Logo
Builders' Plan Gallery  |  Hip Pocket Web Site  |  Contact Forum Admin (Account/Technical Issues)  |  Contact Global Moderator
September 25, 2022, 01:04:12 PM *
Welcome, Guest. Please login or register.
Did you miss your activation email?

Login with email, password and session length
 
Home Help Search Login Register
Pages: [1]   Go Down
Print
Author Topic: How do WS planes with "aft CG" fly at all???  (Read 1361 times)
0 Members and 1 Guest are viewing this topic.
Little-Acorn
Silver Member
****

Kudos: 5
Offline Offline

United States United States

Posts: 221

Topic starter


Ignore
« on: March 27, 2018, 04:41:06 PM »

If the title sounds like I'm frustrated, that's accurate.

Most of my teams' WS planes fly best with the plane (with rubber motor) balanced at the center of the chord, give or take maybe 1/4 in. We've gotten some 2 minute flights with such planes, and are pretty happy. Those planes usually have the wing leading edge around 5" behind the propeller, or more.

There is a neighboring Middle School that uses kits from Freedom Flight Models. I was surprised to see their wing far enough forward, that the leading edge was 3" or less from the propeller. And the planes flew beautifully, smooth and consistent, with a gentle circling climb and then a slow descent as the prop uses up the last of the winds. They let me hold one of the planes, and found it was balanced exactly on the TRAILING edge of the wing.

I was astonished. I thought that a plane with such an "aft CG" would never be stable. It would either nose up and stall repeatedly, or else dive in an ever-steepening descent and hit the ground at a steep angle.

But the plane I balanced, flew beautifully, as I described.

Those planes have a lifting stab, one whose camber is slightly higher in the middle than at the edges (similar to the wing's). Obviously the stabilizer is carrying part of the load, lifting upward just as the wing does.

But with the CG that far aft, I thought that if the plane nosed up just slightly, the wing's increase in lift, times its lever arm to the CG, would exceed the stab's increase in lift times ITS lever arm to the CG, and the plane would nose up further and further - the classic case of static instability.

Yet  this plane apparently doesn't do that.

I noticed, though, that if they just glide the plane with no power, it does nose up abruptly and fall. But when they wind it up and fly with power, it does not.

Since then, I have tried to move my plane's wing forward, and even put together a lifting stab. It balanced exactly where theirs did, at the trailing edge... and when I wound it up and flew it, it either stalled repeatedly or dived in, as I described.

I must admit I am baffled. My major in college was Aerospace Engineering, with emphasis on small aircraft design. I went through all the how-to-calculate-your-plane's-stability-and-performance classes, and even aced most of them. But everything I learned there, said this plane could never fly. Yet I have seen the neighboring school's plane fly very well, consistently, time after time, as long as the propeller had power. And with the CG at the trailing edge of the wing.

I've also seen other planes in YouTube videos, with the wing very close to the front, also flying well.

How do they do it???

Logged
mkirda
Gold Member
*****

Kudos: 20
Offline Offline

United States United States

Posts: 710

WWW

Ignore
« Reply #1 on: March 27, 2018, 04:47:49 PM »

Simple!

The CG is in front of the neutral point.

{ducking}

Regards.
Mike Kirda
Logged

Prosper
Platinum Member
******

Kudos: 67
Offline Offline

United Kingdom United Kingdom

Posts: 1,952



Ignore
« Reply #2 on: March 27, 2018, 07:04:57 PM »

Quote
I noticed, though, that if they just glide the plane with no power, it does nose up abruptly and fall. But when they wind it up and fly with power, it does not.
Hi Little-Acorn; I, like Mike, am going to duck, but only after saying this:

Most theory ignores or does not adequately address real-world effects. It's easy to write some code that calculates static margin and required tail size but only so long as it assumes a 'perfect fluid' and ignores the little problems that crop up in real life, like er, gravity. . .like, eerrrr, viscosity. . .and like eerrrrrr, a propeller - oh, those pesky little things.

