A numerical simulation system and method for detecting silo porosity

Abstract

The present application relates to a kind of granary pore detection numerical simulation system and method belong to the technical field of grain pile pore detection numerical simulation, the system of the present application is by physical model establishment and grid division module, particle accumulation simulation module, DEM-CFD coupling calculation module, data processing module is composed of, can accurately simulate the pore distribution law of arbitrary area or grain layer of grain pile, generates pore distribution nephogram, realizes the rapid, high-precision detection of grain pile porosity, provides key porosity value for the mechanical ventilation of stored grain pile, provides accurate, intuitive porosity information for stored grain practice, it is helpful for relevant worker to find and handle abnormal situation in grain pile in time, take appropriate management measures, to forecast and analyze the change rule of grain state, realize safe storage of grain has important significance, provides a new technical means for the complex stored grain system research of the mutual coupling of multiple factors.

Description

Technical Field

[0001] This invention belongs to the field of numerical simulation technology for grain pile porosity detection, specifically relating to a numerical simulation system and method for grain warehouse porosity detection. Background Technology

[0002] Grain storage in a granary is a complex, multi-factor coupled system. This system includes abiotic factors such as temperature, relative humidity, moisture, pressure, and porosity, as well as biotic factors such as microorganisms and pests. There are complex couplings and relationships between biotic and abiotic factors in the grain storage system. For example, the porosity distribution of the grain pile affects the heat and moisture migration of the grain, and its impact on mechanical ventilation and grain cooling efficiency, all of which collectively determine the stability of the grain storage system.

[0003] Among the many factors influencing grain storage systems, the porosity of the grain pile is a key parameter affecting mechanical ventilation and grain cooling. It determines the airflow channels within the grain pile and directly impacts the distribution of ventilated air and the effectiveness of grain ventilation. During ventilation, low-temperature air tends to pass through areas with higher porosity, facilitating sufficient heat and moisture exchange with the grain in those areas. In areas with lower porosity, however, low-temperature air struggles to penetrate effectively, resulting in poor and uneven ventilation. This leads to airflow driven by temperature differences within the grain pile. While heat transfer occurs, moisture in the pores of the grain pile concentrates from higher to lower temperatures with the airflow, causing localized moisture accumulation and creating low-temperature, high-humidity areas. Once the grain reaches its dew point temperature, condensation occurs, leading to mold growth and heat generation, jeopardizing grain storage safety. Therefore, understanding and mastering the porosity distribution within the grain pile allows for the rational selection of ventilation duct types and the location, number, and size of vents, ensuring that low-temperature air fully covers the entire grain pile, achieving uniform ventilation and cooling. Secondly, understanding the porosity distribution helps to promptly detect and address abnormalities in grain piles, guiding relevant workers to take appropriate management measures during grain storage. For example, based on the porosity differences in different areas, targeted turning can be carried out. This is of great significance for predicting and analyzing the patterns of grain state changes and achieving safe grain storage.

[0004] Currently, grain pile porosity is mainly detected through experiments using self-made test benches and digital imaging technology. Real-world testing establishes small test benches and indirectly calculates the average porosity of each grain layer by acquiring strain distribution values. However, this method cannot directly observe and measure the porosity distribution inside the grain pile, requiring inference, which leads to uncertainty in the measurement results. Furthermore, the experiments are time-consuming and cannot achieve real-time, accurate detection, thus limiting its practical application. Digital imaging technology acquires surface images of grain piles under different pressures. These images contain information about grain particles and pore structures. Image segmentation algorithms are used to separate grain particles and pore structures from the images. By calculating the number of pixels or area of ​​grain particles and pore structures, the surface porosity of the grain pile is further calculated. Digital image processing and analysis involve multiple algorithms and complex calculations, and image quality and resolution are subject to various limitations, such as lighting conditions and shooting angles, thus affecting image quality and usability and making it difficult to guarantee the accuracy of the detection results. Therefore, there is an urgent need to invent a numerical simulation system and method for accurately detecting grain pile porosity, providing a new technical means for the study of complex grain storage systems. Summary of the Invention

[0005] To address the shortcomings of existing technologies and methods in the field of grain storage porosity detection, this invention aims to provide a numerical simulation system and method for grain storage porosity detection. This system can numerically simulate the grain grain entry process under different storage conditions, particle shapes, sizes, and physical properties, forming a grain pile. Through CFD-DEM coupled calculations, it accurately simulates the porosity distribution patterns in any region or layer of the grain pile. Furthermore, a data processing system generates a porosity distribution cloud map, providing accurate and intuitive porosity information for grain storage practices. This numerical simulation system and method can conduct in-depth research on the porosity distribution patterns within grain piles under various storage scenarios, providing a new technical means for studying complex grain storage systems with multiple coupled factors.

