LAX96FA228
LAX96FA228

HISTORY OF FLIGHT

On June 6, 1996, at 0558 hours Pacific daylight time, a British Aerospace Jetstream BA-3201, N926AE, ran off the runway while on takeoff from San Luis Obispo County-Mc Cesney Field, San Luis Obispo, California. The aircraft sustained substantial damage; however; the certificated airline transport pilot and copilot were not injured. The aircraft was being operated as a ferry flight by the JSX Capital Corporation, Washington, D.C., under 14 CFR Part 91 when the accident occurred. The flight was originating at the time on an IFR flight plan to Kingman, Arizona. Instrument meteorological conditions prevailed.

The captain, who was in the right seat, reported that he and the first officer completed the cockpit setup, the before start check, the start engine check, and the after start check. He did not report any abnormalities or interruptions during the procedures. The after start checklist directs that the pilot take the propellers off the start locks by individually pulling each power lever into reverse until the crew sees a rise in torque on the torque meter. There is no other cockpit indication as to the position of the propellers relative to the start locks. (Excerpts from the Jetstream Series 3200 Flight Manual are appended to this report.)

The crew requested their IFR clearance and release from Los Angeles center. They were cleared to Kingman via the Avila 3 standard instrument departure (SID). Ground personnel pulled the chocks and the ground power unit (GPU) and the aircraft prepared to taxi. The first officer reported receiving an emergency briefing from the captain.

Both crewmembers reported checking the aircraft brakes and verifying hydraulic pressure gauges from the respective positions. After taxiing to the departure end of runway 11, the crew completed the before takeoff check, made a visual check for traffic, and then made an advisory call over the tower frequency. Following a check for possible air traffic in the vicinity, the aircraft taxied onto the runway.

The first officer, who was on the controls, applied power as the takeoff roll began. The captain set the power, matching both engines at 100 percent N1. As the aircraft accelerated, the airspeed indicator began to register as the airspeed reached about 40 knots. At this point, the captain told the first officer that he could transition from the tiller to the rudder. The captain saw the first officer's hand come off the tiller as he was making a final power adjustment and moments later he felt the aircraft jerk to the left. He saw the aircraft begin a rapidly increasing left turn and felt the first officer applying full right rudder. The first officer reported that he had rudder authority prior when he took his hand off the tiller, but subsequently lost directional control after feeling the jerk. The aircraft manufacturer recommends that the pilot keep his hand on the tiller until reaching 70 knots.

The captain took the controls and began to abort the takeoff. He brought both power levers back to flight idle and then into reverse as the aircraft continued veering left. He applied full right rudder and right brake. By the time the turn stopped, the aircraft was already headed off the left side of the runway. The captain was unable to steer the aircraft since the tiller is positioned to be reached from the left seat. He remained on the controls as the aircraft left main gear rolled off the left side of the runway. He felt another jolt and noticed the aircraft slowing as left gear encountered the freshly plowed, soft terrain next to the runway. The blades of the No. 1 propeller contacted the ground as the left gear sank into the soft dirt.

As the aircraft continued, it began an uncommanded right turn and rolled back onto the runway. The captain attempted to control the aircraft but it continued in a right turn. The captain then put both propellers into feather and the aircraft coasted to a stop about 5 feet from the right edge of the runway. At this point he reported he had no control over the aircraft steering or brakes. He shut down both engines and notified center that he was disabled on the runway.

The captain stated that he neither saw nor heard any cockpit indications of a problem. He reported that he thought he had lost braking when he first took control of the aircraft because the aircraft did not respond. He also thought he had lost directional control when he applied full right rudder with no apparent effect. The first officer said he did not attempt to use the tiller after relinquishing the controls to the captain. He reported that he did not experience any steering or braking problems during taxi, nor did he notice any control binding or feedback. The aircraft manufacturer reported that rudder effectiveness increases incrementally beginning about 40 knots airspeed with full authority achieved by approximately 70 knots.

PILOT INFORMATION

The captain and first officer were properly certificated in accordance with existing Federal Air Regulations (FARs). Both crewmembers reported that they were in good health at the time of the accident.

The captain had been hired by American Eagle on October 14, 1986. In addition to an ATP certificate he is type rated in the BA-3100, SA-227, and SF-340. He received his initial type rating in the BA-3100 on February 24, 1992. The captain's position with American Eagle is as a check airman and training pilot. His last proficiency check was accomplished on October 23, 1995.

