12.1.5. Discrete element model (DEM)¶
12.1.5.1. Discrete Element - Common¶
These are keywords common to Discrete Element Model (DEM) and Particles In Cell (PIC).
12.1.5.1.1. DES_DIFFUSE_WIDTH¶
Data Type: DOUBLE PRECISION
Applies to solids model(s): DEM
The length scale used to smooth dispersed phase averaged fields by solving a diffusion equation. This approach is typically used when particle sizes near or exceed the size of the Eulerian grid cell sizes.
Mean field diffusion is disabled if DES_DIFFUSE_WIDTH is not specified.
Mean field diffusion cannot be used with the GARG_2012 interpolation scheme.
It is recommended that mean field diffusion be used in conjunction with DES_EXPLICITLY_COUPLED to minimize the computational cost of diffusing field data.
The DES diffusion equation is listed as equation type 10 in the Numerical parameters section.
12.1.5.1.2. DES_EPG_CLIP¶
Data Type: DOUBLE PRECISION
Applies to solids model(s): DEM
Limit at which void fraction calculation is clipped for all Lagrangian models (DEM, PIC, CGP, SQP, GSP). If the computed void fraction in a cell goes below DES_EPG_CLIP, it will be set to DES_EPG_CLIP. Leave DES_EPG_CLIP undefined to turn off the clipping.
12.1.5.1.3. DES_EXPLICITLY_COUPLED¶
Data Type: LOGICAL
Applies to solids model(s): DEM
Enable/Disable explicit coupling of DEM solids and the fluid. This algorithm is presently limited to hydrodynamic simulations.
Name |
Default? |
Description |
|---|---|---|
|
◉ |
The fluid and particles calculate interphase forces at their respective time scales. The fluid phase calculates the interphase coupling forces once per fluid time step. Similarly, DEM particles calculate the interface coupling forces at each solids time-step. The DEM must also bin particles to the fluid grid and recalculate the fluid volume fraction every time-step. |
|
Interphase forces are calculated during the fluid time step and stored for each particle. The interphase forces are then distributed among the solids time-steps. This approach can substantially reduce the computational overhead for coupled simulations. |
12.1.5.1.4. DES_INTERP_MEAN_FIELDS¶
Data Type: LOGICAL
Applies to solids model(s): DEM
Enable/disable interpolation of particle data (e.g., solids volume and drag force) from a particle’s position to the Eulerian grid.
Name |
Default? |
Description |
|---|---|---|
|
◉ |
Assign particle data to the fluid grid cell containing the particle’s center. |
|
Interpolate particle data from the particle’s position to the 27-cell neighborhood surrounding the particle. |
12.1.5.1.5. DES_INTERP_ON¶
Data Type: LOGICAL
Applies to solids model(s): DEM
Enable/disable interpolation of field quantities to a particle’s position. This is used in calculating gas-particle interactions, such as the drag force.
Name |
Default? |
Description |
|---|---|---|
|
◉ |
Use fluid values from the cell containing the particle’s center. |
|
Interpolate fluid values from the 27-cell neighborhood to a particle’s position. |
12.1.5.1.6. DES_INTERP_SCHEME¶
Data Type: CHARACTER
Applies to solids model(s): DEM
Specify the scheme used to map data to/from a particle’s position and the Eulerian grid. This keyword is required when DES_INTERP_MEAN_FIELDS and/or DES_INTERP_ON are specified.
Name |
Default? |
Description |
|---|---|---|
|
◉ |
Do not use interpolation. |
|
Interpolate to/from a particle’s position using the corners (nodes) of the fluid cells. This was the default behavior prior to version 2015-1. See Garg et al. (2012) Documentation of the open-source MFiX-DEM software for gas-solids flows. |
|
|
Divided particle volume method: Information is interpolated to/from a particle’s position using a square filter of size DES_INTERP_WIDTH. |
|
|
Linear interpolation: Hat functions are used to distribute particle information. PIC only. |
|
|
Divided particle volume method using satellites. |
12.1.5.1.7. DES_INTERP_WIDTH¶
Data Type: DOUBLE PRECISION
Applies to solids model(s): DEM
Length used in interpolating data to/from a particle’s position and the Eulerian grid. The interpolation width is only applicable to the DPVM_SQUARE and DPVM_GAUSS interpolation schemes as the GARG_2012 scheme’s interpolation width is determined by the Eulerian grid dimensions.
