DPM Injection Types in ANSYS Fluent 2024 R1

#+CATEGORY: cfd

#+TAGS: DPM Fluent injection types size distributions

#+MATURITY: growing

DPM Injection Types :: Complete Reference for ANSYS Fluent 2024 R1

This page covers every injection type Fluent ships, the size-distribution models available, and how to choose the right injection for your problem. Target application: shisha PM2.5 simulation.

Injection Type Decision Matrix

Surface Injection (Most Common for Shisha)

Surface injection seeds particles from every face of a selected boundary zone. This is the correct choice for your shisha simulation: the smoke source is a surface (the bowl outlet or the shisha chamber inlet), not a point.

Setup procedure: 1. Injections... > Create > Surface 2. Select the surface zone (e.g., bowl_outlet) 3. Set Point Type or Number of Parcels 4. Define material, diameter distribution, velocity, temperature

Face-point methods:

• Face-normal velocity: particles inherit the face-normal velocity component
• Face-centered velocity: particles inherit the velocity vector at face centroid
• From file: read per-face injection data from a sample file
• Number of particles per face: how many parcels seed from each face (default = 1; increase to 5--50 for statistical convergence)

Shisha configuration:

• Surface zone: bowl_chamber_outlet or the smoke_inlet face zone
• Number of parcels per face: 10--50 (PM2.5 needs good statistics)
• Velocity: match the measured draw rate at the inlet
• Total parcels: aim for 10,000--50,000 for a converged concentration field

Single-Point Injection

Inject from a single Cartesian coordinate. Useful for:

• Point-source contamination (a leak)
• Idealized tracer injection
• Debugging particle trajectories

Parameters:

• Position: (x, y, z) in meters
• Direction: unit vector (u, v, w)
• Diameter: single value or distribution
• Number of parcels: how many to release

Conical Injection (Spray Nozzles)

Three sub-types under the conical injector:

Solid Cone

Particles fill the full cone interior. Parameter: Cone half-angle (in degrees). Used for fuel injectors where droplets fill the entire spray envelope.

Hollow Cone

Particles lie on the surface of the cone only. Good for pressure-swirl atomizers where the liquid sheet forms a thin annulus.

Film Cone

Particles start as a liquid film on the cone surface. Requires a wall-film model activation. Used for pre-filming atomizers.

Annular injection: Define inner radius, outer radius, and a radial profile (uniform or custom). Used for air-blast atomizers where the liquid sheet has finite thickness.

Flat-Fan Injection

Sheet-like spray with a flat profile. Parameters:

• Fan angle: spread of the sheet
• Sheet thickness: initial sheet width

Used in paint-spray and agricultural-spraying simulations. Not applicable to the shisha case.

File-Based Injection

Read particle properties from an external .dpm or .inp file. Useful when:

• Experimental PIV data provides initial positions and velocities
• A previous DPM calculation restarts from a saved state
• You want to re-inject particles from a sample-plane output

File format columns (ASCII):

• id: integer particle index
• x y z: position (m)
• u v w: velocity (m/s)
• d: diameter (m)
• T: temperature (K)
• m_flow: mass flow rate (kg/s) of the parcel
• n_components: number of species (1 for inert)
• y_i...: mass fractions for multicomponent

See also: DPM Sample Files for generating compatible files.

Ring Injection

Inject from a ring of specified radius. The ring lies in the plane perpendicular to the injection direction. Parameters:

• Center: (x, y, z)
• Radius: ring radius
• Direction: axis of the ring
• Radial spread: how thick the ring is

Group Injection

Bundle several injections under a single named group. Useful for:

• Multi-nozzle problems (6 identical injectors)
• Applying the same properties to many injections
• Simplifying reports and post-selection

Size Distribution Models

Every injection supports multiple diameter-distribution options. The distribution determines how parcels are split across sizes.

Log-Normal distribution (shisha PM2.5)

The log-normal distribution is the correct choice for PM2.5 smoke aerosols. It arises naturally from condensation and coagulation processes and is the standard model in aerosol science.

Probability density: : f(d) = 1/(d * sigma * sqrt(2*pi)) * exp( -(ln(d) - ln(d_g))^2 / (2*sigma^2) )

Parameters:

• d_g (geometric mean diameter): 0.5--1.5 \mu m for shisha smoke
• sigma (geometric standard deviation): 1.3--2.0 (typical for aged aerosols)

In Fluent:

• Set Distribution type = log-normal
• Mean Particle Diameter = d_g
• Spread Parameter = sigma

Rosin-Rammler distribution

Used for sprays and milled solids: : Q(d) = 1 - exp( -(d / d_bar)^n )

• d_bar: mean diameter
• n: spread parameter (larger = narrower distribution)

Not recommended for PM2.5 aerosols -- this is a spray distribution, not an aerosol distribution.

Parcel Number and Statistical Convergence

The Number of Parcels setting controls how many computational particles the injection emits per pass. This is the single most important setting for concentration accuracy.

Convergence criteria:

• Run the DPM twice with N and 2N parcels
• Compare the concentration field: if the difference is < 5%, N is sufficient
• For shisha PM2.5: start with 10,000 parcels; increase to 50,000 for publication-quality results

Stochastic tracking interaction: Each physical particle path is tracked once per try (stochastic DRW). The Maximum Number of Tries setting multiplies trajectories: total tracked = parcels x tries. For PM2.5 you want both parcels and tries high enough to sample the turbulent-dispersion envelope.

Typical values:

TUI Commands for Injection Management

Shisha PM2.5 Injection Setup Summary

--- :PROPERTIES: :seeAlso: cfd_dpm-setup-activation, cfd_dpm-boundary-conditions, cfd_dpm-sample-files, cfd_dpm-shisha-pm25, cfd_dpm-theory :END: