Oil Flow Image of Separation Bubble Courtesy of Bryan D. McGranahan and Prof. Michael S. Selig (UIUC)

Smoke Flow Image of a Separation Bubble Courtesy of Greg Cole and Prof. Mueller (University of Notre Dame)

Oil Flow Primer

In front of the bubble there is an extended separation zone where the laminar flow lifts and diverges from the surface. The airflow on the surface, beneath the separated laminar stream, is stagnant. This is easily seen in the green fluorescent oil-mist image.

At the rear edge of the bubble, the separated flow plunges back to the surface. The mechanism that causes this and the impact with the surface generates turbulence in the attached flow that follows. This is both bad and good. The bad part is that the energy in the turbulence comes out of the aircraft, resulting in drag. But, the good part is that turbulent flow stays attached to curved or bent surfaces much better than laminar flow and that is very important for the proper operation of control surfaces and for holding off the stall condition at low airspeeds.

Depending on the airfoil and flight conditions, separation bubbles may form to any degree, from no bubble at all to fully developed bubbles that are tall enough to add to the form drag of the wing by forcing the free stream flow to detour around them. At the rear of well formed bubbles there is often a rotational flow of trapped air. The direction of rotation is downstream above and upstream on the surface. The rotation is usually in the form of a flattened irregular oval or triangle, rather than a circle.

Partially developed separation bubbles can actually have beneficial effects. Since they do not develop to a significant height, they do not add appreciably to the form drag, yet they cause some of the laminar flow to be lifted above the wing surface where skin friction drag occurs. In other words, the region of greatest shear gradient is lifted off the surface, reducing loses due to skin friction.

Oil flow tests are useful for indicating the existence and structure of separation bubbles. Oil flows are cheap and easy to perform. To the trained eye of an experienced aerodynamicist, oil flows can give additional tell tale clues about surface flows.

At the front of the wing, where the laminar boundary layer is extremely thin, the oil is wiped back toward the separation zone. As the laminar flow lifts from the surface, the oil beneath is dragged along at ever slower speeds, until the flow is entirely above the oil. The aft edge of well formed bubbles are clearly indicated by a sharp, clean line where the reattaching flow scrubs the oil from the surface. Some of it dribbles downstream. Much of it gets trapped in the circulation at the rear of the bubble, however.

Standard Cirrus #60 Oil Flow Images

Following are thumbnails of oil flows on Standard Cirrus #60, taken in 2001 and 2003. Click them for full size images. It should be noted that this ship has a wing twist of .75 degrees. Somewhere around serial number 175 the twist angle was doubled. This should have an effect on the separation bubble.

Lower Left Outer Wing	Stations: 276", 263", 252	Stations: 240", 228"
Stations: 216", 204"	Stations: 192", 180"	Stations: 164", 156"
Station: 150"	Station: 99"	Station: 26"

---

About the images
Unless indicated otherwise, all of these images were taken after cruising at 70 kts.

The wing diagram depicts in brown the position of the separation bubble on the lower surface of the wing. If you compare the diagram of a separation bubble (above) to the oil flows, it can be seen that the region where the oil flow begins thickening in front of the main line of oil is the region of increasing separation leading back to the thickest part at the rear where the flow reattaches. This makes measuring the exact leading edge of the bubble impossible, since it has no clear leading edge. This situation is complicated by the effect of airspeed changes on the oil flow. As the wind tunnel simulations show, the separation and reattachment points on the lower surface move considerably for moderate changes in airspeed around 70 kts. As you slow down to land the separation point moves aft, blurring the leading edge of the oil flow. The green fluorescent oil-mist image (above) shows no air movement under the bubble, whereas my oil flows show some movement--the thicker the oil, the slower the flow. It may be that most of the flow in this region is due to the reduction in airspeed when landing. For these measurements, a mid-point in the leading part of the bubble was used.

Two broken lines illustrate where not to put turbulator tape. The orange broken line broken line is the .67c line where Peter Masak (Performance Enhancement of Modern Sailplanes, 1991) recommended placing turbulators. This is where I first put them on #60. The results are shown in the bottom row of images above. Clearly this not the place for them.

The green broken line is where I mounted turbulators the second time. This position was based on a misunderstanding of the oil flows in the bottom row above in which the thickest collection of oil was taken for the entire separation bubble. I took drag rake measurements before and after removing that turbulator tape. To my surprise, there was no change in the drag data at any airspeed. The data are illustrated in the graph on the right. Evidently, this was because the turbulator was located to the rear of the bubble in stagnant air or where the reattachment was occurring, depending on movement of the bubble with airspeed.

Turbulators should be placed in the forward separation region at a position that depends on the thickness of the turbulator tape. In other words, it is the height of the turbulator as a fraction of the boundary layer that matters. Thicker turbulators should be mounted farther back. See the Turbulator page for more on this.

