aerodynamic theory

[othmanskn]

I had reviewed a few books on aerodynamic including one for a
university textbook but I was confused by what I read.
When I started reading in the internet, suddenly I was struck
by the simplicity of it all.
These airfoil designers seem to rely too much on experimental
data rather than fundamentals, namely lift to drag coefficient.

The most significant fact that I find is that the most important
factor in flying is the agle of attack. Earlier I was so concerned
by the bernouli effect i.e. differences in pressures but this
did not explain the inverted flying.

Suddenly it occured to me that the most important is collision
theory. I had been toying with this idea for so long trying to
explain inverted flights.

I had written in excel this theory after cracking my head for a week and revising my graduate physics. I have at my disposal
foilsim which is a good software for airfoil designs.

[greenshamrock]

I cant make head or tail what these graphs mean

Would it help if I turned my monitor upside down to get back to non inverted flight

Or if I stood upside down will the extra blood rush to my head inspire me with more knowlege
or if I emigrated to down under australia would this this cure the problem

Mark in right side up Espaina:D :D :D

[e-geezer]

Can we go over this again?:

If you can get lift from an inclined flat wing, why use an aerofoil? I don't think it gives more lift does it?

Is it because an aerofoil works at a bigger range of angles of attack? I think that's the official line.

If so, why would that be a good thing? Is it so the pilot can skid the wing into loops etc without stalling the wing? If that's the only reason, why don't we stick to flat wings for gliders, where the pilot's not going to suddenly jerk the wing into an obtuse angle to the airflow?

(I've built quite a few foam / balsa wings now for my electric 'trainers' and the only reason I'd use a thick wing is that it's stronger. Maybe less drag too? My wings have similar cross-sections to birds' wings. OK I've never seen a bird do a loop - or a roll.)

[Sail 'n Soar]

These airfoil designers seem to rely too much on experimental data rather than fundamentals

I beg to differ. The airfoil designers have been relying on fundamentals, with their mathematical models VALIDATED through experiments, since not that long after the Wright's first flight. Aeronautical science continues to progress, but the basic understanding was well in hand by at least the 1930's, most likely as reflected in the university texts you described. With a modern PC using such programs as XFoil you can do a good job of designing an airfoil that will have reasonable agreement between theoretical and experimental performance (e.g., CL, CM, CD vs. alpha). You can find at least one site on the web which includes both calculated and measured performance for a large number of airfoils, which will support my position.

Much of what you find on the web, including information in this forum, provides a qualitative understanding of what's going on, but is oversimplified in terms of actually designing something or being to provide meaningful CLs vs CDs. That includes collision theory.

[DeaninMilwaukee]

My personal understanding of why not flat airfoils boils down to these three:


1) structural strength

2) greater tolerance of less than perfect angles of attack making for more gentle stall charactoristics.

3) the biggy: drag reduction. A properly designed airfoil will reduce the drag for a given amount of lift.


#3 really hit home for me when I tried a small fun fly plane with first a proper full sym airfoil and then tried it with a flat plate to make construction easier.

They both flew very similarly except that the flat plate needed more power in all situations to perform the same. Flight times were cut about in 1/2 .


Dean in Milwaukee

[Ollie]

The best lift to drag ratio that you can get from a flat plate at model sizes and speeds is about 12.5. The best airfoils for models can produce lift to drag ratios of about 50 or 60. As stated in the previous posts a flat plate has not only aerodynamic but structural disadvantages.

When the flow pattern around the airfoil becomes turbulent, the physics becomes very complex along with the mathematics to describe it. Beyond the stalling angle of attack nonlinear partial differential equations are involved for which exact solutions have yet to be found. It takes super computers to simulate such flow conditions even aproximately. The flow can be chaotic but fortunately it is repeatable so that experimental results are practical and useful. When eddy currents form just outside the boundary of the airfoil the flow beyond the eddy (laminar seperation bubble) behaves as though the shape of the airfoil were changing with angle of attack. This and other boundary layer flow characterisatics, based on experimental measurements but, not on any detailed description of the flow in the boundary layer, can be used to fairly accurately predict the behavior of airfoils up to the stalling angle of attack.

