High Speed Low Aspect Ratio Wings

[mattuk]

My understanding of drag is that the induced drag decreases with speed and all other drag forces increase with speed.

Therefore when designing a model for high speed flight why is a high aspect ratio wing always used? The benefit of a high aspect ratio wing is that it will have lower drag at lower speeds or in a turn (This must be why they are used for pylon racing) but not in a straight line.

So for straight line speed surely the least important thing to consider is induced drag, and therefore a high aspect ratio wing is not needed. Using a lower aspect ratio wing would decrease some of the other types of drag and the increase in induced drag I assume would be negligible? Maybe I have got something completely wrong?! :confused:

So for straight line speed is a typical pylon racer the best choice? (assuming the straight line is long enough for the top speed to be reached)

Hopefully the graph below will help show what I am trying to describe

http://en.wikipedia.org/wiki/Parasitic_drag

Thanks

Matt

[vintage1]

High aspect ratio is still best, because if you want to cut down drag for a plane that only goes fast in a straight line, you reduce wing area..that shifts the wiki curve there up to the right towards higher speeds..

[BMatthews]

Actually for pure straight line speed you're right. A lower aspect ratio or even that delta planform is fine. It's only an issue when you go to turn and the speed bleed is far higher than with a higher aspect ratio.

But for pure straight line stuff the frontal and wetted surface areas are what forms much of the drag. Keep it thin and keep it small. Put the engine cylinder down in line with the wing behind it and fair in the back of the cylinder and then make that wing thin but with a proper airfoil (not the flat or diamond shape that many of the ARF deltas use) and it'll go like a scalded cat.

Using one fin only and making the center of the wing just barely thick enough for the gear to fit inside along with the fuel supply and you'll gain a little more. I'm awestruck that so many extreme speed prop delta models hang the control servos out in the breeze. The reason for only one fin is to avoid any gap drag from the fins being too close together. Also you then have 4 fin to wing joints (2 sides to each fin) and air creates a lot of drag in such corners. So it's better to use just one fin. Or two fins but at the tips so there's only one 90 corner per fin.

[vintage1]

I have to disagree.

In the end, removing the profile drag from the picture, you end up with - at a given high speed, and just enough lift to support the model - the simple equation of wing area, wing shape and angle of attack as inputs to your drag calculation.

Quite obviously your best bet is to use the wing at the optimal lift to drag ratio.

And select a section and planform that has the best lift to drag ratio. That's always going t be a slender beast until you approach transonic speeds.

In short, as speed increases, you dont need to change anything at all except WING AREA to get the slickest wing.

This actually shows up in turboprop airliners..they go fast, but not transonic, and they have long thin wings.

High speed deltas are deltas because those idiots who fly them want speed and ridiculous manoeuverability.

They are not as fast as pylon racers.

[BMatthews]

They are not as fast as pylon racers.

That part I don't doubt at all.

Hmmmmm.... for pure straight line speed perhaps we're missing the overall picture. Oddly enough this same question came up on another forum I'm on and I ended up doing a sketch for a swept flying wing as a possible speed model. Worst still I'd forgotten about having the exact same conversation there as with you here Vintage.

The idea of a flying wing actually sits in with your concept. If we make the wing and wetted area small enough for pure speed the span of a delta gets redicuously small. At that point prop torque could become an issue. And then there's the turns. Maybe not pylon tight turns but at around 200 mph even a big open turn from a modeller's perspective isn't that big and the G forces can add up.

Here's the concept I doodled out. A thin outer wing with a thicker area at the center line to provide room for the fuel tank cylinder fairing and radio gear. To maintain the airfoil shape at the center the stinger just happened but it instantly became a good place to mount the fin. The "fuselage" is the same airfoil as the outer panels but just increased slightly in % at the centerline. The blended body to fin helps avoid the draggy sharp square corner joints. And the odd broken leading edge scimitar shape had a reason as well. I'm just not sure I remember why now. I'll have to go back and review that other thread. Perhaps only because it provides a little more wing area while trying to at least partially avoid the strong tip vortices of a higher G turn.

Wha'dya all think of 'er?

[HugePanic]

looks like a me-163 to me.. :rolleyes:

[mattuk]

Vintage1,

Assuming you are correct, please explain where my understanding is wrong.

I will assume wing area is kept constant.

At high speeds induced drag becomes far less important than all other drag forces.
Therefore the obvious way of reducing drag is by reducing the other forms of drag such as form drag. This can be reduced by decreasing the aspect ratio.

Maybe the benefit of this is outweighed by the increase in induced drag???

Although another small point, looking at this equation, induced drag may not be increased as much as it would immediately appear

http://upload.wikimedia.org/math/5/f/0/5f0051a1f9fb9acfcca7cf4811138884.png
http://en.wikipedia.org/wiki/Induced_drag

On the top of the equation there is lift squared and on the bottom there is aspect ratio. Therefore it appears that weight is a very important factor in reducing induced drag.

I am sure you will agree a wing of lower aspect ratio can be built lighter than a high AR wing.

I am certainly not suggesting that the induced drag would be less on a low AR wing but surely it reduces the detrimental effects of increased induced drag on the overall drag making the other forms of drag even more important (in a straight line).


I am sure you are correct vintage1 as all the high speed planes seem to have high AR wings, but maybe this is because all the planes used for racing have to turn which makes induced drag the most important factor, but please explain why.

[BMatthews]

Vintage1,
.....Maybe the benefit of this is outweighed by the increase in induced drag???
.....

And that's the $64 answer. I'm not sure why Vintage wants to ignore the profile drag since at the very low Cl's of a speed model in level flight the profile drag becomes very significant. But the reality of flying our models is such that they can't fly dead straight and level for very long. Especially a REALLY fast model. So you HAVE to add in the need for turning and the speed bleed that a very low aspect ratio wing will produce. That's speed that then needs to be built up again once you level out. And if it bleeds too much you may never get back up to the final speed by the time you need to pull out or turn again.

Life is all about compromises and this is one of them. My concept may be too much the other way or it may not. Or there may be another concept that is best of all (very likey). But for the real world of model flying where our vision limits the time spent flying in a straight line a model that hits the best compromise between pure straight line speed and one that will turn decently will be best.

Now this doesn't mean that modern pylon racers are the be all and end all. They need to turn far tighter than a pure speed model. So a compromise between a delta with an aspect ratio of 1 or less and the pylon racers with up around 8 or 9 is likely going to hit a good compromise level. Where that best point ends up would require some fancy calculations and some equally fancy piloting to operate within the limits set during the design phase.

[HugePanic]

i just guessed some AR of airplanes that were pretty fast at their time.



BUT: all of these aircraft are designed for ma>1!

medium AR: Bell X-1 (http://www.aviation-central.com/famous/images/ab1gd-x1.jpg)

low-AR: X-15 (http://upload.wikimedia.org/wikipedia/commons/thumb/9/9c/X_15_2.jpg/800px-X_15_2.jpg)

medium-AR: F104 (http://www.fas.org/nuke/guide/usa/airdef/f-104-top.jpg)

i cant remember a fast plane with a high AR.

these planes were build for high speed and not for good turning.

i just found this pretty fast aircraft: X-2 (http://www.uwgb.edu/DutchS/GRAPHIC0/Spacecraft/x-2.gif)
i think this plane was designed also for turns...

at high speeds, the ca goes to zero.
so the drag of the wing and fuselage dominates.

if you don't change the fuselage, you have to find the wing with the smallest drag. (i don't mean induced drag!!)

i could imagine that a high AR wing with a higher span (and area perpendicular to the airflow) creates more drag whan a low-AR wing.
but i don't know!!

maybe someone else collects some formulas and solves them to get drag over AR. :o

[BMatthews]

You're looking at full sized stuff there. The operating regime is far different from our needs as a modeler. Also you've picked planes that are all supersonic. That changes things a lot as well.

Getting a little closer to "home" clipping the wings of Reno unlimiteds has been a time honored tradition. But why they can get away with this is because the planes are no longer expected to carry the weight of the guns and ammo (a very considerable total I might add) and don't need to carry fuel for long distance patrols. So the need for the wing area is reduced. Clipping the wings reduces the area and profile drag while raising the induced drag in the turns. Still, for them it is an advantagious compromise. Again, life and model design is about compromises and balance.

For now I'm with Vintage on the idea that a squat span delta is very likely not the best overall shape for an extreme speed model for the reasons noted above. But my gut feeling is that we don't need a relatively high aspect ratio if we take away the need for pylon turns. Likley something more in the 3 to 6 range will prove to be optimum.

Oh, by the way. I remembered why we ended up with the kink in the outer panesl of the Swift Concept. My first idea is shown below. The idea of a swept wing being a good planform is that a degree to degree and a half of washout is put into the wing so the tips provide stability by lifting less positive than the root. by reducing the lift coefficient at the tips the idea is that the size and power of the tip vortices will be reduced so there will be less tip losses. But then some bright wag brought up the idea of the sweep angle inducing spanwise airflow to the tips thus slamming into the tip fins and producing high drag. So I countered with the stinger fin and kinked tips to try to catch and work with this spanwise flow and squash the other guys protest......

I should'a been one o' dem science guys, eh? Or maybe a politician? :D

The original concept sketch....

[biber]

Some things limiting the optimum AR are Reynolds numbers combined with little wanted surface area, structural issues with slender wings, rollrate (which is limited by the span/airspeed ratio, rather than the AR itself).
Sweep does not help speed unless you aim for mach > 0.7 or so.
On fullsize aircraft that is designed for less than that you hardly find any sweep.
Only if the CG mandates it or controllability and aeromechanical stability on a tailless design require it, you might see sweep there.

Deltas are even worse than just sweep, they don't support laminar flow too well and that really hurts speed a lot.
They only have the advantage of being robust, structurally simple and able to withstand very high speeds without disintergation and no rocket science required for building and operating them.

biber

[MCarlton]

On this subject, it is interesting to look at control line speed models. Years ago, small span, low AR models were the way to go, but more recent developments in composite structures allow a far higher AR. So that is what current CL speed models have.

Here's my rubbish sketch for an RC speed model.

[BMatthews]

On this subject, it is interesting to look at control line speed models. Years ago, small span, low AR models were the way to go, but more recent developments in composite structures allow a far higher AR. So that is what current CL speed models have.......

Ah, but again we're talking apples and oranges. If you look at the current crop of speed models, and especially those used for F2A control line speed you'll find some odd looking models that use much larger inboard wings or in some cases even no outboard wings at all with a hugely long inboard wing. This is done for one reason only. It's to fair in as much of the control line wire as possible.

It's based on the idea that a round "rod" or CL wire has the same drag as a decent shaped airfoil that is 10 times as thick. By fairing in the last 3 feet or so (in the case of the more extreme single sided wings) they are fairing in a large portion of the wire line that is travelling the fastest of all. So in this case they chose the high aspect ratio for a reason far removed from induced drag or other considerations.

[JetPlaneFlyer]

Does the practice of clipping the wing tips of full size racers point us in the right direction :confused: ... I'm thinking of Reno racers racers such as the F8F 'Rare Bear' (http://www.rarebear.com/2007/art/pages.php?directory=.&currentPic=49) and the like. They all have a fair chunk of wing tip cropped off which reduces AR considerably.

Of course it could be that the desired effect was to simply reduce wing area and that cropping the tips was the easiest method to achieve this and that reduced AR was a by-product?.. Regardless; reducing the AR clearly does not slow these aircraft down ;)

Steve

[biber]

JPF, you just gave the answer yourself.

They are existing designs, that are getting modified.
New designs, dedicated to the task might look different.

biber

[JetPlaneFlyer]

They are existing designs, that are getting modified.
New designs, dedicated to the task might look different.

Biber,
I'm sure a built from scratch racer would look a little different but the point still ramains that the fastest piston/prop driven aircraft uses low AR wing and in modifying it to go faster they lowered the AR further. Also if you do look at the full size Goodyear/Formula One race class aircraft (which are custom made for the task) they too generally use medium or low AR.

I'm certainly not saying that low AR is always better but quite clearly in this instance it's not a significant handicap otherwise the Rare Bear would have become slower when they chopped the tips off ;) .

I'd guess that the points made earlier are correct.. At very high straight line speed the vast majority of drag is coming from profile drag and 'wetted surface area' friction. Induced drag will be largely insignificant so AR is relatively unimportant therefore other design considerations dominate.

Steve

[BMatthews]

.... Also if you do look at the full size Goodyear/Formula One race class aircraft (which are custom made for the task) they too generally use medium or low AR.....

Some of the older designs like the Shoestring, Cassut, Rivets, Ol' Tiger, et al had fairly low to medium aspect ratios. However the new trend is to composite wings and high aspect ratios. I've personally seed lots of such designs at Reno the last two seasons. But again these are high G turning racers in a a fairly tight closed course.

[Brandano]

Remember also that while a low AR wing will bleed more speed in a turn, it will make it less likely for the plane to stall out of it. Low AR wings usually have better stall behavior when compared with high AR wings.

[vintage1]

Dynamic soarers probably hold the record for outright RC speed: They are all about energy conservation. They do not use stubby wings.

[BMatthews]

Dynamic soarers probably hold the record for outright RC speed: They are all about energy conservation. They do not use stubby wings.

But they also pull incredible G loads in the turns or loops they perform in cutting across the shear lines. At higher lift coefficients like that there's no doubt at all that a higher aspect ratio is going to be more suitable.

It may well be that we tend to see all the fast planes with higher aspect ratios because the ones we see are all intended to fly in ways that need to deal with high lift coefficients due to high G loaded turns. Or in the case of the control line speed models because it's sort of a rules bending to fair in more of the draggy control line wire. But if we remove the need to lift well in a high G load then it makes sense that the optimum planform for this new style of flying would be different from the optimum planform for a pylon racer. How much it shifts it would need some number crunching that I'm going to avoid though.... :D

[HugePanic]

maybe the best design is virtually NO wing at all.

the X-16's wing is just to land the plane. (my theory)

at high speeds the fuselage alone can provide enough lift easily.
you just need something to attach the ailerons and to control the plane.

just take a look at rocket weapons. they really have small wings with a small AR.
and not all of them fly supersonic only. (but i don't think that makes a big difference here, if it's for supersonic or not)
therea are off course many other things that have to be considerd here, but they still have low-AR wings.

Meteor (http://de.wikipedia.org/wiki/Bild:Meteor_%28Luft-Luft-Rakete%29.jpg)

mfg
me

[Cory]

You really can't compare things like the Meteor or other supersonic weapons. The lift needs and supersonic drag issues are totally different.

I agree with vintage's assesment.

Look at the Tomahawk and ALCM cruise missles. They are not supersonic, need to fly fairly straight for long range, and, sort of like vintage was saying, they don't have short stubby low AR wings but have moderate ARs. Why wouldn't they have high ARs? The military certainly wants them to have all the range they possibly can. One of the design criteria, however, is that the wings need to fold and stow. This would limit the length of the wings, so the only way to get enough wing area is to increase cord a bit.

One of the few modern design Reno racers, the Pond Racer, had a fairly high AR wing: http://www.rcgroups.com/forums/attachment.php?attachmentid=1524718 but as mentioned earlier, it does need to turn.

Also, look at some of the straight wing business jets, such as the Cessna Citation. They are fast, but not fast enough to need sweep. They do not have low AR wings. Of course, they need to land at a reasonable speed and in a reasonably short distance.

The point is, all aircraft design is a compromise. I don't think you will find any aircraft made with only the goal of lowest possible drag in the 200-500ish mph speed range. The plane will either need to land slow, turn better than a battleship, stow the wings, fit in a hangar, etc.

Even when designing an RC plane that is optimized for straight line speed, it's ability to keep it's speed in the turn will make a large difference on the max speed of the plane. The speed that the plane begins it's speed run with will help enable a higher max speed. You don't want to accelerate, get to the edge of your vision, bleed gobs of speed off in the turn, then try to accelerate back the other direction. Nobody has good enough eyesight to make a looooong straight speed run. Maybe if you were designing an RPV fly through a speed trap for a mile or more with a nice long run in to the course you wouldn't care about turn speed. Otherwise, you should.

If you need more proof that vintage's explanation is in the right ballpark, look at the "Fastest Foam Plane Challenge" thread. The current leader, with a top speed around 126 mph does not have a low AR like this: http://www.rcgroups.com/forums/attachment.php?attachmentid=1856992
or this: http://www.rcgroups.com/forums/attachment.php?attachmentid=1883427
It has a pretty high AR: http://www.rcgroups.com/forums/attachment.php?attachmentid=1853885

That body is pudgy to fit the required power system in the most aerodynamic shape possible, but the wing is sure a lot closer to a DS plane's planform than it is to all of the low AR ideas: http://www.rcgroups.com/forums/attachment.php?attachmentid=1956250

I actually find it interesting how close the AR of that little foam prop plane is to that of the Dynamic 160 DSer.

[SteveC68]

I followed the Fastest Foam Plane Challenge thread so I'm familiar with the plane that is the current record holder. My question would be this. All else being equal what would happen to his speed if the span were increased? What would happen if it were decreased? By the way his plane is capable of ROG take-offs.

