Espritmodel.com Telemetry Radio
Reply
Thread Tools
Old Feb 05, 2011, 05:23 PM
Ya gotta land eventually...
Davey NY's Avatar
United States, NY, Mahopac
Joined Sep 2010
1,097 Posts
ok got it all so far, but to make sure I am understanding, I should be able to fly a 3 wing slower than a biplane of similar weight and power system then, but I will loose top end, yes?

So all else being equal..the tri should be a slower flyer/cruiser.

So lift is not related to drag except that the drag created (which may or may not induce greater lift efficiency) must be overcome by thrust. Yes?
Davey NY is offline Find More Posts by Davey NY
Last edited by Davey NY; Feb 05, 2011 at 05:32 PM.
Reply With Quote
Sign up now
to remove ads between posts
Old Feb 05, 2011, 05:35 PM
Wake up, feel pulse, be happy!
C₄H₁₀'s Avatar
United States, AK, Fairbanks
Joined Aug 2009
12,389 Posts
The ability to fly slowly comes from wing area, not number of wings per se. If you have more wing area, you can push against more air and get more lift. The configuration of the wings doesn't matter in most cases, though the real Fokker DR.1 was actually impaired by its middle wing and would have flown better without it. Simple case of too little knowledge of aerodynamics.

If my plane has X wing area and the weight is Y, it doesn't matter if that wing area is spread out on two, five, or three hundred different wings; the loading and stall speed remains the same.

If I have a monoplane with X wing area and Y weight and I make a triplane wing setup for it, changing nothing but the wing configuration so that the wing area remains at X and the weight remains Y, it will not necessarily fly any better.
C₄H₁₀ is offline Find More Posts by C₄H₁₀
Reply With Quote
Old Feb 05, 2011, 05:54 PM
Suspended Account
Joined Jul 2006
22,991 Posts
Quote:
Originally Posted by Davey NY View Post
ok got it all so far, but to make sure I am understanding, I should be able to fly a 3 wing slower than a biplane of similar weight and power system then, but I will loose top end, yes?

So all else being equal..the tri should be a slower flyer/cruiser.

So lift is not related to drag except that the drag created (which may or may not induce greater lift efficiency) must be overcome by thrust. Yes?
I think you've got the underlying concepts. If the tripe has an equal wing area to the bipe, weighs the same, has the same airfoil (we don't want one being a flat plate winged plane and the other being a highly undercambered floater wing).... then if you transfer the bipe motor that was designed for the bipe, it will probably lose a little top end because of the extra draggy airframe of the tripe.

Yep, I'd say that just speaking at this basic level of understanding, you've got yourself in the door pretty solidly...

Chuck
NoFlyZone is offline Find More Posts by NoFlyZone
Reply With Quote  (Disabled)
Old Feb 05, 2011, 07:41 PM
Ya gotta land eventually...
Davey NY's Avatar
United States, NY, Mahopac
Joined Sep 2010
1,097 Posts
Thanks fellas. I appreciate it.

So to button it up on my end, in keeping with the model we are using, once the tripe may flaot along slower but once it hits its stall speed, it should drop out of the sky much harder than the bipe since it has much more drag...so enjoy the slower flying, lighten it up as much as possible to boot..and don't make a final approach to close to your stall speed or you'll have some trouble


David
Davey NY is offline Find More Posts by Davey NY
Reply With Quote
Old Feb 05, 2011, 08:01 PM
Suspended Account
Joined Jul 2006
22,991 Posts
Quote:
Originally Posted by Davey NY View Post
Thanks fellas. I appreciate it.

So to button it up on my end, in keeping with the model we are using, once the tripe may flaot along slower but once it hits its stall speed, it should drop out of the sky much harder than the bipe since it has much more drag...so enjoy the slower flying, lighten it up as much as possible to boot..and don't make a final approach to close to your stall speed or you'll have some trouble


David
Hi David,

How a model behaves once it stalls is not a function of it's drag, but rather a function of other, oftentimes unrelated factors. Therefore we can't say that high drag models necessarily equate to bad handling charactheristics once the plane stalls.

