Aeroengine Safety

Fig. "Bolts and rivets" (Lit. 23.3.1.1-16): Service influences at bolts can get failure effective as a mechanical load in combination with the drop in strength.
Mechanical load acts at a bolt in height and type (bending, tension, shear, torsion). Different to rivets, bolts are not intended for shear loads. Therefore they also should not be used for such a retaining effect. Bending of bolts by a onesided load can be prevented by design. An example is the bolting of axial flanges from pressurized casings. In one case, from the inside a V-bar is counter tensioned (sketch below left). Also high g-loads at flange boltings from rotors can produce markedly bending or shear from corbelling masses like the nut and the bolt shaft of bolt head (sketch above middle). Similar effective act different thermal expansions and elastic, load depending strain of the flanges at the bolting (sketch above right).
Corrosion deteriorates in different ways. Develop pits (pitting corrosion), its notch effect is especially for a dynamic load a problem. Bolt materials of the type 13% Cr steel tend to this failure mode during stand still and watery electrolyte (condensate, sea atmosphere), without sufficient corrosion protection.
For high alloyed austenitic bolt materials (e.g., A286) and Ni alloys (e.g., Waspalloy, IN718, Nimonic alloys) the danger of sulfidation exists. Thereby silver can act promoting (volume 3, Ill. 12.4-14). The sulfur not seldom comes from unsuitable, sulfur containing lubrication greases/pastes (MoS₂ , Fig. "Approved lubrication media").

Oxidation is a main problem at bolts and hot parts. It causes after longer operation periods to stuck the bolts respectively nuts. During loosening of the connection, in spite the use of high temperature lubricants, the deterioration danger of the thread and/or of a torsion overload of the shaft with crack formation up to the fracture exists (Fig. "Jamming causes of bolts and nuts"). This excludes in many cases a reuse. Therefore as a precaution the affected bolt sets are generally exchanged during the assembly. This is caused from the insufficient surem non destructive testability. Typical example are bolts of flanges at turbine rotors.
Oxidation can also have a positive aspect. It offers threads of bolts/nuts from austenitic materials (Ni alloys, high alloyed steels) a certain protection against seizing /galling. In hot parts the former usual silver plating as „lubricant” will more and more be abandoned, because of the deterioration danger. With this, the danger during tensioning of metallic bare threads from new bolts arises. Here a targeted preoxidation in air in an oven at operation temperature can find a remedy.

Fretting wear occurs everywhere, where it comes to micromovements between the components of the bolting. The so developing fresh metallic surfaces and the adrasion are influenced from oxidation (volume 2, Ill. 6.1-4). This can accelerate the wear. In contrast, oxides with the effect of a dry lubricant (volume 2, Ill. 6.1-18) can minimize the wear. Basically the fatigue strength of the usual bolt materials is not so affected by fretting as titanium alloys (volume 2, Ill. 6.1-8).

Liquid metal embrittlement (LME) is a crack formation/embrittlement during sufficient high tension stresses and influence of wetting metal melts (volume 4, Ill. 16.2.2.3-10.1, -10.2, -10.3 and Ill. 16.2.2.3-11). Silver (Fig. "Bolt fracture at hot parts by silver") and cadmium are especially dangerous. This applies also for a deterioration in the solid condition by diffusion (SMIE). This danger decreases markedly at preoxidized/oxidized surfaces, because a wetting contact to the melt is prevented. Influence of foreign objects/force especially occurs as secondary failure at flange boltings from rotors (Fig. "Identifying fractured bolts at rotors"). For example, if during assembly a tool was forgotten (Fig. "Foreign object remained by assembly") or the fracture of a bolt happened (Fig. "Identifying fractured bolts at rotors"). A further possibility is the damage during axial rubbing of the bolt heads or of the nuts.

Material changes during the service can be caused from aging, overheating or diffusion of foreign material. These may be only expected during unusual operation conditions. The selection of the bolt material should certainly prevent such deteriorations.

