No Plan Survives the Enemy

There was a culture among test pilots in the Air Force after World War II and before the advent of computerized aircraft design that required "pushing the envelope" and seeing what an airplane could do with the right pilot flying it. Those days are over and there is now much more sanity in the process. See: Functional Check Flights.

I had FCF duties in the Air Force twice, the first time was in the Boeing 747. In the cockpit and out, I learned you can plan everything to the nth degree but things rarely go as planned. It was a good job, probably the sanest job I had while an Air Force pilot. The goals were clear, the rules were iron clad, and everything was, well, functional. Of course all that had to occur within a very large, military bureaucracy, which military or not, tends to be . . .

"Major Albright," the squadron public address blared, "report to the commander's office." I dropped the uneaten jelly donut back onto its plate and headed for Lieutenant Colonel Larson's office.

"Sir?" I said, peeking through the half-closed door.

"James," he said, "come on in. I thought I would give you a pep talk about Air Command and Staff College."

"I'm not volunteering," I said. "I didn't submit the paperwork."

"Nonsense!" he said. "Of course you're volunteering. Not only that, you are going to be selected. Don't you realize how good this will make the squadron look? Don't you want to do this for us?"

"Not really," I said.

"Pretty funny!" he said. "I'm glad we had this talk, James! Well, I know you are pretty busy with the upcoming test flights. Fly safe, James!"

"Functional check flights," I said. "They aren't test flights."

"Pretty funny!" he said.

Our four Boeing 747s were a mixed bag of airplanes somehow engineered to be identical. The first two were originally built for Qantas Airlines to Australian specifications. A week before the final coat of paint and the Kangaroo logo decals, Qantas ran out of money and Boeing had a good deal to offer the Air Force, which the Air Force gladly took. A few years later we bought a third airplane to look like the first two. A few years after that, just before the 747-200 production line was to come to a close, airplane number four rolled onto our ramp.

Airplane number three was the problem child. As airplanes go, no big airplane is nicer to land than the 747. Except for airplane number three. Or at least that was what everyone said. I wasn't so sure. I do know I hated air-refueling number three. As easy as the 747 is to land, it is difficult to air refuel. Even under the most benign conditions, placing an 800,000-pound airplane a few feet behind a tanker at 25,000 feet and several hundreds of miles per hour can be daunting. For most of our pilots, it was sometimes a joy and sometimes an ordeal. When somebody started an aircraft diary for each airplane, it became clear that more times than not the ordeal was in airplane number three.

It was the "Sick Bag Metric." How many passengers got sick during the air refueling? Airplane number three easily claimed more victims than the rest of the fleet combined. For most of our pilots the primary symptom was misdiagnosed as "pilot induced oscillations," a rapid up and down movement that alternated the pressure on your behind with that against your lap belts and shoulder harnesses. For me, it was simply a feeling of flying sideways. Armed with that newly quantified data (sick bags), we funded a month of test flights at a Boeing 747 repair plant in Texas.

"Test flight" was the wrong terminology. The Air Force really gave up on those years ago. It doesn't make sense to take an airplane into the sky with the outcome uncertain. That costs too much in terms of dollars and lives. The official term is "functional check flight." The experts would work on the airplane until they thought it airworthy and I would fly it to prove or disprove that fact.

For our sideways airplane I asked the squadron to provide their choice of copilot and navigator and I picked a flight engineer who could keep me out of trouble.

Chief Master Sergeant Terry Wilson was flying airplanes since before I was born and knew the 747 from nose to tail. He tended to drink too much on the road and was known for getting air sick in the simulator, but in the airplane he was a rock.

"Terry," I asked after strolling through the aircraft's interior, "why is there a very large rock in the forward cargo compartment?"

"It's for my garden, I picked it up at a local garden supply. We don't have these back home."

Terry wanted that rock and it seemed to him the best place to store it was right on the airplane. I didn't have a problem with it, so long as it was secured so it wouldn't shift in flight and we knew how much it weighed.

"How the hell should I know how much it weighs?" he asked.

"Terry, we either need the exact weight or you need to take it off the airplane until our tests are done."

"Major," he said to me in a professorial tone, "your airplane will weigh at least six hundred thousand pounds during the test flight, my rock ain't gonna make any difference."

"Terry, it's gonna have to be like this." I was trying to avoid the major to sergeant lecture, but my odds were looking slim. "We have controllability issues and I want to rule out center of gravity as a variable. You either weigh the rock or you take it off the airplane."

