Aeroengine Safety

Figure "Influences on parts by coatings": Coatings influence the operating behavior and therefore also the safety of a part in many different ways, including:

• Dynamic fatigue strength
• Wear
• Oxidation
• Corrosion
• Heat flow
• Non-destructive testability

For this reason, deviations and flaws originating in the finishing process are very important. The diagram compiles properties (Ref. 16.2.1.8-6), problems, and flaws that determine the operating behavior of coatings. Every coating type also has specific problems, influences, and flaws that act together to determine the coating characteristics.

Dynamic fatigue strength: This concerns the dynamic fatigue strength of the coating itself, as well as its influence on the dynamic fatigue strength of the base material (Fig. "Fatigue cracks in parts with thermal spray coating"). If the discussion is limited to the dynamic fatigue strength of the coating, in many cases, delamination is caused by LCF loads. These can occur in the form of cyclical centrifugal and/or gas forces, while thermal fatigue is the most common mechanism in hot parts (Volume 3, Chapter 12.6.2.1). In the case of thermal barrier coatings, there is also the special influence of oxide formation between the coating and base material (Volume 3, Ill. 11.2.3.1-5).
Examples of coatings under LCF stress are abradable coatings on spacers, ceramic wear-resistant coatings on labyrinth and blade tips, as well as thermal barrier coatings on hot parts. The separation of the coatings under cyclical loads is highly dependent on the residual stresses in the coating that were induced by the coating process. Minor deviations from the optimized and tested process parameters, especially temperature control, can decisively shorten the life span of a coating which is determined by separation. Residual stresses are also dependent on the coating structure, e.g. its porosity or desired cracks. In thermal barrier coatings, these cracks are segmentation cracks (Fig. "Bond strength of thermal spray coating structures"), while galvanically deposited Cr coatings have typical crack networks (Ills. 16.2.1.8-3 and 16.2.1.8.3-3). These characteristics are also determined by the process parameters.The thickness of the coating also has an effect on the residual stress levels (Fig. "Effects of electroplated coatings on fatigue strength"). Uneven coating thicknesses can cause high local residual stresses and separation. Another important influence is the adhesive strength of the coating, which is in turn closely related to the surface preparation, e.g. roughness and reactivity.
If the dynamic fatigue strength of the base material is affected, coatings can have very different effects.
Under LCF loads, i.e. significant plastic deformation, cracking will occur in brittle coatings. Crack growth from these cracks into the base material will result in a considerable decrease in the LCF strength (Fig. "Danger of a galvanic repair process"). This effect becomes more pronounced with increased coating hardness and bond strength to the base material.
Under HCF loads, i.e. only slightly above the endurance limit, coatings with high compressive residual stress and high strength (hardness), such as case-hardened or nitrided coatings, can increase the dynamic fatigue strength of the part considerably. In contrast, coatings with high tensile residual stresses, such as galvanic Cr coatings (Ref. 16.2.1.8-5) and Ni coatings (Fig. "Damage by Ni coating lowering fatigue strength"), can decisively lower the HCF strength (Fig. "Influence of Cr and Ni coatings at fatigue strength").
The dynamic fatigue strength is indirectly influenced by the FOD resistance of a coating. In this way, deviations in the finishing process will make it more likely for coating cracks or spalling to act as crack-starters into the base material.

Wear, tribologic behavior: The wear properties are of decisive importance for sliding surfaces such as abradable coatings, fretting-resistant coatings, and wear-resistant coatings. They are closely related to finishing-specific coating properties such as roughness, reservoir effects (e.g. cracks in Cr coatings), coating thickness, porosity (size; open; closed), composition (e.g. Ni/graphite abradable coatings). Therefore, deviations in these properties from the specification requirements can unallowably change the operating characteristics.

Oxidation in hot parts can be controlled with diffusion coatings or applied coatings. The concentration and distribution of the important elements (usually aluminum) required for the protective effect of a diffusion coating are important factors, as is the coating thickness (Fig. "Applications of diffusion coatings in engines i").
The properties of thermal sprayed coatings made from MCrAlY are also decisively dependent on the process parameters. If there is a bonding layer (undercoating), sufficiently high roughness becomes an important factor.
On the other hand, if a coating acts as the surface of a part exposed to the flow, the lowest possible roughness is desirable.

Corrosion: Process-specific coating characteristics such as porosity and cracking are important factors in determining corrosion behavior, especially as they interact with the base material (Fig. "Aftertreatment cracks on Cr-coating").

