Quantcast

My Random Ideas

HomeBuiltAirplanes.com

Help Support HomeBuiltAirplanes.com:

addaon

Well-Known Member
Joined
Feb 24, 2008
Messages
1,686
Location
San Jose, CA
I'm starting this thread as a "sneak preview" of an idea I'm working on. It's still very much in the conceptual stage, so absolutely everything is subject to change. Still, I thought having a couple more images on the forum would never be bad.

The goal is to build a rather light weight (900 lbs with payload), medium powered (85hp - 95hp), moderately high speed (60 knot stall, 160 knot cruise) single-seat tail dragger that is quite different than what's out there today. To make that explicit: Some of the very basic design choices are being explored simply because I haven't seen them explored before, and I'm curious what the consequences are. A guiding force in many design choices is mechanical simplicity.

The basic design is a high wing loading forward swept flying wing. Controls are outboard elevons (forward of the CG, so deflection is similar to that of a short-coupled canard) and a vertical tail. Wing span is 18 feet. Longitudinal stability in the current design is marginal at high speed, but positive. Lateral stability has not yet been assessed.

In any case, too much text would be silly, since most of it won't be true any more next week. So here's a top view of the planform. Isometric view blocked on tail sizing and dihedral design.
 

Attachments

orion

Well-Known Member
Joined
Mar 2, 2003
Messages
5,800
Location
Western Washington
At first glance this is an interesting concept but a closer look at the physics and flight requirements suggests several less than favorable characteristics. First and foremost, you are right in that the design will probably need to be configured in a manner similar to that of a canard. As such, it will be of paramount importance that the forward part of the wing span stalls first. Translated, this means that you need to force tip stall, which in an airplane can be anything from an annoying to a downright dangerous condition.

But here the case is further complicated since you are also incorporating roll, which means that at low speed a simple roll control input can instigate the tip stall in an disproportional manner. Since this will most likely happen at low speed, which is mostly in the pattern, an unpredictable stall is simply what you don't need since it might lead to premature contact with the ever present immovable object.
 

addaon

Well-Known Member
Joined
Feb 24, 2008
Messages
1,686
Location
San Jose, CA
Orion, I agree that tip stall is a problem, but I must respectfully disagree that it's a show stopper. Besides tip stall issues, there are a couple other potential difficulties with the design. However, I currently still suspect that they're all solvable.

The first issue, when looking at a forward swept wing, is the potential for aeroelastic divergence (or, approximately, the wings ripping off). The problem is minimized in the design by minimizing forward sweep, increasing wing rigidity, and (primarily) by running at relatively low speeds to decrease forces. Especially compared to a "normal" pure composite design, the total structural weight of this design will be heavier, as it is stiffness critical rather than strength critical. I have not yet run the numbers on just how much weight designing composites for stiffness will add, but I suspect this will be within tolerance.

The second issue is that of adverse yaw. The basic geometry of the airplane tends to exaggerate adverse yaw, as rotation around the vertical (yaw) axis is relatively quick, due to a short fuselage and empennage. The use of elevons makes aileron differential impossible (as this would cause pitch roll coupling, decidedly worse than yaw roll coupling) and the use of Frise elevons is impractical (as drag would be increased throughout most of the flight envelope). I have decided that I'm comfortable solving this particular problem by brute force. The rudder (and ONLY the rudder) will be a torque-motor driven fly-by-wire surface hooked up to a custom system (I'm an EE/CS guy in my other life) to auto-coordinate turns. Manual (pedual?) rudder input will be fed into the system to drive the rudder to deliberately slip or skid, allowing the use of normal rudder inputs when uncoordinated flight is required. No attempt will be made to provide input feel for the rudder.

