Some questions about KF airfoils and other things

Homebuilt Aircraft & Kit Plane Forum

Help Support Homebuilt Aircraft & Kit Plane Forum:

Sockmonkey

Well-Known Member
Joined
Apr 24, 2014
Messages
2,274
Location
Flint, Mi, USA
So I was going back over some stuff about the KF airfoil, and I was thinking that maybe the stability they display is because the notch creates a fixed point where the airflow goes from laminar to turbulent, so the COL doesn't wander around when the AOA changes.
On that note, would regular wings benefit from having a tiny version of that step if it works the way I think?
 

Sockmonkey

Well-Known Member
Joined
Apr 24, 2014
Messages
2,274
Location
Flint, Mi, USA
Okay, but that does create turbulence at that point yes?
Does that specific separation point encourage the airflow to stick to the wing more strongly until then?
 

Norman

Well-Known Member
HBA Supporter
Joined
Nov 28, 2003
Messages
3,132
Location
Grand Junction, Colorado
Okay, but that does create turbulence at that point yes?
The idea of the step is to force a separation bubble although much larger than a natural transition bubble. As with a normal, non-forced, bubble the shear layer will transition to turbulent off-surface if the boundary layer was laminar when it hit the step. If the boundary layer was already turbulent then it will just stay turbulent until it hopefully reattaches after the bubble.

Does that specific separation point encourage the airflow to stick to the wing more strongly until then?
There hasn't been any good investigation into that. The only real wind tunnel test of stepped airfoils that I know of that measured the pressure gradient was that by Fathi Finaish and Stephen Witherspoon but their model's BL was already turbulent long before the step. You can see though that the step does produce a favorable pressure gradient a short distance upstream so if the basic airfoil was laminar to 40% it would probably stay laminar another 10 or 15% until it got to the edge. The pressure coefficient graphs from their paper are reproduced here.
 

Tiger Tim

Well-Known Member
Joined
Apr 26, 2013
Messages
3,932
Location
Thunder Bay
I kind of suspect the KF airfoil doesn’t do anything meaningful to anything bigger than the paper airplane for which it was developed. Even the RC models that use it IMO really use the extra layer of foam that makes the step as a structural member more than an aerodynamic enhancer.
 

Sockmonkey

Well-Known Member
Joined
Apr 24, 2014
Messages
2,274
Location
Flint, Mi, USA
The idea of the step is to force a separation bubble although much larger than a natural transition bubble. As with a normal, non-forced, bubble the shear layer will transition to turbulent off-surface if the boundary layer was laminar when it hit the step. If the boundary layer was already turbulent then it will just stay turbulent until it hopefully reattaches after the bubble.


There hasn't been any good investigation into that. The only real wind tunnel test of stepped airfoils that I know of that measured the pressure gradient was that by Fathi Finaish and Stephen Witherspoon but their model's BL was already turbulent long before the step. You can see though that the step does produce a favorable pressure gradient a short distance upstream so if the basic airfoil was laminar to 40% it would probably stay laminar another 10 or 15% until it got to the edge. The pressure coefficient graphs from their paper are reproduced here.
Ahh, gotcha. That last bit is kind of what I suspected.

Anyhow, here are some thoughts about airflow that may be horribly wrong, but are probably worth thinking about.

The second one is meant to create a sort of "virtual" airfoil defined by the ribs.
 

Aesquire

Well-Known Member
Joined
Jul 28, 2014
Messages
2,964
Location
Rochester, NY, USA
Iirc there was a fad in the 70s ish for KF airfoils on Combat U-Control planes. NASA was talking about them, so... I played with them.

I've built thousands of paper airplanes. Mostly 2 classes. A. Variations of the Barnaby Special. B. Flex wing hang glider/trike wings with sweep and twist.

And a few hundred Darts. ( sometimes a week )

I played the chimes on aspect ratio, twist, tip shapes, etc. Testing stability & performance.

And playing with the origami. Fold under? Over? Tape?

So, naturally, while paper is poor for testing the difference between a laminar NACA & USA 35B, it's perfect for KF vs. Flat plate & cambered foils.

I even tested in the garage wind tunnel. :) Bottom line, my gear wasn't sensitive enough to be sure. I have the semi-intuitive notion that there is a drag reduction at certain narrow speeds & Reynolds Numbers, but that's from looking at a shotgun pattern plot of data that's in the error bars, & could be a Rorschach test... My brain seeing patterns not there? Wish I still had those notes, I bet a supercomputer, or phone app could decipher a trend.

Free flight testing got the same uncertainty, with zero gap (2 to 1 sheet thickness step.) through aprox. 2-5 mm step, being roughly equal. Testing methodology was light rubber band launch off clip board on steps of Gym. ( with ventilation off )

I could, however, see a big jump in drag as you increase the step height that seemed related to Reynolds Numbers.
 
Last edited:

llemon

Well-Known Member
Joined
Apr 29, 2021
Messages
61
Ahh, gotcha. That last bit is kind of what I suspected.

Anyhow, here are some thoughts about airflow that may be horribly wrong, but are probably worth thinking about.

The second one is meant to create a sort of "virtual" airfoil defined by the ribs.
This is not far off dragonfly airfoils, which can work quite well at very low reynolds numbers (<100k). But for a man carrying aircraft the Re will be so much higher that more conventional airfoils are the right choice.

