How do you fine tune V-tail differential on a Chrysalis?

Hi guys, let me first say that I really like my Crysallis. I've been flying now for a couple of months and it flies great.

Lately I've been trying to improve me flight times and noticed that when I use rudder the nose drops alot. I actually have to hold it up by applying elevator.

I tried adjusting the Up/Down ratio to a little more up than down but it seemed like if I got it to where the nose wasn't dropping, the plane would almost barrel roll instead. This caused me to crash my plane and I'm fixing it now (it broke right at the outer panel/inner panel joint).

Should I have been adding more down than up? This is my first VTail so I'm a little confused and annoyed at my inability to figure this out.

I did notice that you recomend 3/8" up and 1/2" down for starters...I think I was more like 1/2" up and 1/2" down at first.

Thanks greatly,
-tom

P.S. Good job on the great products and friendly support!

From Joe Hahn:

Hi Tom. Sounds to me like you may be trying to slow down a bit too much--the Chrysalis prefers to fly fairly fast, especially when turning. I'd say practice technique first, then if things aren't going well, check the wing washout amount. This is the single thing that will either make or break the handling of the ship. If you want to add a little washout greater than what we describe, the handling will improve at a slight performance decrease. I've copied this response to Don, he may add some thoughts for you.

Hope this helps, let us know how it goes!!

Joe

From Don Stackhouse:

Tom, a few more comments. I absolutely agree with Joe's suggestions, but it also sounds to me like you may be overcontrolling a bit. If you barrel rolled into the ground, that sounds like a lot of control sensitivity, with the controls held at a high deflection for way too long. Of course "way too long" is a relative thing. If your controls are set up for a roll rate of 720 degrees per second, then a max authority roll command held for a fourth of a second will be "way too long" (in this case, a 1/4 second would flip the model from upright to inverted). The key is to know how much sensitivity is right for you. Fortunately that's an easy one. If you find that the model is responding faster than you can mentally keep up with it, then it's too sensitive. Move the pushrods further in on the servo arms, or use your fancy programmable transmitter to reduce travel. Dual rate settings on the transmitter are a good way to do this, just remember to check that they're set properly before each launch.

If you find that you're getting bored waiting for the model to respond, then you might like more sensitivity. Just move the pushrod connections further out on the servos, or dial in a more aggressive dual rate setting.

The other big factors on control sensitivity selection are how much max authority you need to deal with your environment (you need enough to deal with the prevailing turbulence and to maneuver safely around obstacles in your landing area), and the need to minimize drag. Deflected controls cause drag. The more smoothly you can fly, and the smaller your control deflections, the better your model's performance will be.

Another thing I should warn you about, on planes that use rudder for roll control, and also on swept wings (both of which apply to some degree to the Chrysalis), the roll authority tends to increase at high angles of attack. When you started playing with tail differential to keep the nose up, you may have also caused an increase in roll sensitivity. No big problem, just be a bit more gentle on the controls (which you should be trying to do anyway), or reduce your control throw to cut down on the sensitivity. Remember, like I discussed above, the controls are set right for YOU when the model is comfortable for YOU to fly. You should not feel like you are constantly "on the edge", holding a tiger by the tail, desperately trying to keep up with the model. You should be comfortable with the model's response. If you're not having fun, you should be. Adjust the control travel until you start having fun.

One other quirk of planes that use rudder for roll control is that as the model rolls into a banked condition, the rudder starts to act a little like down elevator. This effect increases as the bank angle steepens, until at 90 degrees bank the model may have a powerful desire to drop the nose. If the problem that inspired your question happened mostly as you were rolling into a steep turn, this could be the culprit. 2-channel models of this type may have a lot of trouble doing a roll because of this unless they have truly outstanding roll response; they tend to end up in a "graveyard spiral" instead. BTW, the Chrysalis and Monarch 'D' WILL do excellent rolls when properly set up, although they do have this nose-dropping tendency to some extent; it's inherent to varying degrees in any aircraft that uses yaw to create roll control. If you suspect that this may be part of your problem, I recommend re-measuring your dihedral at the center and at both poly breaks to make sure it's correct. If your dihedral (especially at the tip panels) is less than what the plans call for, then you've reduced the desire of the wing to roll when it's yawed. This means that you need extra yaw to get the desired roll response, and that extra yaw shoves the nose down as you roll into a turn.

I don't recommend increasing the dihedral settings to more than what the plans call for, that can lead to "dutch roll" problems. Also very unpleasant, it's when the model wants to wallow from side to side like a falling leaf. It's the result of too much dihedral and/or not enough vertical fin. The flip side of that is spiral instability, where the model wants to slip into a spiral dive when trimmed for a normal turn. Yup, that's the result of not enough dihedral or too much fin. There's a delicate balance between the two conditions, and a lot of our flight testing centers around getting that balance just right.