The third issue is that of tip stall, as you mention. Stalling from the tip is bad, which means we cannot stall from the tips first. Stalling from the root, in a forward swept design, is much worse (lift decreases behind CG, tail drops, unrecoverable deep stall), which means we cannot stall from the root first. As such, the goal of the design is to tailor the stall characteristics (primarily through airfoil choice) so that the stall begins at the 55% outboard spanwise position. In the planform above, this at the inboard edge of the elevon. By using a relatively sharp-nosed airfoil in this section (to cause a sharp stall from the leading edge), and potentially with the addition of a leading edge device, a stall can be induced to start here first. Due to the sweep, the stall seems to propagate primarily towards the root. With up to 30% of the wing stalled in this manner, there is still a well defined nose-down pitch; more than 45 degrees of control surface deflection is required to hold the nose up beyond this point. As of now, it's unclear to me whether the stall behavior of a wing can be tailored carefully enough (especially in skidding or slipping flight) to get the characteristics described here, but I'm still optimistic. In particular, the CL distribution across the span at the stall point is well within the range where it's not hard to pick airfoils that are unstalled for the tip section, and not hard to pick pretty somewhat laminar flow airfoils for the center section that stall sharply at the desired point. The root, due to its increased chord and decreased loading, should be quite easy to make stall after even the tips, and small strakes can improve this if necessary. In summary, difficult, but maybe possible.

Orion, your insightful posts are probably the main reason I've started reading (and now writing to) this board; I welcome additional feedback, and will probably ask more specific questions as this idea matures. (And will probably contact you via e-mail at some point about lofting and such, when the design gets to that point.)
 

addaon

Well-Known Member
Joined
Feb 24, 2008
Messages
1,686
Location
San Jose, CA
One more explicit question for Orion or anyone else... does 160 knots seem achievable for a clean design at 90 HP (10 lbs/hp)? At 160 knots, load distribution is eliptical, and all surfaces are at neutral position. Average CL is about 0.225 due to relatively high wing loading.
 

Topaz

Super Moderator
Staff member
Log Member
Joined
Jul 29, 2005
Messages
14,239
Location
Orange County, California
Just out of curiosity, where is the prop? I'm sure 90hp/160kts is achievable, but can it be done and still retain enough climb rate for safe operations? That'll be the question.

Oh - one last question - why a tail-dragger with this configuration? It's been done (the Quickie series were canard taildraggers, as was the Dragonfly), but it rather artificially extends the runway length requirement, since you'll have to lift down with the wingtips to get the tailwheel off, or wait until enough speed builds up for it to float up on its own. If you're okay with that consequence, then fine, but it's a bit odd with this kind of design.
 
Last edited:

addaon

Well-Known Member
Joined
Feb 24, 2008
Messages
1,686
Location
San Jose, CA
Is the CFD software you're using fairly sophisticated? Enough to adequately model separated flow and turbulence effects? ... Have you been able to adequately model separated flow effects enough to preclude a pitch-up like this?
Nope, this is all speculation at the moment. I've been playing with my own codes (which are an incorporation of various rules of thumb, nothing formal) and VLM codes (which don't do separation). Figuring out if I can actually control stall to the level I need is definitely the next question, and I'll need to get/build new tools for it.

Just out of curiosity, where is the prop? I'm sure 90hp/160kts is achievable, but can it be done and still retain enough climb rate for safe operations? That'll be the question.
Classical tractor prop sitting at the front of the fuselage, where it belongs. Constant speed propeller, no question.

Oh - one last question - why a tail-dragger with this configuration?
Because tail wheels are more fun! Also, I don't really think I'm going to be runway length limited... the shortest runway in my area is 2400 ft of nice concrete, which is an awful lot for a tiny little plane. Also, because the tips are so highly loaded at low speed, the angle of attack at stall (relative to angle of attack at cruise) is only 10 degrees. Acceleration with the tail down (minimum control drag) should be reasonable, and I'm planning on not picking up the tail until reasonably high speed. A locking tailwheel will be used.

That said, I'm not completely wedded to the tailwheel idea, and would change it in an instant if I was convinced that it was impractical. The design as a whole (forward swept flying wing) is not dependent upon this detail.
 

Topaz

Super Moderator
Staff member
Log Member
Joined
Jul 29, 2005
Messages
14,239
Location
Orange County, California
Sorry addaon - I deleted out part of my post that you quoted and hit 'save' some moment just after you posted your reply. My apologies - I decided that the CFD and implications of this kind of configuration really are outside my level of knowledge.

Tractor prop, eh? Destabilizing power-on, you know. ;)

Taildraggers are, indeed, fun. My father and I once had a 1941 Aeronca 65LB Super Chief. Didn't get to fly it a lot, but it was a blast when we did. I personally would put a nosewheel on an airplane like yours, for the simple reason that takeoff will be a simple 'add lift, leave the ground' proposition once you're at flying speed, rather than 'wait for the tail to lift up, then take off.'
 
