23.3.17. Specifying Injection-Specific Physical Models

Drag and breakup models can be specified on a per-injection basis, allowing you to specify the most appropriate models for each injection in your model. These per-injection models are specified on the Physical Models tab of the Set Injection Properties dialog box.

23.3.17.1. Drag Laws

The spherical, nonspherical, Stokes-Cunningham, and high-Mach-number laws described in Particle Force Balance in the Theory Guide are always available, and the dynamic-drag law described in Dynamic Drag Model Theory in the Theory Guide is available only when one of the droplet breakup models is used in conjunction with unsteady tracking. See Breakup for information about enabling the droplet breakup models. The Grace and Ishii-Zuber drag laws described in Bubbly Flow Drag Laws in the Fluent Theory Guide allow to model the drag coefficient in gas-liquid flows in which the bubbles can have a range of different shapes. Both drag laws are only available if the magnitude and direction of the gravity vector are specified in the Operating Conditions dialog box. The remaining three, Wen-Yu, Gidaspow, and Syamlal-OBrien are available only when the dense discrete phase model is enabled (Including the Dense Discrete Phase Model) and the flow regime consists of a dense gas-solid. However, with these models, you cannot verify whether it really is a dense flow or a gas-solid flow. It is up to you to decide. In any case, these drag formulations are suitable for dense gas-solid flows.

If the spherical, high-Mach-number, dynamic-drag, Wen-Yu, Gidaspow, or Syamlal-OBrien drag law is selected, no further inputs are required. If the nonspherical law is selected, the particle Shape Factor ( in Equation 16–75 in the Theory Guide) must be specified. The shape factor value cannot exceed 1. For the Stokes-Cunningham law, the Cunningham Correction factor ( in Equation 16–77 in the Theory Guide) must be specified.

23.3.17.2. Particle Rotation

To set-up particle rotation, follow the steps below:

  1. In the Set Injection Properties dialog box, under the Physical Models tab, enable Particle Rotation.

    The options related to the particle rotation feature appear in the Set Injection Properties dialog box.

  2. From the Rotational Drag Law drop-down list, select the model for calculating the rotational drag coefficient . The following options are available:

    Dennis-et-al

    Uses Equation 16–4 in the Fluent Theory Guide to calculate the rotational drag coefficient .

    none (default)

    Implies zero torque, that is, it excludes the influence of the fluid on the rotational particle motion.

  3. From the Magnus Lift Law drop-down list, select the model for calculating the rotational lift coefficient . The following models are available:

    Oesterle-Bui-Dinh

    Uses Oesterle and Bui Dinh’s formulation (Equation 16–18 in the Fluent Theory Guide).

    Tsuji-et-al

    Is based on the particle spin rate (Equation 16–19 in the Fluent Theory Guide).

    Rubinow-Keller

    Uses a linear dependence of the rotational lift coefficient on the spin rate (Equation 16–21 in the Fluent Theory Guide).

    none (default)

    Implies that the Magnus lift will not be included in your simulation.

  4. Under the Point Properties tab, specify angular velocity components for single, group, and surface injections; and angular velocity magnitudes for cone injections.

  5. For each wall with a reflect boundary condition for the discrete phase, specify the friction coefficient ( in Equation 16–213 and Equation 16–218 in the Fluent Theory Guide) under the DPM tab of the Wall dialog box. For details, see Friction Coefficient.

The particle rotation model options are also accessible through the text user interface, under define/models/dpm/injections/injection-properties/set/physical-models/particle-rotation. Note that the physical-models command becomes available only after you have selected the injection(s) using the prompts of the define/models/dpm/injections/properties/set/pick-injections-to-set text command.

23.3.17.3. Rough Wall Model

To use the rough wall model for walls with the reflect boundary condition type:

  1. In the Injection Settings dialog box, under the Physical Models tab, select Rough Wall Model.

  2. In the Wall boundary conditions dialog box for the appropriate wall, under the DPM tab, select reflect from the Boundary Cond. Type drop-down list.

  3. Specify the following wall roughness parameters:

    Ra

    is the wall mean roughness ( in Equation 16–220 in the Fluent Theory Guide)

    Rq

    is the standard deviation of the roughness structure ( in Equation 16–220 in the Fluent Theory Guide)

    RSm

    is the mean roughness slope (the average distance between peaks of wall material) ( in Equation 16–220 in the Fluent Theory Guide)


    Note:  Note that the DPM Wall Roughness Parameters will only appear in the Wall dialog box with the reflect DPM boundary condition type if the Rough Wall Model is selected in the Injection Settings dialog box.


For information about the rough wall model, see Rough Wall Model in the Fluent Theory Guide.

23.3.17.4. Brownian Motion Effects

For sub-micron particles in laminar flow, you may want to include the effects of Brownian motion (described in Brownian Force in the Theory Guide) on the particle trajectories. To do so, enable the Brownian Motion option under the Physical Models tab. In order to include Brownian motion effects, you must also select the Stokes-Cunningham drag law in the Drag Law drop-down list under Drag Parameters, and specify the Cunningham Correction ( in Equation 16–77 in the Theory Guide).

23.3.17.5. Breakup

To enable the modeling of particle breakup for the injection, select Enable Breakup and choose the Breakup Model (TAB, Wave, KHRT, or SSD) from the drop-down list. A detailed description of these models can be found in Secondary Breakup Model Theory in the Theory Guide.

  • For the TAB model, you must specify the following values:

    • y0 is the initial distortion at time equal to zero ( in Equation 16–371 in the Theory Guide). The default value () is recommended.

    • The number of Breakup Parcels (under the Breakup Model drop-down list), to split the droplet into several child parcels, as described in Velocity of Child Droplets in the Theory Guide. The diameter of the child parcels is sampled from a Rosin-Rammler distribution. This can be switched off in the TUI with the command:

      /define/models/dpm/spray-model/tab-randomize-breakup-parcel-diameter?

  • For the Wave model, you must specify the following values:


    Note:  You will generally not need to modify the value of B0, as the default value 0.61 is acceptable for nearly all cases. A value of 1.73 is recommended for B1.


    The Wave model implementation has been formulated to deal with the initial breakup of a cylindrical liquid jet. In the Rayleigh regime (that is, at low gas Weber numbers), a cylindrical jet breaks up into droplets whose diameters are larger than that of the jet itself. In this regime, the model assumes that the diameter of the continuous, cylindrical liquid jet has been entered as initial diameter in the injection. Therefore, the droplet diameter can grow beyond the initial diameter. For details, see Droplet Breakup in the Fluent Theory Guide.

    If you want to suppress this feature of the Wave model, use the TUI command:

    /define/models/dpm/spray-model/wave-allow-rayleigh-growth? no

    Note that this takes effect unconditionally for all injections using the Wave or KHRT breakup models.

  • For the KHRT model, you must specify the following values:

    The constants B0 and B1 are the same as for the Wave model.

    As for the Wave model described above, the KHRT model is formulated to deal with the initial breakup of a cylindrical liquid jet. The behavior in the Rayleigh regime is as described above for the Wave model and the same TUI command can be used to modify the behavior. In addition, the KHRT model uses the Liquid Core Approximation as described in Liquid Core Length in the Fluent Theory Guide. To suppress the use of the Liquid Core Approximation you can set the Levich constant, , to zero.

  • For the SSD model, you must specify the following values:

For steady-state simulations, you also must specify an appropriate Particle Time Step Size and the Number of Time Steps, which will control the spray density. See Options for Interaction with the Continuous Phase for more information.

Note that you may want to use the dynamic drag law when you use one of the breakup models. See Drag Laws for information about choosing the drag law.


Release 18.2 - © ANSYS, Inc. All rights reserved.