[0006] The present invention discloses a numerical simulation system for grain warehouse porosity detection based on CFD-DEM coupled calculation, comprising module 1, module 2, module 3, and module 4, wherein...

[0007] Module 1 is the physical model creation and mesh generation module, used to create a physical model of the grain silo, set the fluid region and boundary names in the computational domain, generate the mesh and output a mesh format file for use by subsequent modules;

[0008] Module 2 is a particle accumulation simulation module, used to simulate the grain entering the warehouse and forming a grain pile. Based on the geometric parameters of real grain particles, a particle geometric model is established, and a mesh format file is imported as a silo geometric model. Particle size distribution and various contact parameters between particles and silos are set, including: particle and silo density, Poisson's ratio, shear modulus, collision recovery coefficient, static friction coefficient, and rolling friction coefficient between particles and silos, and collision recovery coefficient, static friction coefficient, and rolling friction coefficient between particles. A virtual pellet factory is set up to simulate the grain entering the warehouse and form particle accumulation to obtain the grain pile.

[0009] Module 3 is the DEM-CFD coupled calculation module, used to simulate the flow of fluid-air between pores in a grain pile and obtain pore information; import the mesh format file into Fluent, set various parameters, import the coupling interface udf, connect it with EDEM, initialize it, set the time step to 100 times the EDEM time step, and perform DEM-CFD coupled calculation.

[0010] Module 4 is the data processing module. Through the data processing system, it processes the pore distribution data in the form of volume fraction to generate an intuitive pore distribution cloud map.

[0011] The present invention provides a numerical simulation method for grain silo porosity detection based on CFD-DEM coupled calculation, taking a cylindrical silo as an example, which includes the following steps:

[0012] 1) Use ANSYS DM to create a physical model of the silo and mesh it. Set the names of the fluid regions and boundaries in the computational domain.

[0013] 2) Output mesh format file;

[0014] 3) Using EDEM software, establish a particle geometric model based on the geometric parameters of real grain particles, and import the mesh format file as the geometric model of the silo.

[0015] 4) Through experiments, experience, and theoretical formulas, determine various physical property parameters of grain to ensure that all parameters involved in the CFD-DEM coupled numerical simulation are based on evidence, thus ensuring the reliability and accuracy of the numerical simulation results.

[0016] 5) Set the particle size distribution and various contact parameters between particles and silos in the EDEM, including: particle and silo density, Poisson's ratio, shear modulus, collision recovery coefficient, static friction coefficient, rolling friction coefficient between particles and silos, and collision recovery coefficient, static friction coefficient, and rolling friction coefficient between particles.

[0017] 6) Set up a virtual pellet mill to simulate the grain entering the warehouse, forming a pellet pile; the DEM (Discrete Element Contact Model) of the pellets during the piling process is as follows:

[0018] The vibrational motion during particle contact is decomposed into tangential and normal directions. The normal vibrational motion equation F n for:

[0019]

[0020] During particle contact, the tangential vibration motion manifests as tangential sliding and particle rolling, with its tangential component F s The external torque M acting on the particle is expressed as follows:

[0021]

[0022]

[0023] Where: m 1,2 I represents the equivalent mass of particles i and j, in kg. 1,2 Let i be the equivalent rotational inertia of particles i and j, in kg·m. 2 ; s is the radius of rotation, m; μ n μ s θ represents the normal and tangential relative displacements of the particle, respectively, in meters (m); θ is the particle's rotation angle, in rad; K s K n These are the tangential and normal elastic coefficients in the contact model, respectively; c n c s These are the normal and tangential damping coefficients in the contact model, respectively.

[0024] 7) Use the CFD software ANSYS Fluent to simulate the flow of fluid-air in the pores of the grain pile.

