Dilution particle flow bidirectional coupling modeling method based on lattice-Boltzmann method

Abstract

The invention relates to the technical field of fluid dynamics, in particular to a lattice-Boltzmann method-based thin particle flow two-way coupling modeling method, which comprises the following steps of: performing field domain discretization by selecting spatial resolution, performing collision migration and viscosity coefficient setting, establishing a time step sequence and a fluid initial speed, determining a speed disturbance parameter, and establishing a two-way coupling model; calculating particle volume fraction and density distribution to obtain a speed distribution function set. According to the method, the spatial resolution is selected for field domain discretization, so that flow field division is more refined, the motion resolution of particles in different regions is enhanced, and the numerical error of the particles in a local region is reduced. And through collision migration and viscosity coefficient setting, it is ensured that the particles are evenly stressed in different flowing states, and the stability of the particle motion trail is improved. A time step sequence and a fluid initial speed are established, so that the time advancing process of particle movement is kept consistent, and speed deviation caused by mismatching of time steps is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of fluid dynamics, and particularly relates to a bidirectional coupling modeling method for dilute particle flow based on the lattice Boltzmann method. Background Art

[0002] Fluid dynamics studies the motion laws of fluids and the interaction between fluids and solid boundaries, covering two major directions: hydrostatics and fluid dynamics. This technical field is widely applied in multiple industries such as aerospace, ocean engineering, energy power, environmental science, and biofluids, and involves research directions such as turbulent flow simulation, fluid heat transfer, fluid-structure coupling, and multiphase flow.

[0003] In the prior art, the adaptability to different viscous environments is limited, and the force deviation of particles in low-viscosity or high-shear regions is relatively large, resulting in uneven traction ability of the fluid on the particles. The setting of the time step does not fully consider the synchronization between particles and the fluid, causing a lag effect in the particle motion trajectory and affecting the prediction of particle behavior in unsteady flow. The velocity perturbation parameter is not combined with the particle volume fraction and density distribution, making it difficult to accurately express the velocity fluctuation of particles in local regions and affecting the migration behavior of particles in the fluid. The velocity mapping method of the existing method relies on fixed-grid calculation, and the motion trend of particles in a non-uniform flow field is difficult to match the real fluid environment, affecting the accuracy of the particle flow state. Therefore, improvements are needed. Summary of the Invention

[0004] The purpose of the present invention is to solve the drawbacks existing in the prior art, and a bidirectional coupling modeling method for dilute particle flow based on the lattice Boltzmann method is proposed.

[0005] To achieve the above purpose, the present invention adopts the following technical solution. A bidirectional coupling modeling method for dilute particle flow based on the lattice Boltzmann method includes the following steps:

[0006] By selecting the spatial resolution for field discretization, performing collision migration and viscosity coefficient setting, establishing a time step sequence and the initial velocity of the fluid, determining the velocity perturbation parameter, calculating the particle volume fraction and density distribution, a set of velocity distribution functions is obtained;

[0007] Based on the set of velocity distribution functions, velocity mapping is performed by retrieving the velocity values corresponding to the discrete grid positions, boundary collision information is screened and the collision coordinates are recorded, a shear lift and rotational lift calculation process is established, and the fluid velocity and pressure data are read to obtain the boundary coupling parameter;

[0008] Based on the boundary coupling parameters, by locating the center position of the particles and comparing the coordinate differences between adjacent time steps, collision point checking and pressure gradient difference analysis are carried out to determine the scale of Brownian action and the change in particle volume distribution, and the particle motion characteristics are obtained;

[0009] Based on the particle motion characteristics, the momentum coupling relationship is determined by comparing the particle motion velocity and the local velocity field of the fluid, the velocity perturbation is revised, and the changes in the viscosity coefficient and the turbulence correction parameter are screened, and the particle diameter and the mesh division ratio are checked to obtain the two-way momentum influence amount.

[0010] Preferably, it further includes:

[0011] Based on the two-way momentum influence amount, the fluid velocity is redistributed by collecting the records of fluid velocity and pressure changes, the pressure gradient is revised, and the range of turbulence correction parameters and the particle volume fraction threshold are analyzed, and the velocity distribution after collision migration is sorted out to obtain the coupled flow state;

[0012] Based on the coupled flow state, by comparing the velocity and pressure values in the global dilute particle flow region, the velocity perturbation amplitude and the difference in particle volume fraction are checked, and the distribution information of the viscosity coefficient and the turbulence correction parameter in the results is extracted to obtain the coupled distribution of the dilute particle flow.

[0013] Preferably, the steps for obtaining the set of velocity distribution functions are as follows:

[0014] Based on the data selected according to the spatial resolution, the field area is divided and the collision migration is synchronously controlled, the local density information is screened and an index table is established, the collision corresponding positions and time step information are recorded, and a field synchronization record is generated;

[0015] Based on the field synchronization record, a time step sequence is established and the initial velocity of the fluid is set, the velocity perturbation parameters are calibrated and the particle volume fraction is calculated, the initial value of the viscosity coefficient is retrieved and assigned to the corresponding region, and a density parameter set is generated;

[0016] Based on the density parameter set, the local density distribution is statistically analyzed and the velocity change trend is recorded, the velocity offset data generated by the collision migration is checked and merged into the field index to obtain the set of velocity distribution functions.

[0017] Preferably, the steps for obtaining the boundary coupling parameters are as follows:

[0018] Based on the set of velocity distribution functions, the grid positions are retrieved and the corresponding velocity values are locked, the velocity data is screened to mark the potential collision coordinates and the adjacent velocity differences are registered, and a collision mark summary is generated;

[0019] Based on the summary of the collision marks, extract boundary collision information, correlate the shear lift and rotational lift parameters, check the differences in fluid velocity and pressure data, record the collision points and lift values, and generate a lift-velocity matching table;

[0020] Based on the lift-velocity matching table, read the pressure data, proofread the lift fluctuation range, screen the collision coordinates, output the collision data, and allocate it to the record table to obtain the boundary coupling parameters.

