National Transportation Safety Board
Washington, DC 20594

FOR IMMEDIATE RELEASE: February 26, 2004
SB-04-03

************************************************************
LOSS OF PITCH CONTROL CAUSED FATAL AIRLINER CRASH
IN CHARLOTTE, NORTH CAROLINA LAST YEAR

************************************************************

Washington, DC - The National Transportation Safety Board
determined today that the probable cause of an airliner
crash in Charlotte, North Carolina, last year was the
airplane's loss of pitch control during takeoff. The loss
of pitch control was the result of incorrect rigging of the
elevator control system compounded by the airplane's center
of gravity, which was substantially aft of the certified aft
limit.

"This accident shows how important it is for everyone
involved in the safety chain to do their jobs properly, "
said NTSB Chairman Ellen Engleman-Conners. "It is
imperative that the recommendations we've issued today be
implemented so that tragedies like this not be repeated."

On January 8, 2003, Air Midwest (doing business as US
Airways Express) flight 5481, a Raytheon (Beechcraft) 1900D,
N233YV, crashed shortly after takeoff from runway 18R at
Charlotte-Douglas International Airport. Two crewmembers
and 19 passengers aboard the airplane were killed. One
person on the ground received minor injuries, and the
airplane was destroyed by impact forces and a postcrash
fire.

Contributing to the cause of the accident, the Board
found, were Air Midwest's and the Federal Aviation
Administration's (FAA) lack of oversight of the work being
performed at Air Midwest's maintenance facility in
Huntington, West Virginia. Board investigators found that
the accident airplane entered a maintenance check with an
elevator control system that was rigged to achieve full
elevator travel in the downward direction. However, the
airplane's elevator control system was incorrectly rigged
during maintenance, and the incorrect rigging restricted the
airplane's downward elevator travel to about one-half of the
travel specified by the airplane manufacturer.

Air Midwest contracted with Raytheon Aerospace to
provide quality assurance inspectors, among other
maintenance personnel, for the Huntington maintenance
station. Raytheon Aerospace contracted with Structural
Modification and Repair Technicians to supply the mechanic
workforce. One of these mechanics examined and incorrectly
adjusted the elevator control system on the accident
airplane. The Board stated that the failure of the Raytheon
Aerospace quality assurance inspector to detect the
mechanic's incorrect rigging of the elevator control system
also contributed to the cause of the accident.

The Board found that the FAA's failure to aggressively
pursue the serious deficiencies in Air Midwest's maintenance
training program that were previously and consistently
identified permitted the practices at the Huntington
maintenance facility during the accident airplane's
maintenance check. For example, the Raytheon Aerospace
quality assurance inspector did not provide adequate on-the-
job training and supervision to the Structural Modifications
and Repair Technicians mechanic who performed the
maintenance on the accident airplane's elevator control
system. Furthermore, the quality assurance inspector and
the mechanic did not diligently follow the elevator control
system rigging procedure as written. As a result, they did
not perform a critical step that would have likely detected
the misrig and would thus have prevented the accident, the
Board concluded.

The Board also found that Air Midwest's weight and
balance program contributed to the cause of the accident.
At the time of the accident, the program resulted in
substantially inaccurate weight and balance calculations for
company airplanes. The Board stated that, although Air
Midwest revised its weight and balance program after the
accident, the program is unacceptable because it may still
result in an inaccurate calculation of an airplane's center
of gravity position.

The FAA's weight and balance program guidance at the
time of the accident also contributed to the accident, the
Board noted, because the assumptions in the guidance were
incorrect. If the FAA had performed a survey to determine
average passenger and baggage weights at the time, the FAA
would have realized that these weights were significantly
different from the average weights in its weight and balance
program guidance and in Air Midwest's weight and balance
program. The Board, therefore, concluded that periodic
sampling of passenger and baggage weights would determine
whether air carrier average weight programs were accurately
representing passenger and baggage loads.