[0025] 8) Import the mesh file into Fluent, set the parameters, set the Viscous model to SST k-epsilon, and set the material to grain particles consistent with the EDEM end.

[0026] 9) Open EDEMCouplingserver to prepare for coupling with Fluent;

[0027] 10) Import the coupling interface udf into Fluent and connect it to EDEM so that Fluent can obtain the particle information of the grain pile in EDEM.

[0028] 11) The CFD-DEM coupled calculation method is used to transfer the interaction force between grain particles and the fluid-air mixture in the pores, and to obtain the pore distribution parameters of the grain pile. In the CFD-DEM coupled calculation method, there are two calculation models: the Lagrangian model and the Eulerian model. The Lagrangian model is suitable for cases where the volume of discrete phase does not exceed 15% of the total volume. The Eulerian model is suitable for cases where the proportion of particles to the total volume is greater than 10%. Since the proportion of particles to the total volume in the grain pile exceeds 50%, and particles play a dominant role, the Eulerian model is used for coupled calculation of the grain pile.

[0029] 12) The basic idea of ​​obtaining the pore distribution law through the CFD-DEM coupled calculation method is to give a fluid—air—with a certain initial velocity and let it pass through the grain pile to obtain pore information. The continuity equation of the fluid is:

[0030]

[0031] Where: ρ is the gas density, kg / m³ 2 t is time, s; v is gas velocity, m / s; ε is gas volume fraction; For Hamiltonian operators;

[0032] The differential equation of motion for the fluid is:

[0033]

[0034] Where: P is the pressure on the gas element, Pa; g is the acceleration due to gravity, m / s². 2 μ is the dynamic viscosity of the gas, μ = 1.83 × 10⁻⁶ -5 Pa·s; S is the momentum source term;

[0035] 13) Initialize on the Fluent side, setting the time step to 100 times the EDEM time step;

[0036] 14) Perform CFD-DEM coupled calculations;

[0037] 15) The CFD-DEM coupled calculation process is as follows: the force and position of each particle in the grain pile are solved by EDEM, the flow field related information of the fluid in the pores of the grain pile is solved by Fluent, and the characteristic parameters of the particles and fluids are transferred to each other through the coupling interface. Each iteration process forms a loop until the simulation ends.

[0038] 16) Post-process the data: Use a data processing system to process the pore distribution data, which is expressed as volume fraction, to generate an intuitive pore distribution cloud map.

[0039] The beneficial effects of this invention are as follows:

[0040] 1. Currently, there is no complete, objective, effective, and accurate numerical simulation system and method for grain silo porosity detection. This invention provides a numerical simulation system and method for grain silo porosity detection, which mainly solves the problems that the porosity distribution inside the grain pile cannot be directly observed and measured, the inaccurate inferred porosity, the high cost of experiments, and the inability to achieve real-time and accurate detection. It realizes rapid and high-precision detection of grain pile porosity.

[0041] 2. This invention utilizes the finite element method Fluent software and the discrete element method EDEM software to perform CFD-DEM coupled numerical simulation, realizing the pore distribution pattern of arbitrary regions or grain layers in grain piles with different storage types, particle shapes, particle sizes, and physical property parameters. It generates pore distribution cloud maps, providing accurate and intuitive porosity information for grain storage practices. This provides crucial porosity values ​​for mechanical ventilation of stored grain piles, helping relevant workers to promptly detect and handle abnormalities in grain piles and take appropriate management measures. It is of great significance for predicting and analyzing grain state changes and achieving safe grain storage, and provides a new technical means for studying complex grain storage systems with multi-factor coupling effects.

[0042] 3. All parameters in this invention are set in the simulation software Fluent and EDEM, which can model according to the geometric parameters and gradation of real particles. All parameters involved in the CFD-DEM coupled calculation are based on evidence. The simulation results can restore the essential properties of real grains, ensuring that the numerical simulation results are consistent with reality and have a wide range of applications. At the same time, it is easy to operate and the numerical simulation method is simple and effective. Attached Figure Description

[0043] Figure 1 A framework diagram of a numerical simulation system for detecting porosity in grain silos;

[0044] Figure 2 This is a physical model diagram of a cylindrical silo.

[0045] Figure 3 A diagram of a grain pile formed by the accumulation of particles;

[0046] Figure 4 This is a cloud map showing the porosity distribution of the XOZ cross section of the grain pile;

[0047] Figure 5 This is a cloud map showing the porosity distribution of the YOZ cross section of the grain pile.

