Aerosol particle behavior analysis method, aerosol particle behavior analysis device, aerosol particle safety evaluation method and aerosol particle safety evaluation system

Abstract

The invention discloses an aerosol particle behavior analysis method, an aerosol particle behavior analysis device, a safety evaluation method and a safety evaluation system. The method comprises the following steps: acquiring basic parameters of a to-be-analyzed fluid; constructing a geometric model of a computational domain of the fluid to be analyzed; carrying out meshing on the geometric model to obtain a grid unit data set of the geometric model; obtaining a wall surface function, a turbulence model, a discrete phase model and a continuous random walk model; based on the basic parameters and the grid unit data set, solving the wall function, the turbulence model, the discrete phase model and the continuous random walk model through a solver to obtain working condition distribution data of the to-be-analyzed fluid in the computational domain; and analyzing behaviors of the aerosol particles based on the working condition distribution data. Therefore, the motion behavior of the aerosol particles can be accurately simulated.

Description

Technical Field

[0001] The present invention belongs to the technical field of nuclear power, and particularly relates to a method and device for analyzing aerosol particle behavior, a safety assessment method and system. Background Art

[0002] When a severe accident occurs in a nuclear power plant, a large amount of radioactive fission products will be released into the safe atmosphere. Except for noble gases and some free iodine and ruthenium, most of the fission products exist in the form of aerosols. Therefore, the migration and evolution behavior of aerosols has always been one of the hot topics in nuclear safety research.

[0003] In the transport and deposition of aerosol particles in a turbulent fluid in a pipeline, turbophoresis is one of the main mechanisms of particle deposition. When using the Lagrangian method to simulate turbophoresis, the discrete random walk (DRW) model is usually used for simulation. However, in the wall-bounded turbulence, the discrete random walk model usually overestimates particle deposition because the model assumes that the root mean square values of the pulsating velocities in all directions are the same, which does not conform to the characteristic that the normal velocity pulsation in the wall direction is much higher than that in the mainstream direction. In addition, the way of simplifying the turbulent pulsation to discrete vortices in the discrete random walk model does not conform to the continuity characteristic of the actual turbulent pulsation.

[0004] Therefore, the accuracy of the existing technology in simulating the transport and deposition of aerosol particles in a turbulent fluid in a pipeline is relatively low. Summary of the Invention

[0005] The technical problem to be solved by the present invention is to provide a method and device for analyzing aerosol particle behavior, a safety assessment method and system, aiming at the above deficiencies of the existing technology, so as to improve the accuracy of simulating the motion behavior of aerosol particles.

[0006] In a first aspect, an embodiment of the present invention provides a method for analyzing aerosol particle behavior, including:

[0007] Obtaining basic parameters of a fluid to be analyzed, where the fluid to be analyzed includes aerosol particles;

[0008] Constructing a geometric model of a computational domain of the fluid to be analyzed;

[0009] Meshing the geometric model to obtain a grid cell data set of the geometric model;

[0010] Obtaining a wall function, a turbulence model, a discrete phase model, and a continuous random walk model, where the wall function is used to describe the flow characteristics in the turbulent near-wall region, and the continuous random walk model is used to simulate the continuity of the pulsating velocity of aerosol particles in the wall-bounded turbulent airflow;

[0011] Based on the basic parameters and the grid cell data set, the wall function, turbulence model, discrete phase model, and continuous random walk model are solved by a solver to obtain the condition distribution data of the fluid to be analyzed in the computational domain;

[0012] Based on the condition distribution data, the behavior of the aerosol particles is analyzed.

[0013] In a second aspect, an embodiment of the present invention further provides a safety assessment method for a nuclear power plant, which is applied to a containment vessel. The method includes:

[0014] According to the aerosol particle behavior analysis method described in the first aspect, analyze the behavior of aerosol particles in the containment vessel to obtain an analysis result;

[0015] Based on the analysis result, perform a safety assessment on the containment vessel.

[0016] In a third aspect, an embodiment of the present invention further provides an aerosol particle behavior analysis device, including:

[0017] The device includes:

[0018] A first acquisition module, configured to acquire basic parameters of the fluid to be analyzed, where the fluid to be analyzed includes aerosol particles;

[0019] A model construction module, configured to construct a geometric model of the computational domain of the fluid to be analyzed;

[0020] A grid division module, connected to the model construction module, configured to perform meshing on the geometric model to obtain a grid cell data set of the geometric model;

[0021] A second acquisition module, which acquires a wall function, a turbulence model, a discrete phase model, and a continuous random walk model. The wall function is used to describe the flow characteristics in the near-wall region of the turbulence, and the continuous random walk model is used to simulate the continuity of the pulsating velocity of aerosol particles in the turbulent airflow restricted by the wall;

[0022] A solving module, connected to the second acquisition module, configured to solve the wall function, the turbulence model, the discrete phase model, and the continuous random walk model based on the basic parameters and the grid cell data set through a solver to obtain the condition distribution data of the fluid to be analyzed in the computational domain;

[0023] An analysis module, connected to the solving module, configured to analyze the behavior of the aerosol particles based on the condition distribution data.

[0024] In a fourth aspect, an embodiment of the present invention further provides a safety assessment system for a nuclear power plant,

[0025] Applied to a containment vessel, the system includes:

[0026] The aerosol particle behavior analysis device according to the third aspect, which is used to analyze the behavior of aerosol particles in the containment vessel to obtain an analysis result;

[0027] An evaluation device, which is used to perform a safety assessment on the containment vessel based on the analysis result.

[0028] The aerosol particle behavior analysis method of the present invention accurately analyzes the behavior of aerosol particles by obtaining fluid basic parameters, constructing a geometric model, meshing, obtaining and solving wall functions, a turbulence model, a discrete phase model, and a continuous random walk model. By introducing the continuous random walk model, the continuity of the turbulent pulsation of aerosol particles is ensured, and the pulsation velocity of aerosol particles in wall-bounded turbulence is more accurately simulated, thereby improving the accuracy of the simulation results. Description of the Drawings

[0029] Figure 1 : Schematic diagram of the discrete random walk model in Embodiment 1 of the present invention;

[0030] Figure 2 : Flowchart of a method for analyzing the behavior of aerosol particles in Embodiment 1 of the present invention;

[0031] Figure 3 : Schematic diagram of particle tracking in the boundary layer in Embodiment 1 of the present invention;

[0032] Figure 4 : CRW model calculation process in Embodiment 1 of the present invention;

[0033] Figure 5 : Schematic diagram of the experimental device of Liu and Agarwal in Embodiment 1 of the present invention;

[0034] Figure 6 : Schematic diagram of the geometric model of the numerical simulation in Embodiment 1 of the present invention;

[0035] Figure 7 : Cross-sectional view of the generated mesh in Embodiment 1 of the present invention;

[0036] Figure 8 : Velocity contour map obtained from the model calculation in Embodiment 1 of the present invention;

[0037] Figure 9 : Schematic diagram of the particle distribution at the inlet in Embodiment 1 of the present invention;

[0038] Figure 10 : Schematic diagram of the particle distribution on different cross-sections in Embodiment 1 of the present invention;

[0039] Figure 11 : This is the structural diagram of an aerosol particle behavior analysis device according to Embodiment 3 of the present invention. Detailed implementation manners

[0040] To enable those skilled in the art to better understand the technical solutions of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments.

