ptxman

My Feedback: (2)

In the real world of building wings we always have to truncate the trailing edge of flying surfaces to some amount relative to the typical airfoil theoretical end point coordinate (x=100, y=0). Yet we use the polar data which assumes this perfect profile. It seems that in 'airfoil intensive' pursuits like racing, sailplanes, high performance etc, every effort is made to minimize truncation & extend the profile as much as possible & of course keep it rigid. The thinner the airfoil profile, the harder this is to achieve assuming a minimum acceptable TE thickness.

My question is, where on the airfoil polars would this truncation show up as a degradation? An overall inneficiency loss spread out over the Cl/Cd curve? Or more weighted at high AOA & therefore reduced Cl max? See what Im getting at? Do fat airfoils suffer less than thin airfoils? Is there a magic rule of thumb like "if you achieve less than 95% of the theoretical chord, the polars cannot be trusted & you need to use a fudge factor"? Assuming a truncation is inevitable, would it be better to radius the edge or leave it square?

Tall Paul

Is there a method of determining the real-life performance of any model, and comparing that to a predicted value with any degree of certainty?
There's many "calc" programs around which claim this capability, but all of them make assumptions which may not be real.

rmh

If theory on full scale can relate -- --on the truncated airfoils -- the idea is to create a low pressure on the T edge - which sucks the air down to the wing surface .
How's that for basic?
On the models -- can't proove it one way or another.
At least I can't someone stole all my Reynolds Numbers --
What becomes clear tho --is that the structural change (thicker ) makes it more rigid.
We abandoned the super sharp TE years ago -
Way back -beforetime began - a fellow by the name of KAMM - came up with the novel idea that IF- you could not properly streamline a car body - then you were best off to simply whack it off as short as possible-once you got the front /sides and bottom right .
This also decreased wetted area ---

MadScientist

What was the reason for abandoning the sharp TE? I was curious, as a fellow modeler claims vast improvements in airspeed/handling by feathering down his TE on his planes. I'm mostly referring to 40 size planes with symmetrical airfoils and flat plate elevators, like an Ultra Stick.

banktoturn

ptxman,

Why do you say truncation is inevitable? If performance is really important, you could choose to use harder wood for your trailing edge, and just make it sharp.

If the trailing edge is blunt, the flow will separate. If there are sharp corners (a squared-off trailing edge), the flow will separate right at the corners. If it is rounded, the flow will stay attached partway around the curve, and then separate. The main effect will probably be slightly higher drag. If you look at a plot of upper and lower surface pressure toward the trailing edge, you will see that the pressure difference is getting pretty small in that area. Since the amount of area lost is small, and the pressure difference is small, the amount of lost lift would probably be negligible. Note that if the flow would have separated upstream of where the trailing edge is truncated, then truncating the edge would make almost no difference.

I don't see any reason to intentionally blunt the edge. Sand it as sharp as you can. A few dings won't be a big problem, and if they are, use harder wood. It certainly doesn't weaken the structure to leave the sharp edge on there, and it may save you a little drag.

banktoturn

onewasp

Can't let this one go !

Separation DOES occur before the TE ----well before in most cases. A sharp TE increases the flutter possibilities markedly.

Read Dick Hanson above! Kamm "as in Kamm backed cars etc." determined that there was an "ideal" length of taper to create the least drag. Since this is generally "way beyond" the reasonable limits, he then determined that the best compromise was to cut it off flush at the desired point and that, would yield the best compromise. This was proved in actual wind tunnel test and these data are available.

Do a search on KAMM or hopefully one of the real aerodynamicists on RCU, like Ben Lanterman etal will comment tonight. TallPaul already has commented but in a less than pointed way. He too may re-visit this exchange before the evening is over.

Tall Paul

If your intention is to minimize Cd in flight, then every step that does this is worth the effort.
For a plane which is yanked and banked for most of the flight time, it machs nichts what's at the trailing edge..
But..
If you want to squeeze every mile an hour or minute of flight out of the plane, then ignoring the trailing edge will not get the results you want.
I have yet to see a jet with a blunt trailing edge, or an Expert class pattern plane, or a Goodyear racer, or a thermal duration glider, or competetive slope racer, etc, where speed and or time of flight is paramount.
The StiK/Kaos family.. who cares? Anything works.

Shoe

I have heard the same as Dick Hanson, that in some cases you can actually see a drag REDUCTION with a properly truncated trailing edge. The idea is that the low pressure in the separation bubble behind the trailing edge will reduce the adverse pressure gradient seen by the boundary layer it approaches the trailing edge, thus keeping it attached. Unfortunately, I don't know of any "rules of thumb" or analysis tools that might help here.