Anecdotal Evidence Alert: Some of my bipes (lots of drag) will slow down to below their stall speed, and the only thing that happens is the plane's nose drops, and it starts dropping at a forward angle, all by itself until it picks up enough airspeed to let it 'fly' again. At the other end, I have low drag planes that will stall, then start an ugly looking gyrating dive to the ground, tip stalling along the way and causing one's heart to skip a beat in fear that you might not be able to recover before it plows into the ground.

Chuck
NoFlyZone is offline Find More Posts by NoFlyZone
Reply With Quote  (Disabled)
Old Feb 05, 2011, 08:36 PM
Suspended Account
Joined Jul 2006
22,991 Posts
Davey,

Here's an example of a light, floaty plane that, while it is certainly draggy, is nowhere near as draggy as many other planes.

Very slow flying plane, powered by a tiny little GWS IPS motor pulling about 1.5 to 2 amps maximum from a little 2s Li-Po. 50" wingspan, maybe 6-7 ounces AUW. When this plane hits it's stall speed it's almost a nose first dive to the earth. Very ugly stall.

Next to it is another 6-7 ounce plane, same 50" wingspan, draggy, and powered by the exact same motor. When this plane hits it's stall speed. it lowers it's nose, flies at an angle downward, picks up speed and flies level again. Very pretty stall.

Ugly stalls, therefore, are not strictly a function of drag.

Chuck
NoFlyZone is offline Find More Posts by NoFlyZone
Last edited by NoFlyZone; Feb 05, 2011 at 08:47 PM.
Reply With Quote  (Disabled)
Old Feb 05, 2011, 09:18 PM
Ya gotta land eventually...
Davey NY's Avatar
United States, NY, Mahopac
Joined Sep 2010
1,097 Posts
Hey Chuck.

Ok so that theory is out..but info assimilated all the same. Even as I type I recall other conversations we've had Chuck and what you just posted here gels with everything I've got grinding around, so I think I am good (now who'd of thought that 4 months ago, right

Thanks Chuck et al.
Davey NY is offline Find More Posts by Davey NY
Reply With Quote
Old Feb 05, 2011, 09:28 PM
Suspended Account
Joined Jul 2006
22,991 Posts
Quote:
Originally Posted by Davey NY View Post
Hey Chuck.

Ok so that theory is out..but info assimilated all the same. Even as I type I recall other conversations we've had Chuck and what you just posted here gels with everything I've got grinding around, so I think I am good (now who'd of thought that 4 months ago, right

Thanks Chuck et al.
You're very welcome! Once you are familiar with the "groundwork", so to speak, then you can start to study Don's and Beagle's excellent posts; where they get just a mite more complicated and 'in depth'...

Chuck
NoFlyZone is offline Find More Posts by NoFlyZone
Reply With Quote  (Disabled)
Old Feb 06, 2011, 08:20 AM
Ya gotta land eventually...
Davey NY's Avatar
United States, NY, Mahopac
Joined Sep 2010
1,097 Posts
LOL Chuck.

Just a bit... but I am reading.
Davey NY is offline Find More Posts by Davey NY
Reply With Quote
Old Feb 07, 2011, 02:50 PM
Registered User
Don Stackhouse's Avatar
United States, OH, Bradford
Joined Jun 2005
3,960 Posts
Quote:
Originally Posted by Davey NY View Post
Thanks fellas. I appreciate it.

So to button it up on my end, in keeping with the model we are using, once the tripe may float along slower but once it hits its stall speed, it should drop out of the sky much harder than the bipe since it has much more drag...
Well, um... NO.

Lots of information floating around here, some good, some not so good, some wrong altogether. That includes that Wikipedia article.

Better start with the fundamentals.

If an object is passing through a fluid (or the case of a fluid passing over an object, same results), the two interact with each other, creating forces between them.

