Thursday, July 25, 2013

The Recent Crash Of A Boeing 777.....

 Disclaimer:  The following is a presentation of a recent crash and discussion based on published information. Analysis of this information is that of the author of this report.

The recent crash of a Boeing 777 at SFO was very disturbing when one had the chance to analyze the data presented. Mainly, it seemed that no one was minding the shop. The defense of the flight crew that hinges on the autothrottle is almost ludicrous. I compare it to the driver who is caught speeding along at 75 telling the trooper that “I had the speed control set at 60”. It hinges on responsibility. Whether driving a car or more critically, flying a commercial airliner. The pilot in charge or the driver behind the wheel must be accountable for the actions of their vehicles, whether a plane or a car.

In doing some research on the internet, I was fortunate to come upon a detailed analysis* of a 2008 Boeing 777 crash that occurred at Heathrow airport in Great Britain. Due to the exemplary (and what one should expect) actions of the flight crew, a serious disaster was averted.

Let me present some facts about the Boeing 777 before proceeding further about specific actions. It entered the flight line in 1994. There have been only two or three hull type accidents since inception, the one above included. The recent crash is of particular interest as there was no apparent mechanical problem noted. The only area in question, according to the three pilots, referred to the function of the autothrottle, an automated throttle control linked to the autopilot etc.

Now let me get into some of the features of the B777. I just finished reading that it was designed using the fly by wire (fbw) concept. This protects the airplane from control inputs that might cause structural damage or exceed safe flight conditions e.g. a stall. On the pilots instrument panel as part of the primary display, specific warnings are shown when the plane is flying below the minimum maneuver air speed. The crashed plane at SFO was flying dangerously below safe airspeeds. Therefore I must assume that ample warnings were available. What was done by those pilots? Good question, it doesn’t seem like enough.

By comparison, the flight crew of the* B777 landing at Heathrow in 2008 did everything right. They followed established procedures, learned and practiced over their entire careers. When on final approach into Heathrow, there was a sudden decrease in thrust in both engines occurring almost simultaneously, due to an undetectable ice accumulation in the fuel system. The autothrottle was engaged at the time. The co-pilot, flying at the time advanced the throttles to try to increase engine thrust. However there was no response from the idling engines. The plane started slowing to below minimum safe speeds and the plane started to sink towards the airport outer boundary. The auto-pilot tried to pitch the nose up, but as this would further decay the airspeed, it had to be disconnected by the copilot. While the co-pilot was lowering the nose of the plane to prevent a stall, the captain was wisely reducing the flap setting to 25 degrees to decrease drag. But, alas it was too late to stop the inevitable, and the plane sank further until striking the ground just short of the runway. There was only one passenger injury, no deaths. The plane was a total wreck.

That the flight crew performed admirably is an understatement. When one compares what apparently the flight crew of the Asiana plane did, or rather didn’t do it is appalling. In reading further about the 777, it is stressed that there are all kinds of low speed warnings, prior to “stick shaker”. According to the flight report, there was a “master caution aural warning, indicative of “”airspeed low””.  What more was needed to alert the crew of three including a high time instructor? The accompanying picture above, taken from the cited article is an example of a typical (quite visible) airspeed to the left of the main screen.

Well now I hope you have a good idea of what can be done in a time of crisis with an airliner at near stall speeds only a few hundred feet above the ground. But more importantly I want to discuss responsibility in the cockpit. The pilot in charge (PIC) must know what is happening at all times. That means monitoring lots of data such as airspeed and attitude perhaps the most basic. When the plane is flown with the autopilot engaged, one could easily assume that all is well and become complacent. The same thing with the autothrottle that commonly is linked to the autopilot. These automated systems however are not absolutely infallible and must be monitored. If they do indeed fail or inadvertently disconnect, the pilot must be ready to immediately take over. If not a duplication of what happened on final to SFO could happen again i.e. passively sit by as the airspeed and altitude bleed away as the plane crashes to the ground.

Remember that the pilot in charge is responsible for the safety of his/her crew and passengers. Complacency just won’t cut it. You may get away with it for a while, but eventually some form of  disaster awaits.

*Aircraft Accident Report 1/2010, Report on the accident to Boeing 777-236ER, G-YMMM at London Heathrow Airport on 17 January 2008

Friday, July 5, 2013

Thoughts On A Recent Accident

Recently it was reported that a South Carolina businessman on a recertification flight in a Rockwell Turbo Commander 690B crashed over McClellanville, SC. Two people were killed, the 44 y.o. pilot and the 69 y.o. instructor. They had taken off from John’s island airport and were to operate between 13000 and 15000 feet as cleared by ATC. Supposedly the pilot-owner had been flying this plane for three years. Credentials of the instructor are unknown.

The purpose of the flight as stated above, was to recertify the pilot, presumably for insurance reasons. To comply, generally a set of flight skills are examined. These would typically include: take-off and landings, air work such as steep turns, slow flight etc. and of course instrument flight under simulated low visibility conditions. The other important area of consideration is putting the pilot through simulated engine out procedures. This normally is done by retarding power to flight idle on one engine, usually the so called critical engine (usually the left hand engine on most twins). Finally, the approach to a stall with one engine idled is tested, likely also with the pilot under simulated IFR conditions (pilot hooded). This latter flight test just may be the cause of the recent crash.

Now it has been a few years since I went through a similar procedure. It was when I flew part-time charters for a fixed base operator (FBO) in Burlington, VT. Every six months it was necessary to obtain a check ride with a certified instructor in order for me to legally fly charters. These charter flights included both passengers and/or freight.

Typically the check rides consisted of a variety of maneuvers including slow flight and this key: slow flight approach to a stall with one engine “out”. With an “out” engine being simulated by retarding the throttle of one of the engines, usually the “critical” one (see above). Before entering into this exercise, the correct procedure for carrying it out was clearly discussed with the instructor. The hazard to this maneuver was this*: if while slowing the plane by bringing its nose up, the plane was allowed to stall, a violent falling off towards the “dead” engine most likely would occur. This almost certainly would progress into a flat spin with possible aircraft inversion. If that were to happen, it most likely would be unrecoverable. The goal of the exercise was early recognition of an impending stall. Once the warning of a stall is received as for example by noting some slight vibration in the control yoke, a prompt push of the control yoke is done to start the plane down immediately. Some of the more sophisticated aircraft have various indicators of stall, including: lights, buzzers and other type annunciators. The reason for all of this is that a stall in a complex plane like the above twin has a reasonable likelihood of being a fatal event. (See NTSB report referenced below)

When I was practicing this on my check ride in a B-58 Baron, all went well. I brought the nose of the airplane up until the first indication of an imminent stall, and then quickly shoved the nose down, allowing the plane to regain a safe flying speed. The next week however, another pilot was undergoing the same series of maneuvers, including the approach to stall with one engine ”out”. This time however, the pilot waited too long to push the nose down and the plane entered the stall with the awful result discussed above. The plane, the same one I had flown, entered a flat spin and descended rapidly towards the waiting rough terrain. Fortunately the instructor pilot was a National Guard F-16 instructor pilot and was able to stop the spin just before the plane struck the ground. There were serious injuries, a destroyed aircraft but no deaths.
This scenario is what I think may have occurred over McClellanville.

*NTSB Aircraft Accident Report, North American Rockwell Turbo Commander 690, Wellsburg, West Virginia, 8/14/72 NTSB-AAR73-5, A-73-8 thru 10