KR102441072B1 - Simulation method for flowing implementation of fluid - Google Patents

Abstract

The present invention relates to a simulation method for realizing fluid flow in granular molecular dynamics simulation.

Description

Simulation method for flowing implementation of fluid

The present invention relates to a simulation method for realizing fluid flow in granular molecular dynamics simulation.

Molecular Dynamics simulation (hereinafter referred to as 'MD simulation') is a simulation technique used to predict/interpret the physical properties of materials used at the nanoscale (ig. 10 ^-6 second, 10 ^-9 meter). However, MD simulation has difficulties in expressing the flow of materials such as fluids due to time and space constraints as well as scale limitations.

Meanwhile, Coarse-Grained Molecular Dynamics simulation (hereinafter 'CGMD simulation') is a simulation technique used to predict/interpret the physical properties of materials used in the mesoscale. The CGMD simulation is to predict/interpret the physical properties of a material by assembling molecules of the material (Coarse-Graining) and applying a force field, and unlike all-atom MD simulation, a fluid-like material It can be used to predict/interpret the flow of

On the other hand, in order to generate a flow in the fluid in the pipe, a driving force must be present. As a method of generating flow at the molecular level, there is a case where a temporary force is applied to the left side (※ in the case of a pipe in which the fluid flows from left to right), which is the starting point of the flow by giving an impact to the fluid in a stationary state. It is difficult to represent the continuous velocity distribution or temperature distribution (when heat supply exists) of a Newtonian fluid.

Republic of Korea Patent Publication No. 2010-0029248

An object of the present invention is to provide a simulation method for realizing fluid flow in coarse-grained molecular dynamics simulation.

In order to solve the above problems, the present invention provides a simulation method for realizing a fluid flow, comprising: a) assembling a fluid (Coarse-Graining); b) selecting a force field to confirm the interaction of the assembled fluid; c) creating a simulation model region including a fluid inflow region and an analysis region for realizing the flow of the fluid; d) setting the model domain to a thermodynamic equilibrium state; e) divide the fluid inflow region into a box region, a random region between the box region and the analysis region, set the fluid in the box region to a solidified state, and apply a uniform force while moving the entire box region in the direction of the analysis region moving the fluid in the random area in the direction of the analysis area; f) performing an analysis of the displaced fluid in the analysis area; g) confirming that the flow of the fluid in the simulation model area in the fluid analysis step matches the velocity and temperature distribution of the fluid in scientific research literature; and h) observing the behavior and physical properties of the fluid in the simulation model area determined from steps a to g.

The present invention can efficiently move a fluid to an analysis area in a coarse-grained molecular dynamics (CGMD) simulation, and through this, the flow of a fluid (specifically, a Newtonian fluid) at the mesoscale Because it can be implemented, it can contribute to constructing a system that clearly predicts/interprets the influence of fluid composition/shape, etc. on flow and changes in physical properties.

1 is a flowchart illustrating a process of a simulation method for realizing a fluid flow according to an example of the present invention.
2 to 4 are reference diagrams for explaining an embodiment of the present invention.

Hereinafter, the present invention will be described.

The present invention is a method for realizing a fluid flow of a coarse-grained molecular dynamics simulation (hereinafter, referred to as 'CGMD simulation'), which will be described in detail with reference to FIG. 1 as follows.

Step a) Assembly of the fluid

First, the fluid is coarse-grained. The assembling may consist of selecting a fluid (step a-1)) and setting a bead of the selected fluid (step a-2)).

The selection of the fluid may be made based on the fluid required to implement the flow. Specifically, the fluid may be a Newtonian fluid, and more specifically, the fluid may include at least one selected from the group consisting of water, methanol, ethylene glycol, propylene glycol, glycerol, and nanoparticles. have.

The bead setting of the fluid may be made by bundling several atoms or molecules into one particle (clump).

According to the process of setting and assembling the beads of the fluid as described above, effective scale adjustment of time and space may be possible according to the purpose of CGMD simulation.

Step b) Selection of Force Field

A force field that confirms the interaction of the fluid assembled in step a) is selected. Specifically, the force field may be selected from the group consisting of Martini potential, Shinoda potential, Dissipative particle dynamics potential, and Lennard-Jones potential. have. More specifically, as the force field, a Lenard-Jones potential based on van der Waals interaction may be selected.

