HISTORY OF FLIGHT Use your browsers 'back' function to return to synopsisReturn to Query Page
On December 27, 2000, at 2110 central standard time, an Embraer EMB-135LR, N721HS, operating as American Eagle Airlines flight 230, encountered pitch control problems during initial climb, following takeoff from runway 32L at O'Hare International Airport, Chicago, Illinois. The flight crew made two attempts to land the airplane before finally landing the airplane on the third attempt, on runway 4R at O'Hare. Visual meteorological conditions prevailed at the time of the incident. The scheduled domestic passenger flight was being conducted under 14 CFR Part 121. An instrument flight rules flight plan was on file from Chicago, Illinois, to La Crosse, Wisconsin. The captain, first officer, flight attendant, and 9 passengers on board the flight reported no injuries. The flight was originating at the time of the incident.
In his interview, the captain said that everything was normal up to takeoff. After conducting a thorough briefing with the first officer, the captain said they taxied into position, did the before takeoff check, and advanced the throttles. The airplane accelerated normally, rotated, and "jumped into the air. We were pretty empty, so that didn't surprise me." The captain said, at 800 feet agl (above ground level), "I called for the climb sequence checklist." The captain said that after rotation it is natural to begin to trim, but could not remember exactly when he noticed that the trim was not working. When he did realize the trim wasn't working, he said he knew he needed to bring the aircraft back. "I was having problems trimming nose down." The captain asked the first officer to check his trim switch and insure his control yoke was full forward. He said that his yoke was already full forward. The airplane was cleared to 5,000 feet msl (mean sea level). "We were holding a 3,000 to 6,000 foot rate of climb." The captain was finally able to stop the climb and level the airplane at approximately 8,000 feet. He said that at no time during the flight were they able to hold a true level flight attitude. "We were able to keep the airplane within 300 feet above and below [a particular altitude]." The captain reduced power to bring stability to the aircraft. "At that time, we had no EFIS (Electronic Flight Instrument System) messages at all." The captain said that he had the first officer declare an emergency with Chicago Approach Control and refer to the Pitch Trim 1 and 2 Failure checklists. The captain said that the checklist states to reduce speed. He said they did that. The captain then instructed the first officer to lower flaps to 9 degrees, as specified in the checklist. The first officer was hesitant to do so, but it was in the checklist, so they tried it. "When we brought in the flaps, the aircraft pitched way nose high. We were out of control." When the captain regained control of the airplane, he had the first officer retract the flaps. The captain said that the control yoke remained in the forward position from that time until landing.
"At this time, we were northwest of Chicago. This thing (the airplane) had a mind of its own." The captain told the first officer, "let's bring the gear down." The captain said it improved the stability, slightly. The captain then had the first officer pull the pitch trim 1 and pitch trim 2 systems circuit breakers. "We recycled them, nothing happened. I then told him to pull them out again and count to 10, then reset the circuit breakers. He did. Then I checked my trim switch. He checked his [trim switch] and the back up [trim switch]; nothing worked."
The captain said that about this time, approach control asked them if they wanted to land? "I told them that I could see Chicago very well. They told me, runway 9R is available. We did the descent checklist and the approach checklist. I was lined up for an approach to [runway] 9R, but I was too high and too fast. I did a 360 [degree turn]. I was still high, about 2,000 to 2,500 feet msl. The runway was close by." The captain said that about 10 seconds after he completed the 360-degree turn, he extended the spoilers. He said that the reason he tried the spoilers was because up to that point, everything they had tried had not worked. "The airplane wanted to always pitch up. You could maintain a shallow rate of descent, but not control it. The spoilers should increase the rate of descent. I wanted to bring it (the airplane) in, in one piece. I was going to be at the end of the runway, if I didn't put them (the spoilers) out. That maneuver (extending the spoilers); the airplane pitched way up, so abrupt." The captain said he and the first officer pushed both yokes forward. He advanced the throttles, and retracted the spoilers. "That second, I banked hard left, 50 to 60 degrees as I recall, and chopped the power. It took all our abilities to get the nose down."