Each little piece of the object's surface has fluid molecules bumping into it, scrubbing against it, getting dragged along with it, and transferring the motion and momentum they pick up from the surface to the fluid molecules near them, who pass it on to the ones near them, and so on, until the motion of the object through the fluid (or the fluid past the object) has stirred up the fluid for a large volume surrounding the object. Theoretically the disturbance spreads like ripples in a pond all the way to infinity, although in real life the effects of viscosity tend to limit the total disturbance to somewhat less than that. It's a wild, swirling dance, at once mesmerizing in its complexity, and simultaneously, once you see the grand patterns in the dance, elegant in its simplicity.

All this bumping and rubbing creates a myriad of little forces against and along the object, all over the surface of the object. We can add these all together into a single net force that will be at some angle to the direction of the object's motion through the fluid. Generally speaking, in most cases that net force will tend to oppose the motion of the object.

We can then break that force down into two components.

The component perpendicular to the direction of flow is called "lift". It can be up, down, sideways or whatever, but if it is perpendicular to the direction of the flow, it is called lift.

The component parallel to the direction of the flow of the fluid (NOT to the centerline of the object itself !!), if it opposes the motion of the object through the fluid (which as I mentioned is nearly always the case) is called "drag".

Note, in common aeronautical texts they show a force that opposes drag, and propels the object through the fluid as "thrust". However, thrust can be at any angle, not just parallel to the direction of travel through the fluid. For example, the thrust of the main rotor of a helicopter in forward flight is directed partly upward and partly forward, so that it both supports the weight of the helicopter, and pulls it forwards, overcoming drag. Drag is always parallel to the direction of travel through the fluid.

Drag can be broken down into two subcategories, "parasite" (the result of the fluid just getting out of the object's path, going around it, and closing up again behind it), and "induced" (the natural by-product of making lift using an object that has less than infinite span, which includes pretty much any objects we are going to build).

Let me repeat that:
Drag that is the by-product of making lift is called "induced drag"; all other drag that is NOT the result of making lift is called "parasite drag".

One other term you might hear with regard to airfoils is "profile drag". That's the parasite portion of an airfoil's drag, the drag due to just getting the air safely around the airfoil, independent of any making of lift. Typically a large portion of an airfoil's profile drag is skin friction.

Now lets get back to the subject of lift, which in turn will help us understand induced drag.

A wing makes lift by grabbing chunks of air and shoving them "down", with "down" in this case meaning whatever way is opposite to the direction of the lift. It's Newton's law, the one about action and reaction. To make a given amount of lift, you can either grab a small chunk of air and give it a violent shove, or grab a big chunk of air and give it a gentle push. Hint: that second option results in lower induced drag.

Some of you are probably saying "Wait a minute, what's all that stuff we learned in science class about Bernoulli, and the air moving faster over the top of the wing, with lower pressure on top and higher pressure underneath?" No, unlike some seem to want to believe, there is not some great war between Newton and Bernoulli. A wing makes lift by grabbing chunks of air and shoving (accelerating) them, and as Newton explained, F=Ma (i.e.: Force equals Mass times acceleration), so that accelerating of that mass of air creates the lift force. Bernoulli (the reason behind the higher pressure under the wing and lower pressure above) explains how the wing can grab hold of something as etherial as air in order to do that shoving.

So how big are these "chunks of air"? That's complicated, but we can get a grossly oversimplified visualization. Imagine a cylinder of air, with a diameter equal to the wingspan, and a length equal to the distance the plane flies in one second. The mass of the air inside that cylinder is a representation of the size of the chunks of air that flying surface is "processing" to make its lift.

The volume of a cylinder is the area of its cross-section, times its length. The cylinder gets bigger in proportion to its length, and in proportion to the SQUARE of its diameter. This is why wingspan is so important in the design of gliders, where induced drag is especially important. Even a little increase in wingspan makes the volume of that cylinder of air much bigger.

Alternatively, if we have a lot of forward speed, even if the wingspan and the diameter of the cylinder might be small, the longer length of the cylinder can still give it a respectable volume, resulting in larger chunks of air. This is why very fast airplanes can get away with relatively short wingspans without suffering excessive induced drag penalties (except when they have to slow down for landing and takeoff).

