On April 29, 1993, at 1555 central daylight time, an Embraer EMB-120 RT, Brasilia, N24706, was substantially damaged when it collided with rough terrain during an overrun following a forced landing at the Grider Field airport in Pine Bluff, Arkansas. The forced landing was executed following a stall and loss of control at 17,412 feet during climb which resulted in damage to the left engine and propeller. The airplane, owned by Continental Airlines, Inc., operated by Continental Express, Inc., was using the call sign of Jet Link flight 2733. It was flown by two ATP rated pilots, on a 14 CFR Part 135 scheduled passenger flight from Little Rock, Arkansas, to Houston, Texas. An Instrument Flight Rules (IFR) plan was filed and in effect and visual meteorological conditions (VMC) prevailed at the accident site. Instrument meteorological conditions (IMC) prevailed during the loss of control, descent, and recovery. Of the three crewmembers and twenty seven passengers aboard, the flight attendant and twelve passengers received minor injuries, while the two flightcrew members and remaining fifteen passengers were not injured.

The flight departed the gate at Little Rock, on time, and received takeoff clearance on runway 22R, en route to Houston's Intercontinental Airport, at 1516:00. After a normal takeoff and initial climb, the flight contacted Little Rock Departure Control and was instructed to join the J-180 airway and maintain 10,000 feet. At 1522:09, the flight was handed off to Memphis Air Route Traffic Control Center (ARTCC). Upon contacting the center, the crew reported climbing through 7,500 feet and was instructed to climb and maintain flight level (FL) 220. The cockpit voice recorder (CVR) then recorded a conversation between the captain and the first officer in which they discussed the performance data for a climb to FL260. It was noted on the tape that the captain said "I don't care" in response to a question from the first officer regarding what final altitude the captainwanted for cruise. During this exchange, at 1522:54, while the airplane was passing through about 8,000 feet, the CVR recorded the voice of the flight attendant saying "Hi." Non-pertinent conversation between the flightcrew followed for about one minute. At 1526:26, Memphis Center instructed the flight to continue the climb and gave a final cruise altitude of FL230. The first officer, who was handling the radios, requested and was given FL220 as a final altitude. That was the last radio contact between the flight and air traffic control (ATC) prior to the loss of control. At 1528:49, the flight attendant again entered the flight deck and began a conversation with the captain that lasted until the time of the loss of control at 1533:16, four minutes and twenty seven seconds later. The crew discussed using the windshield wipers to remove something from the wind screen. They later said they were talking about bugs. At 1530:52, the flight attendant requested that the captain "climb faster" as she wanted to begin cabin service and she could not drag the beverage cart "uphill" during the climb. The captain agreed and subsequently said, "Okay we'll try to get up a little more" and "yeah we're almost there another six thousand feet another six minutes." This exchange was followed by non-pertinent conversation between the captain and the flight attendant, during which the first officer commented that they were not climbing very fast. The captain replied, "heavy really heavy" and continued with the conversation. At 1533:11, the captain interrupted the conversation with the flight attendant and said, "Frank hang on something ain't right." This was followed by the sound of the autoflight system disconnect at 1533:16.3, and stick shaker activation at 1533:16.8. At 1533:18, the aural stall warning activated and the captain said, "airspeed." The stick shaker and aural stall warning continued until the end of the recording at 1533:46.7. At 1533:22.7, the captain again said "hang on" and at 1533:24.6, the first officer said, "power up power's." This was followed by increasing engine noise at 1533:25.6 and the beginning of vibrations through the airframe at 1533:34.7. At 1533:39.7, the engine noise decreased and was then no longer heard. At 1534:26, Memphis Center attempted, unsuccessfully, to hand off flight 2733 to Fort Worth Center and at 1534:50, flight 2733 contacted Memphis Center and declared an emergency, stating they had "lost an engine and needed to put her down."