[e-geezer]

What's wrong with birds' wings then, none of which are 'fat'?

[Sparky Paul]

I can only encourage research into the superiority of flat plates over cambered, streamlined surfaces.
An interested world awaits!

[steve lewin]

Originally posted by e-geezer
What's wrong with birds' wings then, none of which are 'fat'?
None of them are anywhere near "flat" either. They have very definite and variable curvature. They also don't create most their lift by being held fixed while the bird moves through the air.

Other than that that they're just like plane wings :(.

Don't confuse camber (curvature) with thickness (fatness). A typical WW1 plane had a thin section but with lots of camber, commonly, if inaccurately, called an "undercambered" airfoil. Like the current crop of single surface foam wings (Pico Stick etc) these don't behave anything like a flat plate.

Steve

[Sail 'n Soar]

Originally posted by Sparky Paul
I can only encourage research into the superiority of flat plates over cambered, streamlined surfaces.
An interested world awaits!

Sparky,

Take your tongue out of your cheek.

Gerry

[e-geezer]

So birds' wings aren't flat plates. And they don't look like 'conventional' fat model wings.

[e-geezer]

a) Why aren't flat plates any good for models?
b) Why aren't model wings like birds' wings?

[steve lewin]

a) Err...that's what this thread has been about. Are you just writing without bothering to read the previous messages ?
b) because model planes don't fly by flapping their wings and find it really difficult to change their airfoils from second to second like even the dimmest bird can do without a thought.

HTH Steve ;)

[vintage1]

(a) flat plates ARE quite good. Its just that other sections are BETTER. Also structurally its difficult to make a big flat plate wing.
(b) part from Brians obvious comments, many model wings ARE quite like bird wings - espcially sailplanes which are rather like seagulls. Hardly surprising since they both spend their time doing similar things. cruising around looking for lift. Not many sailplanes dive for fish tho :D

[e-geezer]

I was trying to get a complete list of disadvantages of flat plates over other shapes. Is this it?:


smaller range of angles of attack for stable flight
heavier for the same strength
more drag


I do read the previous submissions. And thanks for them. I'm just trying to get a summary that everyone agrees with.

[Ollie]

While I agree with your summary list of flat plate airfoil disadvantages compared to well designed airfoils, I have reservations about how to use such a list. How important or unimportant such disadvantages are to a particular type of aircraft and its configuration varies all over the lot. In the case of a low aspect ratio, high powered delta, for example, the three disadvantages don't count for much. In the case of a high aspect ratio sailplane, as another example, the three disadvantages are all important. Moreover, the size and speed of the model affect the relative importance of the three disadvantages greatly.

To put it another way, judging these disadvantages accurately and fairly requires the quantification of the disadvantages and the context in which they are to be applied.

[e-geezer]

Notice I left off


less lift


Is that justified?

I'm thinking of the really thick wings on heavy-weight-lifting transport planes.

Will these thick wings produce no more lift than a flat plate of the same plan?

[Ollie]

A flat plate at model reynolds numbers stalls at a lift coefficient of about 0.7 and a 10% thick symmetrical section like the S8020 stalls at about the same coefficient of lift. However, near the flat plate's maximum lift coefficient it produces several times the drag of the S8020. At low lift coefficients the flat plate's drag is about 1.7 to 2.0 times higher than the S8020. Thick symmetrical sections like the Ultra Sport 1000 (18.5% thick) only have maximum lift coefficients of about 0.8 at model reynolds numbers. It only takes a small mean line camber to greatly improve the maximum lift coefficient. The S8036 only has 1.89% camber yet its maximum lift coefficient is about 1.2. The S8036 is 16% thick.

Heavy weight lifting models use very highly cambered airfoils. The S1223 is 11.93% thick and has 8.67% mean camber. Its maximum lift coefficient is about 2.0.

The airfoils used on full scale heavy weight lifting transport planes don't work very well at model sizes and speeds.

[othmanskn]

Originally posted by DeaninMilwaukee
My personal understanding of why not flat airfoils boils down to these three:


1) structural strength

2) greater tolerance of less than perfect angles of attack making for more gentle stall charactoristics.

3) the biggy: drag reduction. A properly designed airfoil will reduce the drag for a given amount of lift.