The Board's final report includes 21 safety
recommendations directed to the FAA. Among the issues
discussed in the recommendations are improved surveillance
of air carrier maintenance programs, improved maintenance
work card and manual instructions, effective weight and
balance procedures, and air carrier accountability for all
contract maintenance work performed.

Chairman Engleman-Conners noted that 47 NTSB staff
members were assigned to all or part of the Safety Board's
investigation of this crash. "We devoted more than 16,000
hours of staff time to this investigation in order to
complete it in just over a year. I think this was a model
for what I hope will be shorter major investigations in the
future when circumstances permit."

A synopsis of the accident investigation report,
including the findings, probable cause, and safety
recommendations, can be found on the Publications page of
the Board's web site, http://www.ntsb.gov. The complete
report will be available in about 6 weeks.
-30-

NTSB Media Contact: Keith Holloway, 202-314-6100
hollowk@ntsb.gov

'Sloppy' Mistakes Caused Air Midwest Crash
February 26, 2004
Sloppy maintenance and inadequate procedures for calculating and distributing weight led to the crash of an Air Midwest plane in North Carolina that killed 21 people last year, investigators said on Thursday.

In a sharply critical report, the National Transportation Safety Board found that missteps by a contractor were made worse by Air Midwest's use of inadequate weight and balance standards, which left the Beech 1900D turboprop overweight and tail heavy.

"There were a lot of mistakes made here," said Mark Rosenker, the vice chairman of the safety board. "I'll call it sloppy."

John Goglia, a senior board member and former mechanic, said the worst maintenance-related case he had investigated was the 1996 ValuJet crash in Florida. "But this accident, with the sheer number of people who failed to do their job, set a new low," he said.

Brian Gillman, a spokesman for Air Midwest parent Mesa Air Group, said the company was reviewing the safety board's findings and would not comment further.

Doomed Flight 5481 to Greenville, South Carolina, operated for US Airways as a commuter flight by Air Midwest, climbed after takeoff from Charlotte Airport at a dangerously steep angle, rolled left and crashed next to a maintenance hanger. All 19 passengers and two crew members died.

Investigators quickly concluded that cables controlling the plane's elevator, which regulates aircraft pitch, were improperly adjusted.

Records show that mechanics for Raytheon Aerospace, a contractor for Air Midwest, had performed the work in West Virginia two days before the accident. The mechanic who adjusted the cables had not done that work previously.

Investigators said the mechanic had little guidance and steps were skipped, including one that would probably have caught the misaligned cables and prevented the crash.

Investigators also found the work was not adequately supervised. They raised questions about record keeping, Air Midwest maintenance training and oversight, and the clarity of instructions in the aircraft's maintenance manual, which was written by manufacturer Raytheon Aircraft, a unit of Raytheon Co.

Raytheon Aerospace, which changed its name last year to Vertex Aerospace, is not related to Raytheon Co. It no longer does work for Air Midwest.

Compounding the mechanical problem, investigators said, were inadequate standards at Air Midwest for assessing the weight and balance of passengers and luggage. This led to "substantially inaccurate" calculations.

The Beech 1900 had flown several times after the maintenance but not with the type of load it was carrying on its final flight. The plane was 600 pounds (272 kg) overweight and tail heavy. But investigators said the plane, with a fully functioning elevator, should have been able to fly.

http://news.airwise.com/stories/2004/02/1077828553.html

AIR CRASH RESCUE NEWS:

February 26, 2004 - Balance, Maintenance Blamed in N.C. Crash

WASHINGTON (USA) - A maintenance error combined with too much weight in the back of the plane led to the crash of US Airways Express Flight 5481 last year at North Carolina's Charlotte-Douglas Airport, federal investigators said Thursday.

All 21 people aboard were killed in the crash, the deadliest in the United States in nearly 2 1/2 years.

The National Transportation Safety Board recommended the Federal Aviation Administration order changes in maintenance procedures and the way airlines determine weight and load distribution.

The twin-engine Beech 1900 commuter plane operated by Air Midwest took off normally on Jan. 8, 2003. Within seconds, however, its nose pitched up sharply. The aircraft stalled, then rolled left and plummeted into a maintenance hangar. The plane was headed for Greer, S.C.

Lorenda Ward, the investigator in charge of the National Transportation Safety Board probe, said improperly rigged cables that controlled the aircraft's up-and-down motion combined with improper weight distribution led to the crash.

``The simultaneous existence of these two errors resulted in a virtually uncontrollable airplane,'' Ward said in a report presented to the NTSB, which voted to accept the findings.

The pilots estimated the plane was within 100 pounds of its limit when it took off. The cockpit voice recorder transcripts show Capt. Katie Leslie and co-pilot Jonathan Gibbs discussed the issue on the runway. Their conversation turned frantic as the plane took off.

``Help me,'' Leslie said. Gibbs swore.

``Oh, my God,'' said Leslie, who then radioed the control tower, ``We have an emergency.''

A child yelled, ``Daddy!''

Warning horns sounded. ``Pull the power back,'' Leslie said. ``Oh my God.''

Investigators said the plane's tail was too heavy because of the way the passengers and bags were distributed. Too much weight can change a small plane's center of gravity and make it much more difficult to fly. The pilots could not compensate because the cables did not have their full range of motion due to the maintenance errors.

Some airlines weigh passengers and bags to determine weight and load distribution. The NTSB suggested that all airlines operating planes with 10 or more seats at least periodically put passengers on scales to make sure their weight estimates are correct. It also said those estimates should take into account things that would change weight, such as heavier clothing and coats in winter.

NTSB investigators said the Charlotte flight could have flown with improper weight distribution had the elevator cables been properly rigged. Instead, the cables only had half their normal maneuverability because the mechanic had failed to follow numerous steps in the maintenance process.

``There were a lot of mistakes made here,'' NTSB member Mark Rosenker said. ``I'll call it sloppy.''

The NTSB recommended the FAA require a series of improvements to training, oversight and procedures for maintenance personnel. Among them: requiring that work on key flight control systems, including elevator cables, be checked upon completion.

FAA spokesman Les Dorr said it is up to the airline to make sure maintenance work is carried out properly. FAA inspectors visit sites based on where they can best allocate resources to mitigate risk, he said.

Air Midwest contracted maintenance to Raytheon Aerospace (now known as Vertex Aerospace), which hired mechanics from Structural Modification and Repair Technicians Inc. Air Midwest's parent company, Mesa Air Group, announced this week that it will no longer contract out its maintenance. Also, the maintenance manual for the Beech 1900 has been revised to clarify rigging procedures.

From Air Safety Week:

Poor Maintenance Cited as Primary Cause of Air Midwest Crash

Inaccurate Weight and Balance Program Contributed to Being Out of Aft CG Limit
In the chain of events leading to the fatal Jan. 8, 2003, crash of an Air Midwest twin-turboprop, the pilots' inability to prevent the nose from pitching up excessively into a stall was the end result of improper maintenance performed on the airplane's elevator two days before.

However, even with the mis-rigged elevator restricted from its normal 14� nose-down movement to just 8�, the pilots might have been able to recover the situation had not the payload been distributed toward the tail, such that the airplane was a good 5 percent out of its aft center of gravity (CG) limit.

The airplane was some 600 pounds overweight for takeoff, but National Transportation Safety Board (NTSB) investigators determined that being out of the aft CG limit was a more critical factor. If the airplane had simply been overloaded, the pronounced pitch-up after takeoff could have been countered by 7� nose-down elevator, just within the 8� range-of-motion available. However, with the out-of-CG condition, the crew would have needed 9.5� of nose-down elevator, which they did not have.

Investigators were dismayed to discover that even with the heavier average passenger and baggage weights endorsed by the Federal Aviation Administration (FAA) after the accident, it is possible for airplanes like the accident Beech 1900D to be loaded outside of CG limits (see ASW, May 26, 2003). The use of average weights was declared inadequate, given that many regional operators use actual passenger and baggage weights to calculate a more accurate takeoff weight and aircraft CG.