The interpolation half-width cannot exceed the minimum cell dimension because interpolation is restricted to the 27-cell neighborhood surrounding a particle (9-cell neighborhood in 2D).
It is recommended that the DES_INTERP_WIDTH be set equal to the maximum particle diameter when using STL defined boundaries. Field data can be smoothed by specifying DES_DIFFUSE_WIDTH.
12.1.5.1.8. DES_INTG_METHOD¶
Data Type: CHARACTER
Applies to solids model(s): DEM
Time stepping scheme. Note that DEM solids species mass fraction always use firt-order Euler scheme regardless of this setting.
Name |
Default? |
Description |
|---|---|---|
|
◉ |
First-order Euler scheme. |
|
Second-order ADAMS_BASHFORTH scheme. (DEM only) |
12.1.5.1.9. DES_ONEWAY_COUPLED¶
Data Type: LOGICAL
Applies to solids model(s): DEM
Run one-way coupled simulations [default .FALSE.]. If set, the fluid does not see the particles in terms of drag force. The effect of particle volume is still felt by the fluid through non-unity voidage values.
Name |
Default? |
Description |
|---|---|---|
|
◉ |
Two-way particle-fluid coupling. |
|
One-way particle-fluid coupling (fluid does not see particle drag). |
12.1.5.1.10. DESGRIDSEARCH_IMAX¶
Data Type: INTEGER
Applies to solids model(s): DEM
Number of des grid cells in the I-direction. If left undefined, then it is set by MFiX such that its size equals three times the maximum particle diameter with a minimum of 1 cell.
12.1.5.1.11. DESGRIDSEARCH_JMAX¶
Data Type: INTEGER
Applies to solids model(s): DEM
Number of des grid cells in the J-direction. If left undefined, then it is set by MFiX such that its size equals three times the maximum particle diameter with a minimum of 1 cell.
12.1.5.1.12. DESGRIDSEARCH_KMAX¶
Data Type: INTEGER
Applies to solids model(s): DEM
Number of des grid cells in the K-direction. If left undefined, then it is set by MFiX such that its size equals three times the maximum particle diameter with a minimum of 1 cell.
12.1.5.1.13. GENER_PART_CONFIG¶
Data Type: LOGICAL
Applies to solids model(s): DEM
Automatically generate the initial particle position and velocity data based on the parameters specified for each initial condition (IC) region.
Name |
Default? |
Description |
|---|---|---|
|
Generate particle configuration based on the initial condition parameters. Data provided in the particle_input.dat file, if present, is ignored. |
|
|
◉ |
Particle position and velocity data are provided in the particle_input.dat file. A runtime error occurs if this file is not provided. |
12.1.5.1.14. PARTICLES¶
Data Type: INTEGER
Applies to solids model(s): DEM
Number of particles to be read in from the particle_input.dat file. This value is ignored when using automatic particle generation. A simulation with a mass inflow BC can start without solids by setting PARTICLES = 0.
12.1.5.2. Discrete Element Model¶
These keywords relate to the Discrete Element Model (DEM).
12.1.5.2.1. ASPERITIES¶
Data Type: DOUBLE PRECISION
Mean radius of surface asperities that influence the cohesive force. See H. Rumpf, Particle Technology, Chapman & Hall, London/New York, 1990.
12.1.5.2.2. DES_BUFF_RESIZE_FACTOR¶
Data Type: DOUBLE PRECISION
Growth factor when resizing send/recv buffers (Default is 0.5) Increase this value if the resizing fails. This could occur with initial packed bed, and small partition blocks (around 5 cells).
