



Question
A very few design questions, to verify I'm on the right track.
Hello,
I'm in the 'design process' of a bigger TD plane, electric powered. It should be able to circle comfortable, cruise speedy to the next thermal. If things get boring I must be able to fly with higher speeds, make a few "moves" without any fear. The plane's weight should be 66.5kg and have a span of 5.5mtr. It is inspired by euromoldies like the Thermiek XXL or Flying Special 5000. The Extension, build by Wilhelm Niethammer inspires me to build such a plane myself. For the foils I have choosen the MH32 mod. It goes from MH32 12%rooth to MH32 modified thickness and position of the most thick part more forward. For designing the wing I use XFRL5. For sizing the tails and tailboom I use Sailplane calc. My question is about sizing the tails and tailboom. If I look to the planes mentioned above, the tailboom for the Thermik XXL is very short. The tailboom of the Flying special is significant longer, but that is to compensate the Vtail I guess. I have checked the vertical tail size and dihidral criterias of the thermik XXL and Flying Special, but the values are significant lower as suggested in Sailplane calc. BTW, I consider both planes as thermal gliders. Maybe that is not completely correct of me. If I follow the recommendations of the Tailsize checks of Sailplane Calc I get a real big rudder and, for an Xtail rather long tailboom. I know that with increasing aspect ratio the rudder increases and the elevator is not decreasing the way the rudder does. Are the recommended Tailsize check also relevant for bigger gliders? Do you guys have any comment regarding this plane, at a first glance? I know, if I build it the way shown in the picture above, it should be fine. I only feel more comfertable if I get some refeclection or confirmation before starting such a big build..... 



Joined Apr 2010
81 Posts

I can only comment by eye. But I will say that, thats a good choice in airfoil. From an aerodynamcis point of view you clearly know a lot more than me. But just by eye the stab does look quite big for the boom length, and the elevator very small. But if a trusted calculator says thats right, then Id imadgine it would be.
I hope someone more knowledgeable can help you out here. Please do keep us posted on the build, if love to see it . 



What I'm not seeing are the Cl/Cd:Alpha (glide) and the V m/s:Vz m/s (sink) prediction charts. I am trusting that this is a newer version of XFLR? v5 or v6 something with the newest version of Xfoil under the hood?
XFLR can give you so much more than local lift distribution and those numbers are just as important if not more so. to be truthful, one can get XFLR to show good lift distro for a brick, but they don't seem to thermal so great. the chart for V m/s:Vz m/s on said brick just sucks! Give us the rest of the numbers, they will tell us a lot more. like the ones I have here, done for a plank wing. Mark 



More info
If things turn out the way I hope, building should start next winter. First I have to finish a few DLG's and a glass "spare" Supra. There's no hurry, it is a hobby.
I will make a build thread on the dutch forum, with a lot pictures, as usual. Here are some polars for thermal, cruise and speed. (Panel method, wing only.) @muchalucha, the vertical looks less big if you imagine a fuselage in the picture. I hope indeed someone more knowledgeable will climb in. My mane concern is the size of the vertical and his tailboom length. Other issue are, the values of the tailsize check in relation to each other. For example if you have a large wing dihidral, is it then wise to choose a smaller tailsize check number for the vertical? 



I did some searches on RCgroups. Should have done that earlier.. Sometimes I am just to enthousiastic and want to shear thinks I am doing to soon...
Page 13: http://www.rcsoaringdigest.com/pdfs/...SD200408.pdf 



Berrie,
Wonderful project! I'm glad you're finding Sailplane Calc useful. Also make sure you're using the most current version which can be found at www.TailwindGliders.com The link you provided for the RCSD article by Dr. Drela's on Tail Sizing are the formulas I used in Sailplane Calc. I didn't add the radius of gyration because I wasn't sure how he came up with the numbers. If you find how he calculates radius of gyration (rg) or a good rule of thumb for different sized models I'd be happy to incorporate it into Sailplane Calc. Curtis Montana 



