Thursday, November 23, 2006

AV - Tail Flapper Failure

Wednesday 19 May 2004

My tail stopped flapping yesterday. I don't know exactly when. It was okay when I checked it around nine a.m. I was in & out of the shop several times during the day and could hear the motor but I didn't look to see if it was still flapping. When I checked it about ten p.m. this evening the motor was still running but sounded different. The tail was no longer flapping.

Obvious case of Pilot Error.

The motor and its attached gear box is a high-quality item, purchased as new-surplus for just a couple of bucks because the output shaft is metric - - 12mm instead of an honest half-inch. Ami-Ricans go sorta silly when you mention metric, turn their nose up even if it's quality goods, which this unit is. What failed was the grub screw in the disk I've attached to the output shaft.

The disk is a hunka half-inch aluminum tooling plate, about 2.25" diameter. It's turned down to a thickness of .125" except for the hub which is fitted with a #6 set screw and reamed to accept the gear-motor's 12mm shaft, which has a flat for the set-screw. The disk is drilled with a series of .191" holes to accept an AN3 bolt. The holes are in a spiral, each one a precise tenth of an inch farther from the center of the disk, starting just outboard of the hub. I've been using the first hole to flap my tail-hinges. The first hole doesn't give a lot of motion, about three-quarters of an inch, in and out. But lots of power.

The tail I've been flapping isn't much to look at. About 10" wide, mebbe six deep. Just two hinges. Not covered or anything, just the hinge-line spar of 3/4" square wood with a piece of 1/8" ply glued to it, sticking out aft. I put a pound of lead on the plywood, a precise six inches from the center-line of the pivot point. Every time the tail flaps, it raises that pile of wheel weights about four and a half inches. It does that about 20 times a minute and has for the last forty days or so. Good little motor. Nice gear-box. Except for being metric, of course.

The hinges are made out of steel strapping; some junk I hauled home from a job site about a year ago. The strapping is mild steel, .020" thick by .625" wide. It's easy to bend yet surprisingly strong. I did the original calcs using 30kpsi but tear-out tests showed the metal is stronger than that; probably closer to 50kpsi. In theory, the stuff is strong enough to be used for hinges but it's so thin I guessed it would fail at the fastener. Or the thin section would cut through the pivot pin.

But guessing isn't knowing.

I made some C-shaped hinges. For the pivot pin I used AN393-47 clevis pins. That's a pin .191" in diameter and about an inch and a half long. I attached sets of hinges, usually in pairs, to three-quarter inch square stock - - hemlock or Douglas fir, whatever came to hand - - using nuts & bolts. Then I tore them apart using an hydrualic press, guesstimating the amount of force involved by the pressure on the cylinder at the moment of failure.

As a test of fasteners, it was a complete failure because the fasteners didn't. The strapping would tear out around the pivot pin or the wood would fail where it was attached to my puller. I fastened the hinges to the wood with AN3 nuts & bolts, with flat-head machine screws, even with crappy little 6-32 hardware (4 ea). All proved more durable than the material of the hinge.

Which is another way of saying my fasteners were too expensive. And too heavy.

Eventually I worked my way down to SEWING the hinges on using .025 SS safety wire, threaded through six or eight #40 holes. After twisting the wire tight I'd squirt the holes full of urethane glue then let it cure. To provide a bit of lateral support, I butted a scrap of 1/16" plywood about the size of my thumbnail up against the bent end of the hinge, used some more urethane glue. Now I started seeing some failure of the fastener... at about the same time the AN3 hinge-pin would start to tear out. The loads at that point was on the order of five hundred pounds, givertake.

Clevis pins aren't cheap. Actually, at about half a buck they're cheaper than dirt compared to most other aircraft certified hardware but they were also the strongest part of the hinge. Could I find a pin that worked as well but cost less? I donno. The only way to find out was to try different stuff.

I tried steel rod, nails and even clothes-line wire. At a pin diameter of about an eighth of an inch, with a mild steel pin, I began to see evidence of shearing in the pin along with the anticipated tear-out of the strapping. Load (per hinge) at that point was about 350 pounds.

Small diameter pins were something of a problem because they were difficult to secure. Using a nail as a sort of Po' Boy clevis pin, the head of the nail secures one end of the pin but you had to drill the other to accept a small Cotter key or safety wire in a Northrop loop. Eventually I struck upon using the Cotter key itself as the pin. BIG Cotter key... AN380-4-5. That's a Cotter key 1/8" in diameter and about an inch and a quarter long.

Worse-Case load for the elevator I had in mind was 540 pounds distributed across six hinges and I was seeing better than 300 pounds from a hinge that weighed about an ounce and cost less than a dime. It was kinda goofy and fragile looking but it was a surprisingly tough little hinge.

In addition to using the strapping as a pair of identical but overlapping ‘C's, I also experimented with L-shapes and even T-shapes, each of which offers some advantage, either through increasing the strength of the fastening to the wood or increasing the surface area where the pin bears upon the strapping.

I've never heard of anyone using steel strapping for hinges, nor sewing them to the spar. I tried to locate data on similar hinge arrangements and came up dry so I laid out a test program, made up the tail-flapper and started flapping. Of course, just flapping the thing really isn't much of a test. Nothing for UV exposure nor corrosion but I did run some vibration tests. Adding a small motor to the flapper and fastening a short length of steel eccentrically to the shaft gives you an Industrial Grade vibrator that wears out the hinge pins quick like a bunny, along with the motor :-)

The current flapper has been flapping since the afternoon of April third. About 1.3 million cycles. Two months would have given me about two million cycles. Earlier versions usually crapped out after ten days to two weeks due to excessive wear at the pivot point.

I've lubed the hinges with a variety of stuff. Right now I'm using dry tungsten disulfide powder, burnished into the surface of the Cotter key and the bores in which it rests. Seems to work pretty good. The lubricant seems to be the most critical factor although how accurately the samples are made also plays a role.

Anyone else out there trying to push back the darkness? Kevlar roving as the tension members in a fuselage using a Howe truss? Vacuum bagging leading-edge skins? If so, I'd be delighted to hear from you. But privately, please. I don't want to detract from the real purpose of rah.


No comments: