Lies, Damn Lies and Instruments...

After reading Aviatrix's excellent summary of an Air Caraibe incident related to 'bad' air data and instrumentation malfunctions, I was shocked to discover the following information from a Boeing website:

During the past 10 years, more than 300 accidents and incidents have been reported as a result of erroneous flight deck information, including problems with pitot-static probes and air data computers. Several fatal accidents that involved erroneous flight instrument information and six incidents resulting from lost or erratic air data occurred in 1996 alone.

Some fatal and near fatal accidents have been tracked back to something as mundane as a strip of tape over a sensor port.  The Air Caraibe incident involved severe and sudden icing in cruise flight, and may prove to have some common elements with Air France 447.

To a pilot who has become finely tuned and trained to flying by instruments, suddenly having to determine which pieces of conflicting information are right and wrong and how to respond is a huge challenge.  During flight in turbulence at high altitude the margins for error can be very small (see previous posts about airspeed envelopes.)  A wrong decision and control input can quickly put a jet airliner into a situation from which there is little chance of recovery.

Understanding the causes behind Air France 447's demise is extremely important - making it even more urgent that the 'black boxes' are found.

A Day in the Life (38) Eastward Ho!

A description of a typical day’s flight from Montreal to Vancouver and return - as a backdrop for a detailed, non-technical description of what an airline pilot does. (check left hand column for series index).

Time: CYVR  Departure minus 0H00

We’re both a little more awake than we were in the Montreal darkness this morning.  So the checklists and start up procedures flow even more smoothly - and a little faster.

We’re soon waved off and under way and as our final load figures begin scrolling off the printer F/O Paula confirms that we’re loaded to the gunwales. 

We arrive at the button of runway 26L with the checks complete and I make one last check over my left shoulder to ensure no one will try to occupy the same space time coordinates before continuing onto the runway.  My plan to use a rolling takeoff is interrupted by a Beaver puttering southbound along the coastline.  I recite our take off speeds and confirm our initial climbout altitude as we wait.  Then:

“Goose Air, One One Two, you’re cleared take off, runway two six left.”

The thrust levers go up through the clicks to the TOGA slot, there’s a second where it seems as if nothing is happening except for a rise in the noise then slowly the Lumber Jet accelerates.

Rotation speed is 160 kts. and it seems to take forever to get there.  A gentle tug on the stick and we leave the rumbling pavement behind and transition into the rich sea-level air.  After only a slight pause the altimeter begins a determined run up the scale.

“Positive rate.”

“Gear up.”

The usual thumps and whirrs.  Some quick calls and frequency changes and we are soon out over the open ocean west of Vancouver then turning majestically through south towards south-east towards an inspiring view of Mount Baker.

“Goose Air 112, direct to Princeton, maintain flight level 230.”  The air is smoother now than when we arrived.  It’s often surprising how quickly areas of turbulence can well up, then subside.  But up ahead as we near the mountain range again, the clouds convince me to leave the seat belts ‘on’ for a while yet.  I ask FO Paula to make an announcement to keep our new flight attendant crew and passengers in the know.

Above ten thousand feet I put Otto back to work and move my chair closer to the windows for a view of the lower mainland and Fraser valley.  This time of the year the green-drenched ground is a welcome relief from the December blahs dominant in most of the country.

We’re soon level at FL330 with the cruise checks all done.  The countdown to home is underway.  The datalink comes alive with an update from our flight dispatcher.  South-western Quebec along with parts of Ontario and New York state are rapidly deteriorating into a serious winter wonderland.  Visibility continues to drop. Snow continues to fall.

I glance at the estimated fuel overhead Montreal.  The first of many such glances over the next few hours.

Time: CYUL arrival minus 4H30

--------------------------

Max weights of interest:

- Maximum taxi weight............................................................75 900 kg (167 329 lb) 

– Maximum takeoff weight (brake release) ................75 500 kg (166 447 lb) 

– Maximum landing weight .....................................................64 500 kg (142 198 lb) 

– Maximum zero fuel weight ................................................60 500 kg (133 380 lb) 

High Altitude Airspeed Envelope

A friend recently asked me an intresting question, in light of the speculations surrounding the Air France flight 447 disaster:

What would be the stall speed versus the "normal" cruising speed on a plane like this?

At middling altitudes ... like 28,000 there's a ton of spread. Let's say 120 knots stall speed (no flaps) - cruise at 450 knots. But at high altitude cruise it gets a little more complicated because of the speed of sound and incipient shock waves.

As an aircraft climbs above approx. 28,000 feet, the maximum speed the wing can fly gets slower (shock waves start to form). At the same time the lowest speed it can fly at gets faster (thinner air, complicated as well by shock waves).

This phenomenon contributed to some crashes of the original 707 and DC8 jet transport airplanes as the full effects of turbulence and maneuvering on this tightening envelope weren't fully appreciated. At such high altitudes any increase in G-loading, causes these two speed extremes to come even closer together.

The point where they meet, as displayed on a performance chart, has been nicknamed "coffin corner" because of this history.

(Typical Buffet Boundary Chart)

If an A330 is cruising at his maximum altitude for his weight, there could be as little as 10 to 20 knots room to spare on either side of the "envelope". Which is why when we encounter "significant" turbulence or mountain waves, most pilots prefer a lower altitude.

Here is a typical presentation of the speed envelope on a modern "glass" panel. The airspeed tape on the left shows the allowance between high and low speed buffet - about plus/minus 15 knots in this example:

http://en.wikipedia.org/wiki/Coffin_corner_(aviation)

Unfriendly Skies

As the investigation into the tragic loss of Air France 447 gets under way thunderstorms are considered a possible cause.

Isolated thunderstorms can be handled fairly easily in the daytime.  Under those conditions aircraft actually operate near them on a routine basis:

I spent a summer flying hail suppression in the north end of Thunderstorm Alley at Penhold Alberta, Canada.

But crossing a towering line of tropical thunderstorms in the dark of night is on every pilot's list of least favorite pass-times.  Even with the best radar equipment available, encounters with severe turbulence and worst-of-all, hail, are always a possibility.

Were storms a factor in 447's sudden demise?  It's way too early to tell.  Until the flight data recorders are recovered, we're only guessing.  In the meantime investigators will be keeping thunderstorms near the top of the list of suspects.