Fig. "Operation loads of bolts" (Lit. 23.3.1.1-1 up to Lit. 23.3.1.1-4): A bolt connection without forces from the outside is a internal stress stytem. In it the stresses are in a balance (volume 4, Ill. 16.2.2.4-1 and Ill. 16.2.2.4-2). From the outside acting operation forces change this condition, respectively the internal stresses, until balance exists. The bolt gets the tensile loading through the support/flange. It takes the internal stresses from the pretension. As long as this is the case, only a part of the service load acts at the bolt (diagrams above). Is the pretension fully compensated, e.g., the flanges begin to lift off, further increases of the load act fully at the bolt (diagramm above left). So a lower pretension of the bolt causes a higher operation stress of the bolt (Fig. "Avoiding fatigue cracks at bolts"). This especially becomes noticable at dynamically high loaded bolts of the aeroengine design.
Is the bolt, according to its strength, sufficient high prestressed (diagram above right), dynamic stesses can kept sufficient low. Regarding the prestress of the bolt as mean stress, leads its lowering to a markedly higher usable fatigue strength of the bolt material (diagram below). So it is very important, that the in the instructions specified tightening torque, determined from the designer, is kept.

To guarantee the necessary pretension, it must be payed attention at the required lubrication conditions (Fig. "Influence of pretension at fatigue"). Besides the lubricant also the coating plays an important role. It must correlate absolutely the specifications. To guarantee, during unusual high loads, sufficient pretension, reduced bolts/tension screws are used. The relatively low spring stiffness is reached with a waisted shank/reduced shaft (sketch middle right). Typical example of use are `containment loaded' (by impact of fragments) flange boltings (volume 2, Ill. 8.2-15) of casings. Similar bolt connections are also needed at extreme rub processes, e.g. during unbalances from a rotor blade failure (volume 2, Ill. 8.1-15).

Fig. "Influence of pretension at fatigue" (Lit. 23.3.1.1-2, Lit. 23.3.1.1-8 and Lit. 23.3.1.1-16): The efficiency of a bolt/screw is determined by the residual share of the torque for the pretension. These influences can let with the same torque the pretension scatter up to 80 % (Fig. "Influence at the bolting durability"). For this reason it is especially important, to meet exactly the specified procedures during the bolting. So during the design planned minimum pretension can be guaranteed.

Fig. "friction forces on bolt connections" (Lit. 23.3.1.1-2 and Lit. 23.3.1.1-8): The load transmission inside a thread can be simplified, compared with a sloping level. Then the pretension force produced in the thread, correlates a weight in axial/vertical direcction. A circumferential force by the tightening torque, loads the sloping level, e.g. the thread flank (sketch left). In the opposite direction acts the friction. The remaining force, respectively the related tensioning moment, produces the axial pretension in the bolt.
The friction in the thread prevents a loosening/untwisting of the bolt connection. Thereby the clamming friction force is higher than the untwisting component from the pretension. Such a self-locking is given at typical thread pitch of bolts. at least for static loads.
A vibration of the bolting can markedly lower the frictional coefficient of the flanks. This effect is similar a vibrating inclined plane. Is thereby the self-locking canceled, loosens the bolting (sketch right, Fig. "Locking effect at bolting connections").
Because of this reason, actuated by adherend self-locking nuts (e.g., by locally ovalisation or springy tongues) are used. A further possibility are form fitting lockings. For example for shaft nuts bendable locking ears (Fig. "Anti twist devices for shaft nuts") are used. A further example are locking wires (Fig. "Loosening of bolting from hot parts").