The captain stated that immediately prior to the loss of control, he noticed the ball was slewed full left and the rudder trim wheel was trimmed 10 units right, its full limit. According to the digital flight data recorder (DFDR) data, the onset of the loss of control was characterized by an initial constant right yaw with left yoke displacement and then several oscillating bank angles to the right and left that increased in rate and angle severity until the airplane stalled and entered a post-stall gyration. The airplane continued in the gyration for one rotation to the left and then stabilized in a nose down oscillating attitude. Roll oscillations in excess of 111 degrees and pitch attitude values in excess of 67 degrees aircraft nose down were recorded prior to and during the loss of control and descent. Recovery was initiated after the first officer lowered the landing gear. The airplane subsequently entered a secondary stall during the recovery at about 6,700 feet and ultimately returned to controlled flight at 5,500 feet. The loss of control, descent, and recovery all occurred in instrument meteorological conditions (IMC).

The yoke position channel of the DFDR indicated that it was never pushed forward during or following the initial oscillations of the airplane at the beginning of the loss of control. The data also indicated that the control inputs to correct the roll oscillations were out of phase with what the airplane was actually doing throughout most of the recovery. The roll corrections did not phase with the airplane until several seconds after the post-stall gyration completed one revolution and the yoke was never pushed forward. The DFDR and flight data acquisition unit (FDAU) data also indicated that following the initial recovery, the airplane entered a nose high, steep left bank attitude which was maintained until a secondary stall was entered. Control movements during the second recovery were more coordinated than during the first.

Following recovery, the crew contacted Memphis Center, declared an emergency and requested vectors to the nearest airport, stating that they could not maintain level flight. The crew was offered several airports as options and selected Pine Bluff. The captain stated that he ordered the left engine shut down during the descent; he thought he had experienced an over speed on that engine. Following the recovery, the captain noticed the left engine was missing three of the four propeller blades, all of the upper cowlings, and was displaced in the mounts. The crew stated that the airplane would fly at an airspeed of about 125 knots, before the stick shaker activated, and maintain a rate of descent that varied from zero to 500 feet per minute. The crew further stated that they had difficulty turning the airplane to the right.

The airplane broke out of IMC close in to the airport. The captain stated that he overshot the right turn to final due to controllability problems and the airplane touched down with 1,880 feet of wet runway remaining. The captain further stated that he applied the brakes at touchdown and the airplane immediately began hydroplaning and went off the departure end of runway 17, onto wet rough sod, avoiding a vehicle and construction personnel. Tracks found on the runway were consistent with hydroplaning. After passing between the ILS antenna and the ILS equipment building, the airplane came to rest in a rice field, 687 feet from the end of the runway. The crew and passengers immediately evacuated the airplane. The right engine could not be shut down by the crew or aircraft rescue and fire fighting (ARFF) personnel and continued to run in a pool of Jet A fuel for about 15 minutes.


In addition to the crewmembers, statements were obtained from passengers, other pilots, ground witnesses, and air traffic control personnel who were in contact with the flight. Twenty six of the twenty seven passengers responded with written statements, while the other one was contacted directly and her observations obtained. Five pilots were identified that were flying through the same airspace as flight 2733. Four were interviewed by telephone and one provided a written statement. Additionally, seven statements were obtained from the air traffic controllers and six statements were received from ground witnesses located at the Pine Bluff airport.

Several of the passengers commented on the smoothness and lack of turbulence on the flight, up to the beginning of the loss of control. Two specifically noted that they did not observe any lightning or hear thunder. A few characterized the onset of the event as a shudder through the airframe and associated it with turbulence. Some remembered the airplane initially banking left, then right, then left again with greater violence. Others had the sequence reversed. One passenger commented that he observed a whitish substance that appeared to be "snow" on the windshield and another commented that the takeoff roll at Little Rock, appeared to take longer than he was used to in similar airplanes. The passenger seated in the front row, seat 1B, stated that during the climb, she observed the captain put his seat back, unbuckle his seat belt and put his foot up on the console. She further said, "I watched the pilot (captain) turning a wheel which I think made us turn to the right. He kept turning a knob it looked like for balance." A majority of the passengers recalled the flight attendant's pre-takeoff emergency procedure briefing and her pointing out the locations of the emergency exits. They also recalled that prior to the landing she instructed them on assuming the impact position, the locations of the exits, and that the landing would be "hard and fast."

