
- a uniaxial tensile test was performed on each high tensile strength material for learning described above, and point sequence data was created by discretizing the obtained uniaxial tensile stress-strain curve (see Embodiment 1).
- the yield stress (YP), maximum tensile stress (TS), and uniform elongation (u-EL), which are mechanical property values obtained from a uniaxial tensile test were determined for each high-tensile steel for learning (embodiment 2).
- steel type information indicating the steel type of each learning high-tensile material was created (see Embodiment 3).
- FIG. 10A is a graph showing the learning accuracy (regression accuracy in the comparative example) with respect to the model constant Y.
- FIG. 10A As a value indicating the learning accuracy, the prediction error ( prediction accuracy ) and the standard deviation score ( ⁇ ). The prediction error was calculated by dividing the difference between the Y value of the correct data and the predicted value Y e by the Y value of the correct data and multiplying the result by 100 to calculate the percentage value.
- FIG. 10(b) is a graph showing the learning accuracy (regression accuracy in the comparative example) for the model constant A
- FIG. 10(c) is a graph showing the learning accuracy (regression accuracy in the comparative example) for the model constant B.
- the average value (ave) and the standard deviation ( ⁇ ) of the error ratios in Invention Examples 1 to 3 are both smaller than in Comparative Example. This indicates that the learning accuracy for the model constant Y in the trained models of Invention Examples 1 to 3 is slightly higher than the regression accuracy of the Comparative Example. Further, as shown in FIG. 10(b), the learning accuracy of the model constant A in Invention Examples 1 to 3 was not significantly superior to the regression accuracy in Comparative Example. This is because, as mentioned above, the model constants Y and A have a strong correlation with the machine property values YP and TS, and the comparative example using the multiple regression model based on the machine property values also had sufficient regression accuracy. Conceivable.
- FIG. 11 shows the results of predicting the compressive behavior after re-yielding of the high-tensile steel for prediction using a comparative example.
- the "prediction" curve in FIG. 11 is the predicted values Y e , A e , B It is obtained by substituting e into the formula (1).
- the "actual" curve is identified using the formula (1) based on the repeated stress-strain curve obtained by conducting a tension-compression test on the prediction high-tensile material. The closer the "prediction" curve and the "actual” curve are, the higher the prediction accuracy is.
- FIGS. 12 to 14 the results of prediction using Invention Examples 1 to 3 are shown in FIGS. 12 to 14.
- the average value (ave) of the prediction error is 50.2% on average in Invention Examples 1 to 3 compared to Comparative Examples (Invention Example 1 is 31.2%, Invention Example 2 is 30.5%, Invention Example 3 improved by 88.9%).
- the standard deviation ( ⁇ ) of the prediction error was improved by an average of 45.7% for Invention Examples 1 to 3 (42.9% for Invention Example 1, 42.9% for Invention Example 2, and 51.2% for Invention Example 3). From the above, the effectiveness of the present invention was verified.
- the model constants of a material model that expresses the tension-compression reversal load behavior of a metal material are determined based on the results of a simple test that does not require a special test method or jig, and thereby the metal material It is possible to provide a method for predicting the tension-compression reversal load behavior of a metal material that can predict the tension-compression reversal load behavior of a metal material.
