Patent Application
20240319706
The present disclosure relates to a material design support apparatus, a material design support method, and a program.
In conventional design of metallic materials, in order to produce metallic materials with desired characteristics (tensile strength, hardness, toughness, plastic workability, and the like), component compositions of elements in metals and production conditions are determined empirically or by trial and error.
However, as the number of variable items increases in the component compositions of elements in metals and in the production conditions, the human and time burden of designing metallic materials has increased. In order to reduce the human and time burden described above, it has been proposed to design materials through optimization calculations and the like using a computer.
For example, in Patent Document 1, in order to reduce the increase in the computational load on the design of metallic materials, a design support method has been proposed, which includes: a search step of inputting a desired characteristic value in a database that is generated by using at least one mathematical model in which input information including a component composition of elements in metals and a production condition, and output information including a characteristic value of a metallic material, are related, and stores output data of the mathematical model in relation to input data for each mesh in which a range of inputs corresponding to the input information is separated by a plurality of sections, and retrieving a component composition of elements in metals and a production condition; and a presentation step of presenting the component composition of the elements in metals and the production condition corresponding to the desired characteristic value retrieved by the search step.
However, when the number of dimensions of design conditions (the number of combinations of conditions) of materials is large, a large amount of calculation time and resources are required to optimize design conditions. Therefore, an efficient method for optimizing design conditions for materials is required.
One object of the present disclosure is to provide a material design support device, a material design support method, and a program for supporting efficient optimization of design conditions for materials.
The present disclosure includes the following configurations.
According to the present disclosure, it is possible to support efficient optimization of design conditions for materials.
Next, embodiments of the present invention will be described in detail. The present invention is not limited to the following embodiments.
The user terminal 12 is an information processing terminal such as a PC, a tablet terminal, a smartphone, or the like, operated by a user such as a material designer. The user terminal 12 receives an input operation or a change operation of information necessary for optimizing a design condition for a material from the user, and causes the material design support apparatus 10 to execute the processing for supporting the optimization of the design condition for the material. The user terminal 12 receives the result of the processing for supporting the optimization of the design condition for the material executed by the material design support apparatus 10, and displays the result on a display device for checking by the user, for example.
The material design support apparatus 10 is an information processing apparatus such as a PC for executing the processing for supporting the optimization of the design condition for the material. The material design support apparatus 10 executes the processing for supporting the optimization of the design condition for the material using a trained model in which a correspondence between a design condition for a material and a characteristic value of the material is learned. The material design support apparatus 10 transmits the result of the executed processing for supporting the optimization of the design condition for the material to the user terminal 12.
The material design support system 1 illustrated in
The material design support apparatus 10 and the user terminal 12 illustrated in
The input device 501 is a touchscreen panel, operation keys or buttons, a keyboard, a mouse, or the like used by the user to input various signals. The display device 502 includes a display, such as a liquid crystal display, an organic EL display, or the like for displaying a screen, a speaker for outputting sound data, such as voice, sound, or the like, and the like. The communication I/F 507 is an interface used by the computer 500 to perform the data communication.
In addition, the HDD 508 is an example of a non-volatile storage device that stores programs and data. The stored programs and data include an operating system (OS) that is basic software for controlling the entire computer 500, applications for providing various functions on the OS, or the like. The computer 500 may use a drive device (for example, a solid state drive (SSD) or the like) using a flash memory as a storage media, in place of the HDD 508.
The external I/F 503 is an interface provided with respect to an external device. Examples of the external device include a recording medium 503a or the like. Thus, the computer 500 can read from and/or write to the recording medium 503a via the external I/F 503. Examples of the recording medium 503a include a flexible disk, a CD, a DVD, an SD memory card, a USB memory, or the like.
The ROM 505 is an example of a non-volatile semiconductor memory (storage device) that can retain programs and data even when the power is turned off. The ROM 505 stores programs and data, such as a BIOS that is executed when the computer 500 is started, OS settings, network settings, or the like. The RAM 504 is an example of a volatile semiconductor memory (storage device) that temporarily stores the programs and data.