After they got the airplane back under control, the captain said that approach control came back and told them that runway 4R was straight ahead. The captain told the first officer, "There's the airport. I'm going to [runway] 4R." The first officer told the flight attendant to prepare for landing. The captain said he left the airplane configured as it was (landing gear down, flaps and spoilers retracted) and flew a long shallow approach. "I said to myself, God please let me land this airplane. Over the runway, I chopped the power and let it settle on the runway."
Once on the ground, the captain said, "I told the folks, we're in Chicago. We're safe. I taxied to the gate, did the parking checklist, and just sat there. The passengers got off. The aircraft was still powered up. The mechanics came on board. I was numb and angry. I guess that was it."
In his interview, the first officer said, "When we rotated, this was my first thought of a problem. We rotated very rapidly, jumped off of the ground." He said that when they were 800 feet above ground level (agl) the captain called for "flaps up ... speed 200 climb schedule ... and the climb checklist." The first officer said that was when he got the first indications from the captain that there was a problem. The first officer said he checked the trim switch on his control yoke. It didn't work. He said he immediately went to the backup switch on the center console. It didn't work either. He said the captain asked him to help him push the control column forward. "It was against the stop."
The first officer said that they were climbing in excess of 1,000 feet per minute with the control column full forward. He declared an emergency with approach control and requested to come back to the airport and land on runway 9R. The first officer said that it took slow power reductions to bring the nose of the airplane down. The first officer said that the captain had him run the "Pitch Trim Inop[erative] Checklist", which called for "Flaps 9 [degrees]". He said that when they put the flaps down the nose of the airplane began coming up again. He retracted the flaps. He said that he pulled and reset the Pitch Trim 1 and Pitch Trim 2 circuit breakers twice and then tried the trim switches again. "Nothing worked."
The first officer said that when they were 10 miles from runway 9R, they were given clearance for the approach. They put down the landing gear, which seemed to help. "It noticeably lowered the pitch attitude, so that now we were in a slight descent." The first officer said that they were still high and fast, so they performed a 360-degree turn. Following the turn and continuing inbound for runway 9R, the first officer said that the captain then tried the spoilers. "The airplane went nuts. It didn't like that. We put them in and initiated a go-around. The airplane went like a bat-out-of-hell. We were in a steep climb and bleeding [airspeed] off rapidly." The first officer said that the captain rolled the airplane into a 45 to 60 degree steep bank that brought the nose back down. The first officer said that he noticed that the stabilizer trim had moved to 8 degrees after the spoilers were deployed. He also said that it was about this time that they received warning messages that their pitch trim was inoperative.
The first officer said that they maneuvered toward runway 4R. The captain said that this was their chance to land. The first officer ran the No Flap checklist. "We were 185 knots, 2 miles out. We were probably at that speed when we touched down."
The airplane was examined at American Eagle's maintenance facility at O'Hare International Airport on December 28, 2000. An on-ground functional check of the stabilizer trim system was accomplished and showed no anomalies. An on-ground functional check of the airplane's spoiler control unit revealed that when the spoilers were given the command to retract, the unit was not sending an input to the horizontal stabilizer control unit (HSCU) to put in 1 unit of nose down stabilizer trim.
The airplane's digital flight data recorder (DFDR) and cockpit voice recorder (CVR) were removed and sent to NTSB Headquarters, Washington, DC, for readout.
The airplane's HSCU, horizontal stab actuator (HSA), left hand trim switch, right hand trim switch, trim control module, and spoiler control unit were removed and retained for further testing.
The captain said that he had been an employee for American Eagle Airlines, Inc., since 1986. He reported becoming a captain on the EMB-135 in October 1999, and said that he had approximately 800 flying hours on Embraer EMB-135 and EMB-145 airplanes. He reported having approximately 11,500 total flying hours. Company records showed the captain completed a company line check on September 28, 2000.
The first officer said he had become an employee for American Eagle Airlines, Inc., in September 1999. He reported completing his qualification on the EMB 135 in November 2000. He estimated that he had approximately 2,400 total flying hours and 150 hours in the EMB-135. Company records showed the first officer completed a line qualification check on November 4, 2000.