[0048] Figure 6 This is a cloud map showing the porosity distribution of each horizontal section along the Z-axis of the grain pile height.

[0049] Wherein: Figure a is the cross section with Z = 0m; Figure b is the cross section with Z = 0.1m; Figure c is the cross section with Z = 0.2m; Figure d is the cross section with Z = 0.3m. Detailed Implementation

[0050] The present invention will now be described in conjunction with the accompanying drawings.

[0051] like Figure 1 As shown, the numerical simulation system for grain silo porosity detection of the present invention consists of module 1, module 2, module 3 and module 4, wherein...

[0052] Module 1 is the physical model creation and mesh generation module, used to create a physical model of the grain silo, set the fluid region and boundary names in the computational domain, generate the mesh and output a mesh format file for use by subsequent modules;

[0053] Module 2 is a particle accumulation simulation module, used to simulate the grain entering the warehouse and forming a grain pile. Based on the geometric parameters of real grain particles, a particle geometric model is established, and a mesh format file is imported as the grain warehouse geometric model. Particle size distribution and various contact parameters between particles and the grain warehouse are set, including: particle and grain warehouse density, Poisson's ratio, shear modulus, collision recovery coefficient, static friction coefficient, and rolling friction coefficient between particles and the grain warehouse, and collision recovery coefficient, static friction coefficient, and rolling friction coefficient between particles. A virtual pellet factory is set up to simulate the grain entering the warehouse and form particle accumulation to obtain the grain pile.

[0054] Module 3 is the DEM-CFD coupled calculation module, used to simulate the flow of fluid-air between pores in a grain pile and obtain pore information; import the mesh format file into Fluent, set various parameters, import the coupling interface udf, connect it with EDEM, initialize it, set the time step to 100 times the EDEM time step, and perform DEM-CFD coupled calculation.

[0055] Module 4 is the data processing module. Through the data processing system, it processes the pore distribution data in the form of volume fraction to generate an intuitive pore distribution cloud map.

[0056] Implementation Case 2: A numerical simulation method for grain silo porosity detection based on CFD-DEM coupled calculation, taking a cylindrical grain silo and soybeans and their impurities as an example, its features include the following steps:

[0057] 1) Use ANSYS DM to build a physical model of the silo, such as... Figure 2 As shown, the silo has a diameter D = 0.45m and a height H = 0.5m, and is meshed. The fluid region and boundary names are set in the computational domain.

[0058] 2) Output mesh format file;

[0059] 3) Using EDEM software, establish a particle geometric model based on the geometric parameters of real grain particles, and import the mesh format file as the geometric model of the silo.

[0060] 4) Through experiments, experience, and theoretical formulas, determine various physical property parameters of grain to ensure that all parameters involved in the CFD-DEM coupled numerical simulation are based on evidence, thus ensuring the reliability and accuracy of the numerical simulation results.

[0061] 5) Set the particle size distribution and various contact parameters between the particles and the silo in the EDEM: soybean Poisson's ratio 0.23, soybean shear modulus 63 MPa, and soybean density 1230.921 kg / m³. 3 The coefficient of friction between soybeans is 0.292, the coefficient of static friction between soybeans is 0.409, and the coefficient of static friction between soybeans and steel is 0.487.

[0062] 6) Set up a virtual pellet mill to simulate the grain entering the warehouse, forming pellet piles to obtain grain heaps, such as... Figure 3 As shown; the DEM-discrete element contact model of particles during the accumulation process is as follows:

[0063] The vibrational motion during particle contact is decomposed into tangential and normal directions. The normal vibrational motion equation F n for:

[0064]

[0065] During particle contact, the tangential vibration motion manifests as tangential sliding and particle rolling, with its tangential component F s The external torque M acting on the particle is expressed as follows:

[0066]

[0067]

[0068] Where: m 1,2 I represents the equivalent mass of particles i and j, in kg. 1,2 Let i be the equivalent rotational inertia of particles i and j, in kg·m. 2 ; s is the radius of rotation, m; μ n μ s θ represents the normal and tangential relative displacements of the particle, respectively, in meters (m); θ is the particle's rotation angle, in rad; K s K n These are the tangential and normal elastic coefficients in the contact model, respectively; c n c s These are the normal and tangential damping coefficients in the contact model, respectively.

