Discussion in 'Aircraft Design / Aerodynamics / New Technology' started by gschuld, Nov 7, 2009.
kinda looks like the airfoil used on a p-51 mustang....
Deltas are unique animals. Due to the long centerline chord and very low operational lift coefficient, I think you'll find that the Rn and shape affected effects might reduce the extent of the laminar flow however, the benefit of a good shape is still there. The only difference is that deltas, and their lifting capabilities, aren't as affected by airfoil shape as more conventional wings are. One good example of this is the story I relate in my airfoil article, which regards the wing development for the A-4 Skyhawk.
I remember reading that. I suppose one thing that defeats the laminar flow on delta wings is the strong spanwise flow? I'm looking forward to drawing my own airfoils but I saved the one from your delta just in case.
if im understanding this correctly, the argument against the proposed airfoil is the fact that to get the desired low drag of the airfoil is to have the trialing edge to tilt up 5 degrees? but what about the Cl vaules, sufficent? stall carcteristics? just to denie the airfoil because of the trailing edge seems to be shallow at least? sorry for spelling, im on my phone.
Sufficient for what? The big points being made here are:
1) That you need to choose an airfoil suited to your particular design and specification, not slavishly copy what someone else has used.
2) There are very many more important speed-increasing mods you could make than simply changing the airfoil.
As for the NLF-0414f, Cl values have little meaning until they're applied to a wing, and it's the CL (big 'L', meaning wing) which really counts. Same with stall characteristics and drag characteristics. The reason I personally don't like reflex flaps is that they require mixers and other additional weight/complexity in the control system that you could avoid just by choosing an airfoil better suited to the purpose. Trust me, you're not going to be able to tell the difference in this airplane between 60% laminar flow and 70% laminar flow, assuming that you even built a wing accurate enough to get either of those percentages in the first place. We're talking about scratch-building a wing to an accuracy of a very few hundredths of an inch over the entire wing area, top and bottom. That's a more than two hundred square feet of curved surface, every single square inch accurate to the numbers on the plans to about the width of a human hair or two. Are you prepared to do that? Are you experienced enough to do that? If you don't achieve that kind of accuracy, you can specify the most low-drag airfoil in the world and it won't make a whit of difference: the airfoil on your airplane won't be the one that's giving those lovely low numbers in the Cd curve.
The Lancair 235 is already a good sportplane, so why change anything at all? Want to make it faster? Work on cowling shape, cooling drag, prop selection, finding/eliminating areas of airflow separation, and reducing air leaks between the cowling/cockpit/fuselage interior/landing gear wells and the outside airflow. Gap-seal all the control surfaces. Make sure the canopy and gear-doors fit so well you can't slip a business card into them and you can't feel the transition between the fuselage and the part that opens. You'll get far more improvement there than by changing the wing airfoil.
Klaus Savier has been mentioned once or twice here. Fact is, he got the vast majority of his improvements to his aircraft by exactly the kind of things I mentioned above. Just making sure the existing airplane was the best it could be. I believe he also spent some time accurately 'profiling' the flying surfaces in the sailplane sense: making sure they're conforming to the desired airfoil as accurately as is humanly possible, with a minimum of waviness and distortion. If you're building the Lancair from the original kit, odds favor that the wings have warped/crept themselves out of spec over the years to the point that carefully profiling the wing to original spec will gain you more than simply slapping another wing on the thing. If it's the completed aircraft you're showing at the beginning of this thread, then a serious re-profiling is in order. Composites creep over time, even the oven-cured pre-molded ones. Competition sailplane pilots have their wings reprofiled at the start of every season. It's still easier than building a whole new wing, and you'll get most of the improvement you might get from a new airfoil without having to build a new wing.
I don't want anyone to think that I have lost interest in the thread's subject matter. Far from it.
And yes, I have been studying up on every way to improve a small aircraft's efficiency that I can find, from the most obvious(cooling drag) to the smaller stuff (gap seals, excresence drag).
I'm just looking at airfoils and wondering if a higher percentage laminar flow wing could potentially be seen as a speed improvement for a fairly uncompromising(racing) setup. People sometime do radical things for that extra 5kts. I'm curious, there are a good number of Lancairs competing in the sport class division. They all came with 0215(f) airfoil wings. With all the mods these guys have done to them(like the Aerochia racing mods) the "reno wrap" etc, have any of them changed their wing to a different airfoil shape?
