Nov 07, 2018, 09:04 AM
http://bit.ly/2zm6gK7
Discussion

# Parametric CAD setup for Airfoil Studies

So some of you here have been following me as I begin my 'in-depth' studies of KF airfoils (for RC aircraft). If you haven't, you can gloss through these threads to see where I'm heading
- https://www.rcgroups.com/forums/show...troduction-CFD
- https://www.rcgroups.com/forums/show...e-Basics-First

Now, while I'm waiting to see if I can get into a wind tunnel for some 'real' (controlled) testing, I will be running ALOT more CFD simulations (and refining my process). In preparation for the extensive number of tests I plan to run, I needed to develop a method of producing KF airfoil geometry that was consistent in key areas while I vary other portions of the geometry (step location, thickness, etc).

Here is what I have so far. All tests will be run in Imperial units for now, though these are easy to convert later.

Chord: 12in (irrelevant since we will be pulling non-dimensionalized coefficients)
Thickness: varies, parameterized by % of chord, measured from TOP of TE, not bottom of airfoil
Step Location: varies, parameterized by % of chord
LE radius: constant, defined as 0.1in
Top Curve: varies, defined by control-point spline with the main control point at 60% of step location, with end points on LE and step ends
TE Thickness: constant, defined as 0.1in

Photos are attached for reference. As I generate airfoils, I will provide the geometry files for others to play with as well. I am looking for input, mainly on the setup of the airfoils as described here because I plan to start running preliminary CFD sims soon.

I will be documenting the process through threads here on RCG for discussion, but the main 'meat' of the study will be through YouTube videos (which I will include in the threads). I will be plotting drag and moment polars for the KF airfoils as well as some baseline airfoils (Aquila or similar, naca2412 etc) to compare. The first step will to be to refine my simulation settings/meshing until I can get close to published experimental values for my CFD sims (will test on NACA0012 or similar).

Excited to get started!

### Images

View all Images in thread
 Nov 07, 2018, 09:52 AM treefinder Without pondering your parameters in great detail. Most look ok, buy the trailing edge at .1" is even smaller than a dtf one would be and i doubt anyone would actually build a wing with that thin an aft section. I think you will find good support for a .22 or .25" aft thickness/TE as that would represent most fanfold/MPF/depron in use.
Nov 07, 2018, 10:04 AM
http://bit.ly/2zm6gK7
Quote:
 Originally Posted by springer Without pondering your parameters in great detail. Most look ok, buy the trailing edge at .1" is even smaller than a dtf one would be and i doubt anyone would actually build a wing with that thin an aft section. I think you will find good support for a .22 or .25" aft thickness/TE as that would represent most fanfold/MPF/depron in use.
Good point. DTF is what, 4mm thick?

The problem is, I wouldn't model the Aquila airfoil with a DTF sized TE either though; and since I'm trying to be as "apples to apples" as possible, I might consider leaving it 0.1in. that way the CFD program will pick up the drag of the TE but not create such a large disparity when compared to a traditional airfoil with a sharp TE
 Nov 07, 2018, 10:30 AM low tech high tech 5mm-6mm as springer indicated for the most common foamies, 3mm for more specialized smaller and indoor, and on down. I guess it depends on whether you are interested in designing high performance airfoils for model gliders, racers or other specialized aircraft, or on the other hand whether you are interested in designing airfoils for sheet built powered model foamies. The requirements are very different, and so are the trailing edge treatments. Probably a good idea woukld be for you to define a mission for any series of foils you intend. Rapid easy construction is a kind of mission which isn't often addressed in studies of aerodynamic efficiency, yet that has been the primary appeal of KFMs. How rapid, how easy, relatively speaking with other "easy" methods vs specific performance parameters. Frankly there is no such thing as a good airfoil, or a better airfoil by itself. Only in reference to a specific size, use, weight, mission, and material does it begin to have meaning.
Nov 07, 2018, 10:53 AM
low tech high tech
Ferret DLG on the left (KFM foil), PWFerret on the right (PW1211 airfoil).

Question: which is better?

You cannot answer that question, without asking me a number of questions.