12.1.5.2.3. DES_COLL_MODEL¶
Data Type: CHARACTER
Collision model for the soft-sphere approach used in DEM model. All models require specifying the following parameters: DES_EN_INPUT, DES_EN_WALL_INPUT, MEW, and MEW_W.
Name |
Default? |
Description |
|---|---|---|
|
◉ |
The linear spring-dashpot model. Requires: KN, KN_W, KT_FAC, KT_W_FAC, DES_ETAT_FAC, DES_ETAT_W_FAC. |
|
The Hertzian model. Requires: DES_ET_INPUT, DES_ET_WALL_INPUT, E_YOUNG, EW_YOUNG V_POISSON, VW_POISSON. |
12.1.5.2.4. DES_CONV_CORR¶
Data Type: CHARACTER
Specify the Nusselt number correlation used for particle-gas convection.
Name |
Default? |
Description |
|---|---|---|
|
◉ |
Ranz, W.E. and Marshall, W.R. (1952). Chemical Engineering Progress, 48: 141-146 and 173-180 |
|
Gunn D. Transfer of heat or mass to particles in fixed and fluidised beds. International Journal of Heat and Mass Transfer. 1978;21(4):467-476. |
|
|
Wakao N, Kaguei S, Funazkri T. Effect of fluid dispersion coefficients on particle-to-fluid heat transfer coefficients in packed beds: correlation of Nusselt numbers. Chemical engineering science. 1979;34(3):325-336. |
|
|
Tavassoli H, Peters E, Kuipers J. Direct numerical simulation of fluid–particle heat transfer in fixed random arrays of non-spherical particles. Chemical Engineering Science. 2015;129:42-48. |
12.1.5.2.5. DES_EM(PHASE)¶
Data Type: DOUBLE PRECISION
\(1 \le Phase \le DES{\_}MMAX\)
Emissivity of solids phase.
12.1.5.2.6. DES_EN_INPUT(INDEX)¶
Data Type: DOUBLE PRECISION
\(1 \le Index \le MMAX*(MMAX-1)/2\)
Applies to solids model(s): DEM
Normal restitution coefficient for inter-particle collisions used to determine the inter-particle normal damping factor.
Values should be defined for a single dimensional array. For example, a simulation with three solids phases (MMAX=3) needs six values: en11, en12, en13; en22, en23; en33.
12.1.5.2.7. DES_EN_WALL_INPUT(INDEX)¶
Data Type: DOUBLE PRECISION
\(1 \le Index \le MMAX\)
Applies to solids model(s): DEM
Normal restitution coefficient for particle-wall collisions used to determine the particle-wall normal damping factor.
Values should be defined in a single dimensional array. For example, a simulation with three solids phases (MMAX=3) needs three values: enw1, enw2, enw3.
12.1.5.2.8. DES_ET_INPUT(INDEX)¶
Data Type: DOUBLE PRECISION
\(1 \le Index \le MMAX*(MMAX-1)/2\)
Applies to solids model(s): DEM
Tangential restitution coefficient for inter-particle collisions. Values are defined in a one dimensional array. This is required input when using the Hertzian collision model.
12.1.5.2.9. DES_ET_WALL_INPUT(INDEX)¶
Data Type: DOUBLE PRECISION
\(1 \le Index \le MMAX\)
Applies to solids model(s): DEM
Tangential restitution coefficient for particle wall collisions. Values are defined in a one dimensional array. This is required input when using the Hertzian collision model.
12.1.5.2.10. DES_ETAT_FAC¶
Data Type: DOUBLE PRECISION
Applies to solids model(s): DEM
Ratio of the tangential damping factor to the normal damping factor for inter-particle collisions. Required for the linear spring-dashpot collision model.
Name |
Default? |
Description |
|---|---|---|
|
◉ |
For LSD model, if left undefined, MFiX reverts to default value of 0.5. |
12.1.5.2.11. DES_ETAT_W_FAC¶
Data Type: DOUBLE PRECISION
Ratio of the tangential damping factor to the normal damping factor for particle-wall collisions. Required for the linear spring-dashpot model for soft-spring collision modelling under DEM. For the Hertzian model, the tangential damping coefficients have to be explicitly specified and specification of this variable is not required.
