SAN LUIS OBISPO, CA, USA
N313AE
Saab-Scania AB (Saab) 340B
The flight crew reported that all pretakeoff checks were normal and that prior to setting takeoff power the Constant Thrust on Takeoff (CTOT) computer was set to 100-percent engine torque for takeoff. The ensuing takeoff was normal through liftoff and gear retraction; however, during initial climb the left engine experienced compressor stalling and lost power. As the crewmembers were executing the engine failure checklist, the right engine first increased power in Automatic Power Reserve (APR) and then began to similarly stall and lose power. After about 20 seconds the engines recovered sufficient power to enable the flight crew to remain airborne and return for landing on the departure runway. Review of the digital flight data recorder (DFDR) output and internal examination of each engine revealed that the engines had compressor stalled due to accumulated dirt and contaminants on the compressors. A review of the right engine DFDR data revealed that, while the left engine was stalling, the flight crew had retarded the right engine power lever past the 64-degree power lever angle position which tripped off the CTOT system. The resulting transient in fuel flow, in company with an atmospheric temperature inversion, reduced the stall margin of the right engine.
HISTORY OF FLIGHT On May 13, 1997, at 0605 hours Pacific daylight time, Wings West Airlines, Inc. Flight 3319, a Saab 340B, N313AE, experienced a temporary loss of power on both engines while at 200 feet agl on departure climbout from McChesney Field, San Luis Obisbo, California. The aircraft was not damaged and the 2 cockpit crew, the 1 cabin crewmember, and 18 passengers were not injured. The scheduled, domestic passenger flight was operated under 14 CFR Part 121 by Wings West Airlines, Inc., and was departing San Luis Obispo destined for Los Angeles, California. The flight was operating on an instrument flight rules flight plan and visual meteorological conditions prevailed. The flight crew reported the incident as compressor stalls in both engines. The flight crew stated that the flight was routine through takeoff and initial climb. The takeoff was made at 100 percent torque with Constant Torque on Takeoff (CTOT) on and Automatic Power Reserve (APR) armed. During the initial climb, approximately 200 feet agl, there was a "loud hammering noise," accompanied by loss of torque on the left engine. The first officer (the pilot flying) initiated the engine failure memory items and by the time the crew had identified the left engine, the right engine began to similarly lose power. At this point the rate of climb was zero, both engines were "erratic," and the left and right over temperature lights were flashing. The crew was preparing for an off-airport landing when the right engine resumed producing (more) power. The flight crew left the power levers in a reduced power position and was able to return for landing on the departure runway. A passenger reported to the flight attendant that flames were seen coming from the exhaust of both engines and this information was passed on to the cockpit crew. Postflight reduction of stored engine computer data confirmed that the left engine had a power loss and exceeded ITT temperature limits for 7 seconds. The right engine went into APR mode at 104.4 percent torque, and then it had a power loss and exceeded ITT temperature limits for 17 seconds. A Powerplants Group was formed to examine powerplant issues related to the incident. Their report is attached. PERSONNEL INFORMATION The captain and first officer were on the third day of a 3-day schedule flying together. They had remained overnight in San Luis Obispo and this was the first flight of the day for them. The first officer was the flying pilot. The captain upgraded to the Saab 340 in January, 1997, and had accumulated 170 hours in the type aircraft. He had 7,500 hours total time and for the previous 5 years was a BAE Jetstream captain for Wings West. The first officer was hired by Wings West in January, 1995, and had been a Saab 340 first officer the entire time. He had 3,300 hours total time and 1,100 hours in the Saab 340. He is not type rated in the Saab. Flight Crew Interviews The flight crew was interviewed separately at the Safety Board's regional office in Gardena, California, on June 19, 1997. Notes from the interview are attached to the Powerplants Group report. According to the captain, the crew went to the aircraft at 0530. The first officer did the external inspection and the captain did the cockpit checks. The dispatch release was reviewed and the maintenance log checked where they found that the left engine was on derivative power. The passengers boarded at 0545, the crew received their ATC clearance and, winds being calm, taxied to runway 29 for departure. The tower was still closed. They ran the first flight of the day checklists and everything checked normal. Visual meteorological conditions prevailed but there was low fog west and northwest of the airport and low fog at the departure end of the runway. The crew completed the checklist and the first officer determined that a 100 percent torque takeoff was required. On the runway the captain advanced the power levers to 80-85 percent, actually 82 percent, and called "set power." The CTOT was set to "APR" by the first officer and power came up to 100 percent. There was nothing unusual. At 80 knots the first officer took the flight controls and everything was routine through V1 (takeoff decision speed) and Vr (takeoff rotation speed). The first officer called "positive rate, gear up" and the captain retracted the landing gear. Seconds later they heard banging and felt shaking on the left side of the aircraft that felt and sounded like "a washing machine out of balance." The captain said that he had never experienced anything like this before. They didn't know what was wrong but saw the torque on the