8.3.6. Discrete Element

8.3.6.1. Discrete Element - Common

These are keywords common to Discrete Element Model (DEM) and Particles In Cell (PIC).

8.3.6.1.1. 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 overwritten when using automatic particle generation. A simulation with a mass inflow BC can start without solids by setting PARTICLES = 0.

8.3.6.1.2. 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.

Table 8.43 Valid Values

Name

Default?

Description

.TRUE.

Generate particle configuration based on the initial condition parameters. Data provided in the particle_input.dat file, if present, is ignored.

.FALSE.

◉

Particle position and velocity data are provided in the particle_input.dat file. A runtime error occurs if this file is not provided.

8.3.6.1.3. 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.

Table 8.44 Valid Values

Name

Default?

Description

.FALSE.

◉

Two-way particle-fluid coupling.

.TRUE.

One-way particle-fluid coupling (fluid does not see particle drag).

8.3.6.1.4. DES_INTG_METHOD

Data Type: CHARACTER

Applies to Solids Model(s): DEM

Time stepping scheme.

Table 8.45 Valid Values

Name

Default?

Description

EULER

◉

First-order Euler scheme.

ADAMS_BASHFORTH

Second-order ADAMS_BASHFORTH scheme (DEM only)

8.3.6.1.5. 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.

8.3.6.1.6. 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.

8.3.6.1.7. 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.

8.3.6.1.8. 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.

Table 8.46 Valid Values

Name

Default?

Description

NONE

◉

Do not use interpolation.

GARG_2012

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.

SQUARE_DPVM

Divided Particle Volume Method: Information is interpolated to/from a particles position using a square filter of size DES_INTERP_WIDTH.

LINEAR_HAT

Linear interpolation: Hat functions are used to distribute particle information.

8.3.6.1.9. 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.

8.3.6.1.10. 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.

Table 8.47 Valid Values

Name

Default?

Description

.FALSE.

◉

Use fluid values from the cell containing the particle’s center.

.TRUE.

Interpolate fluid values from the 27-cell neighborhood to a particle’s position.

8.3.6.1.11. 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.

Table 8.48 Valid Values

Name

Default?

Description

.FALSE.

◉

Assign particle data to the fluid grid cell containing the particle’s center.

.TRUE.

Interpolate particle data from the particle’s position to the 27-cell neighborhood surrounding the particle.

8.3.6.1.12. 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.

8.3.6.1.13. 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.

Table 8.49 Valid Values

Name

Default?

Description

.FALSE.

◉

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.

.TRUE.

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.

8.3.6.2. Discrete Element Model

These keywords relate to the Discrete Element Model (DEM).

8.3.6.2.1. 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.

8.3.6.2.2. 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.

8.3.6.2.4. 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.

8.3.6.2.5. 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.

8.3.6.2.6. 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.

8.3.6.2.7. 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.

Table 8.51 Valid Values

Name

Default?

Description

LSD

◉

The linear spring-dashpot model. Requires: KN, KN_W, KT_FAC, KT_W_FAC, DES_ETAT_FAC, DES_ETAT_W_FAC.

HERTZIAN

The Hertzian model. Requires: DES_ET_INPUT, DES_ET_WALL_INPUT, E_YOUNG, EW_YOUNG V_POISSON, VW_POISSON.

8.3.6.2.8. 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.

8.3.6.2.9. 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.

8.3.6.2.10. 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.

8.3.6.2.11. 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.

8.3.6.2.12. MEW

Data Type: DOUBLE PRECISION

Applies to Solids Model(s): DEM

Inter-particle Coulomb friction coefficient.

8.3.6.2.13. MEW_W

Data Type: DOUBLE PRECISION

Particle-wall Coulomb friction coefficient.

8.3.6.2.14. 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.

8.3.6.2.15. 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.

8.3.6.2.16. 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.

8.3.6.2.17. 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.

8.3.6.2.18. 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.

8.3.6.2.19. 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.

8.3.6.2.20. 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.

Table 8.52 Valid Values

Name

Default?

Description

UNDEFINED

◉

For LSD model, if left undefined, MFiX reverts to default value of 0.5.

8.3.6.2.21. 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.

Table 8.53 Valid Values

Name

Default?

Description

UNDEFINED

◉

For LSD model, if left undefined, MFiX will revert to default value of 0.5

8.3.6.2.22. EW_YOUNG

Data Type: DOUBLE PRECISION

Young’s modulus for the wall [Pa in SI]. Required when using the Hertzian collision model.

8.3.6.2.23. VW_POISSON

Data Type: DOUBLE PRECISION

Poisson ratio for the wall. Required when using the Hertzian collision model.

8.3.6.2.24. 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.

8.3.6.2.25. 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.