The Leonard-Jones potential is a function expressed using parameters of energy (epsilon, ε) and distance (sigma, σ) between two particles in equilibrium, that is, a function relating to the distance between particles.

When the Leonard-Jones potential is selected as the force field, the parameter that needs to be optimally set may be at least one selected from the group consisting of epsilon (ε) and sigma (σ).

Here, the initial value of the parameter may be set through a calculation process using a value disclosed in scientific research literature or the density of a fluid.

step c) setting the area of the simulation model

Setting the area of the simulation model (that is, configuring the system) specifically sets the area of the simulation model for realizing the flow of Newtonian fluid using the beads (clump particles) formed in step a, for this purpose More than millions of particles are required within the simulation, which requires a very large computer memory source (resource).

In the present invention, for an efficient system configuration, a simulation model is divided into a fluid inflow region and an analysis region, and the entire system is constructed by merging the two regions. Here, the analysis region may have a tubular shape.

When creating the fluid inflow region and the analysis region, a small unit system is first configured and then replicated repeatedly to complete the tube region and the fluid inflow region. In addition, a vacuum region may be added after the analysis region so that the fluid that has exited the analysis region is freed from a periodic system. Accordingly, when it is assumed that the fluid moves from left to right, the entire simulation model area may include a fluid inflow area, an analysis area, and a vacuum area from left to right.

The initial system modeling of each region configuration according to the present invention may be performed using the Materials Studio 2017R2 program.

Step d) setting the simulation model region to a thermodynamic equilibrium state

As a precondition for realizing the flow in the simulation only by the inflow of fluid without applying a separate force in the simulation, the LAMMPS program is used so that the region set in step c can reach the thermodynamic equilibrium state. Energy minimization and NVT simulation 1.4 ns (nano second) are sequentially performed for the fluid particles within the area set in step (here, N: number of beads, V: system volume, T: system temperature).

step e) moving the fluid in the direction of the analysis area

In order to implement a flow by introducing a fluid into the analysis area in the set simulation model, in the simulation model area configured in step c, the fluid injection area is again divided into a box area and a random area between the box area and the analysis area, , by setting the fluid in the box area to a solid state in the simulation and then setting to apply a uniform force while moving the entire box area in the direction of the analysis area to move the fluid in the random area in the direction of the analysis area.

step f) performing the analysis of the fluid

The velocity and temperature distribution are calculated by analyzing the flow of the fluid generated in the analysis area due to the inflow of the fluid formed in step e.

The flow in this analysis region is classified into an inlet region and a fully developed region, and in the fully developed region of the Newtonian fluid in which laminar flow is formed, the velocity distribution in the xy section of the particle (in the simulation, the vertical section of the analysis region in the simulation) is the following equation It can be judged by whether or not it conforms to 1.

[Equation 1]

V _r : fluid velocity at position r

V _avg : average fluid velocity at all r positions

r : position coordinates

R : radius of tube

On the other hand, when there is heat supply (Q) in the analysis region, the ideal heat distribution may be one in which the temperature in the analysis region and the relative temperature distribution at the r position appear uniformly as shown in Equation 2 below.

[Equation 2]

Ts : temperature within the analysis area

Tm : average temperature of the fluid

T(r,z) : temperature of (r,z) coordinate position

According to the present invention, it is possible to consider the case where there is heat supply in the analysis area, and logic judgment is made based on Newtonian fluid standards for both velocity and temperature distribution so that both cases can be realized/expressed with and without fluid particles. It may be desirable to At this time, particle behavior can be analyzed by performing the analysis at intervals of 10 Å and 5 Å, respectively, based on the analysis area (tube-shaped) radial distance and analysis area (tube length (z-axis direction)). .

Step g) Checking the velocity temperature distribution of the fluid

After that, it is determined whether an appropriate flow system has been implemented by verifying that the calculated velocity and temperature distributions agree with the theoretical values in scientific research literature (eg, whether it is a Newtonian fluid flow). If the velocity and temperature distribution do not follow the theoretical values in the scientific research literature, the simulation model can be reconstructed by returning to step c and increasing the size and length of the analysis area.