left engine drop to 20-40 percent. The first officer called "confirm left engine failed and positive autocoursen" which the captain confirmed. The first officer started the engine fail checklist memory items. The captain identified the left power lever and the first officer confirmed the captain's selection. The captain pulled the left power lever to flight idle and identified the left condition lever in preparation for feather of the propeller. As he was pointing to the left condition lever, the right engine started the same banging and shaking. Both sides were the same and both torque gauges fell to 20-40 percent. The captain told the first officer to fly the aircraft. They were over the fog bank now and the first officer identified a clear spot straight ahead in anticipation of an off-airport landing. According to the captain, he didn't know what was happening and so he moved the right and left power levers 2-3 inches with no effect. The over temperature lights were flashing. The captain turned the CTOT off for 1 or 2 seconds, which had no effect so he turned it back on again. Eventually the power came back on one engine and the first officer called "straight ahead or turn." The captain said turn. More power came back in the turn and both engines were delivering partial power on downwind leg. The crew ran the checklists and made a normal landing. In response to a question about which crewmember operated which controls during the flight, the captain said that at the start of the takeoff roll his left hand was on the tiller and his right hand was on the power levers. The first officer, the pilot flying, usually would have had his right hand on the yoke and his left hand near the CTOT switch. The captain brought the power up to 80-plus (80/81) percent power and called "set power." The first officer, with his left hand, moved the CTOT to "APR" position with his right hand on the yoke. At 80 knots the captain called "your aircraft." At this point the captain's left hand was idle and his right hand was on the power levers. The captain said that he did not rest his hand heavily on the levers but rather applied a light pressure. The first officer's right hand was on the yoke and the left may have been backing up the power levers. The captain next called V1 and removed his right hand from the power levers leaving both hands idle. The first officer had both hands on the yoke. The captain called Vr and the first officer then called "positive rate, gear up." The captain raised the gear using his right hand. When the banging started the first officer had both hands on the yoke and the captain had both hands in his lap. The first officer called "confirm left engine fail and positive autocoursen." The captain's hands were in his lap. The first officer called "identify left power lever." The captain placed the index finger of his right hand on the left power lever, and, after confirmation by the first officer, the first officer called "flight idle" and the captain, using his index finger only, pulled the left power lever to flight idle. The first officer called "identify left condition lever" and the captain placed his right hand index finger on the left condition lever. The captain never moved the condition lever because it was at this time that the right engine started banging and shaking. The captain interrupted the engine checklist and started looking for other things. He moved the right power lever 2-3 inches through the midrange, then the left power lever with no response. The first officer, in his interview, described this motion as two-handed, the left engine power lever in the captain's left thumb and forefinger, and the right engine power lever in the captain's right-hand thumb and forefinger. The captain then turned the CTOT off for 1-2 seconds then back on (once). He didn't recall in what position the captain left the power levers. AIRCRAFT INFORMATION The incident aircraft was manufactured in 1992 and had accumulated 9,544 hours total time in service. It was maintained under a continuous maintenance program. The left engine had 7,251 hours time in service and the right engine had 10,150 hours. The aircraft is maintained under a continuous airworthiness maintenance plan. As part of the plan, the engine compressors are washed at 200-hour intervals using a water solution. On the day of the incident the aircraft was operated with the left engine on the operator's approved "derivative power" program. This was the only discrepancy on the aircraft. The derivative power program is utilized for service life extension on engines which do not achieve rated power due to the effects of in-service wear. An engine on derivative power is operated at an adjusted (higher) interstage turbine temperature to achieve rated power using temperature margin built into the flat rating of the engine. According to the manufacturer, the purpose of the CTOT system is to reduce flight crew workload caused by "torque bloom" during the takeoff roll. Without CTOT, if the flight crew set rated torque for takeoff, the crew must frequently retard the power levers during the takeoff roll as the torque "blooms" (increases) due to the effects of ram recovery as the aircraft accelerates. Using CTOT, the crew set the desired takeoff torque value on the CTOT controller. At the start of the takeoff, the flight crew then set the engine power levers to an arbitrary lower torque value and enables the CTOT system. The CTOT system adds an additional amount of (trim) fuel to the engine to achieve the desired takeoff torque; however, as the aircraft accelerates and torque increases (blooms), the amount of trim fuel provided by the CTOT system is pared back to maintain the desired, preset torque. In the event of an aborted takeoff, it is necessary that the CTOT system be disabled. A microswitch is installed in the throttle quadrant for this purpose which disables CTOT when the power levers are retarded aft past the 