Airflow visualisation
 

Ever since i started flying pattern models, i've been curious about their aerodynamics and how the airflow is influenced by the different flying attitudes (horizontal flight, knife edge, vertical uplines, etc). Most of the time i've been too busy with designing, building, practicing and competitions but now, having a 7 weeks break between contests, i finally found time for some tests.

Tufts are an old visualization technique that is used in both flight test and wind tunnel testing to provide diagnostic information about the flow around the model. The fuselage of my Radiance prototype is completely covered in white film, so it was ideal for this kind of tests. Using many many wool tufts on the fuselage and a small keychain camera taped to the wingtip, i made a few video recordings of the airflow around the fuselage.

http://www.rcuniverse.com/forum/atta...mentid=1914142

Taking screenshots from the videos, i put together a few images that cover the whole fuselage. The keychain camera has a narrow field of view, so i had to make separate videos for the nose and tail of the model.

As expected, the camera on the wingtip visibly affected the flying qualities of the model, but it was quite manageable. I noticed that trim was affected for all the controls, the model was flying slower than usual and needed more throttle and fuselage incidence for knife edge flight. Anyway, the camera on the wingtip was far from the fuselage, so the airflow around it was not disturbed.

As a general conclusion, the direction of the airflow doesn't seem to be affected by the spiral propeller slipstream. Please keep in mind that these pictures only show the local direction of the airflow, the speeds and pressures may still be different on the left and right sides of the fuselage.

The following images cover the horizontal flight, vertical uplines and downlines, horizontal knife edge flight and knife edge loops. I am sharing the results of these tests hoping that together we can find new ways to improve our models. As you will see, the vertical tail offers a few surprises which may explain the need for canalisers to improve the rudder authority.

Horizontal flight:

Vertical Upline:

Vertical Downline:

Left Rudder Knife Edge:

Right Rudder Knife Edge:

Left Rudder Knife Edge Loop:

Right Rudder Knife Edge Loop:

As expected for the knife edge flight, the difference in pressure between the 2 sides of the fuselage makes the air leak from the high pressure side to the low pressure side, over the canopy, turtle deck and belly pan.

The big surprise comes from the vertical tail, which acts like a swept wing and develops a strong spanwise flow on the low pressure side, losing lift and increasing drag meaning reduced efficiency.

There are many ways to limit spanwise flow and delay stall in swept wings, some of which may be applied to the fin of our models. A simple way would be to add 1-2 horizontal stall fences or strakes, as Bryan Hebert successfully used for his Valiant.

Here's a good read about swept wings and stall fences: http://www.b2streamlines.com/WingFences.pdf


It's still not clear for me how the canaliser works, but it probably generates a vortex trail that energizes the boundary layer and delays stall on the vertical tail. It also acts like a winglet for the fuselage, preventing the air from leaking from the high pressure side to the low pressure side of the fuselage and increasing the lift of the fuselage this way. My guess would be that it does this only locally and the air still migrates from one side to another over the turtle deck.

There does seem to be some difference in the fuse top tufts in the vertical up and down lines. Were you power-off on the downline?

Alex,
Also an interesting difference between left rud,, ke loop left side and right rud,, ke loop right side.
Above the wing from in front of the wing to mid rear fuz,, .

Great work Alex! It would be very interesting to see the effects of a 'fence' on the vertical fin. Also, adding a 'T' canalizer. Separate experiments on the same airplane to get a good picture of the effects of each.
Thanks for sharing this!
- Will

Alex,

Very nice experiment, it has a lot of answers, are you planing to install a T Canalizer just to see the effect?

Great idea you had.

Regards

Thanks for the great pictures Alex.

I think you would find it interesting to do some spins, upright and inverted. Tufting the wings will add to your understanding of what is happening. Oh, try some stall turns both ways also.

Very cool, thanks for sharing.

Yes, i cut the throttle for the downlines, the brake was activated and set to 30%.


I noticed that too, but i expected some significant differences on the vertical tail to justify the need for 3 deg right thrust, as most model use.

Yes i tested a canaliser a while ago but to be honest i didn't like it. The model needed less rudder for horizontal knife edge, but pulled to the canopy on both sides. When i added more rudder to start the knife edge loop, it pulled to the belly. I tried many incidence settings for the canaliser and different CG positions, but couldn't solve this problem.
Radiance never had problems with rudder authority and can perform a knife edge loop comfortably without canaliser, so for me the small improvements were outweighed by the disadvantages.

Hi Alex,

I think most of us wonder what changes in the airflow, if it changes at all.

The canaliser is easy to put on or remove, so i'll probably test it. The problem is tufting the fuselage again which takes a lot of time and patience, so i may tuft only the fin, rudder and turtle deck.

I find this experiment really interesting. Congratulations Alex. I find also interesting that there is no significant proof that there is a propwash or a spiral flow around the fuselage in any position tested, as per the long discussion of the other day. the flow seems to pe pretty linear from the prop all the way to the rudder. I guess somebody in Luisiana might have been right in the fact that either it does not exist or it is effect is neglictable in pattern ships...

Why do pattern ships have right thrust? A clue: Contra drive pattern ships do not have right thrust.

Jim O

Alex,

Awesome work. Thanks!

To add to Jim's comment....why do the Rasa articulated props, which have a lot less area at the hub than typical prop designs, require significantly less right thrust? I have observed this in constant speed level flight, irrespective of design, mono and bipe.

Regards,

Andre Bouchard

My thought is that it would be hard to see the spiraling air with the tufts being small. We know the air spirals, you can feel it standing behind a plane and glow engines always slime on side of the stab more than the other. But close to the fuse, there is a lot going on: spiraling air is blocked from the opposite side, the 'no slip' effect (the air touching the fuse is moving with the fuse) and subsequent boundary layer, and the inner part of the prop has the highest pitch and lowest diameter, so the spirals are long and stretched out.

Look at the pics of the top of the plane during straight flight and vertical down. The tufts on vertical down seem to split off to either side, but in horizontal flight it seems the tufts are just slightly biased toward the right side of the plane.

A tuft centered on top of the rudder would be the best indication, but still the air will have been straightened out by the fin a little.

There are a lot of things at play that cause a plane to want right thrust:
-spiraling air stream
-torque
-gyroscopic effect
-p-factor

The contra planes combat almost the whole list.

The props with less area at the hub would have less P-factor effect, since the portion of the prop with the highest pitch is eliminated.

Whoever said spiraling air 'doesn't exist' is likely a Republican Scientist who moved to aerodynamics after disproving climate change.

Interesting post and clever use of available technology to the collect data.

I will agree the contra drive does combat that list but I disagree that they all cause the plane to want right thrust.

Torque is a rolling moment and one would use ailerons to trim a rolling moment. Right thrust would not help much.

Gyroscopic precession and P-factor can induce a yaw moment but it is reversed if the maneuver is a push vs. a pull, so right thrust would not compensate for it in both directions.

QED?

Jim O