Fig. "Locking effect at bolting connections" (Lit. 23.3.1.1-8 and Lit. 23.3.1.1-11): Boltings of aeroengine components underlie conditions which lead to a loosening during operation. Already a decreasing pretension can heavily increase a dynamic load if the flanges lift off (Fig. "Operation loads of bolts"). With this the danger of a fatigue fracture rises (Fig. "Appearance of bolt operation failures").
In an extreme case, an unsufficient secured nut can separate. Then the danger of heavy foreign/ own object damages (OOD, DOD) exists.
Following influences can cause a dangerous drop of the bolt pretension:

Untwisting can occur due to unsufficient securing under heavy vibrations (Fig. "friction forces on bolt connections" and Fig. "Loosening of bolting from hot parts").

Setting of the bolting by yielding of the supports. Concerned is a locally plastic deformation caused b

• especially high roughness tips or foreign objects between the contact surfaces, e.g., little metallic chips.
• Formfitting of edges. This is the case if transition radii to the bolt head seat on the bore edge.
• Yielding of a locking profile on the head support. Toothening similar structures which indent as locking into the flange, can further, indent, e.g., during higher operation temperatures or vibrations (see „creep“).
• Slipping, if the thread has unnoticed seized before reaching the demanded support surface (Fig. "Influence of shrink fitting with spline toothing 1" up to -17.3). A similar situation develops, if the flanges before or during bolting, have seized on a centering diameter.

Yielding of threads and/or thread inserts in light metal alloys (Fig. "Bolt connections in light metal castings").

Wear can promote a loosening at all contact surfaces, which influence the pretension. As main mechanism during vibration wear (fretting), micromovements play a role. However a dangerous loosening by wear of the bolting seems to be extremely seldom, at least for flange connections. Rather problems at pipe clamps and its bolt fixtures are to expect (Fig. "Resonance vibration caused by P-clamps" and example 23.5.1-2).

Crack formation at the shaft or the thread from the bolt. Cracks can have different causes, which also appear in combination:

• Production caused (Fig. "Brittle failure modes of bolts and nuts").
• Sulfidation (volume 3, Ill. 12.4-14), diffusion (Fig. "Evaluation of the risk of a broken bolt"), LME (volume 4, Ill. 16.2.2.3-10.1 and Ill. 16.2.2.3-11).
• Vibratiuon fatigue (Fig. "Appearance of bolt operation failures").
• Crack corrosion (stress corrosion cracking, Ill. 23.2.1.2-2 and volume 1, Ill. 5.4.2.2-1).
• Hydrogen embrittlement (Fig. "Brittle failure modes of bolts and nuts", volume 1, Ill. 5.4.4.2-1, volume 4, Ill. 16.2.1.7-14 and 16.2.1.8.3-7).
  

Creep (volume 3, Ill. 12.5-3) is a plastic (lasting) deformation (elongation, compression) during material specific elevated temperature (Ill. 12.5-1). This yielding can occur as well at the bolt (shaft, thread) as also in the thread of a blind hole or in the nut (Fig. "Bolt connections in light metal castings"). Also a yielding of the flanges or of shims can have the same effect.

Differences in thermal expansion are possible during different temperature distribution in the bolting, e.g., during nonsteady operation. Such influences are to expect especially at boltings in the hot gas stream (Fig. "Loosening of bolting from hot parts").
A further possibility is an unfavourable material combination/thermal expansion coefficient between flange material and bolt shaft. Depending of the combination during heating or cooling, prestress can be lowered. Typical problem zones are axial pretensions of rotors, especially for centric tension bolts (Fig. "Tightening torque surface and lubrication", volume 3, Ill. 12.6.3.3-6).

Drop of the E-modulus at temperature (volume 3, ill. 12.4-1). This applies for flanges and boltings. A lowering in stiffness acts at pretension, vibration and micro movements.

Fig. "Bolt prestress decrease during service": From experience, irritations can arise with bolts, which are taken from the storage. This may be the case, if the specified tightening torque is not reached, because the bolt deformed unusual during tightening, respectively deformed plastically unacceptable. If this behaviour can not be assigned the bolt quality, its assembling condition must be checked. Thereby the instructions in the manual must be exactly respected. There are cases, in which an additional lubrication of the thread is not sheduled. For example, shows the thread remains of conservation oil, which before the screwing should have been removed, this can trigger the undesirable effect.