Steve

[JetPlaneFlyer]

What would happen if it were decreased?

If you cropped off the tips (in exactly the same way as the warbird Reno Racers) then I believe it would be a little faster in straight line top speed (Just like it does on the Reno racers)... BUT this would also increase the stall speed which could make the model impractical to take off and land. Conversly edding 'extentions' to increase the span would slow it down.

[Cory]

My question would be this. All else being equal what would happen to his speed if the span were increased? What would happen if it were decreased? Steve

The answer would depend on how quickly the power plant can accellerate the plane. Since it's such a small plane, I'm certain that it will benefit from beginning the measured pass at the highest speed possible.

[vintage1]

The more I read here and think about it, the more convinced I am that long and thin is better.

Basically you need a certain wing area to keep the model up at all..in practice you will gear that wing area to the slowest speed and highest G you need to pull. It mustn't stall under high G, and it mustn't stall when landing.

Given the wing area, what aspect ratio is the lowest drag? high aspect ratio elliptical tips.

Cropping wings only helps if you have more wing area than you need, and its a darned sight simpler than a complete re-design of the wing.

It's no use looking at warbirds: they had manoeuverability and high G turns and reasonable theater duration built in as a design requirement. Not ultimate top speed.

The only full size planes optimised for low drag and high speed that do not fly close to the sound barrier are the slower executive jets and turboprop airliners. None have stubby wings. None have sweepback either.

Probably he place to look for the fastest pre-war stuff is the Schneider trophy racers. All reasonably thin wings.

[biber]

Steve, wing clipping may make a faster topspeed, but you'd get that aswell keeping the span the same but making a higher AR wing.
It's all about throwing away any wetted surface that is not really needed.
Of course the most easy thing to do is to cut off some wingtips and that's why you see that done.

biber

[vintage1]

They did precisely that on one of the Spitfires..the MKV..'clipped, clapped, cropped'..extra power and the shortened wing made it pretty good at low altitudes, but it suffered at higher.

Always a tradeoff..

[Cory]

One of the main reasons for clipping wings on fighters in WW2 was to increase roll rate.

[Brandano]

And perhaps to mitigate their tendency to spin as well. Elliptical wings can be nasty.

[MCarlton]

One of the main reasons for clipping wings on fighters in WW2 was to increase roll rate.

Also, consider that as the war progressed and combat requirements changed, role profiles changed, it was quicker, easier and cheaper to modify and tweak an existing design than try to design mission specific aircraft, especially in the later stages of the WW2.

I'm with V1 on this one I think, I have always been of the opinion that lowest drag comes from higher AR. If you could make really high AR wings, with really thin sections, but make them strong enough to withstand all the loads imposed on them, I believe that wing would be faster (given the same fus/engine/tail etc) than a low AR wing of similar area.


Here's the rub though

One area where low AR does have an advantage (hence delta sport models) is that you can keep the section % thickness down, whilst having a deep enough wing for radio gear, decent spars, undercarriage mounting and the like. Given that section % is probably more important than outright thickness in terms of drag reduction.

By my reckoning, a 1" thick wing, of 20" root chord (typical delta speed model?) has a depth/chord ratio of 5%.

In theory, should that not have less drag than a 0.5" thick wing of 5" chord, that being a 10% thick section?

[MarkusN]

In theory, should that not have less drag than a 0.5" thick wing of 5" chord, that being a 10% thick section?It does, but only marginally so. In sound flow the influence of relative thickness on profile drag is surprisingly low. Basically it comes down to a little bit more wetted surface and slightly thickened boundary layer at the trailing edge. Also: two dimensional profile theory is not valid for massively swept and tapered wings.

The delta, however, has to deal with massively three dimensional flow and separations at very moderate AOA. In terms of wing efficiency the delta is lousy.

[MCarlton]

That makes sense. Thats sort of what was going through my mind, but not in very accurate terms! So I'll stick to my theory that high AR is the way to go for low drag and thus high speed.

Remembering that we are dealing with fairly low Re and fairly low speeds, are speed models designed with swept wings a bit of a red herring? My understanding is that sweep is something which is principally introduced to assist with transonic and supersonic conditions?

Thus, I think a straight spar (ie at 90 degrees to the fus datum viewed from above) with a high AR and a swept tip would be the way to go (like FAI Pylon racers for instance). Does the limiting factor then become one of structure and wing loading? Ie can one build a high AR wing strong enough, and can one keep the wing loading down given the reduced wing area?

Looking at Joe Manor's Dynamic 160 for example, would it have the potential to be faster if it were possible to build enough strength into a wing with double the AR, and keep the loading down to a reasonable level?

[HugePanic]

the argument with the eliptcal high-AR wings sounds logikal.

but the original question had the asumption a veryvery low cl. at this point there should not be much induced drag at any wing.

what about the profile-drag of a high-AR wing??

[MarkusN]

Does the limiting factor then become one of structure and wing loading? That, and Re. Even in speed tasks we are still operating at Reynolds numbers where profile drag changes significantly with Re. So there's definitely an optimum between high AR (low induced drag) and large chord (low profile drag) (Always looking at wings of identical area.)

[Cory]

But, like I essentially said before, low cl flight should not be the only goal if one wants the highest top speed for a model plane. Because of the limits of human vision, the plane needs to turn reasonably efficiently to keep the plane in view. Along the lines of this, I believe that the reasons the RC turbine planes are not as fast as the DS gliders is because the gliders are built to turn well and the turbine planes aren't. In addition to that, the DS flight profile keeps the plane in view, and the turbine planes usually try their speed runs in straight lines. Take a plane like the Dynamic 160, put a sufficiently powerful tubine in it, fly a big circle around the pilot at the limits of pilot vision minus enough cushion for emergencies, and a tubine powered RC plane could go faster than the DSers.

[MCarlton]

I assume that optimum is also an ever changing value, hence birds of prey, especially peregrine, hobby, sparrowhawk et al, have the ability to change wing chord, sweep, area, profile to optimise the wing continuously.

[mnowell129]

If you don't have to turn and don't compromise for T/O and landing,
my vote is low AR.
At high speeds induced drag is a don't care, parasite (form and skin)
drag are everything. The lower AR has the higher reynolds number and lowest form drag. Everything else being equal, low AR should be faster, given all the caveats that you have enough wing area and thrust to fly at low Cl.

I suspect the DS planes would be a lot slower if they had the turbine intakes and/or fuselage dimensions of the turbines. Having a free external power supply lets you go a long way in minimizing drag on the DS's.

[harveyweewax]

One of the main reasons for clipping wings on fighters in WW2 was to increase roll rate.

The book "The Spitfire Story" summarizes a report of a comparative trial between clipped wing and a standard wing Spitfire VBs. The sole difference between the aircraft was that the wing span was reduced from 36 ft 10 in to 32 ft 6 in and the wing area was reduced from 242 square feet to 231 square feet. The removal of the wing tips resulted in a weight reduction of about 30 lbs.

The results of the tests showed that the clipped wing spit was slightly faster (about 5 mph) than the standard spit at altitudes 10,000 ft or lower. Between 15,000-20,000 ft there was no speed difference and above 20,000 ft the standard spitfire was slightly faster. There was no difference in rate of climb at lower altitudes, however, the difference in time during zoom climbs from 20,000 to 25,000 feet was 15 seconds in favor of the standard spitfire.
At all altitudes the clipped wing spitfire had a much better rate of roll, better acceleration and was better (faster) in a dive but, the turning circle increased by about 55 feet at 20,000 feet altitude. There was no difference detected in take off or landing

Regards,

Sven

[Brandano]

But was it also a bit easier to fly? As the war drags on, the number of skilled and trained pilots available tends to drop down. And these are harder to replace than a plane, whatever the plane price might be. Eventually wars die off because there's no people left to fight them.

[vintage1]

I don't think it was easier to fly at all. The spitty was nearing the end of its development at that time, and the war requirements were changing: the longer range Mustangs and P-38's were being used as bomber escorts, and the new generation of fighter bombers like the Typhoon, Tempest, and the Thunderbolt were coming along by then: massive planes of enormous power, equipped with devastating cannons, and capable of very high speed: the 'dog fight' had all but disappeared..speed and rate of climb and ceiling were as important as anything.

WWII really saw the end of the defensive dogfighter as we know it..todays fighters are multi-role. They have to balance manouverability, range, ceiling, rate of climb, speed and weapons load quite delicately.

Essentially you cant outrun an AAM, but you may be able to outurn it...massive AA fire remains the cheapest and best way to defend against low level attack, not fighters..stand-off missiles fired from high altitude stealth platforms counter that..and SAM type attacks..