[0041] It should be noted that in this article, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements not only includes those elements, but also includes other elements not expressly listed, or also includes elements inherent to such process, method, article or device. Without further limitation, the elements defined by the statement "including..." do not exclude the presence of additional identical elements in the process, method, article or device including the said elements.

[0042] Computational Fluid Dynamics (CFD) is a combination of numerical calculation and computational technology, and is a method for numerically solving computational fluid problems.

[0043] FLUENT is the most widely used software in CFD software, mainly used for computer applications to simulate fluid flow and heat transfer. It has various different physical models and can be used to distinguish and judge fluids.

[0044] Source term: The source term is the term that represents the generation or consumption of additional mass, momentum, or energy in the fluid dynamics equation. In this application, the source term represents the loss of mass, momentum, and energy due to water vapor condensation.

[0045] In Computational Fluid Dynamics (CFD), in order to generate flow field data and consider multiple physical phenomena (such as turbulence, thermophoresis, and diffusiophoresis), it is usually necessary to integrate various models (such as turbulence models, thermophoresis models, diffusiophoresis models, etc.) into the overall solution framework. The SIMPLE (Semi-Implicit Method for Pressure Linked Equations) algorithm is one of the commonly used solution frameworks and is usually used to handle pressure-velocity coupling problems.

[0046] A large amount of radioactive fission products will be released into the safe atmosphere during a severe accident in a nuclear power plant. Except for noble gases and some free iodine and ruthenium, most fission products exist in the form of aerosols. Therefore, the migration and evolution behavior of aerosols has always been one of the hotspots in nuclear safety research, including the generation of aerosols, the transport and deposition of aerosols in pipelines, the transport and deposition of aerosols in the containment, etc.

[0047] In the transport and deposition of aerosol particles in turbulent fluid in a pipeline, turbophoresis is one of the main mechanisms of particle deposition. When using the Lagrangian method to simulate turbophoresis, the discrete random walk (DRW) model is currently widely used for simulation, and the ANSYS Fluent software uses this model to simulate turbophoresis. This model assumes that each particle has a series of interactions with turbulent vortices with finite lifetimes and scales, and the pulsating velocity can be obtained through the turbulent root mean square (RMS) statistical characteristic quantity.

[0048] As Figure 1 shown in the schematic diagram of the DRW model. At the initial moment of the vortex-particle interaction, the initial velocity of the particle is u p0 , and it is assumed that the particle is located at the center of the vortex. The characteristic scale of the vortex is L e , and the velocity is u e . During the interaction process, the velocity of the vortex remains unchanged. The vortex translates with the instantaneous velocity, but the particle will form another trajectory according to its motion equation, and the two are not the same. When the lifetime of the vortex terminates, or the particle passes through the vortex (i.e., the distance between the particle and the center of the vortex exceeds L e ), the interaction between the particle and the vortex ends, and it will interact with a new vortex. When using the two-equation turbulent k-ε model, the characteristic length L e and the velocity u e are calculated using the following relationships:

[0049]

[0050] where: C μ is a constant, and generally 0.09 can be taken.

[0051] The interaction between the vortex and the particle is reflected by the drag force caused by the velocity difference between the two. The RANS method obtains the time-averaged velocity field, and the instantaneous velocity can be obtained by superimposing the pulsating velocity, as follows:

[0052]

[0053] where: is the time-averaged velocity; u' i is the pulsating velocity.

[0054] The pulsating velocity u' i is calculated using the following formula:

[0055] u' i = ξ i σ i (4)

[0056] In the formula: ξ i is a random number subject to Gaussian normal distribution to reflect the random effect of turbulence; σ i is the root mean square value (RMS) of velocity pulsation, which is a statistic for measuring the intensity of fluid velocity fluctuation. In the k-ε model, σ i is calculated by the following formula:

[0057]

[0058] Note that for an isotropic turbulence model, σ i is the same in each direction. For the mainstream region far from the wall, this assumption is reasonable. However, for the wall-bounded turbulent fluid, experiments show that the velocity pulsation in the wall normal direction is much higher than that in the mainstream direction, and this assumption is no longer applicable at this time.

[0059] The interaction time between the particle and the vortex is the smaller of the vortex lifetime τ e and the particle transit time t cross and is calculated by the following formulas respectively:

[0060]

[0061] In the formula: T L is the Lagrangian integral time; C L is a constant. For the two-equation k-ε model and k-ω model, Fluent recommends taking 0.30, and for the RSM model, it can take 0.60.

[0062] Considering that the gas turbulence in the pipeline is wall-bounded flow and the turbulence in the boundary layer region is highly anisotropic, the traditional isotropic two-equation turbulence model is no longer applicable to the simulation of particle turbulent swimming. Conducting turbulent swimming simulation based on this turbulence model will significantly overestimate the particle deposition; however, the isotropic two-equation turbulence model has been widely used in engineering practice. Therefore, it is necessary to develop a turbulent swimming model based on the isotropic two-equation turbulence model but capable of reflecting the anisotropic characteristics of the boundary layer.

[0063] In addition, the discrete random walk model treats the turbulent pulsation as a series of discrete vortices, which does not conform to the continuity of the actual turbulent pulsation.

[0064] Therefore, through research, the inventors found that there are mainly two major defects in the existing discrete random walk (DRW) model when simulating the transport and deposition of radioactive aerosols after a nuclear power plant accident: First, in wall-bounded turbulence, since the model assumes that the root mean square values of the pulsation velocities in all directions are the same, which does not conform to the characteristic that the normal velocity pulsation on the wall is much higher than that in the mainstream direction, it will significantly overestimate particle deposition. Second, the treatment of simplifying turbulent pulsations as discrete vortices does not conform to the continuous characteristics of actual turbulent pulsations. These limitations seriously affect the prediction accuracy of the transport and deposition behavior of radioactive aerosols.

[0065] Example 1:

[0066] Based on the above research, to solve the above technical problems, as Figure 2 shown, this example provides an aerosol particle behavior analysis method including the following steps 101 to 106.

[0067] Step 101, obtain the basic parameters of the fluid to be analyzed, where the fluid to be analyzed includes aerosol particles.

[0068] Specifically, the basic parameters of the fluid to be analyzed can be obtained from the relevant data and standards of the nuclear power plant. These parameters include but are not limited to: fluid components: the proportions of non-condensable gases (such as water vapor, air, etc.) and the characteristics of aerosol particles; physical property parameters: the density, viscosity, specific heat, thermal conductivity, etc. of each component; environmental condition parameters: temperature, pressure, humidity, etc., and the parameters to be obtained can be selected according to the actual situation.

[0069] Step 102, construct a geometric model of the computational domain of the fluid to be analyzed.

[0070] Specifically, according to the specific research objectives and conditions, clarify the range of the computational domain of the fluid to be analyzed, including the flow path and boundaries of the fluid in the area to be analyzed; obtain the detailed geometric information of the computational domain, including boundary shapes, sizes, internal features, etc.; then a geometric model of the analysis area (computational domain) can be created using computational fluid dynamics (CFD) software (such as ANSYS Fluent, OpenFOAM, etc.) or 3D modeling software (such as SolidWorks, AutoCAD, etc.).

[0071] Step 103, mesh the geometric model to obtain a dataset of grid cells of the geometric model.