I would expect that if you didn't have a boundary layer that is at or close to separation, truncating the trailing edge would result in a drag penalty.

banktoturn

onewasp,

I'm a little familiar with Kammback cars, and Kamm's work. If you look at cars built for absolutely minimum drag, such as solar powered racers, you will see no Kammback designs. Passenger cars, built to fit in standard parking spaces, be driven on real roads, and carry people & their cargo, cannot afford an optimally tapered back end, and may have to make do with a Kammback. An airfoil designed for minumum drag will not separate before the trailing edge, and will have a sharp trailing edge. The airfoil on model plane may be different matter, but that is not what ptxman was asking.

banktoturn

Johng

My Feedback: (4)

I have heard that in most applications, a small truncation generally doesn't affect the drag but more likely will have a small impact on max lift.

On tiny models, like an HLG, a 1/16" trailing edge will have a much greater effect on flight performance than the same thickness on a 60-sport. That's just a size thing. As far as performance, the more optimized an airfoil or application, the more impact a truncation would have on performance. When building a foam wing, carbon fiber or plywood can be included in the center ply of TE construction to form a sharp edge when the TE is sanded.

I have had the opportunity to look at various full size airplanes, and they don't seem too concerned with a sharp TE. THe Canandair Challeneger trans-continental bizjet has a TE that's at least 3/8" thick and square. Cessna 17x, 18x, & similar have a TE that follows the airfoil contour, but then joggles down flush so the upper and lower skin lap together and get riveted. Plus, you can get a mod that allows a round beaded piece of plastic trim to be rivited between the upper and lower TE skin, leaving the round bead a good bit thicker than the TE. Saves your forehead when you walk into the wing though (and you will). I can't think of any plane that I've bumped up against that had a whole lot of attention paid to a sharp TE. Even modern supersonic fighters don't have a sharp TE, although the F-104 comes to mind as one plane that did have sharp wings on both TE and LE.

BMatthews

I'm not sure if I read it in Soartech 8 or elsewhere but Michel Selig commented on trailing edges and model performance at some point and the end result was if you want to go fast or have lower drag in sailplanes then keep it thin, keep it sharp and keep it stiff.

Obviously the same things would apply wherever speed or efficiency is the goal. The complete antithesis to fun fly's, trainers or Stik's of any sort.....

rmh

Abit more on the Kamm design - the passenger cars are not related to Kamms work - these were all out racing machines -Ithink the Cunningham Coupes were among the better examples.
I had a AH 100 w/ small block Chev - it was spooky much over 100- My Volvo wagon was better ( 5.0 H0 Ford powered.
a shape like the electric cars would simply fly away -worthless as racing cars .
the big low drag LOWSPEED electrics also needed all the solar panel areathey could get.
back before time began -some car company racers were "streamlined at th rear.
These turkeys simply lifted at speed - bad setups.
the current generation of racers have neutral lift -plus adjustale downforce setups - much of which is low belly pressure .
as for the sharp TE on our large aerobatic models
No friggen way thanks -
if in doubt on this -or if you have no familiarity with these --take a look. the 40% models may have TE setups 1/2" thick. on the ailerons -also it is common to see flat plates attachedalong.edge of rudder -to prent hunting .
Sorry guys - Theory is one thing - but when actual application shows otherwise - I have to go with the results .

ptxman

My Feedback: (2)

Tall paul> Is there a method of determining the real-life performance of any model, and comparing that to a predicted value with any degree of certainty?

Hey, you cant answer a question with a question, that’s illegal! <g>. I guess I was hoping someone would say, I was bored at work & truncated airfoils just so happen to fascinate me, so I put 6 identical airfoils into Xfoil, which is a top notch program written by a real smart guy, at varying truncation amounts 100%, 95%, 90%, 85% … etc & attached are the polars which show you blah blah blah…



banktoturn > Why do you say truncation is inevitable? If performance is really important, you could choose to use harder wood for your trailing edge, and just make it sharp.

