Molecular dynamics and lattice Boltzmann domain decomposition coupling method and apparatus
By using a molecular dynamics and lattice Boltzmann domain decomposition coupling method, the problems of low computational efficiency and insufficient accuracy of solid-liquid interfaces at the micro-nano scale are solved, and cross-scale continuous coupling of velocity field and temperature field is realized, thereby improving computational accuracy and efficiency.
Patent Information
- Application Number
- CN202611270046.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-08-20
- Publication Date
- 2026-11-17
AI Technical Summary
Existing technologies struggle to accurately describe molecular motion and energy exchange near the solid-liquid interface at the micro-nano scale. Traditional methods suffer from low computational efficiency and insufficient accuracy at the interface, failing to effectively capture interfacial thermal resistance and temperature jumps.
The computational domain is divided into a molecular dynamics region near the solid-liquid interface and a lattice Boltzmann region far from the interface by a coupling method of molecular dynamics and lattice Boltzmann decomposition. In the overlapping region, velocity and temperature are coupled across scales. Through multi-period iteration, the information on both sides is made consistent, and the velocity field and temperature field are continuously connected.
It improves the accuracy of interface physics capture and overall computational efficiency, ensures the continuous transfer of momentum and energy at cross-scale interfaces, solves the problems of poor momentum coupling stability and non-conservation of energy transfer in existing technologies, and realizes the joint continuous coupling of velocity field and temperature field.
Smart Images

Figure CN123120069A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of convective heat transfer simulation technology, and in particular to a molecular dynamics and lattice Boltzmann domain decomposition coupling method and apparatus. Background Technology
[0002] With the development of micro- and nano-scale electromechanical devices, high heat flux density electronic devices, and micro- and nano-channel heat transfer structures, the characteristic scale of flow and heat transfer processes is constantly decreasing. When the channel scale approaches the molecular mean free path or when interface effects dominate, the traditional continuous medium assumption is no longer fully applicable. The Navier-Stokes equations and conventional macroscopic heat transfer models are unable to accurately describe discontinuous effects such as velocity slip, temperature jumps, and solid-liquid interface thermal resistance in the near-wall region.
[0003] In solid-liquid interface heat transfer problems, significant temperature jumps often occur at the interface, and the corresponding interfacial thermal resistance is known as the Carpitza thermal resistance. For micro / nano-scale liquid film or microfluidic systems, the interfacial thermal resistance significantly affects the overall temperature distribution, heat flow transfer, and heat transfer performance. Therefore, accurately characterizing the molecular motion, energy exchange, and temperature jumps near the solid-liquid interface is a key issue in micro / nano-scale convective heat transfer simulations.
[0004] Molecular dynamics methods can directly describe solid-liquid interactions and interfacial energy exchange processes at the atomic scale, but their computational and time scales are limited, making them difficult to apply directly to entire micro / nano channels or systems with large liquid film thicknesses. Continuous medium methods or lattice Boltzmann methods have higher computational efficiency and are suitable for describing the bulk fluid region, but they struggle to naturally reflect molecular-scale effects near the solid-liquid interface.
[0005] Therefore, a cross-scale coupling method is needed to apply molecular dynamics to the microscopic region near the interface and the lattice Boltzmann method to the bulk fluid region, thereby balancing the physical accuracy of the interface and the overall computational efficiency. Summary of the Invention
[0006] To address the shortcomings of existing technologies, this invention provides a molecular dynamics and lattice Boltzmann domain decomposition coupling method and apparatus that balances the capture of interface physical accuracy with the improvement of overall computational efficiency, and simultaneously achieves joint continuous coupling of velocity field and temperature field at cross-scale interfaces.
[0007] In a first aspect, the present invention provides a coupling method for molecular dynamics and lattice Boltzmann domain decomposition, the method comprising the following steps: The computational domain is divided into a molecular dynamics region corresponding to the upper part of the solid-liquid interface, a lattice Boltzmann region corresponding to the bulk fluid region far from the solid-liquid interface, and an overlapping region between the molecular dynamics region and the lattice Boltzmann region. The overlapping region is divided into a first velocity transfer region, a buffer zone, and a second velocity transfer region along the normal direction. In the first velocity transfer region, the velocity information from the molecular dynamics side is transferred to the lattice Boltzmann side, and in the second velocity transfer region, the velocity information from the lattice Boltzmann side is transferred to the molecular dynamics side. Through velocity exchange, the velocity profile is continuously connected between the two scale regions. Within the overlapping region, the thermal conductivity of the molecular dynamics side and the thermal conductivity of the lattice Boltzmann side are statistically analyzed respectively. The average thermal conductivity of the molecular dynamics side and the thermal conductivity of the lattice Boltzmann side is taken as the target thermal conductivity. The target thermal conductivity is applied to the molecular dynamics side and the lattice Boltzmann side. Through multiple coupling cycle iterations, the thermal conductivity on both sides tends to be consistent, and the temperature distribution and velocity distribution at the convergence of the coupling model are output.
[0008] According to the present invention, a molecular dynamics and lattice Boltzmann domain decomposition coupling method is provided, wherein velocity information from the molecular dynamics side is transferred to the lattice Boltzmann side in the first velocity transfer region, and velocity information from the lattice Boltzmann side is transferred to the molecular dynamics side in the second velocity transfer region, thereby achieving a continuous connection of the velocity profile between the two scale regions through velocity exchange, comprising: When the velocity information from the molecular dynamics side is transmitted to the lattice Boltzmann side in the first velocity transmission region, the molecular velocities within the molecular dynamics sampling units corresponding to the grid size of the lattice Boltzmann are statistically averaged, and the average velocity is used as the boundary condition for solving the lattice Boltzmann problem. When transmitting the velocity information from the molecular dynamics side to the lattice Boltzmann side in the second velocity transfer region, the average velocity of the molecules in the molecular dynamics region is gradually approached by velocity state constraint or velocity recalibration method. By directly exchanging velocities, the velocity profile gradually achieves a continuous and smooth transition between the two scale regions.
[0009] According to the present invention, a molecular dynamics and lattice Boltzmann region decomposition coupling method is provided, wherein the first velocity transfer region is a portion close to the molecular dynamics region, the second velocity transfer region is a portion close to the lattice Boltzmann region, and the buffer zone is an intermediate region.
[0010] According to the present invention, a molecular dynamics and lattice Boltzmann domain decomposition coupling method is provided, wherein the thermal conductivity and heat flux on the molecular dynamics side and the lattice Boltzmann side are respectively statistically analyzed within the overlapping region, comprising: The heat conduction flux on the molecular dynamics side is obtained by removing the convective energy transport term and the stress work term from the total energy flux obtained from the molecular dynamics simulation statistics. A temperature gradient is constructed based on the temperature difference between the inner and outer wall nodes, and the heat conduction flux on the Boltzmann side of the lattice is determined based on the temperature gradient and Fourier's law.
[0011] According to the present invention, a molecular dynamics and lattice Boltzmann domain decomposition coupling method is provided, wherein the target heat flux is applied to the molecular dynamics side and the lattice Boltzmann side, and the heat flux on both sides tends to be consistent through multiple coupling cycle iterations, including: The target heat flux is subjected to an exponential moving average to obtain the processed target heat flux; The processed target heat flux is gradually applied to the molecular dynamics side and the lattice Boltzmann side by means of the relaxation coefficient, and the heat flux on both sides gradually becomes consistent and the interface temperature difference tends to stabilize through the multiple coupling cycle iterations.
[0012] According to the present invention, a molecular dynamics and lattice Boltzmann domain decomposition coupling method is provided, wherein the processed target heat flux is gradually applied to the molecular dynamics side and the lattice Boltzmann side through a relaxation coefficient, comprising: Based on the relaxation coefficient, a preset step is advanced by a molecular dynamics system with a microscopic time step, and the treated target heat flux is applied to the boundary region; Based on the relaxation coefficient, a preset step is advanced by a lattice Boltzmann system with a macroscopic time step, and the processed target heat flux is applied at the corresponding boundary node.
[0013] According to the present invention, a molecular dynamics and lattice Boltzmann domain decomposition coupling method is provided, the method further comprising: After velocity coupling, the energy increment caused by velocity coupling is statistically analyzed, and energy compensation is performed based on the energy increment using the thermal motion velocity scaling method. The particle velocity is decomposed into a local average velocity and a thermal velocity relative to the local average velocity; The thermal motion velocity is scaled while keeping the local average velocity constant.
[0014] According to the present invention, a molecular dynamics and lattice Boltzmann domain decomposition coupling method is provided, wherein the liquid in the molecular dynamics domain is described by a coarse-grained water model, the solid wall is made of silicon material, and the solid-liquid interaction is described by the Lennard-Jones potential function.