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Software Systems (AREA)
- Artificial Intelligence (AREA)
- Data Mining & Analysis (AREA)
- Mathematical Physics (AREA)
- Computing Systems (AREA)
- Computer Hardware Design (AREA)
- Geometry (AREA)
- Biochemistry (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Life Sciences & Earth Sciences (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Medical Informatics (AREA)
- General Health & Medical Sciences (AREA)
- Mechanical Engineering (AREA)
- Health & Medical Sciences (AREA)
- Algebra (AREA)
- Computational Mathematics (AREA)
- Mathematical Analysis (AREA)
- Mathematical Optimization (AREA)
- Pure & Applied Mathematics (AREA)
- Investigating Strength Of Materials By Application Of Mechanical Stress (AREA)
- Shaping Metal By Deep-Drawing, Or The Like (AREA)
Abstract
Description
本発明の実施の形態に係る金属材料の引張圧縮反転負荷挙動予測方法は、金属材料の引張圧縮反転負荷挙動を表現する材料モデルのモデル定数を決定することで引張圧縮反転負荷挙動を予測する方法である。本実施の形態は、図1に示すように、ニューラルネットワークに機械学習させる学習フェーズと、学習フェーズで学習したニューラルネットワークを用いて金属材料の引張圧縮反転負荷挙動を予測する認識フェーズとを含む。以下、本発明の実施の形態1の各フェーズについて具体的に説明する。
学習フェーズは、図2に示すように、ニューラルネットワークの機械学習に用いる学習用データを作成する学習用データ作成ステップと、学習用データ作成ステップで作成した学習用データを用いてニューラルネットワークを機械学習させることにより、学習済モデルを構築する機械学習ステップを含む。
学習用データ作成ステップは、ニューラルネットワークの機械学習に用いる入力データと正解データ(ground truth data)(教師データ(labeled training data)ともいう)を作成するステップである。学習用データ作成ステップにおいては、種々の金属材料の単軸引張試験の結果に基づいて入力データを作成し、同金属材料の引張圧縮試験の結果に基づいて正解データを作成する。具体的には下記のとおりである。
機械学習ステップは、学習用データ作成ステップで作成した入力データと正解データを用いてニューラルネットワークを機械学習させ、学習済モデルを構築するステップである。機械学習ステップで構築される学習済モデルについて、図4を用いて説明する。
認識フェーズは、図5に示すように、予測対象の金属材料に対する単軸引張試験の結果に基づいて学習済モデルに入力する入力データを作成する入力データ作成ステップと、入力データ作成ステップで作成した入力データに基づいて金属材料の引張圧縮反転負荷挙動を予測する引張圧縮反転負荷挙動予測ステップを含む。
入力データ作成ステップは、予測対象の金属材料に対する単軸引張試験の結果に基づいて、学習済モデルに入力する入力データを作成するステップである。本ステップにおける入力データの作成も、学習フェーズの学習用データ作成ステップにおいて機械学習に用いる入力データの作成と同様に行う。具体的には、予測対象の金属材料に対して単軸引張試験を実施し、単軸引張挙動を示す単軸引張応力-ひずみ曲線を取得する。そして、得られた単軸引張応力-ひずみ曲線に対して離散化処理を実施して点列データ(ε1,σ1)~(ε101,σ101)を生成する(図3参照)。
引張圧縮反転負荷挙動予測ステップは、入力データ作成ステップで作成した入力データを学習済モデルに入力し、該入力に基づいて学習済モデルが出力するモデル定数の予測値を取得し、該予測値を圧縮挙動モデルに代入して引張圧縮反転負荷挙動を予測するステップである。
本発明は、金属材料における単軸引張挙動に関連する因子と、引張圧縮挙動に関連する因子(材料モデルのモデル定数)との関係性をニューラルネットワークに機械学習させることで、予測対象の金属材料の単軸引張挙動から引張圧縮反転負荷挙動を予測できるようにしたものである。そして、上述した実施の形態1は、単軸引張挙動に関連する因子として、単軸引張応力-ひずみ曲線を離散化した点列データを用いた。本実施の形態2では、単軸引張挙動に関連する因子として、さらに予測対象の金属材料の機械特性値も用いることで、引張圧縮反転負荷挙動の予測精度をより高める例を説明する。
実施の形態1で説明したように、学習フェーズ及び認識フェーズにおける入力データに単軸引張応力-ひずみ曲線を離散化した点列データを用いることで、引張圧縮反転負荷挙動を精度よく予測することができる。また、実施の形態2で説明したように、単軸引張挙動に関連する因子である機械特性値を上記入力データに加えることで、予測精度の向上が期待できる。さらに、入力データには、単軸引張挙動に関連する因子の他、単軸引張試験に関する試験データ(例えば鋼種情報など)を含めるようにしてもよい。このような例について、以下に説明する。
各学習用ハイテン材における点列データを入力データ、同定したモデル定数Y、A、Bの値を正解データとしてニューラルネットワークに機械学習させ、図4に示したような学習済モデルを構築した。そして、該学習済モデルを用い、実施の形態1の方法で予測用ハイテン材のモデル定数の予測値Ye、Ae、Beをそれぞれ取得した。
各学習用ハイテン材における点列データと機械特性値を入力データ、同定したモデル定数Y、A、Bの値を正解データとしてニューラルネットワークに機械学習させ、図7に示したような学習済モデルを構築した。そして、該学習済モデルを用い、実施の形態2の方法で予測用ハイテン材のモデル定数の予測値Ye、Ae、Beをそれぞれ取得した。
各学習用ハイテン材における点列データと機械特性値、及び鋼種情報を入力データ、同定したモデル定数Y、A、Bの値を正解データとしてニューラルネットワークに機械学習させ、図9に示したような学習済モデルを構築した。そして、該学習済モデルを用い、実施の形態3の方法で予測用ハイテン材のモデル定数の予測値Ye、Ae、Beをそれぞれ取得した。
