Aeroengine Safety

Figure "Suspect unapproved parts" (Lit 20.2.1-1): Behind the terms

• „suspect unapproved parts/products“
• „bogus parts/products”
• „counterfeit parts/products“

can be hidden very different parts respectively products. The following examples may somewhat highlight this.

Parts and components: In some cases these are originals but from scrap and/or accident aeroengines which have been illegal reconditioned and once more put into circulation. With the help of the serial numbers and part numbers, for example by a documentation of the OEM, the origin can be traced back. In case of doubt it can be advised to use this chance.

Not seldom we deal with not approved repairs. For example from a not certificated repair shop or a repair outside the limits, determined in the overhaul manual published by the OEM. From this especially turbine blades/vanes as very expensive and life time limited parts are affected.

The situation becomes very dangerous when rotor components (e.g., disks) with a life time out of date come again into operation. In those cases after a not to foresee operation time, exists the danger of a fracture with catastrophic conse quences. Because those parts with the highest likelihood will not show macrocracks, the possibility of reuse can not be evaluated non destructive. In such a case the documentation must provide information. If there is suspicion, a close, critical examination of the documents, neccessary for the assembly, is essential.

A further source of SUPs are military surplus parts. Even if those have been supplied by the OEM respectively produced new parts the assembly into civil aeroengine types is from the first illegal. Naturally an OEM has also financial interrests, that the parts don't flood the market, bypassing him. However the experience shows, that also the safety aspect and drawbacks for the customer can not be neglected. So for example, it was not possible to operate combustion chambers from military stock in also military aeroengines but of an other helicopter type. Already at the test rig at those „foreign” combustors heavy vibrations occurred. No mean was found for the rerctification with an effort that could be justified.

Machine elements in many cases are delivered as „fill up articles“. To those belong connection elements like bolts and nuts. Their labeling is not individual (not for every single part). So it seems comparatively easy to infiltrate parts, which are not approved /certified for the application. The special problem consists in this, that the parts are often safety relevant and deviations respectively quality deficits can not be identified from the outside (Fig. "Errors due to not understood failure mechanisms" and Fig. "Identification of not approved parts").

Anti friction bearings show indeed significant labellings. But those can be imitated quite easy (Fig. "Identification of a failure"). Here also bad quality can be seen first in a shortened life time (Fig. "SUP features of a friction beating").

Auxilary materials like oils, lubrication greases, cleaning fluids and washing agents can come aware or unaware to application. So it is thinkable, that the `purchasing' has foud an especially cheap supply source whose product does not sufficient correlate the prescribed. For example sulfur compounds which get with grease at hot parts can trigger dangerous hot corrosion (sulfidation, Fig. "Unsuitable lubricant leading to fracture"). Chlorine containing cleaning agents can trigger cracks in components from high strength titanium alloys (volume 1, Ill. 5.4.2.1-8).

Admixtures (e.g., brazing alloys, spray powders) in production and repair for example can determine the strength of brazing joints and welding joints or the characteristics of thermal spray coatings. Even little deviations can unallowable decrease the life time of the part. A later non destructive test, for example of a part that is repaired by brazing, is then hardly possible.

Figure "How SUPs develop" (Lit 20.2.1-3, 20.2.1-7 and Lit 20.2.1-10 up to Lit. 20.2.1-13): Here a survey is given about the manifold reasons to classify a product as suspect (see to this Lit. 20.2.1-5).

„A“ Military origin: Normally these are new parts from storages (surplus parts). They come to the market because the related airplane/helicopter respectively its aeroengines are no more in service. We speak about „breakout parts” if the military order was not directed at the OEM but at more favorable (cheaper) supply source.
Obviously the awareness can not be assumed, that the approval/certification of such parts for civil use is at least questionable and must be clarified. In Lit. 20.2.1-6 we can find important hints about this matter.

„B“ OEM or a certified supplier for the military: As in „A” already described, such parts can get to the assembly. However here parts are meant which originate directly from the producer but who is not scheduled for the supply of the cvivil market.

