A question for all Flying Flea aficionados

HomeBuiltAirplanes.com

Help Support HomeBuiltAirplanes.com:

wsimpso1

Super Moderator
Staff member
Log Member
Joined
Oct 18, 2003
Messages
6,778
Location
Saline Michigan
Hi all,
First, thank you for your many considered replies. And a BIG thank you all for staying on topic. It usually only takes two or three posts before someone hives off into some obscurely related topic, and the original question/comment gets forgotten entirely.

Now - tdfsks (what do these letters stand for, BTW?) - you certainly got me thinking, and backed up by Billski, I think I finally have the basic principles sorted in my head.

This is the configuration of the HM293:
View attachment 98975
Interestingly, the pivot point is 40mm ahead of the c/4, and the Incidence Control Arm is set back 400mm from the pivot. With this arrangement, although a low CM airfoil is advantageous, the pitching moment of the airfoil doesn't matter nearly as much as the lift of the wing itself. Any lift at all will raise the rear of the wing. This is experienced by the pilot as a pull on the stick, and he/she will have to pull back against it. The so-called "living wing" for which Fleas are famous.

The extreme case (the pivot right on the LE), would see the entire wing pivot upwards, and with the increase in lever arm (i.e. 300mm instead of only 40mm), the pilot would in all likelihood not be able to exert sufficient force to keep the wing from pivoting onto its head, as it were.

If the pivot point is *behind* the CoP, then any lift would swivel the wing downwards.
View attachment 98976

As it is, the Mignet "Formula" (as verified by the plans) seems to place the pivot point slightly *ahead* of the CoP, and to place the Incidence Control point rearward of this creating at least a 10:1 moment arm. This now makes a lot of sense.

My own design also uses a 1200mm chord, but places the Incidence Control point on the trailing edge. This means, to keep to the same "formula" (and hence, "feel"), the pivot point needs to be 10% of the Pivot-Incidence distance ahead of the CoP. Like this:
View attachment 98979

This would require a bracket affixed to the spar extending some 160mm or so forwards, with possibly a number of holes to fine tune the pivot point. I *think* my reasoning is sound. Comments welcome, of course.

Regards
Duncan,

You gotta know I am going to try to talk you into another perspective here...

Flying Flea design - With 160 kg (we won't insist on forces being in Newtons right now) supported 0.04 m behind .25c point, the original wing makes 6.4 kg*m of nose down moment. And the mechanism for Incidence control of this foil must counter that moment while resulting in a reasonable force at the stick and within a reasonable stick travel for the entire operating range of that foil.

Now we go to your design - Same chord of 1.2 m, and same lift of 160kg but the distance to pivot from .25c is now going to be 100 mm. The moment to be balanced by stick force will now be 16 kg*m, 2.5 times as much moment as was carried previously. I contend that now your control system must control a device that has the same amount of rotational travel but operates at 2.5 times as much torque. This absolutely means that either your stick forces at the same travel have to be 2.5 times as high or if you play with the gear ratio to get the same stick forces, you will need 2.5 times as much travel. There are an infinite variety of design options between the two mentioned, plus more outside that range.

The location of where the connector attaches to the foil is immaterial to the force and travel choices being presented. You can change ratios anywhere along the system to get a good trade, but if you increase the arm that lift uses to create a moment, the moment must go up. I strongly suggest that if you place your pivot point 100mm forward, you will have heavy controls or inadequate travel or both. If you are really married to placing the pushrod at the trailing edge, set the control forces and stick travel you want and then work the mechanical advantages to have that match the forces and travel of the foil.

Billski
 

rtfm

Well-Known Member
Joined
Jan 3, 2008
Messages
3,233
Location
Brisbane, Australia
Billski,
Ha ha. I'd expect nothing less from you...

Your final suggestion is precisely what I intend to do.

