PCL ALPA Factual Statement 13Jun05

[Inconceivable]

Memo from MEC Chairman Wake Gordon
Re: Pinnacle flight 3701 accident
Date: June 13, 2005

The NTSB is holding a public hearing beginning today, June 13, regarding this
accident. They have listed as issues: 1 ¡V Aircraft and Engine Certification,
and 2 ¡V Operator and FAA Oversight of Flight Operations and Crew Training.

A Public Hearing is a quasi-legal NTSB proceeding which is intended to gather
additional factual information regarding the accident. This hearing takes place
at NTSB headquarters in Washington DC, and is expected to last three or possibly
four days. As its name implies, it takes place in the public eye, and can be
attended by the public, but the public cannot participate in the hearing itself.
However, ALPA, along with the other interested parties to the investigation
(FAA, Pinnacle Airlines, Bombardier and General Electric) will be key
participants in this hearing. In addition, the news media will certainly be
present and reporting on this event. Streaming video and audio of the hearing
is available on the NTSB website (www.ntsb.gov).

In the NTSB¡¦s own words ¡§The hearing is being held for the purpose of
supplementing the facts, conditions, and circumstances discovered during the
on-scene and continuing investigation. This process will assist the Safety Board
in determining the probable cause of the accident and in making any
recommendations to prevent similar accidents in the future. No determination of
cause [ALPA emphasis] will be rendered during these proceedings.¡¨ In addition,
the Public Hearing marks the point in the investigation process when the
existing body of factual information (known as the ¡¥docket¡¦) becomes
available to the general public.

I will caution you right now that you will become aware of certain facts which
will not reflect well on the accident pilots, on you and your fellow pilots, or
on the Company itself. Your first reaction may very well be to jump to the
defense of any or all of these entities. You may be frustrated by ALPA¡¦s
apparent lack of activity in this regard as well. However, like all NTSB
investigations, this investigation is a process with rules and certain
pre-defined steps, and I can assure you that ALPA is doing its best to ensure
that this is a thorough and balanced investigation. In a few months, ALPA will
be able to publicly discuss the issues and causes of this accident, but in
accordance with NTSB rules, that is not the case just yet.

Obviously, both the Pinnacle MEC and ALPA have a lot at stake in this public
hearing and it is therefore imperative that we maintain our integrity and
professionalism during this difficult period, and that we all abide by the
applicable rules and guidelines. Yes it may be quite frustrating, but in the end
we will all benefit by controlling ourselves a short while longer.

It is also important to present a unified appearance in support of our accident
investigation team, led by Captain Aaron Rose, MEC Central Air Safety Chairman.
The team has been meeting with the various investigative groups since the
accident occurred, while also flying. These members have shown great dedication
to determining and understanding the issues related to the accident.

It is essential that we support our team. Thank you Aaron, from the MEC

The following has been written by the Pinnacle Air Line Pilots Association¡¦s
Air Safety committee to inform the general pilot group about Pinnacle Flight
3701. The information contained in this document has been taken from the NTSB
public docket and consolidated from almost 1,200 pages of reports and interview
transcripts. The information contained herein is factual and contains no
analysis. Determination of probable cause and contributing factors into this
accident will be made by the NTSB at a later date. Our team will prepare a
formal submission to the NTSB that draws conclusions and makes recommendations
to improve safety in the airline and industry, based on what was found during
this investigation.

On October 14, 2004 Pinnacle flight 5668, aircraft N8396A, was scheduled to
operate from Little Rock, Arkansas to Minneapolis, Minnesota as a normal
passenger flight. During the takeoff roll the crew received a R 14th Bleed
overheat message. A low speed abort was conducted and the crew returned to the
gate after consulting with SOC. The passengers were rebooked on other flights
and contract maintenance was called. After a period of time contract
maintenance could not determine the cause of the problem so a Pinnacle
maintenance crew from Memphis drove to Little Rock. Over the course of the
afternoon the R 14th Bleed overheat sensing loop was found to have some chaffing
damage where it passed through a rib in the pylon. The loop was replaced in
accordance with the aircraft maintenance manual and the engine was run for
thirty minutes. Before the aircraft could be returned to service the original
crew deadheaded to their domicile due to duty time constraints.