„C“ Approved/certificated producers: To these belong besides the OEM also producer who are approved by the aeronautical authority (FAA). We speak about parts-manufacturer authorization respectively parts-manufactoring approval (PMA, Fig. "Documents for approved parts 1"). Obviously rather the possibility exists, to bring parts of surplus productions, possibly from a deviating production process, on the market.

„D“ Waste, rejections: There may be intent to deceive. Thinkable is, that such parts emerge as repair parts with suitable serial numbers of a no more reparable part.

„E“ Not approved repaired parts: This can be parts from a producer who is not certified for a certain processing step. To these also count parts which have been produced by a process which deviates from the certification.

„F“ Used parts without sufficient testing and documentation: This is exspecially dangerous for lifr limited parts. To those count primarily rotor components. Do these fail during operation it must be rekoned with catastrophic consequences. Unfortunately the end of the cyclic lifetime can not be recognised sufficiensis given t sure, even with sensitive non destructive testing methods which are suitable for series application, let alone by a visual check (Ill. 25.2.2.2-2). On the other hand these are normally very expensive parts. Therefore scepticism is appropriate if the history of the part can not be completely documented (Fig. "Identifying not approved parts"). It is thinkable, that such parts are unconscious once more assembled because of an unsufficient documentation. However also knowingly deception with invalid documents can not be ruled out.

„G“ Not certified producers: In such a case we also speak about „counterfeit parts”. This term, although not generally defined is used for parts which in fact are capable for aviation use, but their documentation was manipulated.
Especially dangerous are parts which only externally look alike original parts, however have been produced consciously cheaper with inferior quality.This can be founded in deviating production processes but also in materials with unsufficient features. In such cases criminal energy can be supposed.

„H“ Flight accident aeroengines: The experience shows that obviously there is a success to introduce such parts into the circuit. Terefore in case of doubt the origin must be completely investigated. Help can come from a registration of the serial numbers at the OEM.

„I“ Parts with wrong, counterfeit or unsufficient papers: Even if the parts are technical acceptable they still are embraced by the term SUP if the documentation (Fig. "Identifying not approved parts" and Fig. "Documents for approved parts 1") shows deficits (Lit. 20.2.1-5). For example the unsufficient possibility of a trace back. This can reach as far as to the rough material supplier.

„K“ Stolen parts: Sure the often high prices of aeroengine components represent a temptation. Depending on where the parts have been stolen and which documents have been also stolen it can be difficult to uncover the illegal origin.

„L“ Bankrupty assets: Such parts can come from an operator (airline), a trading company or a producer, respectively a repair shop. In such cases these may be absolutely not always SUPs. However it is thinkable, that the documentation is unsufficient or the parts are application bound, for example for a certain aeroengine type.

Figure "Visible features of SUPs" (Lit. 20.2.1-10 ): There are part specific potential features for the suspicion of a SUP. Naturally by far not all characteristics are the proof for a SUP. If the suspicion is justified may show not until a detailed investigaston. Basically counts:

Unusual is suspect!
Geometrical features: These are not deviations beyond drawing tolerances and specifications. Even in those, tolerances can be noticed, which lay different as usual at the upper or lower limit. In this sense mating forces of centrings can differ noticeable from the usual. This improves the reconizability, but can also trigger damages by galling/seizing during joining (Fig. "Failures of shrink fitting").
Eye-catching radii geometries and sharp edges can be able to increase undetected local stresses at a damage (notcheffect). This can dangerous lower the vibration fatigue strength (Ill. 20.1-17). Do such failures occur, this can be a hint at such an SUP.

Are flown through openings concerned sharp edges and burrs are quite able to influence the operation behaviour inacceptable (e.g., combustion chambers and control units, volume 4, chapter 16.2.2.2).

Machining characteristics: There are further features besides the already mentioned burrs and badly rounded edges.

