Aeroengine Safety

Figure "Experience first for maintenance safety": This big fan engine of a four engined airliner suffered a bird strike and had to be shut down because severe vibrations. To remove it and bring it to the repair shop is time consuming and costly. This may be the reason, why the maintenance crew decided to fix the fanrotor with belts (seatbelts?). So windmillig with equivalent vibrations should be avoided. Subsequent the flight occurred with the three remaining intact engines. The high air drag of the fixed fan lead to an increased fuel consumption what wa obviously not scheduled. So it was necessary to divert to an airport on the way. There the flight safety authorities forced the exchange of the engine what required 10 days. An inspection of the other 3 engines resulted, that also these had to be exchanged because of deficits.

This case can serve as deterrent learning example. It even can deduced at the mentality inside the maintenance crew, which was absolutely not suitable for the maintenance work at airplanes respectively aero engines

Also the priorisation of economic considerations against safety and expertise throw a bad light at the whole management of the operator and at the corporate culture.

Instructions and manuals of the OEM have obviously been ignored.

Figure "Learning from incidents": The education and training of the mantenance personnel (job title aero engine technic) depends not at least from the training documents (see also chapter 19-1). They should use the experiences by means of concrete and practical examples. Unfortunately primarily only aircraft accidents or particular incidents are so sufficient documented published that they suit for education means. Of high information content and with this espacially instructive are reports about failures (accident, incident reports) of the diverse national (flight) safety administrations like ATSB (australian) and NTSB (USA). Such reports are available by the internet. But they schould evaluated for teaching purposes by an experienced expert.
A much greater number of interresting, especially practical problems happen indeed obviously worldwide. The, for an understanding necessary background informations usually are very scarce, if available at all. They are for the outsider too unsufficient to be instructive. To such documents belong various directives/instructions (Fig. "Maintenance instructions and specifications") like

• Airworthiness directives (AD-Notes),
• Alerts,
• Service Difficulty Reports (Lit. 19.1.2-16).

But just „minor daily incidents” would be for the practical work of the maintenance personnel informative and so of importance for the safety.

Fig. "Job descriptions of repairers" (Lit 19.1.2-2): These job profiles are to assess very demanding. It can be supposed that the the requirements for the characterised activities also trainers are not full aware. Hence arises the question, if the education comes up with the today's requirements (Fig. "Affects at safety due to maintenance"). Therefore in the following will be tried, to describe the necessary knowledge as precondition for certain activities.

Identify and evaluate outer damages: many damages are specific for a part/component and stand in close relation to their function. That counts, e.g., for pipe line/tubes (chapter 23.5). As well as typical regions of parts, at which specific failures like fretting or fatigue cracks occur must be identified, as also the type of failure. So is fretting at titanium tubes relative to CrNi steeltubes expecially critical to evaluate.
The identification of unnormal damages like wear phenomena assumes the knowledge of „usual“, i.e., component specific failures. This comparison is only possible with sufficient experience.

Estimation ot he further use respechtively repair possibility demands exact knowledge of the component function depending from operation caused changes. Especially if those indications in the evaluation documents can not be correlated without doubt (Fig. "Identification of a failure"). To correlate descriptions and limits in the documents like mauals, the failure/damage mechanism must be identified. For this far reaching knowledge, as well theoretically as also based on experience is necessary. Such demands apply also for the „reading and the interpretation of technical documents”.

The knowledge of the systems contains besides complex functions and its consequences also the identification and evaluation of anomalies. Typic example are foreign objects in the lubrication oil (chapter 22.3.3.1). The right approach at magnetic chip and filter controls requires also knowledge about possible causes (chapter 22 and chapter 23).
The diagnosis of problems should determine causes of failures with suitable remedies (repairs, chapter 19.2.5). This requires even more extensive knowledge, because otherwise the danger exists, that merely the systems will be cured.

Approval authorised personnel:
A special qualification is demanded (in Europe) from so called inspectors of aviation equipment, since 1991 labeled as approval authorised personnel. For this an authorisation according to JAR 145 for aircraft engines (with type specific qualification) is required (Lit. 19.2.1-12 and Lit. 19.2.1-13). The necessary level of knowledge must be demonstrated in a specific test, according to the classification into a „category“.
The European aviation authorities (Joint Aviation Authorities = JAA) decided 1991 a harmonisation.
This changes from person orientated testing to system orientated quality insurance.

