Thread: Discussion High AR flying wing for soaring
View Single Post
Old Dec 21, 2010, 12:31 PM
dayhead is offline
Find More Posts by dayhead
San Bernardino, California, United States
Joined Oct 2004
1,967 Posts
While it seems to be a nice little airplane, according to your impressions, it is no longer what I call a flying wing. To me, a "flying wing" is just that- an aircraft where all of the necessary stability is built into the wing all by itself.
That said, I'm glad you've engineered an efficient and fun flying machine. What is interesting is how, like me, you are gaining insight into the aerodynamics of flying wings by experimenting and drawing conclusions based on the results of your experiments.

Keep on Truckin'!

I came up with a free-flight flying wing design in the early '80's, which I patented and later was marketed by others as "Sling Wing". This all sheet foam little glider is equipped with a spring-loaded hinge at it's center, that allows it to be folded for catapult launching. This allows the glider to be launched to as high as 80', so that it's glide characteristics can be observed over time. I taught myself a lot with them during the early months of it's development.

One thing of note is how the shape of the lift distribution curve affects the lateral handling qualities of swept back flying wings. The efficient elliptical curve, which results from using the aft end of the usable CG range, can result in a "control reversal" phenomenon in regards to aileron input. If the CG is far enough forward to cause the outboard area of the elevon or the entire wing tip to be negatively loaded, then the Lift distribution curve becomes "bell shaped" and good predictable roll control will be experienced.
The problem is that the bell shaped curve is inherently less efficient. You'll get the best L/D by using the elliptical curve.
Please keep in mind here that I'm referring to a "pure" flying wing that has no vertical stabilizers.
Anyway, to cut to the chase here, the elliptical curve will work very well (in regards to roll control) if enough anhedral (negative dihedral) is incorporated into the design. The anhedral must be of sufficient value to at least cancel out the dihedral effect found in swept-aft planforms, which is generally considered to be about 1 degree of dihedral for every 5 degrees of sweep.
Also please keep in mind that I'm referring to a flying wing that is intended to be flown up-right, most, if not all of the time. The flying wings usually being discussed on this forum are mostly intended for high speed aerobatic flight. 99% of my flying wings have been non-powered or only minimally powered gliders.

I adjust my anhedral angle so that the glider exhibits a little bit of negative spiral stability, it wants to roll in to the turn, so that while I'm thermalling I'm having to "high side" the stick. So if I'm in a left hand turn, the glider is holding the desired bank angle with the left elevon down slightly and the right side elevon is raised a bit. This yields a very favorable distribution of both lift and drag forces along the span, resulting in the best sink rate and the smallest turning radius at the lowest speed.
Another thing before I go, is the radius of the airfoils' leading edge. You often hear of folks having trouble with tip-stalling when using swept-back wing planforms. Swept-back wings have in addition to an inherent dihedral a wash-in effect as well. A flying wing intended primarily for up-right thermalling flight needs to have some wash-out built into the wing. Insufficient wash-out, combined with a small leading edge radius, will ensure tip-stalling problems. Another culprit here is excessive taper in the planform. It will be noticed that modern swept-aft flying wing designs have minimal taper and are quite often equipped with a constant-chord wing design. If you want a lot of taper because you're trying to fly a scale model of a Horten design you will want to provide plenty of twist (washout) to avoid tip stalls. If your flying wing seems eager to tip stall or spin, the easiest fix is to add washout or if that isn't practical then adding more "roundness" to the leading edge radius may help.
dayhead is offline Find More Posts by dayhead
Reply With Quote