-->

Saturday, July 27, 2013

Unflappable!

When N631S and I departed KVKX last Monday for the trip north to Connecticut, Bridgeport (KBDR) was reporting a 900 foot overcast that was forecast to improve to 1,500 broken around 9AM local time. There were no NOTAM's related to the ILS, so it looked like a good situation.

By the time we were abeam Atlantic City the overcast at KBDR had dropped to 700 feet and by the time we were over the top of JFK it was down to 300 – which is the Decision Altitude for the ILS Runway 6 approach. But there we were, so we flew the approach to have a look. At 300 feet there was nothing but gray in the windshield and we proceeded to fly the missed approach. Checking back on with New York, we heard the expected, "Say intentions." My reply was "Skylane 31 Sierra would like to divert to Oxford." Oxford was reporting a 900 foot overcast.

The controller said, "Maintain 3,000 feet, fly heading 250, vectors for the ILS 36 approach at Oxford." Which will bring us to the point of the story.

New York vectored us onto the final approach course for the ILS Runway 36 approach and handed us off to Oxford Tower. I checked in outside the Final Approach Fix (FAF) and was cleared to land. As usual, I was flying the ILS with 10° of flaps and about 14 inches manifold pressure. This nicely results in about 95 knots and a 600 to 700 feet per minute descent.

We broke out of the schmoo at about 800 feet; there was the runway. At about 500 feet I pitched the nose up a bit to slow N631S and reached over to move the flap control to the fully down position. And...nothing happened. To my surprise, with the flap control fully down, there was no additional flap deployment at all. The flaps just sat there at about 10° extension.

Oxford (KOXC) has about 7,000 feet of runway and landing with 10° flaps is a non-event. In fact, the landing was excellent...a "greaser" (more on that in a bit). I taxied over to the FBO and tried cycling the flaps. Nothing I did would produce more than 10° of movement. So I waited for the ceiling to lift in Bridgeport (which took about an hour), and flew back down there – executing another non-full-flaps landing on arrival. Which led to the next surprise.

The landing at KBDR was not as pretty as the one in Oxford had been; in fact it was a bit of an 'arrival'. With 10° of flaps deployed and the flap control lever in the fully lowered position, the main gear contacted the runway with a hefty 'thump'...and the rest of the flaps deployed very nicely as I rolled down the runway.

Taxiing to the tiedown, I cycled the flaps up and down and got nothing but 'ops normal'. Lever up, flaps up; lever fully down, flaps fully down. All working the way it's supposed to. Drat! I secured N631S and found Tony the Mechanic in the maintenance hangar. He listened to my story and agreed that there was nothing to be done until the failure showed up again.

So...on Wednesday morning, I went back to the airport, opened up N631S, turned the Master Switch ON, and moved the flap control all the way down. The motor began to run, the flaps deployed to 10° – and stopped! Yay! I promptly secured the airplane and went to find Tony.

"Hey, Tony," I called, "Good news – the flaps aren't working!" This drew a strange look from a couple of other people, and a smile from Tony who said, "Good...we'll pull it in here gently, and look for the problem." I guess that it's only in the world of maintenance that the recurrence of a problem is a good thing and its absence is a bad thing. Pilots and mechanics know perfectly well that the problem WILL return...probably when you really don't need it.

The next morning, Tony reported that the problem was fixed. "I turned the master on, and reached up under the panel and moved some wires and the motor started and lowered the flaps. Aha! Pulled the seat out and got under there. A wire to one of the switches was almost broken but not quite – that's why it was intermittent."

Have a look at the picture. When you lower the flap control, it moves the cam and that closes the 'flap down' switch (green in the picture). As the flaps go down, the follower mechanism rotates the switch mount arm and the position indicator until the switch opens and the flaps stop moving – at the ordered position. (It was a wire to this switch that had failed.) When you're done with the flaps, the cam moves the other way, closes the 'flap up' switch, and the process runs in the opposite direction. They've got clever designers out in Wichita!

So N631S is completely healthy again, I get to land with all the flaps I want, and the fix was easy and fairly painless. One of the joys of curating a 36 year old aircraft, I suppose.

Monday, July 1, 2013

Book Review: "QF32" by Capt. Richard de Crespigny

On last Thursday, the Australian Transport Safety Bureau (ATSB) released the final report from its investigation into the causes and consequences of the uncontained catastrophic failure of a Rolls-Royce Trent 900 engine that occurred 4 November 2010. The engine in question was at the time occupying the number 2 position on the wing of a Qantas A380 climbing out from Singapore.

