Aeroengine Safety

Figure "Applications of galvanic and chemical coatings": Metallic, galvanic, and chemically applied coatings are primarily used as corrosion protection and for sliding surfaces. Therefore, as with Cr coatings, their wear resistance may be the most important factor. Typical examples include oil seal surfaces on toothed gear shafts that are subsequently ground to measure and which have floating ring seals running
along them (top right diagram). Their soft elastomer seal edges are prone to trapping hard particles such as oil carbon or dust, which can cut life-reducing grooves even into the extremely hard Cr coating. Bearing seats at risk of fretting are also equipped with Cr coatings.
Cr coatings are used on sliding surfaces in hydraulic cylinders (middle diagram) and actuators in order to lower the coefficient of friction and provide high wear resistance. The typical cracking network in these coatings can also be used as an oil reservoir, in which case this will be specified.
Fundamentally, the use of Cr and Ni coatings (including electroless nickel) is limited due to the reduction of the dynamic fatigue strength caused by high tensile residual stresses in the coating (Fig. "Effects of electroplated coatings on fatigue strength").
Some of the primary applications of galvanic coatings such as cadmium, zinc, and silver are on bolts, nuts, and washers. These coatings are used here in two ways:
1. As glide coatings on threads (silver). They are intended to ensure constant tightening torque during assembly and/or prevent locking during disassembly. Silver is also used on bolts made from Ni alloys that are exposed to high operating temperatures. This use has been reduced, however, since silver can have damaging effects under these conditions (LME, sulfidation, Ills. 16.2.1.4-14 and 16.2.2.3-13).
2. Cadmium and zinc are used as cathodic corrosion protection. Cadmium is used on steel bolts that are not subjected to high operating temperatures. Cadmium-plating causes hydrogen embrittlement (Fig. "Hydrogen embrittlement danger of bolts") and has an embrittling effect on many metals (e.g. titanium alloys and steels) at high temperatures (Fig. "Ni barrier coating preventing cadmium contact").

Figure "Effects of electroplated coatings on fatigue strength" (Ref. 16.2.1.8-7): During hard chrome plating, a reduction in the dynamic fatigue strength (Ref. 16.2.1.8-7) is to be expected. The top left diagram shows the residual stress patterns in the surface of a hard chrome-plated steel specimen (top right frame). The surprisingly low residual stresses are attributed to a relaxation of the Cr coating due to the typical cracking in the coating surface. An important factor for fatigue resistance is the stress peak in the adhesive layer of coating and base material. Dynamic fatigue cracks can be expected in this area. The influence on the fatigue resistance is shown in the bottom diagram in the top right frame. For Cr coatings, this influence is considerably less than it is for the substrate.
Hardening through plastic deformation can be used to raise the fatigue resistance again. This can be done with the aid of shot peening or rolling. It is surprising that in both nickel and Cr coatings, it is more beneficial to treat the coated surface than it is to harden the base material before coating (middle diagrams). this effect is also present with thick coatings, but is slightly less pronounced than in thinner coatings (middle left diagram).
It is also interesting that in the hardened state, the maximum fatigue resistance is found in coating thicknesses between 100mm and 200 mm. Therefore, a continual decrease in fatigue resistance with the coating thickness is not apparent. As expected, the tensile residual stresses in the Cr coating overlay with the operating stresses. This means that the influence of the mean stress on the fatigue resistance is evidently greater in chrome-plated parts than in unplated parts (bottom diagram). This effect must be taken into account when coating highly-stressed parts such as rotors (Fig. "Damage by Ni coating lowering fatigue strength" ).

Figure "Effects of electroplated coatings on fatigue strength" (Refs.16.2.1.8-5 and 16.2.1.8-7): The thickness of galvanic Cr and Ni coatings has a clear influence on the dynamic fatigue strength. However, thicker coatings do not generally have a negative effect. The top left diagram shows that, on smooth specimens, depending on the coating parameters, coatings with a thickness of about
50mm lower the fatigue resistance less than coatings of about 25 mm. The fatigue resistance loss due to Cr and Ni coatings increases with the hardness, tensile strength, and dynamic fatigue resistance of the base material. This effect seems to be more pronounced in Ni coatings than Cr coatings, due to their greater modulus of elasticity. It results in considerably higher stress levels with identical strain. The influence on the strength of the base material is due to the growing difference to the coating strength. This means that cracks will occur in the coating well before the base material is overstressed. The coating is then unable to absorb any more force, and the base material is no longer relieved by the coating. Special care is required if high-strength and, therefore usually highly stressed, materials are coated. The dynamic fatigue strength of low-strength materials is considerably less sensitive to coatings.
Dense cracking in the coating reduces the tensile residual stresses, and therefore has less of an effect on the dynamic fatigue strength. Therefore, the fatigue resistance increases with the crack density (top right diagram).
If notches are coated (bottom diagrams) with thin coatings, the loss of dynamic fatigue strength will be even less than in smooth specimens. In contrast, thick coatings are more damaging to notched specimens.
This means that: coating part zones with stress concentrations (notches) is especially risky in materials with high tensile strength.