To those belongs an unusual orientation of the tool marks. Normally the OEM will pay attention in the specifications, that those marks will not run crosswise to the critical stresses. Deviations from this can be assessed as a sign to look closer at the origin of the part.

Roughness, even inside the specifications can differ quite markedly in the topography (structure/geometry, volume 4, Ill. 16.2.2.1-1 and Ill. 16.2.2.1-2). A possible feature can be a dull or bright/reflecting surface. This is not only a formality. It can be a characteristic of a changed production process and a marked influence at the strength.

For example, a dull surface can be caused by an abrasive blasting process which should cover something like a rework. Dull surfaces develop also by etching/acidly/corrosive treatments. Are those not intended there is the danger, that an attack exists which lowers the vibration fatigue strength unacceptable.

Shot peening surfaces are usually needed to guarantee the vibration fatigue strength. Lack typical indications of a shot peening treatment, is this a dangerous deviation. For example the roots of many turbine rotor blades are shot peened. The boundary or the peening pattern of the shot peened surfaces can be visually evaluated. It is quite alarming if this does not conform with the original parts. To this belongs when the (brittle) diffusion coating which serves as oxidation protection is also peened. Here a checking for the suspicon at a SUP should be advised.
Discolorations are an important feature for the unusual. To this belong distribution, intensity and colouring of annealing colours. It is not necessary that the discolouration itself influences the operation behaviour, but it can be a hint at deviations in the production (volume 4, chapter 16.2.1.1 and chapter16.2.2.6).

Indications of rust can stay in connection with an unsuitable packing which itself creates suspicion.

Rework marks at new parts and repaired parts if unusual should attract attention. It can show at a locally changed look of the surface (machining scratches, polished areas, flat depressions). There is always the suspicion, that there was a damage/fault which was not removed as specified/approved. It is especially alarming if those traces are located at zones at very high loaded new parts which determine the lifetime. At least such features must agree with the corresponding specifications (drawing, manuals and so on).

Welding and brazing: Anomalies of the appearance are noteworthy. A very rough braze joint can point at an insufficient melting caused from deviating brazing parameters or properties of the braze metal (volume 4, Ill. 16.2.1.4-4). Striking wide seams are also suspect. That is also true for unusual bad looking welds (volume 4, Ill. 16.2.1.3-2) whose appearance possibly is connected with an insufficient expertise of the welding personnel.

Coatings can be an important indicator. At first striking deviations of colour and roughness. But also must be payed attention at unusual faults and flaking (volume 4, Ill. 16.2.1.8.2-7). These could be a hint at adhesion problems and with this at deviations of the production process. Possibly in connection with a not certified producer. But also approved producers are not above suspicion. So some time ago a case emerged where turbine blades have been only painted instead of a (diffusion?) coating.

Assembly bahaviour: Among this abnormalities and problems during pusthing together parts/components should be understood. Close tolerances, even in the prescribed tolerance field, can increase assembling forces markedly and so attract attention of the technician, if he did not experience this effect in such an extent till now (Fig. "Failures of shrink fitting"). Also the behaviour of sliding respectively rubbing caused by a changed mashining structure can trigger such effects.

Magnetic characteristics will be noticed rather by chance. Possibly magnetic tools or small parts will stick at a component which rather should be made from a non-magnetic Ni alloy or Ti alloy. Contrary a part known as magnetic can show this feature not so pronounced as estimated. Such effects are indications for a material aberration and with this for a illegal part.

Labelings, markings as „identity card” of a part must not show suspicious deviations (Fig. "Identification of not approved parts"). Attention schould be payed at addings/appendices which can suggest rework or changed versions. Also the marking process can give hints (volume 4, chapter 17.4). Alarming should be if the marking is positioned at an unusual location.

Packings/wrappings are in fact only an indirect indication for the acceptance of the contained part. However aberrances from the usual are worth at least a request. Are new parts concerned the intactness of the original packing is important. Unusual features like noticeable protection oil for storage (e.g., smell, colour, consistency) should also attract attention.