[0072] Specifically, use the mesh generation tools built into computational fluid dynamics (CFD) software (such as ANSYS Meshing, ICEM CFD, OpenFOAM, etc.) or dedicated mesh generation software (such as Gambit, Gmsh, etc.). According to the complexity of the geometric model and the requirements of analysis accuracy, set the relevant parameters for mesh generation, including the type of mesh (such as structured mesh or unstructured mesh), element size, growth rate, etc. Considering the fluid flow characteristics and aerosol particle behavior, perform mesh refinement on key regions (such as boundary layers, complex geometric regions) to improve the accuracy of calculations.

[0073] Run the mesh generation tool to generate the initial mesh elements. During this process, different mesh generation algorithms (such as quadrilateral, hexahedron, triangle, or tetrahedron elements) can be selected to adapt to different geometric structures. After generating the initial mesh, check the quality of the mesh elements to ensure that no distorted, overlapping, or deformed elements are formed. Use the built-in mesh quality inspection function of the software to check indicators such as the orthogonality, uniformity, and element volume change of the mesh elements; according to the quality inspection results, optimize and adjust the low-quality mesh regions, including manual correction or re-meshing, perform local mesh refinement or smoothing on necessary regions to improve the overall mesh quality and calculation stability; after completing the mesh optimization, save the generated mesh element dataset in the format required for numerical calculation (such as the.cas file of ANSYS Fluent, the mesh file of OpenFOAM, etc.).

[0074] The mesh element dataset includes the mesh information of the computational domain, such as node coordinates, element volume, element surface area, and their connection relationships, to ensure that the numerical solution has sufficient spatial resolution.

[0075] Through reasonable mesh generation and optimization, ensure that the mesh element dataset can accurately reflect the details of the geometric model and fluid characteristics, providing a reliable basis for subsequent numerical solutions, thereby ensuring the accuracy and reliability of the simulation results.

[0076] Step 104, obtain the wall function, turbulence model, discrete phase model, and continuous random walk model.

[0077] Among them, the wall function is used to describe the flow characteristics in the near-wall region of turbulence. Exemplarily, the logarithmic law or exponential law model can be adopted.

[0078] The continuous random walk model is used to simulate the continuity of the pulsating velocity of aerosol particles in the wall-bounded turbulent airflow.

[0079] Turbulence models are an important research direction in fluid mechanics, aiming to describe and predict the behavior of turbulence. Turbulence is a complex and irregular state in fluid motion, manifested as rapid changes in fluid parameters such as velocity and pressure in space and time. There are various types of turbulence models. Each turbulence model has its own characteristics and scope of application. The selection of an appropriate model mainly depends on the specific flow problem and computational resources. For example, the k-ε model or k-ω model is used to simulate turbulent flows.

[0080] The Discrete Phase Model (DPM) is a method in Computational Fluid Dynamics (CFD) for simulating the motion behavior of discrete phases (such as particles, droplets or bubbles) in a continuous phase (such as gas or liquid). The Discrete Phase Model (DPM) adopted in this application is based on the Euler-Lagrange method to perform computational simulations on the fluid phase and particle phase.

[0081] Through the combined action of these models, the turbulent characteristics of the fluid, the motion behavior of particles, and the details of the flow near the wall can be described respectively.

[0082] Step 105: Based on the basic parameters and the grid cell dataset, solve the wall function, turbulence model, discrete phase model, and continuous random walk model through a solver to obtain the condition distribution data of the fluid to be analyzed in the computational domain.

[0083] Solve the fluid dynamics equations through numerical methods (such as the finite volume method or the finite element method) to simulate the flow of the fluid and the behavior of aerosol particles. By integrating the above models and running the solver, the condition distribution data of the fluid can be obtained within the computational domain. These data usually include the velocity field, pressure field, temperature field, and the spatial distributions of the concentrations of each phase, the motion trajectories of aerosol particles, etc. Finally, these distribution data can be used for subsequent analysis and engineering applications to optimize the design and verify the flow phenomena.

[0084] Step 106: Analyze the behavior of aerosol particles based on the condition distribution data.

[0085] Once the condition distribution data of the fluid within the computational domain (including the behavior characteristics such as the deposition rate, concentration distribution, and motion trajectories of aerosol particles) are obtained through numerical simulation, the behavior of aerosol particles can be analyzed in detail.

[0086] For example, using the velocity field data, the motion trajectories of particles can be calculated and visualized to describe the migration paths of particles in the fluid; particle velocity: By analyzing the velocity changes of particles at different positions, the effect of fluid flow on particle acceleration or deceleration can be understood. Furthermore, the transport and deposition characteristics of aerosol particles in the fluid can be evaluated, providing a basis for safety assessment.

[0087] Finally, post-processing software such as ANSYS CFD-Post and ParaView can also be used to visualize the previous calculation results and perform detailed analysis (for example, generating velocity vector diagrams, temperature distribution diagrams, pressure distribution diagrams, particle trajectory diagrams, concentration distribution diagrams, etc.).

[0088] In this embodiment, by obtaining the basic parameters of the fluid, constructing a geometric model, meshing, obtaining and solving the wall function, turbulence model, discrete phase model, and continuous random walk model, the behavior of aerosol particles is accurately analyzed. By introducing the continuous random walk model, the pulsating velocity of aerosol particles in wall-bounded turbulence is more accurately simulated, thereby improving the accuracy of the simulation results.

[0089] Optionally, the above method further includes:

[0090] Constructing a continuous random walk model based on the Langevin equation;

[0091] Constructing a continuous random walk model based on the Langevin equation, specifically including:

[0092] According to the turbulent anisotropy characteristics in the geometric boundary layer, creating a boundary layer Langevin equation, which is used to correct the turbulent anisotropy characteristics near the wall surface;

[0093] According to the isotropic assumption in the mainstream region and the effect of turbulent kinetic energy gradient, creating a mainstream region Langevin equation, which is used to control the migration direction of particles in the mainstream region to ensure the continuity of turbulent pulsation;

[0094] Creating a numerical solution equation based on the Euler-Maruyama format, and the numerical solver is used to solve the boundary layer Langevin equation and the mainstream region Langevin equation;

[0095] Among them, the continuous random walk model includes the boundary layer Langevin equation, the mainstream region Langevin equation, and the numerical solution equation.

[0096] The Langevin equation is a differential equation that describes the random motion of particles in a fluid. By introducing the Langevin equation, the random pulsation behavior of aerosol particles in turbulence can be simulated.

[0097] Specifically, first, based on the turbulent anisotropy characteristics within the geometric boundary layer, the boundary layer Langevin equation is constructed. The boundary layer Langevin equation takes into account the turbulent characteristics in the near-wall region, such as the normal velocity fluctuation being significantly higher than that in the mainstream direction. By correcting the turbulent anisotropy characteristics, the boundary layer Langevin equation can more accurately simulate the deposition behavior of particles in the near-wall region. Then, based on the isotropic assumption and the turbulent kinetic energy gradient effect in the mainstream region, the mainstream region Langevin equation is constructed. The mainstream region Langevin equation is used to regulate the migration direction of particles in the mainstream region, ensuring the continuity of turbulent fluctuations and avoiding the simplified treatment of turbulent fluctuations in the traditional discrete random walk model. Finally, the Euler-Maruyama scheme (a method for numerically solving stochastic differential equations) is used to create a numerical solution equation for solving the boundary layer Langevin equation and the mainstream region Langevin equation. The Euler-Maruyama scheme can efficiently solve the Langevin equation, ensuring the stability and accuracy of numerical simulations.

