National Transportation Safety Board
Aviation Accident Final Report
State College, PA
Date & Time:
11/19/2005, 2218 EST
Flight Conducted Under:
Part 121: Air Carrier - Non-scheduled
During the approach, the captain had disconnected the autopilot, but had left the autothrottles engaged in what is known as "mixed-mode" method of flight control in which control of the flight path is mixed between manual and automatic means. During this time, the autothrottle began to track an airspeed which had been previously selected in the mode control panel speed window. The captain disconnected the autothrottles, but did not retard the throttles to idle. Therefore, automatic speedbrake deployment did not take place during the first touchdown. During the resultant bounce, the captain then retarded the thrust levers to idle, the speedbrakes deployed, there was a loss of lift and nose up pitching moment, and a tailstrike resulted. The airplane manufacturer recommended autothrottle use only when the autopilot was engaged. They also advised if the airplane bounced while landing, the pilot should hold or re-establish a normal landing attitude and add thrust as necessary to control the rate of descent, and that thrust need not be added for a "shallow skip or bounce," however, if a "high, hard bounce" occurred, the pilot should initiate a go-around. Information concerning the dangers of mixed-mode operations and bounced landing recovery was available to the operator and Federal Aviation Administration (FAA); however there were no prohibition on mixed-mode operations, nor were bounced landing recovery guidance required to be listed in any of the operator's FAA approved manuals. Two months prior to the accident, the Safety Board issued a recommendation to the FAA; to require air carriers to incorporate bounced landing recovery techniques in their flight manuals and to teach these techniques during training. As a result the FAA issued a Safety Alert for Operators (SAFO), which "recommended" each airline, include bounced landing recovery techniques in their manuals and training; however since the SAFO was not mandatory, the Safety Board requested the FAA survey air carriers to establish how many had adopted the SAFO. The Safety Board advised the FAA that until the results of the SAFO were known, the Safety Board's recommendation would remain classified; Open Acceptable Response.
The National Transportation Safety Board determines the probable cause(s) of this accident to be:
The pilot's improper touchdown and recovery from a bounced landing. Factors to the accident were the operator's failure to provide sufficient information on the use of autothrottles and bounced landing recovery techniques, along with the Federal Aviation Administration's failure to require the inclusion of mixed-mode flight control guidance and bounced landing recovery techniques in operator pilot training programs and flight manuals.
Occurrence #1: MISCELLANEOUS/OTHER
HISTORY OF FLIGHT
On November 19, 2005, at 2218 eastern standard time, a Boeing 737-800, N734MA, operated by Miami Air International as flight 622, was substantially damaged while landing at University Park Airport (UNV), State College, Pennsylvania. The 2 certificated pilots, 1 flight mechanic, 4 flight attendants, and 120 passengers were not injured. Night visual meteorological conditions prevailed, and an instrument flight rules (IFR) flight plan was filed for the flight that departed Capital City Airport (LAN), Lansing, Michigan. The nonscheduled passenger flight was conducted under 14 CFR 121.
According to the flight crew, the en route phase was uneventful. During a "VNAV approach backed up with ILS" to runway 24 at UNV, with the flaps set to 40 degrees, the captain noticed that after the autopilot was disconnected at approximately 1,000 feet above ground level (agl), the airplane appeared to be a "little low on the glideslope." Then after correcting, the airplane appeared to "go a little high" in reference to the precision approach path indicator, as it crossed the runway threshold. Approximately 30 feet agl, the captain began the landing flare and the airplane touched down "a little firm." The airplane then became airborne again, but the captain did not add power. The airplane "floated for a little bit," the spoilers deployed, and the captain tried to arrest the descent rate by adding "flare." The airplane then touched down for a second time.
During the postflight inspection it was determined that the tailskid and the lower portion of the aft fuselage had contacted the ground.
According to Federal Aviation Administration (FAA) records, the captain held an airline transport pilot certificate with multiple ratings including airplane multi-engine land, and a type rating for the Boeing 737. According to records provided by Miami Air International, the captain had a total flight time of 8,141 hours, with 4,954 hours in multi-engine airplanes and 1,258 hours in the Boeing 737. His most recent FAA first-class medical certificate was issued on October 28, 2005.
According to FAA records, the first officer held a commercial pilot certificate with ratings for airplane single engine land, airplane multi-engine land, and airplane instrument. According to records provided by Miami Air International he had a total flight time of 3,072 hours, with 2,281 hours in multi-engine airplanes. His most recent FAA first-class medical certificate was issued on June 16, 2005.