Figure "Quality assurance by process monitoring of chromed parts" (Ref. 16.2.1.8-5): The fatigue strength of galvanic coatings, in this case hard chrome coatings, depends on a large number of parameters. The base material, part contours, and adhesive surface preparation are very important. The risk of dangerous fatigue resistance losses increases with the tensile strength/hardness. Coating areas with stress concentrations (notches) exacerbates the negative coating influence (Fig. "Effects of electroplated coatings on fatigue strength"). Hardening the surface before coating is beneficial, but it is better to harden the coated surface (Fig. "Influence of Cr and Ni coatings at fatigue strength"). The roughness of the surface and the occurrence of grooves, small burrs and sharp slivers (Fig. "Typical flaws on galvanic coatinga") influences the frequency of flaws such as pores and nodules, and therefore also the crack concentration in the coating, as well as the fatigue resistance (Fig. "Effects of electroplated coatings on fatigue strength").

Hard chrome plating process: If coating thicknesses are greater than 0.05 mm, a considerable decrease in fatigue resistance can be expected (Fig. "Influence of Cr and Ni coatings at fatigue strength"). If the coating thickness is less than 25 mm, the tensile residual stresses will primarily affect the dynamic fatigue resistance. As coating thickness increases, the influence of individual coating cracks apparently also increases. On the other hand, as crack density increases, damaging tensile stresses in the Cr coating are
broken down, increasing the fatigue resistance (Fig. "Effects of electroplated coatings on fatigue strength"). These characteristics are affected by the process parameters. The temperature of the chrome bath (electrolyte temperature) has an especially pronounced influence. Excessively high bath temperatures (above 55 °C) can create high tensile residual stresses, thus significantly reducing the fatigue resistance.

Aftertreatment: Hardening (rolling, peening) of the coated surface increases fatigue resistance the most (Fig. "Influence of Cr and Ni coatings at fatigue strength"). However, heat treatments of 100 °C to 300 °C are very problematic. The pronounced decrease in fatigue resistance is traced back to embrittlement of the coating. This effect must be considered with regard to operating behavior. In contrast, higher temperatures may have positive effects if they reduce the coating hardness.

Figure "Typical flaws on galvanic coatinga" (Ref. 16.2.1.8-16): Flaws such as pores and nodules play an important role in Cr coatings, especially precision chrome plating. Nodules can cause notch effects on contact surfaces and/or lead to leakages in sealing surfaces.
Pores form around non-conductive material components and etching pits. Nodules form at sharp slivers on pre-coated surfaces that have been machined (grinding; top left diagram). These act like edges to concentrate the electrical field and lead to increased precipitation. Holes and inner edges have the opposite effect due to a shielding effect (bottom frame).
Even if a surface is optically bright as though polished, it does not ensure that nodules will not occur. If the etching process to remove residual burrs is too short, experience has shown that they will stand up again. This can be explained as follows: The slivers that were torn up by the machining process are pressed down again. When they spring back, strong compressive stresses are induced in their outer zone. Etching removes this outer zone with its compressive stresses. This changes the stress equilibrium so that the sliver stands up again (top right frame).

Figure "Importance of galvanic processes for safe operating": There are two steps in galvanic coating processes that have the greatest influence on the operating safety of a part. These are the masking process against coating and the contacting of the electrical connections.
Masking: Galvanic coatings such as chrome and nickel reduce the life-determining cyclical strength (LCF) and HCF dynamic strength (Ills.16.2.1.8.3-2, 16.2.1.8.3-3 and 16.2.1.8.3-9). Cadmium can lead to embrittlement and cracking through hydrogen (Fig. "Hydrogen embrittlement danger of bolts") or, if it is in contact, SMIE (=Solid Metal Induced Embrittlement). At higher temperatures, molten cadmium can cause LME (= LMIE = Liquid Metal Induced Embrittlement; Ills. 16.2.2.3-10.1 and 16.2.1.8.3-8). This means that reliable masking of critical part zones is highly relevant to safety. Masking is usually done with organic coatings such as waxes and lacquers, resins, or reusable masks made from elastomers such as rubber. Due to their relevance to safety, masking methods require part-specific development and testing. Insufficient contact at edges, pores leading to the base material, and mechanical damages (top frame) are typical weak points of maskings. These flaws must be prevented both during application of the masking and during subsequent handling of the parts. Reusable maskings must be inspected for signs of wear or aging that could affect the masking effect.

Electrical contacts: Given the high currents being transmitted, insufficient electrical contacts can cause local overheating due to contact resistance or arc burning (Chapter 16.2.2.6). Contacts must be affixed securely at non-critical part zones (Ref. 16.2.1.8.3-17, bottom frame). Bores are not suitable electrical contacts because they are highly-stressed part zones! Unfortunately, the highly stressed bores of rotor disks are often seen as contact points because it is possible to run a bolt through them.
The suitability of a contact zone should be confirmed by the department responsible for part safety (strength, design). It is also important that the contact equipment is not damaged or worn. Current-carrying cables must not have any flaws such as areas that are worn and no longer guarantee sufficient insulation. If current-carrying conductors penetrate the insulation, they can cause dangerous arc burning if they come into contact with the part surfaces.
The contact surfaces must be flat, smooth, and clean (bright metal). Fundamentally, electrical contacts are only permitted on specified part zones using specified equipment.

