Graveyard Spiral

The graveyard spiral, or "spiral dive," has claimed the lives of many pilots. In my years as a flight instructor, I've come to find that graveyard spirals are not well understood by many pilots. That's part of the reason graveyard spirals continue to kill people. I believe education is the answer to managing risks in flying, so I'd like to provide some information about spiral dives and how to respond to them and, more importantly, how to prevent them in the first place.

First of all, what is a graveyard spiral? Graveyard spirals usually occur in instrument conditions when the pilot becomes spatially disoriented. The pilot's spatial disorientation could be the result of turbulence, instrument failure(s), lack of instrument training, or simply cockpit distraction. Whatever the case, the pilot loses control of bank attitude and one wing begins to drop. As the bank steepens, the nose falls through the horizon and the aircraft begins descending. The airspeed increases and the bank angle and rate of descent continue to increase. The pilot senses a descent and pulls back on the yoke or stick in an attempt to arrest the descent, but this only steepens the spiral. The load factor increases dramatically and the descent rate becomes incredibly steep. Often in fully developed graveyard spirals the aircraft suffers an in-flight breakup before it reaches the ground. Other times the aircraft strikes the ground in a nose-low attitude at extremely high speed. Accident sites of graveyard spirals are usually breathtaking... Thousands of aircraft fragments (and other things) scattered around as though a bomb went off.

Recovery from a fully developed spiral dive is sometimes impossible. Once the airspeed exceeds a certain value, structural failure will occur during any type of recovery attempt. For this reason it's very important that we never allow a spiral dive to fully develop. The answer to preventing the formation of a spiral dive is to avoid situations where we might become spatially disoriented. Never fly in instrument conditions unless you're instrument proficient, not just instrument current. Minimize cockpit distractions and always remember keeping the airplane under control is your top priority. Never conduct other cockpit tasks during turns because there is a higher risk of losing control of bank during a turn. Use an autopilot, and always maintain a proactive instrument cross-check to detect instrument failures. Luckily, instrument failures are much lesson common now with glass panel systems and laser gyros. We don't have to rely on undependable vacuum pumps anymore for attitude information.

If a spiral dive starts, rely on your unusual attitude recovery training. Reduce power to idle, level the wings FIRST (that's very important because leveling the wings reduces load factor exponentially), and slowly recover from the resulting dive. The most important component of attitude during flight is bank. If you control bank, there's very little pitch can do to hurt you. If the wings are level, positively stable airplanes will fix pitch on their own and the nose will eventually return to a safe attitude after oscillation cycles. Some new airplanes are being built with a "wing leveler" feature for this reason. If the pilot feels like he's losing control of the airplane, he hits a button and the autopilot levels the wings. If the wings are level, a spiral dive can't develop. Remember, spiral dives are the result of runaway bank. Maintain control of bank, and you won't find yourself in one of these terrible situations.

AAL 2

Have you heard about this? On May 4, American Airlines Flight 2, a Boeing 767-200, declared an emergency on approach to JFK after ATC denied their request for landing Runway 31R. Here's part of the ATC audio.

JFK's Runway 31L was closed for repairs, so I believe Runway 31R was being used for departures and Runway 22L (and Right?) was being used for arrivals. When American 2 Heavy checks on with Tower, the controller issues a landing clearance and wind check of 320 at 23, gusting 35. That's a heck of a lot of crosswind (and even some tailwind component!). The crew declares an emergency, makes a turn, and lands Runway 31R. This was an unusual event, and it's difficult to determine whether the crew acted appropriately without knowing the full details of what happened.

I'd like to hear the ATC audio from New York Approach between the controller and this flight crew. The conflict started there, and I'll bet the crew requested landing 31R multiple times and the controller didn't oblige. New York Approach controllers can be stubborn. After hearing the wind check from Tower, the crew, fed up by the Approach controller's lack of assistance, decided their only option was to use their emergency authority and land their aircraft where they deemed fit. I'm assuming there was a minimum fuel situation, and if there was, I wonder if it was communicated to ATC. If there wasn't a minimum fuel situation, why didn't the crew divert to another airport with a suitable runway for landing? Not enough fuel, probably.

After the crew declares an emergency, the Tower issues them a go-around vector which they do not comply with. This seems out of line unless the fuel state was dire. Listening to the Tower audio, it sounds like the controller understood the flight's situation and was attempting to initiate vectors for an approach to 31R. I think the crew was fed with ATC at that point and their brains had switched into "I'm not listening to anything you have to say anymore" mode. That's understandable, and I don't necessarily disagree with the crew's decision to override the controller's instructions but I'm not sure that was totally necessary when the controller was already trying to send them to 31R. Again, maybe fuel state was a factor. Maybe the crew felt they couldn't take more vectoring.