The CPU 506 is an arithmetic unit that reads the programs and data from a storage device, such as the ROM 505, the HDD 508, or the like into the RAM 504, and executes processing to control the entire computer 500 or implement functions thereof. The computer 500 according to the one embodiment implements various functions of the material design support apparatus 10 and of the user terminal 12, which will be described later, by executing programs.
The configuration of the material design support system 1 according to the present embodiment will be described below.
The material design support apparatus 10 of the material design support system 1 illustrated in
The information display unit 20 displays information on the display device 502 or the like. For example, the information display unit 20 displays on the display device 502 a screen for receiving an input operation of information necessary for optimizing the design condition for the design target material, a screen for displaying the result of the processing for supporting the optimization of the design condition for the design target material, and the like. The operation reception unit 22 receives, from the user, the input operation or the change operation of information necessary for optimizing the design condition for the design target material. The request transmission unit 24 transmits a request for the processing for supporting the optimization of the design condition for the design target material, to the material design support apparatus 10. The response reception unit 26 receives a response to the request for the processing transmitted by the request transmission unit 24.
The design condition setting unit 30 sets a range of a design condition for an aluminum alloy, which is an example of a design target material. For example, the design condition setting unit 30 causes the user terminal 12 to display an input screen of the design condition for the user to set the range of the design condition. The design condition for the design target material includes the composition and the production condition of the design target material as information.
The required characteristic setting unit 32 sets a range of a required characteristic of the aluminum alloy, which is an example of the design target material. For example, the required characteristic setting unit 32 causes the user terminal 12 to display an input screen of the required characteristic for the user to set the range of the required characteristic.
The exhaustive prediction point generation unit 34 generates a plurality of exhaustive prediction points within the range of the set design condition for the design target material. For example, the exhaustive prediction point generation unit 34 calculates a plurality of numerical values, for each of the items such as the composition and the production condition included in the design condition for the design target material, at random or in predetermined increments within the range of the design condition for the design target material, and generates all combinations of the numerical values of the items as a plurality of exhaustive prediction points. The exhaustive prediction point generation unit 34 may adjust the time required for optimizing the design condition for the design target material by adjusting the number of exhaustive prediction points to be generated. The exhaustive prediction point generation unit 34 generates, for example, 10,000 exhaustive prediction points.
The prediction unit 36 reads a trained model from the trained model storage 50, inputs the plurality of exhaustive prediction points generated by the exhaustive prediction point generation unit 34 into the trained model, and predicts a characteristic value of the design target material for each of the plurality of exhaustive prediction points. The trained model storage 50 stores the trained model in which a correspondence between a design condition for a material such as an aluminum alloy and a characteristic value of the material have been previously acquired by machine learning and formulated.
As the trained model, a supervised training model such as a neural network or a Gaussian process may be applied. The trained model may be obtained by machine learning using training data that includes, for example, design conditions for a material such as an aluminum alloy read from a known material database and characteristic values of the material.
The design condition-material characteristic table storage 52 stores, in a design condition-material characteristic table described later, the characteristic value of the design target material for each of the exhaustive prediction points predicted by inputting the plurality of exhaustive prediction points into the trained model, in association with the exhaustive prediction point used for predicting the characteristic value. Thus, the design condition-material characteristic table storage 52 stores the design conditions (the exhaustive prediction points) of the design target material, which are inputs to the trained model, and the characteristic values (the predicted characteristic values) of the design target material, which are outputs from the trained model, in one data set in the same row of the design condition-material characteristic table.
The design condition adjustment unit 38 extracts a data set satisfying a range of the required characteristic of the design target material set by the required characteristic setting unit 32 from the design condition-material characteristic table storage 52, and displays an output screen of the range of the design condition satisfying the range of the required characteristic on the user terminal 12 for checking by the user.
The design condition adjustment unit 38 displays, by using a data set satisfying the range of the required characteristic of the design target material extracted from the design condition-material characteristic table storage 52, an output screen of the range of the predicted characteristic value of the design target material satisfying the range of the required characteristic on the user terminal 12 for checking by the user.