[0015] According to the molecular dynamics and lattice Boltzmann domain decomposition coupling method provided by the present invention, in the lattice Boltzmann domain, the flow field adopts the D3Q19 lattice model and the BGK single relaxation scheme to solve the incompressible flow equation; the temperature field adopts the dual distribution function model, in addition to the density distribution function, the temperature distribution function is set to solve the heat transport process.
[0016] Secondly, the present invention also provides a molecular dynamics and lattice Boltzmann domain decomposition coupling device, which includes the following modules: The computational domain partitioning module is used to divide the computational domain into a molecular dynamics region corresponding to the upper part of the solid-liquid interface, a lattice Boltzmann region corresponding to the bulk fluid region far from the solid-liquid interface, and an overlapping region located between the molecular dynamics region and the lattice Boltzmann region. A velocity coupling module is used to divide the overlapping region along the normal direction into a first velocity transfer region, a buffer zone, and a second velocity transfer region. In the first velocity transfer region, velocity information from the molecular dynamics side is transferred to the lattice Boltzmann side, and in the second velocity transfer region, velocity information from the lattice Boltzmann side is transferred to the molecular dynamics side. Through velocity exchange, the velocity profile is continuously connected between the two scale regions. The temperature coupling module is used to statistically analyze the thermal conductivity of the molecular dynamics side and the lattice Boltzmann side in the overlapping region, respectively. The average thermal conductivity of the molecular dynamics side and the thermal conductivity of the lattice Boltzmann side is taken as the target thermal conductivity. The target thermal conductivity is applied to the molecular dynamics side and the lattice Boltzmann side. Through multiple coupling cycle iterations, the thermal conductivity on both sides tends to be consistent, and the temperature distribution and velocity distribution at the convergence of the coupling model are output.
[0017] Thirdly, the present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the molecular dynamics and lattice Boltzmann domain decomposition coupling method as described above.
[0018] Fourthly, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the molecular dynamics and lattice Boltzmann domain decomposition coupling method as described above.
[0019] Fifthly, the present invention also provides a computer program product, comprising a computer program that, when executed by a processor, implements the molecular dynamics and lattice Boltzmann domain decomposition coupling method as described above.
[0020] This invention provides a molecular dynamics and lattice Boltzmann domain decomposition coupling method and apparatus. First, the computational domain is divided into a molecular dynamics region corresponding to the upper part of the solid-liquid interface, a lattice Boltzmann region corresponding to the bulk fluid region far from the solid-liquid interface, and an overlap region between the molecular dynamics region and the lattice Boltzmann region. Then, the overlap region is divided along the normal direction into a first velocity transfer region, a buffer zone, and a second velocity transfer region. In the first velocity transfer region, velocity information from the molecular dynamics side is transferred to the lattice Boltzmann side, and in the second velocity transfer region… The velocity information from the lattice Boltzmann side is transferred to the molecular dynamics side, and the velocity profile is made continuous between the two scale regions through velocity exchange. The thermal conductivity of the molecular dynamics side and the lattice Boltzmann side are statistically analyzed in the overlapping region. The average thermal conductivity of the molecular dynamics side and the lattice Boltzmann side is taken as the target thermal conductivity. The target thermal conductivity is applied to the molecular dynamics side and the lattice Boltzmann side. Through multiple coupling cycle iterations, the thermal conductivity on both sides tends to be consistent, and the temperature distribution and velocity distribution at the convergence of the coupled model are output.
[0021] This invention divides the computational domain into a molecular-scale (MD) region above the solid-liquid interface, a molecular-scale (LBM) region in the main fluid region, and an overlapping region between the two. In the overlapping region, velocity coupling (three-part velocity transfer ensures continuous connection of velocity profiles) and temperature coupling (iterative application after heat flux averaging to make the heat flux on both sides tend to be consistent) are performed. On the one hand, MD is used only near the interface to resolve molecular-scale effects, while LBM is used in the main fluid region to ensure computational efficiency. On the other hand, the independent coupling of velocity and temperature in different regions ensures the continuous transfer of momentum and energy across scales. This balances the capture of interface physical accuracy with the improvement of overall computational efficiency, and simultaneously achieves the joint continuous coupling of velocity and temperature fields at the cross-scale interface. Attached Figure Description
[0022] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.
[0023] Figure 1 This is a schematic flowchart of the molecular dynamics and lattice Boltzmann domain decomposition coupling method provided by the present invention.
[0024] Figure 2 This is a schematic diagram of the MD-LBM cross-scale coupled computational domain provided by the present invention.
[0025] Figure 3 This is a schematic diagram of the data transmission process of MD-LBM cross-scale coupling provided by the present invention.
[0026] Figure 4 This is a schematic diagram showing the comparison between the coupling model provided by this invention and the pure MD algorithm.
[0027] Figure 5 This is a schematic diagram of the molecular dynamics and lattice Boltzmann domain decomposition coupling device provided by the present invention.
[0028] Figure 6 This is a schematic diagram of the structure of the electronic device provided by the present invention. Detailed Implementation
[0029] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.
[0030] The terms "first," "second," etc., used in the specification and claims of this application are used to distinguish similar objects and not to describe a specific order or sequence. It should be understood that such terms can be used interchangeably where appropriate so that embodiments of this application can be implemented in orders other than those illustrated or described herein, and the objects distinguished by "first" and "second" are generally of the same class, not limited in number; for example, a first node can be one or more. Furthermore, in the specification and claims, "and / or" indicates at least one of the connected objects, and the character " / " generally indicates that the preceding and following objects are in an "or" relationship.
[0031] To provide a clearer understanding of the various embodiments provided by the present invention, the technical content involved in the present invention will first be described as follows: Existing molecular dynamics-lattice Boltzmann coupling methods or molecular dynamics-continuum coupling methods are mostly focused on isothermal flow or velocity boundary coupling, and are still insufficient for non-isothermal convective heat transfer problems that simultaneously involve velocity fields, temperature fields, wall heat transfer, interfacial heat flow continuity, and Kapitza thermal resistance.
[0032] On the one hand, existing momentum coupling methods mostly use momentum flux constraints directly, but molecular thermal fluctuations in molecular dynamics simulations can lead to large momentum flux statistical noise, affecting the stability of flow coupling.
[0033] On the other hand, existing energy coupling methods lack a consistent treatment of the physical meaning of heat conduction and heat flux. The total energy flux obtained from molecular dynamics statistics includes convective energy transport terms and stress work terms, while the lattice Boltzmann side calculates pure heat conduction and heat flux based on Fourier's law. The two have different physical meanings, and direct matching will lead to energy non-conservation at the cross-scale interface.
[0034] Furthermore, existing coupling frameworks employ the same partitioning strategy for velocity coupling and temperature coupling within the overlapping region, failing to differentiate for the different physical requirements of momentum transfer and energy transfer.
[0035] In summary, there is an urgent need for a molecular dynamics-lattice Boltzmann domain decomposition coupling method for convective heat transfer problems at the micro-nano scale solid-liquid interface, in order to solve the technical problems of poor momentum coupling stability, non-conservation of energy transfer, single coupling strategy in the overlapping region, and large interference from molecular thermal fluctuations in the existing technology.
[0036] The following is combined with Figures 1 to 6 The present invention describes a molecular dynamics and lattice Boltzmann domain decomposition coupling method and apparatus.
[0037] Figure 1 This is a schematic flowchart of the molecular dynamics and lattice Boltzmann domain decomposition coupling method provided by the present invention, as shown below. Figure 1 As shown, the method includes the following: Step 101: Divide the computational domain into the molecular dynamics region corresponding to the upper part of the solid-liquid interface, the lattice Boltzmann region corresponding to the bulk fluid region far from the solid-liquid interface, and the overlapping region between the molecular dynamics region and the lattice Boltzmann region.
[0038] It should be noted that the execution subject of the molecular dynamics and lattice Boltzmann domain decomposition coupling method provided by the present invention can be an electronic device, or any other molecular dynamics and lattice Boltzmann domain decomposition coupling system capable of realizing the molecular dynamics and lattice Boltzmann domain decomposition coupling method.
[0039] This embodiment provides a domain decomposition coupling method for molecular dynamics and lattice Boltzmann.
[0040] Molecular dynamics (MD) is a computational method that simulates the trajectories of atoms or molecules by numerically solving Newton's equations of motion, directly describing solid-liquid interactions and interfacial energy exchange processes at the atomic scale. The Lattice Boltzmann Method (LBM) is a mesoscopic-scale computational fluid dynamics method that simulates fluid flow and heat transfer processes by solving the evolution equations of particle distribution functions.