So, was this a justified "emergency?" That depends on the fuel remaining. If the flight was at minimum fuel, yes, this was an emergency. They needed to land now at this airport, and the only way they could do that safely was to land on a runway better aligned with the wind than the runway ATC allegedly insisted on giving them. If the flight had fuel to divert, no, this was not an emergency. The FAA will decide.

Crosswind Limitations

Ideally, wind would always be perfectly aligned with runways. When runways are constructed, a consideration of the average wind direction for the area dictates the runway orientation. Los Angeles International, for instance, has east/west runways because of its proximity to the ocean and prevailing east/west winds. Dallas Fort Worth International, on the other hand, has mostly north/south runways because the wind in the area usually blows north/south. The problem is, fronts and other meteorological factors disrupt wind flow sometimes, and uncharacteristic crosswinds make pilots tighten their grip on the controls a bit.

The FAA requires airplanes to be satisfactorily controllable with no exceptional degree of pilot skill or alertness in 90° crosswinds up to a velocity equal to 0.2 Vso. For example, Vso, or the stalling speed in the landing configuration, for the Piper Archer I fly is 47 knots. That means the airplane can be landed by the average Joe pilot who isn't particularly skilled or alert in a direct 9.4 knot crosswind. A non-average Joe test pilot, on the other hand, who is highly skilled and alert, picks a windy day and lands the airplane with as much crosswind as he can handle and that number becomes the maximum demonstrated crosswind component for the airplane. Any airplane certificated after May 3, 1962 must have this number placarded in the cockpit where the pilot can see and be reminded that a test pilot could handle that amount of crosswind, but he probably can't. The maximum demonstrated crosswind component is NOT a limitation, and it can be exceeded. You can legally land with as much crosswind as you want. But if a test pilot could only handle say, seventeen knots, it's reasonable to consider that a limitation, or at least a very strong recommendation not to take on more than that (or even close to it).

When the wind is blowing across the runway and not along it and you're not certain you can land safely in the existing conditions, divert to an airport with a crosswind runway better aligned with the wind. Better yet, avoid that situation in the first place by giving a thorough check of the weather and forecasts before your flight. Wind isn't generally all that difficult to forecast, so if it looks like there's a possibility of encountering too much crosswind, delay the flight until the wind dies down or come back another day.

Looking for Students

Hello, readers. I'd like to take a moment to extend an offer to anyone seeking flight instruction to contact me. I'm taking on new students and would be glad to hear from you. I offer a full range of instructional services from Private through Airline Transport Pilot, including instrument, flight reviews and instrument proficiency checks. Please email me at ee.stuart@gmail.com to discuss your training needs or schedule a lesson. I look forward to hearing from you! --Gene

Salute to Captain Burkill

"It changed my life." That's what Captain Peter Burkill of British Airways Flight 38 said about the crash of his Boeing 777 in London on January 17, 2008. At four hundred feet on final approach into London Heathrow both of the 777's engines failed without warning or explanation. Listen to Captain Burkill's recount of the crash here. Due to Captain Burkill's proper management of the emergency, there were no fatalities.

Captain Burkill and his family suffered through a difficult year following the accident. Nasty rumors began to spread about Burkill's ineptitude as a captain, and how he "froze" on the flightdeck during the emergency. In fact, British Airways cabin crew trainers were propagating the insulting gossip and talking badly about Captain Burkill during recurrent training sessions. Not long after the accident, Burkill was shocked while cruising over the Atlantic when a couple of his flight attendants informed him of what was being said about him during their annual safety equipment training session. Negative media perceptions were formed of the captain, too.

What a shame. Captain Burkill did everything right, in my opinion. The entire emergency lasted less than one minute before impact. During that short time Burkill remained calm and did everything within his power to to manage the situation. He quickly troubleshot and tried to identify what may've caused a sudden loss of thrust on both engines simultaneously. When it was apparent power could not be restored and there was not enough time to continue further with that effort, Burkill made a smart split-second decision to shed drag by reducing flaps by five degrees. This action alone may have been the difference between life and death for passengers and people on the ground. Burkill was aware that reducing the flaps by five degrees would increase the glide range while not dangerously increasing stall speed. And within seconds before impact he made a mayday call to the tower so that firefighting and rescue equipment could be dispatched immediately.

Burkill may have been criticized for not taking aircraft control from his first officer, who was the pilot flying for the landing. Actually, I think that was one of the best decisions he made that day. Burkill knew his first officer was highly experienced in the 777, almost as experienced as he was. And Burkill knew that time was so limited before impact that he needed to have full, undistracted access to his captain's knowledge and wisdom. I think his decision to let his very capable first officer continue flying was a fine display of crew resource management (CRM). Had he taken the controls so low with so little time, Burkill may not have thought to partially raise the flaps and the accident could've ended in tragedy.