Last edited:

addaon

Well-Known Member
Joined
Feb 24, 2008
Messages
1,686
Location
San Jose, CA
Tractor prop, eh? Destabilizing power-on, you know. ;)
Yeah, but the fuselage is relatively short... engine is only about three feet in front of CG. And I just don't see any way to get even marginally acceptable ground clearance with a pusher in this configuration.
 

PTAirco

Well-Known Member
Joined
Sep 20, 2003
Messages
3,562
Location
Corona CA
With all the challenges that this design poses what is the pay-off of such a configuration in real terms? Having a cool-looking airplane is one, I'm sure, and there is the "just to be different" aspect, but in what area does this layout have definite advantages over a conventional design?
 

orion

Well-Known Member
Joined
Mar 2, 2003
Messages
5,800
Location
Western Washington
Well, actually I didn't mean to imply that it wouldn't work - my premise was just that there are specific requirements of this type of configuration that would have to be addressed and very carefully analyzed prior to anyone actually sitting in the pilot's seat. With time and with numerous iterations, it is quite possible that you could configure a solution that would minimize the issues of unsymmetrical stall however, I still think that this will be difficult to do given that you are getting pitch and roll from the same control surfaces - surfaces that have to have a predictable stall behavior.

Many years ago there was a symposium where the subject was breached regarding the Eze configuration. The premise was that it is a layout that could've potentially been a good candidate for this type of control since on the first Ezes the roll control tended to be relatively heavy, especially as compared to pitch. It was therefore suggested that the canard be used for both axees, thus dramatically simplifying the build (of the controls) and potentially removing a possible stall instigation device from the main wings.

The suggestion had some merit but it was eventually proved that a control mixture of that type would have dramatically reduced the pitch authority because now you had to limit the canard flap deflection for either control in order for it to be effective for both pitch and roll yet without the danger of stalling one side (the pricipal result would've been a dramatically larger canard). Given the particular loading of the Ezes (and especially the original design), an unsymmetrical stall with the canard was determined to be a very strong possibility.

Given your speed and the airplane's likely low inertia in pitch, structural divergence was not something I considered significant, especially for a composite arrangement. I know you didn't state the material of construction initially but I just assumed such given the relatively complex wing geometry.

Regarding adverse yaw and yaw stability, yes I missed that one since my thoughts were mainly concentrating on the lift characteristics. And I think you're right, this will probably require a brute force application of a sizable and/or very effective vertical geometry. I rarely suggest this but one possibility might be to incorporate two verticals located at the outboard ends of the centerline trailing edge "V", or maybe even further apart yet.

I might however have a bit of concern when you say the rudder will have no "feel". If the airplane has a bit of a low yaw stability characteristic, the lack of feel may have you chasing the ball, so to speak, rather than providing you with meaningful feedback. Yes, with sufficient modeling and flight testing you might get your software just so a proper amount of input is instigated within your electromechanical controls but don't forget that maneuvering is only a small part of the need for feedback - motion in yaw as a result of turbulence, especially on approach, might be difficult to control without at least some level of being able to gage how much deflection you're generating. Eventually this might come with just spending enough time in the airplane but initially this familiarity wont be there.

I agree with your reasoning about the stall progression but again, keep in mind that you'll most likely have to meet the separation requirement with the controls deflected in any combination, Off the top of my head, this might be tough to do unless you maybe figure out (as you indicate) how to instigate a leading edge stall. But then that could significantly affect your CLmax. You might have to use some form of retractable stall strip that deploys in concert with the control surfaces - but again I'm just thinking out loud.

As far as the performance is concerned, that's a bit more difficult to say since the amount of power you'll need will be a balance between the cruise, your climb and of course, your take-off. Off the top of my head, it might be a bit optimistic but who knows - only until you get deeper into this process will you be able to judge whether this'll meet those requirements. Given the unique configuration and its functionality, it is difficult to make off the cuff evaluations or statements regarding things like how fast it might go.
 

Topaz

Super Moderator
Staff member
Log Member
Joined
Jul 29, 2005
Messages
14,239
Location
Orange County, California
...Many years ago there was a symposium where the subject was breached regarding the Eze configuration. The premise was that it is a layout that could've potentially been a good candidate for this type of control since on the first Ezes the roll control tended to be relatively heavy, especially as compared to pitch. It was therefore suggested that the canard be used for both axees, thus dramatically simplifying the build (of the controls) and potentially removing a possible stall instigation device from the main wings.