All of the statements from the ground witnesses and ATC personnel were consistent with the evidence and tape transcripts. Of the five pilots who responded, two remembered encountering icing conditions in the vicinity of where the loss of control occurred.



FATAL 0 0 0 SERIOUS 0 0 0 MINOR/NONE 3 27 0


The captain, the pilot flying, held an airline transport pilot rating with airplane, single and multiengine land privileges. In addition, he held a Class I medical certificate, issued on November 16, 1992, with a limitation for the use of corrective lenses. He passed his current 14 CFR Part 135 and instrument recurrent proficiency check ride in the EMB-120, on February 12, 1993. He was hired by Continental Express, Inc., on September 11, 1989, and had qualified as a Captain in the EMB-120, on September 13, 1990. Company records indicated that at the time of the accident, he had accumulated a total of 3,600 flight hours, of which 2,600 hours were in the accident make and model. The records indicated that he had flown 130 hours actual instrument time and 40 hours simulated instrument time. During the 90 and 30 day periods prior to the accident, he had flown 204 and 77 hours respectively, all in the accident make and model.

According to the operator's domicile chief pilot, the captain's greatest strength as a pilot was his ability to establish an open cockpit environment with first officers. According to the accident first officer, the captain was easy to get along with and not intimidating. Two first officers, who had flown previously with the captain, agreed and indicated that he set up a relaxed cockpit climate.

A review of the captain's schedule revealed that prior to reporting for the three day trip that culminated in the accident, he had two days off. On the first day of the trip, he flew 4.1 hours, with 9.5 hours duty time followed by 8.5 hours of rest time. On day two, he flew 2.3 hours during 3.8 hours of duty time, followed by 18.6 hours of rest time. On the third day of the trip, he flew 6.1 hours during 10.3 hours of duty time. The captain indicated that during the 51 hour period of the three day sequence, he had slept a total of about 10.6 hours, out of a total of 27.2 hours of scheduled crew rest. During the period, he had one reduced rest period and one extended rest period. The captain stated that he felt well rested prior to departing on the sequence the day of the accident. The accident flight occurred during the seventh and last flight of the day.

The captain had no record of being disciplined by the company for his flying activities, according to the chief pilot. Federal Aviation Administration (FAA) and National Transportation Safety Board (NTSB) records revealed no previous violations or accidents. The captain indicated there had been no major changes in the past twelve months in his financial or personal situations. He was married and had two children.

The first officer, the pilot not flying, held an airline transport pilot rating with airplane, single and multiengine land privileges. He held a Class II medical certificate, issued without limitations, on June 12, 1992. He had completed his current 14 CFR Part 135 and instrument check in the EMB-120, on November 10, 1992. He was hired by the company on June 25, 1990, and qualified as a first officer on the EMB-120, on October 15, 1991, after having served as a captain for the company on different equipment at a different domicile. Company records indicated he had accumulated a total of 3,300 flight hours, of which 700 were in the accident make and model. These included 310 hours in actual instrument conditions and 60 hours in simulated instrument conditions. During the 90 and 60 days preceding the accident, he had flown 199 and 68 hours respectively, all in the accident make and model. It was revealed during the crew interviews that the first officer was an aerobatics pilot and flew in aerobatics competition during his off duty time.

The domicile chief pilot stated that the first officer's greatest strength as a pilot was his ability to give input and demonstrate the principles of crew resource management and assertiveness. The captain of the accident flight stated that the first officer's greatest attribute as a pilot was vigilance and attention to detail in the cockpit. He described the first officer as a "good pilot who loved aviation and who had taught him a lot." The first officer's flight, duty and crew rest schedule was the same as that of the captain for the three day trip sequence. During a post-accident interview, he stated that he had slept about 10.0 hours during the scheduled time off.

The first officer had not received discipline during his employment with the company and audits of FAA and NTSB records indicated no previous violations or accidents. He stated that his financial situation had been downgraded in the recent past when he took a company move to Houston and a downgrade from a captain's position to first officer.