One other thing you can see from this: for any airplane, if you increase the airspeed (all other things being equal), the induced drag goes DOWN, in proportion to the square of the airspeed. Note, it's the square (not just linear) because the size of the chunk of air gets bigger, and the amount you have to shove it to make a given amount of lift goes down, so when you multiply those factors together you get a squared relationship.

On the other hand, parasite drag follows the opposite pattern. At higher speeds, the fluid is being more abruptly pushed aside to let our object go through it, there is more scrubbing of the fluid against the object's surface, and the parasite drag goes UP with the square of the speed.

Now lets consider that seemingly simplest of aircraft (until you try to design one, that is), a glider.

A glider does, in fact, have an engine, and no, it is not the wind. A glider is powered by gravity. It slides downhill through the air like a sled. It's the most reliable engine in the world, if gravity ever quits, we have much more important issues to worry about.

The slope of that invisible "hill" the glider is sliding down is called the "glide ratio", and the glider's flattest possible glide ratio it can sustain is a measure of its efficiency. Glide ratio is typically expressed as how far the glider moves forward for a given loss of altitude. For example, a full-scale Piper J-3 Cub with the engine at idle has a glide ratio around 8:1, meaning it goes forward eight feet for each foot of altitude it loses. If you were at 5,280 feet AGL (Above Ground Level) and killed the engine, you could glide for eight miles before touching down.

BTW, for any of you who happen to be deathly afraid of little airplanes, that gives you a possible area from that altitude of a little over 201 square miles in which to find a safe place to land, and Cubs are pretty flexible on exactly what constitutes a "safe place". If I was going to have an engine failure, a Cub would be an excellent choice to have it in.

Now, the interesting part: For glide ratios flatter than about 5:1, the glide ratio is approximately equal to the ratio of lift to drag. (Note, that assumes that lift is equal to weight, and at glide ratios steeper than about 5:1, that assumption becomes increasingly invalid.) We therefore typically use the term lift-to-drag ratio ("L/D") interchangeably with the term "glide ratio".

The flattest glide ratio a given airplane can sustain, sometimes referred to as L/Dmax, occurs exactly at the point where its parasite drag and induced drag are exactly equal. If you have two mathematical functions (parasite drag vs airspeed, and induced drag vs airspeed in this case), one increasing (parasite drag) and the other decreasing (induced drag), the point where the sum of the two is lowest is the point where their two graphs cross each other, the point where they are exactly equal. In level, unaccelerated flight the lift is essentially equal to weight (assuming the glide ratio is sufficiently flat, better than about 5:1) and therefore constant, so maximizing glide ratio often becomes a matter of minimizing drag. Note, you can improve the L/D by reducing the induced drag or the parasite drag. However, if you reduce the induced drag, the airspeed at which the graphs cross and L/Dmax occurs will be slower, while if you reduce the parasite drag, the speed for best L/D will be faster.

What effect do things like wing area (not span) and wing loading have on all of this?

That's where Bernoulli gets involved. Going back to our fundamentals, the "shoving of the air" (Newton) is the driving force in the making of lift. If this was a car, Newton would be the horsepower of the engine. However, to do all that shoving, we have to get a firm enough grip on that air to do all that shoving. Bernoulli, and things like the airfoil shape and the area of the wing(s) are what provide that grip. In our car analogy, these are the traction of the wheels. It doesn't matter how much horsepower you make in the engine unless you can somehow transmit that power to the road. Otherwise, we're just spinning our wheels.

The aerodynamic equivalent of spinning our wheels is an aerodynamic "stall". The wing has lost its grip on the air, so it can no longer do much accelerating of that air to make our lift. We don't have any "traction".

How much traction our car's wheels have depends on things like tire diameter and width, tread pattern and how well it is matched to road type and road conditions, rubber hardness and characteristics, etc.. Snow tires don't provide as much traction on dry roads as regular tires, but of course are better on snow. Tractor tires would not work well on the Interstate, but car tires aren't much use in a cornfield, and neither of them are optimum on a drag racing track.

