A method for controlling particle transport in a micro-nanofluidic system

Abstract

The application discloses a particle transport control method in a micro-nano fluid system, comprising the following steps: dividing a calculation domain of the micro-nano fluid system into multiple sub-domains, including an atomic domain, a continuous domain and a common region where the two coexist; defining a boundary condition of the common region, transmitting physical quantity information in the atomic domain to the continuous domain by defining a constraint condition on the common region, and transmitting fluid velocity and pressure information in the continuous domain back to the atomic domain; adopting an iterative solution strategy to update the physical quantity information of the atomic domain and the continuous domain; and applying an external field factor to verify the particle transport control performance in the micro-nano fluid system based on the multi-scale coupling calculation method. The application solves the technical problems of low particle cross-domain transport precision, large dynamic flow field prediction deviation and low simulation efficiency in the existing micro-nano fluid system.

Description

Technical Field

[0001] This invention relates to the field of multi-scale simulation and particle transport control technology for micro-nano fluid control systems, and particularly to a particle transport control method for micro-nano fluid control systems. Background Technology

[0002] With the development of multiscale simulation technology, coupled computational models of molecular dynamics and continuum methods have gradually become an important direction for studying complex fluid systems, and have been widely applied in fields such as microfluidic chips, nanomaterial transport, and biofluid simulation. Multiscale simulation technology can be traced back to the 1970s, and its core lies in building an information exchange mechanism between different scales to achieve unified modeling of multiscale fluid behavior.

[0003] Domestic and international researchers have conducted a series of explorations in the theoretical and applied research of multiscale simulation technology, especially the coupling of molecular dynamics and continuum methods. For example, by introducing a "hybrid model," the computational domain is divided into an atomic domain and a continuous domain, and a specific coupling algorithm is used to achieve flux conservation at the interface between the two. In a multiscale method for coupling the continuous domain and the atomic domain suitable for micro-nano fluid flow, a coupling strategy is proposed, which achieves consistency of information exchange between different computational scales by using constrained dynamics in the overlapping region. This successfully simulates the micro-nano scale fluid flow problem and compares the results with analytical solutions and all-atom molecular dynamics simulations. A concurrent coupling method is proposed for soft multifunctional surface flow, combining molecular dynamics and computational fluid dynamics to simulate fluid flow in blood vessels. Regarding micro-nano scale fluid flow, the current status and development prospects of multiscale computational methods based on coupled computational models of molecular dynamics and computational fluid dynamics are discussed, especially the theoretical basis of coupled atomic-continuous domain modeling in computational domains with domain decomposition.

[0004] These studies have not only promoted the development of multi-scale simulation methods, but also provided theoretical support for engineering applications in related fields. However, the existing technologies lack adaptability and accuracy in specific application scenarios of micro-nanofluidics, and have failed to design solutions for specific problems such as boundary disturbances, scale mismatch, and load balancing in micro-nanofluidics. Therefore, it is urgent to propose a particle transport control method in micro-nanofluidics systems to solve the technical problems of low cross-domain particle transport accuracy, large dynamic flow field prediction deviation, and low simulation efficiency in existing micro-nanofluidics systems. Summary of the Invention

[0005] The main objective of this invention is to propose a particle transport control method for micro / nano fluid control systems, aiming to solve the technical problems of low cross-domain particle transport accuracy, large dynamic flow field prediction deviation, and low simulation efficiency in existing micro / nano fluid control systems.

[0006] To achieve the above objectives, the present invention provides a particle transport control method in a micro / nanofluidic system, wherein the particle transport control method in the micro / nanofluidic system includes the following steps:

[0007] S1. Divide the computational domain of the micro / nanofluidic system into multiple subdomains, including the atomic domain, the continuous domain, and the common region where the two coexist;

[0008] S2. Define the boundary conditions of the common region. By defining the constraints on the common region, the physical quantity information in the atomic domain is transmitted to the continuous domain, and the fluid velocity and pressure information in the continuous domain is transmitted back to the atomic domain.

[0009] S3. Use an iterative solution strategy to update the physical quantity information of the atomic domain and the continuous domain;

[0010] S4. Apply external field factors to verify the particle transport control performance of the micro / nano fluid control system based on the multi-scale coupling calculation method.