[0069] 7) Use the CFD software ANSYS Fluent to simulate the flow of fluid-air in the pores of the grain pile;

[0070] 8) Import the mesh file into Fluent, set the parameters, set the Viscous model to SST k-epsilon, and set the material to grain particles consistent with the EDEM end.

[0071] 9) Start the EDEM Coupling server to prepare for coupling with Fluent;

[0072] 10) Import the coupling interface udf into Fluent and connect it to EDEM so that Fluent can obtain the particle information of the grain pile in EDEM.

[0073] 11) The CFD-DEM coupled calculation method is used to transfer the interaction force between grain particles and the fluid-air mixture in the pores, and to obtain the pore distribution parameters of the grain pile. In the CFD-DEM coupled calculation method, there are two calculation models: the Lagrangian model and the Eulerian model. The Lagrangian model is suitable for cases where the volume of discrete phase does not exceed 15% of the total volume. The Eulerian model is suitable for cases where the proportion of particles to the total volume is greater than 10%. Since the proportion of particles to the total volume in the grain pile exceeds 50%, and particles play a dominant role, the Eulerian model is used for coupled calculation of the grain pile.

[0074] 12) The basic idea of ​​obtaining the pore distribution law through the CFD-DEM coupled calculation method is to give a fluid—air—with a certain initial velocity and let it pass through the grain pile to obtain pore information. The continuity equation of the fluid is:

[0075]

[0076] Where: ρ is the gas density, kg / m³ 2 t is time, s; v is gas velocity, m / s; ε is gas volume fraction; For Hamiltonian operators;

[0077] The differential equation of motion for the fluid is:

[0078]

[0079] Where: P is the pressure on the gas element, Pa; g is the acceleration due to gravity, m / s². 2 μ is the dynamic viscosity of the gas, μ = 1.83 × 10⁻⁶ -5 Pa·s; S is the momentum source term;

[0080] 13) Initialize on the Fluent side, setting the time step to 100 times the EDEM time step;

[0081] 14) Perform CFD-DEM coupled calculations;

[0082] 15) The CFD-DEM coupled calculation process is as follows: the force and position of each particle in the grain pile are solved by EDEM, the flow field related information of the fluid in the pores of the grain pile is solved by Fluent, and the characteristic parameters of the particles and fluids are transferred to each other through the coupling interface. Each iteration process forms a loop until the simulation ends.

[0083] 16) Post-process the data: Using a data processing system, process the pore distribution data, which is expressed as a volume fraction, to generate an intuitive pore distribution cloud map, such as... Figure 4 Porosity distribution cloud map of XOZ section of grain pile, Figure 5 Porosity distribution cloud map of YOZ section of grain pile Figure 6 This is a cloud map showing the porosity distribution of each horizontal section along the Z-axis of the grain pile height.

Claims

1. A numerical simulation system for detecting porosity in grain storage facilities, characterized in that... It consists of Module 1, Module 2, Module 3 and Module 4. Module 1 is the physical model establishment and mesh generation module, which is used to establish the physical model of the grain silo, set the fluid region and boundary naming in the computational domain, generate the mesh and output the mesh format file for use by subsequent modules. Module 2 is a particle accumulation simulation module, used to simulate the grain entering the warehouse and forming a grain pile. Based on the geometric parameters of real grain particles, a particle geometric model is established, and a mesh format file is imported as the grain warehouse geometric model. Particle size distribution and various contact parameters between particles and the grain warehouse are set, including: particle and grain warehouse density, Poisson's ratio, shear modulus, collision recovery coefficient, static friction coefficient, and rolling friction coefficient between particles and the grain warehouse, and collision recovery coefficient, static friction coefficient, and rolling friction coefficient between particles. A virtual pellet factory is set up to simulate the grain entering the warehouse and form particle accumulation to obtain the grain pile. Module 3 is the DEM-CFD coupled calculation module, used to simulate the flow of fluid-air between pores in a grain pile and obtain pore information; import the mesh format file into Fluent, set various parameters, import the coupling interface udf, connect it with EDEM, initialize it, set the time step to 100 times the EDEM time step, and perform DEM-CFD coupled calculation. Module 4 is the data processing module. Through the data processing system, it processes the pore distribution data in the form of volume fraction to generate an intuitive pore distribution cloud map.