According to FAA and company maintenance records, the airplane was manufactured in 2000. The airplane's most recent continuous airworthiness inspection was completed on September 29, 2005, and at the time of the accident, it had accumulated 17,022 total hours of operation.
A weather observation taken at the airport about 2 minutes after the accident, included winds from 260 degrees at 4 knots, 10 miles visibility, clear skies, temperature 37 degrees Fahrenheit, dew point 21 degrees Fahrenheit, and an altimeter setting of 30.24 inches of mercury.
According to the Airport Facility Directory, UNV was located at an elevation of 1,239 feet msl. Runway 24 was 6,701-feet-long, 150-feet-wide, and constructed of grooved asphalt. In addition, runway 24 was equipped with an instrument landing system, precision instrument markings, a 4-light precision approach path indicator, and a medium intensity approach lighting system with runway alignment indicator lights.
Cockpit Voice Recorder Information
The 30-minute recording began after the crew had started the descent from cruise altitude for landing at UNV. After performing the descent approach checklist, the crew was assigned a visual approach by air traffic control. After descending through 5,000 feet, the crew received traffic and weather information from University Park Traffic (Common Traffic Advisory Frequency - CTAF).
Two minutes later, the first officer reported runway 24 in sight and canceled the IFR flight plan. The crew next reported right downwind on the traffic frequency, and received a wind and altimeter advisory. The captain requested first flaps 1 degree and then 2 minutes later, flaps 5 degrees. After the first officer reported right base for runway 24, the captain stated he had the field in sight. At a position 5 miles from the runway, the captain turned onto final approach and requested gear extension, 15 degrees flaps, and the final landing checklist. The first officer reported passing "PENUE" at 3,090 feet as the captain requested flaps 40 degrees. As the aircraft approached "minimums," the first officer told the captain that they were above the glideslope.
The airplane touched down on the runway and after the first officer uttered a comment, impacted the runway a second time. As the airplane was slowing, both pilots commented about the landing. The first officer then asked about "manual speed brake." The first officer reported clearing the runway and the captain then made a comment about "the worst landing of his life." After arriving at parking, the first officer read the securing checklist. The recording ended at 30:29 when power was removed from the CVR.
Flight Data Recorder Information
Data from the accident flight was successfully downloaded from the airplane's flight data recorder (FDR). Parameters from the approach and landing phases of the flight were reviewed.
On the accident flight, the FDR data indicated that the speedbrake lever was armed at 5,500 feet radio altitude (RA). The wind transitioned from a 40-knot tailwind, to a 6 to 8 knot headwind at approximately 3,000 feet RA. The autopilot was selected off at 575 feet RA, but the autothrottles remained engaged. At approximately 190 feet RA the auto throttle mode changed, to mode control panel (MCP) speed, and began to track the speed that was in the MCP airspeed window. This was recorded by the FDR as a selected speed of 143 knots. The airplane's airspeed degraded below the selected MCP speed, and the autothrottle commanded a thrust advance. This was followed shortly by a reduction in the throttle resolver angle (TRA) and the rate of descent increased to 900 fpm.
The autothrottles were selected off at 90 feet RA; however, the TRA remained at approximately 50 degrees, and the 6 to 8 knot headwind diminished to zero. The first touchdown occurred 6 seconds later, at a pitch attitude of 8 degrees nose up. The airplane was at 132 knots ground speed with the TRA still at approximately 50 degrees. The airplane became airborne approximately 0.5 seconds later, and the throttles moved to idle. Approximately, 1.4 seconds later, spoilers 3, 4, 9, and 10 all began to increase while groundspeed remained at 132 knots. Another 1.4 seconds later, the airplane touched down again, the spoiler ground speed brake, transitioned from "Not Up" to "Panels Up," and spoilers 3, 4, 9, and 10 fully extended.
During the second touchdown, pitch attitude reached 9.5 degrees nose up, and peak vertical acceleration reached 2.5g's.
External examination by personnel from the airplane manufacturer revealed that the tailskid cartridge had been compressed to its replacement limits and the shoe had been ground off. An aft drain mast had been ground away on its lower edge, and the airplane had suffered damage to multiple skin panels.
Internal examination revealed that multiple frames, doublers, webs, shear ties, stringer ties, clips, splice plates, intercostals, and attach angles had been damaged. Further examination also revealed, that the pressure bulkhead exhibited "oil canning" on the BS1016 web, located just aft between the Y-chord and the circumferential strap inboard of the stiffener.