[0011] One preferred embodiment is that the common area includes buffer 1 area, CFD→MD area, buffer 2 area and MD→CFD area;

[0012] The buffer 1 region is used to store the particle cross-boundary motion in the y direction caused by periodic boundary adjustment during electrostatic force calculation;

[0013] The CFD→MD region is used to set the MD boundary based on the CFD results;

[0014] The buffer 2 area is used to set boundary adjustments to form a spatial buffer;

[0015] The MD→CFD region is used to set the CFD boundary based on the MD results.

[0016] One preferred embodiment is that the MD→CFD region is used to set the CFD boundary based on the MD results, specifically as follows:

[0017] The velocity distribution in the MD main computational region is calculated as the velocity boundary condition of the CFD main computational region. This is achieved by applying the velocity distribution in the MD→CFD region. Atom velocity By performing an arithmetic average, the velocity boundary conditions of the CFD main computation region are obtained, that is, the boundary layer velocity of the CFD main computation region is:

[0018]

[0019] in, This represents the boundary layer velocity of the CFD main computation region.

[0020] In one preferred embodiment, the CFD→MD region is used to implement MD boundary setting based on CFD results. Specifically, the velocity distribution in the main CFD computation region is calculated as the velocity boundary condition of the main MD computation region, and the particles in the atomic domain are coupled to the solution in the continuous domain through constraint dynamics.

[0021] One preferred embodiment is that the process of coupling particles in the atomic domain to solutions in the continuous domain via constrained dynamics specifically involves:

[0022] The CFD→MD region is divided into multiple bin regions in the y direction, and each bin region contains multiple continuum mesh elements.

[0023] Set constraints; the constraints are: the average continuous velocity of the i-th bin region in the CFD→MD region is equal to the average velocity of multiple continuous volume grid cells within the i-th bin region; the average continuous velocity of the i-th bin region is:

[0024]

[0025] in, Let be the average continuous velocity of the i-th bin region. Let be the number of continuum grid cells in the i-th bin region of the CFD→MD region. For the first The velocity of each continuum grid cell;

[0026] Based on the constraints, update the velocity of atom j in the i-th bin region of the CFD→MD region.

[0027] One preferred option is that the velocity of atom j in the i-th bin region of the CFD→MD region after the update is:

[0028]

[0029] in, The updated velocity of atom j in the i-th bin region of the CFD→MD region. Let the velocity of atom j be _____. To constrain the length, Let V be the average continuous velocity of the bin region where atom j is located in the continuous domain. Let i be the number of atoms in the i-th bin region where atom j is located in the CFD→MD region. Let be the velocity of the k-th atom.

[0030] One preferred embodiment, step S3, specifically includes:

[0031] Each iteration first extracts the particle's velocity and position vector from the atomic domain, and then passes the particle's velocity and position vector to the continuous domain to update the fluid's boundary conditions;

[0032] The continuity and momentum equations are solved in the continuous domain using the PISO algorithm, and the results are passed to the atomic domain to update the forces acting on the particles.

[0033] One preferred embodiment is that the continuity equation is solved using the PISO algorithm, specifically as follows:

[0034] Applying Gauss's theorem to the continuity equation, the volume integral is transformed into the flux integral of the control volume surface, and then discretized; the continuity equation is:

[0035]

[0036] The conversion of the volume integral into the flux integral of the controlled volume surface specifically involves:

[0037]

[0038] in, For fluid density, For fluid velocity vector, For time, To control volume, To control the surface volume, It is the unit outward normal vector.

[0039] One preferred approach is to solve the momentum equation using the PISO algorithm, specifically:

[0040] In the continuous domain, Gauss's theorem is applied to the momentum equation to transform the volume integral into the flux integral of the control volume surface. This is then discretized, and during discretization, the continuous field quantities of the momentum equation are distributed to the discrete control volume, ensuring a smooth mapping of the pressure field from the continuous domain to the atomic domain. This allows particles to receive an equivalent pressure gradient force at the coupling boundary. The momentum equation is:

[0041]

[0042] The conversion of the volume integral into the flux integral of the controlled volume surface specifically involves:

[0043]

[0044] in, For fluid velocity vector, To control volume, To control the surface volume, The unit outward normal vector, For fluid density, For fluid pressure, For dynamic viscosity, It is an external force.

[0045] One preferred embodiment, step S4, specifically includes:

[0046] Numerical simulations of Cuyet flow and Poiseuille flow were performed using Lennard-Jones fluid dynamics.