#3 really hit home for me when I tried a small fun fly plane with first a proper full sym airfoil and then tried it with a flat plate to make construction easier.

They both flew very similarly except that the flat plate needed more power in all situations to perform the same. Flight times were cut about in 1/2 .


Dean in Milwaukee

This is the best response among all the other posts and
thank you for giving the above data. I have not tried so many
types of planes yet but just studying the facts.

To enliven the discussion let me comment on your statements.

1) structural strengh.
In modern planes, thin wings are common especially in supersonic jets. Modern materials make them available at greater
costs of course.
My interest is due to the availability of think polystyren cards,
that are about 5mm think that I think I can use as wings to
emulate the modern jet fighters.

2) fine. Just need more skill. With gyros this should be easier.
With more power/ weight ratios this is becoming less significant.
Model planes are not seriously hurt if they stall. No lives will be
lost.

3) More drag for thin wings if you climb steeply.
If you stick to low angles of
attack, thin wings should have less drag.
Common sense and supersonic jet observations.
That collision theory also play a part.
Of course
aerobatics will tend to aggrevate the situation. Your factual
observation is well noted but for me it is less of a concern
compared to easier construction.
I shall try to prove that my assumption is correct,
i.e. I can match the economy and power requirement even
when using thin flat wings.

[othmanskn]

Originally posted by e-geezer
Notice I left off


less lift


Is that justified?

I'm thinking of the really thick wings on heavy-weight-lifting transport planes.

Will these thick wings produce no more lift than a flat plate of the same plan?

More lift at the same angle of attack, wing area and speed.

The reason is the more pronounced Bernouli's effect and
to a much lesser extent, bounyancy effect. If we fill the
wings with helium, we can exploit this bouyancy effect even more.

I am not interested in very slow(relatively) aircraft yet.
When the power/weight ratios were reduced further, all
these will be insignificant. Low drag and weight will be
paramount. Low aspect ratio wings will win in the end.
Of course you need auto pilots(gyros) and VTOL technologies,
which is getting affordable and available.

[Sail 'n Soar]

othmanskn

Common sense and supersonic jet observations.

Supersonic flight and lift are totally different. Such "common sense" is what drove the earliest (pre 1920's) airfoil development the wrong direction until it was learned that slightly thicker isn't necessarily draggier. While thin wings at low angles of attack can have lower drag, if they are flat plates they aren't going to have much lift there either. Their ratio of lift to drag is going to be poor compared to other airfoils unless you go to a highly cambered thin foil. In that case, they can have a very good lift to drag ratio around the design angle of attack, but will fall off rapidly either side of that design angle. Given sufficient strength, such foils will work well for free flight, rubber powered, and rudder only RC planes. They often are also the best choice for very small, low speed planes, but due to other reasons.

That collision theory also play a part.

It's too bad someone raised collision theory in this forum because it just detracts from any real understanding of what's going on. Can you explain just what you mean by collision theory. Why don't you try to calculate the lift on a wing using "collision theory." Then try to explain how any airfoils - model to full scale - can achieve lift coefficients greater than 1.0. If you really want to understand what's happening spend the time deciphering circulation theory. After that a little study of what happens in the boundary layer should make it all clear.


e-geezer

The chances of getting a reply that everyone agrees to on this thread appears to be remote.

[e-geezer]

... and I know consensus wouldn't prove anything. But when I'm too lazy to do the experiments myself, I ask for opinions.

[banktoturn]

Originally posted by e-geezer
I was trying to get a complete list of disadvantages of flat plates over other shapes. Is this it?:


smaller range of angles of attack for stable flight
heavier for the same strength
more drag


I do read the previous submissions. And thanks for them. I'm just trying to get a summary that everyone agrees with.

e-geezer,

Flat plates have worse L/D ( lift divided by drag ) than 'real' airfoil sections. That's pretty much it. Stall characteristics are bad too, but there are other airfoil sections that suffer in this regard as well, so I wouldn't hold that against flat plates. Strength-to-weight ratio is not really a problem with flat plates, it's more a function of thickness. Any thin airfoil section will have issues with strength-to-weight.

