Information processing method, presentation method, program, and manufacturing method

Abstract

To provide an information processing method, a presentation method, a program, and a manufacturing method that, with respect to a molecule constituting a substance, enable a change in a physical property value of the substance to be calculated without an experimental value. An information processing method executed by a computer system (1): acquires a chemical formula of each of one or more molecules constituting a substance and a configuration ratio of the one or more molecules (S1); calculates a molecular feature amount of each of the one or more molecules from the chemical formula of each of the one or more molecules (S2); and generates physical property value change data representing a change in a thermal physical property value of the substance by using the configuration ratio and the molecular feature amount (S3).

Description

Information processing method, presentation method, program, manufacturing method

[0001] The present disclosure relates to an information processing method, a presentation method, a program, and a manufacturing method.

[0002] Conventional thermal fluid calculation and thermal cycle design methods include computational fluid dynamics (CFD), thermal cycle simulations that can calculate the heat circulation of the entire circuit by arranging heat exchangers, piping, etc., and calculating the heat generated in each process, and mechatronics system simulations that can calculate waste heat changes by arranging heat exchangers, piping, etc. designed on a computer. However, calculations that involve macroscopic structures such as flow paths, such as these CFD methods, require visualization of thermal fluid properties such as density and enthalpy, behavior including phase changes such as gas-liquid phases, transient phenomena such as dissolution, and mesoscopic behavior including such dynamic processes, and then appropriately modeling some or all of them.

[0003] As described in Non-Patent Document 1, Gibbs Ensemble Monte Carlo Simulation (GEMC) has been used to understand such mesoscopic behavior. This method calculates the expected value of a physical quantity in statistical mechanics by numerical integration using random numbers to realize the Boltzmann distribution. However, because this method is only applicable in equilibrium states, it is difficult to calculate quasi-static and dynamic processes that change over time, which are essential for thermal cycle and thermal fluid calculations.

[0004] On the other hand, molecular dynamics (MD) is a method for analyzing the non-equilibrium behavior of atoms and molecules. In the molecular dynamics method, it is necessary to set the characteristics of atoms and molecules using various potential parameters called force fields. Conventionally, there has been a method for setting the Leonard-Jones potential parameters among the potential parameters, as in Patent Document 1.

[0005] The Journal of Physical Chemistry B, (US), 2010, Vol. 114, Issue 31, p. 10133-10142

[0006] Japanese Patent Application Laid-Open No. 2021-005344

[0007] However, this method only automatically determines a portion of the many potential parameters included in the MD method, and still requires experimental values ​​obtained from experimental results and previous literature, so it is not necessarily an effective method.

[0008] The present disclosure provides an information processing method, a presentation method, a program, and a manufacturing method that enable calculation of changes in the physical properties of a substance without experimental values ​​for the molecules that make up the substance.

[0009] An information processing method according to one aspect of the present disclosure is an information processing method executed on a computer system, which obtains the chemical formula of each of one or more types of molecules that make up a substance and the composition ratio of the one or more types of molecules, calculates molecular feature values ​​for each of the one or more types of molecules from the chemical formula of each of the one or more types of molecules, and generates physical property value change data that represents changes in the physical property values ​​of the substance using the composition ratios and molecular feature values.

[0010] A presentation method according to one aspect of the present disclosure is a presentation method for thermal fluid calculations executed on a computer system, which accepts input of the chemical formula of each of one or more molecules that make up a substance and the composition ratio of the one or more molecules, generates an image showing continuous data for each physical property value required for thermal fluid calculations for the input chemical formula and composition ratio, and presents the generated image.

[0011] A program according to one aspect of the present disclosure is a program for causing a computer system to execute the above-described information processing method.

[0012] A manufacturing method according to one aspect of the present disclosure is a method for manufacturing data on changes in physical properties of a substance, executed by a computer system, which method obtains the chemical formula of each of one or more molecules that make up the substance and the composition ratio of the one or more molecules, determines molecular feature values ​​for each of the one or more molecules from the chemical formula of each of the one or more molecules, and generates physical property change data that represents changes in the physical properties of the substance using the composition ratio and the molecular feature values.

[0013] Aspects of the present disclosure enable the calculation of changes in the physical properties of a substance without experimental values ​​for the molecules that make up the substance.

[0014] FIG. 1 is a block diagram of a computer system that executes an information processing method according to an embodiment; FIG. 2 is a flowchart of an information processing method according to an embodiment; FIG. 3 is an explanatory diagram of an example of a molecular structure; FIG. 4 is an explanatory diagram of a method for determining a Leonard-Jones potential parameter; FIG. 5 is a graph showing the relationship between the intermolecular distance and potential energy when the orientation angle is changed; FIG. 6 is an explanatory diagram of an example of the physical structure of a substance;

[0015] [1. Embodiment] [1.1 Configuration] An information processing method according to this embodiment is a method executed in a computer system for calculating changes in physical properties of a substance. The information processing method according to this embodiment enables prediction of physical changes in a substance. The information processing method according to this embodiment can be applied to simulation of thermal fluid cycles in geometric shapes, computational fluid dynamics, mechatronics system simulation, or thermal fluid analysis.

[0016] 1 is a block diagram of a computer system 1 that executes an information processing method. The computer system 1 includes an input device 11, an output device 12, a communication device 13, a storage device 14, and an arithmetic circuit 15.

[0017] The input device 11 includes one or more human-machine interfaces for inputting information. Examples of the human-machine interfaces include a keyboard, a pointing device (such as a mouse or a trackball), a touchpad, a position input device such as a touch panel display, etc. The one or more human-machine interfaces of the input device 11 may be built-in or external.

[0018] The output device 12 includes one or more human-machine interfaces for outputting information. Examples of the human-machine interfaces include a display, a speaker, a touch panel display, etc. The one or more human-machine interfaces of the output device 12 may be built-in or external.

[0019] The communication device 13 is connectable to a communication network and has the function of communicating through the communication network. The communication device 13 complies with a predetermined communication protocol. The predetermined communication protocol may be selected from various well-known wired and wireless communication standards.

[0020] The storage device 14 is used to store information used by the computer system 1 and information generated by the computer system 1. The storage device 14 includes one or more storages (non-transitory storage media). The storage may be, for example, a hard disk drive, an optical drive, or a solid-state drive (SSD). The storage may also be an internal, external, or network-attached storage (NAS) type.

[0021] The arithmetic circuit 15 is connected to the input device 11, the output device 12, and the communication device 13, and can access the storage device 14. The arithmetic circuit 15 includes one or more processors (microprocessors) and one or more memories. The one or more processors execute programs (stored in one or more memories or storage devices) to realize various functions. The programs may be pre-recorded in the storage device, or may be provided via a telecommunications line such as the Internet, or recorded on a non-transitory storage medium such as a memory card.

[0022] In this embodiment, the computer system 1 is realized by a single computer. Examples of the computer include a server, a personal computer (desktop computer, laptop computer), a mobile terminal (smartphone, tablet terminal, etc.), etc.

[0023] The information processing method of this embodiment is realized by executing a program by the computer system 1. Fig. 2 is a flowchart of the information processing method according to this embodiment.

[0024] The information processing method includes an acquisition step S1, a molecular feature determination step S2, a physical property change data generation step S3, a geometric shape condition setting step S4, a physical change prediction step S5, and a physical change output step S6.

[0025] The acquisition step S1 acquires information about a substance. The information about the substance includes the chemical formula of each of one or more molecules that constitute the substance and the composition ratio of the one or more molecules. The chemical formula can be selected from one or more of an empirical formula (elementary composition formula, composition formula), a molecular formula, a structural formula, and a rational formula. In this embodiment, the chemical formula is a structural formula. A structural formula not only indicates the number of atoms in each molecule, but also describes how the atoms constitute the molecule and explicitly or implicitly indicates the chemical bonds between the atoms, making it easier to identify the compound and is therefore preferable. The composition ratio is the ratio of one or more molecules that constitute the substance. A substance can be a pure substance or a mixture. If the substance is a pure substance, it is composed of only one type of molecule, and naturally, the composition ratio is 100% of this one type of molecule. If the substance is a mixture, it is composed of multiple types of molecules, and the composition ratio indicates the molar ratio or mass ratio of each molecule. As an example, substance X is identified as containing two types of molecules, A and B, in a molar composition ratio of C:D.

[0026] The molecular feature amount determination step S2 determines the molecular feature amount of each of the one or more molecules from the chemical formula of each of the one or more molecules.

[0027] The molecular feature amount indicates the characteristics of a substance at the molecular level. The molecular feature amount is determined from the chemical formula or structural formula of each of one or more molecules of the substance acquired in the acquisition step S1, based on first principles, by quantum chemical calculation. Examples of quantum chemical calculation methods include density functional theory (DFT), coupled cluster (CC), and multireference configuration interaction (MRCI).

[0028] The molecular feature amount includes a molecular structure and a molecular force field.

[0029] The molecular structure is determined, for example, based on the most stable structure obtained from the chemical formula of each molecule by quantum chemical calculation. The most stable structure is determined based on a structural optimization calculation. In other words, the most stable structure of a molecule can be determined by solving a minimum problem on a potential energy surface. As an example, the potential energy can be determined using, for example, the coupled cluster (CC) method, which is an ab initio molecular orbital method, and a correlation-consistent basis set that includes a polarization function as a basis function.

[0030] In this embodiment, the molecular structure includes at least one of a bond length, a bond angle, or a dihedral angle. The bond length is the distance between two atoms in a molecule. The bond angle is the angle between the bond axes extending from one atom to two atoms when two atoms are bonded to one atom in a molecule. The dihedral angle is the angle between two planes passing through two sets of three atoms sharing two atoms. The bond length, bond angle, or dihedral angle can be determined from the most stable structure. The bond length, bond angle, or dihedral angle can be determined for each combination of atoms that can define them.