The design condition adjustment unit 38 may change, in response to a user operation to change the range of the required characteristic of the design target material, the display of the range of the design condition satisfying the range of the required characteristic displayed on the user terminal 12.
Accordingly, the design condition adjustment unit 38 can extract the range of the design condition that simultaneously satisfies the multidimensional required characteristic, by receiving the operation to set a range of the multidimensional required characteristic from the user. The design condition adjustment unit 38 can adjust the range of the design condition that simultaneously satisfies the multidimensional required characteristic by receiving the operation to change the range of the multidimensional required characteristic from the user.
For example, the design condition adjustment unit 38 may narrow the range of the design condition by receiving the operation to change the required characteristic in the direction to be increased from the user. The design condition adjustment unit 38 may expand the range of the design condition by receiving the operation to change the required characteristic in the direction to be decreased from the user.
Further, the design condition adjustment unit 38 may narrow the range of the design condition for the design target material used for generating the previous exhaustive prediction points, to the range of the design condition that satisfies the range of the required characteristic of the design target material, as the range of the design condition used for subsequently generating the exhaustive prediction points. The design condition adjustment unit 38 can narrow the design space of the design target material, by narrowing the range of the design condition of the design target material used for generating the exhaustive prediction points to the range of the design condition that satisfies the range of the required characteristic of the design target material. The design condition adjustment unit 38 can obtain a more precise range of design conditions by regenerating the exhaustive prediction points within the narrowed range of the design condition and re-predicting the characteristic value of the design target material.
The request response control unit 40 controls a process of receiving a request from the user terminal 12 and returning a response to the request to the user terminal 12. For example, the request response control unit 40 receives a request for the processing for supporting the optimization of the design condition of the design target material from the user terminal 12 and transmits a response to the received request for the processing to the user terminal 12.
The processing by the material design support system 1 according to the present embodiment will be described.
The design condition setting unit 30 of
The input screen 1012 of
The display page switched by the selection operation of the tab 1006 is a display page for the user to set the range of the production condition among the design conditions for the design target material. For example, the minimum and maximum values of items related to the production condition are to be set. Examples of the items of the production condition include: temperature (° C.) and time (h) for each process such as annealing, solution annealing, artificial age hardening, natural aging, hot working, cold working, stabilization, homogenization, and the like as items related to heat treatment; and processing rate, pressing ratio, surface reduction ratio, product shape, and the like as items related to processing conditions.
Details of the display page that is switched by the selection operation of the tab 1002 for the user to set the range of the required characteristic of the design target material will be described later.
Returning to
The exhaustive prediction point generation unit 34 calculates a plurality of numerical values at random for each item within the range of mass percentage of Si in the item related to the composition and the range of annealing time in the item related to the production condition. The exhaustive prediction point generation unit 34 generates all combinations of the plurality of numerical values for each item and outputs them as exhaustive prediction points.
The prediction unit 36 reads the trained model from the trained model storage 50, inputs the plurality of exhaustive prediction points generated by the exhaustive prediction point generation unit 34 into the trained model, and predicts the characteristic value of the 6061 alloy for each of the plurality of exhaustive prediction points. The trained model storage 50 stores the trained model in which the correspondence between input information including design conditions for the 6061 alloy and characteristic values of the 6061 alloy have been previously acquired by machine learning and formulated.
Examples of the items of the characteristic value include tensile strength, 0.2% proof stress, elongation, coefficient of linear expansion, Young's modulus, Poisson's ratio, fatigue strength, hardness, creep property, shear strength, Charpy impact value, specific heat, thermal conductivity, electrical resistivity, density, solidus, liquidus, and the like. Examples of the fatigue strength include tensile strength of a material after repeated loading tests (repeated tensile compression and the like) for 10 million cycles, and the stress value is used as the characteristic value.
The prediction unit 36 inputs the exhaustive prediction points generated by the exhaustive prediction point generation unit 34 into the trained model and stores the predicted characteristic value (output) of the 6061 alloy, in association with the exhaustive prediction points (input), in the design condition-material characteristic table storage 52.