TESTS AND RESEARCH
Autospeedbrake Deployment Logic
According to the airplane manufacturer, automatic deployment of the speedbrakes could only occur when the speedbrake handle was in the armed position, the radio altitude was less than 10 feet, the main gear tires were spun up to at least 60 knots, and both throttle resolver angles were less than 44 degrees TRA.
According to the Boeing 737-NG Flight Crew Training Manual (FCTM), autothrottle use was recommended during takeoff and climb in either automatic or manual flight. During all other phases of flight, autothrottle use was recommended, "only when the autopilot is engaged."
The FCTM also advised that if the airplane should bounce, the pilot should hold or re-establish a normal landing attitude and add thrust as necessary to control the rate of descent. The FCTM also advised, that thrust need not be added for a "shallow skip or bounce"; however, if a "high, hard bounce" occurred, the pilot should initiate a go-around. It stated that the pilot should apply go-around thrust and use normal go-around procedures. It also cautioned the pilot to not retract the landing gear until a positive rate of climb was established because a second touchdown could occur during the go-around.
FAA Human Factors Guidance
According to an FAA document dated June 18, 1996, titled "Federal Aviation Administration Human Factors Team Report on The Interfaces Between Flightcrews and Modern Flight Deck Systems," the FAA's Transport Airplane Directorate, under the approval of the Director, Aircraft Certification Service, launched a study to evaluate the flightcrew/flight deck automation interfaces of current generation transport category airplanes. According to the report, this was a result of a 1994 accident in Nagoya, Japan, as well as other incidents and accidents that appeared to highlight difficulties in flightcrews interacting with flight deck automation.
The report focused on the interfaces that affect flight path management. A team from the FAA and the European Joint Aviation Authorities, as well as technical advisors from Ohio State University, the University of Illinois, and the University of Texas produced the report.
According to the report, the human factors (HF) team received several comments regarding mixed-mode flying. Mixed-mode flying combines elements of automatic and manual control such that the airplane is neither completely under automatic control, nor solely under manual control (e.g., manually controlling pitch, bank, yaw, and flight path while the autothrottle is engaged). Some operators stated they expressly discouraged mixed-mode flying on some airplane types, while others generally encouraged its use as a means to retain manual skills proficiency while minimizing workload and taking advantage of partial task automation (e.g., using the autothrottle to maintain speed control). Possible hazards of mixed-mode flying the team identified were it could lead to unintended mode or configurations changes, cause cross-coupling and inappropriate pitch or thrust responses, mask trends in the airplane's flight path or energy state, or make it more difficult to discern who (or what) was controlling the airplane. The HF team considered mixed-mode flying to be appropriate when conducted in a manner consistent with the airplane manufacturer's design intent and assumptions. They noted, however, that flightcrews should be trained in its advantages, limitations, and proper use. Also, specific procedures should be established and included in training programs. Where mixed-mode flying was not recommended due to potential vulnerabilities, they stated that operators should carefully adhere to the manufacturer's procedures or constraints.
As a result of the study, the HF team recommended that the FAA require operator manuals and initial/recurrent qualification programs to provide clear and concise examples of circumstances in which the autopilot should be engaged, disengaged, or used in a mode with greater or lesser authority, the conditions under which the autopilot or autothrottle will or will not engage, will disengage, or will revert to another mode, and appropriate combinations of automatic and manual flight path control (e.g., autothrottle engaged with the autopilot off).
Flight Safety Foundation Guidance
According to the Flight Safety Foundation (FSF), hard landings can be avoided by conducting a stabilized approach and using proper flare techniques, and by executing a go-around if the approach became unstable or if the aircraft bounced more than 5 feet.
A review by Safety Board investigators shortly after the accident revealed that no prohibition for mixed-mode (autopilot off/autothrottle on) operations, or bounced landing recovery guidance was listed in the operator's 737-800 Operations Manual, Flight Operations Manual, or Flight Operations Training Manual.
Previous Safety Board Recommendations
As a result of an accident that occurred on May 9, 2004, in San Juan, Puerto Rico, the Safety Board issued Recommendation A-05-30 to the FAA, to "Require all 14 Code of Federal Regulations Part 121 and 135 air carriers to incorporate bounced landing recovery techniques in their flight manuals and to teach these techniques during initial and recurrent training."