Fig. "Influence at the bolting durability" (Lit. 23.2.1.1-8 bis Lit. 23.3.1.1-10 ): The achievable prestress of a bolt depends crucial from the „efficiency” of the thread and with this from the friction forces. These are essential determined from the tribological conditions between the thread of nut and bolt:

• Surface quality of the tread flanks.
• Strength of the nut and the bolt (formabilityof the flanks).
• Proneness for seizing.
• Coatings/layers/platings.
• Oxidation/oxide layers.
• Lubricant.

The chart contains relative friction coefficients. 1,00 applies for untreated suefaces. The markedly drop of the friction during the use of a lubricant can be seen. Also an additional coating or the change of a coating can influence the bolt efficiency markedly.So an increase of the friction coefficient about 30 % (gray area in the chart) must be considered in the tightening torque. For this reason a deviation from the coatings, specified by the OEM, is not accptable.
Also a not scheduled additional lubrication can act deteriorating (Fig. "Bolt prestress decrease during service"). In this case, for the bolt exists the danger of overload during tensioning.

Fig. "Tightening torque surface and lubrication": Tension bolts for the axial tensioning of rotors must keep up a minimum pretension also during instationary operatio. This guarantees an exact specified pretensioning process. With this the rotor gets the necessary stiffness and so a vibration behavior, according to the design. The more elastic the pretensioning, i.e. the longer and thinner the elongated length, the better thermal caused expansion differences can be compensated (Fig. "Locking effect at bolting connections", volume 3, Ill. 12.6.3.3-6).
Thereby the long bolt shaft gets not unacceptable twisted during tightening of the nut, if at the side of the nut, spanner flats are applied (mostly square-end). These enable a backing.

Outward located tension bolts (sketch left) can better follow temperature changes as a centric tension bolt (sketch right). At centric tension bolts in the turbine area, the temoerature difference between rim and interior zone (hub) is especially large (volume 3, Ill. 11.2.3.1-8 ). Drops the pretension too much, failures can occur (volume 3, Ill. 12.6.3.3-6). This is the case, when for example during shut down of the aeroengine, the blading and the rim cool down very fast and the relative massive inner zone of the trurbine wheels is much more tardy (volume 3, Ill. 11.2.3.1-8).

Fig. "Rotor tension bolts" (Lit. 23.3.1.1-5, Lit. 23.3.1.1-6, Lit. 23.3.1.1-7 and Lit. 23.3.1.1-16): During tensioning, according the feeling, the scatter band grows about the uncertainty of the tightening torque. From experience the pretension of thinner bolts (e.g., M12) rather above the yield point. In contrast thicker bolts are too low tensioned (e.g., M14). This can be avoided with torque control tightening („A“ ). Usually for this, a torque wrench is used. This can be, depending from the specification, displaying and/or triggering. Of course, such wrenchs must be calibrated as specified. Anyway, caused by the friction forces, for the pretensioning only remains a usuable fraction of about 10% („A”, chart left). Even if the scatter of the torque display lays below ±10%, it is understandable that this can lead due to relative little variations of the friction forces to a large scatter range of the prestress of up to 80%. This correlates a tightening factor (maximum prestress /minimum prestress) of 1,6 bis 1.8. This can be not acceptable for the high loaded boltings of the aeroengine technology.

The meaning of the minimum pretension requires in critical cases a controlled tensioning process. For this several processes are available, which however demand requirements.
Impulse controlled tensioning is carried out with an impact wrench. Here also the tightening torque is adjustable respectively possible to trigger.
This procedure has the advantage,to trigger no response moment. However it underlies many influences at the achieved pretension:

• Elasticity and friction coefficient of thebolting.
• Elasticity of the wrench and the system.
• Strike force and frequency.
• Tightening period respectively number ofimpacts.

