June 25, 1999 

     Pilot vs. NTSB 

     Retired TWA Pilot from
   Southold Interprets Flight 800
         Data Differently 

  It was almost exactly a year ago that
  Howard Mann of Southold, retired
  TWA pilot, outlined in this paper his
  analysis of the Flight Data Recorder
  (FDR), one of the "black boxes," from
  TWA Flight 800. Mann began his career
  with the airline as a mechanic, then
  became flight engineer and, finally, pilot.
  The gist of Mann's analysis concerned a
  line of data that had come to be referred
  to as the "12 second line"; the reading
  that had been taken at 20:31:12 (or 12
  seconds after 8:31 p.m.) on the evening
  of July 17, 1996, when Flight 800, on its
  way to Paris from JFK, came down out
  of the sky, killing all 230 people aboard.

  Further research by Mann, and an as yet
  unreleased addendum by National
  Transportation Safety Board (NTSB) on
  the Flight Data Recorder have brought
  further questions to light. 

  In the December, 1997 hearings on
  Flight 800, the NTSB had presented this
  12 second line of data, but with a line
  through it. Mann said, ³I read a printout
  of those last seconds and used a
  magnifying glass to read the figures that
  had that line through them.² It was
  Mann, among the various private
  individuals involved in investigations of
  the tragedy of Flight 800, who said the
  12 second readings indicated the sudden
  transformation of an airplane from
  normal flight to severe distress. The
  readings, according to Mann, showed
  that an external pressure wave had
  struck the plane. 

  For instance, at eleven seconds after
  20:31, the altimeter shows Flight 800
  was flying at 13,772 feet. In less than a
  second the plane¹s altitude is marked at
  10,127 feet. For an object to plummet
  3,645 feet through the atmosphere in
  one second it would have a speed of .68
  miles a second, or 40.8 miles a minute
  or 2,448 mph. In other words, it was
  not possible that Flight 800 fell at this

  But there is a way in which the altimeter,
  and other instruments on the plane, can
  give ³false readings.² The altimeter
  measures altitude by way of air pressure.
  There is less air pressure the higher you
  go, more as you descend. The altimeter
  of Flight 800 registers, in the last
  moments before the plane is blown
  apart, a pressure wave from outside the
  craft and thus registers it as an increase
  in actual air pressure and translates this
  into an apparent, abrupt, impossible
  drop in altitude. 

  This external pressure wave, Mann
  contends, was further indicated by,
  among other things: the role altitude; the
  angle at which the nose was raised; the
  angle of attack ­ which measures angle
  at which air is striking the wing); and the
  readings for the right rudder trim. 

  Mann and many other investigators said
  this was evidence of a pressure wave
  that could only indicate the plane had
  been struck by a missile, one which
  possibly blew up just outside the plane. 

  But there was an official explanation for
  these aberrant readings, said the NTSB.
  The FDR lays down readings on a spool
  that takes 25 hours of continuous
  information: 43 readings at a rate of 64 a
  second. The key word is continuous: the
  tape does not have a beginning or end to
  its data, but simply records over old
  data, data 25 hours old, as it goes along.

  But the transition from the data from one
  flight to another is not smooth, said the
  NTSB. What had become the infamous
  12 second line was in fact that transition
  point; and, in addition, the Safety Board
  said, those figures were actually for
  Flight 803, Paris to JFK on July 16,
  1996, the day before Flight 800 came

  Mann¹s comment at the time was that if
  the 12 second readings were for another
  plane, that was a plane in trouble. 

  Eventually the NTSB removed that 12
  second line of readings from its website
  report ­ presented at the December,
  1997 hearing ­ on Flight 800¹s Flight
  Data Recorder. 

  * * * 

  Coincidentally, on July 12, 1998, within
  a week and a half of our first interview
  with Howard Mann last year, an
  addendum report on the Flight Data
  Recorder was prepared by Dennis R.
  Grossi of the NTSB. The report was
  officially submitted to the NTSB on
  September 9, 1998. But as of yet the
  report has not been made available to
  the public. 