Name |
Default? |
Description |
|---|---|---|
|
◉ |
For LSD model, if left undefined, MFiX will revert to default value of 0.5 |
12.1.5.2.12. DES_MIN_COND_DIST¶
Data Type: DOUBLE PRECISION
Minimum separation distance between the surfaces of two contacting particles.
12.1.5.2.13. DES_NEIGHBOR_SEARCH¶
Data Type: INTEGER
Flag to set the neighbor search algorithm.
Name |
Default? |
Description |
|---|---|---|
|
N-Square search algorithm (most expensive) |
|
|
◉ |
Grid-Based Neighbor Search (Recommended) |
12.1.5.2.14. DLB_DT¶
Data Type: DOUBLE PRECISION
Time interval at which Dynamic Load Balance (DLB) is performed (sec)
Name |
Default? |
Description |
|---|---|---|
|
◉ |
: UNDEFINED value turns Dynamic Load Balance off |
|
: Any positive value turns Dynamic Load Balance on |
12.1.5.2.15. DLB_EGW¶
Data Type: DOUBLE PRECISION
Eulerian Grid Weight use in Dynamic Load Balance. A value of zero means the balancing only considers particles. A very large value means the partition will balance only the fluid mesh. Finding the optimal value will require trial and error.
Name |
Default? |
Description |
|---|---|---|
|
12.1.5.2.16. DTSOLID¶
Data Type: DOUBLE PRECISION
Solids time step (s). If this value is left blank (UNDEFINED), then it will be automatically computed based on particle properties, and DTSOLID_FACTOR.
12.1.5.2.17. DTSOLID_FACTOR¶
Data Type: DOUBLE PRECISION
Ratio between the collision time and the solids time step (default value is 50). Minimum value is 10.0. It is highly recommended to keep the default value.
12.1.5.2.18. DTSOLID_UPDATE_DT¶
Data Type: DOUBLE PRECISION
Time interval at which the DEM solids time step is updated. Useful when particle’s size or mass change over time. Only applicable to coupled gas/solids flows (this setting is ignored for pure granular flows).
12.1.5.2.19. E_YOUNG(PHASE)¶
Data Type: DOUBLE PRECISION
\(1 \le Phase \le DES{\_}MMAX\)
Young’s modulus for the particle [Pa in SI]. Required when using the Hertzian collision model.
12.1.5.2.20. E_YOUNG_ACTUAL(PHASE)¶
Data Type: DOUBLE PRECISION
\(1 \le Phase \le DES{\_}MMAX\)
Actual Young’s modulus for the particle [Pa in SI]. Used for computing correction terms for DEM conduction.
12.1.5.2.21. EW_YOUNG¶
Data Type: DOUBLE PRECISION
Young’s modulus for the wall [Pa in SI]. Required when using the Hertzian collision model.
12.1.5.2.22. EW_YOUNG_ACTUAL¶
Data Type: DOUBLE PRECISION
Actual Young’s modulus for the walls [Pa in SI]. Used for computing correction terms for DEM conduction.
12.1.5.2.23. FACTOR_RLM¶
Data Type: DOUBLE PRECISION
Effectively increases the radius of a particle (multiple of the sum of particle radii) during the building of particle neighbor list.
12.1.5.2.24. FLPC¶
Data Type: DOUBLE PRECISION
Fluid lens proportionality constant used to calculate the radius of the fluid lens that surrounds a particle. This parameter is used in the particle-fluid-particle conduction model.
12.1.5.2.25. HAMAKER_CONSTANT¶
Data Type: DOUBLE PRECISION
Hamaker constant used in particle-particle cohesive interactions.
12.1.5.2.26. KN¶
Data Type: DOUBLE PRECISION
Applies to solids model(s): DEM
Normal spring constant [N/m in SI] for inter-particle collisions. Required when using the linear spring-dashpot collision model.