And the helicopter has replaced the STOL type aircraft.

Speed simply is not an issue, except with businessmen flying the pond..and they scrapped concorde anyway.

I can't remember what the X-planes finally achieved..4500mph? something like that..and everyone said 'so what' and took their weapons payloads into low space, where they could come in at frightening speeds and almost no defense as ballistic missiles.

Hi tech isn't everything, as 911 demonstrated. Frankly you could probably shoot an airliner down with a Gloster Gladiator..if you had to..and all the hi tech in the world doesn't help you hold ground that you have wiped a military force from..its a funny old world these days, being fought over with car bombs, IED's, suicide bombers AK47s and machetes..

[harveyweewax]

But was it also a bit easier to fly? As the war drags on, the number of skilled and trained pilots available tends to drop down. And these are harder to replace than a plane, whatever the plane price might be. Eventually wars die off because there's no people left to fight them.

In 1942 (when the clipping occured) the Spitfire Mark V was found to be inferior to the FW190, which was just entering service. The FW190 could out-run, out-climb, out-dive and out-roll the Spitfire Mark V. To match the FW190, Supermarine developed the Spitfire Mark IX and the later Griffon engined marks. However; as a stop gap measure the wing tips on the Mark V were removed which evened the odds a bit. This was done solely to increase role rate which improved its combat peformance. There is no mention that clipped wing Spitfire was any easier to fly than the standard spit. If anything, the reports indicate that the later Spitfire marks were more difficult to fly than the early marks.

Regards,

Sven

[MCarlton]

If anything, the reports indicate that the later Spitfire marks were more difficult to fly than the early marks.

Remember of course that the very late marks were twice the weight of the earlier ones. Yes, a lot more power, but wing loading had gone up a lot. An elliptical wing with narrow tips with high wing loading could be a pretty snappy beast.

Its worth noting also that the FW190 originally had a smaller wing than the production variants, but this was scrapped due to handling difficulties and lack of load carrying ability within decent wing loading limits, so there is, as ever, a compromise to be struck.

An interesting point on the subject in hand, the highest performance model of the FW190 legacy was the Ta152 H-0. Take a look at the wing, it is a long, long, high AR wing. Which was faster than the smaller winged FW190D-9 with similar power. Albeit at high altitudes.

[Cory]

Yeah, and that Ta-152 had a top speed in the 470's, which is pretty darn fast for 1945, with that wing that has a much higher AR that a P-51.

[SteveC68]

Cory,

Please post sources for that performance. I've NEVER seen a claim that high and I've got tons of books on that plane. Also keep in mind the role of the Ta-152H. High altitude interceptor. It did have good performance up high, but down low it did not. At 10k ft alt it was around 360 at military power and almost 390 at emergency power. The P-51D was at 400mph at the same alt using military power and almost 410 using emergency power. In fact it wasn't until the thin for the day 25k that the Ta started to come into its own. The P-51 was better at low alt and equal all the way up to around 26k. Not bad for a low AR wing.

[HugePanic]

i am no expert in WWII birds, but i just something interesting at wikipedia.
(i am not claiming that wiki is always right!!)

to the Ta-152 (http://en.wikipedia.org/wiki/Focke-Wulf_Ta_152):
Performance: Maximum speed: 759 km/h at 12,500 m using GM-1 boost (472 mph at 41,000 ft using GM-1 boost)
.....
....
The Ta 152H was among the fastest piston-engined fighters of the war, capable of speeds up to 755 km/h (472 mph) at 13,500 m (41,000 feet, using the GM-1 boost) and 560 km/h (350 mph) at sea level (using the MW-50 boost). To help it attain this speed it used the MW 50 water-methanol injection system mainly for lower altitudes (up to about 10,000 m or 32,800 ft) and the GM-1 nitrous oxide injection system for higher altitudes, although both systems could be engaged at the same time. The Ta 152 was one of the first aircraft specifically designed to employ a nitrous oxide power boost system.

In late 1944, Kurt Tank reported that while flying an unarmed Ta 152H to a meeting at the Focke-Wulf plant in Cottbus, he saw four P-51 Mustangs. He made his escape by engaging the MW 50 boost, opening the throttle wide to gain maximum speed to escape the enemy fighters, and left the four Mustangs behind him. There is no evidence from Allied reports that these P-51s ever saw him.
.....


P51: (http://en.wikipedia.org/wiki/North_American_P-51)
Performance
Maximum speed: 437 mph (703 km/h) at 25,000 ft (7,620 m)
Cruise speed: 362 mph (580 km/h)


that's just what if've found...

[MCarlton]

From Gordon Swanboroughs Book;

Ta152H-1 Junkers Jumo 231E-1 with GM 1 NOS - Max Speed 472mph at 41,010ft

Ta152 V7 (DB603EM - 2,250HP) with MW50 Injection 370mph at sea level.

Note, Junkers Jumo engine also added 485lb residual thrust from exhaust

FW190 V16 450mph at 22,966ft

Not bad for a low AR wing

True, but the P-51 did have a (supposedly) laminar flow wing section.

The low AR wing in itself did not make the P-51 equal in pace to the 152, the engine in the 152 was designed to deliver its power above 26,000ft, low down, where the air is more dense, the lower drag of the laminar flow section would have been more of an influence.

Low level aircraft tend to have a lower AR and a higher wing loading, to keep stability up and negate the effects of low level turbulence, thus giving a more stable, comfortable and more accurate ride.

[SteveC68]

Uh wikipedia isn't a source. Anyone can post anything on wiki.

MCarlton,

You do know that only 3 H-1's were produced? Page 114 of Dietmar Harmann's book Focke-Wulf Ta152. Reason being the GM 1 did not work well when combined with the MW50. All other H's were H-0's without the GM 1. Also the DB603 was NOT put into the Ta-152H it was in the short span 11.0meter (lower aspect ratio 6.2) Ta152C. It was the Jumo engine in all of the H's.

http://img525.imageshack.us/img525/6526/ta152h0xr9.th.jpg (http://img525.imageshack.us/my.php?image=ta152h0xr9.jpg)

Notice there are 3 steps towards the upper right of the graph. Those represent the 3 different GM 1 (nitrous oxide) injection sizes that were tried. The highest value is 415ps (415 horse power). This was experimental and not production. Note the date on the bottom of the chart. 4.9.43. and this was aircraft number 152.000-001. Also even with a colossal 415hp boost notice the increase in speed was less than 10kph. And last notice that the highest speed was 735kph (456mph) at almost 12.5 km (41000 feet). These are actual German flight test reports. I trust them more than any other 'source'. These are also published in Harmann's book.

Anyway I digress the entire point of the discussion is all else being equal what is better high or low aspect ratio for top speed. So if you had the exact same airplane and all you did was start trimming the wings would the plane be faster or slower. Reading through all of the newer posts I've come to the conclusion that as long as the wing is creating enough lift to actually fly then the lower aspect ratio will be faster. This says nothing about low speed handling or anything like that which I didn't see mentioned in the opening posters original question.

Steve

[MarkusN]

So if you had the exact same airplane and all you did was start trimming the wings would the plane be faster or slower. Reading through all of the newer posts I've come to the conclusion that as long as the wing is creating enough lift to actually fly then the lower aspect ratio will be faster.
Yeah, but the point then actually becomes smaller wing area being faster.

If you compare wings of equal area and Re is sound, then, marginal as induced drag at high speed may be, high AR still has an edge.

[Cory]

Sorry Steve, I don't recall sources, other than my first exposure to this plane was through a DML 1/72 scale plastic model kit. The instruction sheet listed the top speed as being in the 470's, but I forget exactly what. Maybe one day I'll look around and find some of the other places I've seen such high speeds for it. I do know that I have never seen a top speed for the TA-152 H-1 published at less than 472, until your posting. I'm not talking internet blogs or wikis. My interest in this plane began way before the internet was common. I don't doubt that your source is good, but I've seen conflicting info in good sources from time to time.

Markus or anyone else, can you name a plane that was designed for lowest drag in a subsonic (probably ought to make that sub-transonic) speed range, and didn't have other design criteria that would shorten the wing? I'm talking a plane designed with one goal in mind, max speed, but yet was constrained by limited thrust which would keep it below a speed where compressability would start to matter. No concern for landing speed, roll rate, or modifying the wing to accommodate such things as retractable landing gear, guns, fuel tanks, etc. Just a plane that was made to go fast, with little concern for landing and take off distance? Remember, if you take a wing of X weight with Y wing area and clip the wings, your plane will have to fly at a higher angle of attack to maintain altitude at a speed of Z. This isn't a baiting/trolling question. I just can't think of anything built like that with a low AR wing, so I tend to think that there's a reason for it.