With this Flea-Bike design, there just isn't room anywhere in front of the pilot for the Incidence Control rods to go. I'm left with two choices: (1) follow the rather clumsy HM360 Incidence Control path - or similar (fitting a bellcrank to the forward wing strut - rather poorly drawn in the plans, unfortunately) - which seems messy to me.
1594097102250.png

(2) Route the control rods under the pilot's seat, and up behind him. This seems the neater option. Especially since the wing trailing edge lines up rather neatly with the back of the seat.
 
Last edited:

wsimpso1

Super Moderator
Staff member
Log Member
Joined
Oct 18, 2003
Messages
6,778
Location
Saline Michigan
While the argument can be made that a trim spring will relieve the control stick loads regardless of size, increased arm - distance between .25 c and pivot - the control gradients (stick force per unit movement of the stick) and energy - stick forces through travel - will also grow with increased arm. Then there is the fact that the pitch response of the foil to vertical gusts will also be substantially increased by increased arm.

I would be very careful about moving away from the well established proportions of the Flying Flea. By all means experiment with more arm, but please do so carefully and do not start with big changes...

As to making the controls work, each of the several bell cranks can be places for making the gear ratio - or mechanical advantage - appropriate for tuning forces and travel of the system.

Billski
 

Victor Bravo

Well-Known Member
Joined
Jul 30, 2014
Messages
7,447
Location
KWHP, Los Angeles CA, USA
Agreed 1000% on the "start with very small changes to what is known to work" approach.

Even better, This would be an ideal case for R/C model testing in my opinion. Make a 1/4 or 1/3 scale quick-built foam model... the cheapest white picnic cooler foam and packing tape with a couple of fiberglass "arrowshaft" tubes, not much more.

Make all the pivots and arms and such with several mounting holes, so you can move the main wing pivot location and control arm attach easily. Just move the pivot bolts into a different set of holes in the same piece of plywood or aluminum angle.

This will give you very high quality data, because the rest of the model is identical... it's the exact same model, but with the pivot moved form X to Y.
 

TFF

Well-Known Member
Joined
Apr 28, 2010
Messages
13,028
Location
Memphis, TN
The only issue for an RC, is the wing will be governed by the servo. The servo will move the wing without regard to effort. It will not float like a person would have to control. So a pivot change for effort is not easily done RC for measurement. You could set it up to where it overpowers the servo, but outcome will not be reusable.
There was a nice 1/4 Scale Pou RC plane back in RCM in the early 80s. 1/4 scale and small at the same time.
 

Malcolm C

Member
Joined
May 18, 2020
Messages
10
The Falconer airfoil has the front spar centre line at 23 cm aft of the leading edge and the rear spar cl 40 cm behind that. I like the idea of having an adjustable pivot point and maybe similar for the actuating rod. I think I will test fly it at the recommended positions first and assuming I survive will fine tune it later. Nice to see there is still some interest in the little "Pou "
 

rtfm

Well-Known Member
Joined
Jan 3, 2008
Messages
3,233
Location
Brisbane, Australia
Hi Billski,
You said:
I would be very careful about moving away from the well established proportions of the Flying Flea. By all means experiment with more arm, but please do so carefully and do not start with big changes...
I will check (as far as possible) but I strongly suspect that this measurement (like almost everything else in the world of Pou's) is different for different designs. It is convenient to talk about the "formula" (or in this case the "well established proportions"), but these are rules of thumb only, and can differ quite widely from one plane to another. It would be important, however, to establish if there is a proportionality of these measurements across different planes.

Question:
I need to brush up on my theory of levers (because, to be honest, I know only the rudiments) - but my understanding is that in the case of the HM293 quoted above which has the pivot 40mm ahead of the c/4, and the control arm 400mm rearwards of the pivot point - would this not be identical to a 100mm - c/4 - 1000mm arrangement, except that the gearing of the bellcranks would need to account for the greater travel required? Would the stick pressure not be identical?