各学習用ハイテン材における単軸引張試験から得られる機械特性値(YP、TS、u-EL)及び同定したモデル定数Y、A、Bの値に基づき、機械特性値(YP、TS、u-EL)を説明変数、モデル定数Y、A、Bをそれぞれ目的変数とする重回帰モデルを作成した。そして、該重回帰モデルを用い、予測用ハイテン材のモデル定数の予測値Ye、Ae、Beをそれぞれ求めた。
Claims (4)
- 金属材料の引張圧縮反転負荷挙動を表現する材料モデルのモデル定数を決定することで引張圧縮反転負荷挙動を予測する金属材料の引張圧縮反転負荷挙動予測方法であって、
金属材料の単軸引張挙動に関連する因子を含む金属材料試験データを入力変数とし、該金属材料の引張圧縮試験に基づいて決定した前記モデル定数の値を出力変数として機械学習させた学習済モデルに対して、
予測対象の金属材料の前記金属材料試験データを入力することによって、当該金属材料の前記モデル定数の値を取得するステップを含み、
前記単軸引張挙動に関連する因子には、単軸引張試験から得られた単軸引張応力-ひずみ曲線を離散化した点列データが含まれている、金属材料の引張圧縮反転負荷挙動予測方法。 - 前記単軸引張挙動に関連する因子には、さらに降伏応力、最大引張応力、一様伸びを含む機械特性値の集合が含まれている、請求項1に記載の金属材料の引張圧縮反転負荷挙動予測方法。
- 前記金属材料試験データには、さらに鋼種情報が含まれる、請求項1又は2に記載の金属材料の引張圧縮反転負荷挙動予測方法。
Priority Applications (3)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
EP22863958.9A EP4378604A1 (en) | 2021-09-03 | 2022-05-30 | Method for predicting tension-compression reverse loading behavior of metal material |
CN202280059666.7A CN117940755A (zh) | 2021-09-03 | 2022-05-30 | 金属材料的拉伸压缩反转负荷举动预测方法 |
KR1020247010508A KR20240056554A (ko) | 2021-09-03 | 2022-05-30 | 금속 재료의 인장 압축 반전 부하 거동 예측 방법 |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2021143634A JP7211461B1 (ja) | 2021-09-03 | 2021-09-03 | 金属材料の引張圧縮反転負荷挙動予測方法 |
JP2021-143634 | 2021-09-03 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2023032370A1 true WO2023032370A1 (ja) | 2023-03-09 |
Family
ID=84982670
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/JP2022/021879 WO2023032370A1 (ja) | 2021-09-03 | 2022-05-30 | 金属材料の引張圧縮反転負荷挙動予測方法 |
Country Status (5)
Country | Link |
---|---|
EP (1) | EP4378604A1 (ja) |
JP (1) | JP7211461B1 (ja) |
KR (1) | KR20240056554A (ja) |
CN (1) | CN117940755A (ja) |
WO (1) | WO2023032370A1 (ja) |
Citations (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2000275154A (ja) * | 1999-03-25 | 2000-10-06 | Toyota Motor Corp | 応力−ひずみ関係シミュレート方法 |
JP2005315703A (ja) * | 2004-04-28 | 2005-11-10 | Nippon Steel Corp | 鋼材の材質予測方法 |
JP2013054001A (ja) * | 2011-09-06 | 2013-03-21 | Jfe Steel Corp | 応力−歪み関係評価方法およびスプリングバック量予測方法 |
JP5582211B1 (ja) | 2013-03-14 | 2014-09-03 | Jfeスチール株式会社 | 応力−ひずみ関係シミュレート方法、スプリングバック量予測方法およびスプリングバック解析装置 |
WO2015170742A1 (ja) * | 2014-05-08 | 2015-11-12 | 新日鐵住金株式会社 | 可塑性材料の評価方法及び可塑性材料の塑性加工の評価方法 |
JP5991278B2 (ja) | 2013-07-12 | 2016-09-14 | Jfeスチール株式会社 | 引張圧縮試験方法および装置 |
JP5991055B2 (ja) | 2012-07-19 | 2016-09-14 | Jfeスチール株式会社 | 引張圧縮試験方法および装置 |
JP6015997B1 (ja) | 2014-11-25 | 2016-10-26 | Jfeスチール株式会社 | 鋼管の製造方法及びその方法に使用するプレス金型 |
CN111339703A (zh) * | 2020-02-27 | 2020-06-26 | 中国科学院金属研究所 | 大应变条件下材料硬化行为的虚拟预测方法 |
Family Cites Families (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPS5940387Y2 (ja) | 1982-12-08 | 1984-11-15 | 壷井 宏純 | 吊持搬送用ベルト |
JPS5991055U (ja) | 1982-12-10 | 1984-06-20 | トヨタ自動車株式会社 | 車両用アンテナアンプの電源供給回路 |
JPS6015997U (ja) | 1983-07-12 | 1985-02-02 | 三菱電機株式会社 | 送風機のケ−シング |
-
2021
- 2021-09-03 JP JP2021143634A patent/JP7211461B1/ja active Active
-
2022
- 2022-05-30 KR KR1020247010508A patent/KR20240056554A/ko active Search and Examination
- 2022-05-30 WO PCT/JP2022/021879 patent/WO2023032370A1/ja active Application Filing
- 2022-05-30 EP EP22863958.9A patent/EP4378604A1/en active Pending
- 2022-05-30 CN CN202280059666.7A patent/CN117940755A/zh active Pending