2. A numerical simulation method for detecting porosity in grain silos, characterized in that... Taking a cylindrical silo as an example, the steps include: 1) Use ANSYSDM to create a physical model of the silo and mesh it. Set the names of the fluid regions and boundaries in the computational domain. 2) Output mesh format file; 3) Using EDEM software, establish a particle geometric model based on the geometric parameters of real grain particles, and import the mesh format file as the geometric model of the silo. 4) Through experiments, experience, and theoretical formulas, determine various physical property parameters of grain to ensure that all parameters involved in the DEM-CFD coupled numerical simulation are based on evidence, and to ensure the reliability and correctness of the numerical simulation results. 5) Set the particle size distribution and various contact parameters between particles and silos in the EDEM, including: particle and silo density, Poisson's ratio, shear modulus, collision recovery coefficient, static friction coefficient, rolling friction coefficient between particles and silos, and collision recovery coefficient, static friction coefficient, and rolling friction coefficient between particles. 6) Set up a virtual pellet mill to simulate the grain entering the warehouse, forming a pellet pile; the DEM (Discrete Element Contact Model) of the pellets during the piling process is as follows: Decompose the vibration motion in the particle contact process into tangential and normal directions, and the normal vibration motion equation F n is: During the particle contact process, the tangential vibration motion is expressed as tangential sliding and rolling of the particle, and the tangential component F s , the external moment M on the particle is respectively expressed as: where: m 1,2 is the equivalent mass of the particle i, j, kg; I 1,2 is the equivalent moment of inertia of the particle i, j, kg-m 2 ; s is the radius of rotation, m; μ n , μ s are the normal and tangential relative displacement of the particles, m; θ is the rotation angle of the particle itself, rad; K s , K n are the tangential and normal elastic coefficients in the contact model; c n , c s are the normal and tangential friction coefficients in the contact model; 7) Use the CFD software ANSYS Fluent to simulate the flow of fluid-air in the pores of the grain pile; 8) Import the mesh file into Fluent, set the parameters, set the viscous model to SSTk-epsilon, and set the material to grain particles consistent with the EDEM end. 9) Open EDEM Couplingserver to prepare for coupling with Fluent; 10) Import the coupling interface udf into Fluent and connect it to EDEM so that Fluent can obtain the particle information of the grain pile in EDEM. 11) The DEM-CFD coupled calculation method is used to transfer the interaction force between grain particles and the fluid-air mixture in the pores, and to obtain the pore distribution parameters of the grain pile. In the DEM-CFD coupled calculation method, there are two calculation models: the Lagrangian model and the Eulerian model. The Lagrangian model is suitable for cases where the volume of discrete phase does not exceed 15% of the total volume. The Eulerian model is suitable for cases where the proportion of particles to the total volume is greater than 10%. Since the proportion of particles to the total volume in the grain pile exceeds 50%, and particles play a dominant role, the Eulerian model is used for coupled calculation of the grain pile. 12) The basic idea of ​​obtaining the pore distribution law through the DEM-CFD coupled calculation method is to give a fluid—air—with a certain initial velocity and let it pass through the grain pile to obtain pore information. The continuity equation of the fluid is: Where: ρ is the gas density, kg / m³ 2 t is time, s; v is gas velocity, m / s; ε is gas volume fraction; For Hamiltonian operators; The differential equation of motion for the fluid is: where: P is the pressure on the gas element, Pa; g is the acceleration due to gravity, m / s 2 ; μ is the dynamic viscosity of the gas, μ = 1.83 x 10 -5 Pa s; S is the momentum source term; 13) Initialize on the Fluent side, setting the time step to 100 times the EDEM time step; 14) Perform DEM-CFD coupled calculations; 15) The DEM-CFD coupled calculation process is as follows: the force and position of each particle in the grain pile are solved by EDEM, the flow field related information of the fluid in the pores of the grain pile is solved by Fluent, and the characteristic parameters of the particles and fluids are transferred to each other through the coupling interface. Each iteration process forms a loop until the simulation ends. 16) Post-process the data: Use a data processing system to process the pore distribution data, which is expressed as volume fraction, to generate an intuitive pore distribution cloud map.

Images