[0047] In the above technical solution of the present invention, the particle transport control method in the micro / nanofluidic system includes the following steps: dividing the computational domain of the micro / nanofluidic system into multiple subdomains, including an atomic domain, a continuous domain, and a common region where both coexist; defining boundary conditions for the common region, transferring physical quantity information from the atomic domain to the continuous domain by defining constraint conditions for the common region, and transferring fluid velocity and pressure information from the continuous domain back to the atomic domain; updating the physical quantity information of the atomic domain and the continuous domain using an iterative solution strategy; applying external field factors to verify the particle transport control performance of the micro / nanofluidic system based on the multi-scale coupled computation method. The present invention solves the technical problems of low cross-domain particle transport accuracy, large dynamic flow field prediction deviation, and low simulation efficiency in existing micro / nanofluidic systems.

[0048] In this invention, a precise boundary coupling mechanism avoids boundary interference and transmission errors in the particle transport process; and through an iterative solution strategy, physical quantities such as velocity magnitude, pressure magnitude, and position vector in the atomic domain and continuous domain are continuously updated, thereby accurately describing the particle transport process in the micro-nanofluidic system and ensuring that the particle transport prediction is consistent with the actual micro-nanofluidic scenario. Attached Figure Description

[0049] To more clearly illustrate the technical solutions in the embodiments of the present 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 only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the structures shown in these drawings without creative effort.

[0050] Figure 1 This is a first schematic diagram of a particle transport control method in a micro / nanofluidic system according to an embodiment of the present invention;

[0051] Figure 2 This is a schematic diagram of the layout of the public area in an embodiment of the present invention;

[0052] Figure 3 This is a schematic diagram of the layout of the common area in the existing multi-scale coupled calculation method of the present invention.

[0053] Figure 4This is a Coulter velocity distribution diagram of Lennard-Jones fluid according to an embodiment of the present invention;

[0054] Figure 5 This is a diagram showing the velocity distribution of the Lennard-Jones fluid in a Poisson's blade flow according to an embodiment of the present invention.

[0055] Figure 6 This is a velocity distribution diagram of a charged fluid in a Coeite flow according to an embodiment of the present invention.

[0056] The realization of the objective, functional characteristics and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0057] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0058] The technical solutions of the various embodiments of the present invention can be combined with each other, but this must be based on the ability of those skilled in the art to implement them. When the combination of technical solutions is contradictory or cannot be implemented, it should be considered that such combination of technical solutions does not exist and is not within the scope of protection claimed by the present invention.

[0059] See Figure 1 According to one aspect of the present invention, a particle transport control method in a micro / nanofluidic system is provided, wherein the particle transport control method in the micro / nanofluidic system includes the following steps:

[0060] S1. Divide the computational domain of the micro / nanofluidic system into multiple subdomains, including the atomic domain, the continuous domain, and the common region where the two coexist;

[0061] S2. Define the boundary conditions of the common region. By defining the constraints on the common region, the physical quantity information in the atomic domain is transmitted to the continuous domain, and the fluid velocity and pressure information in the continuous domain is transmitted back to the atomic domain.

[0062] S3. Use an iterative solution strategy to update the physical quantity information of the atomic domain and the continuous domain;

[0063] S4. Apply external field factors to verify the particle transport control performance of the micro / nano fluid control system based on the multi-scale coupling calculation method.

[0064] Specifically, in this embodiment, see Figures 2-3In multi-scale coupled computational methods, the common region is often divided into CFD→MD region, buffer region, and MD→CFD region. In micro-nano fluidic scenarios, particle (such as drug molecules and ions) transport faces dual interferences. First, particles move across the boundary in the y-direction due to periodic boundary adjustment, directly interfering with the velocity transfer from MD to CFD. Second, the bidirectional data communication between CFD→MD and MD→CFD causes crosstalk, resulting in the inability to synchronize the velocity of the continuous domain with that of the atomic domain. To avoid the occurrence of dual interferences, the common region described in this invention includes buffer1 region, CFD→MD region, buffer2 region, and MD→CFD region.

[0065] The buffer 1 region is used to store the cross-boundary motion of particles in the y-direction caused by periodic boundary adjustment during electrostatic force calculation. The buffer 1 region is a region specifically set for MD calculation. During velocity coupling, the examples in the buffer 1 region are not considered, thereby eliminating the interference of cross-boundary motion on the velocity of the electrolyte solution and reducing velocity transmission error.

