Synopsis
The crew reported a loud grinding sound immediately followed by illumination of the 'engine failure' light. They aborted the takeoff at approximately 30 kt and as the thrust levers were closed the commander was aware of resistance within the No 1 thrust lever system. Subsequent examination revealed significant fire damage centred around the underside of No 1 engine, evidence of penetration from inside the engine casing and a fracture of a second stage low pressure (LP) compressor fan blade at the root attachment. There was no evidence of cowl penetration.
History of the Flight
The crew operated an uneventful flight from Frankfurt to East Midlands airport and shut down the aircraft at 0135 hrs. The First Officer (FO) was the designated handling pilot for the next sector. Before starting engines the crew obtained the current ATIS: surface wind 090°/08 kt, visibility 300 metres in fog, runway visual range Runway 27, 550 metres, cloud scattered at 100 feet, broken at 200 feet, temperature and dewpoint 4°C, QNH 1012 mb. A normal start was conducted for each engine; engine anti-ice was selected ON prior to taxi. ATC cleared the aircraft to taxi at 0320 hrs and 5 minutes later it was cleared to take off on Runway 27. In accordance with published procedures for a takeoff in icing conditions, 70% N1 was set for 15 seconds prior to brake release. The crew later reported that the engine parameters were stable during this period. The brakes were then released and the thrust levers were advanced smoothly to the take-off setting (2.1 EPR for engines No 1 & 3 and 2.0 EPR for engine No 2). At about the time the engines achieved take-off power the crew heard a loud grinding sound, followed immediately by illumination of the 'engine failure' light. The crew estimated their speed at about 30 kt.
The crew aborted the takeoff and the commander took control. As he closed the thrust levers the No 1 engine fire warning activated. Resistance within the No 1 thrust lever system made it difficult to close fully the No 1 thrust lever. He informed ATC of the abort with an engine fire and reminded them of dangerous air cargo aboard. He called for recall items for the engine fire drill. The FO confirmed the No 1 thrust lever was closed and attempted to position the No 1 engine start lever to 'CUT OFF' but could not due to an apparent restriction. He continued recall actions and pulled the No 1 Fire switch. The engine fire warning light remained illuminated; he discharged the first fire extinguisher. After 30 seconds, with the light still illuminated, he discharged the second extinguisher. The flight engineer read the engine fire checklist; when he called for confirmation that the No 1 engine start lever was at 'CUT OFF', the FO was able to place it correctly without difficulty. The aircraft approached taxiway Golf; the No 1 engine fire warning light was still illuminated. As the commander halted the aircraft, the first fire appliance arrived. One fire crew contacted the flight crew on 121.6 MHz and informed them there were flames from the front of No 1 engine and they were applying foam. The commander opened his cockpit window and, unable to see any signs of fire, called for the evacuation drill. As the flight crew completed the drill, the fire crew reported the fire was extinguished. After discussion, the flight crew remained in the aircraft while it was towed to a nearby servicing area with fire crews in attendance.
Airfield Rescue and Fire Fighting Services
Low Visibility Procedures (LVPs) were in force at East Midlands. A fully manned fire appliance was positioned at the Mike 3 holding position to prevent unauthorised movement onto taxiway Alpha. Two other fire appliances were on LVP standby at the fire section. The response was immediate; no problems were encountered locating the aircraft in reduced visibility. The aircraft carried dangerous air cargo. The senior fire officer was provided with a copy of the load manifest on arrival and could determine the most effective response. Airport emergency services, in conjunction with freight operators, had conducted recent exercises involving dangerous air cargo; their effective response was due in part to lessons learned from those exercises.
Flight Recorders
The aircraft was fitted with a Cockpit Voice Recorder (CVR) recording on a continuous 30-minute loop. The circuit breaker was not pulled after the incident, so relevant information was over-written. The Flight Data Recorder, a magnetic tape-based Universal Flight Data Recorder (UFDR), recorded only mandatory parameters (airspeed, altitude, normal acceleration, heading, time) on a 25-hour loop. As the accident occurred below 45 knots, the minimum airspeed registration value, no useful information was obtained.
Engine Examination
The Pratt and Whitney JT8D is a two-shaft, low bypass ratio turbofan engine with a six-stage LP compressor. No external damage was evident, but upon opening the No 1 engine cowling, significant fire damage was found centred around the underside. Sooting, oily residue, wiring damage, and heat damage to fuel control unit input rod bearings were noted. There was evidence of penetration from inside the engine casing in the area of the discharge vanes at approximately the 5 o'clock position (viewed from front). A small piece of debris was found within the cowling. Further examination revealed two penetrations of the engine casing around the 5 and 7 o'clock positions, with damage to the oil tank and a ruptured fuel pipe. Some liberated material penetrated and passed through the engine core, with solidified metal spatter visible in the tail pipe. Visual examination of the LP compressor first stage showed moderate damage to leading and trailing edges of stator and rotors.
A boroscope examination and partial engine strip revealed that a second stage LP compressor fan blade had detached. A portion (approximately 2/3 span) of the blade had been retained within the casing. The blade had fractured at the root attachment, consisting of two clevis-type straps. The attachment pin was intact. The penetration marks on the fan case were consistent with fan blade release kinematics. The initial blade tip impact was in the plane of rotation; the impact of the blade root was aft of the plane of rotation. Both impacts tore the containment case. It is probable that no significant material exited through these penetrations as only one piece of blade strap material was recovered from inside the cowling. There was no evidence of cowl penetration. The fan discharge vanes were severely damaged in the area of the case rupture and also from 1 to 3 o'clock, likely caused by the liberated blade being dragged around just prior to or after final release. Blade tip material likely exited via this route. Some damage to core inlet guide vanes was noted.
Metallurgical Examination
Visual examination of the blade root fracture surfaces revealed extensive post-failure rubbing damage. Where surfaces were not destroyed, evidence indicated ductile overload failure. One fracture had a stepped appearance unlike the other three and is considered a possible initiation site, but post-failure damage had destroyed corroborative evidence. Some twist in the root section indicated sufficient force to deform the blade during failure. It was also probable that the forward part of the root failed first, with the rear failing in overload during twisting.
Previous Incidents
There have been previous incidents where LP compressor second stage blade root fractures resulted in blade release. In some cases, released fragments penetrated the engine cowl, damaged or severed fuel lines, and caused engine fires within the cowl. Manufacturer's Alert Service Bulletin (ASB) 5729 Rev 2, dated July 8 1988, identified this problem. At that time there had been 115 fractures of second stage fan blade root attachments. Twenty resulted in cowl penetration but with no reports of significant aircraft damage. Nine resulted in fires, all extinguished by engine aircraft fire extinguishing systems. The blade root fractures were determined to have required inspection per the ASB. The failed engine had been inspected per the ASB, but failure occurred before a re-inspection was required. A blade redesign is available through a further Service Bulletin; no reported failures to modified blades.