[0098] In the continuous random walk model of this embodiment, the turbulent anisotropy characteristics in the near-wall region can be corrected by the boundary layer Langevin equation, avoiding the overestimation of particle deposition in the traditional discrete random walk model. The continuity of turbulent fluctuations is ensured by the mainstream region Langevin equation, more realistically reflecting the migration behavior of particles in the mainstream region. In addition, the numerical solution equation based on the Euler-Maruyama scheme provides an efficient and stable solution method. In summary, by constructing a continuous random walk model based on the Langevin equation and separately creating the boundary layer Langevin equation and the mainstream region Langevin equation, the accuracy of simulating the behavior of aerosol particles in turbulence is significantly improved.

[0099] Optionally, the above-mentioned boundary layer Langevin equation includes the following formulas (9) to (14):

[0100]

[0101] Among them, is the damping term, is the random term, is the slip correction term, τ i is the time scale, u i is the velocity component of the time-averaged velocity, σ i is the root mean square of velocity fluctuation, ξ i is a random number obeying the Gaussian normal distribution, i is the direction identifier, i = 1 is the mainstream direction, i = 2 is the circumferential direction, i = 3 is the normal direction, u1 is the velocity component in the mainstream direction, u2 is the velocity component in the circumferential direction, and u3 is the velocity component in the normal direction.

[0102] Stk is the Stokes number, defined as τ p / τ L, x2 is the spatial coordinate in the circumferential direction;

[0103]

[0104] Among them, σ1 is the root mean square of the velocity fluctuation in the mainstream direction, σ2 is the root mean square of the circumferential velocity fluctuation, σ3 is the root mean square of the normal velocity fluctuation, y + is the dimensionless wall distance, and u * is the wall friction velocity.

[0105] In this embodiment, formulas (9) to (14) significantly improve the accuracy of simulating the transport and deposition behavior of aerosol particles in wall-bounded turbulence through the modification and extension of the turbulent Langevin equation. First, a slip correction term is introduced in the boundary layer region (formulas 9 - 11), which micro-corrects the particle trajectory by combining the turbulent kinetic energy gradient and the Stokes number, avoiding the non-physical accumulation of small particles in the high turbulent kinetic energy region; secondly, the root mean square (RMS) fitting relationship of the pulsating velocity obtained by direct numerical simulation (DNS) (formulas 12 - 14) can reflect the anisotropic characteristics of the turbulence near the wall.

[0106] Optionally, to achieve accurate calculation of particle motion, coordinate transformation is also required. The transformation of the pulsating velocity from the local body-fitted coordinate system (BFCS) to the global coordinate system (CCS) (see formulas 15 - 17 below) provides a solid foundation for dynamic flow field calculation.

[0107] In an example, referring to Figure 10 the schematic diagram of the cross-sectional particle distribution sparsification shown, the boundary layer Langevin equations (9 - 11) add a slip correction term on the basis of the traditional Langevin equation. The slip correction term is related to the turbulent kinetic energy gradient, effectively driving the particles to migrate to the low turbulent kinetic energy region and eliminating the 21% overestimation error of the DRW model in the mainstream region; secondly, the σ1 - σ3 relationship formula based on DNS data fitting in formulas (12 - 14) reflects the asymmetry of the ratio of the pulsating velocity between the mainstream direction and the wall normal direction, with a 47% improvement in accuracy compared to the traditional isotropic scheme (formula (5) above); finally, through the transformation between the local body-fitted coordinate (BFCS) and the global coordinate (CCS) (formulas 15 - 17), the pulsating velocity components are dynamically adapted to the boundary layer curvature and slope direction, and the deviation rate between the simulated trajectory and the actual streamline is reduced from 14% to 3.8% (see Figure 3 particle tracking verification). Under the combined action, the overall model compresses the prediction error of the particle deposition rate from 38% of the traditional model to within the experimental error band of ±12%, fully ensuring the credibility of nuclear power safety analysis.

[0108] Optionally, the above-mentioned Langevin equation in the mainstream region includes the following formulas (18) to (20):

[0109]

[0110]

[0111] Among them, is the slip correction term, σ is the root mean square of velocity fluctuation. In the mainstream region, it is default that the root mean squares of velocity fluctuations in three directions are the same; ξ i is a random number obeying the Gaussian normal distribution, i is the direction identifier, i = 1 is the mainstream direction, i = 2 is the circumferential direction, i = 3 is the normal direction, x1 is the spatial coordinate in the mainstream direction, x2 is the spatial coordinate in the circumferential direction, and x3 is the spatial coordinate in the normal direction. is the gradient of turbulent kinetic energy, τ L is the Lagrangian time scale, u1 is the velocity component in the mainstream direction, u2 is the circumferential velocity component, u3 is the normal velocity component, τ L .

[0112] Formulas (18) to (20) solve the continuity problem in the transport of aerosol particles in the turbulent mainstream region outside the wall by constructing a normalized Langevin equation for the mainstream region.

[0113] Specifically, first, Formulas (18) to (20) introduce a slip correction term, enabling the particles to migrate towards the wall under the action of the gradient of turbulent kinetic energy. For typical pipe flows, since the turbulent kinetic energy in the center of the mainstream region is low and increases radially, this correction term effectively explains the physical mechanism of the gradual aggregation of particles towards the wall, thus overcoming the limitation of traditional isotropic turbulence models in overestimating the particle deposition rate. Second, Formulas (18) to (20) still retain the description of random motion (Gaussian distribution random numbers) and the damping effect of fluid-particle interaction, ensuring the physical continuity of particle motion.

[0114] Optionally, the numerical solution of the equation includes the following Formula (21):

[0115]

[0116] Among them, n is the current moment, and n + 1 is the next moment; A i represents the slip correction term; u i is the velocity component, τ L is the Lagrangian time scale.

[0117] Specifically, by introducing the Euler-Maruyama numerical format, the above formula (21) realizes the efficient collaborative calculation of the turbulent driving term and the random perturbation term in the aerosol particle motion model. This formula combines the damping term and the slip correction term in the traditional Langevin equation into the numerical solution framework, and maintains numerical stability through dynamic adaptation of the time step. Further, when generating such random numbers, first generate random numbers uniformly distributed between (0,1), and then use the Box-Müller method to generate random numbers obeying the Gaussian normal distribution to accurately simulate the random characteristics of turbulent pulsation. Compared with the piecewise discrete solution method, this unified format enables real-time coupled calculation of the boundary layer region (formulas 9-11) and the mainstream region (formulas 18-20), avoiding the particle trajectory deviation caused by the sudden change of the vortex core switching in the traditional DRW model.

[0118] In this embodiment, formula (21) adopts the Hermite-Euler unified numerical format, enabling seamless coupling of the boundary layer and the mainstream region equations, realizing synchronous solution of the boundary layer and the mainstream region equations, and ensuring continuous calculation of particle trajectories in the flow field mutation region.

[0119] Optionally, the above S105 may specifically include:

[0120] Set boundary conditions according to the basic parameters and the grid cell dataset. The boundary conditions are specific physical constraints imposed on the boundaries of the computational domain in numerical simulations;

[0121] Set the solution parameters of the solver. The solution parameters are key variables used to control discretization methods, iterative strategies, and convergence criteria in numerical simulations;

[0122] Based on the basic parameters, boundary conditions, and solution parameters, perform iterative solutions for the wall function, turbulence model, discrete phase model, and continuous random walk model to obtain converged velocity field, turbulent kinetic energy field, and aerosol particle trajectory distribution data.