The first officer was hired by American Eagle on November 1, 1987. In addition to his ATP certificate he is type rated in the SA-227 and SF-340. A type rating in the BA-3100 is not required for the performance of copilot duties. After retiring from American Eagle on May 15, 1995, the first officer continued his association with the company as a contract consultant. His last proficiency check was accomplished in January, 1995.

AIRCRAFT INFORMATION

The captain reported that he had flown the aircraft on two maintenance test flights in which the No. 2 engine had failed to meet the flight idle torque differential parameters. Trouble shooting revealed that the torque was not adjusted properly. An adjustment was made, but on the second flight the engine still was unable to achieve the proper torque. After maintenance personnel were unable to resolve the problem the No. 2 engine was changed on May 4, 1996.

The aircraft, which had been leased to American Eagle, had, according to maintenance personnel, completed a service check in accordance with the JS3200 Maintenance Schedule on June 5, 1996. The service check was performed as a condition of the lease and was required to be completed before ferrying the aircraft back to its corporate owner in Kingman, Arizona.

As a condition of acceptance, the hydraulic fluid used by American Eagle, MIL-H-83282 was replaced with MIL-H-5606 in accordance with service bulletin No. 29-JK 12078 revision No. 1.

AIRPORT INFORMATION

At the time of the accident a NOTAM was in effect for McChesney Field. The airport was scheduled to be closed from 2345 till 0530 on the morning of the accident for construction. The construction involved plowing the terrain on the left side of runway 11. (A copy of the NOTAM is appended to this report.)

FLIGHT RECORDERS

The flight data recorder (FDR) and cockpit voice recorder (CVR) were secured and removed for readout. The FDR was read by the Air Accidents Investigation Branch (AAIB) of the United Kingdom on July 17, 1996. The readout showed that the airspeed had peaked at 58 knots and the heading had varied from 95 to 170 degrees. (The FDR printout is appended to this report.)

The CVR tape was submitted to the National Transportation Safety Board for a sound spectrum analysis. The examination revealed that one of the engines was at 105-106 percent, while the other engine was at 99-101 percent. These values were obtained from a 2 second segment of the tape, 10 to 11 seconds into the takeoff roll and prior to the engines going into reverse.

The tape was also submitted to the AAIB. Their examination revealed the highest frequencies are those of the No. 2 engine and were calculated to represent 1,541.25 rpm (96.9 percent). The No. 1 engine was calculated to have been running at 1,692 rpm (106.3 percent). The AAIB concluded that the frequencies for the No. 1 engine correlated to expected engine rpm with the propeller in the start lock position.

WRECKAGE AND IMPACT INFORMATION

Scuff marks were found on the runway that corresponded dimensionally to the right and left main gear tires. The marks began approximately 300 feet east of the intersection of taxiway Alpha and runway 11. Additional marks corresponding dimensionally to the nose gear tires were found on the runway east of taxiway Charlie. The aircraft final position of the aircraft on runway 11 was approximately 1,100 feet east of taxiway Alpha on an approximate heading of 170 degrees magnetic. There was a flat spot visible on the outboard sidewall area of the left nosewheel tire. There was evidence of scuffing on the inboard side of the right main gear tires. The aircraft was equipped with an anti-skid brake system.

A postaccident inspection of the aircraft by FAA airworthiness inspectors found that the left wing and left side of the fuselage exhibited skin wrinkling and buckling. The left main landing gear assembly was bent aft about 0.5 inches.

The nose landing gear steering system was inspected for proper rigging and full range of travel. The system was moved both left and right and then back to center. An operational check of the nose landing gear on-wheels squat switch was performed. Maintenance personnel noted that several bypass buttons were found in the bypass position.

The brake system, both normal and emergency, and the anti-skid system, were functionally checked for proper operation with no discrepancies found.

Both engines were inspected for proper rigging and adjustment with no discrepancies found.

Both stop and feather levers were pulled. The No. 2 propeller was found in the feather position, but the No. 1 propeller was found in the start lock position. The No. 1 propeller control system rigging was checked. All four propeller blades were bent aft approximately 12 inches inboard from the blade tips. The leading edges of the bent sections were curled from camber toward the blade face. The blade shim packs were displaced from their retention grooves allowing axial movement of the blades. Some of the balance weight screws were sheared. The propeller spinner was missing from the hub.