The propeller destabilises the aircraft. It destabilises it in both yaw and pitch axes. The bigger the prop or the higher the pitch (higher pitch = more blade area visible in side elevation or plan view) the more destabilising the prop will be. The more forward from the CG the more destabilising the prop will be. If you have a model that flies OK with the prop 3" forward of the CG it may well not fly stably with prop 5" forward of the CG. It's not to do with weight distribution but the "forward fin" effect of the prop. That's a misnomer: it could also be called "forward tailplane".

Quote
But with the CG that far aft, I thought that if the plane nosed up just slightly, the wing's increase in lift, times its lever arm to the CG, would exceed the stab's increase in lift times ITS lever arm to the CG, and the plane would nose up further and further - the classic case of static instability.
Again, perhaps in theory but not in fact. I can't say what's happening, but what if the increase in lift on the forward wing increases its downwash and that puts the rear wing in 'cleaner' air and it lifts more efficiently? Or as the aircraft pitches nose-up it slows down, which increases the effect of propwash, and perhaps the layout of the aircraft favours more lift on the tailplane?

Then there's the effect of the helical propwash, and there's the downwash effect from the wing, and there's the turbulent wake from the wing. The sum of these effects may be allowing your competitor's aircraft to fly well under power but then to stall when it glides with no power.

That's why I'm ducking - and maybe Mike too - all this is very complex and certainly not calculable by some static stability calculator. The exact shape of the prop, its RPM and torque; the exact aerofoil section of the wing and the tail surfaces; the finish of the surfaces; the exact geometric relationship between the wing and tail (and fuselage come to that), the pitch attitude of the aircraft in cruising flight and its weight distribution, all add up to its stability in pitch and yaw. I've probably forgotten some other factors.

At a wild guess, it reads to me as if a propeller 3" from the CG is aerodynamically more stable and is also concentrating the model's overall mass nearer the CG (less work for the tail to do) and perhaps the prop nearer the wing causes better accelerated airflow over the wing (= less turbulent wake) and perhaps the way the prop's helical wash acts on the tail surfaces is optimised too.

Regards,
Stephen.
Logged
mkirda
Gold Member
*****

Kudos: 20
Offline Offline

United States United States

Posts: 710

WWW

Ignore
« Reply #3 on: March 27, 2018, 07:26:29 PM »

Last week I spent at West Baden for the world F1D championships.
The first day and much of the second were the worst conditions for indoor I have ever seen.

Most of the planes could not handle the turbulence well. Flights often launched, stalled out and fell to the ground.
A few that did well tended to be shorter with smaller stabilizer area and the wing further back.
In other words, trimmed to have a more forward CG, but also stiffer than the longer planes.

One thing that was impressed upon me was the importance of having two planes, one that could handle bad air and one for good air (with a more aft CG). Also a lot of different sized props, but I digress from Wright Stuff.
Experience will teach you which to use based on observation of other flier’s planes.

Personally I like the odds of my design over the FF models one, because it is simple, rugged, and consistent.
Match the rubber to the prop and the conditions and you will do great.

Regards.
Mike Kirda
Logged

Tapio Linkosalo
Platinum Member
******

Kudos: 38
Offline Offline

Finland Finland

Posts: 1,373



Ignore
« Reply #4 on: March 28, 2018, 12:21:08 AM »


A long and detailed, but yet in laymans terms written description of the situation can be found in the classic book by Frank Zaic, "Circular airflow". The book dscribes Zaics experimental work to find out how outdoor power models can be flown both fast climb and slow glide, without variable incidence tail. The answer is in largish tailplane, small decalage and aft CofG. Zaic describes how it is not the static lift coefficient ratio between wing and tail, but rather the relative slope of the two as a function of angle of attack that dictates the stability of the model. In essence, also with an indoor model with aft CG, the lift from the wing increases faster as model gains speed and decreases faster when model pitches up and looses speed, hence it remains stable. For indoor it is important to have large enough aspect ratio for the tail to attain this.
Logged
Prosper
Platinum Member
******

Kudos: 67
Offline Offline

United Kingdom United Kingdom

Posts: 1,952



Ignore
« Reply #5 on: March 28, 2018, 07:45:55 AM »

Very interesting lead Tapio, thanks. Online I found this, https://archive.org/details/circularairflowm00zaic and since I have an account with Internet Archive I 'borrowed' the book. I now see the webpage says "this book is on loan" so I guess that's my fault. I believe I have it for two weeks but if I finish with it sooner I expect I can 'unloan' it, for anyone wanting a peek. It looks fascinating!