Yes, I find Sailplane calc usefull, XFRL too. But running numbers is one thing, knowing how they feel in real life is something else.
I have put, both XFRL5 and sailplanecalc, most of my own plane in the programs. That should give me some feeling with the numbers because I know the behaviour of the planes. A nice example of the use of sailplanecalc: I have build a TD glider, 56 years ago, that did not handle the way I want. At that time I did not know of the existing of sailplanecalc, or even the tail volume stuff. I used sailplane calc lately to check what I have done, 56 years ago. The dihidral sizing criteria was 1,78, or EDA 3.63. The glider was not turning nicely. I did build new joiners, who gave the wing now a dihidral sizing criteria of 2.81, or EDA 5.73. And now it circles the way I want. This way I learn to give a meaning to the numbers, more or less. I find it more difficult to determine the behaviour of the tails, in terms of good/better/best or bad... Next step is: designing, with the help of the program and calc.sheet, a plane that flys exactly the way I want. 



For a good understanding of what affects turning control, read Don Stackhouse’s comments over on the Crimson discussion. He makes great points there. Start here, http://www.rcgroups.com/forums/showt...236486&page=57 , then go to the next page.
In pitch response, you have two competing requirements, static pitch stability and dynamic pitch stability. The static case is more critical, and easier to get right. Dynamic instability is harder to keep out of a design, but you can live with it more easily. Static pitch stability is the ability of your model to hold a speed without you having to stay busy on the stick. Dynamic stability is the ability of the model to return to its trimmed speed after a disturbance without you having to get on the stick. In order to have good static stability, you need a “big enough” horizontal stabilizer and your model has to balance comfortably ahead of the aerodynamic center, say 10% static margin for the test flights. Small tails and short tail booms are no problem as long as the CG is far enough forward. For dynamic stability, things get more complicated. Generally, the shorter the tail boom and the smaller the horizontal stabilizer, the worse the dynamic stability problems get. The tradeoff between boom length and tail size is not linear. (That is, you can’t fully compensate for a short tail boom by building a bigger tail.) My rule of thumb for 2 to 3 meter models is this. Measuring from wing quarter chord point to tail quarter chord point, the tail boom should be at least three times as long as the wing’s mean aerodynamic chord, and the horizontal tail volume coefficient should be at least 0.35. You can use average chord in a pinch, but it will be slightly shorter. Also, as the wing aspect ratio goes up, the tail boom needs to get longer. These minimum numbers are slightly unstable, but it takes a while for things to get out of hand. Static pitch stability requires that the CG be moved forward as horizontal tail volume gets smaller and the boom gets shorter, but that makes the dynamic response worse. At some point, the two requirements pass each other and you simply can’t make a short tail model both statically and dynamically stable. Then you have to decide if it’s easier to live with really twitchy elevator response, or stalldivezoom….stalldivezoom….. runaway. The choice usually winds up in favor of dynamic instability, since that mainly develops in turbulence, it develops slowly, and you are actively flying the bumps anyway. You just have to work a little harder to keep things from getting out of control. That said, dynamic instability can become a real problem at speck out altitudes. If things start to get out of hand, the only alternative can become slamming on the spoilers and waiting for everything to sort itself out. (You did build nice big spoilers, didn’t you?) 