Optimizing an airfoil section to get the best L/D requires that we find the best chordwise thickness distribution and the best camber. It would be great if a flat plate were a 'good' airfoil, and we are tempted to say it is for low Reynold's numbers, but it isn't, because it has bad L/D. For low enough Reynold's numbers, it does well enough to justify its convenience for some people and some purposes, but it is not as good as a more conventional airfoil.

banktoturn

[othmanskn]

Elastic collision theory results for similar specs as foilsim is:

3.3 kgf = 32 N

From foilsim emulating a flat plate is this.

FoilSim Plot Data 10/22/03 16:29

Current Conditions:
Density: 1.22859 kg/cu m
Pressure: 0.703 k-pascals
Temperature: 15 °C

Current Settings:
Airspeed: 20 km/hr
Altitude: 0 meters
Angle: 5 degrees
Thickness: 0.05
Camber: 0
Area: 1 sq meters
Lift: 10 newtons

Foilsim is from NASA Lewis centre:
bruce.bream@lerc.nasa.gov

That is my elastic collision theory is out by 300% compared to
foilsim. Or I'd better check my calculations.

[FlightofSong]

Originally posted by othmanskn
That is my elastic collision theory is out by 300% compared to
foilsim. Or I'd better check my calculations.

The fact of the matter is that you're comparing garbage to garbage. This isn't to say that your elastic theory calculations are wrong - just that elastic collision theory is ONLY good for rarefied gas aerodynamics, i.e. drag on satellites in Low Earth Orbit, etc. Fluids do not behave as collections of colliding particles under normal circumstances, they behave as continuous fluids. Air at low alltitudes is definitely a fluid - that's why aerodynamics is a subset of fluid mechanics. So elastic collision theory won't work for aero. Actually, it could work - one of the benchmarks of computer power is how long it would take to solve the flow around a wing using first principles (i.e. particle behavior). I think the number is down to something like 10^13 years now.

Foilsim doesn't get there either - that program is essentially kid stuff, designed by the folks at NASA to teach schoolchildren the basics of lift and drag. It is a 2D, inviscid, camberline-only package that simulates nothing but the most basic of the concepts of lift. It's missing 3D wing effects, boundary-layer effects, wake effects, and numerous other little details that add up to a decent simulation of reality.

The fact of the matter is that all of the explanations of lift, be they momentum theory, Bernoulli theory, circulation theory, etc. all add up to the same thing once you take everything into account. It's that taking everything into account that's tough. Prandtl and his bunch at Goettingen started it, and now folks like Prof. Selig (who I have had the privilege of working with) and Prof. Drela are continuing now (there are lots more people out there - I just shot off the first two that came to my mind.) Airfoils work, and advanced airfoils work well. At extremely low reynolds numbers, the conventional wisom used to be that a flat plate is almost as good, but in the past couple of years, airfoils good down to Re of 50k and less have appeared. Aerodynamics is a very mature science, but there's still more to be done. However, accusing those that are forwarding the science of ignoring basic priciples (as you did in your first post) is simply wrong. These paople are utterly concerned with CL and CD - they design airfoils particularly for certain combinations of them, so how could they not be!?

What you're doing is essentially what the pre-Prandtl crowd though of as the correct ideas for aerodynamics, but the science has moved on to more correct concepts.

For a few very good, concise, understandable discussions of the basics of aerodynamics and flight, let me refer you to a couple of good texts:

Introduction to Flight by John D. Anderson
Aerodynamics, Aeronautics, and Flight Mechanics by Barnes W. McCormick

When you've digested these, if you're still interested in airfoil design, I suggest you check out Theory of Wing Sections, by Abbot and von Doenhoff - it's a classic, but an excellent discussion of classical airfoil design. After digesting that, you'll be properly prepared to examine some more modern airfoil work by Whitcomb and his bunch out at Langley, including the supercritical series and the GA-W series. THEN you'll be properly prepared to read up on the work of the contemporary artists like Selig and Drela.

I guess my point is that before you try to refute an established science, you should undertand its precepts. If you then see a flaw, by all means point it out!

