N78008 – The Boeing Company 777-200 – 2005-03-02 NTSB Accident Report

HISTORY OF FLIGHT

On March 2, 2005, at 1550 eastern standard time, a Boeing 777-200, N78008, operated by Continental Airlines as flight 99, experienced a tail strike departing from the Newark Liberty International Airport (EWR), Newark, New Jersey. There were no injuries to the 2 certificated airline transport flight crewmembers, 2 relief flight crewmembers, 12 flight attendants, or 198 passengers. Visual meteorological conditions prevailed, and an instrument flight rules (IFR) flight plan had been filed for the flight destined for Hong Kong, China. The flight was a scheduled international passenger flight, conducted under 14 CFR Part 121.

According to a representative of the operator, the airplane was departing from runway 4L, an 11,000 foot-long by 150 foot-wide, asphalt runway. As the airplane rotated, during the takeoff, it lifted off the runway, momentarily settled back to the ground, and lifted off a second time. Suspecting that the tail of the airplane struck the ground, the pilot elected to return to the airport, where it landed uneventfully approximately 1-hour later.

According to the captain, who was flying at the time of the accident, the fuel load was approximately 251,000 pounds, which resulted in a gross weight takeoff of approximately 623,500 pounds. The takeoff was planned and briefed with the air conditioning packs "OFF" and with a "full power" thrust setting. The reported winds for takeoff were "very gusty" and variable. The ACCULOAD data for takeoff was planned with a 7-knot tailwind and a "low altimeter setting penalty" (ACCULOAD was a computer based weight and balance system, which was an integral part of the flight operations management system, and was used to generate the pilot weight manifest. Takeoff data was based on the most recent observed metrological and runway conditions).

During the takeoff roll, the indicated airspeed "lagged" slightly before V1 and Vr, but stabilized and accelerated normally after rotation. The rotation was normal at, or after Vr, but the airplane did not "fly away" as he expected. The airplane cleared the ground, but a wind gust forced it back down, and the tail of the airplane struck the ground with a "substantial impact." The "tail strike" EICAS (Engine Indication and Crew Alerting System) message then illuminated. The flight crew completed the takeoff sequence, completed all applicable checklists, jettisoned fuel, and returned to Newark for landing.

In a subsequent statement, the captain recalled that the flight received takeoff clearance from the tower, with the winds reported as 280 degrees at 17 knots.

According to the first officer, he was the "pilot monitoring" for the accident flight. Before takeoff, all normal checklists and briefings were completed in accordance with company flight operations procedures. Before taxiing onto the runway, he noted that the tower controller stated that wind gusts of plus or minus 10 knots were reported by another airplane. The takeoff was briefed with air conditioning packs "OFF," as planned per the ACCULOAD.

Engine spool up was normal and stabilized about 99 percent rpm. As expected, the indicated airspeed was erratic during acceleration due to the gusty wind conditions. Prior to the V1 speed callout, the indicated airspeed stagnated momentarily, and then "jumped" quickly up to the V1 speed. The jump in indicated airspeed was so quick, that the "automated voice" called V1 before the first officer did (synthetic voices annunciate certain normal but time critical operational information, such as the V1 callout). The first officer then announced, "V1, rotate," and recalled that the speed difference between V1 and Vr was one knot. The captain initiated the rotation, while the first officer turned his attention to the airspeed indicator to advise of any further airspeed anomalies. He then observed that the airspeed stopped increasing for a few seconds.

As the airplane "seemed" to pitch up normally, the first officer noticed that the airspeed stagnated, again, just above "V1/Vr." Therefore, his attention was fixated on the airspeed indicators. After several seconds, the airspeed began its normal acceleration toward V2. Once the first officer observed this acceleration, he looked up and noticed a high pitch angle. He then rationalized that with this high pitch angle, the airplane must be airborne. He looked back inside the cockpit to confirm a positive rate of climb and then felt a firm ground impact.

The airplane seemed to fly up and away from the ground as expected; however, shortly thereafter, the "tail strike" EICAS indication illuminated. The flight crew continued the climb out and subsequently returned to Newark, where the airplane landed uneventfully.

In a subsequent statement, the first officer recalled that the flight received takeoff clearance from the tower, which included winds from 280 degrees at 17 knots. The tower provided no further information, and the flightcrew elected to not add a gust factor to the takeoff data.

