Shop our Airplanes Products Drone Products Sales
Thread Tools
Aug 11, 2016, 04:40 PM
Registered User
Discussion

Aero Design for Tailless


To all,
This thread will focus on the aerodynamics of tailless miniature aircraft.
MOST IMPORTANTLY (not SHOUTING) Please give each of your posts a title as this thread is not about any one subject. In this way someone researching design can find a topic of interest more easily than searching within the thread to find it. Thank you.
With a lot of information related to the aerodynamics of tailless aircraft scattered throughout the Nurflugel forum my goal is to have a thread where this knowledge base is concentrated for easier accessibility.
With the advent of the internet there is a wealth of related topics out there so please contribute if you discover something of interest.
I am a tailless design geek so if like me you find this subject entertaining I want this to be the place to go to.
Peter Wick who is well known in the tailless community dropped by in the History thread with his observations on tailless design. Peter is someone who has most importantly practical experience as well as theoretical savvy. I have studied Peters body of work on airfoils and airplanes for years and this is just the type of information that will prove useful. Hopefully we'll see Peter in here.
I do not want this to turn into the debate thread; I want it to be the idea sharing thread. It is ok to differ with the opinion of others let's just make our best effort to be professional and keep it civil.
I'm going to attach two recent projects I spotted online that are of interest.
One is the F3F "SiGh" from an RCSD monthly magazine:
http://www.rcsoaringdigest.com/pdfs/...SD-2013-04.pdf
The other is on electronic "wizardry":
http://e-collection.library.ethz.ch/...h-47495-01.pdf
Chris
Last edited by drive320; Aug 14, 2016 at 02:59 AM.
Sign up now
to remove ads between posts
Aug 12, 2016, 12:35 AM
internet gadfly
nmasters's Avatar

RCSD articales


The "On The Wing" articles provide a lot of information.
Last edited by nmasters; Aug 26, 2016 at 09:21 AM.
Aug 12, 2016, 10:52 AM
Plank addicted
Marcos Quito's Avatar
Great idea !!
Aug 12, 2016, 11:45 AM
Registered User

Flutter and Tailless


It's always body freedom flutter!
Years ago I was diving my balsa structure swept wing virtually straight down targeting me and the visual I had was the wing just growing larger in span as she descended towards me, no right to left motion. I had the wing clocked by a CHP officer with a radar gun at 109 M.P.H., not bad for a 9 ounce balsa structure glider.
On a "record speed" attempt the wing was visually growing larger in span descending rapidly when all of a sudden I saw the trailing edge of the wing (elevons) beginning to flap like a flag slowly up and down, then the flapping began a greater divergence and quicker in frequency until BAM the right wing panel broke free just outside the Horten scallop. The now crippled airframe tumbled down with only a Nyrod keeping the severed panel attached to the center section. Without the Nyrod tying things together it would have been totaled. I repaired the wing and it went on to fly a lot more.
Avery good demonstration of as Norm termed it "body freedom flutter".
Granted Peter has never witnessed a plank flutter and I'm in agreement that a swept planform lends itself to aerolastic affects more so than a plank. Now if flutter did in fact down the Opel is questionable the eye witness describing the wing behaving like a bird flapping its wings (or like a flag on a flag pole) lends me to lean towards flutter but a failure of the drag spar would do the trick.
If control surface flutter sets in on a tailless then they cease to be control surfaces and are now big servo tabs. The result of this is just as serious for a tailed airplane yet even more serious for the reasons Norm illustrated.
WW2 fighter aircraft designers were encountering serious compressibility and flutter issues as speed on their clean and high powered ships built up rapidly.
The US Navy had a requirement that any prototype fighter must dive and then execute a 9 G pull up. That requirement was ok for biplanes but fatal for the test pilots of say a Corsair and was discarded for the "modern" fighters.
Even when losing wings the conventional configured fighter became like an arrow with fins that the pilot might escape if he didn't strike the tail surfaces or get killed by massive dynamic (Q) forces. In a tailless the center body housing the pilot or crew is going to tumble violently rendering the pilot unconscious or killing him by snapping his neck.
Chris
Aug 12, 2016, 12:22 PM
Registered User

Tip Rudders, End Plates and Tip Stall Behavior


Quote Peter:
"A funny history about the use of wing fences.