The comment about extra complexity for negative flap deflection is strange, there's no difference between flaps than can go from 0 to 45 degrees deflection and those that go from -5 to 40 degrees deflection.
I know a couple of serious competition pilots (European champion, 1st and 2nd at world championships) and serious competition pilots don't reprofile their wings. Ever.
That's only done on some of the older gliders (especially LS-3 and ASW-20 are notirous IIRC) where the spar is directly attached to the outer sandwich panel. Hence most new gliders are built with a sandwich core, even over the spar caps and only need a new skin if the gelcoat turns bad. That usually happens after 10-30 years.
Some ballpark blending isn't going to help you. Even on optimized gliders it's extremely hard to achieve laminar flow over anything except the straight wing.
A blended design (with also a lot of crossflow) as yours requires probably endless hours of detailed CFD and wind tunnel analysis and endless improvements since laminar flow is completely non-linear. Also having the requirement that it should behave relatively well (no abrupt stall, predictable reaction to minor turbulence and contamination) makes if very unlikely that a laminar foil is going to help you in any way
The norm for composite airplanes is laminar flow foils, and I indicated that in my last note. Let's not even get into turbulent sections... Within the realm of laminar foils there is a fair sized range in drag reduction. So, I would think that we would be chinning the 0414 against the best of the general purpose laminar flow foils.
I just compliled some data from TOWS and GAA on our current laminar flow foils...
So if you wanted to use Harry's soft stall laminar foils and minimize drag, you benefit 9 counts out of 50 for moving max thickness 10%. That is as calculated using Profil, which has its errors... or 18% improvement for 10% movement in max thickness.
So here you get .0045 at max depth of 35% and 0.0034 with max depth of 50% or about 25% improvement for 15% movement of max thickness. You are into diminishing returns. Extra length of laminar run is getting less benefit for each increment.
So, let's say you get another 6% reduction over the best NACA foils and 31% over the poorest NACA foils. Let's look at that result even closer. At BL0 on the wing (the only place on a 3D wing where the wind tunnel data actually can agree with reality), we can get 6 to 31%. Unfortunately, we have a fuselage in the way for the first 15 to 25 inches either side of BL0, a tractor prop is chopping up the flow out to BL60 or so, and 3-D effects whittle away benefit all of the way to the tip, plus all of the other reality effects. In truth, you would be hard pressed to get half of that 6-31% reduction in drag. So, if you get 3-15% reduction in wing drag - is that going to be significant? Well, if the wing is a third of the airplane's drag, that means you get a 1-5% reduction in real drag, or about 0.5-2.6% less power needed to fly at the same speed for that change. Or on the same power, you just got about a 0.3-1.7% increase in speed... Yeah, Nemesis would have gone more slowly with a 67000 series foil, but only a little. It dominated Formula One because they did a bunch of stuff as well as it had ever been done.
Now in sailplanes, with big aspect ratios and very low other airframe drag, they are likely to capture a bigger fraction of the improvements. At the top of a race game, maybe that 0.3% is significant. And in exchange, you get an airplane optimized for one place on the envelope at the expense of the rest of the envelope, are accepting sharp stall behaviour and most likely, poorer flapped wing performance. Not a good compromise in my book...
You will get far more speed from making sure that the interference drag, cooling drag, gaps at control surfaces, and excrescence drag is minimized.
Given the type of aircraft the OP is talking about, simply reflexing the flaps isn't going to do much (if any) good. You'd want to do the ailerons, too. That means mixers, with all the attendant complexity and weight I mentioned above.
Well, then the technology has changed for top-end gliders in the last few years. I have several friends (with admittedly 7-10 year old designs) who do reprofile their wings. One of them has his glider in the shop up at Crystal right now. Either way, it still doesn't change the original point: The Lancair 235 kit the OP is talking about is 25 years old, minimum, and it doesn't have the latest sailplane refinements. Many of the early molded composite kits (Lancair included) were notorious for arriving at the customer with some built-in warpage and distortion. A quarter-century of sitting probably hasn't improved that situation. The airfoil very likely isn't an NLF-0215(f) anymore, but rather an N-sorta-LF-xx(whatever) now. It's still easier to reprofile the wing than build a whole new one, and unless the OP is ready, willing, and able to build his new wing to competition sailplane tolerances, changing out the wing is likely to make things worse, not better.