[0041] Figure 2 This is a schematic diagram of the MD-LBM cross-scale coupled computational domain provided by the present invention. See [link / reference]. Figure 2 The diagram illustrates the three-dimensional coordinates of X, Y, and Z, where the Z direction is the channel normal direction, and periodic boundary conditions are applied to the X and Y directions. The computational domain is divided into three regions along the Z direction (channel normal direction): the molecular dynamics (MD) region near the solid-liquid interface (upper part of the diagram), the bulk fluid region away from the solid-liquid interface (lower part of the diagram) corresponding to the lattice Boltzmann (LBM) region, and the overlapping region between the MD and LBM regions (shaded area in the diagram). The MD, LBM, and overlapping regions are all simulation regions. The solid-liquid interface is the boundary region where the solid wall contacts the liquid phase; discontinuous effects such as velocity slip, temperature jumps, and Kapitza thermal resistance occur in this region. The overlapping region is a transitional area between the molecular dynamics simulation region and the lattice Boltzmann simulation region, used for cross-scale data exchange and coupling.
[0042] Step 102: Divide the overlapping region along the normal direction into a first velocity transfer region, a buffer zone, and a second velocity transfer region. In the first velocity transfer region, transfer the velocity information from the molecular dynamics side to the lattice Boltzmann side. In the second velocity transfer region, transfer the velocity information from the lattice Boltzmann side to the molecular dynamics side. Through velocity exchange, the velocity profile is continuously connected between the two scale regions.
[0043] Specifically, after dividing the computational domain, velocity coupling (momentum coupling) is further performed in the overlapping region. The normal direction is perpendicular to the solid-liquid interface (Z direction), that is, from the solid wall to the fluid bulk.
[0044] In the Z direction, the overlapping region has different partitioning strategies along the normal direction to adapt to the different physical requirements of momentum transfer and energy transfer.
[0045] See Figure 2 As shown in the right-hand portion (momentum coupling region), the right-hand portion of the overlapping region is divided into three regions along the Z-direction. The upper third is the MD velocity sampling region (i.e., the first velocity transfer region), which is used to sample and obtain velocity information from the MD side (denoted as...). ), and the speed information The velocity information is transmitted to the corresponding grid points on the LBM side; the lower third is the LBM velocity sampling area (i.e., the second velocity transmission area), which is used to sample and obtain the velocity information on the LBM side (denoted as...). ), and the speed information The velocity information is transmitted to the corresponding molecule on the MD side; the middle third serves as a buffer. The buffer does not actively transmit velocity information; it only acts as a transition region to avoid discontinuities in the velocity profile caused by direct conflict between the target velocity on the LBM side and the statistical velocity on the MD side.
[0046] This three-part velocity coupling strategy adapts to different transfer requirements from the continuous medium to the particle domain (LBM→MD) and from the particle domain to the continuous medium (MD→LBM).
[0047] Step 103: In the overlapping region, the thermal conductivity of the molecular dynamics side and the thermal conductivity of the lattice Boltzmann side are statistically analyzed respectively. The average thermal conductivity of the molecular dynamics side and the thermal conductivity of the lattice Boltzmann side is taken as the target thermal conductivity. The target thermal conductivity is applied to the molecular dynamics side and the lattice Boltzmann side. Through multiple coupling cycle iterations, the thermal conductivity on both sides tends to be consistent. The temperature distribution and velocity distribution at the convergence of the coupling model are output.
[0048] Specifically, after dividing the computational domain, temperature coupling (energy coupling) is further performed in the overlapping region.
[0049] See Figure 2 As shown, in the left part of the overlapping region (energy coupling region), the left part of the overlapping region is divided into upper and lower regions along the Z direction. The upper half is the MD heat flow and temperature sampling region, used to sample and obtain the heat conduction flow on the MD side (denoted as ). ) and the temperature on the MD side (denoted as The lower half is the LBM temperature sampling area, used to sample and obtain the temperature on the LBM side (denoted as...). Heat flow on the MD side. After obtaining the total energy flux through LAMMPS statistics, the convective energy transport term and the stress work term are removed to obtain the pure conduction heat flux; the conduction heat flux on the LBM side is calculated from the temperature gradient according to Fourier's law. The heat flux on both sides is averaged and applied as the target heat flux to the coupling interface.
[0050] Subsequently, through multiple coupling cycle iterations, the heat flow on both sides is made to be consistent, and the temperature distribution and velocity distribution at the convergence of the coupled model are output.
[0051] The following is a detailed example 1: This embodiment focuses on the problem of convective heat transfer in micro-nano channels at the silicon-water solid-liquid interface and establishes an MD-LBM cross-scale coupling model.
[0052] First, a global computational domain is established. Periodic boundary conditions are applied in the x and y directions, with the z direction being the channel normal. The upper region near the solid-liquid interface is simulated using molecular dynamics, while the bulk fluid region far from the interface is simulated using the lattice Boltzmann method. A certain thickness of overlap is established between the two to facilitate cross-scale data exchange.
[0053] In the molecular dynamics region, liquid water is described using the mW coarse-grained water model, where a water molecule is treated as a particle to reduce computational cost. The solid walls are made of silicon with a diamond lattice structure. The silicon-water interaction is described using the Lennard-Jones potential. Molecular dynamics simulations are performed using a large-scale atomic / molecular massively parallel simulator (LAMMPS), employing the Verlet algorithm to integrate the equations of motion.
[0054] Within the LBM region, the flow field is solved using the D3Q19 lattice model and the Bhatnagar-Gross-Krook (BGK) single-relaxation scheme to solve the incompressible flow equations. D3Q19 is a discrete velocity model for three-dimensional (D3) simulations, where each lattice point contains 19 (Q19) discrete velocity directions. The temperature field employs a dual-distribution-function model, i.e., an independent temperature distribution function is set in addition to the density distribution function to solve for the heat transport process. The LBM region is used to calculate the bulk flow velocity and temperature fields.
[0055] During velocity coupling, the overlapping region is divided into three parts along the z-direction: a portion closer to the LBM region serves as the region for LBM to transmit velocity information to the MD region (the second velocity transmission region); a portion closer to the MD region serves as the region for MD to transmit velocity information to the LBM region (the first velocity transmission region); and the middle region acts as a buffer. During LBM to MD transmission, the target velocity is extracted from the corresponding LBM grid cell, and velocity constraints are used to gradually bring the average velocity of molecules within the MD region closer to the target velocity. During MD to LBM transmission, the molecular velocities within the MD sampling cells corresponding to the LBM grid size are statistically averaged, and this average velocity is used as the boundary condition for the LBM solution. Through velocity exchange, the velocity profile gradually achieves continuity and smooth transition between the two scale regions.
[0056] During temperature coupling, the overlap region is divided into the MD statistical region and the LBM statistical region. The MD statistical region is used to statistically analyze the average temperature and average heat flux on the molecular dynamics side, while the LBM statistical region is used to read the temperature on the continuous medium side and calculate the Fourier heat flux. At the beginning of each coupling cycle, the heat conduction flux on the MD side and the heat conduction flux on the LBM side are statistically analyzed, and the average of the two is taken as the target heat flux and applied to both sides. The MD system advances several steps with a microscopic time step and applies the corresponding heat flux to the boundary region; the LBM system advances several steps with a macroscopic time step and applies the same cumulative heat flux to the corresponding boundary nodes. Through multiple coupling cycle iterations, the heat fluxes on both sides gradually become consistent, and the interface temperature difference tends to stabilize. Finally, the temperature and velocity distributions at the convergence of the coupling model are output.
[0057] The method provided in this embodiment of the invention first divides the computational domain into a molecular dynamics region corresponding to the upper part of the solid-liquid interface, a lattice Boltzmann region corresponding to the bulk fluid region far from the solid-liquid interface, and an overlapping region located between the molecular dynamics region and the lattice Boltzmann region. Then, the overlapping region is divided into a first velocity transfer region, a buffer zone, and a second velocity transfer region along the normal direction. In the first velocity transfer region, the velocity information of the molecular dynamics side is transferred to the lattice Boltzmann side, and in the second velocity transfer region, the velocity information of the lattice Boltzmann side is transferred to the molecular dynamics side. Through direct velocity coupling, the velocity profile is made to be continuously connected between the two scale regions. The thermal conductivity of the molecular dynamics side and the thermal conductivity of the lattice Boltzmann side are statistically analyzed in the overlapping region. The average thermal conductivity of the molecular dynamics side and the thermal conductivity of the lattice Boltzmann side is taken as the target heat flux. The target heat flux is applied to the molecular dynamics side and the lattice Boltzmann side. Through multiple coupling cycle iterations, the heat flux on both sides tends to be consistent, and the temperature distribution and velocity distribution at the convergence of the coupling model are output.