I’ve miked a few what I would call ‘representative’ all composite high end racers, F3D, F5D, performance soaring wings & the like made from CNC milled molds & using all the cool composite materials & bagging techniques to ensure structural rigidity in its thinnest practical form. The range seems to be between 0.015� – 0.025�. I was involved in a related project & I can tell you with a bit of certainty its actually quite a bit of work to achieve this thickness from the mold milling, layup preparation technique for final product. Someone, somewhere down the line must have figured all this pain was worth the effort because a plain jane 1/16� roundover is dead easy by comparison.


banktoturn > If the trailing edge is blunt, the flow will separate. If there are sharp corners (a squared-off trailing edge), the flow will separate right at the corners. If it is rounded, the flow will stay attached partway around the curve, and then separate. The main effect will probably be slightly higher drag…. If you look at a plot of upper and lower surface pressure toward the trailing edge, you will see that the pressure difference is getting pretty small in that area. Since the amount of area lost is small, and the pressure difference is small, the amount of lost lift would probably be negligible

Great! This is the verbage I was looking for. Thanks.



Tall Paul> I have yet to see a jet with a blunt trailing edge, or an Expert class pattern plane…

You’re right about the jets but I can tell you that my buddies latest ZN design which he flew at the last F3A WC’s have big fat TE’s in the order of 3/16 – 1/4� thick & that is the ‘in’ thing now. I doubt very much it has much to do with eeking out the n-th degree of airfoil performance like in a racer. Ballistic pattern manevers went out in the 80's with the introduction of turnaround. Drag is what they want for downhill breaking & constant speed maneuvers. He tells me it softens/dampens the control response (or something to that effect) & that is a function of their airframes, performance & task at hand. Anyway, different application to what Im talking about.


Bmathews> I'm not sure if I read it in Soartech 8 or elsewhere but Michel Selig commented on trailing edges and model performance at some point and the end result was if you want to go fast or have lower drag in sailplanes then keep it thin, keep it sharp and keep it stiff.

Maybe that’s where I saw it referenced. Did he quantify it in any way? I dont have that publication anymore.

ptxman

My Feedback: (2)

This is interesting. Can you elaborate on why flutter would increase?

Jim Thomerson

Many controline combat airplanes have a piece of 1/4 square spruce for the trailing edge.

Jim

BMatthews

That big fat trailing edge on the combat models is more for toughness during their "arrivals" than for aerodynamic reasons I'm sure.

As for flutter and sharp trailing edges I've only heard that it's the OTHER WAY 'ROUND. Blunt flutters and sharp doesn't. Unless it's not stiff enough. I think you're mixing structural issues with aerodynamic ones.

Tall Paul

.
Yes, adding drag in a useful manner is the other side of the coin. Pattern planes are going away from retracts to fixed gear for that reason... trying for constant up and downline speeds.
As for "softening response" 40 some years ago Walt Good went from streamlined to blunt for his dual proportional planes. On these, the control surface was always in motion, and the plane responded to the average position of the surface. A flat blunt surface worked better; "crisper" response. When proportional control went to the feedback type, the surface stationary. blunt was no longer needed.

Tall Paul

.
Unfortunately. a major portion of the aerodynamic anecdotal information is "seen", and not measured.
There is little genuine quantifiable data that can be depended on to give consistent answers when applied to a model's performance.
Usually we just add more power to make things "better".
Where the performance specs are tight and restrictive, the use of "seat-of-the-pants"/"feel and seem" take second place to genuine information which is generated under test conditions which pick out the good and bad points of the object under test.
Competiition under such specifications weeds out the TLARs and even SWAGs.

davidfee

ptxman,
I just paid the paltry $10 to get my Profili V2.0 upgraded to the full-version. Found out immediately that Stefano is now up to V2.14 with a truck-load of new features. OH MY GOODNESS!!!!!!!! Don't open this program unless you have several hours to spend playing with it!

Anyway, I took the RK40 airfoil and used the Xfoil processing features to thin it to 7.5% to simulate the Avionik D99 wing. Then I used the Xfoil TE thickness tool to increase the TE thickness to 0.5mm. I then made a version with an obscene 2mm TE. I had Xfoil calculate the polars at Re=500000 to see what would happen.

I'm sure that the Xfoil processing works a little differently than just physically truncating the airfoil, but it was interesting to see the calculated results, all the same. It seems you need a very blunt TE to seriously degrade the Cd.

An interesting side-observation is that the Cd becomes very high at very low Cl.

Here's the output:
-David

fainjon

I cant remember who said jets dont have sharp TE but go to an air show and look at all those sharp edges. Round blunt TEs are the most prone to flutter. Sharp TEs are the best for speed and flutter reduction.

