...Fortunately, the only damage to my chrysalis was to the tail assembly - the right side broke off and was hanging by the covering.

Reconstruct the V-tail to its original configuration, but with lightening holes in the ruddervators to reduce their mass and the likelihood that they will flutter. My concern about this is whether I'm right in assuming that the weight reduction will do the trick.

Rebuild the tail section in the conventional configuration, with the idea of winding up with three small control surfaces rather than two large ones. This is do-able but, again, will it do the trick? Also, it's a lot more work, and presents some construction difficulties - first, in establishing a clean mounting surface for the new horizontal stab and second, in re-arranging one of the control cables to get it to where it can operate the rudder.

Your comments/advice?

From : Don Stackhouse

Before diving into a complete structural redesign of the tail, let's check the simple things first. Flutter in the tail is almost always due to slop/flex in the control linkages and/or hinges. Make sure the pushrod ends are securely attached to the pushrods, that they fit the holes well at the servo and at the control horn, that the pushrod guide tubes are securely tied down at both ends as well as along their length, that the unsupported length of pushrod at both ends is as small as possible and that the unsupported section is stiff enough that it doesn't flex under load (if your pushrods are the cable type and not the music wire type, try soldering the last several inches of cable at both ends with low-temp silver solder to stiffen them), and make sure that your hinges don't have too much flex. The Chrysalis does not accelerate fast enough in a dive to get to flutter speed "before I could recover". This tells me that something was wrong, and that your airplane's flutter speed was abnormally low. Linkage flex is by far the most likely culprit.

OK, so you've checked your linkages and hinges, but you're still nervous. What else can we do?

Cutting holes in the ruddervators will probably reduce the stiffness by quite a bit more than the weight that it saves, so if anything it could make things worse.

You might even consider making some new ruddervators with a bit less taper in their trailing edges to stiffen them up a little more in torsion.

If you really want to stiffen them, get some carbon or Kevlar tow (you could even pull some from the edge of a piece of fabric, or peel some from some uni) and run two pieces in an "X" from diagonally opposite corners on both sides of each ruddervator to increase the torsional stiffness.

Mass balancing the ruddervators is a possible fix, but you have to be careful that you don't add so much weight that you create a problem for the stabs or the aft end of the tail section. The most effective place for a mass balance is out near the tip (that's the part with the greatest movement during flutter), and your mass balance should put the C/G of the ruddervator exactly on the hinge line. Probably best to balance the ruddervators while they're off the model so the stiffness of the hinges doesn't interfere with the balance measurement, then hinge them to the stabs.

Stiffen the stabs with some carbon fiber run from the leading edge at the fuselage to the hinge line at the tip. Put it on both the top and bottom surfaces. If you've added 1/2" wide carbon fiber to the inboard wing spar caps, split some of the leftover carbon into 1/4" wide strips and use that. This will create what amounts to an "I" beam along the stab leading edge, which will combine with the basswood along the hinge line to act as an "A" frame, dramatically stiffening the stabs.

Ideally you want the ruddervators, stabs and tail boom to all be very stiff and light, but also of different stiffnesses so that their natural frequencies do not match. Ever notice how many of the planes of WW II vintage used sheet metal fins, stabs and wings, but had fabric covered control surfaces? Yes, those planes in many cases started out with sheet metal covered control surfaces, but they found that this tended to give the control surfaces similar natural frequencies tot he fixed surfaces they were attached to. This was a very bad thing, since this could encourage the two surfaces to excite each other. A number of prototype aircraft were lost due to sympathetic vibration (flutter) between the control surfaces and their adjacent fixed flying surfaces.

Reverting to the "old-fashioned" fabric covering for the control surfaces altered their stiffness and mass characteristics enough to move their natural frequencies away from the frequencies of the wing and tail panels. It also reduced the usage of strategically critical sheet metal, so it could be put to other critical uses such as making sure the troops had their plentiful supply of canned Spam, thereby ensuring the allies that they would have properly nourished military personnel, and that Monty Python would also have plenty of nourishment for their future "products" as well.

If you've now stiffened up the ruddervators and the tail surfaces, the remaining issue could be torsional stiffness of the tail boom. If you pluck the ruddervators like a guitar string, do they vibrate by themselves, or does the whole tail and tail section vibrate with them? If so, you might need to stiffen the tailboom. Some lightweight fiberglass (about .75 oz to 1.0 oz would be about right) on a +/- 45 degree orientation wrapped around about the aft half of the tail boom would help. You could apply it with epoxy, or a lighter option might be to use a water-based polyurethane varnish such as Varathane to apply the dry cloth. You can use water-based acrylic enamels such as the ones Tamiya sells for lexan R/C car bodies to tint the final coats of varnish, eliminating the need for additional paint or covering.

The down side of all of this that you're adding a fair amount of weight to the tail, which means you will also probably need a fair amount of additional weight in the nose to balance it. I'd recommend just checking the linkage stiffnesses first, then doing some shallow dive testing, gradually building up speed and stopping at the first hint of flutter. Proceed to the other heavier and more drastic steps only if you need them.

Don Stackhouse
DJ Aerotech