12.1.5.2.27. KN_W¶
Data Type: DOUBLE PRECISION
Applies to solids model(s): DEM
Normal spring constant [N/m in SI] for particle-wall collisions. Required when using the linear spring-dashpot collision model.
12.1.5.2.28. KT_FAC¶
Data Type: DOUBLE PRECISION
Applies to solids model(s): DEM
Ratio of the tangential spring constant to normal spring constant for inter-particle collisions. Use it to specify the tangential spring constant for particle-particle collisions as KT_FAC*KN. Required when using the linear spring-dashpot collision model.
12.1.5.2.29. KT_W_FAC¶
Data Type: DOUBLE PRECISION
Applies to solids model(s): DEM
Ratio of the tangential spring constant to normal spring constant for particle-wall collisions. Use it to specify the tangential spring constant for particle-wall collisions as KT_W_FAC*KN_W. Required when using the linear spring-dashpot collision model.
12.1.5.2.30. MEW¶
Data Type: DOUBLE PRECISION
Applies to solids model(s): DEM
Inter-particle Coulomb friction coefficient.
12.1.5.2.31. MEW_R¶
Data Type: DOUBLE PRECISION
Applies to solids model(s): DEM
Inter-particle rolling friction coefficient. See Zhou et al. (Physica A 269 (1999) 536-553) model A. Please note that here the rolling friction coefficient is non-dimensional and is equivalent to the Zhou paper’s definition divided by the particle diameter.
12.1.5.2.32. MEW_RW¶
Data Type: DOUBLE PRECISION
Particle-wall rolling friction coefficient. See Zhou et al. (Physica A 269 (1999) 536-553) model A. Please note that here the rolling friction coefficient is non-dimensional and is equivalent to the Zhou paper’s definition divided by the particle diameter.
12.1.5.2.33. MEW_W¶
Data Type: DOUBLE PRECISION
Particle-wall Coulomb friction coefficient.
12.1.5.2.34. MINIMIZE_DES_FACET_LIST¶
Data Type: LOGICAL
Flag to turn on/off optimizing the list of facets at each des grid cell.
12.1.5.2.35. NEIGHBOR_SEARCH_N¶
Data Type: INTEGER
Maximum number of steps through a DEM loop before a neighbor search will be performed. The search may be called earlier based on other logic.
12.1.5.2.36. NEIGHBOR_SEARCH_RAD_RATIO¶
Data Type: DOUBLE PRECISION
Ratio of the distance (imaginary sphere radius) to particle radius that is allowed before a neighbor search is performed. This works in conjunction with the logic imposed by NEIGHBOR_SEARCH_N in deciding calls to the neighbor search algorithm.
12.1.5.2.37. NFACTOR¶
Data Type: INTEGER
The number of iterations of a pure granular simulation to let the initial particle configuration settle before a coupled gas-solid calculation is started.
12.1.5.2.38. REMOVE_ROGUE_PARTICLES¶
Data Type: LOGICAL
Flag to remove rogue particles. A rogue particle is defined as a particle that goes out of the MFiX box outside of a defined pressure or mass outlet BC, or is not located in a fluid cell.
12.1.5.2.39. USE_COHESION¶
Data Type: LOGICAL
Flag to enable/disable cohesion model.
12.1.5.2.40. USE_VDH_DEM_MODEL¶
Data Type: LOGICAL
Flag to use van der Hoef et al. (2006) model for adjusting the rotation of the contact plane. See the MFiX-DEM documentation.
12.1.5.2.41. V_POISSON(PHASE)¶
Data Type: DOUBLE PRECISION
\(1 \le Phase \le DES{\_}MMAX\)
Poisson’s ratio for the particle. Required when using the Hertzian collision model.
12.1.5.2.42. V_POISSON_ACTUAL(PHASE)¶
Data Type: DOUBLE PRECISION
\(1 \le Phase \le DES{\_}MMAX\)
Poisson’s ratio for the particle. Used for computing correction terms for DEM conduction.