This leads to a very bad tightening factor in the region of 2,5-4,0. With an optimizing 1,5-2,5 can be reached. This may not be enough for many cases. Therefore such tools are rather used at the disassembly for the loosening of boltings.

Extension controlled tightening. The elongation is under consideration of the bolt strength direct proportional to the bolt pretension. Thereby the material (modulus of elasticity) and the cross section course assume as a constant. The larger the length of elongation (clamping length), the more exactly is the process. Accuracies better than +-5 % can be expected. This correlates a very good tightening factor of 1,2. Therefore it is predestined for the pretension adjustment of the tension bolts/rods of rotors (Fig. "Tightening torque surface and lubrication"). Precondition for this demanding process is a good accessibility (frame below).
The elongation can also be identified with ultrasonic sound from one side („B2“). Thereby advantage can be taken, influencing the runtime of the sound in the bolt shaft by the tightening process (acusto-elastic effect, Lit. 23.3.1.1-7). This is markedly larger than it correlates the pure elongation (diagram at the left side above right). Requirement for the measurement, i.e. sound reflextion and coupling/contacting are suitable end faces (grinding surfaces) at the head and the bolt end. An advantage of the process is the continuous measuring of the pretension during the tightening process.
During the angle controlled tightening („C”), the bolt is first torque controlled tightened (torque wrench). This is followed by a tightening with a calculated angle of twist until to the beginning plastification. This so called strain controlled tightening (yield controlled tightening) let expect an excellent tightening factotr of 1.0, because the yield point (0,2% plastic elongation) is certeinly reached.
Is the tightening torque above the angle of twist continiually monitored, the reaching of the yield point (also elastic limit, elastic limit) shows with the flattening of the curve. This effect can be used for an electronic control.
Twist angle controlled processes demand an application specific testing and require a sufficient elongation length. Additionally the plastic deformation can exclude a reuse.

Fig. "Type of screw locking" (Lit. 23.3.1.1-3 and Lit. 23.3.1.1-12): Locking devices for bolts and nuts must correlate the specifications in the manual. Unauthorized changed, respectively additional locking measures like adhesives at the thread (Fig. "Problems with additional bolt locking") or a safety wire to a self locking (threaded insert, counter nut) can rather lower the safety.
The pinched off too long ends of a locking wire, can be a special problem. Are these not held as prescribed and so secured against jumping off (Fig. "Assembly as FOD risk"), the danger exists, that these fall in not covered openings of the oil system. As foreign objects, they trigger extensive failures, e.g., in the compressor (volume 1, Ill. 5.2.1.1-6 and Ill. 5.2.1.3-5) or in the oil system.
During application of mechanical lockings like locking wires or shims, the specified procedure must be used. The laying of locking wires must correlate the locking effect. The wires must not be dangerously weakened or get a critical deterioration during service (fretting). They also must not damage other components like pipe lines by fretting (Ill. 23.5.2-5). Damages or unfavourable use of a locking/securing element during deformation can cause cracks and/or fractures (Fig. "Anti twist devices for shaft nuts"). For example a broken locking lug (sketch below right, Fig. "Locks and securings of bolts") can trigger in the gear itself heavy secondary failures, if it gets into the toothing.
Self locking/securing nuts (sketch above left) are, corresponding the introductions in the manual, reusable. From many a time use, the function of the self securing can be deteriorated by wear and/or deformation.

Fig. "Anti twist devices for shaft nuts" (Lit. 23.3.1.1-13): For the fixing of components on shafts frequently can locks are used (sketch below, Fig. "Locks and securings of bolts"). With these, components like gear wheels or rotor disks are tensioned and fixed on the shaft. These nuts have a can lock from sheet metal.This has since long proven in the aeroengine technology. However, during operation exist potential problems, which must be considered. At their symptoms must be payed attention during overhaul.