PERSONNEL INFORMATION

The captain held an airline transport pilot certificate with a rating for airplane multiengine land, and held a flight engineer certificate. The captain held type ratings in Boeing 727, 737, 747, 757, 767, 777, Airbus A-300, and McDonnell Douglas DC-9 and DC-10 airplanes. The operator reported that the captain had accumulated about 17,000 hours of total flight experience, which included 4,100 hours in the Boeing 777.

The captain's most recent Federal Aviation Administration (FAA) first class medical certificate was issued on October 16, 2004.

The first officer held an airline transport pilot certificate with a rating for airplane multiengine land. The first officer was also type rated in the Boeing 777. The operator reported that the first officer had accumulated about 15,000 hours of total flight experience, which included 3,438 hours in the Boeing 777.

The first officer's most recent FAA first class medical certificate was issued on March 15, 2004.

AIRCRAFT INFORMATION

The airplane was maintained under a continuous airworthiness inspection program, and had accumulated about 27,782 total hours of operation, at the time of the accident.

METEOROLOGICAL INFORMATION

The 1451 EWR METAR reported winds from 280 degrees, at 17 knots, gusting to 23 knots. Peak winds from 280 degrees at 26 knots were recorded at 1430.

The 1551 EWR METAR reported winds from 280 degrees, at 21 knots, gusting to 28 knots. Peak winds from 280 degrees at 30 knots were recorded at 1528.

AIR TRAFFIC CONTROL

According to the FAA Air Traffic Control Handbook, 7110.65L, Section 5. Runway Selection,

"3-5-1 Selection

a. Except where a 'runway use' program is in effect, use the runway most nearly aligned with the wind when the wind is 5 Knots or more or the "calm wind" runway when the wind is less than 5 Knots (set tetrahedrons accordingly), unless use of another runway:

1. Will be operationally advantageous, or,

2. Is requested by the pilot.

b. When conducting aircraft operations on other than the advertised active runway, state the runway in use.

NOTE -

1 - If a pilot prefers to use a runway different from that specified, he/she is expected to advise ATC.

2 - At airports where a 'runway use' program is established, ATC will assign runways deemed to have the least noise impact. If in the interest of safety a runway different from that specified is preferred, the pilot is expected to advise ATC accordingly. ATC will honor such requests and advise pilots when the requested runway is noise sensitive."

At the time of the accident, Newark Airport was utilizing an "Informal Runway Use Program."

According to the Newark Airport Informal Runway Use Program:

A. The procedures described in Attachments 1 and 2 shall be used for runway selection except when one or more of the conditions listed below exist. At those times, the runway(s) most nearly aligned with the wind will be assigned, unless use of another runway will be more operationally advantageous.

1. Wind shear reported by either pilot or the "TDWR" or "ITWS."

2. A thunderstorm in the initial takeoff departure path.

3. Snow, slush, ice, or standing water on the runway to be used.

4. Braking actions less than GOOD or reports of hydroplaning or unusually slippery surfaces.

5. If runways are dry, the crosswind component must not exceed 20 knots, and the tailwind component must not be greater than 5 knots.

6. If runways are wet, the crosswind component must not exceed 15 knots, and the tailwind component must not be greater than 3 knots.

B. Participation in the program is voluntary for aircraft operators/pilots. Whenever a request is made for other than the assigned runway, the pilot will be advised that the requested runway is not the noise abatement runway. If the assigned runway is still unacceptable, the pilot will be assigned the runway requested.

Note- Due to the interaction between surrounding air traffic facilities in the metropolitan area, the final authority for runway selection has been delegated to the Operations Manager on duty at the New York TRACON.

Review of Attachment 2 of the Newark Airport Informal Runway Use Program, revealed that the maximum allowable wind velocity, including gusts, for runway 4L, with winds from 280 degrees, was 14 knots.

FLIGHT RECORDERS

The digital flight data recorder (DFDR) was removed from the airplane and sent to the Safety Board's Vehicle Recorder Division in Washington, D.C. The recorder was subsequently downloaded, and contained approximately 61 hours of data. The accident flight was the last recorded event, and its duration was approximately 1 hour and 2 minutes.