[0021] Preferably, the steps for obtaining the particle motion characteristics are as follows:

[0022] Based on the boundary coupling parameters, screen the particle center positions, record the coordinate changes of consecutive time steps, compare the influence of the collision points, and obtain the local pressure data to generate a particle position comparison table;

[0023] Based on the particle position comparison table, perform pressure gradient difference analysis, extract the Brownian action range, check the fluctuations in the particle volume distribution, and compare the collision point results to generate a gradient-Brownian mapping;

[0024] Based on the gradient-Brownian mapping, verify the collision point values, link the particle positions and Brownian characteristics, record the volume distribution trend, and summarize it into a table to obtain the particle motion characteristics.

[0025] Preferably, the steps for obtaining the two-way momentum influence amount are as follows:

[0026] Based on the particle motion characteristics, match the particle motion velocity with the local fluid velocity field, register the momentum difference, record the coupling index to quantify the coupling degree between the particle and the fluid, and generate momentum deviation data;

[0027] Based on the momentum deviation data, perform velocity perturbation correction, screen the change values of the viscosity coefficient, combine the turbulence correction parameters, compare the particle diameter with the grid division ratio, record the corresponding adjustment amounts, and incorporate them into the comparison table to generate velocity-viscosity correction values;

[0028] Based on the velocity-viscosity correction values, perform two-way momentum summation, output the velocity deviation correction results, register the momentum distribution information, and merge it into the global record table to obtain the two-way momentum influence amount.

[0029] Preferably, the steps for obtaining the coupled flow state are as follows:

[0030] Based on the two-way momentum influence amount, collect the fluid velocity, compare the numerical values of the previous and subsequent time steps, record the pressure changes, and summarize the collection results into a unified allocation structure to generate a pressure allocation record;

[0031] Based on the pressure distribution record, perform pressure gradient calibration and retrieve the range of turbulence correction parameters, register the particle volume fraction threshold and compare the collision migration situation, and generate a migration and parameter table;

[0032] Based on the migration and parameter table, re - distribute the fluid velocity and synchronize the pressure numerical change, integrate the velocity data information after collision migration into a unified index, and obtain the coupled flow state.

[0033] Preferably, the steps for obtaining the coupled distribution of dilute particle flow are as follows:

[0034] Based on the coupled flow state, extract the velocity and pressure numerical values of the global dilute particle flow and perform block - by - block comparison, register the difference in velocity perturbation amplitude, and generate a velocity perturbation comparison value;

[0035] Based on the velocity perturbation comparison value, screen the difference in particle volume fraction and extract the viscosity coefficient data, combine with the turbulence correction parameters for block - by - block registration, and generate a parameter and volume fraction record;

[0036] Based on the parameter and volume fraction record, merge the viscosity coefficient and turbulence correction parameters and integrate the velocity and pressure numerical values of all regions, output the distribution information and then archive it to obtain the coupled distribution of dilute particle flow.

[0037] Compared with the prior art, the advantages and positive effects of the present invention are as follows:

[0038] In the present invention, the spatial resolution is selected for field discretization to make the flow field division more refined, enhance the motion resolution of particles in different regions, and reduce the numerical error of particles in local regions. The collision migration and viscosity coefficient are set to ensure the force balance of particles in different flow states and improve the stability of particle motion trajectories. The time step sequence and the initial fluid velocity are established to keep the time advancement process of particle motion consistent and reduce the velocity deviation caused by the mismatch of time steps. The determination of the velocity perturbation parameter is combined with the calculation of the particle volume fraction and density distribution to make the motion change of particles in local regions more accurate and optimize the response of particles under the action of turbulence. The acquisition of the velocity distribution function set makes the motion direction and velocity distribution of particles in the grid more reasonable and improves the integrity of the fluid-particle coupling simulation. The velocity mapping is performed by retrieving the velocity values corresponding to the discrete grid positions to enhance the dynamic characteristics of particles under the traction of the fluid and make the velocity change more consistent with the actual flow state. The screening of boundary collision information and the recording of collision coordinates make the interaction between particles and the fluid boundary more continuous and optimize the motion trajectories of particles at complex geometric boundaries. The establishment of the calculation process of shear lift and rotational lift makes the rotational influence on particles in the flow field more accurate and enhances the reliability of particle trajectory prediction. The extraction of the particle center position is combined with the comparison of the time step coordinate differences to make the trend of particle force change clearer and improve the accuracy of particle motion trajectory prediction. The collision point inspection and the analysis of the pressure gradient difference make the force distribution of particles during the flow process more uniform and optimize the distribution characteristics of particles in different fluid regions. The calculation of the scale of Brownian action and the change of particle volume distribution makes the random motion of micro-scale particles more in line with the actual situation and improves the rationality of fluid-particle interaction. The momentum coupling calculation is combined with the revision of velocity perturbation to make the momentum exchange between the fluid and particles more balanced and optimize the influence process of the fluid on particles. The screening of viscosity coefficients and turbulence correction parameters makes the responses of different particles in different fluid environments more stable and improves the adaptability of simulation results. The inspection of the ratio of particle diameter to grid division makes the motion of particles at different grid scales smoother and reduces the numerical deviation of particles caused by scale mismatch during the simulation process. The redistribution of fluid velocity and the revision of the pressure gradient make the interaction between the fluid and particles more accurate and enhance the stability of the overall simulation. The inspection of the velocity perturbation amplitude and the difference in particle volume fraction makes the motion trend of particles match the change of the flow field and improves the accuracy of fluid-particle coupling calculation. Description of the Drawings

[0039] Figure 1 is a schematic diagram of the steps of the present invention. Detailed Embodiments

[0040] To make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only for explaining the present invention and are not intended to limit the present invention.

[0041] Please refer to Figure 1 , the present invention provides a technical solution, a two-way coupling modeling method for dilute particle flow based on the lattice Boltzmann method, including the following steps:

[0042] By selecting the spatial resolution for field discretization, performing collision migration and viscosity coefficient setting, establishing a time step sequence and the initial fluid velocity, determining the velocity perturbation parameter, calculating the particle volume fraction and density distribution, a set of velocity distribution functions is obtained;

[0043] Based on the set of velocity distribution functions, velocity mapping is performed by retrieving the velocity values corresponding to the discrete grid positions, boundary collision information is screened and the collision coordinates are recorded, the calculation processes of shear lift and rotational lift are established, and the fluid velocity and pressure data are read to obtain the boundary coupling parameters;

[0044] Based on the boundary coupling parameters, by locating the particle center position and comparing the coordinate differences of adjacent time steps, collision point inspection and pressure gradient difference analysis are performed to determine the scale of Brownian action and the change in particle volume distribution, and the particle motion characteristics are obtained;

[0045] Based on the particle motion characteristics, by comparing the particle motion velocity and the local fluid velocity field, the momentum coupling relationship is determined, the velocity perturbation is revised, and the changes in viscosity coefficient and turbulence correction parameters are screened, and the particle diameter and grid division ratio are checked to obtain the two-way momentum influence amount.

[0046] Based on the two-way momentum influence amount, by collecting the records of fluid velocity and pressure changes, the fluid velocity is redistributed, the pressure gradient is revised, and the range of turbulence correction parameters and the particle volume fraction threshold are analyzed, and the velocity distribution after collision migration is sorted out to obtain the coupled flow state;

[0047] Based on the coupled flow state, by comparing the velocity and pressure values within the entire dilute particle flow region, the velocity perturbation amplitude and the difference in particle volume fraction are checked, and the distribution information of the viscosity coefficient and turbulence correction parameters in the results is extracted to obtain the coupled distribution of the dilute particle flow.

[0048] The steps for obtaining the set of velocity distribution functions are as follows:

[0049] Based on the data selected according to the spatial resolution, field region division and synchronous control of collision migration are performed, local density information is screened and an index table is established, the collision corresponding positions and time step information are recorded, and a field synchronous record is generated;

[0050] Based on the field synchronous recording, establish the time step sequence and set the initial fluid velocity, proofread the velocity perturbation parameters and calculate the particle volume fraction, retrieve the initial value of the viscosity coefficient and assign it to the corresponding area, and generate a set of density parameters;

[0051] Based on the set of density parameters, conduct local density distribution statistics and record the velocity change trend, check the velocity offset data generated by collision migration and merge it into the field index to obtain a set of velocity distribution functions.

[0052] Specifically, obtain the original sampling results from the previously selected spatial resolution data, partition the overall field in combination with the grid range marked by the corresponding resolution. When partitioning, set five sections of the same length in both the horizontal and vertical directions respectively, forming a total of twenty-five sub-regions. Then, record the particle displacement at consecutive moments for each sub-region and determine whether the collision migration process is triggered. The determination of a collision can be calculated based on the relative movement amount and relative density ratio of the particle center coordinates at adjacent moments. If the movement amount exceeds the reference displacement standard and the local density ratio is greater than the average level, it is registered as a collision event. The reference displacement standard can be determined by calculating the average displacement μ and obtaining the variance σ from a large number of historical samples, and then superimposing μ and σ according to the principle of μ + σ. For example, in a certain test, if μ = 1.0 mm and σ = 0.4 mm, then 1.4 mm is used as the threshold. Whenever it exceeds 1.4 mm, it is regarded as a collision occurring. Mark all sub-regions that meet the collision conditions one by one and record them in the local data set, and distinguish the collision intensity within each sub-region. The collision intensity threshold can form a new boundary by adding 0.6 mm to the previous displacement threshold. If the actual displacement exceeds this boundary, it is counted as a strong collision, otherwise it is regarded as a weak collision. Then, summarize the local density information in all sub-regions, compare the average particle density of each sub-region with the global density mean value. If it is detected that the average particle density of a certain sub-region is more than twice the global mean value, it is classified as a high-density area and the corresponding flag is added. Subsequently, integrate all collision information and density information into a unified index structure for storing the collision position, collision time, collision intensity, and density status. Record the regions and moments corresponding to the collision events in chronological order and summarize them into a data sequence. Finally, obtain the field synchronous recording.

[0053] After obtaining the field synchronous record, first construct a time step sequence based on the timestamp information therein, and divide the time step sequence into several discrete units. The length of each unit can be set according to the required accuracy and hardware processing capabilities. For example, select 0.002 seconds as a basic time step. This value can be obtained by comparing the computational amounts at different sampling frequencies. When the sampling frequency is set to 500 Hz, the corresponding time step is approximately 0.002 seconds. If higher accuracy is required, the time step can be further reduced. Subsequently, set the initial velocity for the fluid within each time unit. The initial velocity can be allocated based on the average value and fluctuation range of historical flow velocity measurement data. For example, in a set of measurements, if the average flow velocity is 2.0 m / s and the standard deviation is 0.3 m / s, the initial velocity can be set within the range of 2.0 m / s ± 0.3 m / s. Then, perform the calibration of the velocity perturbation parameter. The velocity perturbation parameter can be derived from the measurement of the fluid turbulence characteristics, and a reference threshold T_v can be specified to control this perturbation range within the interval of the average value plus or minus the standard deviation. For example, set T_v = 0.1 m / s. If the measured perturbation is less than 0.1 m / s, the velocity perturbation is not adjusted. If it exceeds 0.1 m / s, the perturbation term is updated according to the measured value. Next, calculate the particle volume fraction by combining the number of particles and the known volume. If the particle volume fraction exceeds the set upper limit, the information of this time period can be additionally recorded in the subsequent process. The upper limit value can be obtained through empirical calculation. For example, in a series of tests, if the flow field fluctuation significantly increases when the volume fraction exceeds 0.15, then 0.15 can be set as the warning value of the particle volume fraction. Finally, retrieve the initial value of the viscosity coefficient and allocate it to the corresponding region, and integrate all the calculated viscosity coefficients and particle volume fractions into a centrally managed parameter list to generate a density parameter set.