The securing/locking naturally demands a suitable `activation'. This usually takes place with a plastic deformation. Thereby a locking lug is bended into a notch (sketch middle) or produces with the specified tool an indentation (sketch above). In this case it must be payed attention, that on one hand a sufficient large deformation guarantees the necessary securing effect. On the other hand no cracks or unacceptable weakening damages like notches may occur. These can be the start of fractures by vibration fatigue. Thereby the locking effect is lost and foreign object damages/own object damages from the fragments must be feared.

Fig. "Jamming causes of bolts and nuts" (Lit. 23.3.1.1-8 and Lit. 23.3.1-14): Stuck/jammed bolts and nuts are especially at hot parts in aeroengines a frequent phenomenon. This can have different causes, which also can occur in combination.
Austenitic materials like Ni alloys and high-alloyed CrNi steels and especially titanium alloys tend during tightening or loosening of bolts and nuts to seizing/galling (Ill. 23.3.2.1-1). Cause is the very thin natural oxide film, which is destroyed from the sliding process under high surface pressure. So it comes to an intense contact of fresh reactive metal surfaces. Such a fusion/welding also occurs at the supporting contact surface of nut and bolt head especially with titanium alloys and Ni alloys. Can first signs of seizure be noticed, every further tightening will increase the deterioration. A typical feature is, that also the untwisting is no more possible. In such a case, the connecrion must be careful loosened and is with unaccptable damages no more usable.

Oxidation („freezing“) appears in the threads of hot parts boltings. The oxides have a larger volume than the base material. This leads to a clamping/jamming effect. Do the oxides have bad sliding properties, this effect can still increase. Unsuitable lubricants can promote the blocking. Corrosion („seizing up by rust”) occurs especially at boltings of steel in the colder region. Here condensate can form and sea atmosphere can act. Also the corrosion may play a role during the growth of the oxide volume. Unsuitable auxilary material, which for example disintegrate at operation temperature (e.g., MoS₂ containing), can even aggravate the corrosion (Fig. "Approved lubrication media").

Fretting (vibration wear) can form on contact surfaces during micro movements. For example such appearances can be found after longer operation periods also at the contact surfaces of boltings from of rotors and flanges. Develops inside the threads oxidizing abrasive (fretting wear, `friction rust', Fig. "Lubrication caused fretting wear", volume 2, Ill. 6.1-3), this can act like corrosion products.

Often a `breakaway' of nuts is possible, however without a total twisting off. This may have several causes, which also intensify in combination:

• Corrosion products, oxides and contaminations with bad sliding properties accumulate during twisting between the flanks of the threads. If such deposits experience through the chemical reactions a volume increase, the clamping effect is intensified.
• Seizing of the thread flanks through a metallic contact during hindered untwisting.
• Stretching of highly loaded thread pitches. Such a pitch error at the nut can cause locally flank overload and seizing.

Remedies against jamming/sticking of nuts and bolts:
Appropriate (specification conform) application of the specified lubricant. If already a coating with dry lubricant is present use only specified procedures. If seemingly necessary, use no additional lubrication.
To be considered:

• Clean contact/bearing surfaces of bolts, nuts washers and flanges. If reused in a maintenance process the bolting components must be cleaned as specified.
• Pay attention at metal chips and abrasives. These can trigger a seizing process during tightening. This is also true for contaminations like abrasive blasting material.
• Lubricant pastes must also applicated in the groove of the thread.
• Bearing surfaces of nuts and bolt heads treating according the specification. They take up to 50% of the torque.
• Preoxidation of bolts and nuts from Ni alloys at operation temperatures (in an oven) on air. The effectiveness depends from the sliding behaviour of the oxides (volume 2, Ill. 6.1-18). This oxide layer prevents metallic contact and with this a seizing of the sliding faces, especially of the thread flanks. But this treatment must conform the specifications, because it influences the effectiveness of the torque, like all influences at the friction conditions. If not explicite planned, the OEM must be consulted.

Fig. "How to loosen bolts and nuts" (Lit. 23.3.1.1-8 and Lit. 23.3.1.1-14): The loosening of sticking boltings can be facilitated with different measures. However these must always be permitted, according to the instructions in the manual.