The airplane was placed in service on July 19, 2000. The airplane was being maintained under a company continuous maintenance program. The airplane had undergone a Phase 2 inspection ("A" Check) on November 4, 2000. The total airframe time recorded at the time of the inspection was 702.4 hours. At the time of the incident, the airplane had 1,002.9 total airframe hours.
The previous evening, December 26, 2001, the crew reported that after takeoff from they received a Pitch Trim 2 Fail "ECAS" (engine indicating and crew alerting system) message. The captain (the same captain who was on the incident flight of December 27) reported that they were unable to restore the system in flight. The first officer's trim switch and back up trim switches were also inoperative. The captain returned to O'Hare airport uneventfully. The maintenance records show that the HSCU was replaced and a successful operations check of the system was performed.
TESTS AND RESEARCH
The HSA (serial number 0334) was examined and tested at Parker Hannifin Corporation's Customer Support Commercial Division, Irvine, California, on January 11 and 12, 2001. An operational test of the HSA, in accordance with the company's acceptance test procedures, showed no anomalies.
Next, an "output force measurement" test was conducted to determine the unit's clutch slip threshold, and the stall load of the HSA. The certified "design stall load" for the HSA, as specified by Embraer, was 1,874 pounds. A gradual force was applied to the HSA, in extension and retraction, until the 28-volt direct current motors reached their peak current design cutout of 33 amperes. The maximum output force applied to the unit's primary channel in extension was measured at approximately 5,600 pounds. In retraction, the force was measured at 4,750 pounds.
On January 12, 2001, the HSA was disassembled to examine its internal parts for wear or damage. The internal inspection of the unit showed no anomalies.
The HSCU (serial number 1250) was examined and tested at Parker Hannifin Corporation's Electronics Systems Division, Smithtown, New York, on January 18, 2001. An operational test of the HSCU, in accordance with the company's acceptance test procedures, showed no anomalies. The HSCU was then cold soaked for 2 hours at -40 degrees Centigrade. An operational test of the HSCU was conducted at the end of the cooling period. The HSCU functioned normally. Following the examination, system schematic diagrams of the EMB-135 were reviewed. The review showed that the spoiler control unit provides a relay closure output for "spoiler deployed" that is electrically in parallel with the copilot trim switch. Its purpose is to advance the horizontal stabilizer trim by 1 unit nose up when the spoilers are deployed. Retracting the spoilers rolls back the trim by 1 unit.
The trim control module (serial number 145276), and spoiler control unit (serial number 145257) were examined and tested at Precision Avionics and Instruments, Inc., Atlanta, Georgia, on January 29, 2001. A functional circuit continuity test of the spoiler control unit, conducted in accordance with the company's test overhaul procedure, showed no anomalies. A functional test of the trim control module also showed no anomalies.
On February 23, 2001, the left and right hand trim switches were examined at Mason Electrics, San Fernando, California. The examination revealed no anomalies with the switches.
Parties to the investigation were the Federal Aviation Administration (FAA) Flight Standards District Office, Schiller Park, Illinois, Embraer Aircraft Corporation, American Eagle Airlines, Inc., Parker Hannifin Corporation, Precision Avionics and Instruments, Inc., and the Air Line Pilots Association.
The horizontal stab actuator was released to Parker Hannifin Corporation. The HSCU was released to American Eagle Airlines, Inc., through Parker's Electronics Systems Division. The spoiler control unit and trim control panel were released to Precision Avionics and Instruments, Inc. The DFDR and CVR were returned to American Eagle Airlines, Inc.
On January 8, 2000, Embraer issued alert service bulletin 145-27-A077 to all operators of EMB-135 and 145 airplanes mandating the installation of a cockpit placard and revisions to the EMB-135/145-series airplane flight manual. The changes established a maximum speed to pitch-trim after takeoff. The placard reads "Airspeed after takeoff/during climb without retrimming max 160 KIAS." The service bulletin directed operators to affix the placard to the instrument panels at the captain's and first officer's stations.
On January 19, 2000, the FAA Transport Airplane Directorate issued emergency airworthiness directive (AD) 2001-02-51 mandating the installation of the cockpit placard and revisions to the EMB-135/145-series airplane flight manual (AFM).