Stephen.
Logged
Hepcat
OOS, January 2019
Platinum Member
******

Kudos: 280
Offline Offline

United Kingdom United Kingdom

Posts: 1,777



Ignore
« Reply #6 on: March 28, 2018, 10:54:53 AM »

Little Acorn,
I have great respect for Tapio and Stephen but I don't think this is a circular airflow matter or a lot of little aerodynamic possibiities. I admire what Zaic did for aeromodelling and 'circular airflow' certainly helped in the case of looping with very high powered outdoor duration models but I can't see it as being relevant to low powered indoor models. I dismiss the lot of little aerodynamic effects because that would not answer why a perfectly flying model would stall violently from a hand glide. Actually the only reason I can see for that happening is that the thrust line is passing a long way above the CG (e.g. a lot of downthrust).
It sounds to me LA that you have done some studying on this subject and would like a better answer but none of us can find one without some sketches with dimensions on them.  I look forward to hearing from you.
John
Logged

John Barker UK - Will be missed by all that knew him.
planesjimmycam
Bronze Member
***

Kudos: 2
Offline Offline

United States United States

Posts: 84



Ignore
« Reply #7 on: March 28, 2018, 11:21:07 AM »

The simplest answer is the long moment arm, combined with a large horizontal stabilizer. A shorter moment arm, with a smaller stabilizer will not fly with an aft CG  at all.  I enjoy designing and flying indoor stick models, and have found that if you have a large horizontal stabilizer, with a long moment arm, then the plane will fly very well with aft CG. If you try to balance this same model at 50%CG, you will find that it needs alot of decalage in the wing to make if fly. It is fun to experiment with sheet balsa stick models, and vary the size of the horizontal stab, and see where it balances. I set the wing at about 3 degrees, and the stab at 0, and see where the CG will end up given the size of the stabilizer. Just my 2 cents.
 Best,
Jim
Logged
strat-o
Gold Member
*****

Kudos: 9
Offline Offline

United States United States

Posts: 488



Ignore
« Reply #8 on: March 28, 2018, 01:32:15 PM »

The majority of aircraft flying today fly with conventional layout and have tails producing negative lift.  These models with lifting tails change everything.  To understand it from your academic perspective, look at it from the point of view of a canard.  The stability of these model planes, I think, have more to do with how canards maintain stability rather than how most planes maintain stability.
Logged
piecost
Gold Member
*****

Kudos: 14
Offline Offline

United Kingdom United Kingdom

Posts: 687



Ignore
« Reply #9 on: March 28, 2018, 04:07:42 PM »

I have been pondering the same problem and decided to approach it from two directions. I used Bernie Hunts stability spreadsheet and also performed glide tests with different sized tails on my Bar Fly Legal Eagle. Glide tests remove the complication from propeller effects. I tried different tails without compensaring for the aft CG movement due to the smaller tail being lighter. The glide tests demonstrated that the smallest tail lacks static stability. In trimming it required adjustments of a fraction of a mm and only achieved a mushy stall or a fast glide. Infact, in a single flight the model alternates between each behaviour. The second smallest tail worked well and trimmed without being too sensitive to adjust and gave a stable flight even when rubber powered. It recovered quickly from a ceiling hit.

The spreadsheet calculator suggested that the oringal Bar Fly had a very low downforce on the tailplane and a relatively large stability. I don't nessesarily believe the absolute numbers from any calculator, but believe it good for comparisons. So, I concluded that the low lift/fownforce wa more important than the fact that the force could be acting up or down!

I concluded that the original tail was not contributing to lift (or downforce) and so made it as small as possible whilst maintaining adiquate stability. This saves weight and drag.

From comparison with the F1D models in the spreadsheet I noticed that they employed a significant lift on their tails. So, I made a new fuselage with the wing moved forward 20mm, employing the large tail. This had a significant lift on the tail (CLT=0.2 ish, compare to 0.3 for F1D models) and low static stability (as a result of moving the wing to increase the tail lift). The lifting tail is not good for efficiency (L/D) (it is most efficient to generate all lift with the wing) but reduces sink rate, thus power to fly, thus endurance (hopefully).