[0123] The specific steps are as follows:

[0124] First, set boundary conditions. According to the basic parameters (such as fluid density, viscosity, temperature, etc.) and the grid cell dataset, impose specific physical constraints on the boundaries of the computational domain, such as inlet velocity, outlet pressure, wall no-slip condition, etc. The boundary conditions are the basis of numerical simulations, ensuring that the simulation results conform to the constraints of the actual physical scenario.

[0125] Second, set solution parameters. Define the key variables in numerical simulations, including discretization methods (such as the finite volume method), iterative strategies (such as the SIMPLE algorithm), and convergence criteria (such as residual thresholds). The solution parameters control the accuracy and efficiency of numerical simulations, ensuring that the simulation results can converge to reasonable solutions.

[0126] Thirdly, iterative solution. Based on the basic parameters, boundary conditions, and solution parameters, iterative solutions are performed for the wall function, turbulence model, discrete phase model, and continuous random walk model. The specific steps include: initializing the velocity field, pressure field, and turbulent kinetic energy field; updating the flow field according to the wall function and turbulence model; using the discrete phase model to track the motion trajectories of aerosol particles; simulating the random pulsation behavior of particles based on the continuous random walk model; checking for convergence, and if not converged, continue the iteration until the convergence criterion is met.

[0127] Through iterative solution, gradually approach the true flow state of the fluid and the motion behavior of aerosol particles, and finally obtain the converged velocity field, turbulent kinetic energy field, and aerosol particle trajectory distribution data.

[0128] To facilitate the understanding of the aerosol particle behavior analysis method provided in this embodiment, the following provides the practical application description of the above method. For specific reference, see the following examples:

[0129] The following steps are implemented based on the computational fluid dynamics software Fluent, specifically as follows:

[0130] Step 1: Geometrically model the computational domain involved in the transport and deposition of aerosol particles in the wall-bounded turbulent airflow to be studied (equivalent to constructing the geometric model of the computational domain of the fluid to be analyzed), and perform mesh division on the geometry. When performing mesh division, it is recommended to encrypt the boundary layer to ensure that the dimensionless height of the first layer of mesh is about 1 (equivalent to meshing the geometric model to obtain the mesh element dataset of the geometric model).

[0131] Step 2: Import the generated mesh file into the Fluent software

[0132] Step 3: Set the fluid medium for the computational domain in the Fluent software (equivalent to obtaining the basic parameters of the fluid to be analyzed);

[0133] Step 4: Select an isotropic two-equation turbulence model (such as the k-ε model or k-ω model) and an enhanced wall function in the software, and select whether to activate the energy equation according to the simulation needs (equivalent to obtaining the wall function and turbulence model);

[0134] Step 5: Set the boundary conditions of the computational domain in Fluent. These boundary conditions can be inlet, outlet, wall, or symmetry plane, etc. (equivalent to setting the boundary conditions according to the basic parameters and mesh element dataset);

[0135] Step 6: Set the solution parameters in the Fluent software, including the solution method, relaxation factor, residual, etc. (equivalent to setting the solution parameters of the solver);

[0136] Step 7: Initialize and start the calculation to finally obtain a converged and stable flow field;

[0137] Step 8: Activate the discrete phase model in the Fluent software, and set information such as particulate properties and injection methods. When necessary, the position, particle size, initial velocity, initial temperature, and flow rate information of silver particles at the inlet can be specified through an external file (equivalent to obtaining the discrete phase model);

[0138] Step 9: Compile the continuous random walk model UDF (User Defined Functions) using the macros provided by the Fluent software, and compile the UDF in the Fluent software and specify the positions where these UDFs act. The custom macros involved in the UDF are DEFINE_ON_DEMAND(), DEFINE_DPM_TIMESTEP(), DEFINE_DPM_INJECTION_INIT(), and DEFINE_DPM_DRAG(). Among them, DEFINE_ON_DEMAND() is used to initialize the user-defined storage space to store information such as the dimensionless height of all near-wall cells, wall stress, and vectors of cells; DEFINE_DPM_TIMESTEP() is used to adjust the time step of particle transport calculations; DEFINE_DPM_INJECTION_INIT() is used to specify that the initial velocity of the particle is the same as the fluid velocity at the position where the particle is located; DEFINE_DPM_DRAG() is used to calculate the drag coefficient between the particle and the fluid, and this macro is the core part of the UDF (equivalent to creating the continuous random walk model).

[0139] The theoretical basis for compiling the continuous random walk model UDF is the Langevin equation, as shown below:

[0140]

[0141] In the formula: τ i is the time scale; dξ i is a series of independent random numbers with a mean of 0 and a variance of dt. Usually, it is assumed that ξ i obeys a Gaussian distribution.

[0142] The above equation is widely used to simulate isotropic turbulence. Under such flow conditions, the RMS of fluid pulsation and the Lagrangian time scale are independent of position. However, in wall-bounded flows, the turbulence is highly anisotropic within the boundary layer, so the above equation needs to be modified. The following equations are established for the boundary layer region and the main flow region respectively:

[0143] Normalized Langevin equation for the boundary layer region (y + less than 100):

[0144]

[0145] In the formula, is the damping term, which will reduce the particle velocity to zero because the particles lose energy over time; is the random term, which is used to describe the random collisions between particles and gas molecules at a constant temperature; the third term on the right side of the equation is the slip correction term. The introduction of this term can effectively avoid the unrealistic accumulation of small particles in regions with high turbulent kinetic energy. For fully developed wall-bounded turbulence, particle deposition is mainly due to the wall-normal turbulence. Therefore, the slip correction term in the mainstream direction (Equation 9) can be ignored; Stk is the Stokes number of the particle, defined as τ p / τ L .

[0146] The pulsating RMS in the above system of equations can be calculated using the following fitting relationships obtained based on direct numerical simulation, which reflect the characteristics of turbulence anisotropy in the boundary layer.

[0147]

[0148] When calculating the boundary layer velocity pulsating RMS using the above fitting relationships and then obtaining the pulsating velocities in each direction according to Equation (3), it should be noted that the pulsating velocities obtained at this time are in the local body-fitted coordinate system (Body Fitted CoordinateSystem, BFCS), as Figure 2 shown. Therefore, the pulsating velocities in the local body-fitted coordinate system need to be transformed back to the global coordinate system (Computational Coordinate System, CCS). The transformation method is as follows:

[0149] 1) Establish the local body-fitted coordinate system where is the unit vector in the mainstream direction, is the unit vector in the wall-normal direction, is the unit vector in the lateral direction.

[0150]

[0151] In the formula: is the fluid velocity vector at the current position of the particle in the global coordinate system.

[0152] 2) Obtain the pulsating velocities U1', U2' and U3' in the three directions in the local body-fitted coordinate system according to Equation (4) and the fitting relationships of the velocity pulsating RMS;

[0155] 3) Convert the pulsating velocity in the local body-fitted coordinates to the pulsating velocity in the global coordinate system according to the following formula.

[0156]

[0157] In the formula: are the unit vectors in the three directions of the global coordinate system respectively.

[0158] For the mainstream region (y + greater than 100), the normalized Langevin equation is as follows:

[0159]

[0160] Similarly, the third term on the right side of the equation is the slip correction term, which is related to the turbulent kinetic energy gradient and can be obtained from the flow field calculation results. For the flow in a pipe, the turbulent kinetic energy in the central region is the smallest and increases gradually in the radial direction. Therefore, the particles in the mainstream region will move towards the wall under the action of the slip correction term.

