What's new
Van's Air Force

Don't miss anything! Register now for full access to the definitive RV support community.

Air flow visualization liquids and interpretation.

BillL

Well Known Member
I hope this is the correct forum.

There have been comments by Dave Anders, SCSmith, and written briefly in the RVAtor. They mention using a fluid on the gear fairings and wheel pants to ensure they are truly aligned with the airflow. Dave Anders mentioned several times of mixing tempera paint powder in some engine oil and (somehow) applying to the gear fairing and getting them aligned. The same for wheel pants.

So, how exactly are these applied (eye dropper?), in a line or literally drops? And what colors work and don't stain the paint? Should they be UV, fluorescent? I have bookmarked 4-5 sites where these can be purchased.

Then, after the flight test and landed again, how are the results interpreted? What am I looking for, even streams all the way to the TE?

Can some of you experienced guys sound off on this, Please.
 
There are a variety of techniques and materials, and a whole treatise (or two) could be written about it.

If you have a white or light-colored surface, just plain used motor oil is best. It has extremely fine particulates and flows smoothly. Adding tempera paint to clean oil will let you change color, so for example if you have a dark surface, use white tempera paint. But it is very hard to mix and filter this paint-oil mix well enough to not have it be somewhat gritty. For gross flow visualization, that may be just fine, depending on your purposes. But the grit will trip a laminar boundary layer early, so it is not useful for visualizing laminar separation, laminar bubbles, or transition.

There are several application techniques too. For streamline visualization, just paint it on over the whole surface. One way to produce really dramatic visualizations of separated flow is to put a stripe of paint-oil along the forward portion of a surface, or a row of dots, and then fly. The paint-oil will flow back, but can not get inside the separation zones, and so will leave those areas completely uncoated. With highly 3-D separation (such as a vortex sheet formation) you may see the oil run around the separation point and fill in behind it, perhaps showing surface vorticity, or creating a number of interesting patterns.

If your goal is to just be sure a surface is aligned to the flow (zero angle of attack) then coat both sides and then check that the attachment line is right on the leading edge. From the leading edge back, it is unlikely that the 'upper' and 'lower' surface will look much different for something like a strut fairing or wheel pant, even if it is at some angle of attack, because there is little if any flow separation, but you will see the attachment line move off the leading edge.

Now the biggest challenge for flight test. You put the stuff on on the ground, and then taxi, take-off, climb, cruise, descend, and land. What part of that flight gets imprinted in the oil flow patterns? It depends. If you put enough on to keep it wet during the whole flight, then it will be biased toward the end of the flight in terms of what has the most influence. Flow character from early in the flight will get 'erased' by additional flow later in the flight.
For some things, you may not care. Certainly landing gear strut alignment to the flow is somewhat sensitive to angle of attack (speed).
Another challenge is gravitational effects on the oil before, during, and after the flight.
A couple of choices: If you put a thin paint-oil mix on, and cruise until it actually dries, then the patterns of flow will be more or less preserved at that flight condition and not corrupted by the landing phase of flight. If you can mount a camera that can see the area of interest, you may be able to record the flow behavior at the desired flight condition, or maybe even more than one condition.

Interpreting results is challenging and interesting, and especially for highly 3-D flows, you can not assume that the flow on the surface is the same as the flow a short distance up off the surface. Examples are spanwise flow on swept wings, especially on the underside in areas of fairly slow streamwise flow component, and areas like the belly just behind the cooling flow exit, where you can very likely see vorticity/recirculation on the surface, but a short distance up off the surface the flow could be perfectly-well organized.

Getting the oil viscosity what you want can also be tricky, and depends on what you want to visualize. If you want to make a pigmented coating that will 'dry' at some point in the flight, then thinner oils are better. If you want to keep everything wet and flowing, heavier oils are better. You will likely find that in areas of high surface shear stress, like the forward portion of an airfoil, the coating may be wiped fairly clean. That is a useful result in itself, it tells you the flow is attached and high velocity near the surface. All that oil has to go somewhere and it will thicken in areas of lower shear stress or lower velocity. Interpreting sudden changes in residual oil can be tricky - it could be transition, it could be a laminar bubble, it could be separation. It takes experience with these to recognize tell-tail signs of each.

I will follow with a few stories, and answer what questions I can.

Oh and full disclaimer - I take no resposibility for whether oil or paint stains or discolors your airplane. I can not promise that a good coat of wax will prevent it, although it seems like it would help, but it may prevent the oil from wetting to the surface, and give different/spurious results.
 
