Thanks to both Blaine and Mark for help on this topic. It is greatly appreciated. If you would not mind, I would like to impose on you for some follow up questions.
The DLG that we have been experimenting with has Vv = 0.048 which is in the ball park of what Bill Watson has found to be successful. We have found that Vv in this area seems to damp as well as a gyro though I don't know how to know for sure. I would think there must be a way to calculate the point where the Damping_From_Large_Vv = (Damping_Increase_From_Gyro)* Damping_From_Smaller_Vv = Perfectly_Damped_Launch. I don't know if this is the way to say it but maybe it gets the question across. This could be hard to do. As a practical matter, it may just be worth it to go with big Vv and simply throw the gyro in on top to make sure you have the least drag possible due to tail oscillation on launch. Blaine writes: "I agree with Mark that a balance between too much tip dihedral and too much root dihedral (that is, beyond elliptical to V-dihedral) is good. A dihedral scheme in which the dihedral angle of each panel is roughly proportional to the distance of the panel MAC from the aircraft centerline gives a good starting point. My current thinking is to bias this slightly towards V-dihedral." Our various flight tests seem to confirm this. Maybe also Mark's. The best wing that we tested had more V-dihedral than the others. So far it seems that with a big rudder on can roll into the turn well and that you need more V-dihedral to easily stay in the turn. Looks like the razor saw might have to be applied again to test some V vs tip dihedrals. At least at my level of pilot skill. For HL it may be the case that more V-dihedral bias is in order since one has to turn tight in small and often light bubbles where falling out even once means you don't get a second chance. Does this make sense? Also, I am not clear on how to calculate your dihedral scheme starting point. Would you consider providing a sample calculation? Blaine writes: "The spiral stability equation that Mark describes works. It is sensitive to wing lift coefficient. If you have a high Cl airfoil, you need more dihedral to be spirally stable. Most modern airfoils thermal best someplace near 0.7 Cl, but small handlaunches might be as low as 0.5. Consider this in the equation!" Ok. This brings up another question that I have been struggling with. That is to try to determine at what cl and V that the HL is actually flying at when doing different things. What I have done to try to figure this our may not be at all the best way. I took xfoil output at re*strt(cl) = 57k. I then took the polar output and determined L/D max from the data. In this case cl at L/D max = .61. Since one thermals at higher cl you know that cl >= .61. Think that I should try to figure out cl at min sink as the best indication of a thermaling cl? Seem like cl could even be more than that when climbing in a thermal. Anyway, it seems pretty straight forward to determine cl for a given hl plane at different velocities, but knowing the velocity or cl at which different maneuvers occur seems difficult. What am I missing? Lastly on the topic of boom length, I am having a hard time making the empirically observed behavior fit with the way things should work. We have two identical fuses and tail groups except one has a 2" longer boom length. The shorter one turns better and tighter using all of the test wings at differnent EDAs. So, I am trying to figure out what could be going on because it does not make sense given the previous comments on this topic. Could it be that since yaw inertia increases by the square of the boom length and damping is linear with the boom length that the reduced yaw inertia is more favorable than the increased damping?? Help greatly appreciated. Tom -----Original Message----- From: Blaine & Deborah Beron-Rawdon [mailto:[EMAIL PROTECTED]] Sent: Monday, October 01, 2001 3:03 PM To: Soaring Cc: Mark Drela; WILLIAM WATSON Subject: Re: [RCSE] How does one calculate parameters for the perfect poly Gents, Mark Drela has invited me to comment on the subject of setting up a polyhedral airplane. Mark was responding to questions nicely put by Tom Clarkson who has gone so far as to do actual experiments. I agree entirely with Mark has to say on the subject. I can add a little though, from experience and thinking about it. When you do an experiment to investigate the best amount of dihedral (equivalent dihedral angle (EDA)), you have to consider several important factors. The vertical tail moment arm, the size of the vertical, the dihedral scheme, the wing airfoil, and the wing weight (especially at the tips) are all important. Moment Arm and Tail Size The combination of the vertical tail moment arm and the size of the vertical stabilizer determine the control power and the damping of the model in yaw. If you have a lot of dihedral, you must have a lot of yaw damping. Damping is proportional to the area of the vertical times the vertical tail moment arm squared. This means that a long boom is especially good for yaw damping. If you have inadequate damping and a lot of dihedral, the model will feel wobbly and imprecise - you will have to fly with smooth inputs for the model to fly smoothly. If you have a lot of yaw damping, you can get a combination of nimbleness and smoothness. If you have adequate yaw damping, you can fly with a lot of dihedral. I have a V-dihedral 100 inch floater that uses a 1/4 inch diameter joiner. For a while, I flew the model with an aluminum joiner. One day, I bent that rod so that the model had 24 degrees per side of dihedral! It flew just fine and was very responsive in roll without wobbling. Mark Drela comes to us from the free flight world (among others) where vertical stabs are incredibly tiny. Also, he has learned how to build very light wing tip panels. As a result, he uses relatively small vertical stabilizers. His recommendation of a vertical tail volume of 0.03 is a bare minimum at best for me. Bill Watson has recently built a series of about six DLG RE models. Bill settled on a vertical tail volume of about 0.06 or more! The discus launch is a factor is sizing the tails, but the models handle very well with about 11.5 degrees EDA. I would recommend the following relationship for vertical tail volume as a function of EDA as a starting point: Vv = 0.02 + 0.002 EDA Example for EDA = 10 degrees: Vv = 0.02 + (0.002)(10) = 0.02 + 0.02 = 0.04. Dihedral Scheme The dihedral scheme will influence how much dihedral you can get away with and how the model behaves. If you use a scheme with relatively flat inboard panels and a lot of tip dihedral, you will find that the model may hesitate to roll up when you bang the rudder, especially when flying slowly. This is the forward tip stalling a bit, and resisting the yaw input. A technique when you squeeze in a little down while rolling up may help, but so would a little less tip dihedral, or a little less EDA overall. This is different that the roll-up stall behavior that Mark describes. I agree with Mark that a balance between too much tip dihedral and too much root dihedral (that is, beyond elliptical to V-dihedral) is good. A dihedral scheme in which the dihedral angle of each panel is roughly proportional to the distance of the panel MAC from the aircraft centerline gives a good starting point. My current thinking is to bias this slightly towards V-dihedral. Wing Airfoil The spiral stability equation that Mark describes works. It is sensitive to wing lift coefficient. If you have a high Cl airfoil, you need more dihedral to be spirally stable. Most modern airfoils thermal best someplace near 0.7 Cl, but small handlaunches might be as low as 0.5. Consider this in the equation! Wing Weight Mark's rule to maintain light yaw inertia is correct, especially for RE models, and even more so for RE models with a lot of EDA. I have found on my models that vertical tail volumes that work well for light built-up models are woefully inadequate for heavier composite models where yours truly used too much structure outboard. Yaw inertia is sensitive to weight and radius squared. If you break up the model into little pieces and then sum up the product of their mass and radius squared, you will get the yaw inertia. (Where radius is the distance to the CG of the airplane.) If you do this for the wing, you will see that the weight of the tip panels dominates the wing inertia. Also the weight of the tails is important, once you have lightened the wing tips. If you somehow make the yaw inertia too high, the plane will be wobbly in yaw. You can fix this in at least two ways. Make the vertical stabilizer bigger. This will fix the wobble and give more yaw control power. Alternatively, consider using a yaw rate gyro. This will improve or fix the wobble, but won't help the control power. I have used this solution