64-degree power lever angle position. According to the operator, in normal flight, inadvertent movement of the power lever aft past the 64-degree switch position will truncate the supplemental CTOT fuel flow and is known to induce compressor stalling. For this reason, the operator's crew training program teaches pilots not to retard power levers during an engine failure after takeoff until above 400 feet agl. METEOROLOGICAL INFORMATION The San Luis Obispo (KSBP) weather observation for 0554, which was 11 minutes prior to the time of the incident, indicated calm winds, 15 miles visibility, and scattered clouds at 25,000 feet. The surface temperature was 9 degrees Centigrade and the dew point was 7 degrees Centigrade. The San Luis Obispo field elevation is 209 feet above mean sea level (msl). The closest weather station making a weather sounding was Vandenberg Air Force Base, approximately 35 nautical miles south of San Luis Obispo. The upper air sounding, executed approximately 2 hours before the incident, showed a surface (367 feet msl) temperature of 9.4 degrees Centigrade and, approximately 2,000 feet msl, a temperature of 24 degrees Centigrade. Weather data is attached to the Powerplants Group report.. FLIGHT RECORDERS The aircraft was equipped with a cockpit voice recorder and an attempt was made to replay the recorder at the Safety Board laboratory; however, the recorder had been permitted to continue to operate during postincident maintenance checks and there was no usable data. The aircraft was also equipped with a Fairchild F-800 digital flight data recorder (DFDR) which was read out at the facilities of Simmons Airlines, another American Eagle carrier. The DFDR samples and records 32 parameters at 1-second intervals. It was determined subsequent to the incident that one parameter, left-hand power lever angle, was not properly recorded due to a broken wire termination at the transducer in the power lever quadrant. The DFDR has time synchronization to UTC standard. The DFDR engine data is presented in the Powerplants Group report. The DFDR provided engine data of the incident for both engines for the period from 8 seconds before the engines lost power until after the engines recovered. The engine parameters recorded were Ng, Np, torque, fuel flow, interstage turbine temperature (ITT), and P3. The following is a sequential description of events recorded on the DFDR. The first torque readings were 98.4 and 98.1 percent for the left and right engines, respectively. The right engine power lever angle (PLA) was set at 67.0 degrees. Beginning at 3 seconds into the recording, the left engine torque dropped over a period of 5 seconds from 100 percent to approximately 94 percent. At 8 seconds into the recording, the left engine torque, P3, and Np all decreased over a 3-second time interval when the airplane was at 587 feet msl. The left engine torque reading went from 94.1 percent to 0 percent, P3 decreased from 225 psi to 40 psi, and Np decreased from 1,400 rpm to 220 rpm. During the time that the left engine parameters were decreasing, between 8 and 10 seconds, the right engine data indicated an increase in torque from 98 percent to 103.7 percent, and a P3 increase from 230 to 240 psi, with no concurrent change in the right engine PLA position. Two seconds after the left engine torque dropped, the right engine PLA changed from 66.6 to 61.1 degrees. One second after the PLA change, the right engine torque dropped to 20 percent and cycled between 20 and 60 percent. Concurrently, Np cycled between approximately 1,300 and 1,400 rpm three times. During this period, the recorded outside air temperature increased from 9.5 to 17.6 degrees Centigrade. Twenty seconds after the left engine torque dropped, the left engine's Np increased from approximately 300 to 1,380 rpm. At 25 seconds after the left engine torque dropped, the left engine was approximately 60 percent torque and had a P3 pressure of 170 psi. At 30 seconds after the left engine torque dropped, the right engine torque recovered and stabilized within 6 seconds and approximately 60 percent. TESTS AND RESEARCH Under the purview of the Powerplants Group, both engines were disassembled and examined at the General Electric overhaul facility in Strother, Kansas. The engine hydromechanical units (HMU) were tested and disassembled for examination at Woodward Governor Company, Rockford, Illinois. The Constant Thrust on Takeoff (CTOT) units were tested at Wings West Airlines, San Luis Obispo, California. Results of these examinations are in the Powerplants Group report. The engine inspections revealed that both engines' axial compressors had dirt and corrosion on the inlet and blade surfaces. The right engine hydromechanical unit (HMU) was found to have internal wear that would have reduced its engine's calculated stall margin by approximately 1 percent. An acceptance test procedure (ATP) was performed on the aircraft throttle quadrant at Adams Rite Sabre International, Glendale, California. An ATP was performed on the CTOT control panel at SAAB Aircraft in Linkoping, Sweden, and the Autocoarsen Computer was functionally tested at FFV Aerospace in Nashville, Tennessee. These units tested within specifications for in-service units. The test results are attached. ADDITIONAL INFORMATION Additional persons (parties) participating in this investigation are: Mr. Pierre Huggins, Airline Pilot's Association, Intl. Washington, D.C. 20036.
Compressor stalls of both engines due to in-service compressor contamination. Factors in the compressor stall of the right engine were an atmospheric temperature inversion and a transient fuel flow condition produced when the constant thrust on takeoff (CTOT) system was disabled by the pilot's inadvertent movement of the right engine power lever during execution of the emergency procedure.
Source: NTSB Aviation Accident Database
Aviation Accidents App
In-Depth Access to Aviation Accident Reports