Latch tools were inserted far enough to contact the latch weights of the No. 1 propeller. The engine was run up and the propeller brought into reverse. The latch tools were fully inserted moving the latch weights clear of the pawl. The propeller was then brought into feather in approximately the proper time interval. The latch tools were removed and the propeller was brought into reverse. With the latches positioned by their springs, the propeller was brought down to the latch position and the latches engaged.

TESTS AND RESEARCH

According to the terms of the lease, American Eagle was to replace the hydraulic fluid before turning the aircraft back to its owner. As part of the process, the hydraulic system was sampled by maintenance personnel who, in turn, submitted the sample to Aviation Laboratories for analysis. According to the results of the analysis, the sample submitted on April 16, 1996, contained particle counts in all size ranges exceeding the maximum allowed. The fluid viscosity was also found to be out of the normal range. Two additional samples taken from the aircraft after the accident showed similar content and failed to meet specifications.

The hydraulic mule used to service the aircraft was an AAR Western Skyways, S/N 5789. It was sampled on October 23, 1996, and failed to meet NAS specifications. According to the aircraft manufacturer, the results of all the analyses grossly exceed the particle size limitations prescribed by NAS 1638 for an in-service fluid sample. (Copies of the Aviation Laboratories reports are appended to this report.)

The steering selector valve, S/N LK9005272, steering jack, S/N SJA460, brake control valve, S/N PB371, and axle maxaret, S/N NV407, were returned to the manufacturer for teardown and analysis. The steering selector valve and steering jack were examined at APPH (Bolton) Ltd. The steering jack passed a production acceptance, but the steering selector valve was not tested because the input shaft was jammed. Both units were subsequently torn down and examined. With the caps removed the bobbin in the liner subassembly would not move when pushed by hand. The bobbin was found to be seized in the liner and when finally removed from the liner was found to be corroded. The input shaft subassembly, including the wobble plate and retaining nuts and bolts were also corroded. The output shaft subassembly rollers and bearings were pitted.

The steering jack piston and gland were free of corrosion but there was extensive corrosion in the main body. The manufacturer concluded that the most likely source of the corrosion was introduced through the hydraulic system. (A copy of the APPH (Bolton) Ltd. report is appended to this report.)

A representative of the aircraft manufacturer stated that the hydraulic actuator in the nose gear steering unit has not experienced any documented hydraulic hardovers. Previously, another BA 3100/3201 aircraft experienced two separate runway excursions during the takeoff roll. In both instances the same aircraft departed the left side of the runway but was brought under control by the crew and avoided damage. The flying pilot described the event as a "hydraulic hardover." The corrective action after both events was the intuitive replacement of steering unit components. During the second repair attempt, maintenance personnel discovered that one of the two hydraulic filters was missing. After the second attempted repair the problem did not reoccur.

The brake control valve and axle maxaret were examined at Dunlop Limited Aviation Division. The brake control valve was subjected to acceptance tests in which it was found that the No. 2 brake was not properly set or synchronized. Seepage was observed from the No. 2 brake port, which indicates a possible leak in the inlet valve. Further leak checks showed the unit was satisfactory with an exhaust leak rate that was within the specified limitations. The fluid bled from the brake control valve was subjected to a particle count and found to meet NAS class 8 standards.

The axle maxaret was subjected to acceptance tests in which it was found the valve block was properly adjusted and operation was satisfactory. However, the unit exhibited an excessive hydraulic fluid exhaust leak of 2 drops per second against the standard of 5 drops per minute. The unit was disassembled and excessive leakage from the valve seat was confirmed. (A copy of the Dunlop Limited Aviation Division report is appended to this report.)

The No. 1 propeller, a Mc Cauley model 4HFR34C653-FH/L106FA-0, S/N 901350, was examined at the Mc Cauley Accessory Division. The propeller pitch change mechanism had originally been found in the start lock position by Golden State Propeller, Inc. However, it was not possible to determine mechanically the operating position of the propeller at the time of the first ground contact. (A copy of the Mc Cauley report is appended to this report.)

ADDITIONAL INFORMATION

The aircraft was towed from the runway by maintenance personnel from American Eagle a few minutes after the accident. The aircraft was released to a representative of the registered owner on May 5, 1997.

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