JAR 145 approved maintenance organisations. The accordant approval authorised certification together withe the JAR 66 rules the internal authorisation for an personel approval authorised certification.To the circle of the „certifying staff” belong according to this directive for the engine relevant area :

• line maintenance certifying mechanic (LMCM, Category A). This one is legitimated to certificate approval authorised attestations for personally executed simple scheduled maintenance work and remedied easy faults.
• Line maintenance certyfying technician (LMCT). For engines category B1. Authorised for approvals according category A and after maintenance work on engines with electrical systems.
• Base maintenance certifying engineer (BMCE, Category C) is authorised to approve overhaul work for the whole aircraft and all systems.
• JAR 66 certifying staff maintenance regulates the requirements for the licence of the approval authorisation together with JAR 145.
• JAR 147 approved maintenance training/ examinations: regulates education and approval/permission of the education facilities.According to the JAR 145 the airworthiness certificate by the overhaul shops is carried out. According to German law this happens by the `Prüfer' (inspector) himself or under his attendance.
  

This was a transition from the person oriented inspection to the system oriented quality insurance. For this the approval authorised personnel must own a licence according to JAR-66 and an internal authorisation.

Besides sufficient practical experience, which also complies with the completed education, extensive theoretical knowledge is necessary. For it come into consideration as well skilled workers (e.g., aircraft engine mechanics) with specific further education as also technicians and engineers with high school education and proved practical experience. The educations are offered from schools/academies which are approved by the responsible aeronautical authority (Lit. 19.2.1-12).
The duties/rights of the approval authorised personnel are very responsible.

Regarding the for this available specialist literature (Lit. 19.2.1-14 and Lit. 19.2.1-15) so the conclusioncan be drawn that to evaluate also complex situations (Fig. "Errors due to not understood failure mechanisms"). For this for example may also belong the knowledge about failure failure developments, failure mechanisms and risks. These thems the book at hand and the other volumes of this row are especially dedicated in an extent like this is in no free available specialist literature the case.

Fig. "Much can be seen during passing by an engine" (Lit. 19.1.1-9): The maintenance personnel and the operator can with attentiveness early recognise indications of problems and if necessary in cooperation with the manufacturer in initiate time suitable remedies (Fig. "Optimal use of human sensations"). In the adjacent illustration typical outer features for developing problems on an engine are shown:
Wire jacket of elastic bellows (A) can get after lonter operation times due to the function caused component movement worn (fretted) through from the inside. This leads to wire fractures and fan out of the jacket, before it comes to a full failing of the pipe joint.

Control jackets for fan outs!
Flange connections (B) vof pressure loaded casings like these of the rear compressor or the combustion chamber and turbine, can show unacceptable leakages (e.g., because of failing of the bolted connection or by crack formation) which show itself with local discoloration of the casing wall.

Look after unusual discolorations and deformations of the casings in the flange area!
At pipelines in the region of fastening clamps (C) there is an icreased danger of wear/fretting marks and crack formation (example 23.5.1-2). Additional the beneficial damping effect of a clamp at vibrations of the pipeline will be minimised by a loosening fit.

Look for wear/fretting marks in the region of clamps!
Crack formation occur on-again at tensioned and vibrating pipelines (D, Fig. "Causes of pipe line vibration").
Therefore is to set a high value on untensioned pipelines !
Leaking brazings or weldings (M).
Brittle crack by forced overload of an (cast) elbow (L).
Wear marks/fretting marks at contact surfaces of lines (E) can be the beginning of a function break down, therefore:

Identify and avoid wear marks/fretting marks on the parts/components!
Leakage of oil, fuel, hydraulic fluid especially in the region of connections/screw joints (F).

Look for leaking media!
At flange clamps (G) the riveted or spot welded straps are prone for crack formation (Fig. "Leakage of accessory devices" and volume 1, Ill. 5.4.2.2-2). But also fractures of the tension bolt in connection with corrosion (stress corrosion cracking = SCC) or embrittlement can not be ruled out (volume 1 Ill. 5.4.2.2-1).

Look at rod strainers on flang clamps!
Bulgings (H) at pressurized casings. At hot parts local annealing colours or increased oxidation can be indications of problems in the inner of the components/parts (e.g., casings of the combustion chamber and the turbine, volume 2 Ill. 9.3-3 and Ill. 9.3-5) .

If casings show bulges/deformations inform the manufacturer/OEM!
Weldings and deformation hindered areas like corners of casings (K) are particular sensitive for thermal fatigue (TF) and pressure cycles (volume 3 Ill. 11.2.2.2-8). Locally noticeable changes like discolorations can give for this regions hits on, at cracks escaping hot gas.

Look for indications of crack formation in the area of casings!

Figure "Errors due to not understood failure mechanisms" (Lit. 19.1.2-8): this flight accident is in detail described in Fig. "Risk of combustion chamber repair". Triggering was the difficulty of the identification of failure causes and processes as well as its influence at the component properties.