[MarkusN]

The Hughes Racer comes to mind. Existed in two span versions that could be selected depending on the task. But again the design principle was to take a design and adapt it to the faster, less enduring version by basically reducing wing area, not primarily AR.

[SteveC68]

The Pond racer is probably the most recent plane designed with a high AR wing however the engines never really produced their designed horsepower and it crashed (killing the pilot) before further speed attempts were made. It did however qualify for the 1991 Silver Unlimited class at ~400mph. Tsunami was another Unlimited class aircraft that used a merlin engine from a P-51 along with the motor mount, spinner and tailwheel. It was clocked at ~490mph and had a low AR wing.

http://members.tripod.com/~tsunam/tsunhist.html

MarkusN,

Please read the 1st post in the thread. The only constraint given to design was AR not wing area, hp or anything else. It was purely which is faster low or high AR wings? Based on that single design criteria I think it is low however there are plenty of very fast planes with high AR wings.

[MarkusN]

MarkusN,
Please read the 1st post in the thread. The only constraint given to design was AR not wing area, hp or anything else. It was purely which is faster low or high AR wings? Based on that single design criteria I think it is low however there are plenty of very fast planes with high AR wings.
Yup, but the conclusion was found pretty quickly that low AR designs usually resulted from other design constraints, often the easy mehtod of clipping wings to reduce their area. Something which probably was also behind the Tsunami's wing.

[MCarlton]

You do know that only 3 H-1's were produced? Page 114 of Dietmar Harmann's book Focke-Wulf Ta152. Reason being the GM 1 did not work well when combined with the MW50. All other H's were H-0's without the GM 1. Also the DB603 was NOT put into the Ta-152H it was in the short span 11.0meter (lower aspect ratio 6.2) Ta152C. It was the Jumo engine in all of the H's.


Nope, I didn't know that, so thanks for enlightening me.

So, the hypothetical test is this, and perhaps someone with a wind tunnel could test this out.

Lets take a really average .40 powered sport model, with a normal lowish AR wing. Ok, lets also make another wing, with say, 50% higher AR, but extended span to keep the wing loading the same (or as near as possible).

Then, fly the model at a pre-determined throttle setting, past a radar gun, without a dive for speed or anything, just max cruise speed.

That should answer the question, is anyone up for the challenge?

[harveyweewax]

Markus or anyone else, can you name a plane that was designed for lowest drag in a subsonic (probably ought to make that sub-transonic) speed range, and didn't have other design criteria that would shorten the wing? I'm talking a plane designed with one goal in mind, max speed, but yet was constrained by limited thrust which would keep it below a speed where compressability would start to matter. No concern for landing speed, roll rate, or modifying the wing to accommodate such things as retractable landing gear, guns, fuel tanks, etc. Just a plane that was made to go fast, with little concern for landing and take off distance?

Check out the Me209. In 1939 it set the world speed record with a top speed of around 470mph.

http://en.wikipedia.org/wiki/Me_209

Interestingly enough, about the same time Supermarine was going for the speed record with the Speed Spitfire which was a ordinary spitfire cleaned up and fitted with smaller wings an a more powerful engine.

http://www.jaapteeuwen.com/ww2aircraft/html%20pages/SUPERMARINE%20HIGH-SPEED%20SPITFIRE.htm

From what I understand the British abandoned the attempt when they found about the Me209.

Regards,

Sven

[SteveC68]

MCarlton,

I would say take a model with a high AR and do a few speed passes to get an average top speed. Then start hacking away at the wing and keep making passes after each hack job getting an average speed at each span. Keep doing this until either the average speed stops changing or the plane becomes unable to fly.

Steve

[MCarlton]

Thats probably a better way to do it. I think there will be a sort of "sweet spot" where decreasing AR further starts to slow things down again.

Interestingly enough, about the same time Supermarine was going for the speed record with the Speed Spitfire which was a ordinary spitfire cleaned up and fitted with smaller wings an a more powerful engine.


True, but aerodynamic theory has moved on a bit since then, since the advent of construction techniques allowing higher AR, more wing section research and the like.

can you name a plane that was designed for lowest drag in a subsonic (probably ought to make that sub-transonic) speed range

The problem with that is that after the advent of supersonic aircraft, jet engines and the like, there was little research into subsonic speeds. Designers had already managed to get into the low transonic zone, by just having a lot of engine power and lowish drag (ie Spiteful, P-38, P-47M etc) and going high, where obviously Mach 1 is at a lower airspeed.

Oh, and according to NASA;

The drag coefficient equation will apply to any object if we properly match flow conditions. If we are considering an aircraft, we can think of the drag coefficient as being composed of two main components; a basic drag coefficient which includes the effects of skin friction and shape (form), and an additional drag coefficient related to the lift of the aircraft. This additional source of drag is called the induced drag and it is produced at the wing tips due to aircraft lift. Because of pressure differences above and below the wing, the air on the bottom of the wing is drawn onto the top near the wing tips. This creates a swirling flow which changes the effective angle of attack along the wing and "induces" a drag on the wing. The induced drag coefficient Cdi is equal to the square of the lift coefficient Cl divided by the quantity: pi (3.14159) times the aspect ratio AR times an efficiency factor e.

Cdi = (Cl^2) / (pi * AR * e)

The aspect ratio is the square of the span s divided by the wing area A.

AR = s^2 / A

For a rectangular wing this reduces to the ratio of the span to the chord. Long, slender, high aspect ratio wings have lower induced drag than short, thick, low aspect ratio wings. Lifting line theory shows that the optimum (lowest) induced drag occurs for an elliptic distribution of lift from tip to tip. The efficiency factor e is equal to 1.0 for an elliptic distribution and is some value less than 1.0 for any other lift distribution. A typical value for e for a rectangular wing is .70 . The outstanding aerodynamic performance of the British Spitfire of World War II is partially attributable to its elliptic shaped wing which gave the aircraft a very low amount of induced drag. The total drag coefficient Cd is equal to the drag coefficient at zero lift Cdo plus the induced drag coefficient Cdi.

Cd = Cdo + Cdi

The drag coefficient in this equation uses the wing area for the reference area. Otherwise, we could not add it to the square of the lift coefficient, which is also based on the wing area.

[SteveC68]

A few more points about a modern low AR full size plane:

Cost of R&D vs. profit from said project. Honestly who is going to plunk down serious money for something like this and are there enough of those people that you would get a return on your investment? I think seriously the best modern example of a ground up modern design would be the Nemesis NXT designed by John Sharp. This is a plane designed to do one thing and that is to go very fast.

http://www.nemesisnxt.com/kit/index.php


Another good example would be the Lancair aircraft. Specifically the IV and the Legacy. Both of which have proved very fast at Reno.

None of these planes would I consider to have a high AR wing.

Another point to keep in mind is that in a full size aircraft the wings typically have to house things like fuel tanks and landing gear. Because of those requirements high AR wings don't really work that well.

One thing that should be defined though is what actually is a high AR wing? Where is the line drawn where one can say everything above this number is high AR and everything below is low AR?

[MCarlton]

Thats a good point Steve, where does one make a distinction?

The Nemesis is not particularly what I would call low AR, yes, the wing is small in comparison to the fus, but if you cover the fus up, it isn't actually that low in AR.

In this scenario, we are talking about powered aircraft, thus

Low AR to me is something around the 4:1 - 6:1 area

High AR is anything over around 12:1

Another good example would be the Lancair aircraft. Specifically the IV and the Legacy. Both of which have proved very fast at Reno

This is over a tight closed course, so roll rate becomes an issue, as does low level stability.

The problem with this conundrum is that it is difficult to get emprirical data regarding the drag/speed difference between high AR and low AR when all other factors are equal, Because all other factors cannot be equal, due to high and low AR having other characteristics which materially affect performance, whilst not being directly related to the relative induced drag differences caused by the AR alone.

This goes back to the fact that aircraft design is a balancing act between conflicting compromises, so no aircraft can ever be designed purely for one performance criteria. Even the Nemesis NXT design is compromised, by virtue of having to incorporate sufficient structural strength, light enough weight, big enough payload/fuel load, cowling, engine and everything else which imposes a limitation on the purity of the design from the outright speed standpoint.

I may be totally wrong on this, but I do not believe that simply making AR lower will yield an increase in speed, because the aerodynamic theories covering AR and induced drag just don't support it. If induced drag is lower with higher AR, then;

Assume wing area, wing section and all other factors to be theoretically equal,

Lower induced drag from a higher AR wing MUST yield a faster aircraft for a given throttle setting.