[0031] Figure 3 is an explanatory diagram of an example of a molecular structure. The molecular structure shown in Figure 3 has atoms A1 to A6. Atom A1 is bonded to atom A2, two atoms A3 and A4 are bonded to atom A1, and two atoms A5 and A6 are bonded to atom A2. The bonding relationships between atoms A1 to A6 can be determined from the chemical formula of the molecule.

[0032] 3, the bond lengths that can be calculated are the bond length between atoms A1 and A2, the bond length between atoms A1 and A3, the bond length between atoms A1 and A4, the bond length between atoms A2 and A5, and the bond length between atoms A2 and A6. The bond angles that can be calculated are the bond angle between atoms A3, A1, and A4, the bond angle between atoms A2, A1, and A3, the bond angle between atoms A2, A1, and A4, the bond angle between atoms A5, A2, and A6, the bond angle between atoms A1, A2, and A5, and the bond angle between atoms A1, A2, and A6. As the dihedral angles, the dihedral angles between the faces of atoms A1, A2, A3 and the faces of atoms A1, A2, A4, the dihedral angles between the faces of atoms A1, A2, A3 and the faces of atoms A1, A2, A5, the dihedral angles between the faces of atoms A1, A2, A3 and the faces of atoms A1, A2, A6, the dihedral angles between the faces of atoms A1, A2, A4 and the faces of atoms A1, A2, A5, the dihedral angles between the faces of atoms A1, A2, A4 and the faces of atoms A1, A2, A6, and the dihedral angles between the faces of atoms A1, A2, A5 and the faces of atoms A1, A2, A6 are determined.

[0033] In Figure 3, r represents the bond length between atoms A1 and A2. Θ represents the bond angle between atoms A1, A2, and A6. χ represents the dihedral angle between the planes of atoms A1, A2, and A3 and A4, i.e., the angle between plane P1 passing through the three atoms A1, A2, and A3 and plane P2 passing through the three atoms A1, A2, and A4.

[0034] The molecular force field is determined based on the molecular structure, and includes at least one of an intramolecular force field and an intermolecular force field (intermolecular interaction).

[0035] The intramolecular force field includes at least one force constant of a bond length, a bond angle, or a dihedral angle. The force constant of the bond length, the bond angle, or the dihedral angle is determined by, for example, normal mode analysis using the most stable structure. In this embodiment, the force constant of the bond length, the bond angle, or the dihedral angle can be determined by performing normal mode analysis on the most stable structure.

[0036] When determining the force constant of the bond length, the bond length of the most stable structure is expanded or contracted, and the potential energy is calculated multiple times. The potential energy can be determined, for example, using the coupled cluster method, which is an ab initio molecular orbital method, and a correlation-consistent basis set to which a polarization function is added as a basis function. The force constant of the bond length can be determined from the relationship between the obtained potential energy and the bond length. For example, the force constant of the bond length can be determined by fitting the relationship between the potential energy and the bond length with the following formula (1). In the following formula (1), E ｒ is the potential energy, r is the bond length, r ０ is the bond length of the most stable structure, k ｒ is the bond length force constant.

[0037]

[0038] When determining the bond angle force constant, the potential energy is calculated multiple times by increasing or decreasing the bond angle of the most stable structure by a predetermined angle (for example, 5 degrees). The method of calculating the potential energy may be the same as above. The bond angle force constant can be determined from the relationship between the obtained potential energy and the bond angle. For example, the bond angle force constant can be determined by fitting the relationship between the potential energy and the bond angle with the following formula (2). In the following formula (2), E Θ is the potential energy, Θ is the molecular angle, Θ ０ is the molecular angle of the most stable structure, k Θ is the molecular angle force constant.

[0039]

[0040] When determining the force constant of the dihedral angle, the phase is changed from the dihedral angle of the most stable structure, and the potential energy is calculated multiple times. The method of calculating the potential energy may be the same as above. The force constant of the dihedral angle can be determined from the relationship between the obtained potential energy and the dihedral angle. For example, the force constant of the dihedral angle can be determined by fitting the relationship between the potential energy and the dihedral angle with the following formula (3). In the following formula (3), E χ is the potential energy, χ is the dihedral angle, n is the number of faces, δ is the phase, k χ is the dihedral angle force constant.

[0041]

[0042] The intermolecular force field can be determined by charge distribution analysis. The intermolecular force field includes at least one of molecular charge and van der Waals potential parameters. The molecular charge is determined so as to reproduce the electrostatic potential of the most stable structure. Mulliken charge, RESP charge, NBO charge, AM1-BCC charge, etc. can be used to determine the molecular charge. In particular, RESP charge is preferable because it is based on the Coulomb potential of molecular dynamics. The van der Waals potential parameters can be determined using, for example, the Lennard-Jones potential or the Buckingham potential. The following equation (4) shows the general form of the Lennard-Jones potential.

[0043]

[0044] In the above formula (4), V ＬＪ is the Leonard-Jones potential, r ｉｊ is the intermolecular distance, ε ｉｊ , σ ｉｊ is the Leonard-Jones potential parameter, p is the order of the attractive term, and q is the order of the repulsive term. There are many forms of the order (p, q), such as the (6, 12) potential or the (6, 9) potential type, but from the viewpoint of ease of handling values ​​and extensibility to molecular dynamics calculation software, it is more desirable to have (p, q) = (6, 12).

[0045] The Leonard-Jones potential parameter ε required for the calculation of the Leonard-Jones potential ｉｊ , σ ｉｊ is also calculated by quantum chemical calculations.

[0046] FIG. 4 is an explanatory diagram of a method for determining the Lennard-Jones potential parameters. Specifically, for example, as shown in FIG. ｉ , m ｊ Prepare two molecules, fix the molecular structure by quantum chemical calculation, and determine the distance between them (intermolecular distance) r at which the potential energy becomes the most stable point. ｉｊ and the intermolecular orientation angle θ ｉｊ Ask for.

[0047] FIG. 5 shows the orientation angle θ ｉｊ When changing the intermolecular distance r ｉｊ 5 is a graph showing the relationship between the orientation angle θ ｉｊ is 90°, the "■" indicates the orientation angle θ ｉｊ is 60°, the "●" indicates the orientation angle θ ｉｊ is 30°, the "▼" indicates the orientation angle θ ｉｊ is 0°.

[0048] The most stable point is the intermolecular distance r ｉｊ and / or the intermolecular orientation angle θ ｉｊ The potential energy is calculated multiple times while changing the intermolecular distance r ｉｊ Leonard Johns Potential V ＬＪ The method for calculating the potential energy may be the same as above. ｉｊ Leonard Johns Potential V ＬＪ The calculation is performed to obtain the Lennard-Jones potential V ＬＪ and the intermolecular distance r ｉｊ The Leonard-Jones potential parameter ε ｉｊ , σ ｉｊ It is better to arrange three or more molecules of the same type, since a more precise Leonard-Jones potential can be obtained, and the accuracy of the parameters may also increase.

[0049] In this way, in the molecular feature determination step S2, molecular feature quantities such as molecular structures including bond lengths, bond angles, and dihedral angles for each of the one or more molecules, and molecular force fields including charges and van der Waals potential parameters, are determined from the chemical formulas of each of the one or more molecules by quantum chemical calculations based on first principles without using experimental values. Therefore, even if there are no experimental values ​​for the one or more molecules that constitute a substance, it is possible to determine the molecular feature quantities for each of the one or more molecules.

[0050] The physical property value change data generation step S3 determines the physical property values ​​(physical state) of the substance based on the chemical formula and composition ratio of each molecule. More specifically, the physical property value change data generation step S3 acquires the physical structure of the substance by molecular dynamics calculation from the molecular feature amount determined in the molecular feature amount determination step S2 and the composition ratio of one or more molecules acquired in the acquisition step S1, and calculates the physical property values ​​of the substance by statistical mechanics calculation using the physical structure of the substance calculated by the molecular dynamics calculation. The calculated discrete physical property values ​​are rewritten into continuous functions based on external input values ​​such as temperature and pressure, and physical property value change data is generated.

[0051] The physical structure of a substance includes at least one of a bulk structure, molecular orientation, a micro-aggregate structure, or phase separation. The initial state of the physical structure of a substance may be determined by using a molecular dynamics method described below to randomly arrange one or more molecules of the substance. The initial bulk structure of a substance can be determined, for example, by performing stable structure calculations based on the composition ratio and molecular feature quantities. As an example, the number of molecules of the substance is set to a desired value, the molecules of the substance are randomly arranged, and then stable structure calculations are performed to obtain the initial bulk calculation structure of the substance. The physical structure of a substance can be used to determine the physical properties of the substance.

[0052] 6 is a diagram showing an example of the physical structure of a substance. As shown in FIG. 6, the physical structure of a substance is given as a 3D representation such as the position coordinates or an image of each molecule.

[0053] Molecular dynamics calculations include NPT calculations in which the pressure and temperature are constant for the initial bulk structure, NVT calculations in which the volume and temperature are constant for the initial bulk structure, NVE calculations in which the volume and energy are constant for the initial bulk structure, NPE calculations in which the pressure and internal energy are constant for the initial bulk structure, BAR calculations that can calculate the solvation Gibbs free energy and excess Gibbs free energy of a substance, and ER calculations.

[0054] 7 is an explanatory diagram of the NPE calculation. In the NPE calculation, a simulation is performed on an initial bulk structure B1 under the condition that the pressure and internal energy are constant, thereby obtaining a structure B2 after the bulk structure has changed. The changed structure B2 obtained in this way can be used to calculate the physical properties of the material.