Note that none of what I just discussed has anything to do with V-tails. The biggest drawback to V-tails is that because they're different, they tend to get blamed for a lot of things that aren't their fault! One quirk they do have, though, is that because the "end plate effects" of the fuselage and the other tail surface are not the same for both surfaces when you input a "rudder" command, the two surfaces don't necessarily generate the same force when the ruddervators are deflected opposite but equal amounts. This imbalance can cause the model to respond in pitch as a side effect of a rudder (yaw) command. This is the reason for using differential on the ruddervators.

To set the differential properly, get the model in level flight, flying away from you. Now watch closely as you apply a sudden, strong rudder input. As the model starts to yaw, does the nose rise or fall? You need to watch the response just as the model starts to yaw, even before it starts to roll. You need to make a strong, sharp rudder input, but you don't need to hold it very long, 20 degrees of bank is more than enough. Do the same thing back the other way to level the wings, and note once again if the nose rises or falls as the model starts to respond. Adjust the differential in the ruddervators until you see little or no pitch change with rudder inputs. There is some personal preference involved here, some folks like to have it absolutely neutral in pitch response to rudder, other folks like to set their model up to naturally drop the nose a bit as they enter a turn. Set yours up so that it feels comfortable to you.

Don Stackhouse @ DJ Aerotech

Part two : 

This is exactly what i was asking. When entering a turn the nose drops. If I was to only turn right, the plane would turn and fly towards the ground. This means I have to banmk slightly and pull back to keep the nose up when entering a turn. On a normal 2 channel ship (no vtail) rudder causes the plane to sort of slide sideways until the poly wingtip is caught and pushed up, and the plane enters a smooth turn (no nose drop).

By adding more up I was hoping to simulate the amount of pull back I give when entering a turn, but the result was that instead of a smooth banking turn, the outside wingtip would flip so high that it was difficult to keep the plane from almost looping. Reducing it only made the nose drop again. IS there a balance somewhere and did I just miss it?

From Don Stackhouse :

Yes, it sounds like that to me. I would also double check that your polyhedral is set properly. If the angles are too low, it takes a greater yaw angle to get the same rolling effect, and that extra yaw shoves the nose down as you get into steeper bank angles.

One thing that maybe I didn't explain clearly, is that if the nose drops immediately as you start to roll into a turn from level flight, it's probably because of too much differential in the ruddervators. If it starts into the turn ok, but starts to noticeably drop the nose after reaching about 20-30 degrees of bank, it's more likely to be a shortage of dihedral.

While you're at it, check to make sure you don't have too much dihedral in the tail. This would act like too much fin on a conventional tail, which causes spiral instability (i.e.: dropping the nose in turns until the model works itself into a spiral dive).

Another possibility is that if you tend to fly right on the edge of a stall, and have a lot of rudder authority, and use all that rudder suddenly as you roll into a turn, you could be stalling the inside wingtip. The maneuver is essentially a classic "snap roll", or what the Brits call a "flick roll", they seem to have their own names for everything over there. Then again, they all drive on the wrong side of the road, too! Anyway, back to the subject, if you yaw the model suddenly you increase airspeed momentarily on the outside wingtip and decrease it on the inside wingtip. If you throw in some significant and sudden "up" elevator at the same time (like you did when you added a whole bunch of "up" differential), the outside wingtip sees a big increase in lift, the middle of the wing sees a moderate increase in lift (which makes the model rise suddenly as it noses up, like the entry to a loop or barrel roll), and the inside wingtip sees a stall (with associated sudden and violent loss of lift). The net result is a sudden and violent roll in the direction of the applied rudder, with an associated pitch-up. Try getting the model into cruising level flight,with plenty of airspeed, then take your hand off the stick and use the trim knob to roll smoothly into a turn. Is the response much smoother and more linear? if so, this could be your problem.

I would also recommend that you re-check your washout NOW, and frequently in the future until you see it holding at the correct setting. Incorrect washout can severely aggravate any tip-stalling behavior. The vast majority of handling problems we've seen with the Chrysalis were directly caused by incorrect washout.

To check the washout, hold the wing upside-down, trailing edge toward you, at arms length. Look across the trailing edge at the high point of the lower surface and rotate the wing in the angle-of-attack sense until you see a part of the trailing edge appear to touch the high point in your field of view. Sometimes it helps to close one eye while you're doing this. If the entire trailing edge appears to touch the high point of the lower surface at the same time, then the washout is zero. If the root trailing edge touches while it still looks like the trailing edge at the polyhedral joint is 1/8" away, then you have 1/8" washout. If the trailing edge at the poly joint touches first while you can still see 1/8" of the underside at the root, then you have 1/8" washin (a very bad thing). BTW, you will need to look at each panel of the wing individually; because of the polyhedral, if you try to judge the outboard panel while your line of sight is centered on the inboard panel, you will not get an accurate reading.

