N491PH
Bell 407
The helicopter, which had a FADEC controlled turboshaft engine installed, was in cruise flight (about 800 feet AGL) over open ocean water when the FADEC FAIL aural warning sounded, followed closely by sound of the LOW ROTOR RPM horn. Simultaneously, the LOW ROTOR RPM, FADEC FAIL, and FADEC FAULT cockpit caution lights illuminated. The 8,300-hour helicopter pilot attempted to regain the RPM's with no result. The FADEC AUTO/MANUAL indicator light/button showed the engine control mode to be in the "AUTO" condition. The pilot recalls that the Ng was approximately 89%. The pilot stated that about 10 seconds elapsed from the onset of the event to cross checking the Ng. The pilot then depressed the AUTO/MANUAL button and switched to the MANUAL mode. He then increased the throttle above the 90% detent to try to regain rotor RPM's. He recalled that the light displayed "MANUAL", and that the FADEC FAIL aural warning ceased after the button was depressed. While descending, on three separate occasions, the pilot attempted to increase the throttle which were accompanied by three uncommanded right yaws, approximately 1-2 seconds apart. During the third uncommanded yaw, the ENGINE OUT audio sounded and the ENGINE OUT light illuminated (these occur when Ng drops below 55%). The pilot then entered a full autorotation, deployed the skid mounted emergency float system and landed upright on the water. Metallurgical examination of the 1st thru 4th stage turbine wheels and 1st thru 3rd stage turbine nozzles, revealed that all associated damage was due to extreme over-temperature operation. The manual mode schedule of the Hydro-mechanical Unit (HMU) was found within limits, and the auto mode schedule had a flow shift of +10 to +15 PPH. Evaluation and testing of the Electronic Control Unit (ECU) and its sub-components revealed a shorted condition on the ECU -15V power supply. Disassembly of the ECU and tests of the Interface (IF) and Power (PWR) circuit boards revealed that the C321 capacitor (p/n CDR33BX104AKUR) on the IF board was found thermally distressed and was measured at .58 ohms. The C321 is a high frequency bypass capacitor from -15V to ground. The PWR board was also visually inspected and the CR423 diode was found thermally stressed. According to the manufacturer, the CR423 diode provides rectification on the -15V power supply and was likely stressed as a result of the shorted C321 capacitor. The C321 capacitor was removed from the IF board and there was no longer a short on the -15V power supply. The ECU was re-assembled, operated, and passed a functional acceptance test after removal of the C321 capacitor. The shorted condition of the C321 capacitor forced the -15V power supply to a low voltage condition and a significant current draw, resulting in the rectifying diode to overheat. According to the manufacturer, the -15V power failure of the ECU resulted in the HMU to revert to manual fuel metering. In the manual mode, the engine would have overspeed protection, but not overtemperature protection.
HISTORY OF FLIGHT On May 11, 2003 approximately 1533 central daylight time, a Bell 407 helicopter, N491PH, registered to and operated by Petroleum Helicopters Inc., of Lafayette, Louisiana, was substantially damaged during a forced autorotative landing into the water in the Gulf of Mexico, following a loss of engine power while in cruise flight. The airline transport pilot and his 3 revenue passengers were not injured. Visual meteorological conditions prevailed and a company flight plan was filed for the Title 14 Code of Federal Regulations Part 135 on-demand air taxi flight. The flight originated from an offshore platform, EI-380, at 1453 and was en route to Morgan City, Louisiana. During an interview with the NTSB investigator-in-charge, the pilot reported that the helicopter was in cruise flight (about 800 feet AGL) when the FADEC FAIL aural warning sounded, followed closely by the LOW ROTOR RPM horn. Simultaneously, the LOW ROTOR RPM, FADEC FAIL, and FADEC FAULT cockpit caution lights illuminated. As the rotor RPM began to decay thru 90%, the pilot attempted to regain the RPM's by lowering the collective with little result. The FADEC AUTO/MANUAL indicator light/button showed the engine control mode to be in the "AUTO" condition. The pilot recalls that the Ng was approximately 89%. The pilot stated that about 10 seconds elapsed from the onset of the event to cross checking the Ng. The pilot then depressed the AUTO/MANUAL button and switched to the MANUAL mode. He then increased the throttle above the 90% detent to try to regain rotor RPM's. He recalled that the light displayed "MANUAL", and that the FADEC FAIL aural warning ceased after the button was depressed. While descending, on three separate occasions, the pilot attempted to increase the throttle which were accompanied by three uncommanded right yaws, approximately 1-2 seconds apart. During the third uncommanded yaw, the ENGINE OUT audio sounded and the ENGINE OUT light illuminated (these occur when Ng drops below 