 

Norman

Well-Known Member
HBA Supporter
Joined
Nov 28, 2003
Messages
3,132
Location
Grand Junction, Colorado
I kind of suspect the KF airfoil doesn’t do anything meaningful to anything bigger than the paper airplane for which it was developed.
Vortex formation behind a step is not Re dependent. It's a problem in architecture that architects deal with all the time, that's one reason buildings are test in wind tunnels.

Even the RC models that use it IMO really use the extra layer of foam that makes the step as a structural member more than an aerodynamic enhancer.
Yeah, strength goes up with the thickness and stiffness goes up as thickness cubed but the stepped plate has better aerodynamic characteristics than just a double thickness flat plate. Whether or not that improvement shows up on a good airfoil is the question. Those wind tunnel tests referenced by the web page I linked to in post #5 indicated that there is improvement at some AoA with degradation in the rest of the envelope, this sugests that a movable cover (like a spoiler) would be better than a fixed step.
 

Sockmonkey

Well-Known Member
Joined
Apr 24, 2014
Messages
2,274
Location
Flint, Mi, USA
Those wind tunnel tests referenced by the web page I linked to in post #5 indicated that there is improvement at some AoA with degradation in the rest of the envelope, this sugests that a movable cover (like a spoiler) would be better than a fixed step.
I think that's what the coverlets on the topside of a bird's wing do at high AOA during landing.

As for the notch itself, figure 10 on your page about KF airfoils makes me think that the inner corner of the notch should should be a quarter-circle rather than a right angle.
 
Last edited:

Norman

Well-Known Member
HBA Supporter
Joined
Nov 28, 2003
Messages
3,132
Location
Grand Junction, Colorado
Anyhow, here are some thoughts about airflow that may be horribly wrong, but are probably worth thinking about.
If you want to explore slot profiles you should look at the Ringleb cusp. Also cutting a step into the ribs weakens them substantially which means you'll need to use a lot more material to get the trailing edge to stay where it belongs so make those steps as shallow as possible. Also if you do multiple steps the distance between steps should probably be around 10% of the chord, as indicated by the favorable pressure gradient forward of the step in the above sighted paper. There are several relevant links in this thread on RCG. There may also be some newer stuff on the net so Googling with the names and concepts in that thread may be fruitful.
 

Attachments

Norman

Well-Known Member
HBA Supporter
Joined
Nov 28, 2003
Messages
3,132
Location
Grand Junction, Colorado
I think that's what the coverlets on the topside of a bird's wing do at high AOA during landing.
Yep!

As for the notch itself, figure 10 on your page about KF airfoils makes me think that the inner corner of the notch should should be a quarter-circle rather than a right angle.
Ringleb's cusp cove is actually elliptical. It came out of wind tunnel research so that's well backed up but a circular notch would be better than a square one, both aerodynamically and structurally
 

Sockmonkey

Well-Known Member
Joined
Apr 24, 2014
Messages
2,274
Location
Flint, Mi, USA
Yep!



Ringleb's cusp cove is actually elliptical. It came out of wind tunnel research so that's well backed up but a circular notch would be better than a square one, both aerodynamically and structurally
How critical is the spanwise flow? Because you could just have notches in between normal-profile ribs.
 

Norman

Well-Known Member
HBA Supporter
Joined
Nov 28, 2003
Messages
3,132
Location
Grand Junction, Colorado
Refer to figure 4 on this page. If the axial flow of a vortex is blocked at both ends it will grow, pop, grow, pop and repeat indefinitely at a frequency that I think can determined with the Stouhol number. The axial jet is faster than the free stream flow and has to be dumped overboard at one end for the vortex to remain intact and act like a roller bearing. If the walls (ribs) are high enough and the rest of your step geometry can trap the vortex and prevent the axial jet from eskaping your step may just be a spoiler. On the other hand it jet can turn downwind and escape from the trench it would probably be fine.
 

Sockmonkey

Well-Known Member
Joined
Apr 24, 2014
Messages
2,274
Location
Flint, Mi, USA
Refer to figure 4 on this page. If the axial flow of a vortex is blocked at both ends it will grow, pop, grow, pop and repeat indefinitely at a frequency that I think can determined with the Stouhol number. The axial jet is faster than the free stream flow and has to be dumped overboard at one end for the vortex to remain intact and act like a roller bearing. If the walls (ribs) are high enough and the rest of your step geometry can trap the vortex and prevent the axial jet from eskaping your step may just be a spoiler. On the other hand it jet can turn downwind and escape from the trench it would probably be fine.
I was worried about that. Most ribs are open lattice structures so the sides of them could be left uncovered to allow the flow through assuming the lattice didn't screw up the flow too much.
Structurally, you might be able to get away with making the part of the rib in the notch completely open inside.
 

Sockmonkey

Well-Known Member
Joined
Apr 24, 2014
Messages
2,274
Location
Flint, Mi, USA
Going off on another tangent for bit, a bit of research on the old Tesla turbines says that to be efficient, they either have to run at structurally impossible RPMs or be used for viscous fluids.
Now I'm thinking they might have a niche in RC models or micro AVs where the small size means those high RPMs are doable and the air viscosity is proportionally greater.
Use it to turbocharge one of those tiny RC engines, or go nuts and make a tiny Tesla turbine jet engine.
One guy made a tiny Tesla turbine that made it up to 87,000 RPM.
 
Top