  Grossi, National Resources Specialist,
  Flight Data Recorders for the NTSB,
  also oversaw the original FDR report
  presented at the hearings. The impetus
  of the addendum is to clarify the issue
  about where the data for Flight 800 ends
  and the data for Flight 803 begins. The
  report reads: ³...a portion of the data
  recorded during TWA803 has
  mistakeningly been analyzed as if it were
  recorded during TWA800.² 

  The addendum goes on to say that it has
  been the normal practice for the NTSB
  ³to include the unsynchronized transition
  data² merely as a marking point to show
  the border or ³transition from the newest
  to the oldest data.² 

  A product called ³Magna See(r)² that
  contains an iron powder in a fast drying
  liquid is placed on the recording tape.
  The powder makes visible the magnetic
  fields on what has been recorded on the

  Apparently, according to the addendum,
  the iron powder was used specifically to
  pinpoint what is called ³the erase-write
  gap² on the tape. Data on the tape is
  erased just a bit ahead of new data
  being placed down on the tape. This
  results in ³a gap between the newest and
  oldest data that approximates the
  distance between the erase and write
  heads,² the report says. The gap is about
  three inches on the tape, or about seven

  In his analysis of the FDR and the
  addendum, Howard Mann says that the
  NTSB¹s Recovery Analysis and
  Presentation System (RAPS) ignores the
  erasure gap. RAPS, the digital system
  used to transcribe the data, makes no
  indication where the gap occurs. 

  ³The point is,² said Mann in a recent
  interview, ³they are saying the data was
  transcribed out of snyc, but the original
  data is still there ­ so why can¹t they just
  go back to it?² 

  ³The amplitude also decreases on either
  side of the erase write gap,² the report
  goes on to say, then adds, ³The
  degradation of the signal did not prevent
  RAPS from recovering the affected

  Mann points out that the place on the
  tape at which the NTSB concludes
  power on Flight 800 was lost to the
  FDR may not be the last moment at
  which power was available to record
  data. The NTSB¹s Sound Spectrum
  Study that analyzed the Cockpit Voice
  Recorder (CVR) on Flight 800 stated,
  ³At time .73 seconds before the end of
  the recording and again at .68 seconds
  before the end,² there was a ³change in
  the background signal as observed on
  the Captain¹s radio channel.² 

  Mann said that each of the 60 kilowatt
  generators on each of the four engines
  operate at the same rotation and speed
  and are linked together by what is called
  the Sync Bus. In his own analysis, Mann
  says, ³Under certain ŒFault Conditions¹
  individual Bus Tie Relays open to isolate
  individual generators and their load
  buses from the rest of the electrical
  system.² Thus one generator may fall out
  of sync, disrupting the even flow of
  power. Mann gave the analogy of
  soldiers purposely marching out of step
  when crossing a bridge in order to
  disrupt the Œharmonic¹ of their
  combined marching and put less stress
  on the structure of the bridge. 

  Citing the .73 and .68 ³loss of upper
  harmonics² on the CVR, Mann says,
  ³We can logically assume that the #4
  Bus Tie Relay has tripped due to some
  sort of fault.² He adds that the ³next
  automatic function would be for the #4
  Main Generator Relay to trip and the
  result is no power to #4 and the
  Essential Load Bus and the FDR.² 

  Mann calculates that the first power loss
  falls within .021 seconds of what the
  NTSB cited as the end of Flight 800
  data and power loss to the FDR:
  20:31:12.031. And then Mann goes on
  to make an interesting connection. 