[0054] On the premise that the density parameter set has been established, it is necessary to statistically analyze the local density distribution and record the trend of velocity changes. First, traverse all previously determined sub-regions, obtain the particle density values and fluid velocity values of each sub-region at different time steps, and compare these data with the global reference values. If the density value of a certain sub-region continuously exceeds the established threshold within a certain period, it is marked as a high-density period. This threshold can be determined by adding the global average density μ_d and the standard deviation σ_d. For example, in a certain statistical analysis, when μ_d = 0.25 and σ_d = 0.08, 0.33 can be set as the threshold, and values higher than 0.33 are regarded as high-density periods. At the same time, track the velocity changes during these periods. If the mean square deviation of the velocity from the previous time step is greater than the set range, it is marked as the velocity anomaly range. The velocity anomaly range can be selected by a similar method and continuously observed in subsequent steps. Then, check the previously confirmed collision events. If it is detected that the collision occurrence period overlaps with the high-density or velocity anomaly range, further adjust the relevant displacement information and record the velocity offset caused by the collision migration. Incorporate this offset into the field index for overall correlation analysis. Finally, after merging the velocity correction data caused by the collision migration into a unified index, comprehensively obtain the complete set of velocity distribution functions.

[0055] The steps for obtaining the boundary coupling parameters are as follows:

[0056] Based on the set of velocity distribution functions, perform grid position retrieval and lock the corresponding velocity values, screen the velocity data to mark potential collision coordinates and record the adjacent velocity differences, and generate a summary of collision marks;

[0057] Based on the summary of collision marks, extract boundary collision information and correlate the shear lift and rotational lift parameters, check the differences between the fluid velocity and pressure data and record the collision points and lift values, and generate a lift-velocity matching table;

[0058] Based on the lift-velocity matching table, read the pressure data and check the lift fluctuation range, screen the collision coordinates, output the collision data, and allocate it to the record table to obtain the boundary coupling parameters.

[0059] Specifically, after obtaining the set of velocity distribution functions, it is necessary to perform a specific search for grid positions and lock the corresponding velocity data. Each grid can be located through a coordinate system, and the corresponding velocity values are recorded in the set of velocity distribution functions. During the search, if the velocity difference between adjacent grids is found to be greater than a certain reference value, then this grid is marked as a potential collision coordinate. This reference value can be obtained through the statistical analysis of the velocity distribution in a typical area. For example, in a dynamic monitoring, if the average velocity difference is 0.5 m / s and the variance is 0.2 m / s, then a marking threshold can be set at around 0.7 m / s. Once the actual grid velocity difference exceeds 0.7 m / s, it is determined that there may be a large relative motion at this position, leading to a collision. All potential collision coordinates are listed one by one and associated with the adjacent velocity differences for correlation analysis. Subsequently, a collision marking summary is generated based on the list summary results, and the information of these grid coordinates is stored in a unified structure for use in the next boundary collision step.

[0060] After generating the collision marking summary, it is necessary to further extract the boundary collision information and associate the shear lift and rotational lift parameters with these collision positions. First, perform a comparison of fluid velocity and pressure data for all coordinates marked as potential collisions. If the difference between the measured pressure value and that of its neighboring grids exceeds the set upper limit, then a relatively high lift may be generated. The set upper limit can be obtained by combining the initial pressure measurement and fluid density calculation. For example, when the initial pressure mean is 1.5 Pa and the variance is 0.3 Pa, the first threshold can be set at 1.5 Pa + 0.3 Pa = 1.8 Pa. If it is higher than this threshold, then it is considered that there is an additional lift contribution at this position. Then, split this lift, including shear lift and rotational lift. The shear lift can be calculated by multiplying the flow velocity gradient by the corresponding shear coefficient, while the rotational lift is estimated based on the spin velocity of the particle and the local fluid viscosity coefficient. The shear coefficient and spin reference values used in these calculations need to be comprehensively set during the previous experiments in combination with fluid viscosity, particle surface characteristics, and flow velocity range. For example, set 0.02 as the shear coefficient, 0.3 rad / s as the spin reference value, and make fine-tuning according to the actual situation. If it is found that the lift value at this collision position is significantly different from that of the surrounding grids, then it is classified into the high-lift set and the specific value is recorded. Then, match and integrate all the recorded lift values and velocity components. Finally, generate a lift-velocity matching table.

[0061] After the lift-speed matching table is constructed, it is necessary to read the pressure data and check the lift fluctuation range. First, retrieve the pressure values at all collision points one by one, and calculate the maximum fluctuation amplitude of these pressure values within the time period they are in. Compare it with the determined fluctuation limit value, which can be set based on the superposition of the pressure mean and variance. For example, when the pressure mean within a certain period is 2.0 Pa and the variance is about 0.2 Pa, then 2.2 Pa can be regarded as the fluctuation limit value. If the actual fluctuation amplitude exceeds this value, it is recorded in a fluctuation anomaly list. Subsequently, screen out all the registered collision coordinate information, cross-confirm the fluctuation anomaly list with parameters such as shear lift and rotational lift, and make independent markings for the collision points where the fluctuation amplitude does not match the lift value. If it is found that the collision points meet the situation where the lift value rises significantly and the pressure fluctuation amplitude is high, they are summarized uniformly. After integrating these data, output them to the main record and generate a final collision data list at the end. Combine all the collision data to update the field index structure, and finally obtain the boundary coupling parameters.

[0062] The steps for obtaining the particle motion characteristics are as follows:

[0063] Based on the boundary coupling parameters, screen the particle center positions and record the coordinate changes in consecutive time steps, compare the influence of collision points and obtain local pressure data, and generate a particle position comparison table;

[0064] Based on the particle position comparison table, conduct pressure gradient difference analysis and extract the Brownian action range, check the fluctuation of the particle volume distribution and compare the collision point results, and generate a gradient-Brownian mapping;

[0065] Based on the gradient-Brownian mapping, conduct numerical verification of the collision points and link the particle positions with Brownian characteristics, record the volume distribution trend and summarize it into a table to obtain the particle motion characteristics.