[0161] Use the Euler-Maruyama format to uniformly solve the normalized Langevin equations in the boundary layer region and the mainstream region. The numerical format is as follows:

[0162]

[0163] In the formula: n represents the current time; n + 1 represents the next time; A i represents the slip correction term.

[0164] The above equation involves random numbers that follow a Gaussian distribution. When generating such random numbers, first generate random numbers uniformly distributed between (0, 1), and then use the Box-Müller method to generate random numbers that follow a Gaussian normal distribution.

[0165] Step 10: Set the boundary conditions for the discrete phase calculation, and set the possible deposition surface of the aerosol particles as the "trap" boundary condition. This boundary condition means that when the distance between the particle and the wall is less than the particle radius, the particle will adhere to the wall and will no longer bounce;

[0166] Step 11: Perform the discrete phase calculation and post-process the calculation results using CFD-Post. The calculation results can obtain information such as particle motion trajectories and particle distributions, and then evaluate the particle deposition situation.

[0167] In some embodiments, the calculation process of the continuous random walk model developed using UDF is as shown in Figure 4 as follows. The specific steps are as follows:

[0168] 1. Pretreatment stage: Calculate the distance between the centers of all fluid cells and the nearest wall cell and store it in the user-defined storage space (UDM).

[0169] 2. Calculate the distance between the current particle and the nearest wall cell, and obtain the dimensionless distance based on the wall friction velocity.

[0170] 3. Calculate the Eulerian statistical quantities σ1, σ2, σ3, and τ L , and generate Gaussian distributed random numbers.

[0171] 4. Position determination: Calculate the dimensionless distance (y + ) between the particle and the nearest wall in real time. Based on the y + value, determine whether the particle is in the boundary layer (y + < 100) or the mainstream region. If not, execute step 5; if so, execute step 6.

[0172] 5. Fluctuating velocity calculation:

[0173] If in the boundary layer, calculate the fluctuating velocity in the local body-fitted coordinate system (BFCS) through the DNS fitting formula (12 - 14), and convert the velocity to the global coordinate system (CCS) through the coordinate transformation matrix (formulas 15 - 17).

[0174] 6. If in the mainstream region, directly calculate the fluctuating velocity based on the isotropic model.

[0175] 7. Force update: Superimpose the fluctuating velocity on the time-averaged fluid velocity, and calculate the drag force in combination with the DEFINE_DPM_DRAG macro.

[0176] 8. Dynamic tracking: Update the particle velocity and position through the Euler - Maruyama format (formula 21), and detect in real time whether the particle touches the wall (triggering the "trap" boundary condition) or leaves the computational domain.

[0177] Through local coordinate transformation and dynamic boundary layer determination, this process achieves accurate capture of the anisotropic characteristics of turbulence while ensuring computational continuity and numerical stability.

[0178] In a specific example, the above method may include the following steps:

[0179] Step 1: Establish a geometric model for the test bench shown in Figure 5 , generate a hexahedral mesh using the o-grid technique in ICEM CFD. When generating the mesh, the requirement that the first layer of cells near the wall is close to 1 is considered, and finally the mesh shown in Figure 6 is generated. Figure 7 is generated.

[0180] This experiment is about the transport and deposition of aerosols in turbulent air flow inside a circular tube. A vibrating and orienting monodisperse aerosol generator is used to generate uniform and spherical olive oil droplets (with a density of 920 kg / m -3 m 3 ) at a rated output of 1.5×10 3 / s. Then the oil droplets are vertically upward passed through a copper tube with a nominal diameter of 0.032 m to reach the top chamber. Subsequently, the particles are transported into a copper tube with a nominal diameter of 0.032 m below the chamber, and then enter a glass deposition tube with a length of 1.02 m and an inner diameter of 0.0127 m. A filter containing a glass filter paper with a diameter of 90 mm is placed downstream of the deposition tube, and the filter is connected to a volume flow sensor, a regulating valve and a pump. The olive oil particles used in the experiment contain less than 10% of urethane, and urethane is used as a fluorescent tracer for quantitative analysis of deposited particles. Once the particles reach the inner surface of the glass tube, they will be captured.

[0181] Step 2: Import the generated mesh file into Fluent software;

[0182] Step 3: Specify the fluid medium as incompressible air in Fluent software;

[0183] Step 4: Select the two-equation RNG k-ε model and enhanced wall function in Fluent software. Considering that the temperature remains constant in the experiment, the energy equation is not activated;

[0184] Step 5: Set the inlet as the velocity inlet boundary condition, the outlet as the free outflow boundary condition, and the wall as the no-slip boundary condition in Fluent software;

[0185] Step 6: Set the solution parameters. The pressure-velocity coupling adopts the SIMPLE format, and the spatial discretization of pressure, turbulent kinetic energy, etc. all adopts the second-order upwind format, and the convergence residual is set to 10 -8 ;

[0186] Step 7: Initialize and calculate to obtain a converged flow field, as Figure 8 shown.

[0187] Step 8: Activate the discrete phase model in Fluent software and set the particle density to 920 kg / m 3 , and inject aerosol particles in a uniform manner at the inlet. The specific implementation method is as follows:

[0188] An inscribed square is constructed for the circular inlet of the deposition tube. The inscribed square is divided into uniform small squares, and then the centers of all small squares are projected onto the inlet circle. A particle is set at the center of each small square whose projection lies in the inner region of the circle, so that a uniform distribution of inlet particles can be approximately achieved. According to this calculation method, a preprocessing tool is compiled using the Python language to generate the.inj file for particle input, and this file can be read in the Fluent particle injection dialog during the calculation.

[0189] Figure 9 An example of the uniform distribution obtained based on this method is given, and the number of aerosol particles in this example is 49080.

[0190] Step 9: Compile the continuous random walk model UDF (User Defined Functions, UDF) using the macros provided by the Fluent software, and compile the UDF in the Fluent software and specify the action positions for these UDFs. The custom macros involved in the UDF are DEFINE_ON_DEMAND(), DEFINE_DPM_TIMESTEP(), DEFINE_DPM_INJECTION_INIT(), and DEFINE_DPM_DRAG();

[0191] Step 10: Set the boundary conditions for the discrete phase calculation, and set the possible deposition surface of the aerosol particles as the "trap" boundary condition;

[0192] Step 11: Perform the discrete phase calculation, and CFD-POST can be used for post-processing. For example, Figure 10 The proportional distribution of the number of particles on different cross-sections is given,

[0193] In this embodiment, by introducing the continuous random walk model (CRW), while retaining the engineering applicability of the isotropic two-equation turbulence model, a slip correction term and a dynamic fitting formula for the boundary layer pulsating velocity are introduced for the anisotropic characteristics in wall-bounded turbulence, significantly improving the physical fidelity and numerical stability of the turbulent swimming simulation. The adopted Euler-Maruyama format realizes the seamless coupled calculation of the normalized Langevin equation in the boundary layer region and the mainstream region. By precisely setting the grid y+≈1 and customizing UDF macro functions (such as DEFINE_DPM_DRAG and DEFINE_DPM_TIMESTEP), the accuracy of the particle trajectory simulation near the wall is ensured. In addition, the CRW model more realistically reflects the continuity of turbulent pulsation, overcomes the defect of overestimation in the traditional discrete random walk (DRW) model when predicting the particle deposition rate, and significantly reduces the deposition rate prediction error from 38% to within the experimental tolerance range (±12%). This technical solution can accurately predict the migration and deposition behavior of aerosols in the pipe, providing an efficient and reliable simulation tool for the risk assessment of radioactive aerosol transport in nuclear power accidents.