The ATSB report is a fascinating engineering document. It covers thoroughly the reasons for the engine's detonation, the nature and extent of the resulting collateral damage, and the very fortunate absence of any significant airframe fire. It does not, however, dwell at any length on the actions and performance of the flight crew during the event and its aftermath. That side of the story is told well by the pilot-in-command, Capt. Richard C. de Crespigny, in his excellent book, QF32. (The title is the Qantas flight number on which the incident occurred.) Capt. de Crespigny provides the reader with both a gripping narrative and a satisfying explication of technical detail. In addition to being a master aviator, the Captain proves to be a gifted explainer.

The author allots roughly the first 40% of the book to telling us about his journey from his youth through his introduction to flying, his training and assignments in the RAAF, his progression through increasingly challenging – and rewarding – flying jobs at Qantas, and (I think, importantly) his founding and nurturing of a successful computer software business. It was a journey that brought him, on that November morning in 2010, to the left seat of an Airbus A380 named Nancy-Bird Walton and to the moment when, four minutes after takeoff and passing through 7,400 feet, he heard a double "Boom!...BOOM!"

Thus began an incredibly intense one hour and 45 minutes of managing a dauntingly complex systems troubleshooting process while nursing a crippled airplane that had sustained an unknown level of damage. If you wanted to specify the perfect aviator to put in charge of these tasks you would design Richard Champion de Crespigny. From the perspective of stick-and-rudder aviating, his experience flying deHavilland Caribou STOL transports and Iroquois helicopters gave him familiarity with taking machines to the edges of their aerodynamic limits. And his intimate knowledge of software design and behavior not only gave him the motivation to understand the Airbus computer protocols at a deep level, but also a well-honed feeling for the ways that software systems respond to degradation and damage.

As fortune would have it there were five pilots on the A380's flight deck that morning – the normal crew of two plus a relief Second Officer, a Check Captain and a Supervising Check Captain. When the Trent 900 'grenaded', taking with it numerous ancillary systems, the resulting flood of messages from the airplane's computers, some logical and some not, kept all hands busy. The author admits to reaching task saturation in the midst of this chaos until (as he puts it), "..I had my epiphany. My mind switched."

"I inverted the logic. I remembered what Gene Kranz, NASA's Flight Director, said during the Apollo 13 mission: 'Hold it! I don't care about what went wrong. I need to know what is still working...' We went back to basics and it became easy..."

From that point, the crew focused on assuring themselves that they had a controllable airplane with minimally functional systems that could safely be landed within the constraints of Singapore's main runway. As they made configuration changes for landing, Capt. de Crespigny disengaged the autopilot and performed 'control checks', a military technique for verifying controllability of an aircraft that has suffered battle-damage. This isn't found in any Airbus manual, but it assured the crew that the airplane would be stable down to the runway.

Runway 20C at Singapore is 4,000 meters long. The author's colleagues on board had calculated that in its damaged condition the A380 would need 3,900 meters to stop if the Pilot Flying executed a perfect touchdown. At the end of a long straight-in approach, Capt. de Crespigny did just that and got the airplane stopped 150 meters short of the runway's end. And that was the end of the beginning.

The big Airbus now sat at the end of the runway, surrounded by emergency equipment. Flammable jet fuel gushed from the perforated left wing. The brakes glowed, nearly white-hot at over 900°C. And the Number 1 engine refused to shut down. The flight was ended, but not the danger.

The decision was taken to not deploy the escape slides for evacuation. The environment outside the aircraft was more hazardous than the inside. So the cabin crew stood by at the armed doors and stairs and ground transport for the passengers were called for. About an hour after touchdown, the first passenger walked down the stairs and to the first bus; the last passenger debarked about an hour later. At last, the tension subsided. It was three hours and 39 minutes since the Number 2 engine had exploded.

In the days and weeks that followed, Capt. de Crespigny and the rest of the flight and cabin crews of QF32 were justly celebrated. Honors and awards were the order of the day. But the author candidly and generously shares his experience with post-crisis psychological reactions to the stressful experience. He sought and received professional help with Post-Crisis Management, and after a couple of months returned to the left-seat on the A380 flight deck.

It's very interesting to compare Richard de Crespigny's experience on QF32 with that of 'Sully' Sullenberger, who famously landed Cactus 1549 in the Hudson River (as described in his book Highest Duty) after losing both engines as a result of bird strikes. To the best of my knowledge, Capt. Sullenberger has never mentioned any post-crisis effects, but his crisis was over within about 15 minutes. Capt. de Crespigny had to operate under extreme stress for over three hours. Soaking in adrenalin for that long has got to have some side-effects.

There are numerous lessons to be taken from the QF32 story. One stands out for me, and it is this: When it all goes pear-shaped, only the attention of highly skilled, highly trained, highly proficient professional aviators stands between the passengers and disaster. Richard de Crespigny and his colleagues proved this. 'Sully' Sullenberger and Jeff Skiles proved this. And sadly, the crew of Air France 447 seem to have proven the converse.

"The Office" of an A380 at FL330 over France on a quiet evening. (Photo courtesy Capt. R. Ch. de Crespigny)