Rust remove, mostly as spray, facilitate, at least after a residence time as well the loosening as the unscrewing of the nut. However here attention is needed. Only for the application approved removers/solvents may be used. For example with Cl containing media the danger of a deterioration of titanium components exists (volume 1, Ill. 5.4.2.1-6 and Ill. 5.4.2.1-8). At high strength steels, stress corrosion cracking can develop. MoS₂ containing media can develop at sufficient operation temperatures sulfidation at Ni alloys (volume 1, Ill. 5.4.5.1-4) or trigger under tension stresses cracking (Fig. "Unsuitable lubricant leading to fracture").

Mechanical shock loading, for example by an impact wrench (Fig. "Rotor tension bolts"). These devices are used during disassembly. Thereby the thread must not be damaged to avoid a hindering of the untwisting process.
Before the untwisting of the nut, protruding thread must be carefully cleaned from corrosion products and firm sticking lubricant remains. Heating the bolt to 50-100 °C. This decreases in many cases the friction coefficient.

Fig. "Dangers to bolts at temperatures" (Lit. 23.3.1.1-17): `High' operation temperature is a relative term. It can, depending from the tribological system and materials, already lay in the range of 200 °C or not before above 400°C. In this temperature range, potential danger of deterioration exists. Thereby failure mechanismy are concerned like:

Embrittlement by diffusion: At temperatures of 600°C, silver can diffuse into the surface of Ni alloys (volume 3, Ill. 12.4-14). For titanium alloys this process begins at about 200°C. Cadmium can get dangerous for steels and titanium alloys, already at temperatures above 200 °C (volume 4, Ill. 16.2.2.3-11). Thereby the melting point of the plating is not reached (SMIE). This is supported from tensile stresses like they appear in the bolt shaft or the thread. Has the diffusion penetrated a critical part of the bearing cross section, a brittle fracture will occur. Is a nut broken off together with the bolt shaft under these operation conditions, also the suspicion of such an embrittlement exists (Fig. "Brittle failure modes of bolts and nuts" and Fig. "Bolt fracture at hot parts by silver").

Crack formation and embrittlement through penetration of a melt: This phenomenon is called `liquid metal embrittlement' (= LME, volume 4, Ill. 16.2.2.3-11). With silver this possibility exists from about 900°C (volume 3, Ill. 12.4-4). For steels, copper is an example. Corresponding with the low melting temperature, lead can act deteriorating already at markedly lower temperatures.

Sulfidation from MoS₂ containing lubricants can act in two different ways:

• Pittings through sulfidation reduce the bearing cross sections and act as notches. Concerned are Ni alloys.
• Crack corrosion under air exclusion. This situation seems at the first sight extremely unlikely in aeroengines. However it develops in threads or at snug fits/fitting seats respectively the bolt shaft, if air entrance is prevented. Then burning will not occur but the decomposition and the release of aggresive acting sulfur. During sufficient high tension stresses, especially, in the region of notches (pores, scratches), spontaneous crack formation can occur. This is a sort of stress corrosion cracking of austenitic steels (Lit. 23.3.1.1-8) and Ni alloys. Also casting alloys like for turbine wheels are endangered (Fig. "Unsuitable lubricant leading to fracture").
  

Under water/condensate and acting of salt (sea atmosphere), MoS₂ seems to promote corrosion at steels. This can dangerous weaken the cross section, as well as making the bolt respectively the nut jamming (Fig. "Approved lubrication media").\
Concluding should be mentioned, that during stand still and acting aggressive condensate, silver can be dissolved from bolts and nuts. Does ths percipitate at operation temperature at other locations (e.g., turbine disks), there pittings can develop, triggered by sulfidation (volume 3, Ill. 12.4-14). This may be a reason besides the risk of diffusion, why OEMs use no silver plated bolts and nuts at hot parts.