[0066] The CFD→MD region is used to set the MD boundary based on the CFD results, that is, the velocity distribution of CFD in this region is used as the velocity boundary condition of the main MD calculation area.

[0067] The buffer 2 region uses the Schwartz coupling approach to set boundary adjustments to form a spatial buffer; the setting of the buffer 2 region can ensure that the communication between the CFD→MD region and the MD→CFD region is relatively independent and not interfered with by each other.

[0068] The MD→CFD region is used to set the CFD boundary based on the MD results. That is, the velocity distribution of MD in this region is used as the velocity boundary condition of the main CFD calculation region, so as to reflect the velocity change in the ion enrichment region of the electrolyte solution and avoid the "velocity lag" of the static boundary.

[0069] Specifically, in this embodiment, the MD→CFD region is used to set the CFD boundary based on the MD result, specifically as follows:

[0070] The velocity distribution in the MD main computational region is calculated as the velocity boundary condition of the CFD main computational region. This is achieved by applying the velocity distribution in the MD→CFD region. Atom velocity By performing an arithmetic average, the velocity boundary conditions of the CFD main computation region are obtained, that is, the boundary layer velocity of the CFD main computation region is:

[0071]

[0072] in, This represents the boundary layer velocity of the CFD main computation region.

[0073] Specifically, in this embodiment, the CFD→MD region is used to realize the MD boundary setting based on the CFD results. Specifically, the velocity distribution in the main CFD computation region is calculated as the velocity boundary condition of the main MD computation region, and the particles in the atomic domain are coupled to the solution in the continuous domain through constraint dynamics.

[0074] Specifically, in this embodiment, the coupling of particles in the atomic domain to the solution in the continuous domain via constrained dynamics is as follows:

[0075] The CFD→MD region is divided into multiple bin regions in the y direction, and each bin region contains multiple continuum mesh elements.

[0076] Set constraints; the constraints are: the average continuous velocity of the i-th bin region in the CFD→MD region is equal to the average velocity of multiple continuous volume grid cells within the i-th bin region; the average continuous velocity of the i-th bin region is:

[0077]

[0078] in, Let be the average continuous velocity of the i-th bin region. Let be the number of continuum grid cells in the i-th bin region of the CFD→MD region. For the first The velocity of each continuous grid cell; during particle transport, averaging the velocity across multiple grid cells can effectively prevent trajectory deviations caused by sudden velocity changes in charge-rich regions.

[0079] Based on the constraints, update the velocity of atom j in the i-th bin region of the CFD→MD region; the updated velocity of atom j in the i-th bin region of the CFD→MD region is:

[0080]

[0081] in, The updated velocity of atom j in the i-th bin region of the CFD→MD region. The velocity of atom j is obtained by performing a standard MD numerical integration on the atomic acceleration. To constrain the length, Let V be the average continuous velocity of the bin region where atom j is located in the continuous domain. Let i be the number of atoms in the i-th bin region where atom j is located in the CFD→MD region. The velocity of the k-th atom; In this invention, the constraint length is 1, indicating that the atomic velocity is corrected entirely according to the continuous velocity, overcoming the limitations of existing technologies. This addresses the time difference in velocity exchange between MD and CFD caused by the inter-value selection, thereby adapting to particle transport processes in micro / nano fluid control systems and avoiding particle velocity response delays.

[0082] Specifically, in this embodiment, in existing micro / nanofluidic systems, the scale difference between MD and CFD can lead to significant errors in the process of using multi-scale coupled computational methods to achieve particle transport control in micro / nanofluidic systems: First, there is a time scale mismatch; the characteristic time scale of particle velocity autocorrelation in MD is t. vv =0.14τ, the characteristic timescale described by the CFD continuous domain is △x△z / μ ~ The difference between 11τ and CFD lies in their magnitudes. Direct data exchange can lead to inaccurate CFD calculations due to significant statistical noise in the MD results, failing to reflect the true particle transport patterns. Secondly, pressure and stress parameters are discontinuous at the coupling interface. In the absence of external forces, the pressure gradient is the core driving force for particle transport in low Reynolds number flow fields. Due to inconsistent spatial scales, pressure averaging is difficult, and interruptions in pressure transmission can result in insufficient mapping accuracy from discrete particles to continuous field variables, leading to a significant decrease in particle transport accuracy in subsequent calculations. Thirdly, particle dynamic tracking is inaccurate. During particle transport, the local flow velocity fluctuates with the dynamic changes in the electric field. Using static or low-frequency position update strategies can cause partial loss of the particle's true trajectory, making it impossible to analyze its transport path. The iterative solution in this invention uses an alternating iterative approach. In each time step, MD simulation is performed first, then the results are transferred to CFD simulation, then CFD simulation is performed again, and the results are transferred back to MD simulation. This continuous data transfer and iteration between MD and CFD continues until a certain convergence condition is met, thereby eliminating the bias caused by scale mismatch.