Review of the DFDR data regarding flight control wheel position (CWP) revealed that at 15:49:55, as groundspeed increased through 8 knots, the CWP was 25.5 degrees counter clockwise (CCW). The captain's CWP decreased to 19.6 degrees CCW at 15:50:04 with a groundspeed of 43 knots. As groundspeed increased over the next 24 seconds, values for the captain's CWP were recorded between 20.2 degrees CCW and 37.5 degrees CCW. The highest captain's CWP recorded prior to pitch increasing was 38.4 degrees CCW at 15:50:28. At this time, the groundspeed was 150 knots and the left inboard and outboard ailerons were approximately 12 degrees trailing edge up.

Review of the DFDR data regarding flight control column position revealed that at the time the takeoff roll began, the force on the captain's control column was approximately -10 pounds (forward pressure on the column). At 15:50:30, at an airspeed of 156 knots, the control column position began to move aft, and the captain's force on the control column was 17.2 pounds (aft pressure on the column). Two seconds later, at an airspeed of 162.5 knots, the control column position reached 5.4 degrees aft, and the captain's force on the control column was 22.7 pounds (aft pressure on the column). Pitch increased to 3.0 degrees airplane nose up (ANU) and the nose wheel indicated no weight on wheels.

At 15:50:34, vertical acceleration reached 1.699 g's, which was the maximum value during the event. Pitch increased to 12.5 degrees ANU, and airspeed was 161.5 knots. The control column position was 6.5 degrees aft, and the captain's control column force was 25.7 lbs. One second later, the control column position reached its maximum value during the event, 9.2 degrees aft. The highest elevator value during the event was also recorded at this time, with the left inboard elevator at 22.3 degrees, trailing edge up (TEU).

At 15:50:36, both left and right main gear indicated no weight on wheels, and pitch reached 17.8 degrees ANU.

Data was also extracted from the recorder for the airplane's previous takeoff for comparison reasons. During the previous takeoff, with an airspeed of 145.0 knots, pitch began to increase to 0.9 degrees ANU and the nose wheel indicated no weight on wheels. Control column position was 5 degrees aft, the captain's control column force was 19.1 lbs, and the highest elevator value during the takeoff was recorded, with the left outboard elevator at 8.2 degrees TEU. At an airspeed of 166.5 knots and a pitch of 6.0 degrees ANU, the left main gear indicated no weight on wheels. The following second the right main gear indicated no weight on wheels.

WRECKAGE AND IMPACT INFORMATION

A Federal Aviation Administration (FAA) inspector and representatives of Boeing and the operator examined the airplane after the accident. Scrape marks were observed on the underside of the airplane's tail section. Further examination revealed substantial damage to the aft pressure bulkhead.

TEST AND RESEARCH

A Performance Study was conducted by the Safety Board's Research and Engineering Division. The Performance Study utilized the data provided to the flight crew for the accident flight in determining calculations.

According to the Performance Study, the airplane's takeoff roll took place in a quartering tail wind, which was recorded by the DFDR, from 280 degrees at 21 knots, with gust to 28 knots. This required the input of 24 to 37 degrees of control wheel input into the crosswind during the rotation. The airplane's rotation began approximately at the planned 158 knot rotation speed, but a gust that produced an approximate 10 knot tailwind component occurred during the rotation. The change in tailwind component represented approximately an 11 percent loss in lift and a reduction in tail clearance.

The Performance Study also stated that, Boeing had developed a method, largely based on flight test, to estimate the effects of airspeed, pitch rate, and lift coefficient (through control wheel input) on aft body clearances during takeoff. The estimation method was based on flight test data, basic aerodynamics, and analytical methods.

The recommended takeoff rotation speeds and pitch rate were set to provide safety margin for stall and/or aft body clearance, which ever was more critical. Anytime a takeoff rotation deviated from the nominal liftoff airspeed, pitch rate, or lift coefficient (by crosswind lateral control input raising the flight spoilers) the aft body clearance would be impacted.

Comparing the airspeed and groundspeed recorded for the accident, it was revealed that a wind gust caused an approximate 7 knot loss in airspeed, the control wheel input during the rotation ranged from 24 to 37 degrees, and the maximum pitch rate was approximately 6.0 degrees/second. Using these numbers in the Boeing estimation method, it was determined that the total tail clearance reduction was approximately 150-170 percent.

It was concluded in the performance study that reducing the pitch rate to the Boeing recommended 2.5 degrees/second would have brought the tail clearance reduction to the 80 percent to 100 percent of nominal level, and from that point, a 1-knot increase in rotation speed would have provided about a 7 percent increase to the tail clearance.

