Dovey Incident

Arthur Dovey filed the following report with the CAA in New Zealand in December 2007

 

On Friday 21st Dec I had the following incident.

I had spent about 10 to 15  minutes doing aerobatics in the Falco over the Hawea flats and was returning to Wanaka to do some circuits.

I did a standard buzz and break from 180 kts into the circuit with a normal landing. I had announced my intention during the base call to do a touch and go.

After landing and during the landing roll, I retracted the flaps from 40° to 20° (normal take-off  setting) and commenced the go-around. Lift off was normal at 70 kts, and I was rolling the elevator trim to nose down to retrim for the go-around when I became aware that the elevator controls were jammed. I applied as much force as I could but was unable to move it.

At this time with full take-off power applied the nose was coming up and at about 30° nose-up I realized that I couldn't prevent further upward movement of the nose and with the speed decreasing through 75 kts drastic action was required. I rolled right to about 80° and cut the power. As the nose came down through the horizon I felt a jerk and was able to force the elevator forward about one inch. Applying about 20” of manifold pressure and in a 30-degree right bank I was able to fly reasonably level, however the elevator controls were still jammed, and I had to alternate from left to right bank turns to keep from climbing.

At this time I found the rudder had jammed as well about half-right rudder. By reducing  power to about 17 inches of manifold which reduced the nose-up moment and and with the turning and crossed controls I was able to maintain roughly level flight and about 85 kts.

I knew by applying full flap I would get a nose-down moment, and as well, by reducing power would achieve the same, so I felt a landing was possible. My first reaction when the jamming occurred was that I'd be able to jump as I was wearing my parachute. (From training on Harvards through Vampires and Canberras to the Yak 3 I've always worn a chute, as well as when I do aeros in the Falco). That was followed immediately by the realization I was too low. When I had completed what was almost a stall turn I was about 250 feet, much too low. I had also given an emergency call when I realized my predicament, which was responded to immediately by Aspiring Air who had heard the engine cut and realized a problem had occurred.

I couldn't work out what had jammed both controls as the elevator has a horn behind the cockpit running directly to the elevator horn, and the rudder only runs through one pulley to cross and give correct movement to the rudder. Given the situation and being unable to climb as more than 20 inches of manifold gave too much nose up movement a landing was the only possibility.

Because the rudder jam had meant right turns were easier than left I was by this time positioned fairly close in on the downwind leg and about 400', so I planned a tight, close-to-glide approach. I lowered the gear, put down full flap which gave a reasonable approach angle and used the throttle to try and keep the approach around a four-degree slope. I was able to use sufficient cross-controls (aileron vs rudder) to continue and landed on the grass vector. A little burst of power assisted the round-out and the landing was passable. When I had taken off the stick forward pressure I found I was able to central the rudder on the ground, so taxiing was not a problem.

When I got out after a few very deep breaths, I went round to inspect the problem.

I found that the rudder balance weight which is a solid steel tube about 1 3/8 inch diameter by 10 inches in length had become loose and had dropped down onto the elevator rod behind the elevator control horn and was the restriction. It would not have mattered how much forward pressure I had applied, any movement  was impossible. It had been held by two screws, but one was missing and the other still in the forward rudder cap strap, and extending about 1/8 inch.

In hindsight when I had applied the rudder, although I wasn't conscious of applying the right rudder during the emergency turn it had moved the steel rod about one inch from centre (being ahead of the rudder hinge line) and this allowed the slight increase forward in stick movement, but had then jammed the rudder as well. Being tightly jammed between the rudder post and the elevator horn any further movement was impossible, The elevator was jammed in about a 5 to 7 degree nose-up position.

Because the rudder is connected to the nose wheel steering it is the only flying control surface which does not give full and  free movement on the ground. With the coving on the rudder mast it is not possible to view the holding screws unless the rudder is full deflection either right or left. I have always viewed the split pins covering the control surface hinges during my preflight checks and will no doubt in future leave the rudder full scale deflection after towing it out of the hangar to include this check as well.

I have corrected the problem by using longer holding screws which give a bigger bite (and Loctited) of at least 5/8 an inch on the balance weight (instead of what was only perhaps 3/32") and a metal washer to give a greater hold on the cap strap. Finally an aluminum plate under the rod will prevent any downward movement, should the screws ever fail.

I can only say that fortunately my backgound and experience allowed me to handle the worst moment in a 45-year, 25,000 hour flying career. From my first passenger jet (B707) through to the B747, there has always been jammed stabilizer emergency checklists to handle this sort of problem and as such I've always been very conscious of the various moment arms in aircraft flight. However I never realized that that experience would probably save my life.

Arthur Dovey
December 2007
adovey@ihug.co.nz

Arthur Dovey
Visit the Falco Hangar for more information on Arthur and his Falco

 

Arthur,

My goodness, you are a lucky man and extra lucky to have been so well-trained to handle a situation like this.

Is there any possibility you could send me some photographs of the rudder balance weight and the problem as you found it?  I can visualize what you are saying, but many people will not be able to do this without a photograph, and so I'm going to be flooded with questions.