55%). The pilot then entered a full autorotation as the helicopter was passing thru 400 feet AGL. The skid mounted emergency float system was deployed and the helicopter landed upright on the water. The pilot recalled that the engine was running after landing. He shut the engine down, and assisted with the egress of his passengers. PERSONNEL INFORMATION The 8,300-hour airline transport rated pilot held ratings for rotorcraft and instrument helicopter. As of the date of the accident, he had accumulated a total of 8,270 hours, all of which were flown in helicopters. He had 1,777 hours of total time in the Bell 407, of which, 158 hours were flown in the last 90 days. The pilot was issued a valid second-class medical certificate on February 5, 2003. AIRCRAFT INFORMATION The Bell model 407 helicopter, s/n 53398, had a total airframe time of 3,360.1 hours as of the accident date. The helicopter was being maintained on an FAA Approved Annual Inspection Program (AAIP). According to maintenance records provided by the operator, the last 300-hour inspection was completed on March 16, 2003. The last 100-hour inspection was completed on May 6, 2003, at 3,340.5 hours. The helicopter was powered by a Rolls Royce 250-C47B turboshaft engine, p/n 23063392, s/n CAE 847082. A review of entries in available maintenance records indicated the total engine time was 5,254.1 hours. The Hydro mechanical Unit (HMU), p/n 23072725, s/n JGALM0315, had 3,572.5 total hours since new and 1,473.1 hours since overhaul. The Electronic Control Unit (ECU), s/n JG8ALK0471, was originally manufactured as a p/n 113500-6B2-5100 in March, 1998. The unit was returned to the manufacturer in March, 2000, after 1,265 hours of operation, to update the configuration (Direct Reversion to Manual). After the configuration update, the ECU's p/n changed to 114395-1A1-5102. There were no records found of previous repair activity on the unit. WRECKAGE AND IMPACT INFORMATION After recovery of the helicopter to the operator's facility, it was noted that the aft section of the tail boom showed evidence of main rotor blade contact. The top of the left vertical fin was severed, the top portion of the tail rotor drive shaft cover was cut and the #5 segment of tail rotor drive shaft was severed. All of the observed damage was consistent with main rotor blade contact during landing. All rigid and flexible oil and fuel lines were checked and found intact, and all fittings were found tight and secured. The engine's HMU and ECU were both intact and secured. Fuel was present at the fuel nozzle delivery line (approximately 2 table spoons). The magnetic particle indicating plugs were found clean and absent of metal particles. Visually, the engine and all components appeared intact and not damaged, other than some thermal signatures on the outside of the cases. The engine was transported to the engine manufacturer's facility in Indianapolis, Indiana, for further examination and teardown. During examination of the helicopter, power was applied and the following cockpit caution lights illuminated; FLOAT ARM, L/FUEL XFER, R/FUEL BOOST, L/FUEL BOOST, AUTO RELIGHT, FADEC FAIL, GEN FAIL, XMSN OIL, CHECK INST, HYD SYSTEM, CYCLIC CENTER, and RPM. After approximately three minutes, the FADEC MAN, and ENG OUT lights illuminated. The FADEC button displayed "AUTO", and after 3 minutes changed to "MANUAL." The collective was noted to be about 1/3 of the way up, the throttle twist grip between 80% NG and Idle, the fuel valve was closed, the fuel gauge read 420 lbs, the voltmeter read 23.8 volts, and the MGT Litton gauge showed an exceedance. Typically, Litton gauge exceedances that are recorded during engine operations are: 727.1 to 779 degrees recorded after 30 seconds; 779.1 to 905 recorded after 12 seconds; and 905.1 and above recorded immediately. Five MGT exceedances were recorded and downloaded. Chronologically, 80 seconds, the recorded MGT exceedance peaks were 981 degrees, 998 degrees, 1,084 degrees, 1,102 degrees, and 800 degrees respectively. Thermal damage was noted on the top of the engine cowl. The engine N1 and N2 appeared to be locked when hand rotation was attempted. The #1 short shaft was found sheared. All cannon plugs on the ECU were found connected and tight, and the unit appeared undamaged. Cannon plug J1 and J2 were disconnected and a slight amount of corrosion was noted on pin #79 on plug J2. A direct ECU download could not be accomplished, as it appeared that the ECU was not communicating properly with the EMC-35A download software. After the attempted file download, the NTSB IIC determined that the ECU and HMU should be removed and transported for further examination/evaluation at the ECU manufacturer’s facility. TESTS AND RESEARCH Under the supervision of the NTSB IIC, the engine was examined and disassembled at Rolls Royce, Indianapolis, Indiana, on May 29, 2003, with the following findings: Both N1 and N2 would not rotate. After the turbine was separated from the gearbox, the gearbox turned freely. Approximately 60% of the first