My $.02 worth:
First, let's clarify that whole stability thing: Static stability is the plane's desire to return to the original attitude and angle of attack if it's disturbed. If you pull the nose up a few degrees and then let go of the stick does the plane try to nose back down to the original pitch attitude? If you shove the nose down a few degrees and let go, does it try to come back up? If so, the plane has positive static stability in pitch. Note, the concept of static stability applies to roll and yaw as well. If the plane has exactly zero tendency to go back to the original attitude, that's called neutral stability. If it tries to increase the original disturbance, pull the nose up and it wants to pull up even more until it stalls, push the nose down and it wants to tuck under into the beginnings of an outside loop, means that it has negative static stability. Static stability involves a number of factors, but the most important one in most cases is the C/G location. The "Neutral Point" or "NP" is the C/G location that results in neutral static stability. C/G's forward of there will give increasing amounts of positive static stability, while C/G's aft of the NP result in negative static stability. The "Static Margin" is the distance between the C/G and the NP, typically expressed as a percentage of the wing's Mean Aerodynamic Chord ("MAC") Typical static margins are around 5% to 10% of the MAC forward of the NP. Dynamic stability is the ability to damp out oscillations. If you pull the nose up and let go, most planes will probably overshoot the trimmed pitch attitude (the "setpoint" in control theory jargon) and go into a shallow dive, then pull out of that and overshoot the setpoint on the way back, going through several oscillations (typically 1.5 to 2.5 cycles) before finally settling down at the original attitude. The key here is whether each oscillation is smaller than the one before it. If so, the plane has positive dynamic stability. If the oscillations stay exactly the same size ("amplitude"), then the plane has neutral dynamic stability, and if the oscillations get bigger and bigger, the plane has negative dynamic stability. The other term you often hear in this topic is "damping". No, NOT "dampening", that's all about getting something wet. Damping is the action of taking the energy of oscillation out of the system, which is precisely what it is that causes the oscillations to get smaller and smaller. In your car's suspension, the damping comes from the shock absorbers. If your shocks are working, they keep the wheel from bouncing on the road. If they are worn out, they don't provide enough damping to keep the wheel from bouncing excessively and for too long after going over a bump, you get a very uncomfortable ride, and in extreme cases it could allow the tire to spend most of its time in the air, not in contact with the road. This interferes with braking, acceleration and steering (kinda tough to steer when the tire isn't even touching the road most of the time), and you could end up in the ditch or wrapped around a phone pole. Positive dynamic stability comes in several flavors. If the damping is enough to make the dynamic stability positive, but not enough to keep it from oscillating a bit before settling back down at the setpoint, it's referred to as "underdamped". Most airplanes are a little bit underdamped. If it returns to the setpoint as quickly as possible with exactly zero oscillations or overshoots, it's called "critically damped". The plane doesn't get back to the setpoint quite as quickly, but in cases where an overshoot would be unacceptable (such as a CNC machine, where an overshoot would take a nick out of the part being machined), it's the best that can be achieved within that "zero overshoots" requirement. If it just sort of "oozes" back to the setpoint, even slower than with "critical damping", it's called "overdamped". In an airplane, once again there are a number of factors, but for dynamic stability the major ones are the tail area in comparison to the size of the wing, and especially the tail moment arm. Generally speaking, the dynamic stability in pitch and yaw are linear with the tail area, but proportional to the SQUARE of the tail moment arm. Double the size of the tail and you get twice the static and dynamic stability. Double the tail moment arm, and you get twice the static stability, but FOUR TIMES the dynamic stability. Contrary to some of the other comments here, in my experience dynamic stability is considerably more important to overall handling qualities than static stability. In fact, I've even set the C/G aft of the NP (so static stability was negative) on a plane with decent sized tail surfaces on a