Thanks, Steve that is a lot of typing!! I downloaded some NASA papers on visualizing the flow, and got some idea of the challenge of viscosity, evaporation and gravity effects. I have some large paint samples I can test for staining. The pants are dark/light so maybe alternating colors. I read to mix pigments to small amounts of oil first in a thick paste to prevent clumping, we shall see how that works.

Well off to this new educational venture. I'll post some photos for interpretation.
 
Continued...

One other tidbit I forgot to mention about oil flow visualization is use of UV pigments. It is not very useful for flight test, unless you want to test at night with no moon, and have a way to bombard the area with UV light. But Aeroshell 100 has a UV pigment that is useful for finding oil leaks on engines, all you need is a hand-held UV light and you can see the green flourescence of the oil. We sometimes use this in wind tunnel flow visualization. It doesn't matter what color the surface is.

Another flow visualization technique that can be useful in flight test, especially if you have very dark colored surfaces is sublimation. Polycyclic aromatic hydrocarbons (PAHs) sublimate (evaporate directly from solid to gas) and the rate of sublimation depends on local pressure, temperature, surface shear stress. You can apply a sublimating coating by dissolving the material in a carrier and spraying it onto the surface. The carrier solvent flashes off leaving a fairly smooth coating. For atmospheric pressure and our typical flight speeds, napthalene works pretty well. For higher flight speeds, you want PAHs that sublimate slower. In supersonic wind tunnels, we use fluorene.

The primary use for sublimation is to identify laminar transition. There is an abrupt increase in surface shear stress where the boundary layer transitions from laminar to turbulent. So if you have a laminar flow airfoil that is coated with napthalene, and you watch the material progressively sublimate off the surface, you will see that the material disappears much more quickly under a turbulent boundary layer, and more slowly under a laminar boundary layer, because of the difference in surface shear stress. As you watch the material sublimation progress, it will later start to disappear in the forward portions of the laminar area as well, because the shear stress is higher there than farther aft. Finally, you may end up with just a stripe of remaining material just ahead of transition, illustrating a laminar separation bubble, where the shear stress is very low. Finally, all the coating will sublimate away. This whole process may take 3--5 minutes. So typically you would image the surfaces of interest in flight, but it is possible to do a very short flight and land and then image the results. Again that complicates the results because of the non-constant flight conditions that the boundary layers have been subjected to. I have used sublimation on sailplane winglets; we flew a constant speed on tow and I imaged the winglet from the cockpit.
 
Just for general interest, I thought I would share one story about a rather unique flow visualization experiment that was done at the (then) NACA Ames Aeronautical Laboratory. (Now NASA Ames Research Center).

Shortly after WW II, the NACA was approached to try to understand why the P-51 Mustang successfully used the NACA 6-series laminar flow airfoils while the Bell P-63 King Cobra, using similar airfoils, did not achieve the significant performance benefit expected. The question fell to a young engineer named John Spreiter. John wanted to do sublimation testing on both airplanes to determine the extent of laminar flow achieved. But how to do that???

Just a short distance from the lab there was a cement plant that, before the days of scrubbers and other polution-controls, would emit an alkaline plume that would extend quite high, especially on cool summer mornings when the marine layer would intrude into the bay area. John found a chemical ( I don't know what it was) that would sublimate rapidly when subjected to the alkaline vapor in the plume. He applied the material to the wing of the airplane, and the test pilot would take off, turn downwind in the pattern, fly through the plume at the cement plant, and land. The team would then rush out and photograph the wings. Areas of laminar flow have much less mixing and less shear stress, thus much less exposure of the surface to the alkaline vapor, whereas areas of turbulent boundary layers have much higher mixing, and so the sublimation coating would preferentially 'clear' in those areas.

His finding was that indeed, the P-63 achieved far less laminar flow, despite using airfoils from the same family. John found that the P-51 was built with heavier wing skins that tended to form smoother surfaces, whereas the P-63 was built with lighter skins that were wavier, especially where riveted to ribs. The P-63 also used thicker airfoils, which make sustaining laminar flow more challenging and more susceptible to surface waviness causing early transition.

John Spreiter had a successful and rewarding career, doing pioneering work in transonic and supersonic aerodynamics, interaction of solar wind with planetary ionospheres, and was a wonderful teacher and storyteller. I was blessed to take several classes from him late in his teaching career, and forever grateful that he agreed to serve on my doctoral committee despite having retired a few years before.
 
Would that be the same cement plant that, for years, was a vfr waypoint for traffic going into San Carlos airport?
 