This method is also very good for spotting warps. For example, if you see 1/8" washout at the poly joint, then you should also see 1/16" washout in the middle of the inboard panel, 1/32" at 25% of the way out from the root, and 3/32" at 75% of the way from the root to the poly joint. Use the procedure described in the instructions to re-set the washout as necessary, and don't be surprised if it takes several iterations of re-shrinking the covering before it holds the desired setting. The Chrysalis was designed to be easy to build and easy to adjust after construction (that's precisely why there is no leading edge sheeting, beginners have HUGE amounts of trouble with that, and if they get it warped initially it's very hard to fix), but the down side of that design decision is that it's also easy for the washout to change inadvertently. You need to keep an eye on it for a while after initial covering, until it gets settled down at the desired setting.

In any case, I suspect that there was a balance there, and yes, you started on one side of that balance and ended up on the other side of it. At this point I would suggest setting your ruddervator differential to zero, i.e.: for rudder inputs, make the travel of the up-going ruddervator the same as the down-going one. If you see the model raise or lower it's nose as you first start to roll, add differential in the correct direction IN SMALL AMOUNTS AT A TIME until it's behaving for you the way you want.

Two other factors could be clouding the issue here. When you roll into a turn, some of the lift of the wing is now being used to pull the model around in a circle, leaving less to do the basic job of holding the model up. To compensate, the wing needs to make more lift. If you do nothing in the way of stick movements or trim knob settings, the model will naturally start to descend faster, which has the effect of increasing both airspeed and angle of attack until there is once again sufficient lift component in the upward direction to support the weight of the model. You can compensate for this by a small amount of "up" elevator while you're in the turn, provided the model is not too close to stall. If this is what's happening to you, then adding "up' to the rudder through differential will give you the correction during the initial roll into the turn (which is BEFORE you need it), but then take it away again when the rudder is centered during the turn. Programming and differential can't help with this one, you just have to learn how much extra "up" to hold while turning.

Complicating this phenomenon on very lightly loaded models (like HLG's) is the curvature of the airflow phenomenon. Because the model is turning, the airflow around it is curved, in both the pitch and yaw sense (because of the bank angle). If the turning radius is small in comparison to the length of the fuselage (which is normally the case for a very light Chrysalis in a very small, tight thermal turn), this curved airflow has the effect of creating some false "down" elevator and "top" (i.e.: away from the turn direction) rudder effect. I've been in situations in extremely tight turns with very light Monarchs (3-4 ft. diameter circle measured at the inside wingtip, about a 45-55 degree bank angle) where in order to keep the model in the turn, I had to use about twice the amount of "up" elevator that it takes to fully stall the same model in level flight. The model was nowhere near stall, actually the angle of attack was fairly low, the extra "up" elevator was just compensating for the curvature of the airflow due to the turn. Once again, there is nothing that transmitter programming can do to influence this issue, you just have to practice and learn to deal with it through piloting skill, or else don't try to make any really sharp turns!

If the nose dropping continues after the model is stabilised in the turn and the controls are centered, these phenomena could be what you're seeing. If that is the case, then it's a matter of piloting technique, not control setup. These are steady-state phenomena that exist continuously during a turn, and the control setups effect only the transient case of rolling in or out of the turn. You can't change steady-state behavior with transient corrections!

You're flying a model that's capable of flying in ways that normal airplanes never experience, and it may be just trying to tell you that there are some new things you need to learn about if you want to "go to the edge of the envelope" like that. That's exactly what any good instructor, airplane or human, should do.

I guess I really need to fly someone elses V Tail so I can see how it behaves. This may be normal and everyone is just used to it from flying VTails for so long.

I personally doubt that it's natural behavior for a V-tail, or any other kind of tail. My personal ones don't act like you describe, if they did we wouldn't be selling kits with that tail design. I've flown both conventional and v-tail versions of the Chrysalis, and they both behave essentially the same. I'm sure your troubles are related to setup and technique, not the design concept, and if we get those sorted out, the problem will go away.

Thanks for taking the time to give me such a detailed response. I think I'll read it again! -tom

No problem at all, you're very welcome. Builder's questions are an essential part of keeping Joe and I sharp as designers, and we welcome all feedback, good and bad. Right now I'm sorry you seem to be having trouble getting your Chrysalis to behave itself, that's probably almost as frustrating for me as it undoubtedly is for you. The basic design is sound, though, and I'm confident that if we stick to fundamentals and make small adjustments at each step, we will be able to sort this out. Good luck, and by all means, please keep in touch!

Don Stackhouse @ DJ Aerotech