Compressor failure in Marbore Engine

Compressor disk failure investigation at“Turbomeca Marbore 6” engine from“Tzukit” (Fuga megister)Maj. Ayal AronsonEngine Failure analysis investigator, IAF

Failure analysis department• Our department is dealing with:– All failure analysis investigations involving flight safety.– All kind of platforms, systems and materials.• The investigators are aeronautical, mechanical andmaterial engineers.• Each investigator specializes in different system(engines, lending gears, transmission gear, structureetc).• We are using advanced equipment:– Optical microscopy – magnificaion capabilities up to X180.– Scanning Electron Microscope (SEM) – Field emission SEM(High resolution).• Engineering and lab supporting departments –metallurgy, analytical lab, measuring lab, NDTcapabilities and more.2

The event details• While "tzukit" airplane (Fuga megister) was training,an explosion was heard.• The explosion happened during the running fortakeoff.3

The event details• The airplane equipped with 2 “Turbomeca Marbore 6" jetengines.• Examination of the airplane and the engine revealed that partsfrom the compressor broke the engine housing and some outerpanels were found over the runway.• The event has high potential for crash and in miracle finishedwithout injuries.4

5The engine that failed after his disassembly fromthe airplane

6Visual examination - Parts breaking outfrom the engine’s housing

7The parts are from the compressor disk

The structure of the compressor• One centrifugal stage.• This stage is an assembly of stainless steeldisk (the leading edge) and aluminum disk.9

The structure of the compressorThe stainless steel diskStainless steeldisk - theleading edgeAluminum disk –the part thatfailed10The aluminum disk

The structure of the compressorThe tenon joints atthe aluminum disk12

Comparing between the disk that failed anda new oneThe undamagedstainless steel disk2 main broken piecesfrom the aluminum disk13

15All the tenons in the disk that failed werebroken/sheared

Fractography• Examination of both main fractures revealed “fingernail“ areas containingbeachmarks that propagating from the tenon radius at the wall that carriesthe loads.“finger nail“ areascontaining beachmarksthat propagating from thetenon radius.5mmX10mm16“finger nail“ areascontainingbeachmarks thatpropagating fromthe tenon radius.10mmX20mm

All the tenons were found to be cracked atthe radius of the wall that carries the loadsThe fracture surfacethe radius of thewall that carriesthe loads17

Fractography• Microscopic examination using SEM revealed fatiguestriations propagating from the tenon radius.fatigue StriationsX250018

Fractography– Corrosion dimples covering the side walls of the tenons.– Corrosion covering the fracture surface, Close to the origins.19

NDT by fluid penetration method.• 15 aluminum disks were examined by fluid penetrationmethod (level 4).• Few disks were found with indication of cracks at the tenon’sradius.• Most of the disks with indications of corrosion pits.20

Metallurgical cross sections• “Wedge” opening shape of the crack.• Transgranular combined with intergranular propagation.• Corrosion findings at the origins and around them.C-CD-D21

Summary and conclusions• The engine failed due cracks that initiated and propagated in thealuminum disk of the compressor stage.• The cracks initiated and propagated in a corrosion fatigue mechanism.– According to the literature – corrosion fatigue occurs when combiningcorroding environment and cycling loading.– The corrosion fatigue mechanism characterized by:• “Wedge” opening shape of the crack.• Transgranular propagation that can be combined with intergranular propagation.• Corrosion findings at the origins and around them.• The initiation occurred at the radius of the tenon, at the side that carriesthe loads.• The anomalies that were found in the measurements of the radiuses andthe corrosion pits that were found at the side walls of the tenons causedstress consecration and accelerated the initiation of the failure.23

Summary and conclusions• The reason that this cracks were not found until now is because themaintenance policy for NDT enabled choosing between inspection underlow magnification or fluid penetration (Zyglo) method.• From the other disks that were found with cracks, it can be concluded thatanother severe event was prevented due to this investigation.• From the amount of corrosion pits and the difficulty to see the coating itcan be concluded that the protective coating was damaged and was notaffective.• The probability of FOD (Foreign Object Damage) as a cause of the failurewas ruled out.– There were no signs for FOD over the stainless steel disk– Other disks were found cracked (with smaller cracks).24

Recommendations• Recommendations for overhaulmaintenance of the disks:– Define the maintenance policy for NDT onlywith fluid penetration method.– Renewal of the protective coating.– Measuring all the disks for anomalies in theradiuses of the tenons.– Examination of all the disks for stronglycorroded disks25