When the wing area is limited it is benificial to use a large lifting tail. When total wing+tail area is limited it is better to use a small tail on a long arm.  Glide and power tests showed that is is adiquately stable and with no worse duration than the previous model.

I like the approach of using a stability calculator, but it must be combined with tests to generate enough data to understand and verify the results.
Attached files Thumbnail(s):
Re: How do WS planes with "aft CG" fly at all???
Re: How do WS planes with "aft CG" fly at all???
Logged
ykleetx
Gold Member
*****

Kudos: 14
Offline Offline

United States United States

Posts: 652




Ignore
« Reply #10 on: March 28, 2018, 07:00:53 PM »

Wikipedia's entry is okay: https://en.wikipedia.org/wiki/Longitudinal_static_stability

You will see that an aircraft can be stable with a lifting tail. That is Lt (tail lift) can be positive.

In indoor modeling, we usually go with the largest stab allowed by the rules, as long as 1) the stab area is not used instead of wing area and 2) the large stab does not increase the model's weight beyond the minimum weight requirement.
Logged
TimWescott
Platinum Member
******

Kudos: 18
Offline Offline

United States United States

Posts: 1,296



Ignore
« Reply #11 on: March 30, 2018, 05:03:43 PM »

The CG is in front of the neutral point.

The CG is way less further forward of the neutral point than on a man-carrying (or RC, or CL) plane, because moving it back desensitizes the plane's pitch axis to changes in speed due to changes in thrust.  It also makes the plane much more sensitive to turning radius and variations in that CG location.  Back in the late 70's Dick Sarpoulis designed a motorized RC glider based on the then-prevalent FF designs, including the CG.  It was a disaster, because it was hugely sensitive to elevator and because when it was trimmed to fly straight, any rudder input would make it dive.  Moving the CG forward fixed all that, but in the end he could have just stuck an engine on the nose of an RC glider.

Get the Frank Ziac books, particularly Circular Airflow.  They're old, he's not an engineer, and yet they spell out the whys and wherefores in a way that an ordinary guy can understand, and an engineer can turn into "engineerese" if you feel so inclined.
Logged
ykleetx
Gold Member
*****

Kudos: 14
Offline Offline

United States United States

Posts: 652




Ignore
« Reply #12 on: March 31, 2018, 10:43:12 AM »

I agree with Tim and Tapio that Zaic's "Circular Airflow" gives a great description of longitudinal static stability. He gives multiple examples, and the description is very well done. I highly recommend it.

I wish the book were titled instead, "Longitudinal Static Stability", to better match the content. The second half of the book covers the airflow to a model while in a banked turn.
Logged
lincoln
Titanium Member
*******

Kudos: 40
Offline Offline

United States United States

Posts: 2,732



Ignore
« Reply #13 on: April 01, 2018, 11:44:15 PM »

-There is no magic change when the tail goes from lifting down to lifting up. The same principles apply. It's the difference between the wing and tail that matters. Look at it this way: Because the c.g. is ahead of the neutral point, it provides a fixed nose down moment. On the other hand, since the tail flies at a lower incidence than the wing, the combination provides a pitching up moment. As the airspeed increases, this nose up moment becomes stronger and overcomes the fixed nose down moment from the c.g. When the model slows down, the reverse applies.

-Testing without props will allow a c.g. somewhat further aft than would work with a prop.

-I suspect that other model mentioned in the first post, with the rearward c.g., had too much incidence in the wing relative to the prop and the stab. (This is the same as too much downthrust combined with too much up elevator.)

-I've had a few indoor models with the c.g. well back. As I recall, some of my EZB's had c.g.'s aft of the trailing edge. This isn't very surprising because EZB's have large stabs on long tailbooms.

-I'm not sure that most rubber powered models can get away with a c.g. that's much further aft than it would be on a full sized aircraft of the same configuration. As the knots unwind unevenly, the c.g. of the rubber can move fore and aft, and so does the c.g. of the model, to a lesser extent. So you need a bit of a safety margin.
Logged
Pages: [1]   Go Up
Print
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.1.21 | SMF © 2015, Simple Machines Valid XHTML 1.0! Valid CSS!