The above scientific research literature is a paper in which research on fluid flow through actual experiments or numerical analytical methods is recorded (for example, 'K. Kikuchi and O. Mochizuk, Meas. Sci. Technol., 2015, 26, 025301' and 'S. E. B. Maiga et al., Superlattices, which describe a study on the velocity and temperature distribution of ethylene glycol fluid containing nanoparticles through numerical analysis methods' Microstruct, 2004, 35, 543-547').

Step h) Observation of fluid behavior and properties

By going through the steps as described above, the simulation model area is confirmed, and the present invention can efficiently move the fluid to the analysis area in the tube in the coarse-grained molecular dynamics simulation, so that the fluid (Specifically, since the flow of Newtonian fluid can be implemented in mesoscale, it can contribute to constructing a system that clearly predicts/interprets the influence of fluid composition/shape, etc. on the flow and changes in physical properties.

Accordingly, by adding nanoparticles (beads) exhibiting the nano-effect of nanofluids, it is possible to confirm the mechanism of specific behavior and observe the effect on thermal properties.

Nanoparticles are also coarse-grained in consideration of the shape and size of nanofluid beads, and force field parameters that define interactions with fluids and particles (ig. iron beads) in the analysis area is set and added to the system, and the effect of nanomaterials on thermal properties can be confirmed by driving the flow system to which nanoparticles are added according to steps c to g described above. The heat transfer coefficient can be obtained by dividing the heat flux by the difference between the temperature in the analysis area and the average fluid temperature as shown in Equation 3 below.

[Equation 3]

h : heat transfer coefficient

q: heat flux

T _s : temperature within the analysis area

T _m : average temperature of the fluid

According to such a process, the present invention can implement the flow of a fluid so that the flow of the fluid can occur, such as an actual flow phenomenon, so that the present invention provides a change in physical properties ( For example, it can contribute to building a system (eg, a heat transfer system for a vehicle coolant) that predicts/interprets changes in temperature or speed, etc.). In addition, when a driving force in the form of introducing a fluid with an equal force into the analysis area is applied, the flow of the fluid (or nanoparticles) is continuously controlled by adjusting the length and size of the analysis area, which is the area where the fluid is introduced. It may be possible to construct and control an efficient system by preventing unnecessary calculations by establishing a simulation model that can observe and introduces only as much fluid as necessary.

Hereinafter, the present invention will be described in detail through examples. However, the following examples only illustrate the present invention, and the present invention is not limited by the following examples.

[Example 1]

a) step

Water and ethylene glycol were selected as the fluid objects, and four water molecules and one ethylene glycol molecule were set as one bead and assembled. That is, the water bead was set as one bead by tying the state quantity and interaction of the oxygen atom and the hydrogen atom into four molecules, and the state quantity of one bead was set to have the same physical properties as the four water molecules. Ethylene glycol beads were set to have the same physical properties as one ethylene glycol by binding the states and interactions of hydrogen, carbon, and oxygen atoms.

b) step

In order to confirm the interaction between the bead particles, the Leonard-Jones potential represented by Equation 4 below was selected as the force field.

[Equation 4]

(φ: Lennard-Jones Potential, ε: energy parameter, σ: van der Waals radius, r: distance between two particles)

In Equation 4, the initial value of epsilon (ε), which is one of the parameters, is 1.1 kcal/mol (a value that appears mainly in scientific research literature) for both water and ethylene glycol, and the initial value of the other, sigma (σ), is Based on the density of ethylene glycol, it was calculated as follows, and was set to 4.93 Å (water) and 4.53 Å (ethylene glycol), respectively.

* Density of water at room temperature (298K) (0.997 g/cc ≒ 1.0 g/cc): the volume of 4 water molecules

* Density of ethylene glycol at room temperature (298K) (1.11 g/cc): volume of one ethylene glycol molecule

The initial parameters between water and ethylene glycol were set using Equation 5 (Lorentz-Berthelot rule) below.

[Equation 5]

Based on the initial values of epsilon (ε) and sigma (σ) of water and ethylene glycol, the physical properties of water and ethylene glycol (i.e. density, self-diffusion coefficient, thermal conductivity, etc.) The parameters were adjusted.