[0194] Embodiment 2:

[0195] This embodiment provides a safety assessment method for a nuclear power plant, which is applied to a containment vessel and specifically includes the following steps:

[0196] According to the aerosol particle behavior analysis method provided in any of the above embodiments, analyze the aerosol particle behavior in the containment vessel to obtain an analysis result;

[0197] Based on the analysis result, conduct a safety assessment of the containment vessel.

[0198] Specifically, by simulating the behavior of aerosol particles in the containment vessel, the concentration, distribution, and migration path of aerosols at different positions can be accurately predicted, the release rate and amount of radioactive substances in the containment vessel can be precisely calculated, and the radioactive risk of the containment vessel can be quantitatively evaluated.

[0199] The safety assessment method of the nuclear power plant in this embodiment can accurately assess the diffusion and deposition risks of radioactive substances in the containment vessel, helping to quickly and accurately conduct risk assessment during an accident, and thus making faster and more effective emergency response decisions, such as evacuation plans and radiation protection measures. Through refined aerosol particle behavior analysis, the risk assessment and accident response capabilities of the nuclear power plant containment vessel can be significantly improved.

[0200] Furthermore, the above analysis results can also be used to optimize the design and operating parameters of the containment vessel, such as the position and specifications of the ventilation system, spray system, and filtration device, thereby improving the accident mitigation ability. By simulating different accident scenarios, reliable basis is provided to improve the existing operation strategies and emergency plans.

[0201] Example 3:

[0202] As Figure 11 shown, this embodiment provides an aerosol particle behavior analysis device 1100, including:

[0203] A first acquisition module 1101, configured to acquire basic parameters of the fluid to be analyzed, where the fluid to be analyzed includes aerosol particles;

[0204] A model construction module 1102, configured to construct a geometric model of the computational domain of the fluid to be analyzed;

[0205] A mesh generation module 1103, connected to the model construction module, configured to mesh the geometric model to obtain a mesh cell data set of the geometric model;

[0206] A second acquisition module 1104, which acquires a wall function, a turbulence model, a discrete phase model, and a continuous random walk model. The wall function is used to describe the flow characteristics in the turbulent near-wall region, and the continuous random walk model is used to simulate the continuity of the pulsating velocity of aerosol particles in the turbulent airflow restricted by the wall;

[0207] A solution module 1105, connected to the second acquisition module 1104, configured to solve the wall function, the turbulence model, the discrete phase model, and the continuous random walk model through a solver based on the basic parameters and the mesh cell data set, and obtain the condition distribution data of the fluid to be analyzed in the computational domain;

[0208] An analysis module 1106, connected to the solution module 1105, configured to analyze the behavior of aerosol particles based on the condition distribution data.

[0209] Optionally, the above device 1100 further includes:

[0210] A creation module, configured to construct the continuous random walk model based on the Langevin equation;

[0211] The creation module specifically includes:

[0212] A first creation unit, configured to create a boundary layer Langevin equation according to the turbulent anisotropy characteristics in the geometric boundary layer, and the boundary layer Langevin equation is used to correct the turbulent anisotropy characteristics near the wall;

[0213] A second creation unit, configured to create a mainstream region Langevin equation according to the isotropic assumption in the mainstream region and the turbulent kinetic energy gradient effect, and the mainstream region Langevin equation is used to control the migration direction of particles in the mainstream region to ensure the continuity of turbulent pulsation;

[0214] A third creation unit, configured to create a numerical solution equation based on the Euler-Maruyama format, where the numerical solver is used to solve the boundary layer Langevin equation and the mainstream region Langevin equation;

[0215] Wherein, the continuous random walk model includes the boundary layer Langevin equation, the mainstream region Langevin equation, and the numerical solution equation.

[0216] The above-mentioned solution module includes:

[0217] A first setting sub-module, configured to set boundary conditions, where the boundary conditions are specific physical constraints imposed on the boundaries of the computational domain in numerical simulations;

[0218] A second setting sub-module, configured to set the solution parameters of the solver, where the solution parameters are key variables used to control discretization methods, iterative strategies, and convergence criteria in numerical simulations;

[0219] An iterative calculation module, configured to iteratively solve the wall function, turbulence model, discrete phase model, and continuous random walk model based on the basic parameters, the boundary conditions, and the solution parameters, to obtain converged velocity field, turbulent kinetic energy field, and aerosol particle trajectory distribution data.

[0220] Optionally, the boundary conditions include an inlet boundary and an outlet boundary, and the solution parameters include a discretization method, a relaxation factor, a coupling method, a time step, and a residual convergence criterion.

[0221] Wherein, the time step is the length of a time step.

[0222] In this embodiment, by obtaining fluid basic parameters, constructing a geometric model, meshing, and obtaining and solving the wall function, turbulence model, discrete phase model, and continuous random walk model, the behavior of aerosol particles is accurately analyzed. By introducing the continuous random walk model, the continuity of the turbulent pulsation of aerosol particles is ensured, and the pulsation velocity of aerosol particles in wall-bounded turbulence is more accurately simulated, thereby improving the accuracy of the simulation results.

[0223] It should be clear that the present application is not limited to the specific configurations and processes described above and shown in the figures. For the sake of brevity, detailed descriptions of known methods are omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of the present application is not limited to the specific steps described and shown, and those skilled in the art can make various changes, modifications, and additions, or change the order between steps after understanding the spirit of the present application.

[0224] The functional blocks shown in the above structural block diagram can be implemented as hardware, software, firmware, or a combination thereof. When implemented in hardware, it can be, for example, an electronic circuit, an application specific integrated circuit (ASIC), appropriate firmware, a plug-in, a functional card, and so on. When implemented in software, the elements of the present application are programs or code segments used to perform the required tasks. The program or code segment can be stored in a machine-readable medium or transmitted via a data signal carried in a carrier wave on a transmission medium or a communication link. "Machine-readable medium" can include any medium capable of storing or transmitting information. Examples of machine-readable media include electronic circuits, semiconductor memory devices, ROM, flash memory, erasable ROM (EROM), floppy disks, CD-ROMs, optical discs, hard disks, fiber optic media, radio frequency (RF) links, and so on. The code segment can be downloaded via a computer network such as the Internet, an intranet, and so on.

[0225] It should also be noted that in the exemplary embodiments mentioned in the present application, some methods or systems are described based on a series of steps or devices. However, the present application is not limited to the order of the above steps, that is, the steps can be executed in the order mentioned in the embodiments, can be different from the order in the embodiments, or several steps can be executed simultaneously.

[0226] As described above with reference to the flowcharts and / or block diagrams of methods, apparatuses (systems), and computer program products according to embodiments of the present disclosure. It should be understood that each block in the flowchart and / or block diagram, and the combination of blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable scheduling devices to generate a machine such that these instructions executed by the processor of the computer or other programmable scheduling devices enable the implementation of the functions / actions specified in one or more blocks of the flowchart and / or block diagram. Such a processor can be, but is not limited to, a general-purpose processor, a special-purpose processor, a special application processor, or a field programmable logic circuit. It can also be understood that each block in the block diagram and / or flowchart, and the combination of blocks in the block diagram and / or flowchart, can also be implemented by dedicated hardware that performs the specified functions or actions, or can be implemented by a combination of dedicated hardware and computer instructions.