Alfred Scott
Sequoia Aircraft Corporation
alfred@seqair.com

 

I will take a photograph of the tail, and you will be able to see the marks on the elevator control horn, but getting the rudder balance weight out would mean taking the rudder off again, disconnecting the rudder control cables, (and I have to get them tensioned and signed off by a certified engineer which I have already had done), taking off the aluminum plate, etc. However I think a couple of photos will give everybody the right idea.

Arthur
 

I enclose pictures taken yesterday. Originally the rudder balance weight was in the rudder just above the lower hinge, forward of the rudder mast and retained by 2 screws through the forward rudder cap strap. The lower screw had come out and the upper was still in the cap strap protruding by about 1/8 inch. I suspect that on the touch down from the first landing the balance weight had come loose and dropped down onto the elevator rod. As the speed reduced and the nose lowered onto the runway I was reducing the flap to 20° from 40° and wasn't aware of any problem. It was only after putting on power and after lift off when I'd run the elevator trim forward to retrim and then tried to lower the nose a bit more that I became aware of the jammed control stick in forward and aft motion position.

The first picture shows where the rudder balance weight was pushed in against the elevator horn. And how it grated against the elevator horn.

This picture also shows the aluminum plate that I screwed in below the balance weight to ensure it can never drop again.

You may just be able to make out the new bottom retaining screw in the photograph above. The screw is very difficult to see in the photo. It's about 3" above the cut out for the hinge area. Even with full left or right rudder on the ground this screw is virtually out of sight, so it will always be difficult to see except in really good lighting.

The photo above shows the approximate position of the elevator after I got back on the ground. About 5 to 7 degrees nose up.

Because the rudder balance weight was ahead of the hinge line, when the right rudder was applied it moved the balance weight laterally and enabled the elevator horn to be moved slightly further aft, thus reducing the nose up movement slightly.

I can't remember thinking about what to do, when the full realization of the jammed stick sunk in, I think it was mainly instinctive, knowing that only by reducing power and getting on a lot of bank, that the nose could be brought down and preventing what would have been a nasty stall at low level. I do remember thinking as the nose came down through the horizon that I had to get most of the bank off or the same thing was going to happen again. Fortunately I was able to balance bank angle and power quite quickly to gain a semblance of control.

Arthur Dovey
 

I'd like to offer a few comments on the construction of this Falco. First, and most important, let me emphasize that the F.8L Falco design does not include a rudder balance weight.

This Falco was built by a man with an independent mind, shall we say, when it came to following the Falco drawings. A little knowledge is a dangerous thing, and this is a good example. The original builder, Arthur Witzke, was an engineer who worked in the automotive field. He was an engineering manager in engines at Ford Motor Company, handling the design and development of engines through all of the phases.

For a number of years, perhaps four in a row, while he was building the Falco Art would visit us in the Falco booth. Like all builders who deviate from the Falco drawings, he always said that he didn't build this or that the way "you" designed it, with no mention of Stelio Frati. He said he didn't like the gear door design and preferred the design he was on Luciano Nustrini's famous Falco, I-ERNA. This was a design that Stelio Frati had used on his early designs and it was a bit like the gear door design of a Staggerwing Beech, where the half-round section that covered the wheel was hinged and folded out parallel to the ground when the gear was down. In fact, Mr. Frati had long since abandoned this design. I never heard a specific reason given for this change, but Luciano Nustrini had long ago abandoned the design for the one that Mr. Frati designed and which we should in the Falco plans.

But Art didn't want to hear anything about that, and every year he droned on and on with the saga of building the Falco. I didn't like the gear door design so I went with the one used on Nustrini's Falco. But then I ran into a problem and I had to change the profile of the wing. Then this caused another problem and I had to change the contour of the center section of the fuselage. This caused another problem and the problems snowballed to the point that he had to change the entire design of the bottom of the Falco right up to the firewall, and I can't remember if it stopped there, because I tuned out on him and spent much of the time hoping he would find some breath-mints. He became a caricature of himself, and I marveled at his inability to see the absurdity of it all.

As mentioned above, the F.8L Falco does not have a rudder balance weight as part of the design. The leading edge has enough overhang from the hinge line so that typically no balance weight is needed to meet the balance requirements. I have no idea why this airplane has a rudder balance weight. It may be that it was required to meet the balance requirements or it may also be that the builder thought -- as so many homebuilders do -- that a "100% balance" on all controls is the best way to do it, when it is not.

In any event, a balance weight should be securely fastened into the control surface. Typically we recommend that the weight be permanently glued in place with epoxy, and the leading edge at bottom of the rudder offers a good place to do all of this. Installing a balance weight as a tube directly above the elevator controls is obviously a poor design because of the possibility demonstrated here that the weight could come loose and jamb the airplane's controls. You really need to have multiple pin fasteners loaded in shear in a design like this which is questionable at best. The aluminum plate now installed below the balance weight is also a poor design because in resisting the downward movement of the balance weight, the screws would have threads in tension, always a poor design and particularly when you are dealing with sheet metal screws with the threads into wood.