Great info Steve

I really enjoy reading this info, Steve. More reminders of how much there is to learn about this amazing hobby!
 
Would that be the same cement plant that, for years, was a vfr waypoint for traffic going into San Carlos airport?

HI Bob,
very likely a different one. There used to be several batch plants on the peninsula. The biggest of course was always the Permanente plant in the Los Altos-Cupertino hills. But this one was at the corner of Moffett Blvd and US-101 in Mt View. I think the one you are referring to was near Harbor Blvd and US-101 in Redwood City? Its been a long time!
 
Report the Cement Plant

Would that be the same cement plant that, for years, was a vfr waypoint for traffic going into San Carlos airport?

I thought exactly the same thing when I saw this. But I think it is more likely to be the Permanente Ridge cement plant that is directly South West of Moffett. https://en.wikipedia.org/wiki/Permanente_Quarry

I'm based out of KSQL, so "Report the Cement Plant" is still something I have stuck in my head. I believe you can still get a shirt from the San Carlos Flight Center that says exactly that.
 
Oil Flow Testing for Sailplane Wings

I found some examples of flow visualization testing on the Standard Cirrus sailplane. These were done using old motor oil. I also remember Dick Johnson used this testing method in many of his sailplane reviews.

http://www.standardcirrus.org/OilFlows.php

An example of flow separation
70ktR99.JPG
 
Steve,

Thank you for your interesting posts on surface flow visualization techniques.

Many years ago, a co-worker and I were dispatched on short notice by the boss to the Mojave desert to do flow vis testing inside a turbofan aircraft engine on a test stand. It was a short test window availability situation.

He supplied us a kit full of oils of different viscosities and black & white colorants. We were trying to determine if there was flow separation on the inner fan duct of a high bypass turbofan engine.

However, when you are in Mojave heat in the summer, and the surfaces are inside an engine, even the most viscous oil is going to get moving early in a throttle up cycle.

We managed to obtain some interesting results at ground idle power. The oil clearly showed flow separation in the fan duct and in front of the upper bifurcator pylon. However, it was pointless to try to obtain any results at higher power settings because the most sticky, viscous oil available had already flowed and been blown out the nozzle at ground idle power.

We also saw that oil drops moved aft down the fan duct, around the fan outlet guide vane airfoils, radially outward along the entire trailing edge of the outlet guide vanes, then streaked downstream along the outer fan duct. I couldn't believe the droplets stayed together long enough, (in hurricane force winds), to make this journey, but they did.

Sometimes even engineers get to have some fun.

-Paragon
Cincinnati, OH
 
This is a great story, thanks. Illustrates again the special challenges in doing tests like this, even on a test stand. Imagine trying to do that in flight!

I'm glad I'm not the one that had to clean the oil out of the acoustic liner in the inlet or bypass duct!

An alternative to oil flow that can sometimes be very useful is tufts. We have used very fine nylon monofilament glued on with a drop of CA glue in a syringe. Then illuminate the area with UV light and the nylon fluoresces blue.
For an engine inlet duct test, you would have to do it at night, but the micro-tufts are durable enough to withstand that environment. Otherwise, conventional white thread tufts might work for daylight illumination and photography.

Steve,

Thank you for your interesting posts on surface flow visualization techniques.

Many years ago, a co-worker and I were dispatched on short notice by the boss to the Mojave desert to do flow vis testing inside a turbofan aircraft engine on a test stand. It was a short test window availability situation.

He supplied us a kit full of oils of different viscosities and black & white colorants. We were trying to determine if there was flow separation on the inner fan duct of a high bypass turbofan engine.

However, when you are in Mojave heat in the summer, and the surfaces are inside an engine, even the most viscous oil is going to get moving early in a throttle up cycle.

We managed to obtain some interesting results at ground idle power. The oil clearly showed flow separation in the fan duct and in front of the upper bifurcator pylon. However, it was pointless to try to obtain any results at higher power settings because the most sticky, viscous oil available had already flowed and been blown out the nozzle at ground idle power.

We also saw that oil drops moved aft down the fan duct, around the fan outlet guide vane airfoils, radially outward along the entire trailing edge of the outlet guide vanes, then streaked downstream along the outer fan duct. I couldn't believe the droplets stayed together long enough, (in hurricane force winds), to make this journey, but they did.

Sometimes even engineers get to have some fun.

-Paragon
Cincinnati, OH
 
Not really flow separation. This is a laminar separation bubble followed by turbulent reattachment.