  In the Grassley hearings last month in
  Washington D.C., regarding the FBI¹s
  conduct and procedure in the Flight 800
  investigation, one of the exhibits
  introduced was a March 28, 1997 letter
  from John C. Gannon, the CIA¹s
  Deputy Director for Intelligence to the
  FBI agent in charge of the Flight 800
  investigation, James Kallstrom. The
  letter was accompanied by a CIA
  analysis that read, in part, ³Using the
  eyewitnesses¹ visual and sound
  observation ­ combined with tracking
  data from the radars and infrared data
  from an intelligence sensor ­ CIA
  analysts were able to reconstruct the
  approximate path of Flight 800 from the
  instant its recordings ended until it hit the

  The agency¹s analysis arrives at the time
  of ³the initial explosion² on Flight 800 as
  occurring ³at 8:31:07.5 p.m.² In other
  words, 4.5-5 seconds before the FDR
  apparently lost power. 

  The 4.5-5 second lapse could be the
  result of a normal few seconds time
  before the power was affected, or, as
  Howard Mann postulates, the ³initial
  explosion² eyewitnesses and infrared
  intelligence from a surveillance satellite
  show is actually a missile launch. (CIA
  public affairs spokesperson Anya
  Guilsher confirmed that the ³infrared
  data² was indeed obtained via satellite.)
  If Flight 800 were at approximately
  13,800 feet at the time, Mann says, ³The
  difference from 7.5 seconds to 12.5
  seconds is about right for something
  climbing to 1370 feet at Mach 2.5
  (2750 feet per second.).² 

  Evidence that power returned to the
  FDR at least for another second comes
  from the transponder signal from Flight
  800 that was picked up by Mega Data
  of Bohemia at 20:31:13. An airplane¹s
  transponder records hits from radar.
  Mega Data receives transponder
  information from flights overhead and
  records them for the airlines. 

  Another pertinent point Mann brings out:
  the point at which the NTSB fixes as the
  beginning of data from Flight 803, the
  reading that records the date is absent,
  as well as the time of day. The date and
  time are recorded in sequence: a
  numeral, in this case ³7² marking the
  month of the July, which is then followed
  by the date then the time. 

  ³You have the FDR reading for the
  month but not the day. If we knew if this
  was the 16th, when Flight 803 was
  recording, or the 17th for 800, and or if
  we had the time, the whole problem
  would be solved.² 

  Mann further questions the NTSB¹s
  assigning the 12 second line to Flight
  803 and not Flight 800 by his analysis of
  the stabilizer trim parameter readings.
  Mann says the horizontal stabilizer
  moves one unit every seven seconds.
  The point on the tape at which the
  NTSB claims is the gap between Flight
  800 and Flight 803 shows what would
  be a normal rate of change for readings
  from one flight. A gap between readings
  assigned to two separate flights would
  not, except by extreme coincidence, be
  in agreement. But if the erasure gap
  occurred later on the twelve second line,
  where Mann contends it does, then the
  divergent readings that are given make
  more sense. 

  The altitude data ­ referred to at the
  beginning of this article ­ on the 12
  second line has another revealing
  dimension to it, Mann adds. There are
  two types of readings the FDR records:
  fine altitude and coarse altutude. The
  coarse alttude reading is in increments or
  brackets of 5,000 feet. In other words it
  shows if a plane is between 0-5,000
  feet, then 5,000-10,000, then
  10,000-15,000 and so forth. The fine
  altitude shows to the foot the height of
  the plane. The fine altitude reading on
  the 12 second line is 10,127. 

  Mann said, ³If the fine altitude is
  incorrect, it would still be in the bracket
  of 10,000-15,000. Yet the NTSB says
  that this is data from Flight 803, which,
  according to their own report, was at
  33,000 feet at that point on the tape.² 

  As of presstime, Dennis R. Grossi of the
  NTSB had not returned calls to discuss
  his addendum to the Flight Data
  Recorder report. 

  Ultimately, Mann once more contends,
  the controversy over Flight 800 data
  could be resolved if the readings that
  were supposedly transcribed out of sync
  can be lifted again, this time correctly. ³If
  the NTSB says it¹s there, why don¹t
  they just go back to it and clear up some
  of these questions that people have been
  wondering about for nearly three years?²

  ­Jerry Cimisi