[0055] In molecular dynamics calculations, a large number of molecules must be placed in a certain compartment (cell) for calculation. The number of molecules placed in a compartment is preferably 200 or more, more preferably 500 or more. When the number of molecules is 200 or more, intermolecular interactions can be calculated, and when the number of molecules is 500 or more, intermolecular interactions can be stably calculated and potentials can be stably obtained by any NPT calculation, NVT calculation, NVE calculation, NPE calculation, BAR calculation, and ER calculation. The material physical structure obtained by calculation in this way can be used as time-series data, or, for example, structural data and data groups at a certain time can be extracted and used in subsequent statistical mechanics calculations.

[0056] The physical properties of a substance are obtained by performing statistical mechanics calculations based on the physical structure of the substance. These properties include at least one of the following: density, vapor pressure (saturated vapor pressure), kinematic viscosity, diffusion coefficient, latent heat (of vaporization / liquefaction), specific enthalpy, thermal conductivity, heat capacity at constant volume, heat capacity at constant pressure, excess Gibbs free energy, Gibbs free energy of solvation, and Gibbs free energy. Among these, the properties obtained during the equilibrium process are density, vapor pressure (saturated vapor pressure), kinematic viscosity, diffusion coefficient, latent heat (of vaporization / liquefaction), specific enthalpy, thermal conductivity, heat capacity at constant volume, heat capacity at constant pressure, and the boiling point and dew point of the pure substance. Among these, the physical properties obtained at equilibrium within the system are density, vapor pressure (saturated vapor pressure), latent heat (of vaporization / liquefaction), and specific enthalpy. On the other hand, the physical properties obtained in a non-equilibrium process within a system or between the physical structures of multiple substances are the dynamic viscosity, diffusion coefficient, thermal conductivity, heat capacity at constant volume, heat capacity at constant pressure, excess Gibbs free energy, solvation Gibbs free energy, and Gibbs free energy. However, the physical properties of substances are calculated as discrete values ​​linked to temperature and pressure.

[0057] NVT calculations are a computational technique that performs molecular dynamics calculations by assigning conditions for the number of molecules, pressure, and temperature. They can be used to determine, for example, density, (specific) enthalpy, constant-pressure heat capacity, latent heat, diffusion coefficient, and excess Gibbs free energy for the physical structure of the resulting substance. NVT calculations can be used to determine, for example, vapor pressure, constant-volume heat capacity, and the density of two layers in a gas-liquid mixture.

[0058] More specifically, in the case of density, it is obtained as an ensemble average of the volume of the calculated molecular structure for the number of molecules incorporated into the molecular dynamics calculation over the calculation time.

[0059] In the case of (specific) enthalpy, it can be calculated as the ensemble average over the calculation time of the sum of the internal energy U, which is the sum of the potential and kinetic energy due to intramolecular and intermolecular interactions using a molecular force field, and the product pV of the applied pressure and the volume of the molecular structure obtained by calculation.

[0060] In the case of constant pressure heat capacity, it can be calculated from the change in specific enthalpy when the temperature is changed slightly under constant pressure in the NPT calculation. More specifically, the small specific enthalpy difference and the temperature difference at that point during the calculation of the physical structure state are recorded, and the ensemble average of the divided results is calculated for the calculation time.

[0061] In the case of latent heat, the pressure is fixed at a desired value, and the NPT calculation is performed while sequentially changing only the temperature to obtain the specific enthalpy. Figure 8 is a graph showing a method for determining latent heat. As shown in Figure 8, in a graph with temperature on the horizontal axis and specific enthalpy on the vertical axis, the difference in specific enthalpy between the high-density state F1 and the low-density state F2 is defined as the latent heat and is calculated.

[0062] In the case of the diffusion coefficient, the mean square transition of the molecules during the NPT calculation is output over time, and the ensemble average of the slope is used to calculate it. The calculation time required for these ensemble averages is 1 ns or more to stabilize the numerical results, and 5 ns or more is more preferable to achieve even greater numerical stability.

[0063] Excess Gibbs free energy can be calculated using the following method. The physical structures of N molecules and one target molecule are obtained by molecular dynamics calculations, and the energy change caused by adding one target molecule to a gap in the physical structure of the N molecules is estimated. This is repeated M times to obtain an ensemble average of the energy when one target molecule is added to N molecules. Meanwhile, the physical structure and energy of N+1 molecules, obtained by previously adding one target molecule to N molecules, are calculated using molecular dynamics calculations. In reality, when one target molecule is added to N molecules, the surrounding N molecules are rearranged to become stable, so the energy of the N+1 physical structure should be lower. Therefore, excess Gibbs free energy is the energy obtained by subtracting the energy caused by adding one target molecule to a gap in the N molecules from the energy of the physical structure of N+1 molecules. Gibbs free energy can also be obtained by adding the Gibbs free energy of an ideal gas to the excess Gibbs free energy. The Gibbs free energy of an ideal gas can be easily calculated using statistical mechanics.

[0064] NVE calculations are a computational method that performs molecular dynamics calculations by assigning conditions for the number of molecules, volume, and temperature. By performing statistical mechanics calculations on the physical structural state of the obtained material, NVE calculations can be used to determine the thermophysical properties of adiabatic processes, such as entropy.

[0065] The NPT, NVT, NVE, NPE, BAR, or ER calculations may be used in combination to simulate a desired state or process of a material. For example, NVT, NPT, NVE, NPE, BAR, or ER calculations may be used in combination to obtain a suitable bulk structure or to perform a calculation to obtain physical properties after annealing the bulk structure.

[0066] By performing statistical mechanics calculations using the material structure determined by molecular dynamics calculations as described above, it is possible to determine the dynamic viscosity coefficient, density, latent heat, (specific) enthalpy, thermal conductivity, heat capacity at constant volume, heat capacity at constant pressure, diffusion coefficient, vapor pressure, (specific) enthalpy change due to adiabatic compression, boiling point, dew point, excess Gibbs free energy, solvation Gibbs free energy, or Gibbs free energy.

[0067] As described above, the physical properties of a material can be calculated by performing statistical mechanics calculations using the material structure determined by molecular dynamics calculations. Typically, material characteristic values ​​are obtained for each material structure. That is, the physical properties are typically obtained as a list of fragmented or discretized physical properties. Such physical properties may be difficult to apply to thermal fluid cycle simulations, computational fluid dynamics, mechatronics system simulations, or thermal fluid analysis. Thermal fluid cycle simulations, etc., require physical state values ​​corresponding to various conditions. For example, in computational fluid dynamics, temperature may change continuously. With only a list of physical properties, a physical state value corresponding to the required temperature may not exist in the list, and separate considerations must be given to handling such missing values. Furthermore, in computational fluid dynamics, the temperature may be considered to be a nearly constant condition. The thermal fluid in this embodiment can be applied to fluid calculations both when the temperature changes and when it is fixed.

[0068] Therefore, for physical properties, multiple values ​​(discrete values) obtained with respect to temperature or pressure may be acquired, and the physical property change data may include a continuous function with respect to temperature or pressure. When converting such physical property change data into a continuous function, for example, the discrete values ​​of the physical property may be converted into a continuous function by approximation through interpolation or fitting between the discrete values, or by using other software. For example, by using machine learning techniques such as deep learning, symbolic regression, or Bayesian regression, or the least squares method, it is possible to convert the physical property into a continuous function in an appropriate form based on the discrete values ​​of the physical property. Using such a continuous function conversion process makes it possible to output physical property values ​​corresponding to various inputs.

[0069] Next, an example of continuous function generation will be described with reference to FIGS. 9 to 11 . FIG. 9 is an explanatory diagram of continuous function generation by fitting density to the inverse of temperature. As shown in FIG. 9 , in the case of density, density is plotted against calculation points where the temperature condition is changed relative to the inverse of a certain temperature. A fitting function can be determined to satisfy these plot points. Next, a temperature vs. density graph is drawn while sequentially changing the pressure, and the parameter changes of the previously determined fitting function are determined. By using the fitting function on a graph drawn by setting the pressure used in the calculation on the x-axis and the parameters of the obtained fitting function on the y-axis, the fitting parameters can be determined by temperature and pressure, resulting in continuous function generation. FIG. 10 is an explanatory diagram of continuous function generation by fitting kinematic viscosity to temperature. FIG. 11 is an explanatory diagram of continuous function generation by fitting vapor pressure to temperature. Kinematic viscosity and vapor pressure can also be converted into continuous functions using the same method described above.

[0070] As described above, in this embodiment, the physical property value change data generation step S3 includes a step of obtaining discrete physical property values ​​by performing statistical mechanics calculations using the material structure determined by molecular dynamics calculations, and a step of converting the physical property values ​​into continuous functions based on the discrete physical property values. This makes it possible to generate physical property value change data including physical property values ​​converted into continuous functions.

[0071] The computer system 1 may have a program capable of presenting (outputting) the physical property change data obtained in the physical property change data generation step S3. More specifically, by inputting the name or chemical formula of a substance and its constituent ratio into the software, molecular feature amounts may be calculated by quantum chemical calculation based on the input name or chemical formula and constituent ratio, physical property values ​​required for thermal fluid calculation may be calculated from the constituent ratio and molecular feature amounts by molecular dynamics calculation and statistical mechanics calculation, a library or images of continuous data representing changes in the calculated physical property values ​​may be generated, and a thermal fluid cycle simulation, a computational fluid dynamics simulation, or a mechatronics system simulation may be performed using the generated library or images.

[0072] FIG. 12 is an explanatory diagram of an example of a presentation screen 100 for presenting physical property value change data.

[0073] The presentation screen 100 includes an input area 110 and an output area 120 .

[0074] The input area 110 is used to input information about a substance whose physical properties are to be calculated. The input area 110 includes a first area 111 and a second area 112.