Regards,
Duncan
 

tdfsks

Well-Known Member
Joined
Aug 29, 2005
Messages
93
Question:
I need to brush up on my theory of levers (because, to be honest, I know only the rudiments) - but my understanding is that in the case of the HM293 quoted above which has the pivot 40mm ahead of the c/4, and the control arm 400mm rearwards of the pivot point - would this not be identical to a 100mm - c/4 - 1000mm arrangement, except that the gearing of the bellcranks would need to account for the greater travel required? Would the stick pressure not be identical?
Billski is right .... the control forces will change.

The best way to think about this is to use the engineering principle of "Work" and treat the control system as a black box containing some combination of bellcranks and pushrods to achieve the desired drive ratio. Work = Force x Displacement and is, in effect, a measure of energy. The input and output work in this control system must be the same (ignoring friction that is). The input work is the pilot force on the stick x displacement of the stick. The output work can either be:
1. The force in the control rod attached to the wing x the displacement of this rod
2. The lift force on the wing x the distance we are moving that.

If you consider the case of the push rod actuating the wing, you have scaled the distance of the pushrod from the pivot and the lift (c/4) from the pivot so that the force remains unchanged in the control rod. However, for a given wing rotation angle (hence given change in lift) you have to move that rod further when at the trailing edge compared to the original location at 400mm. In fact you have to move it 1000/400 = 2.5 x as far. So the work done is higher (i.e. same force x 2.5 times the distance). Since the output work has increased, the input work at the control stick also needs to increase and therefore you will either have a higher stick force for the same movement or a bigger stick movement with the same force.

Consider the other case where the output is the movement of the lift. If you pull back on the stick, the result is that the wing lift has to move a distance downward as the wing rotates. Assume the lift acts at a point on the wing. If the wing rotates, the point at which that lift vector acts on the wing moves. The work done is the lift x the distance moved (the distance moved should be the component of the total distance moved along the direction of the lift vector noting that a point on the wing is actually following an arc). Anyway, the distance that the lift vector moves increases with the distance from the pivot. So if you increase the pivot position from 40mm to 100mm the lift vector translates 100/40 = 2.5 times as far for a given rotation of the wing. So you need 2.5 x as much work. As for the previous case input work needs to increase be a corresponding amount.

I hope this way of explaining it makes this clearer rather than trying to visualize / analyse a complex arrangement of bellcranks .....

Whether a 2.5 x increase in stick force is objectionable depends on what the stick force was in the first place. That is another story that does, I think, demand a more detailed analysis of the control system ....
 
Last edited:

rotax618

Well-Known Member
Joined
Oct 31, 2005
Messages
864
Location
Evans Head Australia
Provided you stick with the accepted Flea airfoil I think your Flea will have acceptable control forces. I never flew my Flea enough to investigate all of its flight regime, I had no bungee and the stick load was light when you reached flying speed, I was never game to let go of the stick but found the Flea to be very stable in pitch, my problem was the delayed yaw/roll couple. My flea was a home brewed HM14/293.67A8FC73-EACF-4073-BAE7-A43E7F821BC2.jpegECF54BB7-ABE6-45F1-AB20-347125CCD503.jpeg
 

rtfm

Well-Known Member
Joined
Jan 3, 2008
Messages
3,233
Location
Brisbane, Australia
Nice looking Flea!
I need to revisit my Flea thread - I think I called it "Duncan's got a Flea" or "Duncan's doing a Pou" - Ha ha.

Regards,
Duncan
 

rtfm

Well-Known Member
Joined
Jan 3, 2008
Messages
3,233
Location
Brisbane, Australia
I think what I might do is mock it all up, and do some testing. TFF put his finger on the problem with testing this sort of thing on an RC platform. But some full-size MDF-based mock-ups will do the trick I think.

Regarding my comment about every Pou being different, and that there is actually no definitive "formula" - I note that the Falconar plans say 230mm from the LE, the HM293 plans say 260mm, and the HM360 plans say 275mm. But they are all in the ballpark. I can't wait to test this rig out in the workshop. If I think about it hard enough, I'm sure I can come up with a way to photograph the process and tabulate the results.

Duncan
 
Top