A JAL Japan Airlines Boeing 767-300, registration JA8980 performing flight JL-632 from Kumamoto to Tokyo Haneda (Japan) with 209 passengers and 8 crew, was climbing through about 6000 feet out of Kumamoto's runway 25 when the crew observed vibrations from the left hand engine (CF6) as well as increasing exhaust gas temperatures. The crew reduced the engine to idle thrust, stopped the climb and returned to Kumamoto for a safe landing on runway 25 about 25 minutes after departure.
A ground witness, director of a local hospital, reported they heard noise, glass of a car parked outside the hospital as well as a window of the hospital shattered when metallic pieces fell onto the ground. None of the 10 patients affected by the broken window was injured.
Japan's Ministry of Transport reported metallic pieces, size up to 5cm by 5cm (2 inches by 2 inches) of the left hand engine, parts of turbine, were found about 7km southwest of the aerodrome and 9 other locations below the flight path of the aircraft. A post flight examination revealed the engine debris had also scratched the horizontal stabilizer. The occurrence was rated a serious incident and will be investigated by the JTSB.
The airline reported the passengers were rebooked onto the next flights.
On Jul 31st 2020 the JTSB released their final report concluding the probable causes of the serious incident were:
It is highly probable that this serious incident was caused by the fractured blade #13 on HPT (high pressure turbine) stage 2 of No. 1 engine (left side), when the Aircraft was climbing, that damaged blades and stator vanes of aft stages, fragments of which collided with LPT (low pressure turbine) casing and generated a hole (crack).
It is highly probable that the fractured blade #13 was caused by cracks that were generated on TA (Turning Around (branching and turning around of cooling air flowing inside blades)) area and progressed thereafter.
It is somewhat likely that cracks generated on TA area were caused by hot corrosion swelling (blister) generated on the coating layer of the blades and low-cycle fatigue initiating from the cracks.
The JTSB reported about 400 fragments of the engine were found on the ground following the flight. The low pressure turbine (LPT) of the engine usually weighs 726kg, the remaining LPT weighed only 653kg, hence 73kg were lost as exhausted fragments.
The damage to the engine began at the high pressure turbine (HPT) stage #2, where a number of blades showed damage with blade #13 having fractured at the shank. Downstream there was damage to "great portion of blades and nozzles" of the LPT. The JTSB further wrote: "There was a hole (crack), which was about 9 cm long and about 2cm wide, generated near LPT stage 4 (in the direction of around three oclock if seen from the engine aft). Besides, the core cowl outside the hole generated on LPT casing was free from damage."
The JTSB described the sequence of events:
During the climb at an altitude of about 7,500 ft at an airspeed of about 250 kt after the take-off from Runway 07 at the Airport, an abnormal noise accompanied by vibration occurred, and the instrument panel indicated reduced rpm of No. 1 engine (left side) and the increases in exhaust gas temperature (EGT) and engine vibration; therefore, the captain switched over to the PF from the PM and conducted the items to deal with Engine Limit or Surge or Stall on the non-normal check list after setting engine thrust idle.
When doing this, the captain and the FO confirmed that the engines instrument panel indicated that figures were reduced to the normal, and the vibration and abnormal noise became lower after reducing engine thrust.
As the captain slowly increased engine thrust accoording to the procedure in the check list, the vibration and noise were increased, therefore, he immediately returned engine thrust to idle.
Although the vibration and abnormal noise from the engine became lower, the vibration still continued; and besides, as there was an available departure aerodrome for landing, the Aircraft returned to the Airport after air traffic control priority was granted, and landed at the Airport at 16:17.
The JTSB analysed:
(1) Engine Fracture Initiating Point
Turbine blades were fractured at various radial locations from HPT stage 2 and aft. HPT stage 2 had four fractured blades of #13, 12, 11 and 10 in the order of having less remainder. Among these, fractured surface of blade #13 had marks of low-cycle fatigue initiating from the TA area. From this, it is highly probable that the engine was damaged initiating with the fractured blade #13.
(2) Cracks Generated on the TA Area of Blade #13
It is somewhat likely that the cracks on the TA area of blade #13 were generated by low-cycle fatigue initiating with the damage generated on the coating layer because swelling (blister) caused by hot corrosion and cracks were generated on the coating layer (see Figure 6).
It is somewhat likely that cracks generated on the coating layer of blades, which progressed to the base materials and finally led to fracture of #13 blade, were involved by complex factors of increased cycles in service, steep rising shape of cooling air passage wall in the TA area (small curvature radius of TA area) (see Figure 8) and the thick coating layer.
i) Cycles in service Generally, the more cycles in service increase, the more quality of coating deteriorates, thereby leading to generation of and progress to hot corrosion and cracks. Accordingly, it is somewhat likely that increased cycles in service affected generation of cracks on the coating layer as well as progress to the base materials.
ii) Rising Shape of Cooling Air Passage Wall in TA area Blade #2 (G14 Type) and #11 (G15 Type), which had more gradual shape of cooling air passage wall in TA area than #13 (P03 Type), were free from generation of cracks on the coating layer (see Table 1 and Figure 8). Accordingly, is it somewhat likely that the steep shape of cooling air passage wall in the TA area of the blade generated stress concentration to the pertinent area and affected generation of cracks on the base materials.
iii) Thickness of the coating layer
Comparative investigation into the condition of TA area of blade #3, 5, 7 and 9, all of which are P03 Type, same as #13, revealed that cracks were generated in the coating layer of #3 and #5 only, which had a relatively thicker coating layer (see Table 1 and Figure 8). It is somewhat likely that the thicker coating layer affected generation of cracks on the base materials because aluminide coating of blades has a low ductility with tendency to be prone to generate cracks when coating layer is thick.
(3) Damage to LPT Casing
It is probable that damage to HPT stage 2, LPT and LPT casing was initiated by the fractured blade #13 of HPT stage 2, followed by a chain of fractures of blades and stator vanes of aft stages, fragments of which collided with LPT casing near LPT stage 4 and generated a hole.
Although the hole was generated on LPT casing, it is probable that fragments of the engine did not penetrate LPT casing because the inner surface of core cowl covering LPT casing was free from marks of collision with the fragments.
It is probable that fragments of the engine dropped on the ground were exhausted from exhaust nozzle of the engine, not from the hole of LPT casing.