When we made the Vision 87 in 1987 the flying wing was tip stalling very very badly (just -1 degree of twist). For an overview of the wing look at the drawings at https://www.aerodesign.de/

The performance was quite good. The intention was to fly F3B with it. But every third launch was a crash. Then we started to build several type of winglets and even removed them totally and used a central fin. With the central one, the tip stall was cured, but the performance was somewhat gone also. So we tried a very nice rounded corner for the winglet. Duration flight time was improved, but the stall was even worse compared to the normal winglets. On some hollidays in the Alps we all together were thinking about how to cure the tip stall or if we just should burn the plane. Michael Wohlfahrt came up with the idea of Grenzschichtzäune, german for wing fences - so we took the hotel "do not disturb" label and used it as a fence between the wings at halvspan. The fence went around the nose of the wing and extended on the upper and underside somewhat longer than the thickest point of the airfoil...around 40% of the chord.
The next day we flew it and the tip stall was totally cured!!! You could hold elevator for a very long time and the plane would just nick up and down. It was never flown again without. Interestingly the performance was still ok, probably a little bit lower than before, but not very significant."
Peter,
Relating to your website unfortunately my German language skills are zero. My Son spent one year as an exchange student in High School in Germany and just completed his Junior year in College at Ludwig Maximillian University in Munich so I may have to chase him down...
I am very interested in your findings relating tip stall and end plates or "winglets". Your experience is that end plates promote tip stall.
Everything I've read on end plates suggests you are correct.
Aside from fancy winglets and such the end plate is similar (not exactly) to testing a wing in a wind tunnel that bridges from wall to wall creating an "infinite aspect ratio" in contrast to a wing in free flight that has tip losses. The vortices on a wing tip effectively create a down wash on the tip area reducing the angle of attack and lift am I correct?
Now if my scenario is correct it would reason that end plates increase the aerodynamic angle of attack at the wing tip and therefore promote tip stall. I had read your experiences (or was it another German's) that moving the vertical fins aft of the wing tip eliminated tip stall.
When I read of attempts to retrofit winglets to an existing airframe the bending moments increase indicating that lift is increased at the tips a result of an increase in aerodynamic angle of attack at the tip?
All of this leads me to what Norm had previously posted in the "Do winglets change the CG/ %MAC position on a wing?" thread.
Quote Norm:
"Winglets move the center of lift outboard which is also aft on a swept wing. Since MAC is that chord where the aerodynamic center is at 25% this means that MAC and the neutral point are farther aft but that doesn't necessarily mean that the CG can also be farther aft. Winglets increase the efficiency of the wing near the tip by moving the tip vortex off the surface of the wing and add some pressure differential of their own. Neither of these effects increase CL max of the wing airfoils so tip stall would occur at the same angle of attack with or without winglets. The NACA did some experiments with theoretically elliptically loaded wings and the results are in TN 606. Figure 2 in that report shows the difference between the theoretical lift distribution of two wings of different aspect ratios and their actual lift distributions. Attached is my defacement err "annotation" of that illustration. The deviation from the theoretical curves along the whole span is from the pressure equalization at the tip and those horrible drag inducing bumps at the tip are from the tip vortex scrubbing the wing surface. Winglets cure both of those problems plus add some extra pressure so the lift distribution is much flatter but that does not change the stall AoA of the wing..."
So if I'm understanding Norm correctly winglets do not promote tip stall (unless I'm just not understanding what he is attempting to convey) which is exactly the opposite of what you experienced in real world testing.
Any thoughts from either of you?
Chris
Aug 12, 2016, 10:39 PM
Registered User
iron eagle's Avatar

Tip plates vs wingletts


Some data from my Stingray.
Last edited by iron eagle; Aug 13, 2016 at 01:26 AM.
Aug 13, 2016, 02:59 PM
Registered User

