Numerical Investigation of Circulation Control Airfoils

1
Numerical Investigation of Circulation Control
Airfoils

• Byung-Young Min, Warren Lee
• Robert Englar, and Lakshmi N. Sankar
• School of Aerospace Engineering
• Georgia Institute of Technology, Atlanta, GA,
  30332-1550

2
Outline

• Background
• Research Objectives
• Configurations studied
• Mathematical and Numerical Formulation
• Results and Correlation with Experiments
• Effects of formal spatial accuracy
• Effects of jet turbulence intensity
• Effects of grid density
• Effects of the inclusion of plenum and nozzle
  geometry in the model
• Effects of turbulence model
• Conclusions and recommendations

3
Background

• Noise pollution from the large aircraft has
  become a major problem that needs to be solved.
  NASA proposed a plan to reduce the noise by a
  factor of four (20dB) by 2025.
• A major source of large aircraft airframe noise
  during take-off and landing is the high-lift
  system - namely flaps, slats, associated with
  flap-edges and gaps.
• The high-lift system also contains many moving
  parts, which add to the weight of the aircraft,
  and are costly to build and maintain.
• These devices for generating high lift are
  necessary for large aircraft that use existing
  runways.

4
Boeing 737 Wing/Flap System (Paper by Robert
Englar)
5

• An alternative to conventional high-lift systems
  is the Circulation Control Wing (CCW) technology.
  
• The CC wing can generate the same high lift with
  much less complexity compared to the high-lift
  system, and many noise sources such as flaps and
  slats, can also be eliminated by the CC wing.
• For example, as shown in previous figure, there
  are just 0-3 moving elements per wing for a
  Circulation Control wing, compared to 15 moving
  parts of a conventional Boeing 737 wing with
  high-lift systems.

6
Circulation Control Wing Concept

• Circulation Control Aerodynamics In this
  approach a tangential jet is blown over a highly
  curved aerodynamic surface (the Coanda surface)
  to increase or modify the aerodynamic forces and
  moment with few or no moving surfaces.
• Figure (Taken from paper by Englar) shows a
  traditional Circulation Control Airfoil with a
  rounded trailing edge.

7
Circulation Control Wing Concept

• In general, the driving parameter of Circulation
  Control is the jet momentum coefficient, Cm,
  which is defined as

• At very low momentum coefficients, the
  tangential blowing will add energy to the slow
  moving flow near the surface. This will delay or
  eliminate the separation, and is called Boundary
  Layer Control.
• When the momentum coefficient is high, the lift
  of the wing will be significantly increased. This
  is called Circulation Control.
• The lift augmentation, which is defined as ?CL /
  ?Cm, can exceed 80.
  

8
Prior Work CCW Airfoil with a Sharp Trailing Edge
9
Prior WorkLift Coefficient vs. Cm
Angle of Attack 0 degrees, Integral Flap at 30
degrees
10
Prior WorkLift Coefficient vs. Angle of Attack
11
Research Objectives

• Extend a previously developed 2-D Navier-Stokes
  based approach for CCW airfoils with sharp
  trailing edge to CCW sections with rounded
  trailing edge.
• Assess the effects of several physical and
  computational parameters on the predictions.
• Grid density
• Formal accuracy of the algorithm
• Turbulence models
• Detailed representation of the plenum and nozzle
  geometry
• Jet turbulence intensity
• Draw conclusions and make recommendations for
  future computational experimental studies.

12
Mathematical and Numerical Formulation

• A 3-D multi-block compressible Navier-Stokes
  solver is used.
• 2-D configurations may be modeled as a special
  case.
• The inviscid flux derivatives are modeled using
  3rd order, 5th order, or 7th order accurate
  weighted essentially non-oscillatory
  interpolations.
• The viscous terms are modeled using standard
  second order central differences.
• The equations are solved by marching in time
  using a temporally first order accurate LU-SGS
  scheme.
• Time-accurate modeling as well as local time
  stepping are available as user-supplied options.
• A variety of turbulence models are available
• Spalart Allmaras (SA) and SA-Detached Eddy
  Simulation (SA-DES) models
• Classical k-w model
• k-w/k-e blended Baseline (k-w BSL, Menter) model
• k-w SST (Menter) model
• This solver was extensively modeled for AGARD
  standard test cases (e.g. RAE 2822 supercritical
  airfoil) prior to its use in the present study.