The suggestion had some merit but it was eventually proved that a control mixture of that type would have dramatically reduced the pitch authority because now you had to limit the canard flap deflection for either control in order for it to be effective for both pitch and roll yet without the danger of stalling one side (the pricipal result would've been a dramatically larger canard). Given the particular loading of the Ezes (and especially the original design), an unsymmetrical stall with the canard was determined to be a very strong possibility....
Rutan actually tried it on... I'm almost positive it was the prototype VariEze, but it may have been the prototype LongEze - one of his many attempts at simplifying the control system, along with the 'rhino' rudder tried out on the LongEze and Defiant. I've seen drawings of the pitch/roll control setup for that 'elevons' on the canard - extremely simple. However, the flight control characteristics were poor, and IIRC, pretty much exactly along the lines you've listed here. It was abandoned after only a few flights.
 

mikemill757

Well-Known Member
Joined
Feb 29, 2008
Messages
50
Location
Portland metro, Oregon
I've toyed with this idea for some time, almost to the point of building a 1/4 scale R/C model and seeing how it does, but the adverse yaw looks insurmountable - the thing will want to wander and never fly straight. Maybe a swept aft flying wing would be better, but that's already been done...Dyke's Delta and a delta Kitten come to mind, neither seem to have caught on. Good luck, and seriously consider a nosewheel - why add more complication to a unproven design?
 

ragflyer

Well-Known Member
Joined
Apr 17, 2007
Messages
234
In addition it appears to me the span efficiency would be terrible, what with the mid span highly loaded, followed by the tips and then the root.
 

addaon

Well-Known Member
Joined
Feb 24, 2008
Messages
1,686
Location
San Jose, CA
Ragflyer, note that this design is within 1% of an elliptical load distribution at cruise speed.
 

Topaz

Super Moderator
Staff member
Log Member
Joined
Jul 29, 2005
Messages
14,239
Location
Orange County, California
One thing you should consider - an elliptical load distribution is most needed at the best-rate-of-climb airspeed, where induced drag plays the largest part in performance. At cruise, parasite drag in all its forms predominates, and induced plays a much smaller part.

As such, you'll need to make that happen with the elevons deflected for trim at best rate of climb. Being a flying wing, the lift distribution should be fairly constant for unaccellerated flight, but you'll want to optimize it for climb, not cruise.
 
Last edited:

ragflyer

Well-Known Member
Joined
Apr 17, 2007
Messages
234
Ragflyer, note that this design is within 1% of an elliptical load distribution at cruise speed.
Are you saying it has a span efficiency e of 0.99? Cannot see how you could do this and be stable and have good stall charecteristics. I guess you could play around with a vlm code like lamdes to work out the optimal twist.

Also, as topaz says you need to calculate e at climb cl (not cruise), as that is where induced drag has the biggest impact.
 
Last edited:

Hugh Lorimer

Well-Known Member
Joined
Feb 14, 2008
Messages
298
Location
Stair, Ayrshire, Scotland.
Hope this works? I attach a schematic of an elevon mixer idea I had some time ago on a cold Scottish night. It was for a close coupled single seat joined wing canard. A model glider was made and flown, but due to the lack of access for a normal sized pilot the idea was not continued.
 

Attachments

Hugh Lorimer

Well-Known Member
Joined
Feb 14, 2008
Messages
298
Location
Stair, Ayrshire, Scotland.
Hope this works? I attach a schematic of an elevon mixer idea I had some time ago on a cold Scottish night. It was for a close coupled single seat joined wing canard. A model glider was made and flown, but due to the lack of access for a normal sized pilot the idea was not continued.
TRy again to post mixer.

Hughie
 

Attachments

Hugh Lorimer

Well-Known Member
Joined
Feb 14, 2008
Messages
298
Location
Stair, Ayrshire, Scotland.
Hope this works? I attach a schematic of an elevon mixer idea I had some time ago on a cold Scottish night. It was for a close coupled single seat joined wing canard. A model glider was made and flown, but due to the lack of access for a normal sized pilot the idea was not continued.

See if I can improve the quality of the mixer Drg.?
 

Attachments

Top