Discussion Thread: The design of a tailless flying wing

[Aerowerx]

Norman knows the theory, he successfully specified the flap size and location for one of my constant chord swept 'wing rc gliders.

With my higher aspect ratio 'wings I've found it's better if the flapped area goes all the way to the root. If the flap doesn't
extend fully inboard, I found you can experience a stall of the central, unflapped area of the wing. This manifests itself
as a sudden, uncontrollable pitch down of the nose just when you're getting ready to flare for touchdown.

The problem is that this would be right in front of the prop. OK for a sailplane, but not good for a powered plane.

I have the formula for a nontapered wing, but want to know what it is with a taper.

I thought I remember seeing a design that did not go all the way to the root, but can't find it.

 

[Hugh Lorimer]

Mike, any idea where? when? for first test?.of your wee "plank"......Hughie.

 

[Mike W]

Mike, any idea where? when? for first test?.of your wee "plank"......Hughie.

Hugh

I am hoping to have the aircraft finished around New Year and am hoping to be allowed to test it from Elvington airfield near York. This was a V bomber base with a 2.5 mile concrete runway, so that we can do short test hops without being committed to flying a circuit first time.

I will keep you informed

Mike

 

[Aerowerx]

I've been playing around with different parameters and happened to come upon a very good combination.

I didn't like what cutting the twist in half looked on the lift distribution, so I changed it to 66%. I also put the wing sweep back to 24 degrees from 20 degrees, and the cruise speed to 90 mph. This would then need a CL of 0.322.

At 10 degrees flaps I played around with the length and discovered that starting at span = 2 ft with a length of 6 feet, the pitching moment was 0.001 and the speed dropped to 75 mph with a CL=0.435.

With 30 degrees flaps the speed dropped to 60 mph, CL=0.722, and pitching moment=0.002.

So essentially these are pitch-neutral self-trimming flaps.

60 mph would make a nice landing approach speed, and 75 mph a decent rotate-on-takeoff speed.


Here is a screen snip of the wing center section, right rear quarter view, with 30 degrees of flaps.


That is showing the air stream. Lots of swirly things going on, which is to be expected.

The model isn't perfect, as I am getting a small amount of change in the zero-moment AoA with the different flap settings. I am also planning on having some small trim flaps to take care of these kind of problems (also any variations in CG), but so far things are looking pretty good.

 

[StarJar]

@ Aerowerks. The 66% sounds like a more practical amount.
In my opinion the final result supersedes the shape of the lift distribution.

I've been messing with the Roncz Low Drag Flying Wing Airfoil. I think with higher sweep angles it gains even more advantage because the sharper nose does not push the air sideways like a blunter nose (like a snowplow).
I have a fabric skin, but it seams to do OK for me at standard roughness. With your composite wing you could get the drag real low, I think.

 

[Aerowerx]

I have been playing around with the ailerons.

Here are some close-ups of the wing tips, with left aileron applied, showing the induced drag...

Right wing^^^....................Left wing^^^^

Note that there is an area of negative induced drag on the right wing tip, in other words, induced thrust, which will pull the right wing into the turn---a good example of proverse yaw.

 

[Aerowerx]

@ Aerowerks. The 66% sounds like a more practical amount.
In my opinion the final result supersedes the shape of the lift distribution.

The whole purpose of all that twist was to get proverse yaw, and I did not like the way things looked when I cut the twist in half. 66% looks a lot better. Also lowering the wing incidence to get the CL=0.2 didn't look good either.

I've been messing with the Roncz Low Drag Flying Wing Airfoil. I think with higher sweep angles it gains even more advantage because the sharper nose does not push the air sideways like a blunter nose (like a snowplow).
I have a fabric skin, but it seams to do OK for me at standard roughness. With your composite wing you could get the drag real low, I think.

That's one thing I need to look into, to see if there is a better airfoil than the Horten-like one.

At this time I am considering wood construction. Maybe a geodesic method like yours, except covering it with a thin plywood, or perhaps strips of plywood with no gap. Wasn't there a German WW1 plane built like that, and I remember seeing some newer one but don't remember where. Likewise I am thinking about a geodesic structure for the wing, to increase stiffness, and covering it also with thin plywood.

I just had an epiphany, Starjar. How about two thin layers of plywood on either side of strips of foam (for the pod)? That should be really strong and stiff.

 

[StarJar]

@Aerowerks;
Well the problem I have with plywood is that a 1/16 sheet weighs someting like 6 lbs. My wing for example is 150 sq ft. area so 300 sq ft surface. That's about 9 sheets and 54 lbs. The lattices would come out to about 18 lbs.total. My proposed design is a UL so, that pretty much made up my mind.

 

[skysoarer]

Hi Plncraze,
They are complete plans from Debreyer.
A French gentleman on this forum built one but cannot get any replies now. I may be wrong but believe it was difficult to be build down to the weight specified.
Hopefully he will see this?

 

[skysoarer]

Mike, any chance of seeing a picture of your plane please.
Hughie, apologies see how your steering and rudder system now work, that is, steering bar and rudder pedals. Will take picture once all installed

 

[Aerowerx]

StarJar,

I took a look at the Roncz airfoil that you mentioned.