[0066] Specifically, based on the boundary coupling parameters, first disassemble and extract all the registered boundary coupling parameter lists, which include the pressure records and collision position indexes at the current moment and adjacent time steps, and obtain the registered central coordinate information of the particles from this list. Subsequently, pair and compare these coordinates in chronological order, determine the displacement of the particles by calculating the coordinate differences between adjacent time steps, and compare the displacement with a pre-set reference threshold. This reference threshold can be set according to the statistically obtained average displacement μ d and variance σ d and can be expressed as T d = μ d + 1.5σ d , for example, in a certain statistics, if μ d = 0.5 mm and σ d = 0.2 mm, then T d= 0.5 + 1.5×0.2 = 0.8 mm is regarded as the threshold value. If it is found that the actual displacement is greater than 0.8, it is marked as abnormal movement. Within the coordinate range of the abnormal movement, the impact data of the collision point is compared. If the historical record of the collision point also shows a relatively high pressure value and the pressure difference from its adjacent grid is greater than the threshold value T p determined by the average pressure μ p and the standard deviation σ p = μ p + 2σ p , then this group of particles is associated and identified with the collision position, and its central position and local pressure value are additionally registered. Then, the central coordinates and pressure information of all particles are integrated into a data index in ascending order of time. Finally, a particle position comparison table is generated based on these coordinates and pressure information.

[0067] Based on the particle position comparison table, the difference in pressure gradient at different time steps is gradually extracted for each record, and it is compared with the marked collision position. During the processing, first calculate the average horizontal increment and vertical increment of the local gradient, and then add the two to determine whether it is greater than the gradient threshold obtained from historical statistics. Denote this threshold as T g and it can be set in the form of μ g + βσ g . For example, in a batch of measurements, if the average gradient μ g = 0.05 Pa / m and the standard deviation σ g = 0.01 Pa / m, let β = 3, then T g = 0.05 + 3×0.01 = 0.08 Pa / m. If the calculated instantaneous gradient difference exceeds 0.08, it is marked in the data. Then, combined with the time period and position coordinates listed in the particle position comparison table, determine whether these high gradient difference values overlap with the previously registered collision point records. If they match, the corresponding time period is extracted and the particle volume distribution fluctuation is calculated internally. The particle volume distribution fluctuation can be obtained by dividing the particle count within the time series by the corresponding local volume. If its value is greater than a certain upper threshold set by long-term observation, it can be registered as an abnormal distribution situation. This upper threshold can be obtained by fitting the average volume fraction and variance. For example, adding the average volume fraction μ v and its standard deviation σ v gives T v = μ v + 2σ v . Subsequently, these high gradient differences are mapped and registered with the abnormal distribution positions, and then the gradient and Brown mapping can be generated.

[0068] Based on the gradient and Brownian mapping, the data related to the collision position are numerically verified, and the time series and coordinate indexes corresponding to these collision points are confirmed in the particle position comparison table. Then, these indexes are compared with the previously extracted Brownian action range data. The Brownian action range can be set by observing the random motion amplitude of particles at a very small scale and setting several segmented thresholds. If the random motion amplitude exceeds the set minimum threshold, the Brownian action is considered to be obvious. For example, when counting a large number of particles, if the average random deviation μ b =0.02mm and standard deviation σ b =0.005mm, T b =μ b +1.2σ b =0.026mm is used as the preliminary judgment standard of Brownian action. If the random offset value of a particle is higher than 0.026 in multiple time step records and has a strong overlap with the coordinates of the collision point, the number of the particle and its center position are marked together. Subsequently, the Brownian feature information is linked in the data record to form a corresponding relationship between multiple particles and their Brownian ranges, and the volume distribution trend is checked in sequence. The volume distribution trend can be evaluated by the slope of the local volume fraction curve in several consecutive time steps. If the slope is positively increasing and has exceeded the baseline value, it is regarded as an upward trend in the volume distribution. These results are summarized in a table to obtain the particle motion characteristics.

[0069] The steps to obtain the bidirectional momentum influence are:

[0070] Based on the particle motion characteristics, the particle motion velocity is matched with the local velocity field of the fluid and the momentum difference is registered. The coupling index is recorded to quantify the degree of coupling between the particle and the fluid, and the momentum deviation data is generated.

[0071] Based on the momentum deviation data, the velocity disturbance is revised and the viscosity coefficient change value is screened. The particle diameter and the grid division ratio are compared with the turbulence correction parameters. The corresponding adjustment amount is recorded and included in the comparison table to generate the velocity and viscosity correction values.

[0072] Based on the velocity and viscosity correction values, the bidirectional momentum is summarized and the velocity deviation revision result is output. The momentum distribution information is registered and merged into the global record table to obtain the bidirectional momentum influence.

[0073] Specifically, based on the particle motion characteristics, the particle motion velocity is first compared with the local fluid velocity field in all recorded time steps, and the instantaneous difference between the two is registered at each spatial grid point. If the difference exceeds the historical average velocity μ s and standard deviation σ s Set the threshold T s It is marked as the momentum difference concentration area. For example, if μ s =1.2m / s and σ s= 0.2 m / s, T can be taken as s = 1.2 + 2×0.2 = 1.6 m / s as the threshold. Once the velocity difference at a certain point is greater than 1.6, it is classified into the momentum difference concentration area. Then, in these concentration areas, record its position and the corresponding time step according to the particle number. Set a coupling index by comparing the momentum exchange between the particle and the fluid. The coupling index can be taken in the form of κ = (M k / M f )×100%, where M k is the measured value of particle momentum and M f is the measured value of fluid momentum. The specific measurement method can be calculated in segments according to the particle mass and fluid density in actual applications. After that, sum up these momentum differences one by one and make numerical records, and finally generate momentum deviation data.