The parameters of iron (tubular material) were set using the first optimal values of epsilon (ε) and sigma (σ) of water and ethylene glycol adjusted to each property value. The sigma (σ) value between water, ethylene glycol and iron was established by calculating and comparing the radial distribution function (RDF) in molecular dynamics simulations and granular molecular dynamics simulations. The epsilon (ε) value was determined as the first optimal value through contact angle comparison. As a result, epsilon (ε, kcal/mol)/sigma (σ, Å) values between water and iron and ethylene glycol and iron were set to 0.7/3.74 and 0.6/4.276, respectively.

c ) step

The area setting of the simulation model was performed using the Materials Studio 2017R2 program. Referring to FIG. 2( a ), the analysis area is composed of a fluid (blue), a cylindrical tube that applies heat (grey), and wall particles (orange). Wall particles are used to fill the empty space that occurs when there is a cylindrical tube in a cuboid-shaped system. As shown in FIG. 2 , the analysis region has a tube shape included in a rectangular parallelepiped with a diameter of 400 Å and a length of 800 Å, and the thickness of the tube is 30 Å. Fluid injection region It is a rectangular parallelepiped with a diameter of 400 Å and a length of 600 Å containing only the fluid (not including the wall particles). From this configuration, the simulation model had a size of 520 x 520 x 2200 Å ³ (total number of beads: 5,065,840).

As a complete system with a size of 520 x 520 x 2200 Å ³ , referring to FIG. The fluid was made visible only water (blue).

d) step

In order to bring the simulation model region formed in step c to the thermodynamic equilibrium state (equlilbrium state), energy minimization and NVT simulation 1.4 ns were sequentially performed only for the fluid particles included in the model region using the LAMMPS program. (N: number of bead particles, V: system volume, T: system temperature).

e ) step

Referring to FIG. 3 , in the simulation model region configured in step c, the fluid injection region is again a box region, and the box region and the analysis region in order to implement a flow by introducing a fluid into the analysis region in the simulation model. It was divided into random regions between (see Fig. 3 (a)).

After setting the fluid in the box area to a solidified state in the simulation, it is set to apply a uniform force while moving the entire box area in the direction of the analysis area to move the fluid in the random area in the direction of the analysis area (Fig. 3(b)). ) to (c)).

Specifically, a force of 0.0001 kcal/mol/Å was added to each particle in the right direction (z-axis direction) in the right direction (z-axis direction) for every 20 fs to the solidified particles in the box region in order to realize the flow in the tube for a time of about 12 ns. Accordingly, it can be confirmed that the particles in the box area gain driving force into the analysis area, and the particles in the random area between the box area and the analysis area flow into the analysis area, thereby realizing the flow of the fluid in the analysis area.

At this time, NVE simulation of fluid particles was performed while maintaining the tube temperature at 298 K for flow formation for 0-4 ns and maintaining the tube temperature at 325 K for 4-12 ns after the flow was formed in the tube (NVE). : number of bead particles, V: system volume, E: system energy). In order for the energy exchange between the tube particles and the fluid particles at a constant temperature to occur, the particles in the analysis region move. In order to prevent the fluid from penetrating into the particles in the analysis area when the fluid flow is implemented, a repulsive force with the fluid particles is set for the tube particles and wall particles (refer to FIG. 2 (b)) without vibration.

f) step

In the analysis area in which flow was implemented, the case of heat supply was considered, and the logic judgment was performed based on the Newtonian fluid standard for both velocity and temperature distribution. The particle behavior was analyzed by performing the analysis at intervals of 10 Å and 5 Å, respectively, based on the radial distance and length (z-axis direction) of the analysis region.