Likewise, the amount of "grip" our wing has on the air depends on airfoil shapes and how well they are matched to the application. The thicker airfoils that provide more max lift at full-scale Reynolds numbers typically provide less lift than a well-designed thinner airfoil at smaller model Reynolds numbers, and too much or too little camber is critical as well. Also important are things like wing planform and twist distribution, and of course how much wing area we have reaching out to grasp the air. There is only so much pressure differential each square foot of wing can support, so to make a given amount of lift you have to have enough surface area to stay within those limits, and the span, airfoils, planform and twist to use that area efficiently and effectively.

How do biplanes and triplanes enter into this? That's where thing start to get really complicated (yes, I know, you thought it was already complicated), with implications for both Newton and Bernoulli, as well as structural and operational considerations.

Lets begin, as before, with Newton. Imagine our monoplane wing we've been discussing so far, but now we add a second wing the same size, at some spacing from it, in a biplane arrangement. Take our cylinder of air, as before, but slice it in half lengthwise, right down the middle, and insert a rectangular brick of air in between, the same height as the spacing between our two wings. The resulting oblong-cylinder-of-air is now larger than the purely cylindrical chunk of air we had before. Bigger chunk of air, the induced drag goes down. That's right, a biplane can have less induced drag than a monoplane of the same span.

Note also that inserting a third wing in between the top and bottom ones does not really have any significant beneficial effect as far as Newton is concerned, unless you have so much space between the top and bottom wings that they would start to act as independent wings. For that to happen, they would need to be more than a wingspan apart. From an airframe design standpoint, that would be, literally, a very tall order.

Where multiplanes really get into trouble is in the Bernoulli department. Airfoils making lift create a low pressure zone on top, and a high pressure zone on the bottom. When you start stacking them on top of each other, the high pressure zone of the one on top starts filling in the low pressure zone on the top of the one below, cancelling out some of the lift of both. The one on the bottom gets hit the worst, but both suffer. The losses can be quite severe if they are less than a chord length apart. It's best to try to have at least a chord and a half between them, and to essentially eliminate the problem you would need at least about a half-span, which is pretty impractical. Trying to pack three wings into a reasonable vertical space just makes the problems worse.

The net result is that for the same wing area, you can't squeeze as much lift per wing out of two or three wings as you can out of one. You can squeeze a lot more area into the same space than you can with a monoplane. If you do everything right, you can make some gains in the induced drag and the total lift you can squeeze from a given span. However, you will pay for it in more whetted area, which adds more skin friction, a form of parasite drag. At low speeds you can still come out ahead on the deal, with a lower stall speed and tighter turning radius, maybe a small net benefit in total low-speed drag, but at high speeds (where parasite drag dominates), you will almost certainly lose, probably by quite a bit.

As always, there are other considerations. In the case of the Fokker DR1, spreading the total wing area over three wings instead of two meant that each of those wings, for the same span, had less chord. This improved visibility. The top wing and bottom wing were well above and below the pilot's main field of vision, and their narrow width reduced the amount of visual obstruction they caused. The middle wing was nearly edge-on in the pilot's view, reducing its visual obstruction. The increase in overall height (and the corresponding chunk of air) allowed for a small reduction in span, which should have reduced the roll damping and roll inertia, which should have improved roll rate. Unfortunately, the necessary vertical spacing put the mass of the top and bottom wings farther from the C/G in the vertical direction, which added to the plane's inertia about the roll axis, negating that benefit of the reduced span.

In addition, the DR1's ailerons were only on the top wing, and not well designed. As a result, the plane had a rather lousy roll rate, especially when inverted. Once in a bank, it could turn very tightly, but could not roll into or out of a bank very quickly. In comparison, the Sopwith Camel had ailerons on all four wings, and a faster roll rate. In a dogfight, the smart Camel pilot would do well to try to get into a jinking fight, where the plane's faster roll rate had an advantage, while a Triplane pilot would be better off trying to keep things in a continuous max-effort turn, or try to loop over the top and get behind someone attacking from the rear (a loop is just a tight turn tilted into the vertical plane), maneuvers that would take advantage of the Triplane's tighter turning radius. BTW, these same tactics apply in dogfights between our Roadkill Series Fokker DR1 and Sopwith Camel, as Joe and I have had considerable fun demonstrating.