On January 30, 2001, the FAA Aircraft Certification Office (ACO) held a meeting in Atlanta, Georgia, to reveal what they had discovered through their discussions with Embraer. Present at the meeting were representatives from the NTSB, Embraer, and ALPA. The ACO revealed that they had received reports of 10 incidents involving the temporary loss of pitch trim command after takeoff and during the climb phase on EMB-135/145 series aircraft. They stated that during the investigation of the reported incident, the HSA was determined to be inadequate to move the horizontal stabilizer in all flight conditions. "The incidents are most likely caused when the flight crew fails to trim the airplane after takeoff before reaching a certain airspeed where the air loads on the stabilizer may overpower the trim actuator, resulting in the horizontal stabilizer not responding to the pitch trim command from the flight crew." The ACO stated that, "Embraer has recently presented flight test data to verify that the load on the horizontal stabilizer actuator (HSA) is directly proportional to the force on the control column. The HSA will reach its force limit at approximately 66 lbf (pounds-force) on the control column."
The ACO listed several contributing factors:
Approximately one year ago, a change was made to the bellcrank in the control column that reduced the control column forces by approximately 30 percent nose up, and 20 percent nose down. All 10 incidents have occurred on aircraft which have this modified bellcrank installed. It is theorized that the reduction in the control column forces on some aircraft may have lead flight crews to delay adjusting the aircraft trim until higher airspeeds have been reached. Preliminary review indicates that full certification tests were done with the new bellcrank to verify compliance with all Federal Aviation Regulations, including minimum stick force per "g" limit of 1 pound per 6 knots.
Because of a fatigue problem, a modification was made to the HSA motor that resulted in a lower output capability of the HSA (from 1840 kilograms-force to 1480 kilograms-force maximum). This lower minimum limit was still significantly above design specification, which was based on the load for the maximum out-of-trim takeoff condition (i.e., takeoff with the horizontal stabilizer initially set at the limit out-of-trim condition within the takeoff "green band"). Based on the analysis that showed the reduction in force was still above design specification, no flight tests were done for this modification. Four of the 10 incidents occurred on aircraft with the HSA with the higher force limit; 6 on those with the lower limit HSA.
For all of the 10 incidents where the flight crew followed the AFM procedure to reduce airspeed for pitch trim inoperative, pitch trim control was regained. On the incident in Chicago on December 27, the airspeed was never reduced low enough to regain control of the trim (but the flight crew had also pulled the circuit breakers so they would not have had trim even if they had reduced airspeed). However, the previous AFM procedures for pitch trim inoperative were not completely clear, thus the AD to mandate changes to clarify the correct procedures and limitations.
The current design logic for presenting a CAS message for a "stall" of the HSA requires that the flight crew first command pitch trim four times for three seconds each time before it senses a failure. The logic prevented any indication to the flight crew of a failure of the trim system.
The ACO reached the following conclusions:
The interim actions of AD 2001-02-51 significantly minimizes the exposure of experiencing pitch trim failure while the AFM revisions provide necessary instruction for the flight crew to regain pitch trim authority if a pitch trim failure is experienced. FAA pilots have stated that the procedures provided in the AFM revision are clear, and intuitive. The limitation imposed by this AD to start trimming below 160 KIAS will be terminated once permanent actions are determined and are fully implemented.
The DAC, CTA-IFI (the Brazilian Airworthiness Authority for Certification) will develop and provide some suggestions for revising certification test methods to prevent similar problems on future aircraft. The FAA will work with them on this effort.
The FAA is currently planning to mandate the fleet inspection of all the horizontal stabilizer actuators to ensure they meet the current design load limit as soon as that procedure is developed and approved by the CTA. This remains an interim action until such time as permanent design changes are made.
Final design changes to be implemented will be presented to the CTA. The FAA and CTA will review the changes together to approve them as soon as possible. They will include at a minimum:
- Redesign of the pitch trim failure indications,
- Redesign of the HSA to increase limit loads, and,
- Ergonomic improvements of the yoke trim switches.