very long tail moment arm (our electric Chrysalis 2meter, which is very slightly overdamped in pitch and yaw), and handed the Tx to a firsttime beginner. Despite the static stability being a little bit negative, because the dynamic stability was sufficiently high, she had no trouble staying mentally ahead of the plane. Other planes with this combination of moderately small tail areas on an unusually long tail moment arm, so that static stability and control authority were in the "normal" range but dynamic stability was unusually high, include the Ryan Navion (famous as being one of the best instrument trainers ever built, with an exceptionally smooth ride in turbulence), the legendary Bucker Jungmeister aerobatic plane, and the Supermarine Spitfire, which contrary to what the narrow landing gear might suggest, was noteworthy for having some of the best ground handling behavior of any of the WW II single engined fighters. So why don't we all make our tail moment arms four or five times the wing span, and have ridiculously high amounts of dynamic stability? In the "old days", that would mean having excessively high weight and whetted area in the tail cone, adding drag and hurting control response. Modern podandboom construction can help mitigate that problem, but for planes like the Beech Bonanza it was a serious issue, and a key reason why they opted for a tooshort tail moment that left them with a fishtailing problem in turbulence. However, the other problem is what it does to turning behavior, a major issue for model and fullscale sailplanes alike, but especially so for models, particularly if they don't have ailerons. In a turn, the wing tip on the inside of a turn sees a lower airspeed than the one on the outside. For a very lightly loaded thermal soarer, that difference could be as much as 2:1, which means the inside tip needs as much as four times the lift coefficient as the outside tip to keep the plane balanced in roll. One of the things required to make this happen is more angle of attack on that inside wing tip. If you have ailerons, you can just add a little outside ("top") aileron. However, for a twochannel setup, you need to yaw the plane towards the outside of the turn. This interacts with the wing's dihedral to increase the angle of attack at the inside tip (and lower it at the outside tip), so the plane stays in balance in roll. This is where things get tricky. When the airplane is in a turn, the airflow past the plane is curved. If the local airflow at the wing is perfectly aligned with the wing's chord, the "relative wind" at the tail might be blowing inward and upwards by fifteen degrees or more! This tends to put the nose down, and to yaw the plane towards the outside of the turn. If the tail moment arm is exactly matched to the aerodynamics of the rest of the plane and to the geometry of the turn, this yaw will be exactly correct to create just the right amount of additional angle of attack at that inside wing tip, and the plane will go around the turn all by itself like it was on rails. If the tail moment arm is too long, the plane will always be trying to roll out of the turn. It will also have about as much maneuverability as an overloaded school bus. Some of our competitors made exactly this mistake on some of their designs, and when we had a chance for some headtohead flight tests we saw exactly this problem. They would go in a perfectly straight line all day long, but trying to get them to enter and then hold a thermal turn required an act of congress plus about two weeks advance notice. Also, if the tail moment is too long, the required size of the tail surfaces becomes so small that they start to run into problems with excessively low Reynolds numbers. Also, you will find that the optimum tail moment depends in part on the bank angle. The plane will tend to roll out of turns on one side of that operating bank angle (excessively positive spiral stability), and roll into the turn (spiral instability) on the other side of the optimum. The trick is to adjust things so that the plane happily holds the turn when at the bank angle most folks typically thermal at with that model. The bad news is that rules of thumb are not really very good in this situation. There is enough variation between design details from one design to the next that you really need to consider each design on its own merits. Although analysis can help nail things down, don't be surprised if you find the need to finetune your prototype based on flight experience. Joe and I have even gone as far as building a prototype with an adjustable length tail boom to make fine tuning easier. It can be worth the trouble if you're trying to come up with a truly optimum design. 