8.3.6.2.26. USE_COHESION

Data Type: LOGICAL

Flag to enable/disable cohesion model.

8.3.6.2.27. VAN_DER_WAALS

Data Type: LOGICAL

Flag to turn on the Hamaker van der Waals forces.

8.3.6.2.28. HAMAKER_CONSTANT

Data Type: DOUBLE PRECISION

Hamaker constant used in particle-particle cohesive interactions.

8.3.6.2.29. WALL_HAMAKER_CONSTANT

Data Type: DOUBLE PRECISION

Hamaker constant used in particle-wall cohesive interactions.

8.3.6.2.30. VDW_OUTER_CUTOFF

Data Type: DOUBLE PRECISION

Maximum separation distance above which van der Waals forces are not implemented.

8.3.6.2.31. VDW_INNER_CUTOFF

Data Type: DOUBLE PRECISION

Minimum separation distance below which van der Waals forces are calculated using a surface adhesion model.

8.3.6.2.32. WALL_VDW_OUTER_CUTOFF

Data Type: DOUBLE PRECISION

Maximum separation distance above which van der Waals forces are not implemented (particle-wall interactions).

8.3.6.2.33. 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).

8.3.6.2.34. 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.

8.3.6.2.35. DES_CONV_CORR

Data Type: CHARACTER

Specify the Nusselt number correlation used for particle-gas convection.

Table 8.54 Valid Values

Name

Default?

Description

RANZ_1952

◉

Ranz, W.E. and Marshall, W.R. (1952). Chemical Engineering Progress, 48: 141-146 and 173-180

GUNN

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

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

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.

8.3.6.2.36. DES_MIN_COND_DIST

Data Type: DOUBLE PRECISION

Minimum separation distance between the surfaces of two contacting particles.

8.3.6.2.37. 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.

8.3.6.2.38. DES_EM(PHASE)

Data Type: DOUBLE PRECISION

  • \(1 \le Phase \le DES{\_}MMAX\)

Emissivity of solids phase.

8.3.6.2.39. 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.

8.3.6.2.40. 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.

8.3.6.2.41. 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.

8.3.6.2.42. VW_POISSON_ACTUAL

Data Type: DOUBLE PRECISION

Poisson’s ratio for the wall. Used for computing correction terms for DEM conduction.

8.3.6.2.43. MINIMIZE_DES_FACET_LIST

Data Type: LOGICAL

Flag to turn on/off optimizing the list of facets at each des grid cell.

8.3.6.2.44. 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.

8.3.6.2.45. DTSOLID_FACTOR

Data Type: DOUBLE PRECISION

Ratio between the collision time and the DEM time step (default value is 50). Minimum value is 10.0. It is highly recommended to keep the default value.

8.3.6.2.46. 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).

8.3.6.2.47. 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).

8.3.6.2.48. DLB_DT

Data Type: DOUBLE PRECISION

Time interval at which Dynamic Load Balance (DLB) is performed (sec)

Table 8.55 Valid Values

Name

Default?

Description

UNDEFINED

◉

: UNDEFINED value turns Dynamic Load Balance off

>0

: Any positive value turns Dynamic Load Balance on

8.3.6.2.49. 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.

Table 8.56 Valid Values

Name

Default?

Description

>=0

8.3.6.3. 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.

8.3.6.3.1. FRIC_EXP_PIC

Data Type: DOUBLE PRECISION

Volume fraction exponential scale factor in frictional stress model.

8.3.6.3.2. PSFAC_FRIC_PIC

Data Type: DOUBLE PRECISION

Pressure linear scale factor in frictional stress model.

8.3.6.3.3. MPPIC_COEFF_EN1

Data Type: DOUBLE PRECISION

An empirical dampening factor for the frictional stress model.

8.3.6.3.4. FRIC_NON_SING_FAC

Data Type: DOUBLE PRECISION

Non-singularity term in frictional stress model.

8.3.6.3.5. MPPIC_COEFF_EN_WALL

Data Type: DOUBLE PRECISION

Normal coefficient of restitution for parcel-wall collisions.

8.3.6.3.6. MPPIC_COEFF_ET_WALL

Data Type: DOUBLE PRECISION

Tangential coefficient of restitution for parcel-wall collisions.

8.3.6.3.7. 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.

8.3.6.3.8. 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.

8.3.6.3.9. PIC_CFL_PARCEL_FRACTION

Data Type: DOUBLE PRECISION

Parcel fraction threshold to invoke a change in solids time step based on solids CFL value.

8.3.6.3.10. 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.

8.3.6.3.11. PIC_COLLISION_DAMPING

Data Type: LOGICAL

Flag to turn on/off PIC collision damping model.

8.3.6.3.12. PIC_CD_E

Data Type: DOUBLE PRECISION

Restitution coefficient used with collision damping