Patent Citations (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2000275154A (ja) * | 1999-03-25 | 2000-10-06 | Toyota Motor Corp | 応力−ひずみ関係シミュレート方法 |
JP2005315703A (ja) * | 2004-04-28 | 2005-11-10 | Nippon Steel Corp | 鋼材の材質予測方法 |
JP2013054001A (ja) * | 2011-09-06 | 2013-03-21 | Jfe Steel Corp | 応力−歪み関係評価方法およびスプリングバック量予測方法 |
JP5991055B2 (ja) | 2012-07-19 | 2016-09-14 | Jfeスチール株式会社 | 引張圧縮試験方法および装置 |
JP5582211B1 (ja) | 2013-03-14 | 2014-09-03 | Jfeスチール株式会社 | 応力−ひずみ関係シミュレート方法、スプリングバック量予測方法およびスプリングバック解析装置 |
WO2014141794A1 (ja) * | 2013-03-14 | 2014-09-18 | Jfeスチール株式会社 | 応力-ひずみ関係シミュレート方法、スプリングバック量予測方法およびスプリングバック解析装置 |
JP5991278B2 (ja) | 2013-07-12 | 2016-09-14 | Jfeスチール株式会社 | 引張圧縮試験方法および装置 |
WO2015170742A1 (ja) * | 2014-05-08 | 2015-11-12 | 新日鐵住金株式会社 | 可塑性材料の評価方法及び可塑性材料の塑性加工の評価方法 |
JP6015997B1 (ja) | 2014-11-25 | 2016-10-26 | Jfeスチール株式会社 | 鋼管の製造方法及びその方法に使用するプレス金型 |
CN111339703A (zh) * | 2020-02-27 | 2020-06-26 | 中国科学院金属研究所 | 大应变条件下材料硬化行为的虚拟预测方法 |
Non-Patent Citations (4)
Title |
---|
F. YOSHIDAT. UEMOR, INT. J. PLASTICITY, vol. 18, 2002, pages 661 - 686 |
KUWABARA TOSHIHIKO, YOSHIDA KENGO: "Plastic deformation characteristics of light metal sheets—methods of measurement and modeling—", JOURNAL OF JAPAN INSTITUTE OF LIGHT METALS, KEIKINZOKU GAKKAI, JP, vol. 65, no. 5, 1 January 2015 (2015-01-01), JP , pages 164 - 173, XP093042270, ISSN: 0451-5994, DOI: 10.2464/jilm.65.164 * |
LIN JIANPING; HOU YONG; MIN JUNYING; TANG HUIJUN; CARSLEY JOHN E.; STOUGHTON THOMAS B.: "Effect of constitutive model on springback prediction of MP980 and AA6022-T4", INTERNATIONAL JOURNAL OF MATERIAL FORMING, SPRINGER PARIS, PARIS, vol. 13, no. 1, 14 January 2019 (2019-01-14), Paris , pages 1 - 13, XP036986711, ISSN: 1960-6206, DOI: 10.1007/s12289-018-01468-x * |
THE JOURNAL OF THE IRON AND STEEL INSTITUTE OF JAPAN, vol. 70, no. 11, 1984, pages 1551 - 1558 |
Also Published As
Publication number | Publication date |
---|---|
EP4378604A1 (en) | 2024-06-05 |
KR20240056554A (ko) | 2024-04-30 |
CN117940755A (zh) | 2024-04-26 |
JP2023037100A (ja) | 2023-03-15 |
JP7211461B1 (ja) | 2023-01-24 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Chongthairungruang et al. | Experimental and numerical investigation of springback effect for advanced high strength dual phase steel | |
Joun et al. | A new method for acquiring true stress–strain curves over a large range of strains using a tensile test and finite element method | |
Chongthairungruang et al. | Springback prediction in sheet metal forming of high strength steels | |
Akiniwa et al. | Notch effect on fatigue strength reduction of bearing steel in the very high cycle regime | |
Paul | Non-linear correlation between uniaxial tensile properties and shear-edge hole expansion ratio | |
Casagrande et al. | Relationship between fatigue limit and Vickers hardness in steels | |
Eller et al. | Determination of strain hardening parameters of tailor hardened boron steel up to high strains using inverse FEM optimization and strain field matching | |
Julsri et al. | Study of springback effect of AHS steels using a microstructure based modeling | |
