Sunday, January 18, 2015

How flying a glider makes me a better airline pilot


Like any endeavor, a well rounded education makes you more insightful, if not better at it. 
Can soaring (i.e., glider flying) make you a better pilot airline pilot? I think so.

Every glider flight, whether by aerotow or winch launch, starts out with a precision flying exercise. Being connected to the airplane in front of you with a rope does command a certain degree of concentration. One must avoid the wake of the tow plane, maintain proper tension on the tow rope (the tow plane and the glider often don't encounter rising or sinking air at the same time), and maneuver to the outside in turns. 
Flying a glider requires an appreciation and understanding of a whole set of factors that an engine makes it easy to ignore. Keeping a glider aloft for hours at a time requires putting the glider in air that, on average, is going up faster than the glider is going down.  So, there are two factors: can you find air that is going up, and how you control how fast the glider is going down. 

The first requires an understanding of the micro-meteorology. 


In soaring, I found an appreciation for how it works, for seeing a cloud not just as a puffy object, but adding to it all the airflows that make it and surround it. Essentially asking "what makes that cloud that shape, and how will it change over time?"   As airline pilots, we fly through clouds all the time—mostly we just plow right through. But we're also looking for a smooth, efficient, and safe ride even if we don't have to try to harness the power of the airflow around it to stay up. The thermals, updrafts, downdrafts, and waves make those clouds and turbulence. In the same way that a lion tamer should know about lions,  I think that appreciation, that better understanding of the environment I fly in makes me a better pilot as a result. 
The second factor, how you control how fast the glider is going down, requires an understanding of aircraft performance. 



When there's no engine, efficiency is the name of the game. It would be easy if there were just one optimum speed to fly—but there isn't. There is a speed that provides a minimum sink rate—that's good for thermalling, flying tight little circles trying to get the most altitude out of the rising air. But that very slow speed (just a few knots above stall) isn't a good one to make a distance over the ground. That is another speed. But, that changes with the wind and how fast the air your in is rising or sinking. Counter-intuitively, in air that is sinking the best approach is to push the nose down! That's to go faster and get out of that sinking air quicker. One also needs to go faster with a headwind than tailwind. (When you're only going 50 mph, it really makes a big difference.)
Those same principles apply equally to a jet airplane!  For example, in the event of an engine loss at altitude, the question is: "what speed to fly". Do I need to minimize my sink rate (to avoid traffic below), make the best distance per altitude (to clear that mountain range), do I need to make the most fuel efficient diversion, or the fastest one?

All airplanes could be gliders, and there have been a few classic examples when some big ones unexpectedly became a glider. Here's three:
  • Air Transat #236  An A330 that ran out of fuel over the Atlantic and glided to a safe landing in the Azores.
  • Air Canada Flight 143 aka, "the Gimi Glider.   A Boeing 767that ran out of fuel at 41,000 feet , about halfway through its flight originating in Montreal to Edmonton, and glided to a safe landing.
  • USAirways 1549 The "Miracle on the Hudson"
In each case, the captain was an experienced glider pilot. Hard to argue with success!
Along those same lines, energy management—a key element in the instances above—is always on a glider pilot's mind. You always have to be able to make it back to the airport (or to a different airport) and a go-around is not an option!

In a recent simulator session  we ended up shortly after takeoff with no engines operating (a fire in one and failure in the other). The objective was apparently to do a ditching drill. But, why put it in the water, when you can land on the runway—which is what I did instead. (A glider pilot is always aware of when and how he can turn back to the airport in case the tow rope breaks!)

I've also found that gliders (or other small acrobatic airplanes) can be a great resource to expand a pilot's attitude envelope. This may come in quite handy in the event of an upset event.  When you're upside down for the first time, hanging by your seatbelt, and all the dirt and other objects that aren't tied down are falling up in front of you-—it can be a little disorienting! "Tunnel vision" comes to mind.  Training in aerobatic capable aircraft can prepare you to handle an extreme upset—like an inadvertent case of upside-down. The intuitive answer is not always the right one!

Then there's the aspect of no-autopilot. Somehow that should count for triple the time spent in an airplane for hour requirements! Of course there would be few times when an autopilot would be of any use, as a glider pilot is almost constantly changing speed and direction to maximize the flight.

It won't happen magically when the new glider pilot solos or gets that new ticket. It will take some time, some effort, some thinking about it, and a lot of fun along the way!

Saturday, January 10, 2015

Comparing QZ8501 to Air France 447

On December 30, 2014 two days after the loss of Air Asia Indonesia (QZ) flight 8501, CNN asked me to provide a comparison between QZ 8501 and Air France 447.
The following was published 12/31/14

On Sunday, all contact with Air Asia flight 8501 was lost over the Java Sea as a wide area of thunderstorms covered the area. The discovery of floating debris on Tuesday about 100 miles from its last known position in combination with an analysis of ocean currents will give investigators clues where to search for the remainder of the aircraft. From its cruise altitude, the airplane’s gliding distance would also be about 100 miles, but consider that for the debris to drift that same 100 miles it would only take a drift rate of 2 knots, yielding a wide range of possibilities as to  the nature of the aircraft’s descent to the water below.