BMatthews

Davidfree, what you show in the charts is a very convincing argument FOR sharp trailing edges. Sure the upper edge of the lift coefficents is better with the blunt TE but look at the difference in the low drag bucket area. It may not SEEM like a lot but from reading lots of polar charts and trying to equate them to real lift the differences shown in your chart is HUGE. A lot of the changes in general sailplane airfoil usage is driven by differences of much less than shown in your chart. The gap between the green line and the others in the low drag area is directly due to the thick trailing edge and the turbulence it causes.

And back to the general audience-
You must realize that the only time we approach a Cl=1 or higher is when we're trying to stretch a landing approach or hang the model nose up into the wind and sort of hover there. The rest of the time it can be shown that our models opperate at between Cl= 0.07 to 0.5. This equates to generally bombing around in sport flying mode at a speed that generously avoids any chance of an unexpected stall. Of course there are transients up to very high values during sharp maneuvers but for the most part we spend more time just flying around.

Model sailplanes probably fly at between Cl=0.07 in high speed mode to 0.5 in a slow thermal turn with possible excursions up to 0.8 if the pilot isn't watching his speed and lets it get too slow. At these values the drag shown in your chart equates directly to lost altitude. Altitude for gliders being synonimous to fuel in the tank. For sailplanes that persue speed related challenges this is not an option and that is why you will always see super sharp trailing edges on high performance sailplanes if the builder knows what is good for them.

I determined these numbers by plugging in typical values for a few different models into FoilSim. It was very educational. I had always assumed, in error apparently, that my models spent much more time up in the 0.5 to 1.0 area. 'Tain't so though.

http://www.grc.nasa.gov/WWW/K-12/airplane/foil2.html

But like most things in life there's always an amount of something that comes across as "the best bang for the buck". So in reality there is probably an acceptable degree of bluntness that even racers and sailplanes can live with. I would suggest that something like a 20 to 30 thou flat on a 10 inch chord would be sharp enough for most folks. So 0.2 to 0.3% chord?

But if you have a big fan up front and top speed or efficiency isn't an issue then it doesn't really matter does it.....

Jimmbbo

I recall reading somewhere in the dim past that a truncated trailing edge actually can reduce drag since the TE will shed vortices alternately from the upper and lower sides and tend to keep the upstream flow less turbulent, if not attached... I also recall that this effect is Reynolds Number dependent.....

Does this ring a bell with anyone??

Cheers!

Jim

Tall Paul

That "shedding vortices" is the usual rationalization for the almost rabid defense of the blunt trailing edge on -everything-!
Howsomever, blunt edgers appear to seldom try the sharp edge.
I have many planes with blunt edges, and sharp edges, and it isn't the shape that creates problems, it's a poor installation of the surface control system beginning with the servo and ending at the hinges, and excessive weight of the surface that makes a surface flutter.
For instance, a 1/4" sq. trailing edge spar has much more mass aft of the hinge than a 1/4" strip tapered to a sharp edge. Mass aft of the hinge=bad!
Stiff and streamlined is the better option.

banktoturn

Jim,

A body with a blunt trailing edge will shed vortices as you describe, over a range of Reynold's numbers. Below and above that range of Reynold's numbers, the vortices will not be shed. The shed vortices do not reduce drag, however.

banktoturn

Tall Paul

Mentioning shed vortices brought this aeronautical oddity to mind...
The Klein-Folgelman airfoil uses a step in the upper or lower surface at 60% chord to change the smooth profile at that point.
The step changes the surface contour from the normal surface to the zero-chord line for the trailing part of the surface.
The claim is that the air tumbing over the step acts as a pusher to the wing, and adds thrust .
There is some theory and wind tunnel testing which has been done, and I found a reference to a home-built single-plane Sonnaire which had this profile.
Dan Santich of "Hots" & "Contender" fame built a Contender with the shape.
Unfortunately his flight results are too subjective.... "feels good", etc..
This image shows the wing I built for my long suffering Goldberg Mirage test article.
50"x9". Uses the NACA 23102 airfoil, with the K-F step at 60% on the bottom.
OS 10 power.
It didn't distinguish itself in flight, appearing no better or worse than the standard Mirage wing profile which comes from the Gentle Lady.
A telling commentary on the full-scale use is after lots of testing and attempts to make it work; the "safety" aspect of the profile got the most push. Apparently the shape while having more drag than a similar normal wing would have can have a slower stall speed.
As an economic incentive to use the thing, it didn't fly. Going someplace using more power than otherwise needed is NOT a sales point, regardless of the "safety" thing..
It reminds me of the idea that you can make a sailboat go faster by having a fan attached to the boat blow into the sail.