Aviation Accident Summaries

Aviation Accident Summary FTW03FA118

Broadus, TX, USA

Aircraft #1

N175PA

Bell 407

Analysis

While establishing a 125-foot out of ground effect hover during a search and recovery mission of Columbia Space Shuttle debris, the helicopter lost power without warning, descended rapidly into a forest with 80-foot tall trees, and impacted the ground. There were no reported radio communications or distress calls. During post accident component examinations, while performing the rigging procedure for the Hydro-Mechanical Unit (HMU) during an engine test cell run, anomalies were noted with the actual position of the throttle Position Lever Angle (PLA) and the readings obtained from the Electronic Control Unit (ECU). During the test cell run, the Full Authority Digital Electronic Control System (FADEC) controlled engine operated erratically in the Auto mode when rigged with its original Electronic Control Unit (ECU) and (HMU). The engine control was then changed from Auto to Manual mode, and the engine responded to throttle input as required; however, the ECU readings were erratic. After a slave ECU was installed, the engine operated with similar erratic readings. A slave HMU was installed, and the engine operated normally without erratic readings. The ECU's nonvolatile memory was downloaded, and revealed no faults on the accident flight. Extensive electrical and mechanical testing of the HMU revealed severe signal fallout on the PLA signal, which was found to be random in location, bi-directional, and present during both rotational and stationary operation of the PLA input of the HMU. The source of the fallout was traced to the HMU potentiometer. The PLA potentiometer examinations consisted of mechanical measurements, electrical testing, inspection of the conductive epoxy joints, and microscopic examination of the three lead wire connections to the potentiometer elements. The PLA potentiometer was found faulty due to insulation breakdown between the rotor and shaft, resulting in a single-point failure that induced erratic fuel metering to the engine. As a result of the findings from this investigation, the Safety Board issued four safety recommendations (A-03-18 through A-03-21), on May 27, 2003, to the FAA that addressed the PLA potentiometer deficiencies.