You also want to remember that any part of a control surface may be subject to flutter in which the loads are violent and much greater than the loads involved in aerobatics. I don't remember the exact standards of FAR Part 23, but I seem to remember that a balance weight like this should be able to withstand loads of up to 50g's so that the control surface does not self-destruct in the event of control surface flutter.

As for Arthur Dovey's handling of this situation, all I can say is wow. There are very few pilots who would have been able to deal with this situation, so I think this is one for the history books.

Alfred Scott
Sequoia Aircraft Corporation
alfred@seqair.com

 

That was some incredibly level-headed flying.  Reminds me of the famous Neil Williams incident, in a Zlin, when he had the main spar snap partway through.  Neil instantly rolled the airplane inverted, which of course reversed the load on the spar and kept the wings "from clapping hands over his head," as James Gilbert once put it, and flew some sort of pattern and landed, rolling upright at the last possible moment.  The wings failed upward just as he touched down.

George Larson's and my pal Tom Block once saw a freight DC8 go straight in at JFK--he was walking to his car after a trip--and it turned out to have been caused by jet blast tossing a stone into the space between the elevator and the stabilizer.  When the pilot rotated, the stone became firmly wedged in the space and the airplane tried to climb out with full rotate up-elevator and of course went as close to vertical as it could before paying off.  Block forever after gave his yoke a full down-to-up movement as his last act before releasing the brakes, something he taught me that I always did as well.

Stephen Wilkinson

 

A lucky man to be able to think his way through it.  I doubt I would have survived the events.

George Larson

 

Greetings from the birthplace of George Bush!

More to the point-why would you want to balance the rudder? I don't recall doing that.

Cecil Rives

 

It's really an astonishing story but more then this I like to think to Mr. Arthur Dovey flying skills.... remarkable!!!... I'm not sure many others would have been so "cool" handling such an emergency!

Andrea Tremolada

 

Remember the uncontrolled pitchup accident of a Beach 1600D  in Charlotte, NC about two years ago?  My first thought was to roll the aircraft to control the pitch up.  You would still have a major problem but possibly a survivable one, as in the case of the Falco and Niel William's Zlin. It is in preperation for unusual situations such as this that unusual attitude and situation training should be in every pilots training.  Untrained and unexpected does happen.  A case in point is the UAL DC-10 accident at Souix City Iowa where the aircraft was controlled with throttles alone in roll and pitch.  CB

Jan Kennedy
CKBeau

 

Amazing airmanship Alfred... plus a lot of luck.  His instincts were right on.  Those are the kind of maneuvers we trained for both in the military and in the airlines.  I'm thankful I never had to use the skill.  Thanks for the info.

Al Aitken

 

As you may remember, I was looking at buying that same Falco when 9/11 occurred and not one aircraft insurance company would write a policy for me at the time.  I feel like sending them all thank-you notes.  I am grateful to be alive to read how he handled the emergency.  Kudos to Arthur Dovey and his amazing flying skills.

Jack Amos

 

I just had to write to state the obvious in regards to the Arthur Witzke/Arthur Dovey Falco incident.  There is a lesson to be learned:  It ain't broke. . . don't fix it!  Stick to the plans!  And in addition, Arthur Dovey must be living right because he is one lucky (and skilled) s.o.b..

Rick Fitzwater

 

Update

The balance weight installed in this Falco was a solid steel rod, about 1 3/8" in diameter and about 10" long. It weighed 4.8 lbs and by Arthur Dovey's calculation and measurement, it should have been 4.7 lbs, so it wasn't far off.

The balance weight was installed directly aft of the leading edge capstrip of the rudder and touching it. It was retained by two 3/32" screws that went through the capstrip and which were installed in two tapped holes in the weight. There was a hole in the Sta. 1 leading edge rib through which the weight could be removed.

The design and installation of the balance weight was clearly deficient in three ways:

1. As is obvious because it failed, it was retained poorly by screws, in design and/or construction technique, probably both.

2. There was an obvious lack of failure analysis in this design. As we can all see, when it did fail, it jammed the controls and this would have been a fatal accident with an average pilot flying the airplane.

3. The center of gravity of the balance weight itself was further aft than it needed to be. With a better design, the weight could be reduced.

As a result of our discussions and analysis with Arthur Dovey, he has removed the balance weight and has replaced it with lead shot set in epoxy below the hinge. He drilled at 3/4" hole in P/N 414-1 rib and then poured the required weight of lead shot and epoxy in place. The rudder was positioned nose-down when the epoxy was hardening with the result that the new balance weight was installed further forward and with a greater moment arm, it needed only 2.8 lbs. A check of the balance gives it a balance of 1 lb 8 oz [24 oz] at the trailing edge, as called for in the Falco plans.

With such a design it's important that the epoxy/lead mixture be installed so that it is retained by mechanical means as well as with the adhesion of the glue, so that it is locked in place even if the glue joint fails. The weight should be able to survive an incident of control surface flutter and not shake loose.

Alfred Scott
Sequoia Aircraft Corporation
alfred@seqair.com

 

 

We will be happy to post any comments you may have on this. Please send them to me at alfred@seqair.com

Alfred Scott
Sequoia Aircraft Corporation
alfred@seqair.com