Neither the aft CG configuration nor the limited pitch control alone was enough to cause the crash, but the combination proved deadly.

Poor maintenance came under intense scrutiny at the NTSB's final hearing Feb. 26. If the maintenance manual, since revised extensively, had been available for the elevator control cable rigging work done on the airplane during maintenance two days before, the cables more likely would have been rigged properly. For want of a good manual, the work was done wrong. For want of properly rigged elevator cables, pitch control authority was restricted. For want of 1� of elevator motion, the airplane and all 21 passengers and crew aboard were lost.

In the nine months since the NTSB's May 2003 fact-finding hearings on the crash, two major developments are worth mentioning. First, Air Midwest has decided to stop outsourcing its routine maintenance. "After an accident like that, you reassess," said Jonathan Ornstein, chief executive of Air Midwest's parent company, Mesa Air Group. Bringing the maintenance back in-house is a cost-effective way to facilitate more direct control of the work, Ornstein said last week. At the time of the accident, the carrier's maintenance work was outsourced to Raytheon Aerospace, whose people performed the work at a facility in Huntington, W. Va. (Raytheon Aerospace, since renamed Vertex, is not related to aircraft manufacturer Raytheon Aircraft Corp., manufacturer of the Beech 1900D). On the night of Jan. 6, 2003, the elevator control system underwent a cable tension check. The NTSB found that the five contract mechanics who performed the work at Huntington on the accident airplane had virtually "zero experience" on the Beech 1900.

The Air Midwest crash, and the company's decision to do more of the maintenance work "in-house," challenges the industry position that there is no link between contract airline maintenance and maintenance-related accidents.

The second major development since those NTSB hearings last year involves a major rewrite of Chapter 27 of the Beech 1900 maintenance manual. Chapter 27 deals with elevator control cable rigging. During the NTSB hearings last year, there was considerable discussion (and frustration) over the perceived difference between cable "tensioning" and cable "rigging." The technicians working on the airplane were under the impression that if only "tensioning" was involved, they could skip steps in the procedure. The incorrect, confusing and imprecise wording of the manual aggravated this interpretive conundrum. Shortly after the accident, Air Midwest advised the NTSB of its many problems with the manual used by the technicians on the night the work was done on the accident aircraft. The problems were not corrected in Air Midwest's own maintenance manual, an omission that was privately criticized by one NTSB member following the Feb. 26 hearing.

Nonetheless, Air Midwest's Feb. 12, 2003, letter highlighted many problems with the manual. Extracts of that letter:

"A review of the .... Maintenance Manual revealed a lack of explicit instructions which, if included, could have avoided the elevator mis-rigging."
"Step] (a) [in the manual] Removal of the autopilot servo cable. [The accident aircraft] did not have [an] autopilot so this could not be accomplished."
"[Step] (d) Requires the removal of passenger compartment seating, carpet and floorboards to gain access to the elevator turnbuckles. These turnbuckles are located in the tail of the aircraft, making this procedure not required and inaccurate."
"[Step] (f) Requires the adjustment of the push-pull tube between the control column and the forward elevator bellcrank.... No procedure is given to perform this adjustment."
"[Step] (p) Tighten the elevator down cable until the elevator rises to 0� ... No procedure is given to accomplish tightening of the turnbuckle."
"We would also like to point out that the ... detail inspection only requires the cable tensions to be inspected with a reference to Chapter 27. There is nothing in Chapter 27 that instructs the mechanic on the procedure for checking the cable tensions. And there is nothing in the inspection program or in the rigging procedures that require the mechanic to re-rig the elevator system if the cable tensions are low."
Since that letter, Raytheon Aerospace has rewritten Chapter 27 of the Beech 1900D maintenance manual. A copy of the latest version, obtained by this publication, shows a dramatic improvement in clarity, ease of understanding, and logical flow of the procedures.

Matthew Thurber, editor of sister publication Aviation Maintenance magazine, described the rewrite as "a remarkable change for the better." He applauded the evident effort necessary to undertake virtually a complete rewrite.