About the damage/failures of hot parts like the walls of combustion chambers after a long operation time should exist sufficient imagination (volume3, Ill. 11.2.2.2-1). For example high operation temperatures and thermal fatigue loads let expect a damage of the material (structure, oxidised cracks) which doesn't permit the condition of a new part after repair.
Basic knowledge about crack growth should include also the interaction with other component specific operation loads (HCF, inner pressure). This should in the case on hand protect to look only for TF (thermal fatigue) and not for TMF (thermal mechanical fatigue, see on this volume 3, chapter 12.6.2).

The criterion of a failure degree, e.g., as `severe' demands the transfer at the component/part. For it also experience is necessary, even if graphical material/photos are existing.

About crack formation schould exist enough knowledge that the characterisation by the crack width instead the crack length is at least problematic (danger of confusion, Fig. "Risk of combustion chamber repair"). An evaluation on theis basis conflicts with the insights of the fracture mechanics. It is, if at all, only thinkable if there is sufficien case specific experience.

This example schows how important the specific knowledge of the mantenance crew is (Fig. "Job descriptions of repairers")

Figure "Identification of a failure": at the inspection or overhaul of the engines hot part (upper sketch) of a helicopterengine the suspicion of an overtemperature during operation existed. The further use of such wheels of the gas generator turbine (lower sketch) was evaluated accordant to the optical indication. This was based on the meager informations in the overhaul manual: evidenceof heavy overheating at the blades of the 1st stage turbine wheel rule out the further use of this component and of the following 2nd stage wheel. So in this case both parts have to be scrapped.

Unfortunately lacks in the overhaul manual the sufficient description what to understand under the term heavy overheating. Especially was not clear, in which failure this failure symptoms become manifested. So the leading edges of the blades were inspected with the binocular for overheating indications according to the instructions. Thereby in many cases a picture like in the shown detail, that demands the selection of the turbine wheels because of a damage due to overheating was evaluated. After over a longer period the costs of such hot parts increased unacceptable, the materials specialists were assigned with the assessment of the failures.

The optical investigation indicated, that the leading edges of the blades showed a burr and erosion flutes, what can not be explained by an overheating. Oviously it was to be about erosion, probably by coke particles out of the combustion chamber (see volume 3 Ill. 11.2.2.3-3).

A destructive laboratory investigation (metallography, volume 4 Ill. 17.3.2-5) confirmed this finding. It showed no signs of a risky material damage due to overheating. The damage on hand could have been removed with an easy mechanic rework.

So the high repair costs by scrapping the wheels over the years, have not been necessary. The evaluating personnel must have realized with sufficient expertise the failure features, especially the circumferential orientated burrs and flutes as not typical for a heavy overheating. As remedy suitable pictures were added into the overhaul manual.

Figure "Misinterpretation of operation damage" (Lit. 19.1.2-10): Schortly after the start, in about 50 m hight, the pilot heared a loud bang and a roaring noise, which he identified as a compressor surge. The airplane tilted to the right. The pilot adjusted the propeller at feathering position. The aircraft speed dropped. The airplane made an emergency landing and was thereby heavily damaged. After the, in the following described malfunction, the right engine was repaired in the night before the accident.

At a formerly start attempt „the engine was stuck“ and the turbine entrance temperature reached for about 10 seconds 800°C. After this it was shut down and about further 10 seconds cranked by the starter. Thereby flames escaped from the exhaust tube. A second start attempt had also to be aborted. The otherwise free rotating propeller could no more turned. A disassembly showed, that the tips of the gas generator turbine wheel blades are „burned” respectively partial molten. The molten material splashed at the guide vanes and the wheel of the power turbine. Therefore, following the evaluation by the OEM a gas temperature of about 700°C is needed. The overhaul manual on hand contained no sufficient criteria to determine, if in the engine emerged an overtemperature. Merely the instruments display for the pilot indicated this. Afterwards the affected parts were peened with glass beads. This approach beither suggested by the OEM, nor scheduled in the overhaul manual. Thereupon the entire gasgenerator turbine, combustion chamber exit, thermocouples and fuel injector have been replaced, but not the the power turbine. During the following test run, all instrument indications were normal and there were no anomalies.

At the question after the accident, which criteria lead to the further use of the power turbine was the answer of the responsible: there were no informations about the cleaning of the parts in the OEM overhaul manual. So ta large certificated repair shop was asked. After this information he assessed the serviceability of the power turbine as operable, because during cleaning no indications of an overheating, like discolorations and deformations at the blades were discovered (volume 3 Ill. 11.2.3.1-9). Comparisons with other engines for the evaluation of the parts were not possible. The OEM was not consulted.