[0075] The first area 111 is an area for inputting the chemical formula of each of one or more molecules that make up the substance. The first area 111 allows input of any chemical formula. In FIG. 12, the chemical formula C represents ethane as molecule 1 that makes up the substance. ２ H ６ is input, and the chemical formula CH ４ is input. Note that a name may be input in place of a chemical formula in the first area 111. In this case, a table linking the name and the chemical formula is prepared in advance.

[0076] The second area 112 is an area for inputting the composition ratio of one or more molecules. The second area 112 allows input of any numerical value. In Fig. 12, 50% is input as the composition ratio of molecule 1 and the composition ratio of molecule 2.

[0077] The output area 120 includes an explanation area 121 and a presentation area 122 .

[0078] The explanation area 121 displays a character string indicating the name of the physical property value. In Fig. 12, thermal conductivity and viscosity are displayed as the physical property values.

[0079] The presentation area 122 presents images G1 and G2 showing continuous data of physical properties calculated based on the information of the substance input in the input area 110. In FIG. 12, image G1 shows a continuous function of thermal conductivity, and image G2 shows a continuous function of viscosity. Note that images G1 and G2 may also show graphs as continuous data of the corresponding physical properties. Images G1 and G2 may be either still images or moving images.

[0080] In this way, the computer system 1 may have a program that executes a presentation method for thermal fluid calculations. This presentation method accepts input of the chemical formulas of one or more molecules that make up a substance and the constituent ratios of the one or more molecules, generates an image showing continuous data for each physical property value required for thermal fluid calculations for the input chemical formulas and constituent ratios, and presents the generated image.

[0081] In the geometric shape condition setting step S4, geometric shape conditions are set. The geometric shape conditions are used to predict physical changes in the physical change prediction step S5. The geometric shape may be one-dimensional, two-dimensional, or three-dimensional. The geometric shape may be any shape, such as a cylinder, a sphere, an n-gonal prism such as a square prism, a pentagonal prism, or a hexagonal prism, a circle, an n-gon such as a square, a pentagon, or a hexagon, or a combination thereof. Furthermore, within the geometric shape, a wall that blocks the penetration or flow rate of a component, a permeable membrane that allows passage in only one direction, an inflow portion through which a component can flow, or an outflow portion through which a component is discharged to the outside of the system can be arbitrarily arranged.

[0082] In the physical change prediction step S5, boundary conditions are applied to the geometric shape in order to use the Navier-Stokes equation, continuity equation, heat conduction equation, and diffusion equation (described later). When considering flow velocity or pressure, for example, a condition for the inflow velocity is applied to the inflow section. For the outflow section, the pressure at the outflow section is determined. A condition that the velocity component is zero is applied to wall surfaces other than the inflow and outflow sections of the geometric shape. For a permeation membrane, a condition that the velocity component in the permeation membrane direction is zero or greater is applied. When considering temperature, for the heat inflow section, the heat quantity is applied by multiplying the density, the temperature difference at the inflow section, and the constant-pressure specific heat by the inner product of the normal component and the inflow velocity. In this case, the physical property values ​​obtained in the physical property change data generation step S3 can be used for the constant-pressure specific heat and density. Furthermore, for the inflow section of temperature itself, rather than heat, the temperature at the inflow section is set to the desired temperature. For the heat outflow section, the temperature gradient with respect to the perpendicular component at the outflow section is set to zero. Wall surfaces other than the inlet and outlet sections of the geometric shape are set as thermal insulating sections, and more specifically, the temperature gradient component in the normal direction to the wall surfaces is set to 0. For the permeation membrane, a condition is imposed that the temperature gradient in the permeation membrane direction is 0 or more. Finally, when considering the concentration of a component, the component concentration is set to the desired concentration in the inlet section. For the outlet section, the dot product of the component in the perpendicular direction to the outlet section and the concentration gradient is set to 0. For wall surfaces other than the inlet and outlet sections of the geometric shape, the concentration gradient in the component perpendicular to the wall surfaces is set to 0. For the permeation membrane, a condition is imposed that the concentration gradient in the permeation membrane direction is 0 or more. In this way, it is possible to set the geometric shape, inlet section, outlet section, wall surfaces, permeation membrane, etc. in advance.

[0083] The physical change prediction step S5 predicts a physical change of the substance using the physical property value change data generated in the physical property value change data generation step S3.

[0084] Predicting physical changes includes at least one of calculating changes in physical properties for use in thermal cycle simulation, computational fluid dynamics, mechatronic system simulation, or thermal fluid analysis of a geometric shape, calculating changes in physical properties for design optimization of a thermal fluid cycle simulation or computational fluid dynamics or mechatronic system simulation of a geometric shape, or calculating changes in physical properties to generate physical properties for use in a thermal fluid cycle simulation or computational fluid dynamics or mechatronic system simulation of a geometric shape.

[0085] In the physical change prediction step S5, in the thermal fluid calculation or thermal cycle simulation, in order to obtain outputs such as pressure distribution, velocity distribution, temperature distribution, and substance concentration distribution, the input of the physical property values ​​(physical property change data) calculated in the physical property change data generation step S3 and the input of the geometric shape conditions set in the geometric shape condition setting step S4 are used.

[0086] In the case of computational fluid dynamics calculations, the material is assumed to be a fluid flowing within a geometric shape, and the Navier-Stokes equations and the continuity equation, which describe the motion of the velocity or pressure distribution of the fluid, can be used. In the case of a compressible fluid, the Navier-Stokes equations and the continuity equation are expressed as the following equations (5) and (6), respectively.

[0087]

[0088]

[0089] In equations (5) and (6), v is the fluid velocity, ρ is the fluid density, μ is the fluid dynamic viscosity, χ is the bulk viscosity, and g is the gravitational field acting per mass of the fluid. Here, the fluid density ρ and dynamic viscosity μ can use the values ​​calculated up to the physical property change data generation step S3. Since the bulk viscosity χ is generally a very small value, it is set to 0. Furthermore, for example, in the case of an incompressible fluid whose density does not change, it can be easily calculated by setting ∂ρ / ∂t = 0 in equation (6). Therefore, the information processing method of this embodiment can be applied not only to compressible fluids but also to incompressible fluids. By using the Navier-Stokes equation and the continuity equation, two variables, the velocity v and pressure p of the fluid of a material, can be output, making it possible to predict the fluid motion of the material.

[0090] Regarding the thermal fluid analysis, the energy conservation equation can be used for the temperature change. The energy conservation equation is expressed by the following equation (7).

[0091]

[0092] In equation (7), T is the temperature, ρ is the density, c ｐ is the constant pressure heat capacity, and can also be used by adding the gas constant R to the constant volume heat capacity. Furthermore, k is the thermal conductivity. Q indicates the amount of heat per unit time.

[0093] For example, the reaction enthalpy h of a substance produced by a chemical reaction rea and latent heat Δh lat In particular, when the calculation system assumes adiabatic changes that cause rapid chemical reactions or state changes, the specific enthalpy that occurs is Δh, which is the result used in the NPE calculation of molecular dynamics and statistical mechanics calculations. adi Using this, the amount of heat Q per unit time can be expressed as the following equation (8).

[0094]

[0095] However, the latent heat Δh lat is the amount of heat generated only when passing through the boiling point or dew point determined in the step S3 of determining the structure of the substance and generating the data of changes in physical properties.

[0096] Assume that substances A, B, C, and D undergo a chemical reaction such as that shown in formula (9).

[0097]

[0098] In this case, for example, for substances A, B, C, and D, the specific enthalpy is h Ａ , h Ｂ , h Ｃ , h Ｄ and Gibbs free energy G Ａ , G Ｂ , G Ｃ , G Ｄ The reaction Gibbs free energy generated by the chemical reaction of equation (9) is G rea Then, G rea can be written as in equation (10).

[0099]

[0100] Reaction Gibbs free energy G rea When the value of is negative, the reaction proceeds spontaneously. Therefore, the reaction enthalpy (heat of reaction) generated by a chemical reaction is called h rea Then, h rea can be calculated using equation (11).

[0101]

[0102] The reaction enthalpy thus obtained h rea into equation (8) to find the amount of heat Q per unit time, and by incorporating this Q into equation (7), the change in temperature distribution caused by the heat of reaction can be calculated.

[0103] As described above, by using the density, constant pressure (constant volume) heat capacity, thermal conductivity, specific enthalpy, latent heat, and specific enthalpy generated by adiabatic compression obtained by molecular dynamics calculation and statistical mechanics calculation, and the flow velocity calculated from the Navier-Stokes equation and the continuity equation, equation (7) can be solved to output the temperature T due to heat transfer in the material, making it possible to predict the temperature distribution of the material. Therefore, equation (7) can be solved, and the temperature distribution of the material over time and space can be calculated.

[0104] To determine the concentration distribution of a substance, a diffusion equation is used, which is expressed by the following equation (12).

[0105]

[0106] In equation (12), ρ is the density of the substance, c is the concentration of the substance, t is time, v is velocity, D is the diffusion coefficient, M ｎ is the molecular weight of the substance, R rea is the amount of material generated or reduced by a reaction, etc. The density ρ and diffusion coefficient D are calculated using the calculation results up to step S3 of determining the structure of the material and generating data on changes in physical properties. The velocity v is calculated using the results of the Navier-Stokes equation and the equation of continuity. M ｎ can be automatically determined if the chemical formula of the substance acquired in the acquisition step S1 is given. Therefore, it is possible to solve the above equation (12), and calculate the concentration distribution of the substance with respect to time and space.

[0107] The prediction of physical changes is used in simulation of thermal fluid cycles, computational fluid dynamics, or mechatronic system simulations by inputting geometric shapes.