Captain Jesse Rhodes was hired by Pinnacle Airlines on 2/24/03 as a first
officer. At the time of the accident he had 6900 hours total time, 5055 hours
as pilot in command, 150 hours as pilot in command CRJ and 823 hours second in
command CRJ. He held a first class medical certificate and an Airline Transport
Pilot certificate with type ratings in the CL-65 and BE-1900. His last PC was
conducted on 8/10/04 and was 31 years old.

FAA records show Jesse Rhodes received a notice of disapproval at several points
during his flight training, during his BA-4100 first officer initial and during
his ATP / BE-1900 type rating. All certificates were completed after additional
training. One additional sim session was required during CRJ upgrade training
due to checklist difficulties. There is no history of driver¡¦s license
suspension or revocation or FAA certificate action. Pilots and instructors
interviewed after the accident had favorable comments about his flying abilities
and he had received a letter of commendation for his handling of an emergency at
Gulfstream Airlines. He had not worked or engaged in unusual activity in the 72
hours prior to the accident and the accident flight was his first duty
assignment of the day.

First Officer Peter Richard Cesarz was hired by Pinnacle Airlines on 4/26/04 as
a first officer. At the time of the accident he had 761 hours total time with
222 as SIC in the CRJ. He held a first class medical certificate. His last PC
was conducted 6/27/04.

FAA records show Peter Cesarz received a notice of disapproval during his multi
commercial check ride and passed the next day. There is no history of driver¡¦s
license suspension or revocation or FAA certificate action. Pilots and
instructors interviewed after the accident had favorable comments about his
flying abilities and he had received several favorable letters of recommendation
from former instructors. He had flown a trip on October 11th and 12th, sat home
reserve on October 13th and the accident flight was his first duty assignment of
the day on October 14th.

Tissue specimens from the captain and first officer tested negative for ethanol
and a wide range of drugs, including major drugs of abuse.

Captain Jesse Rhodes and First Officer Peter Cesarz were deadheaded from Detroit
to Little Rock in order to reposition aircraft 8396 to Minneapolis. They
arrived in LIT at approximately 2000L. They blocked out of Little Rock at 2054L
as Flagship 3701. The flight was operated as a FAR 91 flight and dispatch had
issued release paperwork in the same manner done for FAR 121 flights.

The flight route for them was as follows:
FLG3701 CRJ2/F 0434 LIT P0200 330
LIT..COU..ALO.KASPR2.MSP/0145

Airplane Information: Weight and Balance

Operating Empty Weight (OEW) 31,436
Passenger Weight 0
Baggage/Cargo Weight (Manifest) 0
Fuel Weight 8,100
Taxi Fuel Burn 200
Actual Takeoff Weight 39,336
Estimated Fuel Burn to Accident Site 2,200
Takeoff Center of Gravity (CG) 21.6% MAC
Takeoff CG Limits 9.0 ¡V 35.0
Takeoff Stab Trim 6.0

Examination of the FDR data showed portions of the ascent where sharp pitch up
maneuvers occurred. Pitch- up maneuvers were identified at three separate times
between liftoff and the final level off altitude at 41,000 feet. The initial
rotation occurs at 02:21:45, up to a pitch angle of about 6 degrees. Four
seconds later, a larger column deflection to 8 degrees raised the pitch of the
aircraft to 22 degrees, and generated close to 1.8 g¡¦s of vertical
acceleration. A single stick pusher discrete is recorded on the FDR, followed by
a large column deflection towards nose down (-7„a, where ¡V11„a is the limit)
and reduction of pitch angle.

At approximately 02:27:17, while level at 15,000 feet, the autopilot is
disconnected and a pull up is initiated with a column deflection. Over 2.3 g¡¦s
of vertical acceleration is generated as the airplane pitches up to 17 degrees
over the next several seconds. At 02:27:26, a nose down push is initiated,
requiring a calculated 20-pound force in the nose down direction. During this
push, the vertical acceleration drops to close to 0.3 g¡¦s. This is followed by
another pull of about 26 pounds. Approximately 70 seconds later, a rudder
doublet is performed, consisting of a large rudder input to the left, then to
the right. This generates close to 0.16 g¡¦s of lateral acceleration on the
first deflection left, than 0.34 g¡¦s lateral acceleration to the right, than
0.18 g¡¦s back to the left. After these maneuvers, the autopilot is reconnected
at 02:29:27 as the airplane ascends through 22,300 feet.