Angle of attack and induced drag


Don,
You stated: "induced drag is related to lift coefficient, not alpha."
The induced drag formula does not have the term for angle of attack in it however it is in it:
CL*2 / pi x AR x e
I am asking questions here not making statements; I'll look for you to correct me on what I'm asking you.
A wing with a symmetrical airfoil is at an angle of attack of zero degrees thus the wings CL is zero.
The wing has profile drag but no induced drag as it is not generating any lift.
There are no tip losses as there is no pressure differential between the upper and lower surfaces of the wing.
To increase the CL one must increase the wings angle of attack.
One might assume that deploying high lift devices such as trailing edge flaps increases lift without altering the main wings angle of attack.
This assumption is false.
The angle of attack is defined as the angle between the airfoils chord line and the relative wind.
The chord line is the straight (not camber) line between the leading and trailing edges of the airfoil.
A lowered flap increases the angle of the chord line as now the chord line has become the straight line between the leading edge of the airfoil and the trailing edge of the lowered flap.
The instant I am required to support the weight of the aircraft it has to fly at some angle of attack.
In order to increase lift I can increase velocity or angle of attack or both.
Increasing velocity increases lift provided the lifting surface has an angle of attack to the relative wind.
When I increase angle of attack I increase the CL of the wing, the two are intrinsically intertwined.
As the Sinatra song goes "You can't have one without the other".
As angle of attack increases the "physical" lift is progressively tilted aft in the direction of relative flow and is greater in magnitude than the weight of the aircraft.
The vector of lift directly opposing gravity (weight) is equal in length to weight vector.
Geometrically the vector of physical lift tilted aft is greater than the lift vector directly opposing the weight of the airplane a function of the lift opposing the weight divided by the cosign of the angle of attack.
There is a third vector that "connects" the vectors of lift that opposes the weight and the vector of the physical lift that is tilted aft. This connecting vector opposes the aircrafts flight path comprising one of the components of drag. The connecting vector is defined as induced drag or the penalty one incurs for generating lift.
The greater the angle of attack the longer the resulting vector ("Drag, D" in attached) opposing thrust or a vector of gravity if it's a glider.
Because you told me "induced drag is related to lift coefficient, not alpha" here is where things get muddy for me.
I have two identical aircraft excepting one has a swept wing and the other has a straight wing (with no sweep).
Your paper on what happens when a wing is swept states that due to the effects of wing sweep angle a swept wing for every degree of angle of attack must fly at a greater angle of attack (a function of the cosign of the sweep angle) to generate the identical CL of a straight wing.
For the swept wing there is no net loss of lift just less CL per unit degree of alpha.
Ultimately the swept wing achieves the same max CL it just happens later at a greater angle of attack.
According to the paper if we sweep a wing 20 degrees then using the cosign of 20 degrees we reduce the CL six percent per degree. If the straight wing generates a Cl of .1 per degree then the swept wing is generating a CL of .094 per degree.
In order to generate a CL of 1 the straight wing must increase its angle of attack to 10 degrees.
For the swept wing to generate a Cl equal to 1 it must increase its angle of attack to 10.64 degrees.
It then would stand to reason that the swept wing is generating greater induced drag to achieve the same amount of lift as the straight wing at all angles of attack.
To fully grasp the performance penalty associated with the reduction per degree of the swept wing and the resulting penalty associated with induced drag I'll revisit the induced drag equation from above.
Reducing the CL per degree from .1 for the straight wing to a six percent less CL of .094 per degree for the 20 degree swept wing the induced drag penalty is increased 13 percent.
If we estimate that at best L/D parasitic drag equals induced drag then this penalty is nearly seven percent of total drag.
This is a substantial drag deficit compared to a straight wing.
The reduced L/D is significant for tailless soaring and long range aircraft like bombers.
This swept wing induced drag penalty is a physical property that cannot be altered with any electronic wizardry.
Chris
Last edited by drive320; Aug 15, 2016 at 03:11 AM. Reason: correction to induced drag ramifications
Aug 18, 2016, 12:40 AM
internet gadfly
nmasters's Avatar