Yilmaz et al. | The prediction of mechanical behavior for steel wires and cord materials using neural networks | |
Li et al. | Stress relaxation in tensile deformation of 304 stainless steel | |
Shikalgar et al. | Analysis of p-SPT specimens using Gurson parameters ascertained by Artificial Neural Network | |
Nurcheshmeh et al. | The effect of normal stress on the formability of sheet metals under non-proportional loading | |
Narita et al. | Evaluation of strength of stainless steel bolt without heat treatment considering Bauschinger effect during manufacturing process | |
Lee et al. | Calibration of distortional plasticity framework and application to U-draw bending simulations | |
Lee et al. | Evaluation of loading-path-dependent constitutive models for springback prediction in martensitic steel forming | |
Pham et al. | Application of a graphical method on estimating forming limit curve of automotive sheet metals | |
WO2023032370A1 (ja) | 金属材料の引張圧縮反転負荷挙動予測方法 | |
Shim et al. | An effective formability model of pearlitic steel wires in multi-stage drawing process based on the stress-based forming limit criterion | |
Martínez-Palmeth et al. | Formability limits of high-strength H240LA steel sheets under stress/strain gradients | |
Jung et al. | Constitutive modeling of asymmetric hardening behavior of transformation-induced plasticity steels | |
Joun et al. | Tensile test based material identification program afdex/mat and its application to two new pre-heat treated steels and a conventional cr-mo steel | |
Gothivarekar et al. | The Influence of Post-Necking Strain Hardening Behaviour on Fatigue Lifetime Prediction of Cold-Formed High Strength Steel. | |
Sumikawa et al. | Stress state dependency of unloading behavior in high strength steels | |
Junior et al. | Numerical simulation of a rapid fatigue test of high Mn-TWIP steel via a high cycle fatigue constitutive law | |
Nakwattanaset et al. | Formability prediction of high‐strength steel sheet using experimental, analytical and theoretical analysis based on strain and stress forming limit curves and its application to automotive forming parts: Vorhersage der Umformbarkeit von hochfesten Stahlblechen mittels experimenteller, analytischer und theoretischer Analyse auf der Grundlage von Grenzformänderungskurven für Dehnungen und Spannungen und ihre Anwendung auf die Umformung von Automobilteilen |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 22863958 Country of ref document: EP Kind code of ref document: A1 |
|
WWE | Wipo information: entry into national phase |
Ref document number: 2022863958 Country of ref document: EP Ref document number: 2401001294 Country of ref document: TH |
|
WWE | Wipo information: entry into national phase |
Ref document number: 202280059666.7 Country of ref document: CN |
|
ENP | Entry into the national phase |
Ref document number: 2022863958 Country of ref document: EP Effective date: 20240228 |
|
ENP | Entry into the national phase |
Ref document number: 20247010508 Country of ref document: KR Kind code of ref document: A |
|
NENP | Non-entry into the national phase |
Ref country code: DE |