I am now trying to understand dynamics....
to say it short: static stablitly can be managed with a short tailboom. Dynamic stablity prefers a longer tailboom. (Am I right?)
The dynamic issue reminds me after an RCSD artikle: http://www.rcsoaringdigest.com/pdfs/...SD201111.pdf It started at page 21. Dutch rolling is:
Yesterday I did some testflying at my field. I compared my homebuild Supra with my own design TD glider that I build before I build the Supra. I recognize the tailwagling in more turbulant air of my own design glider. The Supra behaves more calm. (Both planes fly well btw, ony has my own design has some habits that I would like to understand) To be continued, with small steps. edit: these doc's shall be studied: http://www.xflr5.com/docs/XFLR5_and_...y_analysis.pdf http://xflr5.sourceforge.net/docs/XF...ts_Rev_0.1.pdf 




Quote:
You can get static stability with any length tailboom, including one where the distance between the trailing edge of the wing and the leading edge of the stab is zero. That particular case is a plank style flying wing, where the last 20% or so of the chord in effect acts as the horizontal tail, and the reflex needed in the plank flying wing's airfoil trailing edge is analogous to the "decalage" between the wing and tail of a tailed aircraft. So, yes, as long as you have enough pitch authority to trim the plane, you can get a given amount of static pitch stability from essentially any length tail moment arm by finding the required C/G location. That said, the shorter the tail moment arm, the more down force you need at the tail for a given amount of static stability. In addition, it also takes more tail area to make that down force, and that means more whetted area in the tail. OTOH, if you make the tail moment longer, you have less whetted area and weight in the horizontal tail, but more whetted area and weight in the tail boom. This is the tradeoff Beech (and others using similar construction philosophies) faced in the design of the Bonanza and similar aircraft. That type of construction results in a lot of weight and whetted area in the tail cone, so making it longer carries some significant penalties. One approach to optimizing this is to choose a parameter; weight, whetted area, material and labor costs, whatever; and then look for the optimum combination. When I was designing a Spitfire Mk22 Quarter 40 pylon racer, I chose the tail moment arm based on what combination of tail cone and tail surfaces, for given tail volume coefficients, resulted in the lowest total whetted area, and therefore the lowest skin friction drag. The result actually came out surprisingly close to scale. The typical method for managing the static stability and control authority vs tail area and moment arm question is with tail volume coefficients. Basically we put everything that helps the tail do its job and put that in the numerator, and things that make the tail's job more difficult go in the denominator. There are some articles in the "Ask Joe and Don" section of our website www.djaerotech.com that go into the details. For example, the formula for the horizontal tail volume coefficient is: Vht = (tail area / wing area) * (tail moment arm / wing MAC) where MAC is the wing's Mean Aerodynamic Chord. Typical numbers in model sailplanes for this are around 0.45 to 0.55 . Now, if we want to look at the dynamic stability question, simply square that second factor, to create what I refer to as a "Dynamic Tail Volume Coefficient": Vhtd = (tail area / wing area) * (tail moment arm / wing MAC)^2 Volume coefficients are a way of attaching a number to the effectiveness of any particular airplane's tail design. Find examples of planes whose control response and static and dynamic stability you like, calculate their static and dynamic tail volume coefficients, and with enough examles you should see a pattern emerge, an typical range for those coefficients that result in the behavior you are looking for. There will be variations caused by things like the amount of inertia about the pitch and yaw axes due to masses in the extremities, additional tail area needed to counteract aerodynamic pitching moments due to flaps, extra fin area in multiengined planes due to the need for enough rudder authority to overcome asymmetric thrust in an engine failure situation, etc., but assuming your examples are sufficiently similar to your situation and to each other, there should be a pattern, a range of numbers that seem to work well. Use that as a guideline for your design. 



Joined Oct 2010
347 Posts

It's looking good! I normally use a slightly larger tail plane and a slightly smaller fin, but that is more what I like than science.
I build a new model in the Xplane simulator to test it, before I build it, to see, if it is something that I like flying. 



@sablatnic, that is nice way to design, virtually test flying before the build!
How do you feel about Xplane compared to real world flying? Well, I have now the following design: Dihidral sizing spiral stability: 2.81 Dihidral roll controll: 0.06 Equivalent Dihidral angle: 6.02 Vertical tail volume: 0.021 Hoizontal tail volume: 0.41 It only looks like an Amigo in frontview, at the otherhand, I can feel it circle if I let it turn on my screen in XFRL5...... 
Thread Tools  
Similar Threads  
Category  Thread  Thread Starter  Forum  Replies  Last Post 
Question  A Few Questions on the HAM Radio Test  DJO  FPV Talk  13  Feb 13, 2012 07:50 PM 
Help!  few questions ondesigning a ornithopter for senior design class  sharkforce  Ornithopters  9  Sep 13, 2011 05:04 PM 
Question  Am I on the right track? (Brand new build, what am I missing)  molish  FPV Talk  22  Sep 09, 2011 09:29 AM 
flies very well inverted, cant track straight up right?  juantrinidad  Electric Plane Talk  6  Dec 23, 2002 12:27 AM 