 

[Inconceivable]

While at 25,000 feet, a climb is initiated with the autopilot engaged and
vertical speed selected at 600 feet/min. With the autopilot engaged, the column
is moved aft to 4 degrees airplane nose up by 02:32:40. This increases the pitch
to over 10 degrees, and generates 1.87 g¡¦s vertical acceleration, and increased
to the climb rate to over 5000 fpm for several seconds. At 02:32:46, the
autopilot is disconnected and the column is pulled with a calculated force of 34
pounds, increasing the pitch to close to 15 degrees. The column position remains
between 3 and 4 degrees over the next 8 seconds. This column force is released,
and another pulling force is gradually applied starting at 02:32:58, reaching a
calculated peak of 44 pounds as the column moves to 5.3 degrees (the max column
movement nose up is 13.8„a). The autopilot is then re-engaged at 02:33:10 with a
target climb rate of 3000 ft/min, which is reduced to 1400 ft/ min, and then
1000 ft/min. Shortly after this the crew requests and is issued a climb
clearance to 41,000.

The weather at 41,000 feet at the time of the accident was -47.1¢X or ISA +
9.4¢X. Winds were 315 at 48 knots and light to moderate turbulence had been
forecasted between 27,000 and 35,000 feet. Using Bombardier performance charts
an aircraft of 38,000 pounds and ISA +10, about 40,000 feet using 300 FPM climb
capabilities is the maximum altitude.

During the climb from 34,000 feet to 41,000 feet the autopilot engaged. While
below 35,000 feet, the airplane was climbing at a vertical speed of 1000 feet
per minute (fpm), as selected in the autopilot. The engines maintained an N1
setting of 95.5%, and Mach number varied slightly between 0.60 and 0.61. As the
airplane passed through 35,000 feet, the autopilot vertical speed selection was
reduced stepwise to zero, and the aircraft leveled off at 36,500 feet at
02:41:00. During this level off, the calculated Mach number began to increase.

Shortly after 02:42:00, a vertical speed of 500 fpm was selected on the
autopilot. The engines increased in N1 by about 0.5%, and Mach number continued
to increase. By 02:44:00, the aircraft was passing through 37,400 feet and the
Mach number reached 0.644. The airplane continued this climb at a vertical speed
of 500 fpm over the next 7 minutes, until reaching the final altitude of 41,000
feet. During these 7 minutes, the calculated Mach number showed a continuous
decay to 0.57. The pitch angle had increases from about 3 degrees to about 6
degrees, and N1 on both engines had reduced to 94.7%.

At 0252, the controller instructed FLG3701 to contact ZKC sector 30 on
frequency125.67. The pilot acknowledged and did so, checking in at 0252:08 level
at FL410. Once the airplane reached 41,000 feet, vertical speed reduced to zero
as the selected altitude was captured by the autopilot, and the airplane leveled
off. Over the next 3.5 minutes, calculated Mach number continued to decay to
about 0.53, pitch angle and angle of attack continued to increase to close to 7
degrees, and N1 in both engines continued to reduce by another 0.5%.

At 0254:31, the crew requested descent, saying, ¡§yeah just as you said that
looks like we¡¦re not even going to be able to stay up here uh look for maybe uh
three nine oh or three seven.¡¨ The controller instructed the crew to stand by
and initiated coordination for descent with sector 29. The sector 29 controller
asked, ¡§He don¡¦t like forty one?¡¨ At 0254:49, the R30 controller replied, ¡§I
don¡¦t think he had enough gas up there he is so slow,¡¨ and concluded the
coordination.

While the coordination call was in progress, the pilot of FLG3701 called the R30
controller. According to the controller¡¦s statement, the pilot¡¦s voice sounded
stressed and it sounded like he said ¡§emergency.¡¨ When the call was over, the
R30 controller stated, ¡§FLG3701 I was off frequency say again.¡¨ The pilot did
not respond. At 0255:03, the controller again called
FLG3701. Two seconds later, the pilot transmitted, ¡§emergency stand by.¡¨

At 02:54:34, the airspeed had decayed to 150 knots (calculated Mach = 0.526). A
single discrete for stick shaker on the right side switched to ¡§on¡¨, and the
autopilot disconnected. The control column moved to close to zero degrees by
02:54:35. 3.5 seconds after autopilot disconnect, a pull on the control column,
calculated to be 25 pounds, occurred which was then released. Over the next few
seconds, the control column moved again to 4 degrees in the airplane nose up
direction, with airplane nose up column forces calculated to be up to 26 pounds.
During this column movement, the pitch increased to close to 8.5 degrees.