[0058] This invention divides the computational domain into a molecular-scale (MD) region above the solid-liquid interface, a molecular-scale (LBM) region in the main fluid region, and an overlapping region between the two. Within the overlapping region, velocity coupling (three-part velocity transfer ensures continuous velocity profile connection) and temperature coupling (iterative application after heat flux averaging to make the heat flux on both sides tend to be consistent) are performed independently. On the one hand, MD is used only near the interface to resolve molecular-scale effects, while LBM is used in the main fluid region to ensure computational efficiency. On the other hand, the independent coupling of velocity and temperature in different regions ensures continuous cross-scale transfer of momentum and energy, thus balancing the capture of interface physical accuracy with the improvement of overall computational efficiency. At the same time, it achieves joint continuous coupling of velocity and temperature fields at the cross-scale interface.
[0059] It should be noted that each implementation method of this application can be freely combined, rearranged, or executed individually, and does not need to rely on or depend on a fixed execution order.
[0060] According to the present invention, a molecular dynamics and lattice Boltzmann domain decomposition coupling method transfers velocity information from the molecular dynamics side to the lattice Boltzmann side in a first velocity transfer region, and transfers velocity information from the lattice Boltzmann side to the molecular dynamics side in a second velocity transfer region. This velocity exchange enables a continuous connection of the velocity profile between the two scale regions, including: When transferring the velocity information from the molecular dynamics side to the lattice Boltzmann side in the first velocity transfer region, the molecular velocities within the molecular dynamics sampling cells corresponding to the grid size of the lattice Boltzmann are statistically averaged, and the average velocity is used as the boundary condition for the lattice Boltzmann solution. In the second velocity transfer region, when transferring the velocity information from the lattice Boltzmann side to the molecular dynamics side, velocity state constraints or velocity recalibration methods are used to make the average velocity of molecules in the molecular dynamics region gradually approach the target velocity. By directly exchanging velocities, the velocity profile gradually achieves a continuous and smooth transition between the two scale regions.
[0061] Specifically, in some embodiments, the velocity coupling process includes the following steps: Among them, velocity state constraint means that when transmitting macroscopic velocity information to the molecular dynamics region, instead of directly constraining molecular momentum flux, the statistical noise caused by molecular thermal fluctuations is avoided by constraining the average molecular velocity. Velocity recalibration is a technique that scales or adjusts the overall velocity of molecules within the molecular dynamics region so that its average velocity gradually approaches the target velocity.
[0062] In practical applications, when the velocity information from the molecular dynamics side is transferred to the lattice Boltzmann side in the first velocity transfer region, the molecular velocities within the molecular dynamics sampling cells corresponding to the grid size of the lattice Boltzmann are statistically averaged, and the average velocity is used as the boundary condition for solving the lattice Boltzmann problem.
[0063] In the second velocity transfer region, when transferring the velocity information from the lattice Boltzmann side to the molecular dynamics side, velocity state constraints or velocity recalibration methods are used to gradually bring the average velocity of molecules in the molecular dynamics simulation region closer to the target velocity.
[0064] Since the velocity state constraint acts directly on the average velocity of molecules rather than momentum flux, it can avoid the momentum flux statistical noise caused by molecular thermal fluctuations and improve the stability of flow coupling.
[0065] Through direct velocity exchange, the velocity profile gradually achieves a continuous and smooth transition between the two scale regions.
[0066] The following is a detailed embodiment 2: In the first velocity transfer region (the part near the MD region), the MD region is divided into several sampling cells according to the LBM grid size. The molecular velocities within each sampling cell are ensemble averaged, and the average velocity is used as the boundary condition for the LBM solution. Through calculations over multiple coupled cycles, the velocity profile gradually achieves continuity and smooth transition between the two scale regions.
[0067] In the second velocity transfer region (the part near the LBM region), after extracting the target velocity from the LBM grid, the velocities of molecules in the MD region are adjusted through velocity recalibration: the current average velocity of molecules within the MD sampling unit is calculated, compared with the target velocity, and a uniform correction is applied to the velocity of each molecule, so that the overall average velocity gradually approaches the target velocity. This velocity state constraint method does not directly constrain momentum flux, therefore the instantaneous momentum fluctuations generated by molecular thermal motion are not amplified, and the coupling process is more stable.
[0068] The method provided in this invention uses velocity state constraints or velocity recalibration methods (instead of directly using momentum flux constraints) when transmitting velocity information from LBM to MD to gradually bring the average velocity of MD molecules closer to the target velocity. This avoids the strong statistical noise caused by molecular thermal fluctuations when directly calculating molecular momentum flux. At the same time, when transmitting from MD to LBM, the molecular velocities within the sampling unit are statistically averaged and used as LBM boundary conditions. Thus, while ensuring accurate transmission of velocity information, the method effectively suppresses the interference of molecular thermal fluctuations on momentum coupling and improves the stability of velocity profile iterative convergence.
[0069] According to the molecular dynamics and lattice Boltzmann region decomposition coupling method provided by the present invention, the first velocity transfer region is the part close to the molecular dynamics region, the second velocity transfer region is the part close to the lattice Boltzmann region, and the buffer zone is the middle region.
[0070] Specifically, the first velocity transfer region is the portion of the overlap region closest to the molecular dynamics region, where velocity information is transferred from the MD side to the LBM side. The second velocity transfer region is the portion of the overlap region closest to the lattice Boltzmann region, where velocity information is transferred from the LBM side to the MD side. The buffer zone is the intermediate region located between the first and second velocity transfer regions, used for smooth transition and to avoid abrupt changes in velocity information.
[0071] The following is a detailed embodiment 3: Along the Z-direction (channel normal direction), the right side of the overlapping area is sequentially divided into three consecutive sub-regions (e.g., Figure 2 As shown on the right): The lower third of the LBM region is the LBM velocity sampling area (i.e., the second velocity transfer area, V). LBMThe middle section is the buffer zone, and the upper third of the MD region is the MD velocity sampling area (i.e., the first velocity transfer area, V). MD (Regions). The thickness ratio of the three regions is adjusted according to the differences in grid resolution and time scale between MD and LBM. In a typical implementation, the thickness of the first velocity transfer zone, the buffer zone, and the second velocity transfer zone are approximately equal (each occupying one-third of the overlapping area), but the specific ratio can be optimized according to the actual flow characteristics. The buffer zone does not actively transfer velocity information; it only serves as a transition zone to avoid discontinuities in the velocity profile caused by direct conflict between the target velocity on the LBM side and the statistical velocity on the MD side.
[0072] The method provided in this invention clearly divides the overlapping area along the normal direction into a first velocity transfer zone near the MD region, an intermediate buffer zone, and a second velocity transfer zone near the LBM region. The buffer zone is used as a transition zone to avoid direct numerical conflicts between the target velocity on the LBM side and the statistical velocity on the MD side. This prevents the discontinuity of the profile caused by direct data exchange between the two sides during the cross-scale transfer of velocity information, and further ensures the smoothness and stability of velocity coupling iteration.
[0073] According to the present invention, a molecular dynamics and lattice Boltzmann domain decomposition coupling method is provided, which separately statistically analyzes the thermal conductivity and heat flux on the molecular dynamics side and the lattice Boltzmann side within the overlapping region, including: The heat conduction flux on the molecular dynamics side is obtained by removing the convective energy transport term and the stress work term from the total energy flux obtained from molecular dynamics simulation statistics. A temperature gradient is constructed based on the temperature difference between the inner and outer wall nodes, and the heat conduction flow on the Boltzmann side of the lattice is determined based on the temperature gradient and Fourier's law.
[0074] Specifically, in some embodiments, the process of statistical analysis of heat conduction flux in temperature coupling includes the following steps: The total energy flux is the energy flux statistically obtained using the heat / flux correlation method of LAMMPS, including contributions from convective energy transport, stress work, and pure conduction heat flux. The convective energy transport term represents the energy transport contribution carried by the overall motion of particles (macroscopic flow), and is related to the translational kinetic and potential energy of the particles. The stress work term represents the energy contribution corresponding to the work done by the interaction forces between particles, and is related to the interaction forces and relative velocities. Fourier's law is the macroscopic law describing the relationship between conduction heat flux and the temperature gradient; that is, the heat flux density is proportional to the temperature gradient, and the proportionality constant is the thermal conductivity.