g) step

Referring to FIG. 4 , in the simulation model, as a result of analyzing the analysis region for 8 to 9 ns, from the start of the analysis region, from 5 Å (lengthwise direction) before (entry region) and after 5 Å (lengthwise direction) A developmental region (a region where the flow of fluid is fully developed) appears and the velocity distribution as known theoretically is confirmed. The temperature distribution also confirmed that the temperature increased from the center of the analysis area toward the tube surface in the analysis area. In addition, it was confirmed that when the temperature of the fluid reached the temperature of the tube (325K), the temperature of the tube center and the tube surface became similar. where the velocity and temperature distribution along the length (z-axis coordinate) of the analysis region. The Z-axis represents the z-axis coordinates of the simulation model set in step c, and the 600 Å position represents the entrance to the analysis area. In literature, the velocity distribution in the fully developed region of Newtonian fluid requires that the central velocity of the pipe section be approximately twice the average velocity of the pipe section as shown by the fitting equation in Fig. The central velocity of the section was 0.098 Å/ps, and the average velocity was 0.051 Å/ps.) The temperature distribution should show a tendency to increase toward the vicinity of the tube. (Refer to the temperature graph of FIG. 4 ) Since this is consistent with the velocity and temperature distribution of the fluid shown by this simulation model, it was confirmed that the Newtonian fluid was driven.

h ) step

When the behavior of the Newtonian fluid is confirmed from the result of FIG. 4, the simulation domain model obtained from the steps a to g is confirmed, and nanoparticles (beads) showing the nano effect of the nanofluid are added, and these nanoparticles are In the case of inclusion, the mechanism of specific behavior was identified and the effect on thermal properties was observed.

In this case, the nanoparticles are also subjected to coarse-graining as in step a in consideration of the shape and size of the nanofluid bead, and the force field parameters defining the interaction with the fluid and the interaction with the tube constituent particles. set and added to the system. The force field between the nanoparticles and the tube was set so that the graph shape of the radial distribution function of the molecular dynamics simulation could be simulated in the granular molecular dynamics simulation, and the epsilon (ε, kcal/mol) and sigma (σ) of the Leonard-Jones potential , Å) were 0.5 and 4.365, respectively.

Then, according to steps c to g, the flow system after adding 580 nanoparticles was operated to check the effect of nanoparticles on thermal properties. The heat transfer coefficient was obtained by dividing the heat flux by the difference between the tube temperature and the average fluid temperature as shown in the following equation. As a result, a heat transfer coefficient of 7.27×108 W/m ² *K was obtained.

[Equation 3]

h : heat transfer coefficient

q: heat flux

T _s : temperature within the analysis area

T _m : average temperature of the fluid

Claims (8)

A simulation method for realizing a fluid flow, comprising:
a) coarse-graining the fluid;
b) selecting a force field to confirm the interaction of the assembled fluid;
c) creating a simulation model region including the fluid inflow region and analysis region;
d) setting the model domain to a thermodynamic equilibrium state;
e) divide the fluid inflow region into a box region, a random region between the box region and the analysis region, set the fluid in the box region to a solidified state, and apply a uniform force while moving the entire box region in the direction of the analysis region moving the fluid in the random area in the direction of the analysis area;
f) performing an analysis of the displaced fluid in the analysis area;
g) confirming that the flow of the fluid in the simulation model area in the fluid analysis step matches the velocity and temperature distribution of the fluid in scientific research literature; and
h) A simulation method for realizing fluid flow, comprising the step of observing the behavior and physical properties of the fluid in the simulation model area confirmed from steps a to g. The method according to claim 1,
Step a) is,
a-1) selecting a fluid; and
a-2) A simulation method for realizing fluid flow comprising the step of setting a bead of the selected fluid. The method according to claim 1,
The force field in step b) is selected from the group consisting of Martini potential, Shinoda potential, Dissipative particle dynamics potential, and Lennard-Jones potential. A simulation method for realizing fluid flow. The method according to claim 1,
A simulation method for realizing a fluid flow that the force field in step b) is a Lennard-Jones potential. The method according to claim 1,
A simulation method for realizing a fluid flow, wherein the fluid is a Newtonian fluid. The method according to claim 1,
A simulation method for implementing a fluid flow in which the fluid includes at least one selected from the group consisting of water, methanol, ethylene glycol, propylene glycol, and glycerol. 7. The method of claim 6,
A simulation method for realizing a fluid flow further comprising nanoparticles in the fluid. The method according to claim 1,
The simulation method for realizing a fluid flow, wherein the analysis region is in the shape of a pipe.

Images