Both flightcrew members indicated they had received initial training in crew resource management (CRM) during their initial ground school training, even though it was not required for 14 CFR Part 135 operators. In addition, the operator's director of flight operations and director of inflight services indicated that flight crews and flight attendants trained together on CRM during recurrent training in a program that had been in place for about two years.

The flight attendant was hired by Continental Express, Inc., on December 18, 1992, following her successful completion of initial training on December 14, 1992. She was flying the same trip sequence and rest periods as the flightcrew during the three days prior to the accident. There was no record of her having been disciplined by the company during her employment.


The airplane, an Embraer EMB-120 RT, Brasilia, serial number 120.093, was certificated in the transport category, with a maximum certified takeoff gross weight of 25,353 pounds. It was equipped with two Pratt & Whitney-Canada PW-118, 1,800 shaft horsepower engines and two Hamilton-Standard model 14RF-9, four blade propellers. A review of the airplane's maintenance records indicated that it was being maintained under a FAA approved continuous airworthiness program and had last been inspected on April 19, 1993, at a total airframe time of 10,324.5 hours, 73.5 hours prior to the accident.

An audit of the airplane maintenance records did not reveal any outstanding discrepancies that would have affected its airworthiness. The left propeller was on a watch list for a leaking hub. The hub had been inspected earlier on the day of the accident in Houston and found to be within limits. The leak inspection did not require that the blade de-ice leads be disconnected and evidence indicated that all the leads were attached when the three blades departed the hub.

The flight dispatch records indicated the airplane was refueled in Little Rock, and departed with 2,701.6 pounds of Jet A. Takeoff gross weight was estimated to be 25,333.3 pounds, which included 752 pounds of carry on and checked baggage. Passengers estimated their carry on and checked baggage to weigh about 552 pounds. Company personnel initially boarded thirty passengers at the gate, however, three passengers were off loaded prior to departure for weight and balance considerations. The data showed the airplane to be within the prescribed limits for weight and center of gravity at takeoff, at the time of the loss of control, and at landing.

The aircraft performance group used recorded radar data, weather data, CVR data, and DFDR information to develop a time history on the flight parameters of flight 2733. Correlation of the data indicated behavior consistent with a normal takeoff, initial climb, and a steady climb to altitude with a normal power reduction to 90% (climb power) at about 12,400 feet. The data showed a steady state climb pitch attitude of about 3.2 degrees and an airspeed of about 185 knots until 1530:58, about fourteen and one half minutes into the flight. At that time, the airspeed began to decrease and the pitch angle began to increase until it ultimately reached 6.5 degrees at 1531:54. The point at which these parameters began to change corresponded to the point in time when the flight attendant asked the captain if he could climb faster and he agreed. At the beginning of the loss of control, the airspeed had decreased to 138.57 knots and the pitch angle had increased to 6.5 degrees airplane nose up. There was no increase in engine power after the reduction at 12,400 feet. The crew stated that the autoflight system was configured in the pitch and heading hold modes throughout the climb. A review of the manufacturer's performance data for the gross weight and atmospheric conditions that existed indicated that for an international standard atmosphere (ISA) +0 day, climb speed should have been 155 knots indicated airspeed which would have resulted in a climb rate of 1,333 feet per minute (FPM). In ISA +10 conditions, the target climb speed would have remained the same and the rate of climb would have been 800 feet per minute. The performance data also indicated that for the conditions present, the stick shaker should have activated at 127 knots indicated airspeed (IAS) and the airplane would have stalled at 116 knots IAS. Engineering simulator derived performance data, provided by Embraer after the accident, indicated that the abnormal climb performance and loss of control at a higher than expected airspeed were consistent with aerodynamic degradation of the wings and/or full right rudder deflection.