A number of sources have tried to make an issue of the Triplane's almost nonexistent yaw stability, and its ability to make "flat" turns using the rudder, without banking. This is only a transient possibility. At some point you need lift to change the plane's direction of flight, and trying to get it from the fuselage is an extremely inefficient and ineffective way to do that. Unless the plane can tolerate being flown backwards without shedding all sorts of flying surfaces (something I would not want to try with anything that came out of the Fokker factory, which during WW I was notorious for flimsy construction and shoddy workmanship), just trying to boot the plane around with the rudder alone is not likely to be useful. It might buy some time to let the plane's ineffective ailerons catch up with the situation, but a better set of ailerons would be a bigger advantage. The reports I've read of the Sopwith Triplane (which did have better ailerons) support this.

So if biplanes and triplanes have so many pitfalls, why did they totally dominate aviation's early years? For that we can probably thank the Wrights. As I indicated above, at low Reynolds numbers (small planes and/or slow speeds, such as birds, and our models), thinner airfoils actually work better. However, at full-scale Re's, thicker airfoils typically have an advantage. Of course the Wrights' wind tunnel was testing small airfoils at low speeds, which favored thin wings, too thin to make the spars deep enough for cantilever, internally braced wings to be practical. Even the monoplanes of those days relied on extensive external bracing. In fact, Fokker and Junkers were some of the first to explore thicker airfoils and internally-braced cantilever wings, but not early enough to really reap the advantages until after the war.

Meanwhile, if you are going to have external bracing, a biplane or triplane layout is better from a structural efficiency standpoint. You can get a lot more strength and rigidity for less weight and less wires than you can with a typical externally braced monoplane. The geometry naturally tends to give the wires better angles for better mechanical advantage. Multiplanes also have an easier time getting some structural redundancy, which helps the plane withstand battle damage.

What finally ended the reign of the externally-braced biplane was the advent of all-weather operations by the airlines in the early 30's. This meant fitting de-icing equipment to the airplanes. That's practical for an internally braced airplane, but not really a viable option if you also have to somehow de-ice a whole cat's-cradle of struts and wires.

And what about your other question regarding stall characteristics? No, you can't make any really sweeping generalizations about biplanes or triplanes on that subject. Stall characteristics are the result of complex interactions of a host of factors, beginning with the airfoil characteristics.

In a stall, large areas of the airflow separate from the airfoil, no longer following its surface. There is typically some major loss of lift, but exactly how much depends on the details. The shape of the airfoil's Cl vs alpha graph (lift coefficient vs angle of attack) will tell you what to expect.

For example, the Cl vs alpha plot for a USA-35B airfoil (the airfoil used by the Piper J-3 Cub, and the wing tip of a DeHavilland DHC-1 Chipmunk) looks almost like a semicircle, very gently and smoothly rounded off at the top, with a gradual decrease in Cl after the stall. You still can get a significant amount of lift from it well beyond the alpha for max lift. That rounding off at the top also indicates a lot of gradually increasing separated flow before and after the stall, which in a full-scale airplane means the pilot will feel a lot of shaking and buffeting before the airflow gives up in disgust at these unreasonable demands from management for more lift, and goes out on strike.

The full-scale DeHavilland Chipmunk is a perfect example of this. It's one of the most delightful and honest airplanes I've ever flown, and you would have to be brain-dead to miss its advance warnings. I had the privilege of getting my aerobatic training in one, and it's been literally a let-down flying anything else ever since. Besides the USA-35B tip airfoil, whose gentle stall meant that the airplane could be quite deep into a stall and still have decent aileron response, the NACA 2412 at the root (which also has a very nice, although more well-defined stall of its own) had stall strips on the leading edge about 18" long, about a foot out from the fuselage on each side. As the plane approached stall, the stall strips would begin to trip the flow at the root.

Note, those are the characteristics of those airfoils at full-scale Re's, and you would have to make some adjustments to their shape to get similar characteristics at typical model Reynolds numbers.