"You know how many guys it took to get that thing on the airplane?" I said nothing. He took the rock off the airplane.

The Boeing engineer's first theory about our controllability issue was a mis-rigged aileron system. Every control surface on the 747 is hydraulically powered with no cable backups. The actuators were controlled by levers and pulleys, of course, but the forces were all a function of hydraulic piping diameters, pump forces, and artificial feel systems. Any link in that chain could have been at fault.

I brought the schematics to my hotel room, trying to formulate my questions for the next day's session with the Boeing technicians. I was deep into the flight control rigging manual when the phone on the generic night stand rang.

"James, it's Lieutenant Colonel Larson. We need you to submit a volunteer letter for school."

"I don't know that I want to go to Air Command and Staff College."

"If you don't volunteer," Larson insisted, "you ain't going. So you should at least volunteer, you can always turn it down."

Of course if I did that, I would be stealing someone else's opportunity. Air Command and Staff College, what we all called "colonel charm school," was a guarantee for promotion but also a good way to end a flying career. Getting promoted would be nice, but very few lieutenant colonels in the Air Force get to fly airplanes. "I'll think about it."

The next day I was only thinking about stall speeds.

"Okay guys," I said over the interphone system, "pre-stall checks are complete, I am hand flying the airplane at twenty-five thousand feet, the airplane weighs six-hundred, twenty-four thousand pounds, the air is calm, and the configuration is clean. I just need a target stall speed."

"Looks like about one-thirty." I could see copilot Dave Winters chasing through the airplane performance charts with the end of his index finger; not the kind of precision I was looking for.

"I don't want about, I want an exact number."

"But, James." I stared at him and he stopped midsentence.

Stalling a Cessna 150 is something all private pilots learn to do so that they may forever avoid doing just that on larger aircraft. The only reason we were doing it on our considerably larger airplane was to validate its low speed flying characteristics. I had done it many times and the process was methodical. Because it was methodical, it was safe. I would decelerate the airplane at precisely two knots per second while keeping the wings level. The moment I felt a tremor in the controls, I knew the airflow over the wings was starting to separate. That would be the "stall onset" and Boeing provided data to predict exactly when that would happen. If it didn't happen or I somehow missed it, we would risk missing the stall onset and flying into the stall. As long as I kept the airplane in coordinated flight - no adverse yaw - the airplane would continue to fly. But since we had an airplane with controllability issues to begin with, we couldn't guarantee an adverse yaw free condition. Every pilot knows a stall plus adverse yaw equals a spin. Every 747 pilot knows a spin on an airplane this size cannot be survived. We obviously couldn't risk that. We had to know at what point to abort the maneuver.

"One-thirty-one point five," he said.

"Good. Start timing. Pulling half a knob on two."

I pulled the two inboard throttles back about half a knob-width, about a half-inch. Dave read off the airspeed and it looked like about two knots per second. At 150 knots I felt a tremor in my left hand. It had to be a mistake. At 146 the tremor was a rumble.

"Onset at 146," I said while pushing the throttles forward slightly, "that can't be right."

We repeated the entry four times, each time with the same results. It was time to tear the airplane's flight controls apart.

After two days of looking into inspection panels, measuring hydraulic pressures, and testing individual control units, we had nothing. I alternated hours with the mechanics and technicians with phone calls from Lieutenant Colonel Larson.

"We need a highly qualified major from this squadron to volunteer for ACSC," he said with no sense of irony, "and I'm going to volunteer you."

"What about next year?"

"What about it?"

The next theory for our too-early stall was something pilots have known about for a while, but usually isn't a problem on a 747: spoiler blow up. We have several panels on the top of each wing that have three functions. First, they assist the airplane's ailerons when rolling the airplane left or right. In a right turn, for example, the right spoilers deflect upward, spoiling the lift on only one wing. The result is a crisper turn. On the ground all the spoilers go up and act as ground spoilers to help bring the airplane to a stop. In the air they can also be used together as speed brakes, to help the aircraft descend. During slow speed flight, the air on top of the wing becomes a vacuum that can suck the spoilers out of position, the so-called spoiler blow up. That would explain our too-early stall.

Since you can't see the wings from the cockpit of a 747, we got permission to put a company technician in a window seat next to each wing. "Tell me what the spoilers are doing."

I snuck the throttles back and Dave called the airspeed decay. I felt the rumble. "Anything on the wings?"

"Left wing, nothing, sir."