to good effect on the overbuilt model outlined above. That's all for now. Blaine Beron-Rawdon Envision Design San Pedro, California http://members.home.net/evdesign/ Date: Sun, 30 Sep 2001 22:24:00 -0400 From: Mark Drela <[EMAIL PROTECTED]> To: [EMAIL PROTECTED] Cc: [EMAIL PROTECTED] Subject: Re: [RCSE] How does one calculate parameters for the perfect poly handlaunch glider turn? Message-ID: <[EMAIL PROTECTED]> >EDA_min = 5 * CL * Span / Length > >Length I believe is the length between the 1/4 MAC point >of the wing and the horizontal stab. Vertical stab, actually. >1) Could it be true that the ideal EDA for a poly rc hlg = EDA_Min? That's a good starting point. A slight amount of spiral instability is not necessarily all that bad. >if you rolled into a bank and then took >your hands off the sticks the 9.6 wing would expand the radius of its turn >until almost going straight. The 5.6 would tighten its turn and you would >have to control the plane pretty quickly. This depends also on the trimmed CL. If you add up-elevator trim (increase CL), then the 9.6 wing would tend to roll out less and less. In any case, releasing the stick should result in the glider slowly rolling out, since you normally want to hold a touch of rudder into the turn, no? Spiral stability will also depend on the size of the vertical tail. For the stable EDA's, a bigger vertical tail will make it more stable. For the unstable EDA's, a bigger vertical tail will make it more unstable. Also, the larger the EDA, the smaller the required inward sideslip in a sustained turn. Less sideslip = less drag. >1) Could it be true that the ideal EDA for a poly rc hlg = EDA_Min? That's a good starting point. Keep in mind that the formula is approximate. The spiral stability changes somewhat when in a sustained turn, so some tweaking may be warranted. >2) It seems that wings with more dihedral where the wings join >seem to fly better than those that are flatter in the middle >even with equivalent EDA. The spanwise Cl distributions depend on the dihedral distribution and on the type of maneuver. Vee dihedral... increases the inside-tip Cl when rolling into a turn decreases the inside-tip Cl when in a sustained turn Tip dihedral... decreases the inside-tip Cl when rolling into a turn increases the inside-tip Cl when in a sustained turn My V-dihedral Apogee will drop a wing if the glide is slow and the turn entry is very sharp, but once in a tight turn it is virtually impossible to tip stall. My original composite Allegro had lots of dihedral in the tips, and would do the opposite... almost bulletproof on turn entry, but you had to be more careful when a sustained turn. I find that the poly layout of the Allegro-Lite seems to strike a good balance. >3) To what extent is boom lenght a factor? Some people think that >longer boom lenghts are the way to go. A longer tailboom will improve pitch and yaw damping for a given amount of pitch stability and yaw power. Normally this is a good thing. I have yet to see a poly glider with too much yaw damping. A longer boom will require more up-elevator and inside-rudder bias in a steady turn. Again, this is rarely a problem for most "long-tail" designs. A longer tail will also allow smaller tails with less drag, and less EDA for the same amount of spiral stability. >Based on flight experience it seems that you want to >achieve the smallest turning radius with the least bank angle. Bank angle only depends on flight speed and turn radius. The tail has nothing to do with it. To get to your initial question... >The definition of best that I am trying to determine how to >calculate is a turn that is nimble, instantly controllable, >and loses the least energy and alititude in a turn. Three rules: 1. Low yaw inertia 2. Low yaw inertia 3. Low yaw inertia :-) On a magical poly glider with zero yaw inertia the rudder would have slightly faster roll response than ailerons, without the attendant drag penalty of TE device deflections. Oh yea, and use adequate EDA and a big vertical tail. A Cv of 0.03 or more is good. There is little downside to a big vertical tail on a poly other than a tiny bit of drag and weight. Blaine Rawdon can probably pipe in here too. - Mark RCSE-List facilities provided by Model Airplane News. Send "subscribe" and "unsubscribe" requests to [EMAIL PROTECTED] RCSE-List facilities provided by Model Airplane News. Send "subscribe" and "unsubscribe" requests to [EMAIL PROTECTED]