[0075] In practical applications, at the beginning of each coupling cycle, the heat conduction fluxes on the MD side and the LBM side are first statistically analyzed. The heat conduction flux on the MD side is obtained by removing the convective energy transport term and the stress work term from the total energy flux obtained from molecular dynamics simulations (such as the heat / flux correlation method in LAMMPS), thus obtaining the pure heat conduction flux. The purpose of this removal operation is to ensure that the physical meaning of the heat flux on the MD side is consistent with that of the Fourier heat flux on the LBM side, both being pure heat conduction fluxes, rather than the total energy flux including macroscopic flow contributions.
[0076] The heat conduction flow on the LBM side is obtained by constructing a temperature gradient based on the temperature difference between the inner and outer wall nodes, and then determining the heat conduction flow on the lattice Boltzmann side based on this temperature gradient and Fourier's law.
[0077] The following is a detailed example 4: On the MD side, in the upper half of the MD heat flux sampling region on the left side of the overlap area, the total energy flux is calculated using the computeheat / flux command in LAMMPS. In LAMMPS, the total energy flux is defined as the sum of the products of energy and velocity of each particle plus the potential force contribution from the interparticle forces. This total energy flux includes convective energy transport and stress work terms. Subtracting these two contributions from the total energy flux leaves the pure conduction heat flux. Specifically, the convective heat flux and the potential force heat flux are extracted from the LAMMPS output and then removed from the total energy flux.
[0078] On the LBM side, in the lower half of the LBM temperature sampling area on the left side of the overlap region, the inner wall node near the interface and the outer wall node far from the interface are taken in the overlap region. The temperature difference ΔT between them is calculated and divided by the node spacing Δz to obtain the temperature gradient ∇T. Then, the heat conduction flux is calculated according to Fourier's law q=-K∇T, where K is the thermal conductivity of the fluid.
[0079] The method provided in this invention obtains pure conductive heat flux by removing the convective energy transport term and the stress work term from the total energy flux obtained from LAMMPS statistics on the MD side. This ensures that the heat flux on the MD side and the conductive heat flux calculated by the LBM side according to Fourier's law have completely consistent physical meanings (both are pure conductive heat fluxes). This avoids the problem of non-conservation of cross-scale interface energy caused by the mismatch of the physical meanings of the heat fluxes on both sides, and guarantees the physical consistency and conservation of energy transfer.
[0080] According to the molecular dynamics and lattice Boltzmann domain decomposition coupling method provided by the present invention, a target heat flux is applied to both the molecular dynamics side and the lattice Boltzmann side, and the heat flux on both sides tends to be consistent through multiple coupling cycle iterations, including: The target heat flux is processed by exponential moving average to obtain the processed target heat flux. By applying the treated target heat flux gradually to the molecular dynamics side and the lattice Boltzmann side through the relaxation coefficient, the heat flux on both sides gradually becomes consistent and the interface temperature difference tends to stabilize through multiple coupling cycle iterations.
[0081] Specifically, in some embodiments, the target heat flux application step includes the following steps: Exponential moving average is a method for smoothing time series data, assigning higher weights to recent data, which decay exponentially over time, effectively suppressing instantaneous fluctuations. The relaxation coefficient, a factor between 0 and 1, controls the rate at which the target heat flux is gradually applied, preventing abrupt changes in heat flux that could lead to computational instability.
[0082] In practical applications, the target heat flux is first processed using an exponential moving average to obtain the processed target heat flux. Specifically, let the target heat flux obtained from the current coupling cycle be q(t), and the heat flux after moving average in the previous coupling cycle be q̄(t-1). Then, the heat flux after moving average in the current cycle is: q̄(t) = α·q(t) + (1-α)·q̄(t-1) Where α is the moving average coefficient (0 < α < 1), q̄(t) is the heat flow after moving average in the current coupling cycle, q(t) is the target heat flow obtained statistically in the current coupling cycle, and q̄(t-1) is the heat flow after moving average in the previous coupling cycle.
[0083] Then, the treated target heat flux is gradually applied to both the molecular dynamics side and the lattice Boltzmann side using a relaxation coefficient. That is, the actual applied heat flux is q. applied =β·q̄(t), where β is the relaxation coefficient (0<β≤1).
[0084] Through multiple coupling cycle iterations, the heat flow on both sides gradually becomes consistent, and the interface temperature difference tends to stabilize.
[0085] The following is a detailed embodiment 5: In each coupling cycle, the target heat flux is obtained by averaging the MD-side conductive heat flux qMD measured in the upper half of the MD sampling area on the left side of the overlap region and the LBM-side conductive heat flux calculated according to Fourier's law in the lower half of the LBM sampling area on the left side of the overlap region. Since the instantaneous heat flux fluctuations caused by molecular thermal motion in the MD simulation can be several times the average value, directly applying this target heat flux would lead to severe oscillations in the temperature field. Therefore, an exponential moving average is applied to the target heat flux, with a moving average coefficient α = 0.15, i.e., q̄(t) = 0.15·q(t) + 0.85·q̄(t-1), effectively suppressing instantaneous heat flux fluctuations.
[0086] Subsequently, the processed target heat flux is applied gradually using a relaxation factor. In the first 80 coupling cycles, the relaxation factor β is set to a small value of 0.2 to avoid sudden application of heat flux that could cause system instability. From the 81st coupling cycle onwards, β is maintained at 0.35, representing the fully applied target heat flux after moving average. This smooth and gradual application strategy significantly improves the stability of the coupling calculation.
[0087] The method provided in this invention suppresses instantaneous heat flux fluctuations caused by molecular thermal motion in MD simulation by performing exponential moving average processing on the target heat flux, and gradually applies the processed target heat flux to the solvers on both sides through relaxation coefficients, thereby smoothing the cross-scale transfer process of heat flux in the time dimension, effectively reducing the disturbance of macroscopic temperature field by molecular-scale thermal fluctuations, and improving the numerical stability of heat flux coupling iteration.
[0088] According to the present invention, a molecular dynamics and lattice Boltzmann domain decomposition coupling method is provided, which progressively applies the processed target heat flux to the molecular dynamics side and the lattice Boltzmann side through a relaxation coefficient, comprising: Based on the relaxation coefficient, a preset step is advanced by a molecular dynamics system with a microscopic time step, and the treated target heat flux is applied to the boundary region; Based on the relaxation coefficient, a preset step is advanced by a lattice Boltzmann system with a macroscopic time step, and the processed target heat flux is applied at the corresponding boundary node.
[0089] Specifically, the microscopic time step is the time step used in molecular dynamics simulations, typically on the order of femtoseconds (fs) (e.g., 1 fs), used to integrate the equations of motion of particles. The macroscopic time step is the time step used in lattice Boltzmann simulations, typically several orders of magnitude larger than the microscopic time step, depending on the mesh size and fluid properties.
[0090] In practical applications, the treated target heat flux is gradually applied to the molecular dynamics side and the lattice Boltzmann side using a relaxation coefficient, including the following steps: Based on the relaxation coefficient, the molecular dynamics system advances a predetermined step with a microscopic time step and applies the treated target heat flux to the boundary region. The lattice Boltzmann system advances a predetermined step with a macroscopic time step and applies the same cumulative heat flux to the corresponding boundary nodes. Since MD and LBM use different time steps, the number of steps advanced on both sides is different, but the total applied heat flux (heat flux density × time) remains consistent to ensure energy conservation.
[0091] The following is a detailed example 6: Within one coupling cycle, the MD system advances 500,000 steps (500 ps) with a microscopic time step of 1 fs, applying the treated target heat flux to the boundary layer (molecular layer near the overlap region) of the MD region. Simultaneously, the LBM system advances 5,000 steps (500 ps) with a macroscopic time step of 100 fs, applying the same cumulative heat flux to the corresponding boundary node of the LBM region. Since the time span of 500,000 steps (500 ps) for MD and 5,000 steps (500 ps) for LBM is the same, the total applied heat (heat flux density × time) is equal on both sides, thus ensuring energy conservation. As the coupling cycle increases, the relaxation coefficient gradually increases, and the applied heat flux gradually approaches the target heat flux.
[0092] The method provided in this invention advances the MD system with a microscopic time step and the LBM system with a macroscopic time step based on the relaxation coefficient, and applies the same cumulative heat flux (heat flux density × time) to their respective boundaries. This ensures that even if the solvers on both sides use different time step systems, the total heat applied within a coupling cycle remains strictly equal. Thus, while taking into account the time scale differences between MD and LBM, the conservation of energy exchange at the cross-scale interface is guaranteed.
[0093] According to the present invention, a coupling method for molecular dynamics and lattice Boltzmann domain decomposition is provided, the method further includes: After velocity coupling, the energy increment caused by velocity coupling is statistically analyzed, and energy compensation is performed based on the energy increment using the thermal motion velocity scaling method. The particle velocity is decomposed into local average velocity and thermal velocity relative to the local average velocity; The thermal velocity is scaled while keeping the local average velocity constant.