### Images

View all Images in thread
 Nov 07, 2018, 04:37 PM Build straight - Fly twisty I believe the 'traditional' squared off trailing edge is an important component in a KFm airfoil. My contention is that the turbulent wake it produces forms a virtual trailing edge, and this cord extension is the reason that a KFm can often work with a more rearward CG than is usual. It is also one of the reasons that some users feel the sections make for greater stability. Their 'geometrically correct' CG position is made to be a more forward position by the action of the virtual trailing-edge extension.
 Nov 07, 2018, 05:13 PM B for Bruce I'm posting this as a bit of a WAG but I'm pretty sure that the point of CFD and other predictive software is to arrive at mathematical equations for the computer to use which describes real life. It may be something like the Eppler code which is specific to model airfoils. Or it may be trying to mathematically describe how air moves in near and far fields around any generic object. The second situation seems to be what I've mostly seen in connection with research on KF airfoils. But all those things don't mean anything to us wanting to use it as an airfoil if we can't also include what that other air movement means in terms of lift and drag to our rather specific needs. And without that in some manner of trusted computer form we're left with using a wind tunnel to directly measure these specific behaviors. And then perhaps the CFD software can be tweaked until we have a mathematical model programmed into the software that predicts those same results for a few of the directly tested samples. Once we get to that point THEN we can use the new tuned up CFD program to obtain decently trustworthy predictions of other stepped airfoils or even fuselages. Yes? No? Latest blog entry: Sharpening carving knives for balsa or...
Nov 07, 2018, 06:23 PM
treefinder
Quote:
 Originally Posted by corsair2014 Good point. DTF is what, 4mm thick? The problem is, I wouldn't model the Aquila airfoil with a DTF sized TE either though; and since I'm trying to be as "apples to apples" as possible, I might consider leaving it 0.1in. that way the CFD program will pick up the drag of the TE but not create such a large disparity when compared to a traditional airfoil with a sharp TE
You know far more than i about the math, but a 0.10" TE is something that essentially doesnt exist in the real world of kfm wings and flyers ( in the same way that a conventional airfoil with a step cut in it is done so seldom simply because it violates the fundamental reason why we pick KFM- ease of construction. Why go to all the trouble of building a conventional wing then cut out the top half?) Being apples to apples with something no one does isn't particularly useful to me. Even when i build a conventional wing, the TE is more than .1" thick, to improve it's surviveability. I probably sacrifice efficiency (would be nice to know how much) but a TE that stays looking nice is better than one efficient in principle but looks trashed after a couple flights.
 Nov 07, 2018, 11:30 PM low tech high tech re. Ferret and PWFerret....which is better.......some of many potential questions you might ask me, besides the obvious ones related to foil aerodynamics: Were these be used for thermal soaring or slope soaring, or both? How did the foil shape affect the CG location -- was ballast required on one compared to the other with the same gear? How did the foil shape effect the AUW? How does wear and tear affect the performance or longevity per foil? Does a potential builder own a hot wire setup? Does the builder have access to thick foam, or thin foam sheets? What is the time required to build? What are the skills needed to build? Which plane will have better performance if not built entirely true to form? I can say right now that I prefer to throw the PWFerret for thermal hunting, but prefer building and always recommend to others to build the Ferret. The PWFerret also required additional ballast compared to the KFM foiled plane because of the distribution of mass in the foils themselves. A foil forces design considerations beyond its listed polars, and these can affect performance. Which is the better plane? Neither. The question is too general. It is important to be very specific about intents and mission(s) to avoid generalizing from data.
Nov 09, 2018, 09:25 AM
http://bit.ly/2zm6gK7
I apologize for not responding sooner, I didn't see the notifications for this thread (and Autdoesk CFD was quickly destroying my confidence in it's abilities once again so I was preoccupied with that).

I think I will change the TE thickness to 5mm for now, since as alot of you mentioned this is about what most wings would be built around anyway. Also, I will work on defining the basic "mission" of these airfoils, which will probably be for FPV/camera ships (flying wings or twin boom type planes etc) that are easy to build (and crash lol).

Quote:
 I believe the 'traditional' squared off trailing edge is an important component in a KFm airfoil. My contention is that the turbulent wake it produces forms a virtual trailing edge, and this cord extension is the reason that a KFm can often work with a more rearward CG than is usual.
That is really interesting, and makes sense! I will look for this when doing testing.

Quote:
 I'm posting this as a bit of a WAG but I'm pretty sure that the point of CFD and other predictive software is to arrive at mathematical equations for the computer to use which describes real life. It may be something like the Eppler code which is specific to model airfoils. Or it may be trying to mathematically describe how air moves in near and far fields around any generic object.
If i understand correctly, you are commenting on the validation/accuracy of the CFD program right? I agree, I have NO idea how accurate it currently is lol, so that is why I am currently running validation tests using a NACA 0012 airfoil at various AoA and Re#. However, since Autodesk CFD as crapped out for the 20th time in the last couple weeks, I think I am going to force myself to learn Fluent now. Will take longer now that I will have to use a new program and won't be able to Cloud Solve my cases, but at least it won't crash on me every other simulation... more to come on this.

Quote:
 Even when i build a conventional wing, the TE is more than .1" thick, to improve it's surviveability. I probably sacrifice efficiency (would be nice to know how much
I will be thickening the TE to 5mm, but that would be interesting to run the same airfoil but with 5mm TE compared to 1mm TE... perhaps in the near future we can do a 'quick' analysis.
Nov 09, 2018, 09:27 AM
http://bit.ly/2zm6gK7
Quote:
 Originally Posted by vtdiy Ferret DLG on the left (KFM foil), PWFerret on the right (PW1211 airfoil). Question: which is better? You cannot answer that question, without asking me a number of questions.

Now, this has perked my interest in building a small glider again haha. Have any links to more info on these? Anyone power them, or only gliders? Looks fun
 Nov 09, 2018, 09:58 AM low tech high tech Corsair, the Kferret thread is here: https://www.rcgroups.com/forums/show...-maiden-report The PWFerret isn't an existing named design -- is just something I built myself, using the same planform and a PW1211 foil. I just call it that. On this scale, and this complexity of model, wind tunnels aren't needed, because it's easy to test actual size, and not expensive also. It does show up differences in airfoils under real world conditions -- maneuvering, wind shifts, recoveries from displacements, thermals vs obstruction lift, etc, etc. And it does that in a purer form than tests with a motor, its slipstream, throttle setting, etc. I think simple small flying wing DLGs are an excellent testbed for aerodynamic theory in a real world application. I am opposed to generalizing from that very much however -- it just tells you about small flying wing DLGs in the conditions you have where you live, and in the way you throw. But there, yes it's possible to compare and draw personal conclusions.
Nov 09, 2018, 10:31 AM
http://bit.ly/2zm6gK7