Figure "Hydrogen embrittlement danger of bolts" (Ref. 16.2.1.8-20): As indicated by the damage symptoms, this accident developed from the fracture of a bolt head (bottom frame) due to hydrogen embrittlement. This allowed fuel to escape, causing a fire.
The hydrogen was probably released during the Cd coating process, and was evidently not dealt with using an embrittlement process. In addition, the inspection revealed that the broken bolt, as well as neighboring bolts, did not correspond to specifications and were considerably too hard. The parts were unrated (bogus parts). The greater the hardness/strength of the bolt material (heat treated steel), the more sensitive it is to hydrogen embrittlement (see Volume 1, Ill. 5.4.4.1-5).

Figure "Ni barrier coating preventing cadmium contact" (Ref. 16.2.1.8-19): A compressor disk (right frame), probably made from martensitic steel, fractured during takeoff. The subsequent inspection revealed that the disk fracture originated in a crack on the inside of the annulus. Cd was found in the crack initiation zone in contact with the disk material. This evidently led to cadmium embrittlement. The cadmium evidently came from the NiCd coating that was used for corrosion proctection on the hub. Further investigations led to the conclusion that an insufficiently thick intermediate Ni layer had been applied. It was unable to prevent direct contact between the base material and the subsequent Cd coating during the diffusion process. It can be assumed that the cadmium embrittlement involved the diffusion of solid cadmium in contact or the inclusion of molten cadmium (LME, aka LMIE, Ills. 16.2.2.3-10.3 and 16.2.2.3-11).
In general, it must be noted that in the case of NiCd coatings, there is a risk of contact between Cd and the disk material. Therefore, this coating should be replaced by unproblematic alternatives.

Figure "Danger of a galvanic repair process" (Ref. 16.2.1.8-18): Shortly after takeoff, the 2nd stage high-pressure turbine disk (right frame) failed with uncontained fragments. The turbine housing was breached and there were extensive damages to the turbine.
Investigations revealed that the turbine disk had a unique configuration and originated in an early stage of the production period. It was reworked for later use by removing a flange in order to affix a centering collar with a guide lug. This was done in order to install a labyrinth spacer ring onto a fixed receiver (Fig. "Damage by Ni coating lowering fatigue strength"). The contact surfaces required nickel coating. The disk had previously been installed in a flight testing engine that had twice registered overtemperatures. Subsequent penetrant testing revealed cracking around the nickel coating. However, these findings were mistaken for penetrant residue on the coating transition. Therefore, it was decided to rework these zones and approve the disk for reuse.
Laboratory investigations revealed that the final fracture originated in a 14 cm circumferential crack (bottom left diagram) along the contact zone of the disk and the labyrinth ring. Multiple cracks were found in the nickel-coated area. In addition, the nickel coating had pronounced galling. The cracks in the nickel coating that spread into the base material of the disk were caused by dynamic fatigue. Investigations using part-specific specimens showed that the dynamic fatigue strength had been reduced by half relative to the uncoated disk material (Fig. "Damage by Ni coating lowering fatigue strength").

Figure "Damage by Ni coating lowering fatigue strength" (Ref. 16.2.1.8-18): This diagram explains the dynamic fatigue strength reduction caused by a nickel coating, which resulted in the disk failure described in Fig. "Danger of a galvanic repair process". The crack in the disk originated in the transition zone around cracks in a galvanically applied nickel coating (diagram). The damage was promoted by the fact that the location of the coating and transition on the part made crack detection with the aid of penetrant testing difficult. This led to misinterpretation of the findings (Fig. "Failed detection of large cracks"). Investigations using specimens that corresponded to the conditions of the damage case revealed that a nickel coating with a thickness of 1 mm reduces the dynamic fatigue strength to a mere 50% of that of the uncoated material (diagram).

Figure "Cracking in steels processed in burnishing baths": The case-hardened steels commonly used in gearboxes (shafts, gears) in turbine engines are usually burnished. In this process, a hot alkaline solution (NaOH + NaNO₃ ) creates a thin brown coating, the positive effects (oil storage, corrosion protection) of which are questionable. There have been reported cases in which centimeter-long cracks were found in these parts after burnishing. These cracks showed signs of stress corrosion cracking. If the influence of the alkaline solution is significant, it is known as caustic embrittlement. This occurs if there are sufficiently strong tensile stresses. In order to prevent this from happening, parts are shot peened before burnishing. The induced compressive stresses provide adequate protection. However, problems can arise in part zones that are not suitable for shot peening, such as the spline toothing of gear shafts. The corners in the grooves act as notches and are difficult to reach with the peening shot. Therefore, they are a weak point for burnishing cracks. In addition, the groove geometry makes magnetic crack detection difficult. For these reasons, it is understandable that, in several cases, cracked parts were installed in engines and failed following dynamic fatigue crack growth, causing extensive consequential damages.
Incidents such as these are probably one reason why some OEMs have prohibited the burnishing of toothed gears.