Evaluations of the Horten H-IV


Wasn't sure whether to post this in your history or design thread because these are historical technical documents. The Horten H-IV was evaluated twice, once by DFS in 1944 and again by MSU in '59. In the 15 years between those tests it had been crashed and repaired several times, not always, possibly never, with access to the original drawings. It had also crossed the Atlantic on a boat and then gone all the way to Calafornia befor finaly going back to Mississippi . Anybody know what effect several weeks of exposure to salt and humidity have on a wood, aluminum, and magnesium airplane? MSU modified the center section to streamline the landing skid and accommodate a much heavier pilot. These modifications reduced directional stability by moving the center of lateral area forward, increased induced drag simply by making the gross weight higher, and increased parasite drag by increasing the wetted area and messing up the streamlining behind the canopy. Naturally with all these "improvements" MSU got poorer results than DFS. Figure 13 in the MSU report shows their glide polar (curve #1) and several projected curves if various improvements were made. Not surprisingly curve #3 "induced drag reduced" is pretty close to the DFS data. After they finished with it MSU traded it for a box of radio equipment and then it changed hands a couple more times until the final crash during an aborted car tow in California.

--------.~.
--------/V\
------//----\\
-----/(------)\
----(^^)---(^^)--Norm
Last edited by nmasters; Aug 26, 2016 at 09:23 AM.
Aug 18, 2016, 12:43 PM
Registered User

Horten H-IV and SB13


[QUOTE=nmasters;35512593]Wasn't sure whether to post this in your history or design thread because these are historical technical documents. The Horten H-IV was evaluated twice, once by DFS in 1944 and again by MSU in '59. In the 15 years between those tests it had been crashed and repaired several times, not always, possibly never, with access to the original drawings. It had also crossed the Atlantic on a boat and then gone all the way to Calafornia befor finaly going back to Mississippi . Anybody know what effect several weeks of exposure to salt and humidity have on a wood, aluminum, and magnesium airplane? MSU modified the center section to streamline the landing skid and accommodate a much heavier pilot. These modifications reduced directional stability by moving the center of lateral area forward, increased induced drag simply by making the gross weight higher, and increased parasite drag by increasing the wetted area and messing up the streamlining behind the canopy. Naturally with all these "improvements" MSU got poorer results than DFS. Figure 13 in the MSU report shows their glide polar (curve #1) and several projected curves if various improvements were made. Not surprisingly curve #3 "induced drag reduced" is pretty close to the DFS data. After they finished with it MSU traded it for a box of radio equipment and then it changed hands a couple more times until the final crash during an aborted car tow in California.

Norm,
If you are referring to me I'd like this to be not "my" aero thread but every ones and for all to benefit from.
My hope is this thread will focus on aerodynamic design and theory to advance the state of the art and provide a place for the tailless designer to readily obtain information.
It would help if each posting has a title (placed in the title box) that will tie it to similar postings.
If it's not too much trouble could you "edit" your last two post with titles?
I would relish having you, Stackhouse, Stokley, Suter, Wick and others getting in here.
It's difficult to separate theory when discussing historical practice but as soon as the theory part kicks in I'd like to see it wander over from history here to aero.
The Horten video you submitted in the history thread is killer especially when who pops up there is Rudy Opitz. Classic stuff and perfect in the history thread.
All of what you talk about related to the H-IV is spot on.
I did not know about the car aero tow and subsequent crash, what a shame as the H-IV deserved better.
I've looked at the DFS data and MSU data but was the H-IV superior in performance or equal to that of its contemporaries in performance?
The Darmstadt D-30* flying alongside the H-IV that is in the pictures you posted flying side by side is so cool. How did the H-IV without "wear and tear" compare to the performance numbers for the D-30 I lifted from Wikipedia compare:
0.5 m/s (98 ft/min) minimum, at 62 km/h (39 mph; 33 kn)*
The D-30 was likely the pinnacle of wood and fabric (some aluminum structure like the spar) conventional design of that pre war period.
How did the two stack up? I do not know that is why I'm asking.
I just came across your thoughts here so I need to take a closer look:
https://www.rcgroups.com/forums/show...1564973&page=2
That leads to the "modern" day version in what Peter Wick posted the SB13.
Had the SB13 team gotten past all the challenges they encountered in flutter and tip stall (I see fences on the SB13 in the museum) how would it have stacked up against one of its contemporaries of 1980's design? If the SB13 was clearly superior to a conventional wouldn't it have lead to more advanced design?
You have opened up an aero discussion that deserves more focus.
Come to think of it should this be in the history thread?
Chris
* https://en.wikipedia.org/wiki/Akafli...dt_D-30_Cirrus
Last edited by drive320; Aug 18, 2016 at 12:48 PM.
Sep 04, 2016, 10:28 AM
internet gadfly
nmasters's Avatar