The exhaust casing of the right accident engine was torn from the inside behind the gas generator, alongside the containment ring(!). Five blades of the power turbine were broken, others damaged.
Additional the actuator rod of the propeller controller showed at the end a fatigue fracture in the thread to the ball joint. Fracture causative faults have not been found.

Conclusions: a blade of the power turbine failed already at the start. The developed unbalance overloaded its already by the overtemperature and incipient melting weakened blades. So it came to the fracture of several blades. The still running gas generator drove the damaged power turbine further. The to the propeller delivered power was however no more sufficient. Severe vibrations lead to the fracture of the actuator rod of the propeller controller. This prevented as a secondary failure the feather position of the propeller.

Summary: the assessment of the hot part damages by the maintenence personnel was wrong. From these sufficient experience should be expected to identify, if in an engine gas temperatures above 1090 °C over the 2 seconds limit had occurred that damaged the power turbine (Fig. "Identification of a failure"). Thereby they should have identified, independent from the lacking evaluation documents, that overtemperatures made the exchange of the power turbine necessary. In spite of the engine failure, the pilot could have avoided an emergency landing with the right action.

Figure "Understanding technical terms is safety relevant": Flow disturbances in the compressor can develop in different manners and get failure. Therefore the maintenance personnel of aero engines should have understood the basics of its mechanisms. This themes are discussed in detail in volume 3, chapter 11.2.1.1 and chapter 11.2.1.2 and are here only shortly addressed.

Rotating Stall ( = rotating flow separation) is an effect at which the flow separates from single lightly damaged or influenced profiles (frame above left). This separation then migrates against the rotation along the circumference of the rotor. By this, the blades are short-term unloaded , what can trigger severe vibrations. This high frequency process can not be recognized from the outside of the compressor and can therefore last over a dangerous long time till a fatigue crack/fracture occurres.

Surge (german `Verdichterpumpen') stands for the total flow interruption which spreads in axial direction from the concerned stage (sketch above right). Thereby it comes to a repeated pulsating load of the compressor components. It must be rekoned by this surge buffets with short term overloads. In an extreme case rotorblades and stator blades can be deflected so wide that they touch each other. Thereby LCF fractures/cracks (volume 3 Ill. 12.6.1-6) of the blades/vanes (example 19.1.2-1) and forced fractures of the main bearings are possible.

Note:
Frequently the literature does not properly distinguish between the term stall (flow separation at single blades) and surge (interruption of the whole compressor flow).

Example 19.1.2-1 (Lit. 19.1.2-11): shortly after the start during climb the crew of an airliner heared a loud bang as it is typical for a compressor surge. The cockpit instruments indicated a fast rotation speed drop of the low pressure compressor and the high pressure compressor of one engine. The exhaust gas temperature exceeded the scale of the instrument display. The pilot at once shut down the engine and undertook an emergency landing.

The subsequent investigation of the aviation authority found, that all operation data of the concerned engine, till the loud bang had been normal. The turbine showed features of an extreme overheating (Fig. "Understanding technical terms is safety relevant" bottom right). The turbin entrance guide vanes (nozzles) merely exhibited light signs of overheating at the trailing edges. The high pressure turbine rotor blades behind as well as all blades of the low pressure turbine had be molten down. The solidified melt was concentrated in the exhaust pipe.

The day before the accident during accelerating the engine to idle a surge occurred that repeated at least 5 times during rolling of the aircraft. After that, according to the maintenance manual, the bleed valves had been exchanged. The mantenance manual (touble shooting guidelines) didn't demand an inspection of the compressor, so this didn't happen. A following engine start seemd to confirm the successful failure elimination.

The airplane returned to operation for a flight of about 3 hours. However it was indicated, that the turbine entry temperature (EGT) had increased about 20°C and the high pressure rotation speed had been dropped about 2% compared to former data. From this it could be concluded that adverse aerodynamic conditions due to a failure in the compressor and/or the turbine had been settled.
An investigation showed, that several compressor rotor blades had failed in the dovetail roots by vibration fatigue. A majority of other compressor blades showed cracks in the roots. The, for this causative extreme bending vibrations are typical for a compressor surge (volume 3 Ill. 11.2.1.2-1).

Evaluation by the investigating administration:

• Even though by the OEM in his „trouble shooting guidelines“ not demanded, a performance analysis in line with a ground run with high performance should have been necessary. A following boreskope inspection should have discovered the damaged compressor blades.

• The maintenance personnel and inspection personnel should have been sufficient informed about the identifying features and the consequences of flow separations (stall and surge).

• Operators have to instruct their personnel how to indikate the consequences or a compressor surge.