[0074] Based on the momentum deviation data, it is necessary to read the registered velocity difference and the corresponding time step information item by item. On this basis, perform velocity perturbation revision. First, select those particles that showed a relatively high momentum deviation in the previous step, calculate the difference between their actual movement velocity and the local fluid velocity, and compare this difference with the reference perturbation threshold. The reference perturbation threshold can be obtained comprehensively according to the historical measured average perturbation μ r and variance σ r to get T r = μ r + 1.5σ r . For example, if μ r = 0.3 m / s and σ r = 0.05 m / s, then T r is about 0.375 m / s. When the difference exceeds 0.375, the particle velocity needs to be revised again. When revising, perform interpolation calculation on the actual measured velocity and the fluid velocity, and then compare with the change value of the viscosity coefficient. If the viscosity coefficient correction amount is greater than the reference offset, record this part of the data additionally. Then, combine the turbulence correction parameters to compare the particle diameter and the grid division ratio. For example, when the particle diameter is about 0.2 and the unit length of the grid division is 0.5, the relative ratio can be recorded as 0.4. If this ratio falls outside the pre-set threshold range, continue to record the correction result. Incorporate all the obtained correction amounts into the comparison table, and finally generate the velocity and viscosity correction values.

[0075] Based on the velocity and viscosity correction values, summarize the two-way momentum data at each time step and each grid point. First, count the velocity and viscosity coefficient conditions of all particles after the above-mentioned correction, and cross-compare them with the velocity field data of the fluid. Whenever it is found that the difference between the particle movement velocity and the local fluid velocity field is greater than the threshold T s set in the previous stage, record the momentum information of this point again, and compare the fluctuation of the viscosity coefficient. For the fluctuation amplitude, it can be μ η + γση Set a threshold value, where μ η is the average viscosity coefficient and σ η is its variance, and γ is an empirical factor. If it is detected that the change in the viscosity coefficient has exceeded this threshold, supplementary registration is performed in the look-up table. Subsequently, all momentum information and particle velocity correction parameters are integrated into a unified record. Finally, the velocity deviation revision result is output, and the momentum distribution information is registered at the end and merged into the global record table to obtain the two-way momentum influence quantity.

[0076] The steps for obtaining the coupled flow state are as follows:

[0077] Based on the two-way momentum influence quantity, fluid velocity is collected and the values at the previous and subsequent time steps are compared. The pressure change is recorded and the collection results are summarized into a unified distribution structure to generate a pressure distribution record;

[0078] Based on the pressure distribution record, pressure gradient correction is performed and the range of turbulence correction parameters is retrieved. The threshold of the particle volume fraction is registered and the collision migration situation is compared to generate a migration and parameter table;

[0079] Based on the migration and parameter table, the fluid velocity is redistributed and the pressure numerical change is synchronized. The velocity data information after collision migration is integrated into a unified index to obtain the coupled flow state.

[0080] Specifically, based on the two-way momentum influence quantity, first measure or read the fluid velocity at all time steps. For example, in a certain collection, record the fluid velocity values of each grid at an interval of 0.001 seconds, and obtain the velocity value of the same grid at the previous time step. Then, take the difference between the two to find the velocity change amount. If this change amount exceeds the threshold T v estimated from previous measurements, then mark it as a large change position. The threshold T v can be calculated in the form of μ v + kσ v where μ v is the average value of the velocity change amount, σ v is the standard deviation of the velocity change amount, and k is an empirical coefficient. For example, in a specific scenario, μ v = 1.0 m / s and σ v = 0.2 m / s. When k = 2, then T v = 1.4 m / s. Immediately list all the positions marked as large changes and include them in the comparison list. Then, collect the pressure records for each corresponding position, obtain the pressure values at the current and previous moments at this position through sensors or numerical simulations, and subtract the two to get the pressure change amount. If the pressure change amount is also greater than the threshold T p set according to the previous statistics, mark it additionally in the data. The threshold T p can also be obtained through μ p +

[0081] k′σ p Formal calculation, where μ p is the average value of the pressure change, σ p is the standard deviation, and k′ is another empirical coefficient. For example, when k′ is taken as 3, if μ p = 1.5 Pa and σ p = 0.2 Pa, then T p = 2.1 Pa. After completing the analysis of the velocity and pressure differences for all grids, these collected results are centralized and summarized into an allocation structure to form a list containing multiple records such as grid numbers, time steps, velocity changes, and pressure changes. Finally, it is named the pressure allocation record.

[0082] Based on the pressure allocation record, traverse and analyze the pressure difference data listed therein. First, calculate the spatial gradient value corresponding to each record and compare it with the historical mean. If the gradient value exceeds the threshold T grad obtained from the previous collection results, perform calibration. The threshold T grad can be set in the form of μ grad + cσ grad where μ grad is the historical average gradient, σ grad is its standard deviation, and c is an empirical factor. For example, if μ grad = 0.05 Pa / m and σ grad = 0.01 Pa / m, when c = 2, T grad = 0.07 Pa / m is obtained. During calibration, the records exceeding this threshold are updated to the corrected pressure gradient values. Then, retrieve the range of turbulence correction parameters, and obtain the possible parameter intervals by querying the previously stored set of turbulence parameter settings. For example, configure and mark them incrementally by 0.05 in the interval from 0.1 to 0.3. Subsequently, register the threshold of the particle volume fraction, which can be calculated from the mean μ vf and variance σ vf obtained from long-term observations. For example, if μ vf = 0.08 and σ vf = 0.01, T vf = μ vf + 3×σ vf = 0.11 can be used as the volume fraction threshold. Finally, check against the collision migration situation. If a certain record shows spatio-temporal overlap in the collision data and the volume fraction also exceeds 0.11, it is classified into the high migration risk category, and the comprehensive information is integrated to generate the migration and parameter table.