Figure "Suspicious packing indicating at SUP": Already the packing can be suspicious. Naturally this is only possible in comparison with the usual and approved.

Figure "SUP features of a friction beating" (Lit. 20.2.1-14, Lit. 20.2.1-15 and Lit. 20.2.1-16): Anti friction bearings for aeroengines are markedly more expensive as such for cars or the common engineering. So the motivation can be corresponding high to use this cost advantage.

To achieve the high reliability, especially the highest attention must be payed at the pureness of the material for the rolling elements and the races/rings (diagram below left). It can be recognised that the life time drops exponential with decreasing purity grade, that means increasing number of impureness particles. Also the heat treatment plays a major role (diagram below right). Especially the scatter of the lifetime is markedly influenced. Thus it is possible, that the likelihood of a bearing failure with extensive, safety relevant and costly consequences increases.

Already this especially high material quality influences the price. Additionally for the rolling process alloys of the highest loading capacity are used. Just the production of the races demands optimised parameters and processes. For example a state of internal stresses as beneficial as possible and the required surface quality must be guaranteed. Those features can not be recognised at parts, approved for aeronautical use, compared with less expensive bearings. Differences, if existing, are limited therefore at outwardnesses (Ill above). They can merely produce indirect suspicion but don't represent a technical deficit. To identify such aberrances from the „usual” especial experience a specialist knowledge is essential.

Figure "SUP bolts with deviations" (Lit. 20.2.1-17 and Lit. 20.2.1-18): Bolts and nuts can be `called bulk parts' or `chute parts'. They are especially prone to get a SUP.

The problem is then obviously if the bolt already fails during tightening. For example in some cases the bolt head separates. Experienced technicians can identify a blolt as SUP because of its unnormal elongation behaviour. However it is quite possible that the bolt shows no anomalies, but anyway has not the necessary operation strength.

Of special influence on the vibration fatigue strength of a bolt is the production of the head and the thread. Are the critical zones plastically deformed strain hardening and internal compresson stresses can markedly increase the fatigue strength (diagramm below left). This effect is used especially at high strength bolts of aeroengine technology.

The „fibre” (orientated material structure) of an upset forged bolt head (sketch left) follows nearly the transition radius to the shaft (detail). That can be estimated beneficial for the vibration fatigue strength. In contrast the `broached' fibre is not optimal (detail) like it is typical for bar stock (sketch right). Here the `fibres' can act as weak points of the especially high stressed surface (notch effect, bending) and promot a vibration fatigue crack (Fig. "Appearance of bolt operation failures").

A rolled thread behaves positive (sketch and detail right). The influence of the roller application on the strength ot the threads hows the diagram below right. Especially the root radius with its notch stresses is of importance for the vibration fatigue strength of the bolt. It can especially benefit from rolling.

Figure "Identification of not approved parts" (Lit. 20.2.1-10): In the cases sketched obove a technician noticed during a routine overhaul an unusual discolouration of the distance ring from a main shaft. Thus to notice this the normal look of the part must be aware from experience. This needs the necessary specialist knowledge about the function of the part in the aeroengine and the consequences of a possible failure.

This part is highly safety relevant. If a malfunction or failure occurres the lubrication oil can be contaminated (e.g., wear chips) up to the fracture of a turbine shaft, triggering the total destruction of the aeroengine by overspeed. An investigation at the OEM showed, that a counterfeit part was concerned.

The lower sketches show an example where the SUP was discovered due to the faulty marking. In this case the correct stamp of the OEM lacked.

Figure "Identifying during assembly a too soft bogus nut" (Lit. 20.2.1-19): The technician noticed one of 200 nuts during the assembly of the low pressure compressor rotor (detail below left) of this turboprop-aeroengine (upper sketch). Its colour differed from the usual nuts (sketches in the middle). He reported his observation and thereupon the nut was investigated. During a tightening test the thread was pulled out far below the claimed fastening
torque. That means the threads have been sheared off (detail below right). A material test showed, that the nut was made instead from the specified high strength steel of a much softer steel.