At about 8 knots above the stall, the turbulence coming off the stall strips would begin to tickle the elevators on the tail, which the pilot could feel as a little "nibble" in the control stick. As the angle of attack was increased (an airfoil stalls at its max lift coefficient, which is the result of angle of attack, not airspeed, you can stall at any airspeed if you pull back or push forward hard enough), the nibble grew stronger, until about 4 knots above stall the control stick was shaking strongly. At that point there would begin a faint vibration in the seat. This would increase, along with more shaking of the stick, until when the plane finally reached stall, the whole airplane would be shaking like a car on a bad gravel road, but still fully controllable on all three axes. At that point there would be a firm but gentle and moderate stall break. Just ease off the back pressure on the stick and the plane would be flying again. The plane would go through this entire litany, practically phoning ahead for a reservation, whether it was at half a "G" at the top of a loop, or at three G's at the bottom of one.

On a model you would not be able to feel the buffeting, but you might see some if you were paying attention. Most likely your best advance warning would be the way the controls start to feel less effective, and the plane doesn't seem to pick up as much additional lift when you pull the stick further back. It feels "mushy" on the controls. However, the biggest issue would be making sure the root stalls well before the tips, and that the stalled region of the wing spreads slowly and gradually out along the span as the plane gets deeper into a stall. When I'm designing a wing, I check the local lift coefficients along the span and compare them to the stall Cl's of the local airfoil shapes along the span as one way to keep an eye on this. Some good planform design software can be a big help in this.

Note, from the explanation above, it should be obvious that the Chipmunk's wonderful handling characteristics did not just happen by accident. Someone did a lot of engineering and development work making them happen. Occasionally when designing an airplane, full-scale OR model, you might encounter a "happy accident", but it's very poor engineering practice to count on that. Good airplanes are almost always the result of doing your homework, not just counting on rules of thumb, TLAR ("That Looks About Right"), and blind luck.

Now, for another example, consider the NACA 23012 airfoil. This is very popular on a lot of full-scale general aviation aircraft, partly because it has a respectable L/D for a non-laminar-flow airfoil of that generation. It also has some of the worst stall characteristics of any I've seen. The Cl increases steadily , almost perfectly linear, with very little rounding off as the alpha approaches the stall angle, so there's very little buffeting or "mushy" feeling prior to the stall. Then, at the stall, the Cl suddenly drops like going over a cliff, to a very low value, with further rapid decreases from there at alpha angles beyond stall. No significant warning, it just quits flying, all of a sudden. It's the mark of a "leading edge stall", where the separation at stall begins at or near the leading edge and takes out all the upper-surface flow at once, rather than a trailing-edge stall that then moves slowly and progressively forwards, like the airfoils I described above.

Some of the full-scale airplanes I've flown do indeed behave something like that. However, I have flown some that use the NACA 23012 and yet have fairly gentle stall characteristics. Not as gentle or predictable as the Chipmunk's, but still not bad. This is typically because even though the airfoil itself quits flying all at once, the designers used things like washout (twist in the wing that puts the tip at a few degrees lower angle than the root), wing planform (such as constant-chord or modest amounts of taper), or stall strips (like the Chipmunk) to force one part of the wing to stall before the rest, so that the stall begins at the root, and moves slowly and progressively outboard from there.

Again, good handling was the result of careful design by someone who knew that an airplane needs good handling, not just good performance, to be successful, and did not rely on luck.

All of this applies just as much to model design as it does for full-scale. Someone earlier tried to take some of my comments out of context and claim that what I referred to as "cruise" was a concept that applied to full-scale airplanes but was irrelevant to models. They missed the point.

Every airplane design, be it model or full-scale, has something we want it to do, some purpose for being, some "mission profile" to perform. "Cruise" is just one example.

Maybe it's to fly as long as possible on a specified size battery.

Maybe it's just to fly around ("cruise"), look cool, and be fun to fly. In which case it's still an advantage to have good aerodynamic efficiency, since that will let it use a smaller, cheaper and lighter battery, and fly longer and better on that battery. In this model, good handling is even more crucial.

Maybe it's to be able to climb to 200 meters in 30 seconds, then shut down the motor and become a very efficient sailplane for an ALES (Altitude Limited Electric Soaring) competition.

