Metallurgical failure analysis of gears of accessory gearbox of an aero-engine

A few components belonging to the engine accessory gearbox of an aeroengine were forwarded to this laboratory for investigation. Examination revealed that there was fatigue failure in the bevel gear on O11 axis bearing part No. 104.06.52.008 and Sl. No. /set No. II 054027. Multiple fatigue cracks were found to have initiated at the concave side root fillet of one of the teeth. After initiation, these cracks have propagated progressively over a distance of about 45 mm in the hub before giving way to overload failure. The failure has resulted in dislodgement of a sector of the gear encompassing 5 teeth. Detailed examination revealed that the fatigue cracks were initiated during the start-up of the engine, that is, when the concave side of the teeth was the loading flank. It was also discovered that after initiation, these cracks have propagated only when the concave side of the teeth acted as the loading flank, that is, during the start-up of the engine. Similar fatigue cracks were seen in another tooth of the fractured segment. The fracture surface of the gear had distinctive bands that were resulted due to change in the crack propagation plane. The change in crack propagation plane generally occurs due to sudden change in the state of stress ahead of the crack tip such as during starting of the engine. Examination revealed that there were about 175 bands on the fracture surface. Assuming that each band is associated with a starting cycle of the engine and the A few components belonging to the engine accessory gearbox of an aeroengine were forwarded to this laboratory for investigation. Examination revealed that there was fatigue failure in the bevel gear on O11 axis bearing part No. 104.06.52.008 and Sl. No. /set No.  054027. Multiple fatigue cracks were found to have initiated at the concave side root fillet of one of the teeth. After initiation, these cracks have propagated progressively over a distance of about 45 mm in the hub before giving way to overload failure. The failure has resulted in dislodgement of a sector of the gear encompassing 5 teeth. Detailed examination revealed that the fatigue cracks were initiated during the start-up of the engine, that is, when the concave side of the teeth was the loading flank. It was also discovered that after initiation, these cracks have propagated only when the concave side of the teeth acted as the loading flank, that is, during the start-up of the engine. Similar fatigue cracks were seen in another tooth of the fractured segment. The fracture surface of the gear had distinctive bands that were resulted due to change in the crack propagation plane. The change in crack propagation plane generally occurs due to sudden change in the state of stress ahead of the crack tip such as during starting of the engine. Examination revealed that there were about 175 bands on the fracture surface. Assuming that each band is associated with a starting cycle of the engine and the average duration of each starting cycle as 50 s, the crack propagation time, after initiation, is estimated to be approximately 2.5 hr. Also, since the engine had a total of 288 starting cycles, it is, therefore, stated that the fatigue cracks were initiated in the gear after about 115 starting cycles after the engine was inducted in service. The primary failure in the engine accessory gearbox (EAGB) has been identified as the fatigue failure of the bevel gear on O11 axis. Once the fracturing took place in this gear, the smooth running of the bevel gear pair in O11 and O1 axes was affected. This has led to resistance in power transmission from engine to the EAGB. During this time, the quill shaft on O11 axis (part No. 104.06.52.003) has failed by torsional overload. Hence, the failure in the quill shaft is secondary in nature and it occurred subsequent to the primary fatigue failure of the bevel gear on O11 axis. The other damages seen on bevel gear on O1 axis (part No. 104.06.52.200) and the spur gear (part No. 104.06.52.025) are also secondary in nature. \ud Fatigue crack initiation regions in bevel gear on O11 axis were largely damaged due to post fracture impact. However, the available fracture signatures and the location of fatigue crack origins did not suggest any specific abnormalities in the material which could be responsible for premature fatigue crack initiation. This was further confirmed through metallurgical examination of the gear wherein the material of construction was found to conform to specification with respect to composition, microstructure, surface hardening, hardness and case depth. Hence, the fatigue failure of the gear appears to be because of reasons other than metallurgical factors. \ud From the laboratory examination and the analysis, it was not possible to identify the exact cause(s) for the premature fatigue crack initiation in the bevel gear in question. However, based on the failure evidences and the analysis, one or more of the following could be responsible for the fatigue crack initiation. It is suggested that these aspects be examined thoroughly by the investigation committee/court of inquiry before the conclusions are drawn.\ud (a) The fatigue crack has initiated and propagated only when the bevel gear on O11 axis has acted as a driven gear, that is, during the engine start-up. Hence, it needs to be examined whether or not there was any situation during the service life of the engine wherein the torque during starting cycle exceeded the design value. Examination of the run-down time in the engine log may provide valuable information in this regard. It is important to note that once the cracks initiate, they continue to propagate even if the engine running conditions are normal thereafter.\ud (b) The meshing between the bevel gears of O11 and O1 axes needs to be examined. In this regard, the contact patterns on the flanks of the failed gear need to be analyzed. In case of abnormal meshing, tooth bending stresses at the fillet may be unfavourable and this can promote premature fatigue crack initiation in the gear. \ud (c) Examination revealed discrete tool marks on the root region of the failed bevel gear. Comparative study suggested that such tool marks were not present in the root region of the bevel gear of O1 axis. Possible stress concentration effect because of the presence of rough machining marks leading to premature fatigue crack initiation also needs to be examined