In my opinion, Captain Burkill's handling of that unprecedented and very challenging emergency was a fine display of airmanship. The man deserves respect and recognition for a job well done. Many people could've died that day, but everyone survived. It disappoints me how disrespectfully Captain Burkill was treated after the accident. I personally would like to extend to him my respect and congratulations for his work that day. Well done, Captain Burkill.

Directing Attention

I keyed the mic, "Direct DUSTT, hold as published, maintain 3,000, EFC 0315, Nine Four Charlie." The world outside was dark and dimensionless. Pitch black, actually. It was a low IFR night and I was single pilot. The engine noise had just transformed from a relaxed purr to a roar as I executed a missed approach. During instrument training, pilots learn to expect a sensation of tumbling backward during a throttle up in IMC. I felt that sensation that night and responded as I was trained -- disregard the sensation and hunker down on the instruments. Even our own bodies lie to us sometimes. It's important not to allow that unusual sensation to cause hesitation during application of climb power. I was climbing, turning toward the missed approach fix, configuring things for the hold, cleaning up the airplane (retracting flaps and completing the after takeoff checklist), and I had no autopilot. I was task saturated. Suddenly, I stopped.

I've trained myself to do this whenever I feel the current workload is becoming overwhelming. I stop what I'm doing, put my non-flying hand on my leg, and just fly. I transition back to the flight instruments and reallocate my attention to attitude-instrument flying. "Am I sending the airplane to where it needs to go? To a safe place?" Anything else can wait. The most important thing is to keep the airplane in a safe attitude and positively under control. I have programmed a mental alarm of sorts that monitors how much of my attention is on flying the airplane. When that figure dips below the 75% range, the alarm trips, and I respond by suspending all tasks that don't directly involve flying the airplane. That night, I climbed to the assigned altitude, established on-course to the missed approach fix, and resumed other cockpit tasks only after I was straight-and-level. If the airplane departs from controlled flight, all those other tasks don't matter anyway.

During single pilot IFR operations, this type of thought discipline is especially important. Too many airplanes are lost to loss of control accidents in instrument conditions. Remember, if you don't fly the airplane, no one is. During high workload or task saturated situations, remember that the only thing that matters is to first stash the airplane in a safe place in the sky, nicely under control. There are some situations in instrument flying where there is simply too much for a single pilot to think about all at once. That's where workload management tactics, or task prioritization, comes in. We all know how many accident reports cite a loss of situational awareness as a cause or contributing factor. The most important aspect of your situation is always the attitude of the airplane and its position in space, particularly in relationship to terrain and obstructions. Although there are many other important situational elements to maintain awareness of, they are all secondary to this. The airplane must remain under control for any of those other tasks to even stay possible. Pilots have been lost when their attention was diverted from basic airplane control because they were fixated on another task or problem.

I recently completed an instrument currency session in a flight training device. The flight was approach intensive, as most instrument currency flights are. There were lots of rapid re-configurations, frequency changes, OBS resets, GPS tasks, and checklist duties. I was a busy pilot. I noticed my mental "stop and redirect" procedure was being used frequently due to the high workload. I'd be programming the GPS for the next approach and would suddenly and very deliberately move my hand away from the dial mid-rotation and put it on my leg. It almost always feels unnatural to do this. Often the GPS would be left with a flashing highlight on a selection screen. No matter, it can wait. It won't go anywhere. I needed to focus on flying and keeping the airplane heading in a safe direction.

This is my own personal workload management tactic that I've found to be very useful. It's effective at breaking a fixated thought pattern. If I feel that the GPS or finding a number on a chart is consuming too much of my precious attention--of which there is only so much of to go around--or that I'm becoming fixated, I break that thought pattern by forcing my hand to leg. It's a way of not letting my mind address anything at that moment other than flying the airplane. This procedure reminds me of Cesar Millan's, The Dog Whisperer, technique of touching a dog in a specific place on the body and making a "Sh!" sound when the dog is acting out. He's breaking the fixated thought process and redirecting the brain's attention.

I had a sharp flight student who I would occasionally challenge with task saturation. I wanted to see how he'd respond when the environment was asking too much of him all at once. I'd be barking commands at him and playing ATC as he was running checklists and flying. When the workload became too great, he'd begin to shed lower priority tasks and focus on flying. I was so proud when he'd ignore me and hunker down on the instruments. After he was sure he had the airplane under control, he'd begin responding to me. He reacted perfectly.

Pilot-in-command authority is more than just a term in the regulations. It's a mindset. It's an attitude. We have to be able to take charge of the situation, to be the "alpha dog" up there. When we are bombarded by too much input, we have to take charge by confidently selecting which tasks to focus on and which to address at a later time, and we must do this authoritatively. When we feel our minds losing situational awareness, we must recognize this and react by immediately eliminating distractions to flying the airplane. Get back to the rest when it's safe to do so.