[0094] Specifically, in some embodiments, during the momentum coupling process, the energy of the system may change due to velocity constraints and stress work. This embodiment further calculates the energy increment caused by momentum coupling and uses the thermal motion velocity scaling method for compensation.
[0095] After velocity coupling, the energy increment caused by velocity coupling is statistically analyzed. Energy compensation is performed using a thermal velocity scaling method: the particle velocity is decomposed into a local average velocity and a thermal velocity relative to the local average velocity; the thermal velocity is scaled while keeping the local average velocity constant.
[0096] This thermal motion velocity scaling energy compensation method only corrects the thermal motion velocity of particles without changing the local average velocity, thereby compensating for the energy changes caused by stress work, while maintaining the stability of the momentum field and avoiding non-physical interference of velocity coupling on the energy field.
[0097] The following is a detailed embodiment 7: During velocity coupling, due to the application of velocity constraints or velocity recalibration to the molecules in the MD region, the velocity distribution of the molecules changes, and the kinetic energy and internal energy of the system may deviate from their physical equilibrium values. Specifically, let the velocity of the i-th particle in a certain MD sampling unit before velocity coupling be v. i The local average velocity is: in, For local average velocity, The velocity of a particle in a certain MD sampling unit before velocity coupling is given, and N represents the total number of particles in the molecular dynamics (MD) sampling unit (i.e., the total number of molecules contained in the sampling unit).
[0098] The thermal velocity of the i-th particle for: After velocity coupling, the energy increment ΔE of the statistical system is calculated. If ΔE > 0 (system energy increases), the thermal velocity needs to be reduced; if ΔE < 0 (system energy decreases), the thermal velocity needs to be amplified. The scaling factor λ is determined based on the energy increment. in, To compensate for the target internal energy, It is the translational kinetic energy (determined by the local average velocity). It is kinetic energy generated by thermal motion.
[0099] Then the thermal velocity of each particle is scaled: in, Let be the scaled thermal velocity of the i-th particle. Let be the thermal velocity of the i-th particle. This is the scaling factor.
[0100] The particle's new velocity is: The velocity of the i-th particle is v i Due to local average velocity If the velocity field remains unchanged, the macroscopic velocity field obtained by velocity coupling is unaffected, while the internal energy of the system is corrected, and the temperature field is restored to physical consistency.
[0101] The method provided in this invention, by statistically analyzing the energy increment caused by velocity constraints or velocity recalibration after velocity coupling, and by decomposing the particle velocity into local average velocity and thermal motion velocity, scaling only the thermal motion velocity while keeping the local average velocity unchanged, corrects the system energy deviation caused by velocity coupling without affecting the macroscopic velocity field. Thus, while compensating for the non-physical energy changes caused by stress work during velocity coupling, it maintains the stability of the momentum field and avoids secondary interference of energy compensation operations on the velocity coupling results.
[0102] According to the present invention, a molecular dynamics and lattice Boltzmann domain decomposition coupling method is provided, wherein the liquid in the molecular dynamics domain is described by a coarse-grained water model, the solid wall is made of silicon material, and the solid-liquid interaction is described by the Lennard-Jones potential function.
[0103] Specifically, in some embodiments, the liquid in the MD region is described using a coarse-grained water model. In this embodiment, the mW coarse-grained water model is preferred. This model treats a single water molecule as an equivalent particle and simulates the tetrahedral hydrogen bond structure of water molecules by introducing a non-bond angle dependent term. The mW model only includes short-range interactions, but it can reproduce the thermodynamic properties and phase transition behavior of water with better accuracy than most all-atom models, while the computational cost is less than 1% of that of all-atom models.
[0104] Among them, the coarse-grained water model is a simplified model that treats multiple atoms or molecules as a single particle, used to reduce the computational cost of molecular dynamics simulations. The mW (monatomic water) model is a representative coarse-grained water model that treats a single water molecule as a single particle.
[0105] The solid walls are made of silicon, with a diamond lattice structure. The solid-liquid interaction is described using the Lennard-Jones potential function. The Lennard-Jones potential function is a mathematical model that describes the interaction potential energy between two neutral atoms or molecules, containing attraction and repulsion terms, and is widely used to describe solid-liquid interactions.
[0106] The method provided in this invention reduces computational costs by using the mW coarse-grained water model in the MD region to represent a single water molecule as a particle. At the same time, it uses the Lennard-Jones potential function to describe silicon-water interaction, thereby significantly reducing the computational cost in the MD region while ensuring an accurate description of the intermolecular interactions at the solid-liquid interface. This allows a solid-liquid interface system containing tens of thousands of particles to be simulated within a reasonable time.
[0107] According to the molecular dynamics and lattice Boltzmann domain decomposition coupling method provided by the present invention, in the lattice Boltzmann domain, the flow field adopts the D3Q19 lattice model and the BGK single relaxation scheme to solve the incompressible flow equation; the temperature field adopts the dual distribution function model, and the temperature distribution function is set in addition to the density distribution function to solve the heat transport process.
[0108] Specifically, in some embodiments, within the lattice Boltzmann simulation region, the flow field is solved using the D3Q19 lattice model and the BGK single relaxation scheme to solve the incompressible flow equations. The D3Q19 model contains 19 discrete velocity directions, enabling accurate description of momentum transport in three-dimensional space. The BGK single relaxation scheme uses a relaxation time τ to control the approach of the distribution function to the equilibrium state; the relaxation time is related to the kinematic viscosity of the fluid.
[0109] Among them, the D3Q19 lattice model is a three-dimensional nineteen-velocity lattice model, one of the most commonly used three-dimensional flow field models in the lattice Boltzmann method, containing 19 discrete velocity directions. The BGK single-relaxation scheme, namely the Bhatnagar-Gross-Krook single-relaxation approximation, is the most basic collision model in the lattice Boltzmann method, using a unified relaxation time to describe the process of the distribution function approaching the equilibrium state.
[0110] The temperature field adopts a dual distribution function model, that is, in addition to the density distribution function, an independent temperature distribution function is set up to solve the heat transport process. The density distribution function is used to solve the density and velocity of the flow field, and the temperature distribution function is used to solve the temperature field. The two are coupled through the velocity field (convection term) to achieve the joint solution of flow and heat transfer.
[0111] The method provided in this invention solves the flow field in the LBM region using the D3Q19 lattice model and the BGK single relaxation scheme, and solves the temperature field using a dual distribution function model (with an independent temperature distribution function in addition to the density distribution function). This couples the velocity field and the temperature field through the convection term in the distribution function, thereby accurately describing the fluid momentum transport and heat transport in three-dimensional space while achieving a joint solution for flow and heat transfer.
[0112] Figure 3 This is a schematic diagram of the data transmission process for MD-LBM cross-scale coupling provided by the present invention, as follows: Figure 3 As shown, the data transfer process for MD-LBM cross-scale coupling is as follows: Specifically, the upper part of the figure illustrates the process of inputting atomic parameters. These atomic parameters, after being calculated using molecular dynamics simulations, generate the MD average velocity. Simultaneously, the Lattice Boltzmann simulation yields the LBM velocity distribution. The MD-side calculation corresponds to the microscale, while the LBM-side calculation corresponds to the mesoscale.
[0113] The coupling mechanism of the upper overlapping region is shown in the middle of the figure. In this upper overlapping region, the heat flux q is obtained by sampling on the MD side. MD and temperature T MD Temperature T was obtained by sampling from the LBM side. LBM T in the diagram h T represents the temperature at the high-temperature end of the coupling interface. l This represents the temperature at the low-temperature end, and the temperature difference between the two (T). h -T l Used to calculate the heat flux q on the LBM side based on Fourier's law. LBM MD-side heat flow q MD With LBM side heat flow q LBM The average heat flux is obtained by averaging. The average heat flux The target heat flow is applied to the coupling interface; at the same time, the temperature on the MD side and the temperature on the LBM side are coupled to achieve the co-evolution of the temperature fields on both sides.
[0114] The lower part of the figure illustrates the coupling mechanism of the overlapping region. In this lower overlapping region, the velocities on the MD side and the LBM side exchange velocity, with the LBM side transmitting velocity distribution information to the MD side. The velocity information from the LBM side is provided as boundary conditions to the MD side to constrain the motion of molecules in the MD region; simultaneously, in the upper overlapping region, the statistically averaged velocity from the MD side serves as the boundary condition for solving the LBM. The figure also shows the boundary temperature q. b It is used to characterize the temperature boundary at the solid-liquid interface.
[0115] By coupling heat flow and temperature, as well as velocity, in the overlapping region, this method enables complete velocity-temperature-heat flow joint data exchange between the MD and LBM.