As illustrated in
The required characteristic setting unit 32 of
The design condition adjustment unit 38 extracts a data set from the design condition-material characteristic table storage 52, and by using the data set, displays an output screen as described later on the user terminal 12 for checking by the user.
”.
By checking the output screen 1112 of
By the selection operation of the tab 1104, the output screen 1112 of
For example, in the box-and-whisker diagram in the lower part 1114b, three bars indicate the maximum, median, and minimum values of the range of the predicted characteristic value satisfying the range of the required characteristic of the 6061 alloy, the lower end of the box indicates the first quantile, and the upper end of the box indicates the third quantile.
By checking the output screen 1114 of
For example, in the example of the output screen 1114 illustrated in
By the selection operation of the tab 1106, the output screen 1112 illustrated in
When the user terminal 12 displaying the output screen such as the output screen 1112 illustrated in
The design condition adjustment unit 38 extracts, for example, an exhaustive prediction point that satisfies the range of the required characteristic of the 6061 alloy set by the required characteristic setting unit 32, and narrows the range of the design condition based on the extracted exhaustive prediction point.
For example, the design condition adjustment unit 38 proposes the range of the design condition corresponding to the exhaustive prediction point that satisfies the range of the required characteristic to the user, as the narrowed range of the design condition. The design condition adjustment unit 38 may expand, by a predetermined percentage, the range of the design condition corresponding to the exhaustive prediction point that satisfies the range of the required characteristic, and propose the expanded range of the design condition to the user, as the narrowed range of the design condition.
By performing a selection operation such as pressing a “reflect proposed material” button 1208 on the input screen 1212 of
By performing a selection operation such as pressing a “calculate” button 1210 on the input screen 1212 of
When the exhaustive prediction point generation unit 34 of
The design condition-material characteristic table storage 52 restores the characteristic value of the 6061 alloy for each of the re-predicted exhaustive prediction points in the design condition-material characteristic table, in association with the exhaustive prediction point used for the re-prediction of the characteristic value.
The design condition adjustment unit 38 extracts a data set from the design condition-material characteristic table storage 52, and by using the data set, displays an output screen as illustrated in
In the output screen 1112 illustrated in ”.
By checking the output screen 1112 of
By the selection operation of the tab 1104 illustrated in
By checking the output screen 1114 of
For example, in the example of the output screen 1114 of
The user can adjust the range of the required characteristic as illustrated in
In the output screen 1114 of
The operation for increasing the lower limit value of the tensile strength, which is an example of the required characteristic, may be an operation for extracting the upper data (such as the data of the upper 2.5%) among the tensile strengths of the plurality of exhaustive prediction points predicted as one of the characteristic values of the 6061 alloy.
The output screen 1114 of
When the output screen 1114 illustrated in
When the output screen 1114 illustrated in FIG. 15 receives the user operation to change the upper limit and the lower limit of the required characteristic, both the state before the user operation is received and the state after the user operation is received may be displayed in different display formats (different colors, different concentrations, and the like).
Thus, while adjusting the range of the required characteristic of the 6061 alloy, the user can narrow the range of the design condition for the 6061 alloy to satisfy the range of the required characteristic of the 6061 alloy.
When the user terminal 12 displaying the output screen such as the output screen 1112 of
The design condition adjustment unit 38 extracts, for example, the exhaustive prediction point that satisfies the range of the modified required characteristic of the 6061 alloy, and narrows the range of the design condition based on the extracted exhaustive prediction point. For example, the design condition adjustment unit 38 proposes the range of the design condition corresponding to the exhaustive prediction point that satisfies the modified range of the required characteristic to the user as the range of the narrowed design condition.
By performing a selection operation such as pressing the “reflect proposed material” button 1208 on the input screen 1212 of
By performing a selection operation such as pressing the “calculate” button 1210 of the input screen 1212 of
When the exhaustive prediction point generation unit 34 of
The design condition-material characteristic table storage 52 restores the characteristic value of the 6061 alloy for each of the re-predicted exhaustive prediction points in the design condition-material characteristic table, in association with the exhaustive prediction point used for the re-prediction of the characteristic value.