[0083] Based on the migration and parameter table, extract the corresponding fluid velocity setting information and re - distribute it in combination with the pressure change situation at the current time step. First, retrieve the previously recorded grid numbers and collision migration status, and perform local velocity updates for scenarios with signs of collision and where the turbulence correction parameter has not exceeded the predefined maximum value. If it is found that the number of particles in the grid is large and the volume fraction is above the previously set threshold T vf above, then when re - distributing the velocity, a proportional interpolation is made between the actually measured value and the historical mean to obtain the new velocity result. The ratio can be calculated based on the local number of particles and the maximum number of particles that the grid can accommodate. For example, when the maximum number of particles is 500 and the current measurement is 450, the interpolation coefficient α = 450 / 500 = 0.9 can be set. Thus, the measured velocity and the historical mean velocity are combined at a ratio of 0.9:0.1. Subsequently, all the new velocity data after collision migration are added to the index structure, and the pressure value is synchronously updated during this process. The pressure data and the velocity data are paired and placed in the same record. Finally, a comprehensive entry covering the grid number, turbulence correction parameter, volume fraction, and updated velocity is formed to obtain the coupled flow state.

[0084] The steps for obtaining the coupled distribution of dilute particle flow are as follows:

[0085] Based on the coupled flow state, extract the velocity and pressure numerical values of the global dilute particle flow and compare them in blocks, register the difference in velocity perturbation amplitude, and generate the velocity perturbation comparison value;

[0086] Based on the velocity perturbation comparison value, screen the difference in particle volume fraction and extract the viscosity coefficient data, and register them in blocks in combination with the turbulence correction parameter to generate the parameter - volume fraction record;

[0087] Based on the parameter - volume fraction record, merge the viscosity coefficient and the turbulence correction parameter and integrate the velocity and pressure numerical values of all regions. After outputting the distribution information, archive it to obtain the coupled distribution of dilute particle flow.

[0088] Specifically, based on the coupled flow state, uniformly extract the velocity and pressure numerical values within the global range and compare them in blocks in the horizontal and vertical directions. First, divide the entire space into several regions of equal size and sequentially traverse the velocity numerical values of each region, compare their differences with the surrounding regions. If the velocity difference between a certain region and the average velocity of the adjacent region is greater than the velocity perturbation threshold T dist set by historical tests, then register this difference value, T dist can be obtained in the form of μ dist +dσ dist where μ dist is the average value of the velocity difference, σ dist is the standard deviation, and d is an empirical coefficient. For example, μ dist = 0.5m / s, σ dist= 0.2 m / s, when d = 2, T dist = 0.9 m / s. Then, the velocity differences in each region are summarized, and the corresponding pressure levels are statistically calculated in a block manner. If there are obvious velocity differences in a certain block area, these difference values are listed separately and the coordinate indices of the regions to which they belong are marked. To achieve an overall view of the dilute particle flow in the entire domain, the extracted pressure data also needs to be compared in blocks. If it is found that the pressure deviates significantly from the mean value, it is included in an independent abnormal set. After all the block results are statistically calculated, the velocity perturbation amplitudes are registered one by one, and finally the velocity perturbation comparison values are generated.

[0089] Based on the velocity perturbation comparison values, traverse the list of difference amplitudes in each block, and check the entries marked as exceeding the velocity perturbation threshold with the previously saved particle volume fraction information. If it is found that the volume fraction in the current block area has exceeded the previously set threshold T vf and the perturbation difference is also higher than T dist , the corresponding grid or region is marked as a potentially unstable section. At the same time, retrieve the locally registered viscosity coefficient data and summarize them into a newly created comparison item. The viscosity coefficient here can adopt the average value μ η and the standard deviation σ η Calculate and mark the upper and lower limits of the interval. For example, when μ η = 0.0012 Pa·s and σ η = 0.0003 Pa·s, the allowable variation range can be set as μ η ±2σ η , that is, 0.0006 - 0.0018 Pa·s. Then, in combination with the already registered turbulence correction parameters, perform block registration according to their block positions, pair the turbulence parameter values of each region with the viscosity coefficient and volume fraction, record them as a segmented list, and then merge them into a whole structure, and finally generate the parameter and volume fraction record.

[0090] Based on the parameter and volume fraction record, perform a merging process on all the registered viscosity coefficients and turbulence correction parameters. First, compare the parameter values in the same region or adjacent regions. If the fluctuation of a certain parameter entry between multiple time steps exceeds the fluctuation threshold T para set by the previous measurement, then take its average value as the final parameter record. T para can be calculated in the form of μ para +fσ para , where μ para is the historical average value of this parameter, and σ parais the standard deviation, and f is the empirical constant obtained from the actual measurement. Then, all the finally obtained viscosity coefficients and turbulence correction parameters are combined with the velocity and pressure values of the previous regions. The combination method can follow the criteria of the same coordinate grid or adjacent position combination, integrating the velocity, pressure, volume fraction, and the corresponding viscosity coefficient and turbulence correction parameter of the same grid into one record. Finally, the combined results of each region are summarized and archived in sequence to obtain the coupled distribution of dilute particle flow.

[0091] The above is only the preferred embodiment of the present invention, and it does not limit the present invention in other forms. Any person skilled in the art may use the technical content disclosed above to make changes or modifications into equivalent embodiments with equivalent changes and apply them to other fields. However, as long as it does not depart from the technical solution content of the present invention, any simple modification, equivalent change, and modification made to the above embodiments based on the technical essence of the present invention still fall within the protection scope of the technical solution of the present invention.

Claims

1. A two-way coupling modeling method for dilute granular flows based on the lattice Boltzmann method, characterized in that, It includes the following steps: By selecting the spatial resolution for field discretization, conducting collision migration and setting the viscosity coefficient, establishing a time step sequence and the initial velocity of the fluid, determining the velocity perturbation parameter, calculating the particle volume fraction and density distribution, a set of velocity distribution functions is obtained; Based on the set of velocity distribution functions, velocity mapping is performed by retrieving the velocity values corresponding to the discrete grid positions, boundary collision information is screened and the collision coordinates are recorded, the calculation processes of shear lift and rotational lift are established, and the fluid velocity and pressure data are read to obtain the boundary coupling parameter; Based on the boundary coupling parameter, by locating the center position of the particle and comparing the coordinate differences of adjacent time steps, collision point inspection and pressure gradient difference analysis are carried out to determine the scale of Brownian action and the change in particle volume distribution, and the particle motion characteristics are obtained; Based on the particle motion characteristics, by comparing the particle motion velocity and the local velocity field of the fluid to determine the momentum coupling relationship, velocity perturbation is revised and the change amounts of the viscosity coefficient and the turbulence correction parameter are screened, and the particle diameter and the grid division ratio are checked to obtain the two-way momentum influence amount.