[othmanskn]

Originally posted by FlightofSong
The fact of the matter is that you're comparing garbage to garbage. This isn't to say that your elastic theory calculations are

...

concepts of lift. It's missing 3D wing effects, boundary-layer effects, wake effects, and numerous other little details that add up to a decent simulation of reality.

...

(as you did in your first post) is simply wrong. These paople are utterly concerned with CL and CD - they design airfoils particularly for certain combinations of them, so how could they not be!?

...


We are honoured to have such a useful comment with references. I would certainly look up these books in preference
to others even above those found in the internet.

My original post was not to disprove the established work.
Most of the work that I came across refer to established
CL and CD coefficients for particular airfoils which were found
from experiments. I had one such book at home. Sorry can't remember the author or tittle but it is from a University in UK.
No data for flat plates.

Also most of the formulae that I can decifer refer to airfoil sizes
that are much larger than electric modelling sizes now being
possible, less than 6.7 dm2, which is much lower than the lowest
that even foilsim can give.

That elastic theory of aerodymanic is only useful for flat plates,
but what is important is to know its error margins. We in the hobby field is not really interested in absolute perfections.
I admire work by Takao who built the plywood electric aircraft.

You indeed mention that the elastic collision theory can
be applied with better accuracy in the outer space. This is not
the intention but it shows that by learning basics, we can still
find some uses for them somewhere.

My estimates for the accuracy of the elastic theory was about
90% for flat plates. 70% lower for thick wings depending on
configurations.
Benouli effects, viscosity and circulation, and other effects,
should contribute the other accuracies.
I used to think that Bernouli contribute to 90% of the lifting effect
making it difficult for me to justify the inverted flights. I must have
been misled considerably. NASA and other aerodynamic sites
had succeeded in convincing me that other effects are more pronounced.

I'd love to have access to a very accurate aerodynamic software
that work closest to basic principles so that we can experiment
with all types of ridiculous shapes such as the flat plates
at such low reynould numbers as currently possilbe with electric
planes, cheaply. foilsim is the most convenient that I can find,
at least at the moment.

[FlightofSong]

For an accurate software for evaluating airforl sections, try xfoil or Profili (which is a front end to xfoil). Xfoil is free, and can be downlaoded at raphael.mit.edu/xfoil

Bernoulli effect, if integrated over the entire section, accounts for all of the possible lift conditions on an airfoil - inverted flight is simply acheived by the angle of attack imposed (in that case, negative), which induces a circulation, and voila, faster flow on the "bottom" of the wing, resulting in a pressure differential between the bottom and the top - negative lift...

Any of the three basic explanations of lift (Bernoulli, momentum, or circulation) will explain lift in any condition, so long as they are fully applied.

[othmanskn]

Originally posted by FlightofSong
For an accurate software for evaluating airforl sections, try xfoil or Profili (which is a front end to xfoil). Xfoil is free, and can be downlaoded at raphael.mit.edu/xfoil

...
Any of the three basic explanations of lift (Bernoulli, momentum, or circulation) will explain lift in any condition, so long as they are fully applied.

Again many thanks. I have got xfoil but it was too difficult
to use. I remember getting lost in one of the commands.
Now I forgot completely how to use it. foilsim although
much simpler is also much easier to use for people like
us, hobbyists.

Enclosed for further discussions if possible are graphs where
electric plane reynold numbers are valid.
Taken from my current reference:

martyn Pressnell: Aerofoils for aeromodellers
Published by Pitman.

These figures are mostly from published data in journals.

[othmanskn]

http://c:/apln/aircraft design/reynold-max_lift.bmp

[othmanskn]

Oops. I tried to display it immediately.
I shall try to get a php encoder.

[Mike James]

There's more to this issue of flat plates vs. airfoils than just the shape...

I'm not an aerodynamics expert, and when I have problems I can't solve, defer (and beg info from!) gentlemen such as Ollie. What I can tell you here is based on recent actual experience, and has been proven to me several times over the years.

I recently designed, built, and flew a 1/3rd scale model of the Long EZ. Even at 1/3rd scale, the canard chord was only slightly over 4 inches. All scale airfoils were tried first, out of curiosity.