Thus there must be a plottable curve of AR vs Induced Drag, which I would expect to be shaped roughly like the attached sketch, and I would guess that you would find that most powered models and full size GA aircraft would fit within a fairly small distribution around the centre, with high performance sailplanes towards the bottom LH corner and very low AR "fun fly" models and even WIGE craft at the other end.

My diagram is not very good and is based on supposition and what feels right.

[SteveC68]

I think that the opening poster's point was that when you lowered the AR you also lowered frontal area which has a greater impact in drag as speed increases vs. induced drag. That was the main reason behind my thought process on just taking a high AR wing and start hacking the tips off in stages and getting speed measurements. Adding the constraint/requirement that wing area remains constant would add an additional level of complexity/difficulty in performing this experiment. This was also a contributing factor in my idea.

Steve

[vintage1]

The Lockheed U-2 was one aircraft optimised for low drag subsonic cruise..
As is any airliner.

Up to about 400mph slender wings are the way to go: above that some sweepback seems advantageous - presumably to prevent localised transonic airflow.

[mnowell129]

Assume wing area, wing section and all other factors to be theoretically equal,

Lower induced drag from a higher AR wing MUST yield a faster aircraft for a given throttle setting.


I disagree. This is a good conclusion based on a errored assumption. At very low Cl, i.e. very high speed flight, the induced drag goes to 0. Thus the induced drag of a high AR and low AR wing are both the same, that is 0. The drag is then dominated by form drag and skin drag. The form drag is a function of the frontal area, so for a given wing area the low AR has less frontal area, thus less drag. Also the lower AR has a longer chord, thus a higher reynolds number, and thus more laminar flow for longer along the chord, thus lower drag. Thus all things being equal for highest speed flight the low AR will be faster.
A caveat is that the wing loading is such that the high speed flight is at sufficiently low Cl.

[MarkusN]

I would say take a model with a high AR and do a few speed passes to get an average top speed. Then start hacking away at the wing and keep making passes after each hack job getting an average speed at each span. Keep doing this until either the average speed stops changing or the plane becomes unable to fly.
That's again going the very same route that favors low AR wings for high speed: Taking a design constraint (the original wing) and then optimising wing area with AR as a byproduct, instead of AR by itself.

[MarkusN]

At very low Cl, i.e. very high speed flight, the induced drag goes to 0. Thus the induced drag of a high AR and low AR wing are both the same, that is 0.
Nope. It becomes small, but not zero. The drag is then dominated by profile drag.

The form drag is a function of the frontal area, so for a given wing area the low AR has less frontal area, thus less drag.Even if you want to attribute the elusive "form drag" to a wing (this is usually a term used for draggy bodies with lots of turbulence in their wake), given that wing area and relative thickness are the same, frontal area is also the same.

Also the lower AR has a longer chord, thus a higher reynolds number, and thus more laminar flow for longer along the chord, thus lower drag. Thus all things being equal for highest speed flight the low AR will be faster.That depends a whole lot on the Re range you are in. For high Re it's actually the opposite. On a high Re wing transition is further upstream. As sinking Re approaches critical we run into laminar separation problems, which are a can of worms in their own right. However, not usually a problem in engine powered models because of the speeds involved and the transition forced by vibration. You are right, however, that drag usually decreases with growing Re. (Thats because relative thickness of boundary layer decreases) That's why I said that there is an optimum AR somewhere between induced drag and profile drag. Increase of drag with lowering Re is more prominent at lower Re, so in the case of an RC model the optimum is narrower.

[HugePanic]

can we simplify the model by reducing it to a flat-plate-wing with zero lift???

there should be far may data available if we do so...

[vintage1]

I disagree. This is a good conclusion based on a errored assumption. At very low Cl, i.e. very high speed flight, the induced drag goes to 0. Thus the induced drag of a high AR and low AR wing are both the same, that is 0. The drag is then dominated by form drag and skin drag. The form drag is a function of the frontal area, so for a given wing area the low AR has less frontal area, thus less drag. Also the lower AR has a longer chord, thus a higher reynolds number, and thus more laminar flow for longer along the chord, thus lower drag. Thus all things being equal for highest speed flight the low AR will be faster.
A caveat is that the wing loading is such that the high speed flight is at sufficiently low Cl.

The induced drag cannot go to zero except in a vertical dive, where no wings at all is nbets.

The induced drag in any plane using its wing to fly, is the weight divided by the lift to drag ratio.

Ergo, what is needed is a wing that has the lowest lift to drag ratio at the designed airspeed. That is a slender wing, up to transonic regimes..

[pmackenzie]

The form drag is a function of the frontal area, so for a given wing area the low AR has less frontal area, thus less drag.

For a given airfoil section and wing area, frontal area is independant of AR.
It will always be wing area x % thickness

Pat MacKenzie

[MarkusN]

The induced drag in any plane using its wing to fly, is the weight divided by the lift to drag ratio.
Nitpicking, but quite significant here: Drag, not induced drag. Of which induced drag is a part, but a small one at small CL.

I do agree with your conclusion, however.

[mnowell129]

Nope. It becomes small, but not zero. The drag is then dominated by profile drag.

I should have said "goes to 0". But you make my point, the drag is dominated by profile drag ( the combination of form and parasite drag, thus ...)


Even if you want to attribute the elusive "form drag" to a wing (this is usually a term used for draggy bodies with lots of turbulence in their wake), given that wing area and relative thickness are the same, frontal area is also the same.

Wings have form drag and parasite drag, the combination of which is usually referred to a profile drag. Form drag comes from integrating the total pressure around the wing in the drag direction. Form drag is what the wing makes when it's at Cl=0, excluding parasite drag. And no the frontal area of a long wing or a short wing of equal thickness is not the same. Wetted area is the same, frontal area is not.

[MarkusN]

And no the frontal area of a long wing or a short wing of equal thickness is not the same. Wetted area is the same, frontal area is not.Yes it is. I said RELATIVE thickness. See also Pat above.

[pmackenzie]

And no the frontal area of a long wing or a short wing of equal thickness is not the same. Wetted area is the same, frontal area is not.

Note that I said equal percentage thickness.
If you still disagree, please show your work :)


Here is mine:
Wing area is the span wise integral of :

chord(s) * ds

Frontal area will be the span wise integral of the thickness which will be:


thickness(s) * ds

but thickness(s) = chord(s)* % thickness,

substitute and rearange to get :

%thickness * chord(s) *ds

Basic calculus tells you that you can move the constant %thickness outside of the integral.
Therfore frontal area is always exactly the wing are times the thickness in percent.

[Brandano]

I don't actually think induced drag decreases with speed. I think it stays constant, since it's only proportional to the amount of lift produced by the wing, and this in level flight always equals the weight of the plane. Vintage, if I might add something to your "vertical dive" statement, there should be also no induced drag in a vertical climb, or any other flight attitude where there is no lift produced by the wing, like a parabolic 0g flight.

[MarkusN]

I don't actually think induced drag decreases with speed. I think it stays constant, since it's only proportional to the amount of lift produced by the wing, and this in level flight always equals the weight of the plane.In absolute values you are right. Other drags increase with the square of the speed however, so contribution of ID to total drag decreases significantly.

[vintage1]

I don't actually think induced drag decreases with speed. I think it stays constant, since it's only proportional to the amount of lift produced by the wing, and this in level flight always equals the weight of the plane. Vintage, if I might add something to your "vertical dive" statement, there should be also no induced drag in a vertical climb, or any other flight attitude where there is no lift produced by the wing, like a parabolic 0g flight.

Now this is getting close to what I feel are the salient issues. For a constant L/D the induced drag is a constant..indeed.

Which means that the best high speed wing will always be one with the best L/D ratio. It just gets smaller, the faster you go, so that you do end up operating at the best L/D ratio.

In short the issue resolves not to profile drag or anything else - that is simply there whatever else you do - its 'what is the best planform for the highest L/D ratio at 200mph' or whatever.

This is where my desire to get things right meets the boundary of my sphere of competence..I would GUESS that the wing would be very thin, maybe only 5%-10%, operate at a low angle of attack, and be very slender as well, probably with elliptical, or using other means, to reduce tip vortices.

i.e pretty much the way a fast carbon glider looks. And as little fuselage area as possible...

What we need is teh way in which L/D varies with chord thickness and angle of incidence and airsp[eed, to get the section right, and a way in which planform affects it.

Someone probably has the graphs to hand: I don't. I don't attempt to buiid high performance models: Just nice to fly ones.