The FAA replied by issuing Safety Alert for Operators (SAFO) 06005 on June 9, 2006. The SAFO recommended that each 14 CFR Part 121 and 135 airline check to see that bounced landing recovery techniques were included in the manuals used by their pilots, and in their initial ground training for each of the airplane types that the airline flew; and that those same techniques were reinforced by briefings and debriefings during flight training, supervised operating experience, and line checks. The SAFO also included instructions on how to develop information on bounced landing recovery techniques if it was not already addressed by the carrier.
The Safety Board noted that the SAFO was not a requirement; rather, it was a recommendation from the FAA. In order to ensure that issuance of the SAFO fully met the intent of the recommendation to require that airlines train their pilots in bounced landing recovery techniques during initial and recurrent training, the Safety Board asked the FAA to survey Part 121 and 135 carriers to determine how many had adopted the recommendations in SAFO 06005.
On February 15, 2007 the Safety Board advised the FAA that pending results from the survey of 14 CFR Part 121 and 135 air carriers, Safety Recommendation A-05-30 would remain classified; Open Acceptable Response.
On December 19, 2005, the operator advised the Safety Board that all of their flight crews attended a safety briefing regarding the accident, a pilot "read file" was issued regarding bounced landing procedures, and a Boeing produced tail strike video was required to be reviewed by all their pilots.
On June 25, 2007, the operator advised the Safety Board that they would also comply with the recommendations on autothrottle use contained in the Boeing 737-NG Flight Crew Training Manual, and issued a company bulletin to their flight crews, which addressed "Autothrottle Mixed-mode Flying" on July 12, 2007.
Airline Transport; Commercial
Multi-engine Land; Single-engine Land
Other Aircraft Rating(s):
Seatbelt, Shoulder harness
Second Pilot Present:
Class 1 With Waivers/Limitations
Last FAA Medical Exam:
Last Flight Review or Equivalent:
8141 hours (Total, all aircraft), 1258 hours (Total, this make and model), 5575 hours (Pilot In Command, all aircraft), 113 hours (Last 90 days, all aircraft), 71 hours (Last 30 days, all aircraft), 4 hours (Last 24 hours, all aircraft)
Multi-engine Land; Single-engine Land
Other Aircraft Rating(s):
Seatbelt, Shoulder harness
Second Pilot Present:
Class 1 Without Waivers/Limitations
Last FAA Medical Exam:
Last Flight Review or Equivalent:
3072 hours (Total, all aircraft), 748 hours (Pilot In Command, all aircraft), 105 hours (Last 90 days, all aircraft), 52 hours (Last 30 days, all aircraft), 4 hours (Last 24 hours, all aircraft)
Aircraft and Owner/Operator Information
Year of Manufacture:
Landing Gear Type:
Retractable - Tricycle
Date/Type of Last Inspection:
09/01/2005, Continuous Airworthiness
Certified Max Gross Wt.:
Time Since Last Inspection:
2 Turbo Fan
Airframe Total Time:
17022 Hours at time of accident
Installed, not activated
International Lease Finance Corporation
Miami Air International
Operating Certificate(s) Held:
Flag carrier (121)
Operator Does Business As:
Operator Designator Code:
Meteorological Information and Flight Plan
Conditions at Accident Site:
Condition of Light:
Observation Facility, Elevation:
UNV, 1239 ft msl
Distance from Accident Site:
Direction from Accident Site:
Lowest Cloud Condition:
4 knots /
Turbulence Type Forecast/Actual:
Turbulence Severity Forecast/Actual:
30.24 inches Hg
3°C / -6°C
Precipitation and Obscuration:
No Obscuration; No Precipitation
Lansing, MI (LAN)
Type of Flight Plan Filed:
State College, PA (UNV)
Type of Clearance:
Type of Airspace:
University Park Airport (UNV)
Runway Surface Type:
Runway Surface Condition:
ILS; RNAV; Visual
6701 ft / 150 ft
Wreckage and Impact Information
Investigator In Charge (IIC):
Todd G Gunther
Additional Participating Persons:
Alan W Erickson; FAA/FSDO; Miami, FL
John E Feliu; MAPA; Miami, FL
Mark H Smith; The Boeing Company; Seattle, WA
NTSB accident and incident dockets serve as permanent archival information for the NTSB’s investigations. Dockets released prior to June 1, 2009 are publicly available from the NTSB’s Record Management Division at firstname.lastname@example.org, or at 800-877-6799. Dockets released after this date are available at http://dms.ntsb.gov/pubdms/.