4th AIAA CFD drag Prediction workshop

Venue : 27th Applied Aerodynamic Conference, San Antonio, TX
Date: 20-21 June, 2009

Objectives

  1. Provide an impartial forum to evaluate RANS solvers.
  2. Identify areas in CFD needing research and development.

Focus

Conduct blind drag prediction accuracy to test state of the art CFD methods and follow up with wind tunnel tests after workshop.

Configuration

The NASA Common Research Model (CRM) transonic wing-body-tail model.

Test Cases :

Case 1a. Grid convergence study
  1. Grid convergence study at Mach 0.85, CL = 0.500 (±0.001).
  2. Tail incidence angle = 0 degrees
  3. Chord Reynolds number = 5 million based on reference chord of 275.80 inches
  4. Coarse, Medium and Fine grids to be used for the study
Case 1b. Downwash study
  1. Downwash study at Mach 0.85, CL = 0.500 (+-0.001).
  2. Medium grid for computations.
  3. Drag polars for alpha iH = 0.0°, 1.0°, 1.5°, 2.0°, 2.5°, 3.0°, 4.0° degrees.
  4. Tail Incidence angles iH = -2°, 0°, +2° degrees and Tail off.
  5. Trimmed Drag polar of tail off v/s tail on
  6. Chord Reynolds number = 5 million based on reference chord of 275.80 inches
Case 2 (optional). Mach Sweep study
  1. Drag polars at Mach 0.70, 0.75, 0.80, 0.83, 0.85, 0.86, 0.87.
  2. Drag rise curves ar CL = 0.400, 0.450, 0.500 (±0.001 or extracted from polars).
  3. Untrimmed, Tail Incidence angle, iH = 0°
  4. Medium grid for computations
  5. Chord Reynolds Number = 5 million based on reference chord of 275.80 inches
Case 3 (optional). Reynolds number study
  1. Reynolds Number ar Mach =  0.85, CL = 0.500 (±0.001).
  2. Tail Incidence angle, iH = 0°
  3. Medium grid for computations.
  4. Compare Reynolds number = 5 million results with Reynolds number 20 million results based on reference chord of 275.80 inches.

 

Inferences

Important concluding remarks by organizing committee:

  1. More scatter from unstructured methods than from structured grid methods. Suspect is more on the grid than the code.
  2. The number of unstructured codes that provided results within the span and accuracy of the structured grids have increased. Less outliners.
  3. Still a fair amount of scatter in separation bubble, but the bubble itself is smaller on this configuration.
  4. Reynolds number increment predictions were very consistent. Correlated better with turbulance model than with grid type.
  5. As with previous DPW's, there are no clear trends for turbulence model. Clearly, SA and SST models dominate.

Computations using HiFUN

Cut section Tail - Fuselage junction Wing - Fuselage junction
Fine Mesh

Grid Type : Unstructured Hybrid grid.

Elements : Prisms and tetrahedrons

Grid Coarse Medium Fine
Field Cells 6244147 21288317 58076968
Boundary Cells 171374 407710 748150

Results

Grid Convergence Study :

Pressure distribution Grid Bubble
Tail 0 configuration results

 

Drag coefficient Moment coefficient
Blue line indicates the HiFUN results

Reynolds Number Study :

Re = 5.0 million Re = 20.0 million

Note :

  1. Smaller separation bubble at Wing Body junction for Re = 20.0 million.
  2. No separation near wing trailing edge for Re = 20.0 million.

References

  1. AIAA 4th Drag Prediction workshop
  2. Ravindra K, Nikhil Vijay Shende, N Balakrishnan, ``CFD Computations for commen Research model using the code HiFUN –09", Proceedings of fourth AIAA drag prediction workshop, SAN Antonio, TX
.