[mnowell129]

I don't actually think induced drag decreases with speed. I think it stays constant, since it's only proportional to the amount of lift produced by the wing, and this in level flight always equals the weight of the plane.
Cl lowers with speed. L = cl * v*v * area, etc.
For example if L = 1, v = 1 cl = 1, if you double speed, v = 2, L still
equals 1, v*v = 4, thus cl = .25
If Cdi = k* Cl * Cl, when v = 1 Cdi = k *1 *1 = k*1, When v = 2 Cdi =k * .0625

let X = 1/2 * rho * S
Then total drag, D ~= Cdo * v * v * x + Cdi * v * v * x
From the above example
D at v = 1 = Cd0 *1 * x + 1 * 1 * x
D at v = 2 = Cd0 * 4 * x + .0625 * 4 * x

I think the point is that Cdi goes down fast with speed, and once it gets down to the order of Cd0, Cd0 starts to dominate quickly.

[mnowell129]

Note that I said equal percentage thickness.
If you still disagree, please show your work :)
.
Agree based on % thickness.
I was assuming same absolute thickness. I didn't feel constrained to hold the percentage thickness constant when changing AR.
I was considering the basic question of what "planform" gives lowest drag for a high speed plane. So all I was holding constant was wing area.
Based on this assumption and only being concerned with highest straight speed with no compromises, I would choose a low aspect ratio planform with thin section over a high aspect ratio planform with the same absolute thickness.

[vintage1]

Agree based on % thickness.
I was assuming same absolute thickness. I didn't feel constrained to hold the percentage thickness constant when changing AR.
I was considering the basic question of what "planform" gives lowest drag for a high speed plane. So all I was holding constant was wing area.
Based on this assumption and only being concerned with highest straight speed with no compromises, I would choose a low aspect ratio planform with thin section over a high aspect ratio planform with the same absolute thickness.

So every turboprop airliner manufacturer has in fact got their sums wrong?

They should indeed be flying Lockheed Starfighter shaped aircraft..

Hmm. ;)

[mnowell129]

So every turboprop airliner manufacturer has in fact got their sums wrong?

They should indeed be flying Lockheed Starfighter shaped aircraft..

Hmm. ;)
You've ignored my basic caveats and tried to put words in my mouth that I didn't speak. My assumption is no compromises to takeoff, landing, turning or other necessary conveniences for other purposes. Just a pure speed plane. I.E. what is the lowest drag for a craft designed specifically for no other purpose other than to go as fast as possible in a straight line. Then you end up with F104 shapes.

Mission profiles for turboprop airliners include many different flight phases, takeoff/landing distance, ROC, engine out performance, fuel economy, etc. etc.

[harveyweewax]

My assumption is no compromises to takeoff, landing, turning or other necessary conveniences for other purposes. Just a pure speed plane. I.E. what is the lowest drag for a craft designed specifically for no other purpose other than to go as fast as possible in a straight line.


Well, if that's the design criteria then I would design a small plane with a big engine and wings small in area that would only provide the minimum amount of lift necessary. But a wing small in area doesn't imply that I would need to use a wing with a low aspect ratio. Correct me if I am wrong but a high aspect ratio wing is a more efficient lift producer then a low aspect ratio wing. Then, if that assumption is correct, then the high aspect ratio wing would be smaller in area than the equivelant low aspect ratio wing and less draggier.

Regards,

Sven

Regards,

Sven

[vintage1]

You've ignored my basic caveats and tried to put words in my mouth that I didn't speak. My assumption is no compromises to takeoff, landing, turning or other necessary conveniences for other purposes. Just a pure speed plane.

That is precisely what an airliner is. It seldom turns, has a pathetic rate of climb, uses several miles of runway to get aloft..

I.E. what is the lowest drag for a craft designed specifically for no other purpose other than to go as fast as possible in a straight line. Then you end up with F104 shapes.


Only if you wish to go trans/supersonic.

Mission profiles for turboprop airliners include many different flight phases, takeoff/landing distance, ROC, engine out performance, fuel economy, etc. etc.

No, the overriding factors in airliners are fuel efficiency, speed and range. A few other issues like maintenance cost and structural strength are there also, but by and large you want a slick airframe that will go as fast as possible on the least fuel, so you get more pasenger miles per interests rate repayment.. Turboprops are more efficient than jets fuel wise, but are slow. That means on longer hauls they take too long, and they cant do the flight times. Therefore jets are used up to the point at which drag starts to increase sharply in the transonic region.

Once into the transonic region, all bets are off., Its a completely different ball game altogether.

[HugePanic]

i just made 2 calculations:

aircraft weight: 1000kg
roh = 1.225
v = 220m/s

Wing surface: 10 m^2
wing thickness 0.1m

I made two wings: one with AR=10, and one with AR=1.
(One wing with 10x1m, one wing with 3.2x3.2m)

In the result the AR=1-Wing has less thant half the combined drag (D = Df + Di) of the AR=10-Wing

Can anybody calculate with the sama data???
(I can post my calculation-sheet later, but i want to wait for verification of my result first)

thx

[mnowell129]

HugePanic,
Thank you, thank you, thank you.
What value did you use for Cd0 (or Cdf) ?
I'll see if I can come up with a similar answer.

[HugePanic]

HugePanic,
Thank you, thank you, thank you.
What value did you use for Cd0 (or Cdf) ?
I'll see if I can come up with a similar answer.

my cd0 is called cf (form). assumed 0.03 for that.

edit: i just double-checked --> 0.03 is maybe too less, 0.05 seems to be better. but doesn't change the result too much...

[vintage1]

i just made 2 calculations:

aircraft weight: 1000kg
roh = 1.225
v = 220m/s
x
At that speed (500mph) the wing is into transonic airflow over parts of it: all bets are off.

The accepted wisdom is that up to about 350 mph a straight high aspect ratio wings is the lowest drag: then a swept wing into the transonic region, and then a highly swept or delta configuration for super and hypersonic flight.

Also, there is no particular reason to keep the wing thickness the same for high and low aspect ratio wings. For speed you probably want a very thin section whatever the aspect ratio, 10% or less.

If you consider parasitic drag of two wings of equal area, one of high aspect ratio, and one of low, but of equal percentage thickness to chord you will see that not only is the surface ('wetted') area the same on both cases, but for a constant percentage chord thickness (let's leave taper of the depth out) the frontal area is also identical.

Ergo the parasitic drag of any wing of a given area, and identical aerofoil profile is the same whatever the aspect ratio.

Which means the only way to reduce drag is a thinner section, to reduce parasitic drag, or a higher aspect ratio, to reduce induced drag.

In practice both lead to structural problems, increasing aircraft weight, so there is a practical limit.

Above around 350 mph, transonic effects start to become relevant: efforts have to be made to reduce soickwave formation along the span and leading edge (where maximum airflow acceleration takes place) ..


These issues have been well worked out for years. at slow speeds and light weight a very slender wing works very well, as evinced by todays sailplanes. As you get faster, the very slender high aspect ratio wings get very hard to make to be capable of withstanding loads, and wings tend to come down in aspect ratio: about mach 0.7 or so, sweepback is introduced, and beyond mach1, the shapes are very weird indeed.

[mnowell129]

At that speed (500mph) the wing is into transonic airflow over parts of it: all bets are off.

Speed of sound at sea level is 340 m/sec. 220/340 = .65 mach, well below critical mach. So this is valid incompressible subsonic calculation.

[mnowell129]

Which means the only way to reduce drag is a thinner section, to reduce parasitic drag, or a higher aspect ratio, to reduce induced drag.


Thinning the section reduces form drag, not parasite drag.
Profile drag is form drag plus parasite drag.

[vintage1]

form drag is part of parasitic drag. According to the articles I perused.

[mnowell129]

form drag is part of parasitic drag. According to the articles I perused.
Form drag and parasite drag are both separate components of profile drag.
Form drag is the drag due the pressure on the airfoil in the drag direction. Parasite drag is skin friction due to shearing friction in the boundary layer.
Profile drag is the drag the airfoil is making when it is not producing any lift, hence the drag coefficient for this component is Cd0 (Coefficient of drag @ lift = 0). It's hard to determine how much of the profile drag is form drag and how much is parasite, so it's generally easier to just use the combined profile drag.
You can see the profile drag directly on L/D plots for airfoils since the Cd does not go to zero when Cl = 0.

[MarkusN]

That calculation is very much in favor of the low AR wing as it assumes same absolute thickness of the wing. Not a common scenario at all. Sure, a low AR wing can be built with less relative thickness, given identical wing loading.

Then, calculating wing drag with assumption of the same Cd based on frontal area can not yield reliable results. Cd would be much higher for a wing section with small relative thickness if based on frontal area (as most of it is skin friction.) There is a reason that cd/cl data for wings are usually normalized to wing area.