[0108] The physical change output step S6 outputs the physical change of the substance derived based on the physical property values. More specifically, the physical change output step S6 outputs the physical change predicted in the physical change prediction step S5.

[0109] For example, physical changes can be steady states or unsteady states, including time-series changes in non-equilibrium phenomena such as fluid flow, heat, and mass transport. In this case, outputting physical changes includes visualizing, numerically or graphically, temperature distribution, pressure distribution, flow velocity distribution, concentration distribution, and different physical property values ​​used in the calculation, including non-equilibrium processes that change over time. Techniques for graphically visualizing such changes in physical properties include images such as videos and still images. Videos or still images showing changes in physical properties can also be used to output parameters required for models such as nucleation, nucleation growth, or film thickness of a phase change region that satisfy the changes in physical properties shown in the video or still image, or to obtain physical properties using autocorrelation functions. Dynamic visualizations such as videos do not necessarily need to include everything from the initial conditions to the equilibrium state.

[0110] For example, physical changes include changes in the physical structure of a substance. In this case, outputting a physical change includes visualizing different states of the physical structure of a substance. Visualizing the phase change or molecular structure of a substance makes it easier to understand the physical structure of a substance. Phase changes include liquid bubbles in a gas phase, a solid phase in a liquid phase, or a liquid phase in a solid phase. A molecular structure is not a single molecule, but a structure in which multiple molecules gather and loosely interact to form a kind of macromolecule-like shape. Examples include aggregates of polymers such as proteins or surfactants. Quasi-static and dynamic changes such as the diffusion of such aggregates or the aggregation or dispersal of molecules may be output as videos, etc. Furthermore, parameters necessary for dispersal and aggregation may be extracted from images of the physical structure of a substance, or the diffusion rate may be quantified using image recognition technology, etc. The numerical values ​​of the parameters obtained in this way may also be converted into continuous functions.

[0111] [1.2 Application Examples] The information processing method according to the present embodiment described above can be used to visualize the thermal process in a heat exchanger that uses a substance as a working medium.

[0112] First, information on the substance to be used as the working medium is acquired (acquisition step S1). The information on the substance includes the chemical formula of each of one or more molecules that make up the substance and the composition ratio of the one or more molecules.

[0113] Next, molecular feature quantities of the substance are determined (molecular feature quantity determination step S2). Here, the molecular feature quantities include a molecular structure and a molecular force field (intramolecular force field and intermolecular force field) for each molecule. The molecular structure includes bond lengths, bond angles, and dihedral angles. The molecular force field includes force constants for the bond lengths, bond angles, or dihedral angles, RESP charges, and Lennard-Jones potential parameters.

[0114] Next, the physical structure of the substance is obtained by molecular dynamics calculation, and physical property values ​​are determined by statistical mechanics calculation (physical property value change data generation step S3). Note that the physical property values ​​include heat capacity, thermal conductivity, and specific enthalpy.

[0115] The heat capacity may be, for example, a constant pressure heat capacity (constant pressure molar heat capacity) or a constant volume heat capacity (constant volume molar heat capacity).

[0116] In determining the constant pressure heat capacity, molecular dynamics calculations can calculate the internal energy, which is the sum of the kinetic energy of molecules and the potentials within and between molecules, based on the calculated physical structure. Therefore, an NPT calculation is performed under the desired temperature and pressure conditions, and the (specific) enthalpy H is calculated in principle as the sum of the internal energy and the product of the applied pressure and temperature, and the constant pressure heat capacity is calculated using the following relational expression (16). In expression (16), C ｐ is the constant pressure heat capacity, H is the calculated enthalpy, and T is the temperature.

[0117]

[0118] A canonical ensemble is used to determine the constant volume heat capacity. Specifically, an NVT calculation is performed under certain temperature conditions, the temperature difference and the molar internal energy in the system being calculated are calculated, and the constant volume heat capacity is calculated using the following equation (14). In equation (14), C ｖ is the heat capacity at constant volume, E is the energy, and T is the temperature.

[0119]

[0120] To determine the thermal conductivity, an arbitrary NPT calculation is performed, and the ensemble average of the autocorrelation function of the heat flux in the system calculated according to the Green-Kubo formula from the difference between the heat quantity and temperature generated in the subspace within the cell is calculated, and the differential value with respect to the temperature difference is taken to determine the coefficient.

[0121] Next, the physical property values ​​of the heat capacity, the thermal conductivity, and the specific enthalpy are obtained (physical property change data generation step S3).

[0122] These physical property values ​​are used to predict physical changes in the substance (physical change prediction step S5). The lists of heat capacity, thermal conductivity, and specific enthalpy obtained up to the physical property change data generation step S3, as well as the physical change data converted into functions, are output to, for example, multiphysics simulation software, making it possible to reference the physical property values ​​required for calculations. In this case, the physical property values ​​may be entered directly into the software, or the required physical property values ​​may be read each time the software performs the necessary calculations. In this case, the physical change prediction of the substance mainly predicts physical properties, so for example, thermal fluid calculations may be used.

[0123] In a thermal fluid calculation process using a substance as a working medium, the time changes in temperature distribution, pressure distribution, flow velocity distribution, and substance concentration distribution can be obtained by using the kinematic viscosity coefficient, density, latent heat, (specific) enthalpy, thermal conductivity, heat capacity at constant volume, heat capacity at constant pressure, diffusion coefficient, vapor pressure, excess Gibbs free energy, solvation Gibbs free energy, and Gibbs free energy calculated up to the physical property change data generation step S3. Furthermore, the time changes in temperature distribution, pressure distribution, flow velocity distribution, and substance concentration distribution can be obtained from physical change data.

[0124] Finally, the predicted physical changes of the substance are output (physical change output step S6). As an example, it is possible to visualize the temperature distribution, flow velocity distribution, and time changes in inflow and outflow temperatures in a flow channel of any geometric shape.

[0125] In this way, by using the information processing method of the above embodiment, it is possible to provide guidelines for equipment design, necessary materials, or composition design through temperature, flow velocity, pressure distribution, etc. The information processing method of the above embodiment can be applied, for example, to elucidating the behavior of a car engine, verifying necessary physical properties, and elucidating or verifying necessary physical properties of a thermal cycle including piping, compressors, heat exchangers, etc. of an air conditioner or air purifier. Furthermore, by using the information processing method of the above embodiment, it is possible to provide physical property values ​​necessary for multiphysics simulation or mechatronics system simulation, etc., and to perform highly accurate simulation or numerical analysis by linking them to material properties in a first-principles manner without experiments.

[0126] [1.3 Example] Calculations were carried out to confirm the effects of the information processing method according to this embodiment, and comparisons were made.

[0127] [1.3.1 Example 1] In Example 1, the well-known refrigerant R1234yf was used. R1234yf is 2,3,3,3-tetrafluoropropene, and is a pure substance. The chemical formula of R1234yf is C ３ H ２ F ４ The composition ratio of R1234yf is C ３ H ２ F ４ is 100%. The quantum chemistry calculation software GAMESS was used to calculate the most stable structure of the molecule using the coupled cluster method, which is an ab initio molecular orbital method, and a correlation-consistent basis set with a polarization function added as the basis function, to determine the molecular length, molecular angle, and dihedral angle. Furthermore, the force constants for the molecular length, molecular angle, and dihedral angle were determined using the above formulas (1), (2), and (3). Table 1 below shows the bond length force constants [kJ / (mol nm )] for the double bond between carbon atoms (C=C), the single bond between carbon fluorine atoms (C-F), and the single bond between carbon atoms (C-C) determined in this manner. ２ ) ] is shown.

[0128]

[0129] In Table 1, the experimental values ​​are shown as reference examples. The experimental values ​​in the reference examples are values ​​obtained from the literature (Zeitschrift für Naturforschung B (ZNB), (Germany), 2014, Vol. 69b, pp. 379-387).

[0130] A comparison between Example 1 and the Reference Example in Table 1 confirmed that the calculated values ​​of molecular feature quantities obtained by the information processing method according to this embodiment were in good agreement with the experimental values. In particular, the calculated values ​​of molecular feature quantities obtained by the information processing method according to this embodiment had an error of within 5% from the experimental values.

[0131] In Table 1, for Comparative Example 1, the most stable molecular structure of R1234yf was calculated using the quantum chemistry calculation software GAMESS, using the HF method approximated by one Slater determinant, and the molecular length, molecular angle, and dihedral angle were determined. Charge was assigned using AM1-BCC. The force constants and Lennard-Jones potentials for the molecular length, molecular angle, and dihedral angle were defined using the Amber force field. A comparison of Comparative Example 1 and Comparative Example 2 reveals that the calculated values ​​differ significantly from the experimental values. This is thought to be due to the fact that, in the Amber force field, molecular force field parameters are defined for a specific type of molecule and are not specific to that molecule. In other words, in Example 1, which is the information processing method according to the present embodiment, a specific molecular force field is determined from the chemical formula of the molecule, whereas in Comparative Example 1, a default molecular force field preset for the type of molecule is used, rather than being specific to the molecule itself, and therefore the calculated values ​​differ significantly from the experimental values.

[0132] Furthermore, in Example 1, the RESP charge and the Lennard-Jones potential were determined as an intermolecular force field. The Lennard-Jones potential of the molecule was determined so as to satisfy the relationship between the distance and the energy value by performing energy calculations multiple times while changing the distance and angle between two different molecules.

[0133] Using the molecular feature quantities calculated from the first principles by quantum chemical calculations, physical property values ​​were calculated using molecular dynamics calculation software Gromacs. Regarding the bulk structure of the molecule, the number of molecules was set to 400, and after random structural arrangement using Pymol, stable structure calculations were performed. After annealing for 100 ps in the NVT calculation, NPT calculations were performed at a temperature of 273 K and a pressure of 3 MPa. The density [kg / m ３ ], specific enthalpy [kJ / kg], and specific heat at constant pressure [kJ / (kg·K)] are shown in Table 2 below.

[0134]

[0135] In Table 2, the reference values ​​are values ​​obtained from the Reference Fluid Thermodynamic and Transport Properties Database (REFPROP) provided by the National Institute of Standards and Technology (NIST). Since REFPROP is a database that compiles experimental results, the values ​​obtained from REFPROP are used as reference values ​​showing experimental values.

[0136] From a comparison between Example 1 and the Reference Example in Table 2, it was confirmed that the calculated values ​​of the physical properties obtained by the information processing method according to this embodiment were in good agreement with the experimental values.

[0137] In Table 2, Comparative Example 1 shows the density [kg / cm 3 ] obtained by the same method as in Example 1 using the molecular feature amount obtained using the Amber force field as described above. ３ ], specific enthalpy [kJ / kg], and specific heat at constant pressure [kJ / (kg·K)]. A comparison between the Reference Example and Comparative Example 1 reveals that the calculated values ​​are significantly different from the experimental values. This is thought to be because the calculated values ​​of the molecular feature quantities in Comparative Example 1 are significantly different from the experimental values.

[0138] Next, in Example 1, using the bulk structure obtained based on the NPT calculation, an NVE calculation was performed for 3 ns at the same volume, and the dynamic viscosity and thermal conductivity were calculated from the autocorrelation function obtained in the last 1 ns. In addition, the diffusion coefficient was calculated using Gromacs from the autocorrelation function using the bulk structure obtained by the NPT calculation at a temperature of 303.2 K and a pressure of 1 atm. The dynamic viscosity [cm ２ / s], thermal conductivity [mW / (m·K)], and diffusion coefficient [ / cm] are shown in Table 3 below.

[0139]

[0140] In Table 3, the dynamic viscosity coefficient and thermal conductivity of the reference example are values ​​obtained from REFPROP. The experimental value of the thermal diffusion coefficient of the reference example is a value obtained from the literature (Netsu Bussei, 2020, Vol. 34, Issue 1, pp. 10-14).

[0141] A comparison of Example 1 and the Reference Example in Table 3 confirmed that the calculated values ​​of the physical properties obtained by the information processing method according to this embodiment were in good agreement with the experimental values. In particular, the calculated values ​​of the physical properties obtained by the information processing method according to this embodiment had an error of about 10% from the experimental values.

[0142] In Table 3, Comparative Example 1 shows the dynamic viscosity coefficient and thermal conductivity obtained by the same method as Example 1 using the molecular feature obtained using the Amber force field as described above. A comparison between the Reference Example and Comparative Example 1 shows that the calculated values ​​are significantly different from the experimental values. This is thought to be because the calculated values ​​of the molecular feature of Comparative Example 1 are significantly different from the experimental values.

[0143] In Table 3, Comparative Example 2 corresponds to the simulation of GEMC described in Non-Patent Document 1, but the dynamic viscosity coefficient, thermal conductivity, and diffusion coefficient were not theoretically calculable, and therefore are indicated as "ND."

[0144] In Example 1, the vapor pressure was calculated. Specifically, an NPT calculation was performed at a temperature of 303 K and a pressure of 0.1 MPa, and then the temperature was fixed and the density was calculated at 100 kg / m ３ The pressure intervals at which the pressure changes are calculated are then calculated in more detail for the pressure intervals obtained, and the density is calculated as 100 kg / m ３ The pressure that changed significantly was calculated as the vapor pressure. The vapor pressure [MPa] obtained by this calculation, the latent heat [kJ / kg] obtained by NVT calculation while changing the temperature so that the vapor pressure became 1 atm, and the liquid density [kg / m ３ ] and gas density [kg / m ３ ] are shown in Table 4.

[0145]

[0146] In Table 4, the reference values ​​are those obtained from REFPROP.

[0147] From a comparison between Example 1 and the Reference Example in Table 4, it was confirmed that the calculated values ​​of the physical properties obtained by the information processing method according to this embodiment were in good agreement with the experimental values.

[0148] In Table 4, Comparative Example 2 is a value obtained by a GEMC simulation described in Non-Patent Document 1. A comparison of the Reference Example and Comparative Example 2 reveals that the calculated values ​​differ significantly from the experimental values. This is thought to be because the calculated values ​​of the molecular feature quantities in Comparative Example 2 differ significantly from the experimental values. Furthermore, in Example 1, the excess Gibbs free energy and Gibbs free energy were calculated. Specifically, NPT calculations were performed for 500 molecules and 501 molecules for 2 ns at a temperature of 243 K and a pressure of 2 MPa, and an additional 10 ps calculation was performed as the main calculation. Furthermore, an NVT calculation was performed for 1 molecule in a cell with the same volume as the 500 molecules for 100 ps as the main calculation. The excess Gibbs free energy was calculated using ER calculations using three different calculation systems. The excess Gibbs free energy [kJ / kg] and Gibbs free energy [kJ / kg] obtained at this time are shown in Table 5.

[0149]

[0150] In Table 5, the excess Gibbs free energy and Gibbs free energy of the Reference Example are values ​​obtained from REFPROP. A comparison of the Reference Example and Comparative Example 1 shows that the calculated values ​​are significantly different from the experimental values. This is thought to be because the calculated values ​​of the molecular feature amounts of Comparative Example 1 are significantly different from the experimental values. In comparison, the excess Gibbs free energy and Gibbs free energy of Example 1 are quite close to those of Reference Example 1.

[0151] A thermal fluid calculation was performed on the pipe for Example 1 obtained as described above. The material of Example 1 at 200 K was flowed through a pipe with a radius of 0.5 m from the left at a flow rate of 1.0 m / s. The material was heated to 400 K from 1.5 m to 2.5 m from the left end, and latent heat was generated at the boiling point (dew point) at the vapor pressure calculated in Table 4, causing vaporization. The liquid density and gas density calculated in Table 4 were used as the density at this time. The kinematic viscosity, thermal conductivity, and diffusion coefficient were calculated in the same way as in Table 3, and the constant pressure specific heat and specific enthalpy were calculated in the same way as in Table 2 to determine the required temperature and pressure.

[0152] Next, the results of thermal fluid calculations for a multiphase flow of gas and liquid components are shown. FIG. 13 is an explanatory diagram of gas distribution in the thermal fluid calculation of Example 1. FIG. 14 is an explanatory diagram of flow velocity distribution in the thermal fluid calculation of Example 1. FIG. 15 is an explanatory diagram of pressure distribution in the thermal fluid calculation of Example 1. FIG. 16 is an explanatory diagram of temperature distribution in the thermal fluid calculation of Example 1. All of FIGS. 13 to 16 show distributions 5 seconds after the start of calculation. By utilizing physical property values ​​obtained by the information processing method according to this embodiment, physical property values ​​can be predicted without experiments, and physical changes such as those from thermal fluid calculations can be predicted.

[0153] [1.3.2 Example 2] In Example 2, R454B was used as the substance. R454B is a mixture because it contains multiple types of molecules. More specifically, R454B is a mixed refrigerant that contains R1234yf and R32. R32 is difluoromethane, and its chemical formula is CH ２ F ２ The molecular feature amount was determined in the same manner as in Example 1 for R1234yf.

[0154] In R454B, the composition ratio of R1234yf to R32 was set to 68.1:31.9 by mass. The number of R454B molecules was determined by this composition ratio, and after random structural arrangement using Pymol, stable structure calculations were performed. After 100 ps annealing in the NVT calculation, NPT calculations were performed at a temperature of 273 K and a pressure of 3 MPa. The density [kg / m ３ ], specific enthalpy [kJ / kg], and specific heat at constant pressure [kJ / (kg·K)] are shown in Table 6 below.

[0155]

[0156] In Table 6, the reference values ​​are those obtained from REFPROP.

[0157] From a comparison between Example 2 and the Reference Example in Table 6, it was confirmed that the calculated values ​​of the physical properties obtained by the information processing method according to this embodiment were in good agreement with the experimental values.

[0158] In Example 2, the solvation Gibbs free energy and Gibbs free energy were determined. Specifically, at a temperature of 243 K and a pressure of 2 MPa, an NPT calculation was performed for 500 molecules of 2 ns and 501 molecules with one molecule of R1234yf added, followed by an additional 10 ps calculation as the main calculation. Furthermore, an NVT calculation was performed for 1 molecule in a cell with the same volume as the 500 molecules for 100 ps as the main calculation. Using three types of calculation systems, the excess Gibbs free energy was calculated using ER calculations. The solvation Gibbs free energy [kJ / kg] and Gibbs free energy [kJ / kg] at this time are shown in Table 7.

[0159]

[0160] In Table 7, the solvation Gibbs free energy and Gibbs free energy of Reference Example are shown as values ​​obtained from REFPROP. The solvation Gibbs free energy and Gibbs free energy of R454B in Example 2 are judged to be quite close to those in Reference Example 1.

[0161] [1.4 Effects, etc.] The information processing method described above is an information processing method executed by a computer system 1, which acquires the chemical formula of each of one or more types of molecules constituting a substance and the composition ratio of the one or more types of molecules (S1), calculates molecular feature amounts for each of the one or more types of molecules from the chemical formula of each of the one or more types of molecules (S2), and generates physical property value change data representing changes in the physical property values ​​of the substance using the composition ratios and molecular feature amounts (S3). This configuration makes it possible to calculate the physical property values ​​of the molecules constituting the substance without experimental values. Furthermore, it makes it possible to output the calculated physical property values.

[0162] The information processing method further includes executing a simulation using the physical property change data as an input based on the geometric shape condition setting (S4), and predicting a physical change of the material (S5). This configuration enables prediction of a physical change of the material.