According to Boeing, the typical liftoff attitude of the 777-200 was 8.5 degrees, the minimum tail clearance was 37 inches, and the tail strike pitch attitude was 12.1 degrees. Boeing also reported that turning the control wheel resulted in the upward deflection of the respective wings outboard aileron, flaperon, and 6 of 7 spoilers.

ADDITIONAL INFORMATION

According to the ACCULOAD manifest used for the takeoff, the V1, VR, and V2 speed were calculated as, 157, 158, and 164 knots respectfully.

After the airplane landed, all of the baggage on board the airplane was counted and weighed. It was discovered that an additional 2,992 pounds of baggage was loaded on board the airplane than what was originally calculated for use by the flight crew. This resulted in a gross takeoff weight of approximately 626,523 pounds for takeoff. The maximum gross weight for takeoff was 629,200 pounds.

Takeoff Training and Procedures

The operator provided pilots with company Boeing 777 flight manuals for training and operational use. The information provided in the flight manuals was derived from numerous sources, to include the Boeing 777 Flight Crew Training Manual (FCTM).

The Boeing 777 FCTM was an advisory manual, which provided information and recommendations on maneuvers and techniques. A revision to the Boeing 777 FCTM, dated and effective October 31, 2004, was issued to all Boeing 777 operators. Included in the revision was additional information and flight training guidance concerning rotation techniques and gusty wind and strong crosswind conditions.

The operator received the Boeing 777 FCTM change during December 2004, and scheduled the revision changes to be added to the company flight manuals on May 1, 2005, as revision 14.

According to the October 31, 2004, revision of the Boeing FCTM, Takeoff and Initial Climb section,

"Takeoff speeds are established based on minimum control speed, stall speed, and tail clearance margins. When a smooth continuous rotation is initiated at VR, tail clearance margin is assured because computed takeoff speeds depicted in the QRH, airport analysis, or FMC, are adjusted to provide adequate tail clearance. Above 80 knots, relax the forward control column pressure to the neutral position. For optimum takeoff and initial climb performance, initiate a smooth continuous rotation at VR toward 15 degrees of pitch attitude. The use of stabilizer trim during rotation is not recommended. After liftoff use the flight director as the primary pitch reference cross checking indicated airspeed and other flight instruments.

Note: Do not adjust takeoff speeds or rotation rates to compensate for increased body length.

With a consistent rotation technique, where the pilot uses approximately equal control forces and similar visual cues, the resultant rotation rate differs slightly depending upon airplane body length.

Using the technique above, liftoff attitude is achieved in approximately 4 seconds. Resultant rotation rates vary from 2 to 2.5 degrees/second with rates being lowest on longer airplanes...

Takeoff and initial climb performance depend on rotating at the correct airspeed and proper rate to the rotation target attitude. Early or rapid rotation may cause a tail strike. Late, slow, or under-rotation increases takeoff ground roll. Any improper rotation decreases initial climb flight path. An improper rotation can have an effect on the command speed after liftoff. If the rotation is delayed beyond V2 + 15, the speed commanded by the flight director is rotation speed up to a maximum of V2 + 25. An earlier liftoff does not affect the commanded initial climb speed, however, either case degrades overall takeoff performance."

Crosswind and Tailwind Training and Operations

According to the October 31, 2004 revision of the Boeing FCTM,

"Crosswind guidelines are not considered limitations. Crosswind guidelines are provided to assist operators in establishing their own crosswind policies."

According to the operator's Boeing 777 flight manual, dated November 1, 2004, revision 13, which was the revision provided to the flight crew, and did not contain the October 31, 2004 Boeing 777 FCTM changes, it stated,

"CROSSWIND TAKEOFF

The crosswind takeoff characteristics are typical of most swept-wing transports. The upwind wing will tend to rise as the takeoff roll begins. This may be corrected by using aileron as required or by pre-setting a fixed amount of aileron into the wind prior to takeoff roll. Maintain a slight forward pressure on the control yoke until approaching rotation speed. In either case, large control wheel oscillations and inputs should be avoided.

Another indication of a crosswind condition is the tendency of the aircraft to weather vane into the wind, requiring rudder application for directional control. As speed increases, the aileron deflection requirement will decrease. Continue to maintain directional control with smooth rudder application. This will result in a cross control condition that must be maintained through liftoff. During rotation, hold the control wheel in a displaced position as required to keep the wings level. When airborne, aileron and rudder cross control should be slowly and smoothly relaxed.