The airframe manufacturer undertook engineering simulator and flight tests in order to determine the reasons for the airplane's performance deterioration. Normal climb, climb with ice formation, and climb with rudder trim set to 10 units right scenarios were simulated and/or flown. Performance degradation was present during both the icing and abnormal trim simulations. The icing simulation came closest to duplicating the DFDR and FDAU data recovered from the accident flight. The manufacturer's tests did not include a test during which icing and an out of trim condition were introduced simultaneously. METEOROLOGICAL INFORMATION

The 1300 National Weather Service (NWS) surface analysis showed a weak cold front in northwestern Arkansas with a low pressure area in northeastern Texas. The 1600 NWS surface analysis showed a stationary front in northwestern Arkansas with a low in northeastern Texas. An outflow boundary was moving through west central Arkansas; the area in which the accident flight was conducted.

The 1530 special observation taken at Little Rock, reported the weather as 500 feet scattered, measured ceiling 2,500 feet broken, 4,000 feet overcast, visibility 2 miles in light rain showers and fog, winds from 090 degrees at 5 knots, altimeter setting 30.01 inches, with a note that tower visibility was 3 miles.

The 1550 record observation taken at Pine Bluff, reported the weather as estimated ceiling 4,500 feet broken, 10,000 feet overcast, visibility 5 miles in light drizzle and fog, temperature 68 degrees, dew point 62 degrees, winds from 100 degrees at 3 knots, altimeter setting 30.00 inches, with a note that the drizzle was intermittent. A special observation, taken at 1556, immediately after the accident, reported that the wind had changed to 080 degrees at 3 knots, and the altimeter to 29.99 inches. The upper air data taken at Little Rock, indicated the freezing level was at about 11,500 feet and the temperature at 17,400 feet was about minus 11 degrees centigrade. A Meteorological Impact Statement issued by the Memphis ARTCC Weather Service Unit (WSU) called for occasional moderate icing in clouds and precipitation between 12,000 and 20,000 feet.

During the investigation, seven aircraft were identified operating in the same air traffic control sector as the accident airplane at approximately the same time. Five of the seven were tracked down and the captains were interviewed regarding their recollections of the weather conditions they had encountered, especially icing conditions. Two of the captains recalled encountering icing conditions; the captain of United Airlines Flight 421 recalled encountering light to moderate icing conditions and the captain of an Arkansas Power and Light shuttle flight, flying a Beech 1900, recalled encountering a trace of light icing while flying at 15,000 feet, en route from Little Rock, to New Orleans, Louisiana.

The accident flightcrew members did not recall seeing evidence of icing prior to the loss of control. The captain stated that he recalled last looking for ice as the flight passed through about 12,000 feet. Only one of the passengers recalled seeing any evidence of ice. He stated that about 10 minutes after takeoff, the flight attendant commented to other passengers about the snow on the pilot's windshield. He further stated that he looked and saw a whitish substance that appeared to be snow about 8 to 10 inches above the windshield wipers. The wiper blades were mounted vertically. The flight attendant did not recall making any statement about snow.

Review of the enclosed meteorological data showed that the potential for icing existed from 11,500 feet to above 19,000 feet in the area of the loss of control. Review of the data also showed conditions in the area were conducive to the production of large super cooled water droplets and possible icing away from the leading edge of the wing.


Transcripts of the ATC communications between the airplane and the various controlling agencies were reviewed. During the initial stages of the emergency, the crew did not specify the exact nature of their problem and the Memphis Center controller had to work through Little Rock Approach Control in order to determine the availability of adequate airport facilities, approaches, and current weather. Once the crew decided to divert to Pine Bluff, the center controller was not aware of the notices to airmen (NOTAMS) which listed the approach aids as being out of service, or that there was construction equipment on the runway and its availability was limited to the north 3,000 feet. The Memphis Center controller would not normally have had access to the NOTAM, as the facility was not in his area of control.


Pine Bluff, Arkansas, was served by Grider Field, a general aviation airport, which had a single 6,008 foot long by 130 foot wide asphalt runway, oriented 17/35. The airport was served by ILS Rwy 17, VOR Rwy 17, and VOR/DME Rwy 35 instrument approaches. Runway 17 was equipped with medium intensity runway lights, MALSR approach lights, and a two bar VASI. The airport was also equipped with a control tower that was not operational at the time of the accident. Minimum ARFF equipment was available with prior notification.