The design condition adjustment unit 38 extracts a data set from the design condition-material characteristic table storage 52, and by using the data set, displays an output screen as illustrated in
”.
By checking the output screen 1112 of
By the selection operation of the tab 1104 illustrated in
By checking the output screen 1114 of
For example, in the example of the output screen 1114 of
By repeating the above process, the user can check, for example, the range of the composition satisfying the range of the required characteristic of the 6061 alloy and the re-predicted range of the characteristic value satisfying the range of the required characteristic of the 6061 alloy, as illustrated in the output screen 1114 illustrated in
In step S10, the design condition setting unit 30 causes the user terminal 12 to display the input screen 1012 as illustrated in
In step S12, the required characteristic setting unit 32 causes the user terminal 12 to display the input screen 1014 as illustrated in
In step S14, the exhaustive prediction point generation unit 34 generates a plurality of exhaustive prediction points, such as 10,000 points, within the range of the design condition for the 6061 alloy that the design condition setting unit 30 has received from the user. In step S16, the prediction unit 36 selects one unselected exhaustive prediction point from the plurality of exhaustive prediction points generated by the exhaustive prediction point generation unit 34.
In step S18, the prediction unit 36 reads out the trained model from the trained model storage 50. The prediction unit 36 inputs one exhaustive prediction point selected in step S16 into the trained model and predicts the characteristic value of the one exhaustive prediction point selected in step S16. In step S20, the design condition-material characteristic table storage 52 generates a data set in which the one exhaustive prediction point selected in step S16 is associated with a characteristic value predicted by inputting the one exhaustive prediction point into the trained model, and stores the data set in the design condition-material characteristic table.
In step S22, the prediction unit 36 determines whether there is an unselected exhaustive prediction point in the exhaustive prediction points generated by the exhaustive prediction point generation unit 34. When there is an unselected exhaustive prediction point, the prediction unit 36 returns to the process in step S16 and continues generating the data set.
When there is no unselected exhaustive prediction point, the material design support system 1 according to the present embodiment proceeds to an information display process in step S24 and performs the process for narrowing the design condition using the design condition-material characteristic table.
In step S40, the design condition adjustment unit 38 extracts a data set satisfying the required characteristic from the design condition-material characteristic table storage 52. In step S42, by using the extracted data set, the design condition adjustment unit 38 causes the user terminal 12 to display the output screen 1114 illustrated in
For example, when the user terminal 12 displaying the output screen such as the output screen 1114 of
For example, when the user selects the narrowed range of the design condition for the 6061 alloy proposed in step S44, the user performs a selection operation such as pressing the “reflect proposed material” button 1208 on the input screen 1212 illustrated in
In step S46, the design condition adjustment unit 38 determines whether the narrowed range of the design condition for the 6061 alloy proposed in step S44 has been selected by the user. When the narrowed range of the design condition for the 6061 alloy proposed in step S44 has been selected by the user, the design condition adjustment unit 38 proceeds to the process in step S48, changes to the narrowed range of the design condition for the 6061 alloy proposed in step S44, and returns to the process in step S14.
Therefore, in the material design support system 1 according to this embodiment, the exhaustive prediction points can be regenerated from the narrowed range of the design condition for the 6061 alloy, and the characteristic value of the 6061 alloy can be re-predicted for each of the regenerated exhaustive prediction points.