stage turbine wheel blade airfoils were missing, and about 50% of the second stage turbine wheel blade airfoils were missing. Impact damage and some metal splatter was observed on the fourth stage wheel, and the third stage shroud was missing. The first stage nozzle was in good condition, the second stage nozzle was destroyed, and the third stage nozzle had minor damage (third stage blade track area had some debris impact damage). Impact damage from debris was noted in the fourth stage nozzle and fourth stage blade track area. The compressor shroud had minor impact damage. Shafting and bearings throughout the engine were found intact. All four thermocouples were found with their respective tips burned. Metallurgical examination of the 1st thru 4th stage turbine wheels and 1st thru 3rd stage turbine nozzles, revealed that all associated damage was due to extreme over-temperature operation. A functional acceptance test of the HMU was performed and completed at the manufacturer’s facility in West Hartford, Connecticut. The manual mode schedule was found within limits, and the auto mode schedule had a flow shift of +10 to +15 PPH. According to the manufacturer, these findings indicated a distressed gear roll pin, and examination of the pin revealed that it was slightly bent. The load piston stop was found out of limits at 0.577 volts (limit 0.69 +/- 0.02 volts), which also is consistent with a bent roll pin. All other functional acceptance test were found within limits. Evaluation and testing of the ECU and its sub-components were also performed in West Hartford, Connecticut. Visual inspection of the unit revealed the J1 and J2 connectors had signs of corrosion at the base of several I/O pins, and 2 of the pins (39 ARINC Xmit1 & 79 not used) of the J2 connector had signs of a white residue. The manufacturer stated that the J1/J2 connector pin condition was not believed to have contributed to any ECU malfunction. The ECU was then installed on Systems Engineering Closed Loop Test Equipment (CLTE), and a CLTE test HMU was used for the tests. Earlier field observations (attempted download at the operator's facility) that the ECU was not communicating was confirmed with the CLTE computer. When the ECU cover was removed, an odor similar to thermally stressed components was noted. Loading was measured in ohms at 460, 116K, and .853 for the 5V, 15V, and -15V supplies respectively, indicating a shorted condition on the -15V power supply. A decision was made to attempt to download/retrieve Incident Recorder (IR) data. Subsequently, the ECU computer (CPU) board was removed for installation in a test ECU. The CPU board power supply inputs were checked for shorts in ohms, and no shorts were detected. The CPU board was then installed in the test ECU for EEPROM download of IR data. The IR data was found to be clear. No faults or incidents were found in non-volatile memory which indicated that the power supply damage likely prevented and data from being recorded. The ECU was disassembled to isolate the source of the -15V power supply short. When the Interface (IF) and Power (PWR) boards were removed, the C321 capacitor (p/n CDR33BX104AKUR) on the IF board was found to be thermally distressed and was measured at .58 ohms. The C321 is a high frequency bypass capacitor from -15V to ground. The power board was also visually inspected and the CR423 diode was found thermally stressed. According to the manufacturer, the CR423 diode provides rectification on the -15V power supply and was likely stressed as a result of the shorted C321 capacitor. The C321 capacitor was removed from the IF board and there was no longer a short on the -15V power supply. The .5 ohm short continued to be present in the C321 capacitor after it was removed. The ECU was re-assembled, operated, and passed a functional acceptance test after removal of the C321 capacitor. The data collected during the fault isolation indicated that a short circuit of the C321 capacitor on the IF board occurred. The shorted condition of the C321 capacitor forced the -15V power supply to a low voltage condition and a significant current draw, resulting in the rectifying diode to overheat. According to the manufacturer, the -15V power failure of the ECU resulted in the HMU to revert to manual fuel metering. In the manual mode, the engine would have overspeed protection, but not overtemperature protection. ADDITIONAL INFORMATION The helicopter was released to the owner's representative.
The short circuit of the C321capacitor in the Electronic Control Unit (ECU) that resulted in a single-point failure of the ECU's -15V power supply which disengaged/reverted the Hydro-mechanical Unit (HMU) from automatic to manual fuel control. Factors contributing to the accident were the pilot's attempted remedial actions in the manual mode that resulted in the engine over temperature and loss of power, and the lack of suitable terrain for the forced landing.
Source: NTSB Aviation Accident Database
Aviation Accidents App
In-Depth Access to Aviation Accident Reports