[0116] Optionally, the coupled model can be validated. First, Couette flow and pure heat transfer examples are performed in a smaller computational domain, and the velocity and temperature profiles obtained from the coupled model are compared with those from the pure MD algorithm. The results show that the MD–LBM hybrid model can reproduce the velocity and temperature distributions of the pure MD algorithm well. Subsequently, the length of the LBM region is expanded to construct a larger-scale computational system, further validating the stability of the method in long channels. The results show that the velocity field gradually forms a linear Couette distribution, the temperature gradients on both sides of the MD and LBM gradually converge, and the interfacial heat flow stabilizes after several coupling cycles, indicating that the method can achieve stable momentum and energy transfer.
[0117] For example, Figure 4 This is a schematic diagram comparing the coupled model provided by this invention with the pure MD algorithm, as shown in the figure. Figure 4As shown, the two figures on the left, (1) and (2), focus on the smaller region of the z-axis (0-500 Å), aiming to verify the continuity and accuracy of the coupling model at the microscopic interface. In Figure (1), the horizontal axis represents the z-axis coordinate in space, used to describe the spatial position of the system in the vertical direction, with the length unit being Å (angstroms). The vertical axis represents the velocity (m / s). The curves of the coupling algorithm (red solid line) and the pure MD algorithm (light-colored scatter plots) almost completely overlap. In Figure (2), the horizontal axis represents the z-axis coordinate in space, used to describe the spatial position of the system in the vertical direction, with the length unit being Å (angstroms). The vertical axis represents the temperature (K). The curves of the coupling algorithm (red solid line) and the pure MD algorithm (light-colored scatter plots) almost completely overlap. Among them, the light-colored scatter plots represent the results of traditional molecular dynamics simulations, serving as a high-precision benchmark reference. The red solid line represents the calculation results of the coupling model proposed in this invention. The MD part is the microscopic region calculated using the MD method in the coupling model, and the LBM part is the mesoscopic / macroscopic region calculated using the LBM method in the coupling model.
[0118] The two figures on the right, (3) and (4), show the overall distribution of the large z-axis region (0-1000 Å), highlighting the stability of the coupled model at different time steps or periods. In Figure (3), the horizontal axis represents the z-axis coordinate in space, used to describe the spatial position of the system in the vertical direction, with the length unit being Å (angstroms). The vertical axis represents the velocity (m / s). The curves show a high degree of agreement between the macroscopic and microscopic scales. In Figure (3), LBM-1 to LBM-3 represent the calculation results of the LBM part of the coupled model at three different time steps: 600 ps, 1.6 ns, and 4.4 ns. MD-1 to MD-3 represent the calculation results of the MD part of the coupled model at the same macroscopic scale. In Figure (4), the horizontal axis represents the z-axis coordinate in space, used to describe the spatial position of the system in the vertical direction, with the length unit being Å (angstroms). The vertical axis represents the temperature (K). The results of the coupled algorithm and the pure MD are basically consistent at different periods. In Figure (4), LBM-1 to LBM-3 represent the calculation results of the LBM part of the coupled model under three different periods of 20, 70, and 134, while MD-1 to MD-3 represent the calculation results of the MD part of the coupled model under the same conditions. This shows that the coupled model can not only accurately simulate local microscopic phenomena, but also accurately reflect the macroscopic velocity and temperature gradient distribution of the fluid.
[0119] The molecular dynamics and lattice Boltzmann domain decomposition coupling device provided by the present invention will be described below. The molecular dynamics and lattice Boltzmann domain decomposition coupling device described below can be referred to in correspondence with the molecular dynamics and lattice Boltzmann domain decomposition coupling method described above.
[0120] Figure 5 This is a schematic diagram of the molecular dynamics and lattice Boltzmann domain decomposition coupling device provided by the present invention, as shown below. Figure 5 As shown, the molecular dynamics and lattice Boltzmann domain decomposition coupling device 500 includes the following modules: The computational domain partitioning module 510 is used to divide the computational domain into a molecular dynamics region corresponding to the upper part of the solid-liquid interface, a lattice Boltzmann region corresponding to the bulk fluid region far away from the solid-liquid interface, and an overlapping region located between the molecular dynamics region and the lattice Boltzmann region. The velocity coupling module 520 is used to divide the overlapping region along the normal direction into a first velocity transfer region, a buffer zone, and a second velocity transfer region. In the first velocity transfer region, the velocity information of the molecular dynamics side is transferred to the lattice Boltzmann side, and in the second velocity transfer region, the velocity information of the lattice Boltzmann side is transferred to the molecular dynamics side. Through velocity exchange, the velocity profile is continuously connected between the two scale regions. The temperature coupling module 530 is used to statistically analyze the thermal conductivity of the molecular dynamics side and the thermal conductivity of the lattice Boltzmann side in the overlapping region, respectively, and take the average of the thermal conductivity of the molecular dynamics side and the thermal conductivity of the lattice Boltzmann side as the target heat flux. The target heat flux is applied to the molecular dynamics side and the lattice Boltzmann side. Through multiple coupling cycle iterations, the heat flux on both sides tends to be consistent, and the temperature distribution and velocity distribution at the convergence of the coupling model are output.
[0121] This invention divides the computational domain into a molecular-scale (MD) region above the solid-liquid interface, a molecular-scale (LBM) region in the main fluid region, and an overlapping region between the two. In the overlapping region, velocity coupling (three-part velocity transfer ensures continuous connection of velocity profiles) and temperature coupling (iterative application after heat flux averaging to make the heat flux on both sides tend to be consistent) are performed. On the one hand, MD is used only near the interface to resolve molecular-scale effects, while LBM is used in the main fluid region to ensure computational efficiency. On the other hand, the independent coupling of velocity and temperature in different regions ensures the continuous transfer of momentum and energy across scales. This balances the capture of interface physical accuracy with the improvement of overall computational efficiency, and simultaneously achieves the joint continuous coupling of velocity and temperature fields at the cross-scale interface.
[0122] According to the present invention, a molecular dynamics and lattice Boltzmann domain decomposition coupling device 500 is provided, wherein the velocity coupling module 520 is specifically used for: When the velocity information from the molecular dynamics side is transmitted to the lattice Boltzmann side in the first velocity transmission region, the molecular velocities within the molecular dynamics sampling units corresponding to the grid size of the lattice Boltzmann are statistically averaged, and the average velocity is used as the boundary condition for solving the lattice Boltzmann problem. When transmitting the velocity information from the lattice Boltzmann side to the molecular dynamics side in the second velocity transfer region, the average velocity of the molecules in the molecular dynamics region is gradually approached by velocity state constraint or velocity recalibration method. By directly exchanging velocities, the velocity profile gradually achieves a continuous and smooth transition between the two scale regions.
[0123] According to the present invention, a molecular dynamics and lattice Boltzmann region decomposition coupling device 500 is provided, wherein the first velocity transfer region is a portion close to the molecular dynamics region, the second velocity transfer region is a portion close to the lattice Boltzmann region, and the buffer zone is an intermediate region.
[0124] According to the molecular dynamics and lattice Boltzmann domain decomposition coupling method provided by the present invention, the temperature coupling module 530 is specifically used for: The heat conduction flux on the molecular dynamics side is obtained by removing the convective energy transport term and the stress work term from the total energy flux obtained from the molecular dynamics simulation statistics. A temperature gradient is constructed based on the temperature difference between the inner and outer wall nodes, and the heat conduction flux on the Boltzmann side of the lattice is determined based on the temperature gradient and Fourier's law.
[0125] According to the molecular dynamics and lattice Boltzmann regio decomposition coupling device 500 provided by the present invention, the temperature coupling module 530 is further used for: The target heat flux is subjected to an exponential moving average to obtain the processed target heat flux; The processed target heat flux is gradually applied to the molecular dynamics side and the lattice Boltzmann side by means of the relaxation coefficient, and the heat flux on both sides gradually becomes consistent and the interface temperature difference tends to stabilize through the multiple coupling cycle iterations.
[0126] According to the molecular dynamics and lattice Boltzmann regio decomposition coupling device 500 provided by the present invention, the temperature coupling module 530 is further used for: Based on the relaxation coefficient, a preset step is advanced by a molecular dynamics system with a microscopic time step, and the treated target heat flux is applied to the boundary region; Based on the relaxation coefficient, a preset step is advanced by a lattice Boltzmann system with a macroscopic time step, and the processed target heat flux is applied at the corresponding boundary node.