And Billski's analysis above is quite right. This isn't a sailplane we're talking about, where a few percent of wing drag reduction might actually be measurable in terms of aircraft performance. A good third of this wing is buried in fuselage and turbulent prop blast.
I am really enjoying reading these post. Billski, your latest post was great. Thank you. Assuming your calculated(guestimated) numbers are fairly accurate, and they sure sound reasonable, it gives a great overall perspective.
Do others here have any reason to disagree with Billski's calculations?
Let's just assume, for a moment or two, that a whole new wing needed to be designed/built for a Lancair 200 kit. A totally clean sheet design, head to toe, and the resulting construction of this wing would be dead nuts perfect(competition sailplane standards accuracy/quality). Given the goals are to be a bit different than the original plane, that is, the plane would be geared much more heavily toward competition (with a 100-120hp engine), would anyone hazard a guess for an appropriate airfoil to replace the 0215(f)?
I'd think we'd need some more information to give you a real selection:
CLmax required, with what kind of flaps (if any)? At what Reynold's number?
How low a drag do you need to meet your performance specification - how much can you trade off for other parameters?
How much real-world suitability does this airfoil require? (sensitivity to dirt/rain/bugs, etc.) You've mentioned that you'd still be using the airplane as a personal sportplane, so this isn't a trivial consideration.
How are you tailoring the stall behavior? Will the wing shape be doing most of it or is the lift distribution near-elliptical for good energy retention through hard turns, and so the airfoil will need a soft stall on its own?
I personally could say, "Go with one of Riblett's 40-series 'foils" and you'd have a nice range to choose from, but that's pulling fluff out of the air, not a rational choice. Picking an individual airfoil comes at the end of wing aerodynamic design from what I'm finding, not at the beginning.
The most optimized glider profiles do perform considerably better which isn't strange if you know that there's 95 and 75% laminar flow at high speeds (below 150 kts and around 3 ft of chord)
I don't know which Reynolds numbers are used in Billski's numbers, but I'm pretty sure they're a factor 3 or so higher than the ones I mentioned above. I'm also quite sure that drag will rise a bit at higher Re, but whether that's 1 or 50%, I don't have the slightest clue.
That's not enough. To achieve laminar flow you need turbulators, specialized ailerons, flaps and especially seals of those two and the airplane has to be kept extremely clean. You also need very high quality of the gelcoat surface.
In doing that you can probably achieve laminar flow in the whole wing, except 1.5 times the propellor diameter in the middle.
In that section is also your biggest opportunity, an optimal fairing can give you a huge decrease in drag, probably more than you can achieve by going to the most optimal airfoil. That's a hard and complicated job however.
As for the best laminar airfoil, the best around are the DU profiles by Boermans, but you need to check whether they do work as well at your estimated speed and chord length.
Wittman Tailwind has the same equivalent drag area as Lancair Legasy FG. Retractable gear legasy has less drag only thanks to the retracred gear.
Wittman uses basically a conventional 4-digit NACA airfoil on a wooden wing and a simple tube and fabric fuselage. Makes 220-230mph on O-320. Outraces Longeze with the same engine.
Another plane, which has the same drag area is Dynaero MCR-01ULC/ This plane is all-composite, but the wing uses NACA23015 section. All these planes have very close wing areas. The clear winner is Wittman.
Sure, they claim that. According to pilot reports it's more in the 180 region in reality. That's about 80% more drag compared to the Lancair...
There are many sources confirming that Tailwind can fly faster than 200mph with O-320 engine/ For instance, CAFE report and Airventure Cup races results. Equivalent drag area of well-built W10 is 2.03 sq. ft. as determined at CAFE tests. One can find that this figure is very close to that od fixed gear Lancair and Glasair. The overall sporead is under 10%. Other high-performance planes with similar wing area differ mostly by their retracted gear. This proves that not much is dependent on the airfoil and the skin gloss.
According to the CAFE report they got 217 mph squized out of it, the Lancair Legacy 293.. Not to mention the Legacy has close to two times the range, space and useful load. If you make a fair comparison to planes that are actually comparable it's still considerably draggier, by dozens of percents as for example the Glasair IIFG (231 mph and still 40% more useful load)
As for that Vari-Eze, they do 258 mph... on 100 hp.
It proves? Except that your assumption and statements above are completely wrong, it doesn't prove anything.