By 02:54:45, the angle of attack had increased sufficiently to activate both
the stick shaker and stick pusher, and the column moved forward rapidly.
Bombardier Aerospace indicated that the stick pusher moves the column forward
with a force of 80 pounds. The action of the pusher reduced the pitch angle to
¡V3.5 degrees and the angle of attack to 0 degrees. The vertical acceleration
reduced to near zero g¡¦s. As the vertical acceleration was reducing, the angle
of attack reduced, the stick pusher cancelled, and the control column moved back
beyond the neutral point to about 5 degrees airplane nose up. The aircraft
pitched nose up to 8 degrees pitch angle and the angle of attack increased to 11
degrees, leading to another stick pusher activation.

Over the next 8 seconds, this same procedure is repeated two more times, with
the stick pusher moving the column forward, and after pusher release from
reduction of angle of attack, movement of the control column beyond the neutral
point to airplane nose up. During the pitch up in the second of the cycles at
02:54:51, the first drop in engine 2 N1 is noted, from a steady value of 94%
down to 84%. After 02:54:53, both engines N1¡¦s and fuel flows begin to reduce
in value.

These subsequent pitching cycles cause the airplane nose down pitch angle in
each pusher activation to increase, reaching ¡V19.5 degrees by 02:54:55. The
subsequent pull back on the control column went to a higher nose up value each
cycle, reaching 6.8 degrees on the third pull at 02:54:56. The nose up movement
also caused the vertical acceleration to reach lower values in each cycle,
reaching a maximum of ¡V0.8 g¡¦s. During these cycles the airspeed had increased
to 160 knots.

By 02:54:57, the stick pusher had activated a fourth time, pushing the column to
nearly full nose down, where it remained for the next 4.5 seconds. During this
push, the angle of attack reached the maximum measurable value of 27 degrees.
Engine fuel flow had reduced to near zero by 02:54:58. Pitch angle reached a
maximum value of 29 degrees at 02:54:59. The recorded airspeed was only 74 knots
at this time, and the altitude is increasing. However, according to Bombardier,
the static source errors corrections programmed into the aircraft Air Data
Computers (ADC) are not calibrated for the high angle of attack experienced at
this time (> 27 degrees) and the recorded values of airspeed and altitude may
not be accurate.

As the pitch angle began to reduce, a left rolling motion started, eventually
reaching 82 degrees left wing down and the pitch reached ¡V32 degrees at
02:55:06. Over the next 14 seconds during the recovery from this upset, there is
considerable control column, control wheel, and rudder movement. The stick
pusher activated three more times during this recovery. Engine fuel flow
remained near zero, and both engines N1 exhibited decays in value. The FDR
stopped recording data by about 02:55:20 when both engine driven AC generators
fell off line and the ADG deploys.

 

[Inconceivable]

The FDR resumes recording data at 02:59:10 when AC power is reestablished from
the APU. Both engine N2 and fuel flows are at zero and remain that way until
the end of the flight. The sensors for N2 cannot detect rotation less than
0.4%, about 60 RPM. Some turning actually was determined on further examination,
at the very end of the flight. One wind milling relight attempt is made at
approximately 21,000 feet with a maximum airspeed of 235 KIAS. Below 13,000
three attempts are made at APU engine starts. Passing 8500 feet AGL the crewed
declared to ATC a dual engine failure and requested vectors to the nearest
airport. The were given vectors for Jefferson, Missouri (KJEF) were the weather
was recorded at 0253Z wind 290 at 6, visibility 10 miles, overcast skies 4400,
temperature 10 dew point 5, altimeter 29.63. The FDR recording ceases at
03:15:05 with the left wing 42 degrees nose down. The last airspeed recorded
was 152 KIAS.

The first point of impact was determined to be the left wing hitting a large oak
tree approximately 450 feet away from the main impact point. When the left wing
hit the tree, the wing was severed near its midpoint and the aircraft rolled
left. The aircraft hit several more trees prior to hitting the ground inverted
behind a row of houses. The cockpit was destroyed as the aircraft further
struck rising terrain and a garage with a concrete base. The impact forces and
the ensuing fire destroyed the forward section of the aircraft leaving only the
tail section remaining approximately 150 feet from the cockpit section impact
point.