The solid black vertical line at the 164 inch (417 cm) span station marks the end of the inner (transition) panel and the beginning of the outer (aileron) panel.

The red broken line marks the span station where I have been using a drag rake for preliminary tests on the Sinha Flexible Composite Surface Deturbulator.

Observations
There is a strong separation bubble on the lower wing surface over the transition (inner) panel. The leading edge curves rearward in the middle of this area. This is mostly due to the changing airfoil and wing twist. The changing Reynolds number with chord length also plays a part. The outer panel shows only weak separation activity. It is doubtful that turbulators on the under side of the outer panel will benefit enough to justify the cost in skin friction drag. This is consistent with Steve Willits' report that the same airfoil (Wortmann FX 66-17 A II-182) was tested in a wind tunnel at NASA Langley by Dan Somers with the result that turbulators provided no advantage, and even a little loss.

Finally, at the wing tips I have installed soft bumpers for better protection. This keeps the leading part of the wing tip from scraping on uneven surfaces and prevents the aileron from contacting the ground at full deflection. That, there is a price to pay for these advantages is seen in the flow streams that divert around the bumper. I left the original steel skid in place for insurance. It isn't pretty or efficient, but it's practical.

Position of Separation Bubble
The following tables give the bubble position as the distance from the trailing edge of the wing to the front of the bubble. As noted above, this position is about half way along the front part of the bubble where the separation is gradually increasing. The span position in bold font marks the end of the inner (transition) panel and beginning of the outer (aileron) panel.

          Inches
span         from TE
  0      30.3 = 30 2/8
 12      28.3 = 28 3/8
 36      24.9 = 24 7/8
 72      21.1 = 21 1/8
 99      19.2 = 19 2/8
108      19.1 = 19    
144      18.6 = 18 5/8
164      18.4 = 18 3/8
180      17.3 = 17 2/8
216      14.8 = 14 6/8
252      12.2 = 12 2/8
276      10.5 = 10 4/8

  Centimeters
 span   from TE
  0        77.0		
 50        68.9		
100        61.8		
150        55.8		
200        50.7		
250        48.8		
300        48.2
350        47.6
400        48.0
416.6	   46.8
450        44.4
500        40.9
550        37.4
600        33.9
650        30.3
700        26.8

Upper Left Outer Wing	Upper Left Tip
Stations: 276", 263", 252"	Winglet, 52 kts
Stations: 204", 192", 180"	Stations: 164", 156"
Stations: 15", 32", 62"

---

About the images
The winglet image was taken after cruising at 52 kts and the others were taken after cruising at 70 kts. These images illustrate a well defined separation bubble on the suction side of the wing. The wing diagram depicts in blue the position of the separation bubble. As for the pressure side, the leading edge of the bubble is really a poorly defined region where the separation is gradually increasing.

Observations
The situation on the upper surface stands in sharp contrast to the lower surface. Notice the well defined bubble at all span stations. This suggests that our gliders could benefit from turbulators on the upper wing surface. Notice the dramatic sharpening of the bubble when a winglet is installed. This seems to indicate that winglets cause the flow near the tips to more nearly resemble mid-span flow. Perhaps the strengthened bubble near the tip is a penalty imposed by winglets.

Position of Separation Bubble
The following tables give the bubble position as the distance from the trailing edge of the wing to the front of the bubble. As noted above, this position is about half way along the front part of the bubble where the separation is gradually increasing. The span position in bold font marks the end of the inner (transition) panel and beginning of the outer (aileron) panel.

          Inches
span         from TE
  0      20.3 = 20 3/8
 12      20.2 = 20 2/8
 36      19.9 = 19 7/8
 72      19.5 = 19 4/8
 99      19.2 = 19 2/8
108      19.1 = 19 1/8 
144      18.7 = 18 5/8
164      18.4 = 18 3/8
180      17.3 = 17 2/8
216      14.8 = 14 6/8	
252      12.2 = 12 2/8	
276      10.5 = 10 4/8

  Centimeters
 span   from TE
  0        51.6
 50        51.1
100        50.5
150        50.0
200        49.4
250        48.8
300        48.2
350        47.6
400        48.0
416.6      46.8
450        44.4
500        40.9
550        37.4
600        33.9
650        30.3	
700        26.8

Fuselage oil flows were taken just to see if anything bad would show up. The most obvious feature is the stream of thick oil following the flow lines down and back from the wing. On close examination, this stream is fed from above where a thick stream of oil is seen following along the gap seal tape. It goes around the trailing edge of the wing, then runs inward to the fuselage and forms the main stream on the side of the fuselage. This path seems to work as a channel for air from the turtle deck area above. These oil flows show in a remarkable way how lift comes from redirecting the momentum of the air stream.

Careful examination of the 2nd and 3rd images shows a region above the wing fillet where there seems to be a slight flow separation from the surface.