Go read this
That's an optimized plane where the wing profile drag accounts for 33% of the total drag (fixed gear). Using the numbers for a turbulent airfoil - even an efficient one - raises total drag considerably by 20 or 30%. That's much.
As for the MCR-01, it has an awful small wing and huge fowler flaps. If they would've gone for a clean laminar wing about 50% bigger and simple flaps they would've had the same drag, only a simpler construction...
If we want to compare apples to apples, we have to look first at Lancair 320 against W10 as they have the same engine, O320. Moreover, the fuel burn at maximal speed flights by CAFE was the same, 11.8gph. Lancair 320 got 233mph TAS at 5975' densuty altitude against 217 of W10 at 8666'. The IAS values are 211 and 191mph, respectively. If we assume the engine power was equal in both cases (according the fuel flow the same engine type), the equivalent drag area of Lancair 320 should be 74% of that of W10. Lancair 320 has retractable gear and the drag diference agaiunst W10 is very close to the fixed gear drag.
Fixed gear Lancair legasy uses 180hp Lycoming O-350 or IO-360 and typically cruises at 210mph. Also Lancair has a bit smaller wing area than W10.
Another W10 Tailwind, N374WT got 224.43mph against 221.8mph of Longeze at 2007 Airventure Cup race.
Lancair Legacy tested by CAFE had not only retractable gear but 310HP TCM IO-550N engine and got 257mph IAS at that particular density altitude. This number is again in line with the ratio of the engine power together with very low SFC of fuel injected TCM engine.
Some comparative data on drag areas of high - performance two-seat planes are also given here:
Note that the drag area of W10 was accurately determined in CAFE tests by glide at controlled zero thrust. The same value was obtained for Wittman's own W8 in 1950s by propless glide tests arranged by August Raspett. Certainly not bad.
The higher useful load and corresponding longer range of Lancair and Glasair is accompanied by considerably heavier airframes and means nothing for drag area comparison.
The low drag area of Arnold Ar-5 is also a remarkable example of overall optimization as well as Tailwind. Ar-5 is a single place plane with small wing and also small fuselage wet area. For comparison, the wing profile drag portion in the equivalent drag area of a similarly optimal two seater plane with fixed gear is always less than 33% and can be as small as 20%. Indeed 33% is a good estimation of wing profile drag account of a retractable gear two-seater like a lancair.
It is easy to undrrstand that Vary-Eze cannot reach 258mph with stock O-200 engine and any ariframe mods. This requires even smaller drag area than that of Ar-5, which has about one square foot. The O-200 engine of Klaus Savier is highly modified and there is no disclosed information on its altitude performance, and it is known that O-200 performance can be highly boosted. Look at the Formual One racers.
MCR-01ULC has exactly the same wing are as W10 per plans. Its wing is well-optimized for both low cruise drag, low stall and acceptable weight. It is an european microlight plane which must have the stall speed under 40mph and takeoff weight limit is 991lb. The use of laminar arfoil would result in larger area due to lower CL and heavier airframe weight.
You're still comparing a high versus a low wing, a composite versus traditional, taildragger (fg) against trigear (fg).
You were the one to mention the Legacy that had similar drag to the Wittman. Though correct the Legacy is indeed twice a Wittman in size and weight and power..
Sure, but every engineer knows that's far from relevant since the prop flow completely alters the flow around your fuselage, a good 60% of the total drag..
You didn't do engineering did you? More range (payload), more room and more useful load do require more structures, volume and THUS also more wing and tail area. That all adds up in condiserably more drag.
Did anyone say that tailwind wasn't boosted too? No.
Further on, the AR-5 has no turbo, so it all depends on the altitude. Arnold estimated the AR-5 to put out 60 HP, that would require about 110 HP and Savier claims to have gained a few HP. Sounds very possible..
Again, you're clearly no engineer.
Let's assume that you're going to double the wing area and keep the aspect ratio constant. That will lower your wing weight (much thicker wing) As I said before, a laminar airfoil has considerably lower drag.
That's the very reason we're not only referring to airfoils with their pure Cd, but also to their (CLmax/CDmin) which is much more interesting. If you look at some laminar glider airfoils they have a maximum Cl of around 1.7 and a minimum Cd of 0.0025. The Naca 23015 normally does around 0.058 and has a similar Clmax. Even with double slotted fowler flaps Clmax of the Naca 23015 won't surpass 3. If you double your wing area your total drag is still lower with a laminar airfoil and you need only a slight increase in area, not a doubling.