We talk about resource management in flying. To me, resource management extends to the most valuable resource on the airplane, the pilot's brain. That brain has limited capacity. Too many demands can start to pull attention away from flying the airplane, and most of the brain's resources should be continuously attached to that effort. If attention starts peeling away from aircraft control, it should activate a mental alarm and that attention should be restored and re-attached to flying immediately. Be aware of where your attention is directed. Keep enough attention in the right places, that's what situational awareness is all about. If things start slipping, stop and just fly.

Brief Yourself

You might be thinking, "He's talking about briefings again?" Yes, I am. Briefings in GA are widely underused and under-rated! I've previously written about approach briefings, which are arguably the most important of all briefings. But, there are others that deserve mention, too. Several different briefings are used routinely in airline operations, and we need to beef up our "briefings culture," for lack of a better term, in GA. Today, let's discuss departure briefings.

Remember, it's important to conduct briefings even in single-crew (single pilot) operations. In multi-crew environments, briefings benefit the briefer just as much as the brief-ee. It's been proven that information that is verbalized tends to stick better in our brains, which is why it's critical to brief aloud, even if you're the only one listening. In a moment, I'll give an example of a self-briefing I conducted for a departure out of Arcata, California on a low IFR morning. First, let's go over some departure briefing basics.

A departure briefing should be performed before every departure, no matter how simple or routine it may seem. VFR, IFR, good weather, bad weather, familiar/unfamiliar airport, no matter the case, conduct a briefing. Pilots screw up even simple departures sometimes, and they pay the ultimate price. The more organized your mind is, the more sharp your flying becomes. A departure briefing should be performed just prior to engine start and should cover these items:

Current position on airport and anticipated taxi route

Determine your current position using the airport diagram and brief the taxi route you anticipate (consider highlighting during your preflight planning). Discuss any taxiway closures.

Departure runway information

Determine which runway will be used for takeoff, available runway distance for takeoff, runway lighting, etc.

Runway conditions

Is the runway dry, wet, icy, contaminated? Abnormal runway conditions may increase takeoff roll distance.

Takeoff procedure and special considerations

Discuss what type of takeoff is required: normal, short/soft-field, crosswind. Discuss aircraft configuration requirements (flap settings, etc.).

DP walkthrough/departure transition and automation mode (if applicable)

If a departure procedure will be flown, brief it the same way you would an instrument approach procedure. If a DP will not be flown, discuss the plan for a custom departure transition. That could be as simple as "left turn on course, up to 5,000." Whatever your plan is for transitioning from the runway into the enroute structure, brief it. If autopilot will be used, discuss when and what modes.

Terrain/obstructions

Any terrain or obstructions of interest should be noted. Keep it simple, otherwise you'll forget anyway. Something like "high terrain to the east" works. The idea is to know which way NOT to turn in the event of an engine failure or other emergency.

Emergency considerations

Discuss the plan for an emergency return to the airport if one becomes necessary. I don't recommend departing from an airport where the weather is too low to permit an emergency return, but if there's a good alternate nearby, discuss the route there and the approach to be used.

Here's an example of a departure briefing I conducted for a flight in a Piper Archer out of Arcata Airport in California:

"We're currently at the GA ramp here. We'll plan to join Alpha via Delta from the ramp and taxi southbound on Alpha to Runway 32. Looks like there is a run-up pad at the approach end of Runway 32, so we'll conduct a run-up there. We'll use Runway 32 for takeoff. We've got six thousand feet of takeoff distance available. We do have edge and centerline lighting for this runway. The runway is wet but we have plenty of available distance for the roll. We'll do a normal takeoff, little bit of a left-to-right crosswind. We're cleared for the HOCUT THREE departure, Crescent City transition. I have the HOCUT THREE departure, Arcata, California, Alpha Charlie Victor. This chart is current eleven March 2008 to eight April 2008. For departure Runway 32, it'll be a climbing left turn to join the 250 outbound Arcata to position HOCUT, then a right turn to join Victor 27 northbound to TRIAL, then Crescent City. We'll climb unrestricted to 9,000. I'll hand fly until established outbound on the 250, then engage NAV GPS roll steering and Vertical speed modes. We'll use GPS as primary for navigation with VHF data as backup. I have Arcata set and positively identified on NAV2 with 250 set on the OBS with Crescent City in the standby. Once we turn northbound to join Victor 27, it will be the Crescent City 161 and 341 on the OBS. We'll be out over the ocean almost immediately with high terrain to the east. If we have to return we'll proceed directly to KNEES for the ILS 32 and continue climb to 5,900. Departure briefing complete."

I've always enjoyed briefings and found them to be very helpful. They help instill a clear mental picture of the game plan and help things go smoothly as planned. There's a lot to keep straight up there, so verbalizing a clear and concise plan will help keep you on the straight and narrow. Blog post complete!