Factual Information

HISTORY OF FLIGHT On March 27, 2003, at 1636 central standard time, a Bell 407 helicopter, N175PA, registered to I Inc., of Kirkland, Washington, and operated as a Public Use aircraft under contract to the US Forest Service (USFS), was destroyed when it crashed into heavily wooded terrain near Broadus, Texas, while conducting low level flight operations in support of the Federal Emergency Management Agency's (FEMA) mission to support an inter-agency (NASA, Texas Forest Service, USFS) search/recovery effort of Columbia Space Shuttle debris. The pilot and 1 crewmember were fatally injured and 3 other crewmembers sustained serious injuries. Visual meteorological conditions prevailed and a VFR flight following plan was filed for the Title 14 Code of Federal Regulations Part 91 Public Use flight. The flight originated at 1515 from the search operation's Helibase, located at the Angelina County Airport, 7 miles south of Lufkin, Texas. The helicopter was assigned to the search mission in the Angelina National Forest on March 23, 2003. On March 25th, the 5-person crew was assigned, and flights took place on that day. On March 26th, the helicopter was out of service for scheduled maintenance. On the morning of March 27th, the pilot performed a routine post maintenance evaluation flight of 30-minutes duration. After the successful evaluation flight, the helicopter and crew were dispatched to commence their assigned search grid, refueled, and returned to the grid to complete the day's mission. After the last fly over of the grid, the helicopter made a right turn and was in the process of establishing a high hover approximately 125 feet above the ground to establish GPS coordinates and verify completion of the search grid. During post accident interviews, the three surviving occupants reported that while establishing the 125-foot hover, the helicopter lost power without warning, descended rapidly into 80-foot tall trees, and impacted the ground. There were no reported radio communications or distress calls. PERSONNEL INFORMATION According to the USFS, the pilot held a valid FAA medical certificate and was qualified and current for the type of mission flown. He had over 8,200 total flight hours in helicopters, with over 150 hours in the Bell 407. The pilot's initial ground and flight training in the 407 was conducted at the Bell Textron training facility in November, 1997. He was designated an FAA check airman in the 407 on February 20, 2002, and his last FAA competency/proficiency check flight in the 407 was completed on May 23, 2002. In addition to FAA airmen proficiency requirements, the Office of Aircraft Services (OAS) and USFS requires all of its contract pilots to meet flight standards for the type operations in which they act as pilot-in-command. This requirement was met by the issuance of a USFS qualification/certification "card." The accident pilot's USFS certification card was issued on February 21, 2003, after the pilot's successful completion of ground and flight checks by a USFS check airmen. The pilot was well known within the USFS and his documented performance on previous assignments was positive. A review of the pilot's 72-hour history revealed no unusual circumstances or medical issues. Also, all USFS agency requirements for flight and duty time were met. AIRCRAFT INFORMATION The helicopter, serial number 53154, was manufactured on June 3, 1997, and delivered to Papillon Helicopters on June 5, 1997. At the time off the accident, the helicopter was in a standard passenger configuration; however, the rear cabin doors had been removed to facilitate the search and recovery mission. The operator reported a total airframe time of 3,339.5 hours at the time of the accident, and the most recent 100-hour inspection was completed on March 26, 2003 ( completed on the morning of the accident). Review of the helicopter's service records revealed that it was certified for USFS use on December 11, 2002. Maintenance records reviews did not reveal anomalies or uncorrected defects. The helicopter was within weight and balance requirements for the flight, and there were no known or reported performance issues. The helicopter was equipped with a Rolls Royce Model 250-C47B series IV engine (S/N CAE847172), which incorporates a Full Authority Digital Electronic Control System (FADEC) that electronically controls engine fuel flow via a Hydro-mechanical Unit (HMU), and Electronic Control Unit (ECU). The FADEC can operate in automatic or manual mode. By default, the FADEC operates in automatic mode, using the collective twist-grip throttle position angle (PLA), among other inputs to determine the engine fuel metering requirements necessary to maintain efficient engine operation. PLA is transmitted via a mechanical link between the cockpit and a shaft contained within the HMU. A potentiometer in the HMU detects the PLA and generates a signal that is sent to the ECU, which processes the PLA signal and other engine information and generates a signal that is sent to the HMU to control fuel flow. When operated in the manual mode, the ECU deactivates the auto/manual solenoid on the HMU, eliminating electromechanical metering of the fuel. The fuel flow is then hydro-mechanically controlled via the throttle position and the HMU. The FADEC system automatically switches from automatic to manual mode if it detects certain faults. Otherwise, the pilot may chose to switch to manual mode by pushing a FADEC mode button located on the forward instrument panel. In