When the user does not select the narrowed range of design condition for the 6061 alloy proposed in step S44, the design condition adjustment unit 38 proceeds to the process of step S50. In step S50, the design condition adjustment unit 38 determines whether or not a user operation to adjust the range of the required characteristic has been received via the output screen 1114 of
When the user operation to adjust the range of the required characteristic is received, the design condition adjustment unit 38 proceeds to the process in step S52. In step S52, the design condition adjustment unit 38 extracts a data set satisfying the adjusted required characteristic from the design condition-material characteristic table. In step S54, by using the extracted data set, the design condition adjustment unit 38 causes the user terminal 12 to display the output screen 1114 illustrated in
When the user operation to adjust the range of the required characteristic is not received, the design condition adjustment unit 38 skips the process in steps S52 and S54. In step S56, the design condition adjustment unit 38 determines whether the user operation to end the screen display is received. When the user operation to end the screen display is not received, the design condition adjustment unit 38 returns to the process of step S44. When the user operation to end the screen display is received, the design condition adjustment unit 38 ends the information display process illustrated in
According to the process illustrated in
The process illustrated in
In step S100, the design condition setting unit 30 causes the user terminal 12 to display the input screen 1012 as illustrated in
In step S102, the required characteristic setting unit 32 causes the user terminal 12 to display the input screen 1014 as illustrated in
In step S106, the exhaustive prediction point generation unit 34 generates a plurality of exhaustive prediction points, such as 10,000 points, within the range of the design condition for the 6061 alloy, which the design condition setting unit 30 has received from the user. In step S108, the prediction unit 36 inputs the exhaustive prediction points generated by the exhaustive prediction point generation unit 34 into the trained model to predict the characteristic value of each of the plurality of exhaustive prediction points. The design condition-material characteristic table storage 52 generates a data set in which the exhaustive prediction point is associated with the characteristic value predicted by inputting the exhaustive prediction point into the trained model, and stores the data in the design condition-material characteristic table.
In step S110, the design condition adjustment unit 38 extracts a data set satisfying the required characteristic from the design condition-material characteristic table. The design condition adjustment unit 38 can analyze, by using the extracted data set, the range of the design condition satisfying the range of the required characteristic of the 6061 alloy and the predicted range of the characteristic value.
In step S112, the design condition adjustment unit 38 determines whether the design objective such as “maximizing tensile strength” set in step S104 has been achieved. When the design objective has not been achieved, the design condition adjustment unit 38 proceeds to the process in step S114. In step S114, the material design support system 1 according to the present embodiment regenerates a plurality of exhaustive prediction points within the range of the design condition narrowed to the range satisfying the range of the required characteristic, and returns to the process in step S108. Further, when the design objective is, for example, “maximizing tensile strength”, a step may be added in which the data having the highest tensile strength (the top 2.5% of the data, and the like) is extracted and a plurality of exhaustive prediction points are regenerated within the range of the design condition.
Thus, the material design support system 1 according to the present embodiment repeats the processes of steps S108 to S114, narrowing the design objective to the range of the design condition satisfying the range of the required characteristic, until the design objective such as “maximizing tensile strength” set in step S104 is achieved. When the design objective is achieved, the material design support system 1 according to the present embodiment ends the processing of the flowchart of
According to the processing illustrated in
The results of the processing for supporting the optimization of the design condition for the material, performed by the material design support apparatus 10 according to the present embodiment, may be used as input information to a production apparatus for generating a design target material, such as an aluminum alloy and the like. The results of the processing performed by the material design support apparatus 10 according to the present embodiment may be input to the production apparatus. As a result, technical parameters of the production apparatus are determined such that design target materials in a range satisfying the required characteristic are produced, and the production process can be controlled.
In the present embodiment, an example has been described in which the exhaustive prediction point generation unit 34 generates one set of a plurality of exhaustive prediction points within the set range of the design condition for the design target material, but a plurality of sets may be generated, and the set that best satisfies the required characteristic may be selected and used.
The material design support system 1 according to the present embodiment can provide a material design support apparatus, a material design support method, and a program for supporting efficient optimization of the design conditions for the design target materials.
Although the embodiments are described heretofore, various modifications to the form and details may be made without departing from the subject matter and scope of the appended claims.
Although the present invention is described with reference to the embodiments, the present invention is not limited to the embodiments described above, and various modifications can be made without departing from the scope of the subject matter recited in the claims. The present international application is based upon and claims priority to Japanese Patent Application No. 2021-120261, filed on Jul. 21, 2021, the entire contents of which are incorporated herein by reference.
| Number | Date | Country | Kind |
|---|---|---|---|
| 2021-120261 | Jul 2021 | JP | national |
| Filing Document | Filing Date | Country | Kind |
|---|---|---|---|
| PCT/JP2022/027887 | 7/15/2022 | WO |