[0083] Specifically, in this embodiment, the iterative solution over time typically requires coordinating the time steps of two different simulations to ensure the stability and accuracy of the coupled system. Clearly, there is a difference of two orders of magnitude between these two scales in this invention. For data exchange in overlapping regions, a specific time interval MΔt is used. MD Time averaging was performed to reduce statistical noise in the MD results. Numerical results show that when M=100, the two time scales are comparable and can produce appropriate coupling.

[0084] Specifically, in this embodiment, to achieve control of particle transport in the micro / nanofluidic system, it is necessary to ensure consistency of physical quantities such as particle velocity, pressure, and position vector between the atomic domain and the continuous domain; step S3 specifically involves:

[0085] Each iteration first extracts the particle's velocity and position vector from the atomic domain, and then passes the particle's velocity and position vector to the continuous domain to update the fluid's boundary conditions;

[0086] The continuity and momentum equations are solved in the continuous domain using the PISO algorithm, and the results are passed to the atomic domain to update the forces acting on the particles. Through this iterative process, the velocity and force can be updated and the particle position can be corrected, thereby accurately describing the particle transport process in micro / nanofluidic systems, such as flow field fluctuations caused by sugar chain extension / folding.

[0087] Specifically, in this embodiment, solving the continuity equation using the PISO algorithm involves:

[0088] Applying Gauss's theorem to the continuity equation, the volume integral is transformed into the flux integral of the control volume surface, and then discretized; the continuity equation is:

[0089]

[0090] The conversion of the volume integral into the flux integral of the controlled volume surface specifically involves:

[0091]

[0092] in, For fluid density, For fluid velocity vector, For time, To control volume, To control the surface volume, It is the unit outward normal vector.

[0093] Specifically, in this embodiment, the momentum equation is solved using the PISO algorithm, as follows:

[0094] In the continuous domain, Gauss's theorem is applied to the momentum equation to transform the volume integral into the flux integral of the control volume surface. This is then discretized, with the continuous variables of the momentum equation allocated to the discrete control volume during discretization. This ensures a smooth mapping of the pressure field from the continuous domain to the atomic domain, allowing particles to receive an equivalent pressure gradient force at the coupling boundary. Pressure data is transmitted back to the computational domain in real time to update the pressure gradient force on the particles, thus avoiding "particle accumulation" caused by static pressure. The momentum equation is:

[0095]

[0096] The conversion of the volume integral into the flux integral of the controlled volume surface specifically involves:

[0097]

[0098] in, For fluid velocity vector, To control volume, To control the surface volume, The unit outward normal vector, For fluid density, For fluid pressure, For dynamic viscosity, It is an external force.

[0099] Specifically, in this embodiment, the micro / nanofluidic system simulation needs to handle multi-component systems with up to 10 atoms. 6 -10 8 In terms of scale, existing technologies face two major bottlenecks: First, static task allocation results in the computational load of the atomic domain being 1.2-2.5 times that of the continuous domain, causing an overall load imbalance and insufficient parallel efficiency. Second, common areas require explicit processor allocation, with communication latency accounting for 30% of the total computation time, failing to meet the "multi-parameter rapid iteration" requirements of micro / nano fluid control devices. This invention employs a distributed memory architecture, utilizing a message passing interface to distribute computational tasks to multiple computing nodes. Each node is responsible for processing a specific subdomain, including both atomic and continuous domains, i.e., for a given number of processors N... p The entire continuous medium is decomposed into N in OpenFOAM. p Each region, and the entire atomic domain is also decomposed into N in LAMMPS. pEach processor can independently process its assigned subdomain, enabling parallel computing and significantly improving computational efficiency. In coupled molecular dynamics and computational fluid dynamics calculations, handling common regions is a critical issue. This is addressed by having all processors check if they contain a portion of a common region and by communicating between processors that do. Specifically, each processor communicates its local values ​​with other processors, and then calculates the average value of the common region through a reduction operation. This average value is stored as a variable (e.g., a velocity setpoint) and used as a setpoint in the calculation of another domain. This approach ensures the continuity and consistency of the common region while avoiding explicit allocation of it, thus simplifying the implementation of parallel computing. Compared to traditional serial computing, the distributed memory architecture allows for real-time optimization of subdomain ranges based on flow field changes (such as particle enrichment region shifts). For example, when particles migrate from the calyx layer to the channel center, the atomic domains adjust with the particle migration direction, ensuring a consistently balanced processor load and avoiding local overload.

