An Experimental Investigation of Turbulent Boundary Layer Flow over Surface-Mounted Circular Cavities

1
An Experimental Investigation of Turbulent
Boundary Layer Flow over Surface-Mounted Circular
Cavities

• Jesse Dybenko
• Eric Savory
• Department of Mechanical and Materials
  Engineering
• University of Western Ontario, London, ON
• May 24, 2006

2
Flow Geometry
3
Motivation

• Cavities are found on aircraft and automobiles
• Landing gear wheel wells
• Recessed windows
• Sun roofs
• Symmetric geometry, asymmetric mean flow
• Not well researched
• A better understanding of these flows could lead
  to drag and noise reduction for airframes

4
Background

• Peak in cavity drag at h/D 0.5

5
Background

• Cavity Feedback Resonance (Rossiter, 1964)

6
Background

• Vortices shed from upstream cavity lip

7
Background

• Vortices convected downstream

8
Background

• Vortex impinges on downstream lip

9
Background

• Acoustic pulse radiates upstream

10
Background

• Pulse disturbs shear layer causes vortex to be
  shed. Feedback loop is closed.

11
Background

• Frequency associated with this mechanism can be
  estimated using Rossiters Formula (Rossiter,
  1964)
• f is predicted oscillation frequency, m is
  integer mode number, is vortex-sound pulse
  lag-time factor, M is free stream Mach number,
  is ratio of vortex convection velocity to free
  stream velocity

12
Background

• Oscillation can also occur according to the depth
  scale of the cavity depth-mode resonance
• Can also estimate frequency due to this
  mechanism
• f is predicted oscillation frequency, N is
  odd-integer mode number, c is speed of sound in
  air, h is cavity depth

13
Major Objectives

• To understand the causes of abnormal flow in
  cavity and in its wake for h/D 0.5
• What causes this flow to differ from flow at
  other depth configurations?
• To investigate the fluctuating nature of the
  flows at various cavity depths and their
  relationship with resulting cavity drag

14
Experimental Setup
15
Experimental Techniques

• Three cavity depth ratios were used for
  measurements
• h/D 0.20, 0.47 and 0.70
• Cavity depth was only variable
• Three systems were used for measurements
• Pressure transducers
• Surface pressure distribution
• Microphones
• Acoustic response of cavity
• Two-component hot-wire anemometry
• Mean Velocity and Turbulence Profiles in wake

16
Experimental Variables

• U0 27.0 m/s, d 55 mm (d/D 0.72), ReD
  1.3 x 105

17
Results and Discussion

• Mean pressure distributions on sidewall

18
Results and Discussion

• Mean surface pressure distributions on cavity
  base

19
Results and Discussion

• Vortex Skeleton Diagrams h/D 0.2

20
Results and Discussion

• Vortex Skeleton Diagrams h/D 0.47

21
Results and Discussion

• Vortex Skeleton Diagrams h/D 0.70

22
Results and Discussion

• RMS pressure distributions on sidewall

23
Results and Discussion

• RMS pressure distributions on cavity base

24
Results and Discussion

• Drag coefficient comparison

25
Results and Discussion

• Wake velocity profiles Stream-wise velocity

26
Results and Discussion

• Wake velocity profiles Stream-wise turbulence

27
Results and Discussion

• Frequency analysis
• Estimate Frequencies
• Cavity Feedback Resonance
• Predicted first-mode f 145.5 Hz

28
Results and Discussion

• Frequency analysis
• Estimate Frequencies
• Depth Mode Resonance
• Predicted first-mode frequencies are
  depth-dependent, for three depths tested
• h/D 0.20 ? f 2329 Hz
• h/D 0.47 ? f 1512 Hz
• h/D 0.70 ? f 1164 Hz

29
Results and Discussion

• Frequency analysis Microphone in base

30
Results and Discussion

• Frequency analysis Microphone in base

31
Conclusions

• Pressure Measurements
• RMS pressure patterns show maxima at shear layer
  reattachment points and vortex centres
• Mean pressure patterns agree well with those done
  by previous investigators
• Integrated drag coefficients also match well
  with previous data

32
Conclusions

• Wake Flow Analysis
• Symmetric velocity and turbulence profiles for
  h/D 0.20
• Asymmetric for h/D 0.47, showing clear,
  circular trailing vortex feature, which can be
  disturbed to switch sides
• In this feature, mean streamwise velocity is at
  a minimum, turbulence at a maximum

33
Conclusions

• Frequency analysis
• Possible link between cavity feedback resonance
  and abnormal flow behaviour at h/D 0.47
• Depth mode oscillations occur for h/D 0.47 and
  0.70 0.20 not deep enough

34
Recommendations

• Aerodynamic Design
• If circular cavity required on vehicle frame,
  shallow holes are best (h/D 0.20 or less)
• low drag, low noise in high frequency band, no
  resonances

35
Acknowledgements

• Technicians at BLWTL
• Prof. Gregory Kopp
• University Machine Shop
• Advanced Fluid Mechanics Research Group
• Tom Hering and Rita Patel

36

• Questions?
• For additional information on research done by
  the AFM Research Group, try our website
• http//www.eng.uwo.ca/research/afm