12.1.5.2.43. VAN_DER_WAALS¶
Data Type: LOGICAL
Flag to turn on the Hamaker van der Waals forces.
12.1.5.2.44. VDW_INNER_CUTOFF¶
Data Type: DOUBLE PRECISION
Minimum separation distance below which van der Waals forces are calculated using a surface adhesion model.
12.1.5.2.45. VDW_OUTER_CUTOFF¶
Data Type: DOUBLE PRECISION
Maximum separation distance above which van der Waals forces are not implemented.
12.1.5.2.46. VW_POISSON¶
Data Type: DOUBLE PRECISION
Poisson ratio for the wall. Required when using the Hertzian collision model.
12.1.5.2.47. VW_POISSON_ACTUAL¶
Data Type: DOUBLE PRECISION
Poisson’s ratio for the wall. Used for computing correction terms for DEM conduction.
12.1.5.2.48. WALL_HAMAKER_CONSTANT¶
Data Type: DOUBLE PRECISION
Hamaker constant used in particle-wall cohesive interactions.
12.1.5.2.49. WALL_VDW_INNER_CUTOFF¶
Data Type: DOUBLE PRECISION
Minimum separation distance below which van der Waals forces are calculated using a surface adhesion model (particle-wall interactions).
12.1.5.2.50. WALL_VDW_OUTER_CUTOFF¶
Data Type: DOUBLE PRECISION
Maximum separation distance above which van der Waals forces are not implemented (particle-wall interactions).
12.1.5.3. Particles in Cell¶
These keywords relate to the Particle in Cell model. Please note that the PIC model is currently not supported by the GUI, because PIC support is undergoing a rewrite in the solver.
12.1.5.3.1. FRIC_EXP_PIC¶
Data Type: DOUBLE PRECISION
Volume fraction exponential scale factor in frictional stress model.
12.1.5.3.2. FRIC_NON_SING_FAC¶
Data Type: DOUBLE PRECISION
Non-singularity term in frictional stress model.
12.1.5.3.3. MPPIC_COEFF_EN1¶
Data Type: DOUBLE PRECISION
An empirical dampening factor for the frictional stress model.
12.1.5.3.4. MPPIC_COEFF_EN_WALL¶
Data Type: DOUBLE PRECISION
Normal coefficient of restitution for parcel-wall collisions.
12.1.5.3.5. MPPIC_COEFF_ET_WALL¶
Data Type: DOUBLE PRECISION
Tangential coefficient of restitution for parcel-wall collisions.
12.1.5.3.6. MPPIC_VELFAC_COEFF¶
Data Type: DOUBLE PRECISION
Solids slip velocity scale factor. This term can be used to scale the bulk solids velocity when calculating parcel/bulk solids slip velocity. Scaling is uniform in all three directions.
12.1.5.3.7. PIC_CD_E¶
Data Type: DOUBLE PRECISION
Restitution coefficient used with collision damping
12.1.5.3.8. PIC_CFL¶
Data Type: DOUBLE PRECISION
Solids CFL (Courant-Friedrichs-Lewy) value. This term can be used to arrest solids time step through a local examination of parcel velocity. Leave undefined (blank) to disable time step control.
12.1.5.3.9. PIC_CFL_CONTROL¶
Data Type: CHARACTER
Control variable for managing solids CFL value. Solids time step will be based on MAX or AVG calculated solids CFL.
12.1.5.3.10. PIC_CFL_PARCEL_FRACTION¶
Data Type: DOUBLE PRECISION
Parcel fraction threshold to invoke a change in solids time step based on solids CFL value.
12.1.5.3.11. PIC_COLLISION_DAMPING¶
Data Type: LOGICAL
Flag to turn on/off PIC collision damping model.
12.1.5.3.12. PSFAC_FRIC_PIC¶
Data Type: DOUBLE PRECISION
Pressure linear scale factor in frictional stress model.