As the airspeed approaches V1, the slight forward control column pressure is relaxed to neutral, allowing for a smooth rotation to begin at Vr. Takeoff and initial climb performance depend on rotating at the correct airspeed and proper rate, to the rotation target attitude. Early, rapid, or over-rotation may cause aft fuselage contact with the runway. Late, slow, or under-rotation increases takeoff ground roll. Any improper rotation decreases initial climb performance. Aft fuselage contact occurs at a pitch attitude of 12 degrees with wheels on the runway and landing gear struts extended.

For optimum takeoff and initial climb performance, initiate a smooth continuous rotation at Vr toward 15 degrees of pitch attitude. Rotate smoothly at an average pitch rate of 2 degrees/second. A 10 degree body attitude will be achieved in approximately 5 seconds.

Note: The flight director pitch command is not used for rotation.

An improper rotation can have an effect on the commanded speed after liftoff. If the rotation is delayed, the speed commanded by the flight director will be liftoff speed plus 15 knots limited to a maximum of V2 + 25. An early liftoff does not affect the commanded initial climb speed; however, either case will degrade overall takeoff performance.

When a positive rate of climb has been verified on the IVSI and altimeter, either pilot will call, "POSITIVE RATE." When a positive rate of climb is confirmed, the PF will call "GEAR UP," stabilize airspeed at V2 +15 to 25 knots, and transition to the F/D pitch command. Cross check indicated airspeed and other flight instruments. If the flight director is not used, indicated airspeed and attitude become the primary pitch references."

"...Tail strikes have resulted in major aircraft damage or unnecessary down time. Airplanes at light takeoff gross weights are most vulnerable to tail strike incidents. The rapid acceleration may catch pilots by surprise and unprepared to perform a smooth, steady rotation. Rapid rotation to an initial target climb speed will almost certainly result in a tail strike.

Avoid any tendency to rapidly rotate to a 10 degree pitch attitude and hold it until lift-off. This technique invites a tail strike.

Aft fuselage contact will occur at a pitch attitude of 12 degrees on the B777.

After the issuance of October 31, 2004, Boeing 777 FCTM revision 3, the following information was added to the operator's Boeing 777 flight manual, dated May 1, 2005, revision 14,

"For takeoff in gusty or strong crosswind conditions, consider using a higher thrust setting (up to maximum if appropriate) consistent with takeoff weight, weather conditions, runway length, and aircraft performance. When the prevailing wind is at or near 90 degrees to the runway, the possibility of wind shifts resulting in gusty tailwind components during rotation or liftoff increases. The use of a higher thrust setting reduces the required runway length and minimizes the airplane exposure to gusty conditions during rotation, liftoff, and initial climb. Avoid rotation during a gust. If a gust is experienced near Vr, as indicated by stagnant airspeed or rapid airspeed acceleration, momentarily delay rotation. This slight delay allows the airplane additional time to accelerate through the gust and the resulting additional airspeed improves the tail clearance margin. Do not rotate early or use a higher than normal rotation rate in an attempt to clear the ground and reduce the gust effect because this reduces tail clearance margins. Limit control wheel input to that required to keep the wings level. Use of excessive control wheel may cause spoilers to rise which has the effect of reducing tail clearance. All of these factors provide maximum energy to accelerate through gusts while maintaining tail clearance margins at liftoff. The airplane is in a sideslip with crossed controls at this point. A slow, smooth recovery from this sideslip is accomplished after liftoff by slowly neutralizing the control wheel and rudder pedals."

"Proper rotation will result in an airspeed of approximately V2 + 15 knots at average gross weights. Initially, the flight director will command 15 degrees nose up. If the vertical speed is acceptable, adjust the pitch attitude smoothly to maintain an airspeed of V2 + 15 to 25 knots. As the radio altitude and vertical speed increase, the F/D will command pitch to maintain V2 + 15 knots, allowing the pilot to transition to the F/D and maintain V2 + 15 until reaching initial flap retraction altitude."

Gust Factor

According to Boeing and the operator, there were no procedures or policies for adding a gust factor to the takeoff performance planning data for the Boeing 777.

Tail Strike Protection System

According to Boeing, an electronic tail strike protection system was installed on the 777-300ER and 777-200LR. The system was a function designed to reduce the probability of the tail of the airplane striking the runway on takeoff and landing, and provided a command to the elevator in order to prevent a tail strike only if a tail strike was imminent. The system was not installed on the 777-200 model airplane.