either case, the full effect of the transition to manual mode may take 1.6 to 5.8 seconds, depending on the altitude and power setting of the engine in the Bell 407. Once the manual mode is commanded, the fuel flow is reduced to 282pph. Assuming no other changes, this would reduce the rotor rpm. After the transition period, the fuel flow is controlled by the pilot. The ECU ((P/N 114395-2A1-5202, S/N JG6ALK0147) was manufactured by Goodrich Pump and Engine Controls (GPECS), located in West Hartford, Connecticut. The function of the ECU is to control, and monitor the engine while maintaining rotor rpm. Two electrical connectors protrude from the front face of the ECU. The first connector, J1, receives and outputs signals to the engine and HMU via an engine wire harness. The second connector, J2, receives and outputs signals to and from the aircraft via an airframe wire harness. The ECU contains an engine monitoring system to record and store engine and system fault information in nonvolatile memory. The HMU was manufactured by GPECS. The accident helicopter's HMU (P/N 114070-03A5, S/N JGALM0224) had accumulated a total of 3,187.4 hours since new, and 1,461.2 hours since its last overhaul on August 21, 2001. The HMU consists of a gear-box mounted duel-element fuel pump, fuel metering valve, manual fuel control, three solenoid valves, a power lever input shaft, and two feedback potentiometers. The HMU provides/receives signals to/from the ECU and controls fuel flow in either automatic or manual modes. WRECKAGE AND IMPACT INFORMATION Accident Site Examination: The helicopter wreckage was located in a heavily wooded marsh in the Angelina National Forest, east of Lufkin, Texas. The helicopter was found on its right side at the base of an 80-foot tree on a heading of 078 degrees magnetic. First responders to the accident site found the helicopter with the cockpit section crushed against the tree, with the main cabin mostly intact and the tail boom separated. Several pieces of the main rotor blades were observed scattered on the ground and in shallow standing water in an area south and east of the fuselage. No fire damage was observed on or around the wreckage, which was surrounded by 80-foot mature trees. Trees adjacent to the wreckage exhibited broken and cut branches near their tops, and the tree closest to the wreckage contained a fresh groove on its trunk which extended vertically downward from its upper portion, consistent with having been scraped by the helicopter as it descended downward toward the ground. The concentration of tree damage and position of the wreckage were consistent with the helicopter descending into the trees with little or no forward airspeed. The engine compartment was intact, and the engine exhibited N1 and N2 shaft continuity throughout the drive train. All rigid and flexible oil and fuel lines were checked and found to be intact, and all fittings were found to be tight. All engine controls, HMU, ECU, and CEFA were intact. Fuel was found present in the fuel nozzle line. The magnetic chip indicating plugs were examined and were absent of any particles. Visually, the engine and accessories appeared undamaged. The electronic control unit (ECU) was removed on-scene for data download. The ECU's connectors (J1 and J2) were found tight and secure with no damaged pins. The NTSB IIC and representative parties concurred that the engine and airframe should be transported to Air Salvage of Dallas, Lancaster, Texas, for detailed examination. Wreckage Examination at Air Salvage of Dallas: Fuselage - The fuselage exhibited heavy damage to the cockpit and nose area, with severe crush deformation on the floor section between the chin bubbles. The nose section was heavily deformed and fractured, and the cockpit damage appeared more severe on the right side. The passenger cabin was essentially intact and relatively undamaged. All rear cabin seats were intact and the occupant restraints appeared undamaged. The skid landing gear was attached to the fuselage, with the forward section of the right skid overload fractured and forward section of the flight step was bent upward about 10 degrees. The transmission and engine were intact and remained mounted to the cabin roof. Cockpit Controls - Flight controls in the cockpit displayed damage consistent with structural disruption of the forward fuselage. The instrument panel was partially detached from the cockpit from overload fracture of the support structure. The pilot's collective lever was overload fractured at the base and the throttle twist grip was found positioned at the 78% bevel position. Examination of flight control linkages from the cockpit to the hydraulic servos and tail rotor revealed multiple overload fractures from impact forces. Tailboom and Tail Rotor - The tailboom was detached from the fuselage from overload fracture of the structure at Boom Station 68. The first two sections of the tail rotor drive shafts aft of the oil cooler exhibited damage at the Thomas couplings. The coupling damage was consistent with misalignment at the time of tailboom separation. All tail rotor drive shaft hanger bearings were intact, operable, and appeared to be in good condition. Hand rotation of the 90-degree gearbox input produced smooth rotation of the tail rotor. The tail rotor was intact and blades were not damaged. Both tail rotor flapping stops were found reduced by blade flapping impacts. The tail rotor pitch change system operated normally by hand. Hydraulics and Main Rotor Control System - The