Shortening the story greatly, the model would not rotate, although several different airfoils and angles of attack were tried. Increasing the chord to 5 inches, increasing the span about 10%, and again trying different airfoils, the model would rotate and fly, but required full up elevator (canard) and full power to maintain level flight.

When I finally increased the canard chord to 5.25 inches, the plane instantly changed to a model with excellent all-around performance, and had plenty of elevator authority, even during deadstick landings.

This tells me (again) what many of us have found, and that is that, at typical model sizes and Reynolds numbers, the airfoil doesn't seem to matter much, below a chord of about 5 to 5.5 inches. Once over this "hump", the varying of airfoil sections creates performance differences you can readily see.

(The article is now re-posted on my non-commercial site, at http://homepage.mac.com/mikejames/rc_reviews/berkut01.html )

[Sail 'n Soar]

Originally posted by othmanskn
I have got xfoil but it was too difficult
to use. I remember getting lost in one of the commands.


Did you down-load and print the XFoil manual? If not, give it a try.

[RCAV8R13]

Gentlemen, gentleman..please. The lighter you build an airplane the more aerodynamic murder you can get a way with. There are several avantages to a flat plate wing. If you build light enough, a flat plate can be strong enough (It can even contribute to the goal of reducing the weight!). The lighter the plane the less strength it needs anyway. Add elevator to flapperon mixing and you have variable camber wings (just like the birdies). Todays 3D aerobatic planes seem to spend most of their lives in a post stalled condition, hovering, etc. In this condition a nice smooth flow of prop wash over the tail, undisturbed by having to go around an airfoil, is very useful indeed.
Coefficient, shmoefficient, build light and soar with the birds!
RCAV8R13
Example:

[e-geezer]

Here are the stats on opinions among Dean, Ollie, Vintage1, Othmanskn and banktoturn regarding flat plates:

Less strength per weight (for given material) : 4 out of 5
Greater range of Angle of Attack (before stall): 2 out of 5
Increased drag at all but 0 degrees: 5 out of 5
Reduced lift at low angles of attack: 0 out of 5

... if I'm interpreting you right.

:-) Oh, and the usual sprinkling of conservatives saying all was obvious, or the question was ill-posed, or go read these great books. :-)

[Ollie]

The hard facts are contained in the polar diagrams. You can't get at the hard facts by counting the votes based on verbal answers to your questions. You must learn to read the polar diagrams or be forever short of facts about airfoils.

[Sparky Paul]

What Ollie says!

[banktoturn]

Originally posted by e-geezer
Here are the stats on opinions among Dean, Ollie, Vintage1, Othmanskn and banktoturn regarding flat plates:

Less strength per weight (for given material) : 4 out of 5
Greater range of Angle of Attack (before stall): 2 out of 5
Increased drag at all but 0 degrees: 5 out of 5
Reduced lift at low angles of attack: 0 out of 5

... if I'm interpreting you right.

:-) Oh, and the usual sprinkling of conservatives saying all was obvious, or the question was ill-posed, or go read these great books. :-)

e-geezer,

Ollie gave it to you straight; the polar diagrams tell the story, at least as far as aerodynamics. There are other parts of the trade-off, as several posters have mentioned, such as simplicity, strength, ability stall the wing intentionally, etc. You need to decide what you're after, and make a choice. If you are flying fine with a flat plate, then there is no need for a debate. If you really need to optimize performance, then I doubt that a flat plate is the best solution for any type of plane.

banktoturn

[e-geezer]

You mean radius / angle r-theta diagrams? - as distinct from X-Y? I think this is the first time these have been mentioned on this thread (?).

[Ollie]

egeezer,

The polar diagrams relevant to airfoil behavior are:
1. Coefficient of lift versus coefficient of drag with reynolds number and angle of attack as parameters.
2. Coeficient of lift versus angle of attack.
3. Pitching moment coefficient versus angle of attack.

They are called airfoil polars because angle of attack is either a parameter or a coordinate.

You simply must read an elementary text on aerodynamics like Model Aircraft Aerodynamics by Martin Simons or you will never get off the starting line to understanding.

You will only be defeating yourself if you dig your heels in against this sort of advise which comes from many knowledgable people.

[e-geezer]

OK Pop.

[Ollie]

That's Grandpa to you. ;)