A CVR sound study was initiated shortly after the accident to examine several
anomalies with CVR sound recording and FDR recording of radio transmissions.
During the last eight minutes of flight a feedback squeal is heard from the left
side of the cockpit, the left CVR channel becomes inoperative and the FDR
records the left microphone keyed for the remainder of the flight. The CVR
sound study determined that this was due to the left audio panel being placed in
the emergency mode. Using ATC audio tapes, it was determined that when climbing
through 25,000 feet, the voice on the ATC transmission associated to each
cockpit microphone changed. When the right cockpit microphone was keyed the
captain¡¦s voice was broadcast to ATC and when the left cockpit microphone was
keyed it was the first officer¡¦s voice. This remained the case through the
upset and then on the descent, while passing 11,000 feet the sound of cockpit
seats being adjusted is heard and the first officers ATC transmissions are
associated with the right side cockpit microphone.

Post accident teardown of the engines revealed that the number two engine had
severe heat damage in its turbine section. FDR data shows that prior to spool
down, its ITT reached 1250 degrees C, the highest recordable value for the ITT
sensors. Both engines showed shiny areas within the grooves for the interstage
turbine seals. These marks are indicative of a condition known as ¡§core lock¡¨
within General Electric. Core Lock is understood to be caused by the contact of
the High Pressure Turbine (HPT) inter-stage seal (ISS) static honeycomb
component binding with the rotating seal teeth of the Outer Torque Coupling
resulting in a frictional sticking (stiction), that results in the rotor being
temporarily inhibited from rotating. No prior cases of core lock are reported to
have been stickier than the air turbine starter¡¦s capacity to overcome it.

This concludes the factual briefing prepared by your Air Safety Committee.


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

 

[Smacktard]

It's a sobering account that hits close to home because most of us in this forum fly the aircraft. Other than that, however, it's almost foreign to me what these guys were doing. It is so far out of the realm of what I would ever consider doing in the plane, that I have a hard time believing two people would be stupid enough to play like that.

The only thing I can't find reference to that I am curious about is if the FO made any comments about being uncomfortable with the way the captain was flying (I assume the Captain was flying since the FO was doing the radios.) I know he was agreeable to going up to 410, but other than that, were there any occurences of him trying to input some sensibility?

It should be required reading for any new hire. Actually, it should be required reading for anyone in this industry.


And no, I didn't fly the 1900 at Gulf$tream.

 

[BoilerUP]

Very sobering, indeed.

I hope universities with aviation programs incorporate this accident into their CRM/High Performance Aircraft courses.

 

[Ben Franklin]

Using ATC audio tapes, it was determined that when climbing through 25,000 feet, the voice on the ATC transmission associated to each cockpit microphone changed. When the right cockpit microphone was keyed the captain¡¦s voice was broadcast to ATC and when the left cockpit microphone was keyed it was the first officer¡¦s voice. This remained the case through the
upset and then on the descent, while passing 11,000 feet the sound of cockpit
seats being adjusted is heard and the first officers ATC transmissions are
associated with the right side cockpit microphone.

No opinion inserted by the author, but are they trying to say(without saying)that the F/O was in the left seat for a portion of the flight?


Sincerely,

B. Franklin

 

[climb2slow]

"No opinion inserted by the author, but are they trying to say(without saying)that the F/O was in the left seat for a portion of the flight?"

I got the same impression. Also sounds like alot of "yank and Bank" going on during the climb. Not terribly flattering information.

 

[Beechnut]

Wow.

Professionals.

S.

 

[Broke in CVG]

This follows under Darwins Law.... just glad no inocent parties were taken with them.

 

[chperplt]

Reread the portion I've quoted... YES.. THE FO WAS IN THE LEFT SEAT

Using ATC audio tapes, it was determined that when climbing through 25,000 feet, the voice on the ATC transmission associated to each cockpit microphone changed. When the right cockpit microphone was keyed the
captain¡¦s voice was broadcast to ATC and when the left cockpit microphone was keyed it was the first officer¡¦s voice. This remained the case through the upset and then on the descent, while passing 11,000 feet the sound of cockpit seats being adjusted is heard and the first officers ATC transmissions are associated with the right side cockpit microphone.