[0100] Specifically, in this embodiment, external field factors, such as flow field and electric field, are applied to verify the particle transport control performance of a micro / nanofluidic system based on a multi-scale coupling calculation method. Existing multi-scale coupling schemes are only applicable to simple fluids, such as Lennard-Jones fluid, and cannot meet the actual scenarios of many micro / nanofluidic systems, such as the transport of charged particles / biomolecules. To verify the accuracy of particle transport control in the micro / nanofluidic system based on the multi-scale coupling calculation method of this invention, and thus greatly expand the applicability of the multi-scale coupling calculation method, numerical simulations of Cuyet flow and Poiseuille flow are performed using Lennard-Jones fluid. The simulation results are shown in [reference missing]. Figures 4-5 Couette flow describes fluid flow between two parallel plates, one of which is semi-fixed and the other moving parallel to it at a constant velocity. The entire fluid domain is initialized to zero velocity, and at t=0, the velocity of the solid wall suddenly increases to the wall velocity. The analytical solution for the suddenly initiated Cuete flow is:

[0101]

[0102] in, For the analytical solution of the Couette flow, For the wall velocity, The height of the fluid's location. To calculate the size of the domain in the y-direction, For the number of terms, Kinematic viscosity, For time;

[0103] Poiseuille flow describes the laminar flow of fluid between a circular pipe or parallel flat plates driven by a constant pressure gradient. This flow is also known as pipe flow and is a classic problem for studying viscous fluids in cylindrical pipes. The velocity distribution in the pipe flow can be solved using the Navier-Stokes equations, yielding the following velocity distribution:

[0104]

[0105] in, For the analytical solution of the Poisson flow, The height of the fluid region. For dynamic viscosity, The pressure gradient along the x-axis, The height of the fluid's location;

[0106] By comparing the results of particle transport control in a micro / nanofluidic system based on a multi-scale coupled computation method with the analytical solution, it can be found that the simulation results are well coupled in the atomic domain and continuous domain in the common region, and the trend is the same as that of the analytical solution. The error is within an acceptable range. Thus, the particle transport control in a micro / nanofluidic system based on a multi-scale coupled computation method is obtained.

[0107] After verifying the performance of particle transport control in a micro / nanofluidic system based on a multi-scale coupled computational method using Lennard-Jones fluid dynamics, multi-scale coupled computations were performed on charged fluids to simulate their flow. (See also...) Figure 6 The transient velocity of the charged fluid was calculated by simulating the Couette flow within a nanochannel. The coupling results between the atomic and continuous domains, and the comparison with the analytical solution, proved the accuracy of the momentum coupling of particle transport control in the common region of the micro / nanofluidic system based on the multi-scale coupling calculation method proposed in this invention. During the flow process, the simulation results agreed well with the analytical solution, indicating the feasibility of satisfying the nanoscale output characteristics of the charged fluid through the multi-scale coupling calculation method. This proves that the multi-scale coupling calculation method proposed in this patent can accurately reflect the regulatory effect of the electric field on the transport of charged particles. The effectiveness of the scheme in the specific field of micro / nanofluidic technology was ensured through Lennard-Jones fluid (basic flow) and charged fluid (complex physical flow), with the simulation results of the charged fluid deviating from the theoretical value by less than 10%.