Span efficiency and glide of Horten airplanes


This graph and data table was given to me bey Marko Stamenovic a few days ago. Note that most of the aircraft with glide ratios below the trend line had large pods. Fuselages are always bad, both because they create parasite drag and tend to reduce span efficiency. Low AR wings tend to have high span efficiency but high sink rate simply because of the small span relative to the gross weight.
Last edited by nmasters; Sep 04, 2016 at 01:01 PM.
Sep 04, 2016, 12:16 PM
Registered User
Don Stackhouse's Avatar
There was some discussion on another thread (that I no longer participate in) claiming the H-IV was a deficient design because it didn't outperform the Darmstadt D-30, a conventional "tailed" sailplane that had been built as a state-of-the-art max performance experiment.

The comparison used in the comment was somewhat invalid. Yes, both aircraft had the same span, about 60 feet, and similar weights. However, comparison of a BSLD against an elliptical lift distribution of the same span puts the BSLD at an unfair disadvantage.

A bigger issue is that the D-30 had an aspect ratio of 33:1 for an L/D of 39:1

The Horten H.IV had an aspect ratio of only 21.8:1 and a measured L/D of 37.1 according to some tests, 32:1 according to other sources. The highly flawed tests of a badly maintained and substantially modified H.IV at MSU after the war was 29:1

However, the key point is that the H.IV did have a lower L/D than the D-30. BUT, if we refer to the chart Norm just posted, if we compare the Horten design at the SAME aspect ratio as the D-30 (which would be more in keeping with the H.VI), we get an L/D in the low 40's, which is superior to the D-30.

Note, there are still other issues (such as how to properly compare spans between an ESLD aircraft vs a BSLD aircraft in a performance comparison; for a valid comparison the two should NOT be equal), but the point is that in comparisons between the concepts, the Horten puts forth a much better showing for itself than the raw data seems to imply.
Last edited by Don Stackhouse; Sep 04, 2016 at 12:22 PM.
Sep 04, 2016, 02:48 PM
Registered User
Sure Don has a valid point. Changing the premis of what to optimize, makes the span invalid as a constant.

Actually that was the original point of Prandtl...he would suggest to compare given lift and bending moment as a premise.

On the other hand, is bending moment (or payload) a perfomance issue for typical model aircrafts?

Here is by the way a video from the SB13 flying with CG fra back, showing some pretty nasty stall tendencies. Thats probably why they use the fences on the wing.

SB 13 Trudelversuche (1 min 37 sec)
Oct 08, 2016, 02:00 AM
internet gadfly
nmasters's Avatar
This article about a new online aerodynamic analysis program, called MachUp, from Utah State University showed up in my Facebook news feed this morning. I don't know how it stands up to XFLR5 or FLZVortex but it looks interesting and easy to use, as you can see in the attached video. There are also a couple of spreadsheets on this page of the Nest of Dragons web site that should be part of any nurflugeler's tool kit.

--------.~.
--------/V\
------//----\\
-----/(------)\
----(^^)---(^^)--Norm

MachUp: Flying Wing Overview (7 min 12 sec)
Mar 31, 2017, 09:09 PM
internet gadfly
nmasters's Avatar
OpenVSP 3.11.0 has a really neat blending feature that may be interesting to some builders. I personal prefer to avoid compound curves but if you don't mind that this could be a great program.
Blended Wing Tutorial (9 min 45 sec)


Thread Tools