[0163] The information processing method further includes issuing an instruction to execute a simulation using the physical property value change data as input (S5), and outputting predicted data of the physical change of the substance calculated in the simulation (S6). This configuration enables prediction and output of the physical change of the substance.

[0164] In the information processing method, calculating molecular feature quantities includes performing quantum chemical calculations on the chemical formulas of one or more types of molecules and determining the molecular feature quantities from first principles. This configuration makes it possible to obtain changes in the physical properties of a substance through calculation alone, without experimental values, for the molecules that make up the substance.

[0165] In the information processing method, the calculation of molecular feature quantities is determined by determining the most stable point between two or more molecules through stable structure calculations in which the structure is fixed in quantum chemical calculations for the van der Waals potential parameters of the molecular feature quantities. This configuration makes it possible to obtain changes in the physical properties of a substance through calculation alone, without experimental values, for the molecules that make up the substance.

[0166] In the information processing method, generating the physical property change data includes calculating a physical structure that indicates the state formed by the substance by molecular dynamics calculation using the constituent ratio and molecular feature amount. This configuration makes it possible to obtain changes in the physical property values ​​of the molecules that make up the substance through calculation alone, without experimental values.

[0167] In the information processing method, the molecular dynamics calculation is performed by performing at least one of NVT calculation, NPT calculation, NPE calculation, BAR calculation, and ER calculation. This configuration makes it possible to obtain changes in the physical properties of molecules constituting a substance through calculation alone, without experimental values.

[0168] In the information processing method, generating the physical property change data further includes calculating the physical property by statistical mechanics calculation based on the physical structure. This configuration makes it possible to obtain changes in the physical property of a substance through calculation alone, without experimental values ​​for the molecules that make up the substance.

[0169] In the information processing method, statistical mechanics calculations are determined by ensemble averages and / or autocorrelation functions based on the Green-Kubo formula. This configuration makes it possible to obtain changes in the physical properties of a substance through calculations alone, without experimental data for the molecules that make up the substance.

[0170] In the information processing method, generating physical property change data involves calculating a physical structure for each of a total of N+1 molecules, including one target molecule, N molecules other than the target molecule, and the target molecule and the molecules other than the target molecule, by molecular dynamics calculation, and using the physical structures to calculate the energy relaxed when the one molecule is added to the N molecules by subtracting the energy calculated when the one target molecule is added to the N molecules from the energy calculated from the physical structures of the N+1 molecules. This configuration makes it possible to calculate the excess Gibbs free energy.

[0171] In the information processing method, the physical property values ​​include at least one of the following: dynamic viscosity coefficient, density, latent heat, (specific) enthalpy, thermal conductivity, heat capacity at constant volume, heat capacity at constant pressure, diffusion coefficient, vapor pressure, excess Gibbs free energy, solvation Gibbs free energy, and Gibbs free energy. This configuration makes it possible to calculate the physical property values ​​of a substance without experimental values ​​for the molecules that make up the substance.

[0172] In the information processing method, the physical change data representing changes in the physical property value includes a continuous function for the physical property value generated based on a plurality of discrete values ​​of the physical state of the substance. This configuration makes it possible to output physical property values ​​corresponding to various inputs.

[0173] The above-described presentation method is executed by a computer system 1, which accepts input of the chemical formulas of one or more molecules constituting a substance and the composition ratios of the one or more molecules, generates images G1 and G2 showing continuous data for each physical property value required for thermal fluid calculations for the input chemical formulas and composition ratios, and presents the generated images G1 and G2. This configuration makes it easy to grasp changes in the physical property values ​​of the substance. This configuration also makes it possible to transfer the obtained continuous data for each physical property value to other programs or analytical methods for use.

[0174] The above-described information processing method can be realized by causing the computer system 1 to execute a program. This program is a program for causing the computer system 1 to execute the above-described information processing method. This configuration makes it possible to calculate the physical property values ​​of a substance without experimental values ​​for the molecules that constitute the substance.

[0175] The above-described manufacturing method is a manufacturing method of physical property value change data of a substance, executed by a computer system 1, which obtains the chemical formula of each of one or more molecules constituting the substance and the constituent ratio of the one or more molecules, determines molecular feature amounts for each of the one or more molecules from the chemical formula of each of the one or more molecules, and generates physical property value change data representing changes in the physical property values ​​of the substance using the constituent ratios and molecular feature amounts. This configuration makes it possible to calculate the physical property values ​​of the substance without experimental values ​​for the molecules constituting the substance.

[0176] 17 is a block diagram of a computer system 1A according to a modified example. The computer system 1A may be realized by a plurality of computers including a first computer 1a and a second computer 1b.

[0177] Each of the first computer 1a and the second computer 1b can be realized by a server, a personal computer (desktop computer, laptop computer), a mobile terminal (smartphone, tablet terminal, etc.), or the like.

[0178] The first and second computers 1a and 1b each include an input device 11a, 11b, an output device 12a, 12b, a communication device 13a, 13b, a storage device 14a, 14b, and an arithmetic circuit 15a, 15b. The input devices 11a, 11b, the output devices 12a, 12b, the communication device 13a, 13b, the storage device 14a, 14b, and the arithmetic circuit 15a, 15b are similar to the input device 11, the output device 12, the communication device 13, the storage device 14, and the arithmetic circuit 15, respectively.

[0179] In this modified example, the arithmetic circuits 15a and 15b share the tasks of performing the acquisition step S1, the molecular feature determination step S2, the physical property change data generation step S3, the geometric shape condition setting step S4, the physical change prediction step S5, and the physical change output step S6.

[0180] As an example, the arithmetic circuit 15a executes an acquisition step S1, a molecular feature determination step S2, and a physical property change data generation step S3, and the arithmetic circuit 15b executes a geometric shape condition setting step S4, a physical change prediction step S5, and a physical change output step S6.

[0181] In this case, the first computer 1a executes the acquisition step S1, the molecular feature determination step S2, and the physical property change data generation step S3 using the arithmetic circuit 15a to obtain physical property change data. The first computer 1a transmits the obtained physical property change data to the second computer 1b via the communication device 13a. The second computer 1b receives the physical property change data from the first computer 1a via the communication device 13b. In the second computer 1b, the arithmetic circuit 15b uses the physical property change data to execute the geometric shape condition setting step S4, the physical change prediction step S5, and the physical change output step S6.

[0182] In this example, the first computer 1a is responsible for calculating the physical property values, and the second computer 1b is responsible for the simulation using the physical property values. Note that the second computer 1b can be replaced by a computer that runs conventionally known simulation software.

[0183] As is clear from the above embodiment and modifications, the computer system 1 may be realized by a single computer or by multiple computers (capable of communicating with each other). It is not necessary for the multiple functions (components) of the computer system 1 to be concentrated in a single housing; the components of the computer system 1 may be distributed across multiple housings. Furthermore, at least some of the functions of the computer system 1, for example, some of the functions of the arithmetic circuit 15, may be realized by the cloud (cloud computing) or the like.

[0184] It should be noted that the computer system 1A does not necessarily have to include the second computer 1b.

[0185] [3. Aspects] As is clear from the above-described embodiment and modifications, the present disclosure includes the following aspects.

[0186] [Aspect 1] An information processing method executed by a computer system, comprising: acquiring a chemical formula of each of one or more types of molecules constituting a substance and a composition ratio of the one or more types of molecules; calculating a molecular feature amount of each of the one or more types of molecules from the chemical formula of each of the one or more types of molecules; and generating physical property value change data representing a change in a physical property value of the substance using the composition ratio and the molecular feature amount.

[0187] [Aspect 2] The information processing method of Aspect 1, further comprising: executing a simulation using the physical property value change data as an input based on the setting of geometric shape conditions, and predicting a physical change in the substance.

[0188] [Aspect 3] The information processing method of aspect 1 or 2, further comprising issuing an instruction to execute a simulation using the physical property value change data as an input, and outputting predicted data of the physical change of the substance calculated in the simulation.

[0189] [Aspect 4] The information processing method according to any one of Aspects 1 to 3, wherein calculating the molecular feature amount includes performing quantum chemical calculations on the chemical formulas of each of the one or more types of molecules and determining the molecular feature amount from first principles.

[0190] [Aspect 5] The information processing method according to any one of Aspects 1 to 4, wherein the calculation of the molecular feature amount is determined by finding the most stable point between two or more molecules by stable structure calculation in which the structure is fixed in quantum chemical calculation, for van der Waals potential parameters in the molecular feature amount.

[0191] [Aspect 6] The information processing method according to any one of Aspects 1 to 5, wherein generating the physical property value change data includes calculating a physical structure representing a state formed by the substance by a molecular dynamics calculation using the composition ratio and the molecular feature amount.

[0192] [Aspect 7] The information processing method of aspect 6, wherein the molecular dynamics calculation is performed by performing at least one of an NVT calculation, an NPT calculation, an NPE calculation, a BAR calculation, and an ER calculation.

[0193] [Aspect 8] The information processing method according to aspect 6 or 7, wherein generating the physical property value change data further includes calculating the physical property value by statistical mechanics calculation based on the physical structure.

[0194] [Aspect 9] The information processing method according to Aspect 8, wherein the statistical mechanics calculation is determined by an autocorrelation function based on an ensemble average and / or the Green-Kubo rule.

[0195] [Aspect 10] The information processing method according to any one of Aspects 1 to 9, wherein generating the physical property change data includes calculating a physical structure by molecular dynamics calculation for one target molecule, N molecules other than the target molecule, and a total of N+1 molecules including the target molecule and the molecules other than the target molecule, and using the physical structures to calculate energy relaxed when the one molecule is added to the N molecules by subtracting energy calculated when the one target molecule is added to the N molecules from energy calculated from the physical structures of the N+1 molecules.