Claims

1. A particle transport control method in a micro / nanofluidic system, characterized in that, Includes the following steps: S1. Divide the computational domain of the micro / nanofluidic system into multiple subdomains, including the atomic domain, the continuous domain, and the common region where the two coexist; S2. Define the boundary conditions of the common region. By defining the constraints on the common region, the physical quantity information in the atomic domain is transmitted to the continuous domain, and the fluid velocity and pressure information in the continuous domain is transmitted back to the atomic domain. S3. An iterative solution strategy is used to update the physical quantity information in the atomic and continuous domains; specifically: Each iteration first extracts the particle's velocity and position vector from the atomic domain, and then passes the particle's velocity and position vector to the continuous domain to update the fluid's boundary conditions; The continuity equation and momentum equation are solved in the continuous domain using the PISO algorithm, and the solution results are passed to the atomic domain to update the forces acting on the particles. The continuity equation is solved using the PISO algorithm, specifically as follows: Applying Gauss's theorem to the continuity equation, the volume integral is transformed into the flux integral of the control volume surface, and then discretized; the continuity equation is: The conversion of the volume integral into the flux integral of the controlled volume surface specifically involves: in, For fluid density, For fluid velocity vector, For time, To control volume, To control the surface volume, It is the unit outward normal vector; The momentum equation is solved using the PISO algorithm, specifically as follows: In the continuous domain, Gauss's theorem is applied to the momentum equation to transform the volume integral into the flux integral of the control volume surface. This is then discretized, and during discretization, the continuous field quantities of the momentum equation are distributed to the discrete control volume, ensuring a smooth mapping of the pressure field from the continuous domain to the atomic domain. This allows particles to receive an equivalent pressure gradient force at the coupling boundary. The momentum equation is: The conversion of the volume integral into the flux integral of the controlled volume surface specifically involves: in, For fluid velocity vector, To control volume, To control the surface volume, The unit outward normal vector, For fluid density, For fluid pressure, For dynamic viscosity, External force; S4. Apply external field factors to verify the particle transport control performance of the micro / nano fluid control system based on the multi-scale coupling calculation method.

2. The particle transport control method in a micro / nano fluidic system according to claim 1, characterized in that, The common area includes buffer 1 area, CFD→MD area, buffer 2 area and MD→CFD area; The buffer 1 region is used to store the particle cross-boundary motion in the y direction caused by periodic boundary adjustment during electrostatic force calculation; The CFD→MD region is used to set the MD boundary based on the CFD results; The buffer 2 area is used to set boundary adjustments to form a spatial buffer; The MD→CFD region is used to set the CFD boundary based on the MD results.

3. The particle transport control method in a micro / nano fluidic system according to claim 2, characterized in that, The MD→CFD region is used to define the CFD boundary based on the MD results, specifically: The velocity distribution in the MD main computational region is calculated as the velocity boundary condition of the CFD main computational region. This is achieved by applying the velocity distribution in the MD→CFD region. Atom velocity By performing an arithmetic average, the velocity boundary conditions of the CFD main computation region are obtained, that is, the boundary layer velocity of the CFD main computation region is: in, This represents the boundary layer velocity of the CFD main computation region.

4. The particle transport control method in a micro / nanofluidic system according to claim 2, characterized in that, The CFD→MD region is used to set the MD boundary based on the CFD results. Specifically, the velocity distribution in the main CFD computation region is calculated as the velocity boundary condition of the main MD computation region, and the particles in the atomic domain are coupled to the solution in the continuous domain through constraint dynamics.

5. The particle transport control method in a micro / nanofluidic system according to claim 4, characterized in that, The solution that couples particles from the atomic domain to the continuous domain via constrained dynamics is specifically as follows: The CFD→MD region is divided into multiple bin regions in the y direction, and each bin region contains multiple continuum mesh elements. Set constraints; the constraints are: the average continuous velocity of the i-th bin region in the CFD→MD region is equal to the average velocity of multiple continuous volume grid cells within the i-th bin region; the average continuous velocity of the i-th bin region is: in, Let be the average continuous velocity of the i-th bin region. Let be the number of continuum grid cells in the i-th bin region of the CFD→MD region. For the first The velocity of each continuum grid cell; Based on the constraints, update the velocity of atom j in the i-th bin region of the CFD→MD region.

6. The particle transport control method in a micro / nanofluidic system according to claim 5, characterized in that, The updated velocity of atom j in the i-th bin region of the CFD→MD region is: in, The updated velocity of atom j in the i-th bin region of the CFD→MD region. Let the velocity of atom j be denoted as . To constrain the length, Let V be the average continuous velocity of the bin region where atom j is located in the continuous domain. Let i be the number of atoms in the i-th bin region where atom j is located in the CFD→MD region. Let be the velocity of the k-th atom.

7. A particle transport control method in a micro / nanofluidic system according to any one of claims 1-6, characterized in that, Step S4 specifically includes: Numerical simulations of Cuyet flow and Poiseuille flow were performed using Lennard-Jones fluid dynamics.

Images