On the date of the accident, the runway lighting system was undergoing upgrade construction and the south 3,000 feet of runway 17 was closed. All of the runway lighting systems were out of service. In addition, the ILS, and the Pine Bluff VOR and HIWAS were also out of service. A NOTAM had been issued on these service downgrades and was passed on to the crew of flight 2733 at the time they contacted Little Rock Approach Control after the loss of control and recovery.

Little Rock Approach Control notified the airport manager of the inbound emergency about five minutes prior to the airplane's arrival over the airport. The manager immediately took action to clear the runway of construction equipment; however, at the time of the landing, personnel who had not been notified and one vehicle remained on the runway centerline at the departure end of runway 17. FLIGHT RECORDERS

The airplane was equipped with a Sunstrand model 980-4100-FXUS digital flight data recorder. The recorder was not damaged during the accident and the medium was undamaged. The unit was taken to the NTSB's laboratory for readout and evaluation. The data recorded during the accident flight was successfully retrieved along with data from previous flights. In addition, data was obtained from the airplane's FDAU and compared with the DFDR data.

The airplane was equipped with a Fairchild model A-100A cockpit voice recorder. The unit was not damaged during the accident and it was sent to the Safety Board's audio laboratory for readout. The radio channel of the recording was of good quality; however, the channel for the captain's and first officer's hot boom mikes was of poor quality. This was determined to be due to neither crewmember using their headsets, with the boom mikes, during the accident flight. It was also determined that the cockpit area microphone, mounted in the upper right center console, was inoperative.


The crew stated that after touchdown, the airplane immediately began hydroplaning on the wet runway and the crew simultaneously saw the truck and personnel at the end of the runway. The captain said that he steered the airplane to the right to avoid the obstacles. He subsequently steered the airplane between the ILS antenna and the ILS equipment building which were located off of the departure end of runway 17. Both flightcrew members stated that braking action was nil on the runway and braking and steering were nil in the wet grass off the runway.

As the airplane departed the pavement, the right main gear traveled over a three foot deep runway end lighting ditch. The impressions in the grass then veered to the left away from the equipment building. The building was mounted on a three foot high shale pad. The right main landing gear track went up the 45 degree up slope of the pad, while the nose gear track traveled along the left edge of the pad and the left main gear track traversed level ground. In addition, evidence of ground slash marks from the right propeller were found forward of the gear imprint on the pad. After passing over the pad, the airplane ground scar entered a wet rice field. The ground scars indicated that it yawed nose right and came to rest on a measured heading of 220 degrees, about 75 feet beyond the building and 687 feet beyond the departure end of runway 17.

Examination of the wreckage revealed the inboard section of the right wing, between the fuselage root and engine nacelle, was buckled downward about 45 degrees leading edge down and twisted. Both the leading and trailing edges of the right wing were found opened up and the aft spar was broken. The propeller hub and engine inlet of the right engine were found canted downward and partially buried in mud. The right main gear was deflected aft, but not collapsed. All four propeller blades had separated from the right hub and were found in the immediate vicinity of the wreckage. A puncture hole was found on the right side of the fuselage, adjacent to the plane of rotation of the right propeller. The puncture penetrated the outer skin, but did not reach into the cabin area.

The left wing was undamaged. Three of the four blades were found separated from the left propeller hub and have, to date, not been found. The captain stated that he noticed their absence during his survey of the left engine following the recovery. The fourth blade remained loose in the hub in the feathered position. It was free to rotate within the hub. Evidence indicated that the remaining blade made an excursion out of its normal plane of rotation and had impacted the right side of the leading edge of the left engine inlet. Matching scars were found on the trailing edge of the blade and the anti-ice boot on the inlet. All of the engine mounts on the left engine were distorted and displaced. The lower engine deck was broken and rub marks were found between the breaks. The front section of the engine was displaced downward about 5 degrees and left about 5 degrees. There was downward compression buckling on the lower cowling, below the deck. All of the upper cowlings departed the engine and have not been found to date. Examination of the failure mechanics of the cowlings indicated they had departed aft under vibratory and flight loads. This was evidenced by the orientation of the elongation of the attachment holes. Examination of the exterior of the engine itself revealed damage that equated to engine displacement of over 4 inches in all directions in the mounts. No evidence of inflight or post-crash fire was found.