[0196] [Aspect 11] The information processing method according to any one of Aspects 1 to 10, wherein the physical property values ​​include at least one of a dynamic viscosity coefficient, a density, a latent heat, a (specific) enthalpy, a thermal conductivity, a heat capacity at constant volume, a heat capacity at constant pressure, a diffusion coefficient, a vapor pressure, an excess Gibbs free energy, a Gibbs free energy of solvation, and a Gibbs free energy.

[0197] [Aspect 12] The information processing method according to any one of Aspects 1 to 11, wherein the physical change data representing a change in the physical property value includes a continuous function for the physical property value generated based on a plurality of discrete values ​​of the physical state of the substance.

[0198] [Aspect 13] A presentation method for thermal fluid calculations executed by a computer system, comprising: accepting input of the chemical formulas of one or more molecules constituting a substance and the composition ratios of the one or more molecules; generating an image showing continuous data for each physical property value required for thermal fluid calculations for the input chemical formulas and composition ratios; and presenting the generated image.

[0199] [Aspect 14] A program for causing the computer system to execute the information processing method according to any one of aspects 1 to 12.

[0200] [Aspect 15] A method for producing data on changes in physical properties of a substance, executed by a computer system, comprising: acquiring a chemical formula for each of one or more molecules constituting the substance and a composition ratio of the one or more molecules; determining a molecular feature amount for each of the one or more molecules from the chemical formula for each of the one or more molecules; and generating physical property change data representing changes in physical properties of the substance using the composition ratio and the molecular feature amount.

[0201] [Aspect 21] A method for predicting a physical change in a substance, executed by a computer system, comprising: acquiring a chemical formula of each of one or more molecules constituting the substance and a composition ratio of the one or more molecules; determining a molecular feature amount for each of the one or more molecules from the chemical formula of each of the one or more molecules; determining a physical state of the substance including at least one of a structure of the substance and a thermophysical property value of the substance from the composition ratio and the molecular feature amount; generating physical change data representing a change in the physical state based on the physical state; and predicting the physical change using the physical change data.

[0202] [Aspect 22] The prediction method of Aspect 21, wherein the molecular feature amount is determined from the chemical formulas of the plurality of different molecules by quantum chemical calculations based on first principles.

[0203] [Aspect 23] The prediction method according to Aspect 2, wherein the molecular feature amount includes a molecular structure and a molecular force field.

[0204] [Aspect 24] The prediction method of Aspect 23, wherein the molecular structure is determined based on a most stable structure obtained from the chemical formula of each molecule by quantum chemical calculation, and the molecular force field is determined based on the molecular structure.

[0205] [Aspect 25] The prediction method of Aspect 24, wherein the molecular structure includes at least one of a bond length, a bond angle, or a dihedral angle, and the molecular force field includes an intramolecular force field, and the intramolecular force field includes a force constant for at least one of a bond length, a bond angle, or a dihedral angle.

[0206] [Aspect 26] The prediction method according to Aspect 25, wherein the error of the force constants of the intramolecular force field relative to force constants calculated from actual measurements is within 5%.

[0207] [Aspect 27] The prediction method according to any one of Aspects 24 to 26, wherein the molecular force field includes an intermolecular force field, and the intermolecular force field includes at least one of a molecular charge and a van der Waals potential.

[0208] [Aspect 28] The prediction method according to any one of Aspects 21 to 27, wherein the thermophysical property value includes at least one of a dynamic viscosity coefficient, a density, a latent heat, a (specific) enthalpy, a thermal conductivity, a heat capacity at constant volume, a heat capacity at constant pressure, a diffusion coefficient, a vapor pressure, and an entropy.

[0209] [Aspect 29] The prediction method according to any one of Aspects 21 to 28, wherein the physical change data includes a continuous function of the physical state of the substance generated based on a plurality of discrete values ​​of the physical state of the substance.

[0210] [Aspect 30] The prediction method of any one of Aspects 21 to 29, wherein predicting the physical change includes at least one of: calculating the change in the physical state for use in a simulation of a thermal fluid cycle in a geometric shape, computational fluid dynamics, or thermal fluid analysis; calculating the change in the physical state for design optimization of a simulation of a thermal fluid cycle in a geometric shape or computational fluid dynamics; or calculating the change in the physical state to generate thermophysical property data for use in a simulation of a thermal fluid cycle in a geometric shape or computational fluid dynamics.

[0211] [Aspect 31] A program for causing the computer system to execute the prediction method according to any one of aspects 21 to 30.

[0212] [Aspect 32] A method for producing physical change data of a mixture composed of a plurality of different molecules, executed by a computer system, comprising: acquiring chemical formulas of the plurality of different molecules and a mixing ratio of the plurality of molecules; determining molecular feature amounts of the plurality of different molecules from the chemical formulas of the plurality of different molecules; determining a physical state of the mixture, including at least one of a structure and a thermal property value of the mixture, from the molecular feature amounts and the mixing ratio; and generating physical change data representing a change in the physical state based on the physical state.

[0213] [Aspect 33] An information processing method executed by a computer system, comprising: acquiring a chemical formula of each of one or more molecules constituting a substance and a composition ratio of the one or more molecules; determining a physical state of the substance, including at least one of a structure of the substance and a thermophysical property value of the substance, based on the chemical formula and the composition ratio; and outputting a physical change of the substance, including at least one of a change in the structure and a change in the thermophysical property value, derived based on the physical state.

[0214] [Aspect 34] The information processing method of Aspect 33, wherein outputting the physical change includes generating an image showing at least one of the different states of the structure and the different values ​​of the thermophysical property, and outputting the generated image.

[0215] [Aspect 35] A program for causing the computer system to execute the information processing method of aspect 33 or 34.

[0216] Aspects 22 to 29 are optional elements and are not essential. Aspects 22 to 29 can be appropriately combined with Aspects 32, 33, and 34.

[0217] The present disclosure is applicable to an information processing method, a presentation method, a program, and a manufacturing method. Specifically, the present disclosure is applicable to an information processing method for calculating changes in physical property values ​​of a substance, a presentation method, a program for executing the information processing method on a computer system, and a manufacturing method for producing physical property value change data.

[0218] S1 Acquisition step S2 Molecular feature determination step S3 Physical property change data generation step

Claims

1. An information processing method executed on a computer system, comprising: acquiring a chemical formula of each of one or more types of molecules constituting a substance and a composition ratio of said one or more types of molecules; calculating a molecular feature value of each of said one or more types of molecules from the chemical formula of each of said one or more types of molecules; and generating physical property change data representing a change in a physical property value of said substance using said composition ratio and said molecular feature value.

2. The information processing method according to claim 1, further comprising: executing a simulation using the physical property value change data as an input based on the setting of geometric shape conditions, and predicting physical changes in the material.

3. The information processing method according to claim 1, further comprising: issuing an instruction to execute a simulation using the physical property change data as an input; and outputting predicted data of the physical change of the substance calculated in the simulation.

4. The information processing method according to claim 1, wherein calculating the molecular feature amount includes performing a quantum chemical calculation on the chemical formula of each of the one or more types of molecules and determining the molecular feature amount from first principles.

5. The information processing method according to claim 1, wherein the calculation of the molecular feature amount is determined by determining the most stable point between two or more molecules by stable structure calculation in which the structure is fixed in quantum chemical calculation, for van der Waals potential parameters in the molecular feature amount.

6. The information processing method according to claim 1, wherein generating the physical property change data includes calculating a physical structure indicating a state formed by the substance by molecular dynamics calculation using the composition ratio and the molecular feature amount.

7. The information processing method according to claim 6, wherein the molecular dynamics calculation is at least one of an NVT calculation, an NPT calculation, an NPE calculation, a BAR calculation, and an ER calculation.

8. The information processing method according to claim 6, wherein generating the physical property change data further comprises calculating the physical property by a statistical mechanics calculation based on the physical structure.

9. The information processing method according to claim 8, wherein the statistical mechanics calculation is determined by an ensemble average and / or an autocorrelation function based on the Green-Kubo rule.

10. The information processing method of claim 1, wherein generating the physical property change data comprises calculating a physical structure by molecular dynamics calculation for each of one target molecule, N molecules other than the target molecule, and a total of N+1 molecules including the target molecule and the molecules other than the target molecule, and using the physical structure to calculate the energy relaxed when one molecule is added to N molecules by subtracting the energy calculated when one target molecule is added to N molecules from the energy calculated from the physical structure of the N+1 molecules.

11. The information processing method according to claim 1, wherein the physical property values ​​include at least one of the dynamic viscosity coefficient, density, latent heat, (specific) enthalpy, thermal conductivity, heat capacity at constant volume, heat capacity at constant pressure, diffusion coefficient, vapor pressure, excess Gibbs free energy, solvation Gibbs free energy, and Gibbs free energy.

12. The information processing method according to claim 1, wherein the physical change data representing the change in the physical property value includes a continuous function for the physical property value generated based on a plurality of discrete values ​​of the physical state of the substance.

13. A presentation method for thermal fluid calculations executed by a computer system, comprising: receiving input of a chemical formula for each of one or more molecules constituting a substance and a composition ratio of said one or more molecules; generating an image showing continuous data for each physical property value required for thermal fluid calculation for the input chemical formula and composition ratio; and presenting the generated image.

14. A program for causing the computer system to execute any one of the information processing methods according to claims 1 to 12.

15. A method for producing data on changes in physical properties of a substance, executed by a computer system, comprising: acquiring a chemical formula for each of one or more molecules constituting the substance and a composition ratio of the one or more molecules; determining a molecular feature value for each of the one or more molecules from the chemical formula for each of the one or more molecules; and generating physical property change data representing changes in the physical properties of the substance using the composition ratio and the molecular feature value.

Images