"Right wing, nothing, sir."

A few more knots. "Anything?"

Silence, and then, "spoiler deflection right!"

"What about left?"

"Nothing."

How can that be? I aborted the maneuver and ordered the observers to the cockpit. Perhaps the left wing technician didn't know what to look for. "Let's try it again, this time trade wings."

"Spoiler deflection right!"

"Left wing, nothing."

It didn't make sense. Why just one wing? I inched the throttles forward and kept the airspeed steady. The tremor in my left hand was now a constant buzz. I looked at my hand. I was holding the right wing down and the yoke was deflected a good thirty degrees. The spoiler on the right wing was deflected because I was telling it to deflect.

"Look at the ailerons."

"You're causing the deflection," Terry said. "Major, why are you doing that?"

"I'm keeping the wings level," I said, "I guess subconsciously to keep from yawing the aircraft. Of course I'm doing just the opposite."

I pushed the throttles forward a bit. "Let me think about it."

Something was causing the aircraft to roll left and I was counteracting that by raising my left hand to command an opposite roll. So maybe the spoilers aren't the issue at all.

"Okay," I said over interphone, "here's the plan. We'll enter the stall with the airplane in a thirty-degree bank to the right. I'll try to time the airspeed decay so that I can keep the ailerons flush until we've rolled into a thirty-degree bank to the left. We'll record that speed and call it a day."

We got it right on the first attempt. As the speed dipped below 150, I centered the aileron input and the airplane rolled slowly to the left. As the wings were parallel to the horizon the speed was about 140. At thirty degrees bank to the left we were at the precise stall speed and the tremor in the controls returned.

"The airplane is stalling at the correct speed," I announced, "our real problem is an uncommanded roll at low speeds."

We were doing all our flying from an aircraft plant an hour east of Dallas, living in a filthy Holiday Inn, and eating most of our meals at a Burger King. The rest of the crew found ways to keep themselves diverted each night while I looked at schematics and argued with squadron mates over the phone on one topic or another.

It was an open-ended arrangement. We were stuck there until the airplane was fixed. After the second week I got a stack of mail that included an envelope from the Air Force personnel center. I had officially volunteered for Air Command and Staff College.

"But I don't want to go." I called Lieutenant Colonel Larson at his home.

"You didn't tell me that, in so many words."

"Well, that's what I meant."

"Don't worry about it, James. The selection rate is less than one-half of one percent."

"It's a Q-spring on the top rudder," the plant's chief engineer announced, "we are certain of it."

After our fourth check flight and discovery of the roll problem, the aircraft was attacked by technicians and mechanics looking for anything that could cause an uncommanded roll. I remember learning about the Q-spring in school, but hadn't given it much thought ever since.

The rudder on the 747 is huge. Actually there are two, one on top and another below it. They are attached to the vertical fin and are connected to pedals at each pilot's feet. With a press of either pedal the pilot can induce a yawing moment to either assist the coordination of a turn or to align the aircraft with the runway on takeoff or landing. Because the rudders are so huge the hydraulic system provides nearly all the force necessary. As a result, the pilot is left with little or no feedback in his feet and as such, doesn't have a "feel" for what the rudders are doing.

The Q-spring on older Boeings is just that, a spring that gives the pilot a small measure of feedback. On the 747 it is a bit more complicated, but not much.

"The top Q-spring unit was actually seized into one piece so that it only deflects the rudder to the right." The engineer explained that it would be hardly noticeable.

"Would it explain our issues during air refueling and landing?"

"I find that hard to believe," he answered. "But I'm not a pilot. I guess anything is possible."

In two days the entire top rudder actuator assembly, including the Q-spring unit, was replaced. The next day the airplane stalled completely normally. The day after that we ordered a tanker and had the time of our lives air refueling.

"This is easy," Dave said while inching the airplane forward for his second contact with the tanker, "I could do this for hours."

"Congratulations," the general officer on the phone said. When I heard the call was from Montgomery, Alabama I knew I should have been concerned. "You've been selected for next year's Air Command and Staff College."

Postscript

When I was handed the functional check pilot job I thought it was almost fate. I was an engineer turned pilot and I was meant for the job. I was planning to entrench myself into that job for the long term. And then fate dealt me a new hand and I found myself at Air Command and Staff College. Among my five hundred classmates were a number of Army officer who were fond of saying: "No plan survives the enemy." But still you plan. The better the plan, the easier it is to recover from the next act of fate.