[mnowell129]

I'm assuming non-dimensionalizing by wing area, not frontal area.
And I'll restate my basic assumption, that is, that the constant is wing area and what is allowed to vary is AR and thickness. If I'm making a plane to see how fast I can go with no other compromises to takeoff,climb,landing,turning etc. I will choose a thin low aspect ratio wing over a high aspect ratio wing with the same absolute (not %age) thickness, because as speeds go higher and Cl (and thus Cdi) tends toward 0, profile drag (Cd0) will dominate.
Structurally a low AR wing is easier to make than a high AR wing of the same relative % thickness, so the low AR wing can be both thinner (in absolute terms) and lighter tending to increase the speed capability.

[harveyweewax]

This a quote from a Wikipedia article on wing aspect ratio so take it for what it is worth:

"While high aspect wings create less induced drag, they have greater parasite drag, (drag due to shape, frontal area, and friction). This is because, for an equal wing area, the average chord (length in the direction of wind travel over the wing) is smaller. Due to the effects of Reynolds Number, the value of the section-drag coefficient is an inverse logarithmic function of the characteristic length of the surface, which means that, even if two wings of the same area are flying at equal speeds and using equal angles of attack, the section drag coefficient is slightly higher on the wing with the smaller chord. However, this variation is very small when compared to the variation in induced drag with changing wingspan (for example, Airplane Aerodynamics by Dommasch/Sherby/Connolly, published by Pitman Publishing in 1961, calculated on page 128 the ratio of a specific airfoil contour (NACA 23012) at typical lift coefficients: Section Drag Ratio = (chord ratio)^0.129, which means that a 20 percent increase in chord length would decrease the section drag coefficient by 2.4 percent)."

Well, I tied to post a link to the article but I get a bad link :mad: . Anyway, If you are curious do a google search for wing aspect ratio and it should pull up.

Regards,

Sven

[vintage1]

What that means is that the less lift you need to generate, the less wing you need :D

Seriously, until the airfoil IS working at 2.4% induced, and the rest parasitic, there is not much to be gained: if you are at that level anyway, you should simply be making the wing a lot smaller..

[mnowell129]

What that means is that the less lift you need to generate, the less wing you need :D

Seriously, until the airfoil IS working at 2.4% induced, and the rest parasitic, there is not much to be gained: if you are at that level anyway, you should simply be making the wing a lot smaller..
Note entirely true. When you lower the area you must raise the Cl to match increasing the induced drag. You have to consider the whole equation Drag = Cd * 1/2 * rho * v * v * AREA and Lift = Cl * 1/2 * rho * v * v * AREA. Where Cd ~= Cd0 + k * Cl * Cl, where k = some function of AR. Where k*Cl*Cl = Cdi.
Or Drag = (Cd0 + (k/AR)*Cl*Cl) * 1/2 * rho * v * v * area.
The 2.4% number was a relative reduction between two section coefficients due to reynolds number, not the difference due to frontal area, your conclusion based on that number is in error. You need to keep the wing area to keep the Cl low so the Cdi doesn't creep back in. If you make the wing area too small then you must fly at a high Cl, thus high induced drag, so the design direction is to keep enough area to keep the Cl low, but do so at a low enough AR to keep the frontal area low, this is the design optimization for maximum speed.

[admodesi]

Speed of sound at sea level is 340 m/sec. 220/340 = .65 mach, well below critical mach. So this is valid incompressible subsonic calculation.

Ummm the speed of sound (in air) is not affected by pressure or altitude, it is only a factor of temperature. Strange but true! Look it up. In a standard atmosphere (at sea level) at 15 deg C the speed of sound is 661 knots or 1224 kph or 765mph. In the same standard atmosphere the temperature at 38,000 feet is -57 deg C making the speed of sound 572 knots or 1060 kph or 663 mph.

The formula for speed of sound is: 38.98 x square root of (273 + local air temp C) = the answer in knots (because that what full-size aircraft have on their speedos! - and the all above is off the top of my head so correct me if I'm wrong!)

This formula is great to figure out how far away lightning is when you time the difference between the flash and the thunder.

So back on the theory of wings ...just because an aircraft's pitot tube may be indicating a subsonic speed it does not mean that there is no shockwaves forming near on or around the aircraft. At airspeeds well below super-sonic there could be small and weak shock waves forming in front of the nose, at the max thickness point on the top surface of the wing, or around the cockpit and in other similar spots. When these shockwaves occur the plane is said to be traveling at “transonic speeds”. These shock waves form when the local airflow is accelerated to the speed of sound when it is forced to flow quickly around an object. Exactly the same type of acceleration that provides the lift when it flows around a wing. When the aerodynamics of a plane is poor (in a super/transonic sense) these shockwaves will form at a lower speed. So the "Critcal Mach" is extremely dependent on the shape. At higher speeds shockwaves also form a sort of bow-wave in front of the object. At and beyond supersonic speeds you will find 2 shockwaves on a wing. One connected to the leading edge and one connected to the trailing edge. The idea behind designs like a F-104 Starfighter is to contain the entire aircraft within the conical shaped shockwave created by the pitot tube at the extreme front of the aircraft. At speeds beyond Mach 2 the shockwave formed is an extremely acute cone. The F-104 wings are short to fit inside this cone. At subsonic and even transonic speeds I see no benefits what-so-ever to wings with an aspect ratio as small as an F-104. Besides the Lockheed U-2 is a transonic aircraft! Both the U-2 and the F-104 were designed by that aerodynamic genius Clarence Kelly Johnson, and they both have the same basic fuselage designs, however it is a great comparison of how wing aspect ratio affects an aircraft’s flight envelope!

But as we are talking about model aircraft here how the hell are you going to accelerate it up to a speed where transonic problems are going to occur? A propeller propelled aircraft just is not going to do it. A model jet will probably burn all its fuel in the attempt! The only model planes currently exploring anywhere near transonic speeds are the Dynamic Soarers. Oh and guess what, they have high aspect ratio wings!

Keith

[vintage1]

A propeller propelled aircraft just is not going to do it. A model jet will probably burn all its fuel in the attempt! The only model planes currently exploring anywhere near transonic speeds are the Dynamic Soarers. Oh and guess what, they have high aspect ratio wings!

Keith

prop driven aircraft in the second world war experienced transonic issues in dives.

Prop tips can easily go transonic.

[admodesi]

Yes, but it is the extra thrust of gravity that was getting it there. As we both know the thrust drops off and the drag goes up with a prop as we explore the realms of Subsonic Mach Numbers.

[mnowell129]

Ummm the speed of sound (in air) is not affected by pressure or altitude, it is only a factor of temperature.
The speed of sound is usually calculated using the standard atmosphere which provides a temperature profile vs altitude, hence the speed of sound at sea level implies a given temperature. Also the speed of sound varies with density, which does vary with altitude.

[admodesi]

The speed of sound is usually calculated using the standard atmosphere which provides a temperature profile vs altitude, hence the speed of sound at sea level implies a given temperature. Also the speed of sound varies with density, which does vary with altitude.

Are you saying that if I go to into the Stratosphere (using the "international standard atmosphere" as the model) where between roughly 10 to 20 km above the earth that the speed of sound will be different? So I ask you this will the speed of sound be different from say 12km up compared to 18km up?

PS: NO HELP from the others especially you Vintage1 this question is for mnowell129 only!

[mnowell129]

Oooh Goodie a quiz, to prove myself! I love it when I get little quizzes to prove my worth, especially when you put me down by saying that Vin1 is smarter than me and already knows the answer.
No, it doesn't. And it only takes 5 minutes to go remind myself where the temp inversion layer is....
Any other condescending personal insults thinly disguised as questions ?

[admodesi]

Oooh Goodie a quiz, to prove myself! I love it when I get little quizzes to prove my worth, especially when you put me down by saying that Vin1 is smarter than me and already knows the answer.
No, it doesn't. And it only takes 5 minutes to go remind myself where the temp inversion layer is....
Any other condescending personal insults thinly disguised as questions ?YES!
...5 minutes? Had to look it up on google did we? but you said:The speed of sound is usually calculated using the standard atmosphere which provides a temperature profile vs altitude, hence the speed of sound at sea level implies a given temperature. Also the speed of sound varies with density, which does vary with altitude.Please explain? Why doesn't it now? It did before didn't it?

And yes it was a loaded question. My added PS made that quite obvious! Just like in a court of law if you get the witness to admitt they were mistaken then it throws the rest their testamony in to doubt. Welcome to the court of public opinion! :D