Fails respectively loosens such a bolting, heavy vibrations, blade fractures and breakdown of the aeroengine must be expected.

Because of this risk all suspect, already mounted aeroengines of these military cargo aircrafts had to be checked. Because it was necessary, all aeroengines had to be opened and so a large expenditure of time and cost effort arose.

The suspect nuts have been so called `parts provided'. These purchases the military operator at thr market and provides it to the assembly. In the case at hand the parts originated from a certified dealer. Obviously he had to bear the juridicial consequences after this case.

Fig. "Accident by a bogus bolt" (Lit. 20.2.1-21): After the accident the aeroengine (middle sketch) was demounted and the fuel control unit was sent to the producer for an investigation. The producer replaces the bolts at every overhaul. There it was determined, that the bolthead of the bolting of one cover with the duct for an actuating rod was broken off. It clung at he locking wire which was connected with the cover (frame below). The sealing O-ring in the bearing surfac of the flange was damaged in the region of the broken bolt.

A microscopic investigation of the fracture surface showed distinct featues of hydrogen embrittlement (volume 1, Ill. 5.4.4.1-2).

The embrittlement developed with high probability during the cadmium plating (volume 4, Ill. 16.2.1.8.3-7). It is known, that during this galvanic process hydrogen is absorbed in dangerous quantities. The embrittlement effect must be avoided with a short term followed heat treatment (de-embrittlement, volume 1, Ill. 5.4.4.1-7). Is this neglected it can come to delayed brittle fractures also not until a long time period. The embrittling effect gets more pronounced with increasing strength of the bolt material.

The material investigation showed, that the hardness/strength of the bolt was significant higher than as maximum acceptable specified. Additionally the transition from the shaft to the bolthead acts not corresponding with the drawing.

It proved, that the concerned broken bolt and further bolts at the control unit are SUPs. The producer of the control units could assign the bolt to a larger delivery. He got it about 5 years ago from a sub-supplier. The concerned control unit was assembled 1 year after the supply. The inspection of the whole shipment showed, that all bolts lay outside the drawing requirements. This demanded a heat treatment (not to confuse with the de-embrittlement) to lower the hardness. This was documented by an `X' as mark on the boltheads. If this marking lacked at the broken boltm can not be seen from the available literature.

Figure "Identifying not approved parts" (20.2.1-10 und Lit. 20.2.1-22): Besides the in Ill 20.2.1-3.1 described outer identification features especially the accompanying documents (Fig. "Documents for approved parts 1") can give informations about the specification compatibility of a part/component. This illustration shows besides the noticeable price a summary of shipping documents and its information value for a backtracing.

Figure "Documents for approved parts 1" (20.2.1-11 ): The parts manufacturer/producer and the maintainer need approvals (Fig. "Requirements for repair") respectively documents. These are primarily prescribed from the responsible aeronautical authority for the supply of airtraffic parts/components. They are also necessary for the backtracing and/or the identification of suspect parts. This summary tries a clear view of the `document practice'.

Figure "Documents for approved parts 2" (20.2.1-10, Lit. 20.2.1-23 und Lit. 20.2.1-24): As certification of a manufacturer/producer respectively a repair company in the aviation a Parts Manufacturing Approval is needed. Thereby it must be regarded that not-OEM-companies own no licenses respectively no rights of the OEM. This is for example also true for the use of drawings and process specifications of the OEM. Here it is easy to blunder into a „SUP tiwilight zone”. In cases of doubt the aviation authorities (e.g., FAA and JAA) will provide information. For example the FAA edited for this the form 8120-11 (Fig. "Maintenance and repair FAA forms"). It guides the questioner so, that no further enquiry is necessary.

In cases of doubt it can be advantageous to try to ask the OEM if characteristics of the aeroengine are known, from which the part (serial nummer) originates. For example if it was involved in a flight accident with ground contact the condition may be at least questionable. This is especially true for rotor components like disks, rings and shafts.