[0127] According to the present invention, a molecular dynamics and lattice Boltzmann domain decomposition coupling device 500 is provided, wherein the molecular dynamics and lattice Boltzmann domain decomposition coupling device 500 further includes an energy compensation module; The energy compensation module is used for: After velocity coupling, the energy increment caused by velocity coupling is statistically analyzed, and energy compensation is performed based on the energy increment using the thermal motion velocity scaling method. The particle velocity is decomposed into a local average velocity and a thermal velocity relative to the local average velocity; The thermal motion velocity is scaled while keeping the local average velocity constant.
[0128] According to the present invention, a molecular dynamics and lattice Boltzmann domain decomposition coupling device 500 is provided, wherein the liquid in the molecular dynamics domain is described by a coarse-grained water model, the solid wall is made of silicon material, and the solid-liquid interaction is described by the Lennard-Jones potential function.
[0129] According to the present invention, a molecular dynamics and lattice Boltzmann domain decomposition coupling device 500 is provided, wherein in the lattice Boltzmann domain, the flow field adopts the D3Q19 lattice model and the BGK single relaxation scheme to solve the incompressible flow equation; the temperature field adopts a dual distribution function model, in addition to the density distribution function, a temperature distribution function is set to solve the heat transport process.
[0130] Figure 6 This is a schematic diagram of the structure of the electronic device provided by the present invention, such as... Figure 6 As shown, the electronic device may include a processor 610, a communications interface 620, a memory 630, and a communication bus 640. The processor 610, communications interface 620, and memory 630 communicate with each other via the communication bus 640. The processor 610 can call logical instructions from the memory 630 to execute molecular dynamics and lattice Boltzmann domain decomposition coupling methods.
[0131] Furthermore, the logical instructions in the aforementioned memory 630 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.
[0132] On the other hand, the present invention also provides a computer program product, which includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer is able to execute the molecular dynamics and lattice Boltzmann domain decomposition coupling methods provided by the above methods.
[0133] In another aspect, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, is implemented to perform the molecular dynamics and lattice Boltzmann domain decomposition coupling methods provided by the methods described above.
[0134] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.
[0135] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.
[0136] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims
1. A coupling method for molecular dynamics and lattice Boltzmann domain decomposition, characterized in that, include: The computational domain is divided into a molecular dynamics region corresponding to the upper part of the solid-liquid interface, a lattice Boltzmann region corresponding to the bulk fluid region far from the solid-liquid interface, and an overlapping region between the molecular dynamics region and the lattice Boltzmann region. The overlapping region is divided into a first velocity transfer region, a buffer zone, and a second velocity transfer region along the normal direction. In the first velocity transfer region, the velocity information from the molecular dynamics side is transferred to the lattice Boltzmann side, and in the second velocity transfer region, the velocity information from the lattice Boltzmann side is transferred to the molecular dynamics side. Through velocity exchange, the velocity profile is continuously connected between the two scale regions. Within the overlapping region, the thermal conductivity of the molecular dynamics side and the thermal conductivity of the lattice Boltzmann side are statistically analyzed respectively. The average thermal conductivity of the molecular dynamics side and the thermal conductivity of the lattice Boltzmann side is taken as the target thermal conductivity. The target thermal conductivity is applied to the molecular dynamics side and the lattice Boltzmann side. Through multiple coupling cycle iterations, the thermal conductivity on both sides tends to be consistent, and the temperature distribution and velocity distribution at the convergence of the coupling model are output.
2. The molecular dynamics and lattice Boltzmann domain decomposition coupling method according to claim 1, characterized in that, The process of transferring velocity information from the molecular dynamics side to the lattice Boltzmann side in the first velocity transfer region, and transferring velocity information from the lattice Boltzmann side to the molecular dynamics side in the second velocity transfer region, thereby achieving a continuous connection of the velocity profile between the two scale regions through velocity exchange, includes: When the velocity information from the molecular dynamics side is transmitted to the lattice Boltzmann side in the first velocity transmission region, the molecular velocities within the molecular dynamics sampling units corresponding to the grid size of the lattice Boltzmann are statistically averaged, and the average velocity is used as the boundary condition for solving the lattice Boltzmann problem. When transmitting the velocity information from the lattice Boltzmann side to the molecular dynamics side in the second velocity transfer region, the average velocity of the molecules in the molecular dynamics region is gradually approached by velocity state constraint or velocity recalibration method. By directly exchanging velocities, the velocity profile gradually achieves a continuous and smooth transition between the two scale regions.
3. The molecular dynamics and lattice Boltzmann domain decomposition coupling method according to claim 1, characterized in that, The first velocity transfer region is the portion near the molecular dynamics region, the second velocity transfer region is the portion near the lattice Boltzmann region, and the buffer zone is the middle region.
4. The molecular dynamics and lattice Boltzmann domain decomposition coupling method according to claim 1, characterized in that, The step of separately calculating the thermal conductivity on the molecular dynamics side and the thermal conductivity on the lattice Boltzmann side within the overlapping region includes: The heat conduction flux on the molecular dynamics side is obtained by removing the convective energy transport term and the stress work term from the total energy flux obtained from the molecular dynamics simulation statistics. A temperature gradient is constructed based on the temperature difference between the inner and outer wall nodes, and the heat conduction flux on the Boltzmann side of the lattice is determined based on the temperature gradient and Fourier's law.
5. The molecular dynamics and lattice Boltzmann domain decomposition coupling method according to claim 1, characterized in that, The process of applying the target heat flux to the molecular dynamics side and the lattice Boltzmann side, and aligning the heat fluxes on both sides through multiple coupling cycle iterations, includes: The target heat flux is subjected to an exponential moving average to obtain the processed target heat flux; The processed target heat flux is gradually applied to the molecular dynamics side and the lattice Boltzmann side by means of the relaxation coefficient, and the heat flux on both sides gradually becomes consistent and the interface temperature difference tends to stabilize through the multiple coupling cycle iterations.
6. The molecular dynamics and lattice Boltzmann domain decomposition coupling method according to claim 5, characterized in that, The stepwise application of the treated target heat flux to the molecular dynamics side and the lattice Boltzmann side via the relaxation coefficient includes: Based on the relaxation coefficient, a preset step is advanced by a molecular dynamics system with a microscopic time step, and the treated target heat flux is applied to the boundary region; Based on the relaxation coefficient, a preset step is advanced by a lattice Boltzmann system with a macroscopic time step, and the processed target heat flux is applied at the corresponding boundary node.
7. The molecular dynamics and lattice Boltzmann domain decomposition coupling method according to any one of claims 1-6, characterized in that, The method further includes: After velocity coupling, the energy increment caused by velocity coupling is statistically analyzed, and energy compensation is performed based on the energy increment using the thermal motion velocity scaling method. The particle velocity is decomposed into a local average velocity and a thermal velocity relative to the local average velocity; The thermal motion velocity is scaled while keeping the local average velocity constant.
8. The molecular dynamics and lattice Boltzmann domain decomposition coupling method according to any one of claims 1-6, characterized in that, The liquid in the molecular dynamics region is described using a coarse-grained water model, the solid wall is made of silicon, and the solid-liquid interaction is described using the Lennard-Jones potential function.
9. The molecular dynamics and lattice Boltzmann domain decomposition coupling method according to any one of claims 1-6, characterized in that, In the lattice Boltzmann region, the flow field uses the D3Q19 lattice model and the BGK single relaxation scheme to solve the incompressible flow equations; the temperature field uses a dual distribution function model, with a temperature distribution function set in addition to the density distribution function to solve the heat transport process.
10. A molecular dynamics and lattice Boltzmann domain decomposition coupling device, characterized in that, include: The computational domain partitioning module is used to divide the computational domain into a molecular dynamics region corresponding to the upper part of the solid-liquid interface, a lattice Boltzmann region corresponding to the bulk fluid region far from the solid-liquid interface, and an overlapping region located between the molecular dynamics region and the lattice Boltzmann region. The velocity coupling module is used to divide the overlapping region into a first velocity transfer region, a buffer zone, and a second velocity transfer region along the normal direction. In the first velocity transfer region, the velocity information of the molecular dynamics side is transferred to the lattice Boltzmann side, and in the second velocity transfer region, the velocity information of the lattice Boltzmann side is transferred to the molecular dynamics side. Through velocity exchange, the velocity profile is continuously connected between the two scale regions. The temperature coupling module is used to statistically analyze the thermal conductivity of the molecular dynamics side and the lattice Boltzmann side in the overlapping region, respectively. The average thermal conductivity of the molecular dynamics side and the thermal conductivity of the lattice Boltzmann side is taken as the target thermal conductivity. The target thermal conductivity is applied to the molecular dynamics side and the lattice Boltzmann side. Through multiple coupling cycle iterations, the thermal conductivity on both sides tends to be consistent, and the temperature distribution and velocity distribution at the convergence of the coupling model are output.