13
Configuration Being Modeled

• NCCR 1510-7076N airfoil tested at David W. Taylor
  Naval Ship RD Center by J. Abramson in 1977.
• The chord length is 20.34 cm, with the slot
  position at 0.967c.
• A slot height to chord ratio (h/c) of 0.003 was
  selected for current study.
• The freesream dynamic pressure was 957.6 N/m2.
• The freestream static pressure and density were
  assumed to be 101325 pa, and 1.216 kg/m3,
  respectively.
• The corresponding freestream Mach number is
  calculated as 0.116 and the Reynolds number is
  estimated as 5.45105.

14
Numerical Results

• The momentum coefficient was changed over the
  range 0.025 to 0.209
• Systematic Studies were done to assess the
  effects of the following factors on the
  prediction
• grid density
• formal spatial accuracy
• jet turbulence intensity
• inclusion of plenum and nozzle geometry in the
  model
• turbulence models

15
Grid Topology
16
Grid Sensitivity and Spatial Accuracy Studies (Cm
0.209)
For the limited range of grid densities
considered, the solution and the Formal accuracy
of the solution had minimal influence on the
overall loads.
17
Eddy viscosity ( Cm0.209)Higher Order schemes
Resolved the wall jet and the confluent boundary
layers more crisply
18
Effects of Inclusion of the Plenum and Nozzle
Geometry (spatially 3rd order scheme, k-w SST,
Cm0.209)

• Inclusion of the plenum had relatively small
  effect on the flow patterns and overall loads

19
Effects of Jet Turbulence Intensity

• In most experiments, the turbulence intensity
  level of the jet is not measured.
• Our studies indicate this is an important
  parameter and may have a significant effect on
  the computed solutions.

20
Effects of Turbulence Models on Evolution of Lift
with Time
Cm0.025
Cm0.209
21
Effects of Turbulence Models on Evolution of Lift
with Time

Cm0.025
Cm0.209
22
Instantaneous Streamlines at Nominal Steady State
or Limit Cycle, Cm0.025 (top) and 0.209 (bottom)
23
Observations on the Adequacy of Turbulence Models

• Most models (SA-DES, k-w/k-e blended, k-w SST)
  assume that there are two dominant shear layers
  and associated length scales.
• In these models,
• Region close to the wall has eddies of the size
  comparable to the distance from the wall
• Regions away from the wall have length scales
  comparable to shear layer thickness, or grid size
• None of these two layer models properly model CCW
  effects caused by three or more shear layers
  (wall jet, surface boundary layer upstream of the
  slot, mixing layers).
• Among the models tested, the k-w BSL (k-w/k-e
  blended) model performed best.

24
Effects of Turbulence Modeling on Airloads
25
Surface pressure distribution
Cm0.209
Cm0.025
26
Conclusions

• Reynolds-Averaged Nervier-Stokes simulations have
  been done for a circulation control airfoil for
  range of momentum coefficients.
• The effects of grid density, spatial accuracy,
  upstream turbulence level at the jet slot, and
  turbulence modeling have been investigated.
• It was found that turbulence models dramatically
  affected the wall jet behavior and its detachment
  point and hence the overall lift value predicted.
  
• The turbulence level at the jet slot was also
  found to have a noticeable influence on the
  computed solutions.
• For the grids used in this study, use of high
  order spatial accuracy algorithms appeared to
  achieve an enhanced resolution of the wall jet,
  boundary layers, and the mixing layer, but had
  negligible effect on the overall loads.
• The inclusion of the plenum chamber and the jet
  nozzle was found to have negligible effect on the
  overall loads.
• Among the turbulence models tested, the blended
  k-w/k-e model (referred to as k-w BSL) performed
  the best for the entire range of the momentum
  coefficient considered.

27
Recommendations

• These conclusions are based on correlations with
  measured data for the overall lift coefficient
  for the NCCR airfoil and a sharp-trailing edge
  airfoil studied previously.
• For a further assessment of these results, and
  for improved modeling of the CCW flow phenomena,
  it is essential that the turbulent flow behavior
  be characterized through flow visualization and
  hot wire measurements of the turbulent flow field
  downstream of the jet slot.

28
Acknowledgements

• This work was supported by the NASA Langley
  Research Center under the NASA Grant and
  Cooperative Research Agreement (NRA) NNX07AB44A.
• Dr. William E. Milholen is the technical monitor.
• The authors are thankful to Greg Jones of NASA
  Langley Research Center for his interest and
  encouragement throughout this study.