Dennis Román | Project Portfolio

Aircraft Icing

RG-15 IPW2 Reproduction (2D-Intent Workflow)

Separate RG-15 icing campaign to reproduce IPW2-style trends with a 2D-intent setup, including bin sensitivity checks and a level-based grid refinement sweep.

ANSYS FENSAP-ICEFluent MeshingMATLABPython
RG-15 1-bin freezing fraction and instant ice-growth plots

Provided-grid 1-bin behavior used as a baseline in the RG-15 bin-comparison check.

Project scope

  • This is a separate IPW2-related project focused on RG-15.
  • Goal: reproduce IPW2-style comparison behavior with a workflow configured to be as close as possible to a 2D study.
  • Constraint: student environment did not allow native 2D meshing, so the setup was designed as a 2D-intent approximation.

Case and setup highlights

  • Reference context: public IPW2 RG-15 glaze case (CSV-based curve comparison workflow).
  • Runs included:
    • Provided baseline grid with 1/3/7-bin sweeps.
    • Custom grid family L1-L4 plus provided baseline level for trend checks.
  • Example droplet-bin definitions used:
    • 7-bin: 81.3/5, 51.2/10, 36.8/20, 26.8/30, 18.7/20, 12.5/10, 7.3/5 (um / %)
    • 3-bin: 61.2/15, 27.3/70, 10.8/15
    • 1-bin: 26.7/100

Key numerical snapshots

Provided-grid bin study (same grid, 76,891 cells)

Bin setupBeta_maxMass-caught peak [kg/m^2s]Max ice thickness
1-bin0.8250.010730.00254 m (2.54 mm)
3-bin0.8020.010430.00261 m (2.61 mm)
7-bin0.7960.010340.00561 m (5.61 mm)

Grid-family sweep (L1-L4)

Grid levelElement countBeta_maxMass-caught peak [kg/m^2s]Max ice thickness
L180,0000.64000.008320.00397 m (3.97 mm)
L2158,0000.69320.0090110.00412 m (4.12 mm)
L3237,0000.56360.0073270.00323 m (3.23 mm)
L4380,0000.57900.0075270.00372 m (3.72 mm)

Grid-independence evidence

The chart in the carousel summarizes the measured L1-L4 progression and delta behavior between levels.
Interpretation used in this project:

  • Beta and mass-caught are comparatively stable by finer levels.
  • Thickness is the most sensitive output and should be tracked with tighter near-wall/time-step control.
  • The harness is suitable for continued IPW-style studies because setup parity and trend-check infrastructure are in place.

References used for this project

  • Numerical Results (RG-15) (1) (1).pdf
  • Reproducing IPW2 Results (RG-15) (3).pdf
  • IPW overview: 2nd Workshop

Validation

  • Common units and axis consistency for comparisons

    pass

    Curves compared on aligned wall s/c coordinates with thickness in mm-equivalent reporting.

  • Cross-grid trend review (L1-L4 + provided baseline)

    pass

    Mass-caught and Beta trends stabilize by finer levels while thickness remains most sensitive.

Reproducibility

  • Fixed case setup and bin tables

    Same case BCs were used across levels and bin studies, with explicit 1/3/7-bin definitions.

  • Level-structured runs

    Grid family was executed as L1-L4 plus provided baseline to evaluate convergence behavior.

  • Post-processing harness

    MATLAB/Python post-processing regenerates overlays, deltas, and summary metrics from exports.