All of the flight controls remained attached in their respective positions and continuity was established to each. The right aileron cables were bound, as were the right engine control cables, due to distortion of the wing. The flaps were found deployed at 25 degrees. All of the airplane's lift enhancing devices remained attached during the accident.

According to the crew, the cockpit configuration had been changed during the descent and partially secured after the landing. Significant cockpit instrument and system readings noted in the "as found" state were:

Stall Warning Panel - ON, BOTH Left Hydraulic Power - AUTO Right Hydraulic Power - AUTO Rudder Power Green - ON Blue - ON Propeller De-ice - TIMER 2, COLD MODE Engine #1 Inlet - OFF Engine #2 Inlet - OFF De-Ice Monitor - ON TAT Sensors - ON AOA Sensors - ON Sideslip - ON Pitot/Static (all) - ON Windshield Heat (both sides)- ON Windshield Defrost - ON Elevator Trim Setting - 3 UNITS (ANU) Aileron Trim Setting - 8 UNITS RIGHT WING DOWN Rudder Trim Setting - 10 UNITS NOSE RIGHT

The trim settings corresponded to the observed tab settings on the surfaces.


Post-accident toxicology samples, requested by the NTSB, were obtained from all three crewmembers upon their arrival in Houston, at 2300, about seven hours after the accident. The results for all three crewmembers were negative for alcohol and drugs.


This was a survivable accident in that the impact forces encountered did not exceed the human tolerance level. Evidence and statements from the occupants indicated that all but one of the exits had been used during the evacuation. These included the captain's cockpit window, the left main entry door, the left over wing exit, the right emergency exit door, and the right over wing exit. The first officer removed his cockpit window; however, he did not use it as an exit. He stated that prior to leaving the airplane, he noticed that one female passenger remained in the cabin and appeared to be having difficulty with her seat belt. He entered the cabin and assisted the passenger from the airplane using the main entry door. The passenger later stated that she could not release her seat belt because she thought she was supposed to push the latch and she thought the release was on the side, as in her car. Examination of the seat attachments and occupant restraint systems revealed none suffered failures or displacement and all were functional.


Powerplant and Propeller Disassembly and Examination: Both engines and propellers were removed from the airframe as units at the accident site and shipped to the facilities of Pratt & Whitney-Canada. Evaluation of the data available from the DFDR, FDAU, and CVR indicated that both engines were performing within normal parameters up to a point about 20 seconds after the beginning of the loss of control. The recorded torque, RPM, temperatures, fuel flows, and propeller RPM values were matched and nominal up to that time. Propeller vibration values were also nominal until the vibration levels began to increase on the left propeller, about 20 seconds after the loss of control. A tear down examination of the left engine and left propeller was conducted. The right engine and propeller were not torn down as they continued to operate within normal parameters until the collision with the terrain after the landing. The examinations did not reveal any evidence of pre-impact failure or malfunction of the components. All of the abnormalities noted were determined to have been event related.

Aircraft Systems Functional Testing, Disassembly and Examination: The airplane's systems and respective components were evaluated and functionally tested in place. Several components were removed and tested off site. These included the autoflight computers, the autoflight control panel, the captain's and first officer's flight control panels, the rudder power control unit, the air data sensors, the flight control servos, and the rudder signal potentiometer. All of the airplane systems, including the stall warning system, tested on site, performed within operational specifications. All of the components removed were functionally tested and found to operate within normal parameters, with the exception of the rudder signal potentiometer, which was found to have a malfunctioning connector.

Continental Express Operations Procedures: Use of Boom Microphones; The Continental Express Flight Operations Manual stated that on "aircraft equipped with a Cockpit Voice Recorder which can record boom or mask microphone signals, the flightcrew members must use boom microphones below 18,000 feet MSL." The requirement was in accordance with 14 CFR Part 135.151(d). The flightcrew did not use their boom microphones during the accident flight.