Bottomline, by comparing apples to a submarine you're not going to prove that laminar airfoils don't do much.
Fact is that wing drag on highly optimized aircraft is a good 1/3rd of the total drag. Fact is too that laminar airfoils have considerably lower drag than conventinal ones, the very reason a complete glider has the same drag as a pair of Cessna 182 wheelpants...
A Diana II glider needs only 30 HP to fly by at 213 mph...
Now, who is a better engineer?
First, let us remember a definition of equivalent drag area. This is a sum of all drag components referred to respective areas, not including induced drag and effects of interaction of prop with airframe.
The equivalent drag area divided by total wet area is a mean drag coefficent, which can be used as a measure of overall design efficiency. This coefficient for a clean airplane without fixed gear and other considerable local drag
sources consists mostly of a friction drag. Other drag components to be taken into account are interference drag, trim and cooling drag.
They are affected by the plane layout (high, mid or low wing, tractor or pusher, etc).
Generally the equivalent drag area depends on the angle of attack, but for the performance comparison its value is determined at the AOA of maximal cruise speed at optimal altitude and mean weight. This is our case.
All the planes mentioned here except Varyeze and Ar-5 are very close by their dimensions, wing area and total wet area. This means the comparison of their equvalent drag areas is valuable.
Next, we can draw a conclusion of similar propulsive efficiency for the planes with the same engine installation, i.e. a tractor prop on an air-cooled horizontally opposed engine in an efficient cowl.
This arrangement is very efficient and produces near zero prop efficiency loss against free air operating congitions in our region of interest.
The maximal cruise speed of a high performance airplane is close to twice the best L/D airspeed and this means that the induced drag typically comprises just few percent of total drag. Moreover, the trim drag may have the same or higher contribution here.
Moreover, all aircraft I've used for comparison have no turbo.
If we'll remove the 1/3 of the total drag area of a plane (this means our wing has zero profile drag) and apply the same propulsive power, we'll see the airspeed increase by 14.5%. If we'll cut the profile drag component by half (assume we use a highly sophisticated laminar airfoil, which supposedly confirms its numbers), we'll get 6% airspeed increase. If we'll instead double the profile drag, for instance, use a four digit NACA airfoil with fabric cover in Ar-5, the speed will drop by 9%.
If we'll double the power, the speed will increase by 26% while at 50% power the speed will drop by 20%. A good aerodynamicist keeps these numbers firmly embedded in mind as well as many important others.
Finally, let as look at Diana II speed polar at maximal wing load. One can see this sailplane sinks at the rate of 600fpm at 138knots. The net required power for level flight at this speed is 20hp. It is the end of the curve, but one can extrapolate it to 30hp and get the estimated airspeed of 158kt or 182mph. It is certainly well below 213. And 30hp is here again the net power, or the prop efficiency is assumed 100%. Finally we can derive the equivalent drag area and Cdmin from these data. They are respectively 0.72 sg.ft. and 0.0078.
The comparison with Cessna wheelpants is highly hyperbolized. Instead this drag area is close to that of the whole Cessna fuselage.
The Cdmin = 0.0078 is actually 1/3 of that for W10 Taiwind. Indeed, the proportion of profile drag in total drag area of a high performance sailplane is exactly opposite ant it typically contributes to more than 2/3 of the total lift-independent drag. this means that profile drag Cd here is at least 0.005. May be it is 0.004, but certailnly not 0.0025. These numbers are consistent with other real world test data at similar conditions and represent a honest achievement for a designer. Sorry, I wore no pink glasses.
As for the Diana you're correct, my apologies, I took the minimum t/o weight and used the polar of the MTOW :speechles
Which is a significant number from my point of view.
That 30% drop in drag would happen with a composite or well-faired aluminium turbulent profile. A fabric cover is way more draggy.
But you used the wrong assumptions. A 15-meter glider typically has more than 50% of its non-induced drag from its fuselage, tail and interference drag, only the open class ships get to the 2/3rd and they do just. According to Boermans, designer of 80% of the now top sailplanes and according to Schempp-Hirth..
So no not only me, but Billski too is wrong? He posted the drag coefficient of the 671-215 profile to be .0033 and I think that's a correct number. Using the data on the newest airfoils they clearly state a Cd of around 0.0025 as tested in wind-tunnels and real-world environments..
Separate names with a comma.