[0227] The above is only the specific implementation manner of the present application. Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the above-described systems, modules, and units can refer to the corresponding processes in the foregoing method embodiments and will not be elaborated herein. It should be understood that the protection scope of the present application is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present application can easily think of various equivalent modifications or substitutions, and these modifications or substitutions should all be covered within the protection scope of the present application.

Claims

1. A method for analyzing aerosol particle behavior, characterized in that: include: Acquiring basic parameters of a fluid to be analyzed, wherein the fluid to be analyzed includes aerosol particles; Construct a geometric model of the computational domain of the fluid to be analyzed; Meshing the geometric model to obtain a grid unit data set of the geometric model; Obtaining a wall function, a turbulence model, a discrete phase model and a continuous random walk model, wherein the wall function is used to describe the flow characteristics of the turbulent near-wall region, and the continuous random walk model is used to simulate the continuity of the pulsating velocity of aerosol particles in the turbulent airflow limited by the wall; Based on the basic parameters and the grid unit data set, the wall function, turbulence model, discrete phase model and continuous random walk model are solved by a solver to obtain the operating condition distribution data of the fluid to be analyzed in the calculation domain; The behavior of the aerosol particles is analyzed based on the operating condition distribution data.

2. The method according to claim 1, characterized in that: The method further comprises: Based on the Langevin equation, the continuous random walk model is constructed; The continuous random walk model is constructed based on the Langevin equation, specifically including: According to the turbulence anisotropy characteristics in the geometric boundary layer, a boundary layer Langevin equation is created, wherein the boundary layer Langevin equation is used to correct the turbulence anisotropy characteristics near the wall; Based on the isotropy assumption of the mainstream region and the turbulent kinetic energy gradient effect, a Langevin equation in the mainstream region is created, wherein the Langevin equation in the mainstream region is used to control the migration direction of particles in the mainstream region and ensure the continuity of turbulent pulsation; Creating a numerical solution equation based on the Euler-Maruyama format, wherein the numerical solution equation is used to solve the boundary layer Langevin equation and the mainstream region Langevin equation; Wherein, the continuous random walk model includes the boundary layer Langevin equation, the mainstream region Langevin equation and the numerical solution equation.

3. The method according to claim 2, characterized in that: The boundary layer Langevin equation includes the following formulas (9) to (14): in, is the damping term, is a random item, and is the slip correction term, τ i is the time scale, u i is the velocity component, σ i is the RMS velocity fluctuation, ξ i is a random number that obeys a Gaussian normal distribution, i is the direction identifier, i=1 is the mainstream direction, i=2 is the circumferential direction, i=3 is the normal direction, Stk is the Stokes number, defined as τ p / τ L , x2 is the spatial coordinate in the circumferential direction; Among them, σ1 is the RMS velocity fluctuation in the mainstream direction, σ2 is the RMS velocity fluctuation in the circumferential direction, σ3 is the RMS velocity fluctuation in the normal direction, and y + is the dimensionless wall distance, u * is the wall friction velocity.

4. The method according to claim 3, characterized in that The mainstream region Langevin equation includes the following formulas (18) to (20): in, and is the slip correction term, σ is the RMS velocity fluctuation, ξ i is a random number that follows a Gaussian normal distribution, u i is the velocity component, i is the direction identifier, i=1 is the mainstream direction, i=2 is the circumferential direction, i=3 is the normal direction, x1 is the spatial coordinate of the mainstream direction, x2 is the spatial coordinate of the circumferential direction, and x3 is the spatial coordinate of the normal direction. is the turbulent kinetic energy gradient, τ L is the Lagrangian time scale.

5. The method according to claim 4, characterized in that The numerical solution equation includes the following formula (21): Among them, n is the current moment, n+1 is the next moment; A i represents the slip correction term.

6. The method according to claim 1, characterized in that The method of solving the wall function, turbulence model, discrete phase model and continuous random walk model by a solver based on the basic parameters and the grid unit data set to obtain the operating condition distribution data of the fluid to be analyzed in the calculation domain specifically includes: Setting boundary conditions according to the basic parameters and the grid unit data set, wherein the boundary conditions are specific physical constraints imposed on the boundary of the computational domain in the numerical simulation; Setting the solution parameters of the solver, wherein the solution parameters are key variables used to control the discretization method, iteration strategy and convergence criterion in numerical simulation; Based on iteratively solving the wall function, turbulence model, discrete phase model and continuous random walk model according to the basic parameters, the boundary conditions and the solution parameters, converged velocity field, turbulent kinetic energy field and aerosol particle trajectory distribution data are obtained.

7. A safety assessment method for a nuclear power plant, characterized in that: Applied to a containment vessel, the method comprises: According to the aerosol particle behavior analysis method according to any one of claims 1 to 6, the aerosol particle behavior in the containment is analyzed to obtain an analysis result; Based on the analysis results, a safety assessment is performed on the containment vessel.

8. An aerosol particle behavior analysis device, characterized in that: The device comprises: A first acquisition module, used to acquire basic parameters of a fluid to be analyzed, wherein the fluid to be analyzed includes aerosol particles; A model building module, used to build a geometric model of the computational domain of the fluid to be analyzed; A meshing module, connected to the model building module, for meshing the geometric model to obtain a mesh unit data set of the geometric model; The second acquisition module acquires a wall function, a turbulence model, a discrete phase model, and a continuous random walk model, wherein the wall function is used to describe the flow characteristics of the turbulent near-wall region, and the continuous random walk model is used to simulate the continuity of the pulsating velocity of aerosol particles in the turbulent airflow limited by the wall; A solution module, connected to the second acquisition module, for solving the wall function, turbulence model, discrete phase model and continuous random walk model through a solver based on the basic parameters and the grid unit data set to obtain the operating condition distribution data of the fluid to be analyzed in the calculation domain; An analysis module, connected to the solution module, is used to analyze the behavior of the aerosol particles based on the operating condition distribution data.

9. The device according to claim 8, characterized in that The device also includes: Creating a module for constructing the continuous random walk model based on the Langevin equation; The creation module specifically includes: A first creation unit is used to create a boundary layer Langevin equation according to the turbulence anisotropy characteristics in the geometric boundary layer, wherein the boundary layer Langevin equation is used to correct the turbulence anisotropy characteristics near the wall; The second creation unit is used to create a Langevin equation in the mainstream region according to the isotropy assumption of the mainstream region and the turbulent kinetic energy gradient effect, wherein the Langevin equation in the mainstream region is used to control the migration direction of particles in the mainstream region to ensure the continuity of turbulent pulsation; A third creation unit is used to create a numerical solution equation based on the Euler-Maruyama format, and the numerical solver is used to solve the boundary layer Langevin equation and the mainstream region Langevin equation; Wherein, the continuous random walk model includes the boundary layer Langevin equation, the mainstream region Langevin equation and the numerical solution equation.

10. A safety assessment system for a nuclear power plant, characterized in that: Applied to a containment vessel, the system comprises: The aerosol particle behavior analysis device according to any one of claims 8 to 9 is used to analyze the aerosol particle behavior in the containment to obtain an analysis result; An evaluation device is used to perform a safety evaluation on the containment shell based on the analysis result.

Images