2. The two-way coupling modeling method for dilute particle flow based on the lattice Boltzmann method according to claim 1, wherein It also includes: Based on the two-way momentum influence amount, by collecting the records of fluid velocity and pressure changes for re-distribution of the fluid velocity, pressure gradient is revised and the range of the turbulence correction parameter and the particle volume fraction threshold are analyzed, and the velocity distribution after collision migration is sorted out to obtain the coupled flow state; Based on the coupled flow state, by comparing the velocity and pressure values in the global dilute particle flow region, the velocity perturbation amplitude and the difference in particle volume fraction are checked, and the distribution information of the viscosity coefficient and the turbulence correction parameter in the results is extracted to obtain the coupled distribution of the dilute particle flow.

3. The bidirectional coupling modeling method for dilute particle flow based on the lattice Boltzmann method according to claim 1, characterized in that The steps for obtaining the set of velocity distribution functions are as follows: Based on the data selected according to the spatial resolution, field region division and synchronous control of collision migration are carried out, local density information is screened and an index table is established, the collision corresponding positions and time step information are recorded, and a field synchronous record is generated; Based on the field synchronous record, a time step sequence is established and the initial velocity of the fluid is set, the velocity perturbation parameter is proofread and the particle volume fraction is calculated, the initial value of the viscosity coefficient is retrieved and assigned to the corresponding region, and a density parameter set is generated; Based on the density parameter set, local density distribution statistics are carried out and the velocity change trend is recorded, the velocity offset data generated by collision migration are checked and merged into the field index to obtain a set of velocity distribution functions.

4. The two-way coupling modeling method for dilute particle flow based on the lattice Boltzmann method according to claim 1, characterized in that The steps for obtaining the boundary coupling parameter are as follows: Based on the set of velocity distribution functions, grid position retrieval is carried out and the corresponding velocity values are locked, the velocity data are screened to mark potential collision coordinates and the adjacent velocity differences are registered, and a collision mark summary is generated; Based on the collision mark summary, boundary collision information is extracted and the shear lift and rotational lift parameters are associated, the differences between the fluid velocity and pressure data are checked and the collision points and lift values are recorded, and a lift and velocity matching table is generated; Based on the lift and velocity matching table, pressure data are read and the lift fluctuation range is proofread, the collision coordinates are screened and the collision data are output and assigned to the record table to obtain the boundary coupling parameter.

5. The two-way coupling modeling method for dilute particle flow based on the lattice Boltzmann method according to claim 1, characterized in that The steps for obtaining the particle motion characteristics are as follows: Based on the boundary coupling parameter, screen the central positions of the particles, record the coordinate changes at consecutive time steps, compare the influence of the collision points, obtain the local pressure data, and generate a particle position comparison table; Based on the particle position comparison table, perform pressure gradient difference analysis and extract the Brownian action range, check the fluctuations in the particle volume distribution, compare with the collision point results, and generate a gradient-Brownian map; Based on the gradient-Brownian map, verify the numerical values of the collision points, link the particle positions with the Brownian characteristics, record the volume distribution trend, and summarize it in a table to obtain the particle motion characteristics.

6. The bidirectional coupling modeling method for dilute particle flow based on the lattice Boltzmann method according to claim 1, characterized in that, The steps for obtaining the two-way momentum influence amount are as follows: Based on the particle motion characteristics, match the particle motion speed with the local fluid velocity field, register the momentum difference, record the coupling index to quantify the coupling degree between the particles and the fluid, and generate momentum deviation data; Based on the momentum deviation data, revise the velocity perturbation, screen the change values of the viscosity coefficient, combine the turbulence correction parameters, compare the particle diameter with the grid division ratio, record the corresponding adjustment amount, and incorporate it into the comparison table to generate velocity and viscosity correction values; Based on the velocity and viscosity correction values, summarize the two-way momentum and output the revised results of the velocity deviation, register the momentum distribution information, and merge it into the global record table to obtain the two-way momentum influence amount.

7. The bidirectional coupling modeling method for dilute particle flow based on the lattice Boltzmann method according to claim 2, characterized in that, The steps for obtaining the coupled flow state are as follows: Based on the two-way momentum influence amount, collect the fluid velocity, compare the numerical values at the previous and subsequent time steps, record the pressure change, and summarize the collection results into a unified distribution structure to generate a pressure distribution record; Based on the pressure distribution record, revise the pressure gradient, retrieve the range of the turbulence correction parameters, register the particle volume fraction threshold, compare with the collision migration situation, and generate a migration-parameter table; Based on the migration-parameter table, redistribute the fluid velocity, synchronize the pressure numerical change, and integrate the velocity data information after the collision migration into a unified index to obtain the coupled flow state.

8. The bidirectional coupling modeling method for dilute particle flow based on the lattice Boltzmann method according to claim 2, characterized in that, The steps for obtaining the dilute particle flow coupling distribution are as follows: Based on the coupled flow state, extract the velocity and pressure numerical values of the global dilute particle flow, compare them in blocks, register the difference in the velocity perturbation amplitude, and generate a velocity perturbation comparison value; Based on the velocity perturbation comparison value, screen the difference in the particle volume fraction, extract the viscosity coefficient data, and perform block registration in combination with the turbulence correction parameters to generate a parameter-volume fraction record; Based on the parameter-volume fraction record, merge the viscosity coefficient and the turbulence correction parameters, integrate the velocity and pressure numerical values of all regions, output the distribution information and archive it to obtain the dilute particle flow coupling distribution.

Images