control linkage from the hydraulic servo actuators to the swashplate was intact and operable, and the swashplate and collective sleeve assemblies were intact and operable. Both swashplate drive links were overload fractured. Two of the four main rotor pitch links were overload fractured at the upper ends, with the other two pitch links attached to their respective pitch change horns. The hydraulic servo actuators appeared to be in good condition. The hydraulic reservoir was empty, and the two hydraulic filter module indicator buttons were in the retracted position. Main Rotor and Hub and Blade Assembly - The main rotor hub and blades exhibited severe damage consistent with tree and ground impact. The main rotor hub yoke flexures were severely delaminated and "broomstrawed" consistent with impact forces. Two of the yoke flexure arms were partially separated from the center section from impact forces consistent with rotor rotation at impact with trees. All four of the main rotor blades were heavily damaged from impact forces, and fragmentation of the blade structure on outer span locations was consistent with moderate rotational forces when the blades impacted the trees during the accident sequence. Transmission Drive - The main transmission was still mounted to the pylon links, and impact marks were observed on the pylon restraint stops consistent with the transmission being forced in a clockwise direction (opposite main rotor direction) during the impact sequence. The K-flex driveshaft connection was intact between the engine and transmission, and rotation of the driveshaft by hand produced smooth rotation of the main rotor hub. The transmission appeared to be fully serviced with oil, and inspection of the upper and lower transmission chip detector plugs revealed no debris or chips. Engine and Fuel System- The engine was still mounted to the fuselage and appeared undamaged. The fuel line between the HMU and nozzle was removed and drained, and the amount of fuel present in the line (approximately 1 tsp.) was consistent with the line being full. Disassembly of the airframe fuel filter revealed a full filter bowl, with the element being clean. Fuel samples taken from the forward tanks and airframe fuel filter were analyzed at the Bell Helicopter Chemical Laboratory with no anomalies found. The power lever pointer on the HMU was found indicating between 69 and 70 degrees. The NTSB IIC and representative parties concurred that the engine be removed intact and transported to Dallas Airmotive, Dallas, Texas, for an engine test run. TESTS AND RESEARCH Engine Test Run at Dallas Airmotive - The engine was prepared for test cell in accordance with standard maintenance manual procedures, with all of the accident helicopter's original engine components and accessories, including the HMU and ECU. When performing the rigging procedure for the HMU, problems were noted with the actual position of the PLA and the readings obtained from the ECU. The PLA was manually rigged to obtain 0 degrees (minimum position) and 100 degrees (maximum position). Mechanical stops on the test cell throttle controller were then adjusted to 35 degrees PLA (idle) and for 75 degrees (run position). When adjusted as described, the ECU readings were observed to be erratic. At 100 degrees actual PLA position, the ECU indicated 65 degrees. At the cutoff position, the ECU indicated 10-11 degrees. The engine was then started in the Auto mode and engine speed and temperature oscillations were observed at a constant throttle setting. The engine control was then changed from Auto to Manual mode. The engine responded to throttle input as required; however, the ECU readings were erratic. After a slave ECU was installed, the engine operated with similar erratic readings. A slave HMU was installed, and the engine operated normally without erratic readings. The NTSB IIC and representative parties concurred that the HMU and ECU should be transported to capable facilities for further examination and testing. The HMU and ECU were sent to GPECS, West Hartford, Connecticut. A complete test plan was developed and followed for the conduct of test/evaluation the ECU and HMU. ECU Examination - External inspection of the ECU, revealed minor damage to the cover at the vibrator mounting hole. Both electrical connectors (J1 and J2) and contact pins were found to be undamaged with no visible signs of contamination or corrosion. The ECU was downloaded, and revealed no faults on the accident flight. A complete report of the download is contained in the supporting documents to this report. HMU Examination - Detailed external inspection of the exterior of the HMU, including the electrical connector and PLA potentiometer, did not reveal any visible damage or evidence of the anomalies discovered during the engine run tests. Further extensive electrical and mechanical testing of the HMU, in summary, revealed severe signal fallout on the PLA signal, which was found to be random in location, bi-directional, and present during both rotational and stationary operatio

Probable Cause and Findings

The partial loss of engine power due to erratic fuel flow metering to the engine resulting from the single point failure of the PLA potentiometer in the hydro-mechanical fuel control unit. A contributing factor was the lack of suitable terrain to execute a forced landing.

 

Source: NTSB Aviation Accident Database

Get all the details on your iPhone or iPad with:

Aviation Accidents App

In-Depth Access to Aviation Accident Reports