Maybe it's to fly almost 200 MPH on the straights, and retain as much of that speed as possible in a 17 G turn at the pylons for a Quarter-40 racer.

The point is that whether it's a reduced-power "cruise" (which a typical sport model spends most of its time doing), or a climb requirement, or a glide requirement, or something else, full-scale airplanes as well as our models have things we expect them to do. In most cases there are multiple things we expect them to do, which almost always requires juggling of contradictory demands. In addition, if the plane has poor handling, it is not likely to be successful regardless of what sort of performance it has.

To achieve all of this, we can waste a lot of wood, empty our wallets on failed experiments, and generally make ourselves frustrated and miserable as design after design fails while we keep tinkering, always hoping for a "happy accident" that rarely comes.

Or, we can learn about what makes the things happen that we want to achieve, make our design decisions based on educated guesses rather than just blind stabs in the dark, and accomplish far more for the same amount of effort and investment, reaping greater satisfaction and greater rewards in the process. We will also have the satisfaction of knowing why something worked, as well as how to do it again on the next model.

You decide.

Don
(humble student of "the dance")
Don Stackhouse is offline Find More Posts by Don Stackhouse
Last edited by Don Stackhouse; Mar 11, 2011 at 09:29 AM.
Reply With Quote
Old Feb 07, 2011, 02:54 PM
Suspended Account
Joined Jul 2006
22,991 Posts
WOW, Don.... Thanks so much! That is definitely going into my files to keep and study!

Chuck
NoFlyZone is offline Find More Posts by NoFlyZone
Reply With Quote  (Disabled)
Old Feb 07, 2011, 04:13 PM
Wake up, feel pulse, be happy!
C₄H₁₀'s Avatar
United States, AK, Fairbanks
Joined Aug 2009
12,389 Posts
I believe I am learning more about physics from this thread than I am in my physics class at school. I also believe that Don is better at teaching and explaining things than my teacher, and Don is infinitely more patient and he's not getting paid....

Don, I am officially nominating you for "Helpful Genius of the Millenium". I will be saving that post and MEMORIZING it

I know that the next plane I design is gonna be done with me making extensive use of this thread and all the information that's been posted here. Truly amazing.....
C₄H₁₀ is offline Find More Posts by C₄H₁₀
Reply With Quote
Old Feb 08, 2011, 09:28 AM
Registered User
Lilienthal's Avatar
Littleton, CO
Joined Nov 2010
1,239 Posts
Wow! That was exceedingly good. Thank-you Don.
Lilienthal is offline Find More Posts by Lilienthal
Reply With Quote
Old Feb 08, 2011, 05:26 PM
Ya gotta land eventually...
Davey NY's Avatar
United States, NY, Mahopac
Joined Sep 2010
1,097 Posts
well... that's kinda what I meant to say Don

but sincerely thank you very much. I printed this out and am reading it several times now...

David
Davey NY is offline Find More Posts by Davey NY
Reply With Quote
Old Feb 08, 2011, 07:44 PM
Suspended Account
Joined Jul 2006
22,991 Posts
Hey Don,

I just wanted to say thanks for not ripping up the people who genuinely try to help other people and manage to give bad information while doing so. This is precisely why FliesLikeABeagle is one of the most admired people on RCG. He never makes us look like idiots when we screw up in giving advice... and MAN can we ever screw up...

Chuck
NoFlyZone is offline Find More Posts by NoFlyZone
Reply With Quote  (Disabled)
Reply


Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
Question Is my rudder too big? Jefte Foamies (Scratchbuilt) 11 Sep 22, 2010 09:48 PM
Discussion crappy hobby store, and now a motor too big for my esc? evilways Power Systems 10 Sep 19, 2010 04:14 PM
Question How big is too big for a park? Irg 3D Flying 22 Sep 03, 2010 01:47 PM
Motor shaft too big for prop.... OUAndy1807 Power Systems 4 Mar 17, 2004 08:52 AM
Is there such a thing as motor being too powerful/prop being too big? simingx Power Systems 9 Jan 28, 2003 07:36 PM