Sterile Cockpit Procedures: The cockpit door remained open throughout the accident flight, however, there was no federal air regulation or company policy which required that the cockpit door be secured. In accordance with 14 CFR Part 135.100, the flight operations manual stated that "Sterile Cockpit procedures prohibit all activities in the cockpit not required for the safe operation of the aircraft during critical phases of flight. These prohibited activities include non-safety related Company calls, P.A.'s, payroll entries, and non-essential conversations. Critical phases of flight include all ground operation involving taxi (movement of an aircraft under its own power on the airport surface), takeoff and landing, and all other flight operations conducted below 10,000 feet except cruise flight." In addition, the Inflight Manual stated that "During sterile cockpit periods, Flight Attendants are not to enter or call the cockpit unless necessary to relay information regarding emergency situations or incidents involving the safety of the customers and crew. Calls requesting gate information, wheelchairs or any nonessential conversation are prohibited." During the accident flight, the flight attendant interrupted the crew with non-pertinent conversation once as the airplane was passing through 8,000 feet. During a post-accident interview, the director of inflight customer service support stated that flight attendant training stressed that cockpit visits should be limited to flight related conversation. She also stated that a five minute visit by the flight attendant to the cockpit would not be of concern if the conversations dealt with the conduct of the flight.

Autoflight System Description and Operation: The airplane was equipped with a multi-mode autoflight system whose functions were controlled by the autoflight panel. The system's lateral navigation capabilities included roll, heading, navigation, approach, back course, and go around modes. The vertical functions included pitch hold (the default mode), altitude hold, indicated airspeed, vertical speed, altitude preselect, descent, and climb modes. The captain stated that he had selected the heading mode and the pitch hold mode during the flight.

A review of the Continental Express Operations Manual indicated that with the heading select mode engaged, the autoflight system would cause the airplane to turn to and maintain the heading set with the heading bug on the electronic horizontal situation indicator (EHSI). The autoflight system defaulted to the pitch mode if no other vertical modes were selected on the control panel. The operations manual stated that with one of the lateral navigation modes selected and no vertical mode selected, the electronic attitude direction indicator command bars would be in view and display roll commands appropriate to the selected lateral mode and pitch commands to maintain the pitch attitude present at the time of mode selection. It further stated that the pitch command could be changed by engaging the vertical trim toggle switch, located on the center pedestal. Momentary activation would provide a 0.5 degree pitch change. Longer activations would provide a fixed slew rate for the commandbars. The manual also stated "With the autopilot engaged, the airplane responds to the pitch command. If an air data mode is selected, the vertical trim switch provides a fixed incremental change for each actuation." The system description for the climb mode stated that upon engagement, the autoflight system would initiate a gradual climb, stabilizing at an indicated airspeed defined by the climb profile. The climb profile was controlled by the air data computer and was a constant 155 knots up to 20,000 feet and then gradually decreased to about 132 knots at 32,000 feet. If the climb mode was selected at a speed below the climb profile speed, the autoflight system would decrease the rate of climb to 50 feet per minute until the climb profile speed was attained. With the proper power settings selected, the climb mode afforded a stall speed margin throughout the climb envelope, whereas, the pitch mode offered no such speed guarantee. A review of the Continental Express Aircraft Operations Manual, training syllabus, and discussions with the chief pilot, indicated that crews were instructed to climb in either the climb or indicated airspeed modes. Stall Recovery: The Continental Express EMB-120 RT Aircraft Operations Manual, normal procedures section, called for the accomplishment of the following recovery procedures by the pilot flying at the first indication of a stall. "Simultaneously a) apply maximum power, b) level wings, c) hold a pitch attitude to stop deceleration and minimize sink, d) call for flaps 15 degrees." The procedure also stated that the "non-flying pilot must position or leave the flaps at 15 degrees."


Wreckage Release: The main wreckage was released to the operator on June 25, 1993. All of the retained system components were returned to Pine Bluff and placed with the wreckage upon completion of their respective examinations. The engines were released at the facilities of Pratt & Whitney-Canada, as per the operator's instructions. No original aircraft records were retained during the investigation.

Use your browsers 'back' function to return to synopsis
Return to Query Page