The pressure curves certainly look better than the PRANDTL-D airfoil, at least at the nose. But when I tried it on my wing it actually came out slightly worse. Not by much, but there is a difference.

The biggest difference is in the CL value. Changing nothing but the airfoil I get CL=0.263 for the PRANDTL-D airfoil and 0.366 for the Roncz. That doesn't sound like much, but it makes a difference of about 13 mph in the cruise speed.

Changing airfoils is not difficult, but tedious. Because of the nonlinear twist distribution I have to define an airfoil every 1 foot across the span. Fortunately, XFLR5 can interpolate between two airfoils for you, and creates a new airfoil.

 

[Aerowerx]

Interesting....

I have been playing with self-trimming flaps in my virtual wind tunnel.

The more degrees of flaps you have the more stable it becomes.

I have the wing defined at 1 foot intervals, so I haven't been able to tweek it to perfection, but it goes from a speed of 89 mph with no flaps to 46 mph with 40 degrees of flaps with a pitch change of less than 0.8 degrees at zero pitching moment.

 

[pictsidhe]

I've been looking at flaps on my wing project too.
The Nickel section is for a swept rectangular wing with an elliptical lift distribution. I pondered working out a formula for a tapered BSLD, but was lazy and did a crude integration in my spreadsheet instead, gets me in the ballpark. I suspect that for a sin³ distribution on a rectangular wing, the 8/3pi term wants to be 2/3.
I don't think the flaps do need to span the MAC, but the additional lift needs to be at the same longitudinal place.
Other things. I have been looking at the required twist. It's a lot at high Cl, I keep thinking about wing warping as control as a difference in twist of about 13deg between cruise and stall is going to need some compromises. Other than patent issues, were there other reasons for the wing warping demise? My sexy Santa is hopefully bringing me an aero elasticity book...

 

[Aerowerx]

...Other than patent issues, were there other reasons for the wing warping demise?...

I would think that it was mainly because of structural difficulties in making a flexible wing that was strong enough for the higher performance planes.

 

[Aerowerx]

I still have not found an equation or procedure for designing self-trimming flaps on a tapered swept wing.

But I did find this interesting paper that discusses taper, sweep, and aspect ratio with regards to self-trimming flaps on a tailless aircraft. It was first published in 1944, and then re-released after WW2 as part of a collection of war-time research.

Although it does not give any specific design details, there are a number of charts that show how the different parameters interact. And it turns out that what I am finding in my virtual tests agrees pretty close to what is in the charts as "optimum". For example, they say a taper of 0.33 to 0.5 is "structurally optimum" (whatever that means). And a sweep of 20 degrees with flap chord of 30% requires a flap span of about 56%. I am currently using a taper of 0.5 with 18 degrees sweep. My flap chord is 30% with a flap span of 45%.

 

[Hot Wings]

I still have not found an equation or procedure for designing

It's these kind of problems that prompted me to redouble my efforts to become fluent in MatLab. Excel is great for some things but for setting up an iterative process something like MatLab is much easier to 'program' - at least for me.

 

[pictsidhe]

Aerowerx, that is a very interesting paper! For non elliptical lift, the optimum taper is probably different, I'm thinking a bit lower.
I am ballparking my flaps based on the Nickel book. He states that flaps will be pitch neutral if the additional lift from them acts at the same x ordinate as the centre of lift. On elliptical lift, the centre lift is 4/3pi out on the 1/4 chord. For sin³ it is 1/3 out. I call that the B point. Nickel is allergic to BSLD, so most of his stuff needs a little tweaking...
I didn't do a formula as I am looking at non straight wings with a probable variable flap chord.
What I did was for each wing station, calculate where the additional lift would be relAtive to the B point, what additional lift was and get a moment by multiplying them. Sum them across the whole flap and you have the flap moment. root station was halved as I'm just doing half the wing. Tweak the flap till it is zero. Not super accurate, but I just need the ballpark ATM. It's producing believable results.
I have an idea that the moment produced by split flaps depends where they are, move them forward, their moment drops. I cant find any papers on it though.

 

[pictsidhe]

BSLD vs elliptical. The pros and cons.

Induced drag. Di
Structural weight. m
Span. b

Sin³ lift distribution. BSLD
Elliptical lift distrib. Elliptical

b fixed. Space BSLD vs Elliptical

Di. Spaceeeeeeeeeee >
m. Spaceeeeeeeeeeee<

Di fixed

b. Spaceeeeeeeeeeee >
m. Spaceeeeeeeeeee <

m fixed

Di.Spaceeeeeeeeeeee<
b. Spaceeeeeeeeeee >

Yaw Spaceproverse Spac adverse

Twist due to downwash

Spac Varies with Cl Space constant
Space
Edit, 'I carefully spaced all that to have my formatting eaten
Edit 2, 'I forgot one
Edit 3-? Thanks hotwings!

 

[Hot Wings]

Edit, 'I carefully spaced all that to have my formatting eaten

The forum software likes the taste of spaces. Insert place holders of your liking xxxxxxxxxxxxxx and then turn the font white. There is another method that has been posted, but I can't remember what it is.

 

[pictsidhe]

Unless you have a physical span limitation, the only bad thing about BSLD is the required twist varying with CL.
Away from the design point, the angle of attack goes to hell at root and tip.