Many parallels between Air Asia 8501 and Air France 447 in June, 2009 are obvious. Both aircraft were lost in thunderstorm areas of the Intertropical Convergence Zone (ITCZ). Both were found within a few miles of its last known cruise altitude position, both were sophisticated fly-by-wire Airbus aircraft (though different models), and both crashed at sea.

While flying into a thunderstorm is always to be avoided, it not likely the sole cause of the accident.
The weather in the ITCZ has some unique qualities compared to your average thunderstorm over land. The storms are driven by the convergence of airflow patterns between the northern and southern hemispheres of the Earth in addition to the usual factors of warm moist air and unstable atmospheric conditions. The height of the stratosphere –- which tends to put a cap on the height of thunderstorm growth, and averages about 35,000 feet over the mid latitudes (such as that of mainland USA), reaches to 50,000 feet or more, providing for the growth of thunderstorms to great heights and accompanying intensity. These features can lead to some unusual conditions within those storms, making the proper assessment of them with airborne weather radar more difficult.

In the aftermath of the Air France crash significant emphasis has been made in pilot training on the prevention and recovery from similar scenarios. I would say that few pilots, especially of Airbus aircraft would be unaware of AF447’s lessons, almost certainly one with the reported experience of QZ8501’s captain.

There is a recent development however that relates to Airbus A320 series aircraft. A December 10, 2104 Airworthiness Directive (AD 2014-25-51) describes how control of the aircraft could be lost in flight as a consequence of icing of the angle-of-attack probes and an interaction with the airplane's stall protection function. Those probes act like small weather vanes on the side of the aircraft and measure the angle at which the airplane moves through the air--the angle of attack. If the angle is too high the air can no longer flow smoothly around the wings, resulting in an aerodynamic stall. The acceptable range of angles of attack is fairly small, and gets considerably smaller at higher speeds, such as cruise speed.

Simply put, depending on the position of the angle-of-attack probes when freezing occurs and the subsequent speed of the aircraft, the system may be fooled into thinking that the aircraft is approaching a stalled condition-even when it isn’t. In response, the airplanes stall protections pitch the aircraft’s nose down to recover. This erroneous pitch down cannot be overridden by the pilots unless an emergency procedure in the Airworthiness Directive is followed. All pilots flying this model airplane should be aware of this.

The procedure instructs the pilots to shut down two of the three air data computers to render the usual stall protection inoperative an allow recovery of the aircraft. Of course, there is no way, at this stage of the investigation, to know if this played a part but investigators will certainly be looking for evidence of this phenomenon.

Another obvious question is the apparently lack of transmitted position and altitude data after its last known position in cruise. This data is transmitted throughout the flight by a system known as ADS-B (Automatic Dependent Surveillance-Broadcast). This system transmits the airplane’s position and other basic data to ground stations. Though its position is GPS satellite derived, it is not transmitted to satellites, only to ground stations – so the range to the nearest station is a factor.

The apparent sudden loss of this data at cruise could be explained by failures in flight such as an electrical failure, in-flight breakup of the aircraft, or the pilots switching off required data to operate the system such that outlined in the emergency procedure above. However, it could also be that the aircraft simply flew out of range of the ground stations. Flight tracking websites indicate that this routinely occurs in the general area where QZ8501’s last ADS-B transmission was made. I think that is the most likely cause of the end of the data stream and is not necessarily an indication of catastrophic failure in flight.


In the case of Air France 447, the aircraft came down in the Atlantic Ocean where the sea depth exceeded 12,000 feet. While some floating wreckage and a of number of bodies were discovered within a few days on the surface, the extreme depth and rough terrain on the ocean bottom delayed discovery of the remainder of the aircraft and recovery of the flight recorders for two years.

Fortunately, the 100 foot depth of the Java Sea in the area where evidence of QZ8501 was found will almost certainly result in the relatively rapid location of the aircraft and recovery of the two flight recorders. Consideration of ocean currents during the two days between the aircraft’s disappearance and the discovery of floating debris will help lead investigators find  the remainder of the aircraft and its passengers. We should not be subjected to long period of uncertainty such as with AF447 or the continuing lack of information on MH370.

In the aftermath of the Air France crash significant emphasis has been made in pilot training on the prevention and recovery from similar scenarios. I would say that few pilots, especially of Airbus aircraft would be unaware of AF447’s lessons, almost certainly one with the reported experience of QZ8501’s captain.

While any accident investigation will take months to complete, I would expect more information to be available as the search and recovery continues. Clues from the way in which airplane parts were damaged on impact, the flight data and voice recorder contents will provide answers. But like any aircraft accident, the cause is likely to be the result of a chain of events and conditions, the absence of any one of which would have avoided this tragic accident. At this time we can only guess what some of those events and conditions are.