Thanks for the correction. I re read the article and it says the turbulent flow has less drag because it has lower skin friction drag but higher form drag
 
Thanks for the correction. I re read the article and it says the turbulent flow has less drag because it has lower skin friction drag but higher form drag

Other way around. Turbulent boundary layers have higher skin friction. This is because the small eddies embedded in the boundary layer mix higher velocity flow down closer to the surface, so there is more surface shear stress.

But turbulent boundary layers are more resistant to separation, so less chance of causing a bunch of pressure drag (form drag) from a separated flow region. They are more resistant to separation for the same reason as the higher surface shear stress -- the eddies mixing in the boundary layer bring fresh, higher energy flow down into the boundary layer.

The laminar bubble that forms as part of the transition process generally is very small and creates very little drag. Laminar bubbles can get big enough to create some drag, and the alternative is to force transition to a turbulent boundary layer ahead of where the laminar bubble would form. This is done with zig-zag tape or dimpled trip tape, or with pneumatic blow holes. This technique of tripping the boundary layer to avoid laminar bubbles can be quite successful, as it was on the Std Cirrus.
 
When mounting a blade-type transponder antenna underneath my Lancair, I first tried using tufts to determine the direction of the airflow. That was an abject failure because the propwash during takeoff just slammed the tufts all over the belly.

AIsPwa.jpg


So I drilled a small hole in the belly where the coax would go. Then, on the inside of the baggage compartment, I stuck a 3' length of 1/4" Tygon over the hole using modeling clay. On the other end of the tubing I inserted a small syringe filled with Champion spark plug thread lube. I could reach the syringe while flying. Then I flew the plane. When I got up to cruise configuration, I discharged the spark plug fluid out into the airstream. It made a nice, black streak back from the hole at a 7 degree angle to the centerline of the aircraft. That's how I mounted my antenna.

8wvac9.jpg


Z1HTY4.jpg
 
I like that idea, I have 4-5-6 small tubes from the cabin to the engine compartment, I can use this to check the flow exiting the chute!

Steve, update on the UV of the aeroshell 15w50 - I went in the dark and could not see the oil. Is it just the W100 oils that lights up with UV?

I have some oil UV dye and thinking about using that. I'll test a paint sample first to see if it can be cleaned off.
 
Flow Visualization

Many years ago we did a study to determine where the flow transition from laminar to turbulent was on a Blanik L-13 sailplane. We tried two techniques, the oil method and sublimating chemicals. For our testing, the oil method worked quite well. But as Steve mentioned, it may not be useful for 3D flows such as the ones you want to investigate.

What worked the best for us was Mobil 1 motor oil with lamp black mixed in. We simply painted it on the wing, approximately back to the point where we thought the transition would occur, and were towed aloft. The point of flow transition showed as a darker color, and there were obvious v-shapes indicating where bugs and other flaws tripped the flow.

We used a chase airplane and photographed the glider at different angles of attack. Today it is much simpler and better to use properly placed GoPro cameras to record the flow, as the oil does shear off after a period of time and the visual aspect is lost. Takes a bit of clean-up after the test!
 
The presence of turbulent wedges from small roughness elements can be key in correct interpretation of transition. Sometimes we put 'bugs' on the leading edge on purpose to achieve that. Thanks for the good story. Too bad L-13s are now pretty much recycling bait. Really nice-flying glider.

Many years ago we did a study to determine where the flow transition from laminar to turbulent was on a Blanik L-13 sailplane. We tried two techniques, the oil method and sublimating chemicals. For our testing, the oil method worked quite well. But as Steve mentioned, it may not be useful for 3D flows such as the ones you want to investigate.

What worked the best for us was Mobil 1 motor oil with lamp black mixed in. We simply painted it on the wing, approximately back to the point where we thought the transition would occur, and were towed aloft. The point of flow transition showed as a darker color, and there were obvious v-shapes indicating where bugs and other flaws tripped the flow.

We used a chase airplane and photographed the glider at different angles of attack. Today it is much simpler and better to use properly placed GoPro cameras to record the flow, as the oil does shear off after a period of time and the visual aspect is lost. Takes a bit of clean-up after the test!
 
OK, I could not find tempura powder locally, but tried blue snap line chalk. I mixed it with oil and smeared it on a white paint sample for a test. It mixes well, and did not stain the white paint sample.

Tempura is apparently an egg based paint and is not good for any painted surface, so I have some available for face painting.

Next: it gets flight tested on some white parts when weather permits.
 
Back
Top