Heat flow: The insulating effect of thermal barrier coatings makes possible sufficiently long hot part life spans with minimal use of cooling air. Finishing-related coating flaws and deviations such as undesired porosity, spalling due to missing segmentation cracks (Volume 3, Ill. 11.2.3.1-4), and residual stresses, as well as structural anomalies, can worsen the insulating effect and decisively compromise part safety.
On a related subject, the effectiveness of the resistance of ceramic coatings to titanium fires requires optimal porosity.
Non-destructive testibility: The testibility of a coating and the base material below it are highly dependent on the coating properties. The use of magnetic crack detection depends on the magnetic properties of the coating. For example, non-magnetic Cr coatings (Fig. "Aftertreatment cracks on Cr-coating") require penetrant testing. If the Cr coatings are sufficiently thin, then it is possible to use magnetic testing to detect cracks in the magnetic steel substrate beneath the coating.
Many coatings have very limited non-destructive testibility with regard to verifying that they have sufficient operating characteristics. This means that quality assurance must be done through suitable and sufficiently documented process monitoring. Porous coatings such as abradable coatings do not permit ultrasonic testing. Penetrant testing is rarely useful. Thermography is used especially to detect coating separations (Fig. "Methods of qualitative bond strength testing").

Figure "Aftertreatment effects on coatings". (also see Fig. "Influence of finishing at coating properties"): In many cases, aftertreatment of coatings is necessary in order to obtain the desired properties. These aftertreatments are a coating-specific component of the specification drawings. The specification drawings are backed up by optimized and verified (tested) process parameters. It is assumed that these specifications will be strictly observed during the coating process.

Smoothing coatings: In order to achieve minimal roughness (e.g. on aerodynamic surfaces) or specific surface structures (e.g. oil film formation on sliding surfaces), reworking may be necessary. Reworking must not compromise any of the desired properties of the coating. For example, if a sufficient level of roughness is required on a wear-resistant coating on a labyrinth tip in order for rubbing to have a proper chipping effect (low friction heat), then the coating must not be reworked with a diamond file. The smoothing of soft, porous abradable coatings (e.g. metal felt) in housings with the aid of a hardening process (e.g. rolling) can result in dangerous overstressing of blade tips during rubbing (overheating, bending, vibrations). For this reason, coatings with low strength may only be (subsequently) machined if this is specifically allowed. There is also a risk of overstress and cracking (in brittle coatings) and/or local separation of the coating.
Experience has shown that subsequent grinding of galvanic Cr coatings can lead to cracking in the substrate below the coating (Fig. "Aftertreatment cracks on Cr-coating").
Smoothing may be done as afterwork on specified deviations such as powder particles sticking to diffusion-coated parts. If adhering particles must be removed, it must be ensured that no micro-cracking and/or spalling of the surface occurs, especially in the case of brittle coatings. These notches can unallowably lower the dynamic strength or compromise the function of the coating. One example of this is local damage to an anti-oxidation coating. During operation, this area can then experience premature attack of the base material and sub-surface corrosion of the undamaged surrounding coating.

Hardening processes for coatings: Surprisingly, in some cases, shot peening coatings results in greater dynamic strength increases than shot peening the base material before the coating is applied (Fig. "Influence of Cr and Ni coatings at fatigue strength"). Contrary to expectations, this behavior is exhibited by relatively brittle galvanic Cr coatings. The effect is primarily due to the breaking-down of high tensile residual stresses in the coating, and a simultaneous buildup of compressive residual stresses.
In ductile (plastically deformable) coatings (e.g. silver coatings), otherwise non-damaging shot peening (glass beads) can induce sufficiently high compressive stresses to cause coating separation and blistering in zones with minimal adhesive strength (Fig. "Testing bond strength of coatings"). This means that non-destructive testing of the adhesive strength is possible, at least in some cases.

Influencing coating-specific porosity: Chip-removing machining (turning, grinding), hardening (e.g. peening), and abrasive treatments can change the porosity of coatings, thereby affecting their characteristics. If open porosity is smeared or pressed shut, it may affect sliding behavior. This may be due to reduced acceptance of lubricants. Machining dust that has collected in the pores can be released later during operating, contaminating systems with throughflow.
If closed porosity is opened to the surface by material removal, encroaching media from the surrounding environment can affect the oxidation or corrosion behavior.
If porosity inside the coating is partially or completely pressed shut due to plastic deformation (machining, peening), it may have undesirable effects on properties such as elasticity, abradable behavior, or heat transfer.

Heat treatment can decisively alter the properties of a coating. For example, the hardness of certain Ni coatings can be increased considerably. In Al diffusion coatings, the optimal element distribution that results in the desired operating behavior (oxidation resistance with sufficient toughness) is only attained after annealing.
Of course, heating coatings can also have negative effects. Overheating during the grinding of Cr coatings can evidently even cause cracks in the base material (Fig. "Aftertreatment cracks on Cr-coating"). Organic coatings such as elastomers and synthetic resins can undergo unallowable changes even at increased temperatures that are apparently harmless. These effects include embrittlement, loss of strength, and shrinkage, possibly in combination with cracking.

Figure "Aftertreatment cracks on Cr-coating": Toothed gear shafts (middle diagram) made from case-hardened steel with galvanic Cr coatings on the bearing contact surfaces and bearing carriers showed pronounced cracking fields in the course of magnetic crack detection after being ground to size. The cracks were primarily axial, and perpendicular to the direction of grinding (left diagram). A metallographic examination revealed ingranular crack growth typical of grinding cracks.
At first, the magnetic crack indication was misinterpreted as acceptable, process-specific cracking in the chrome coating. However, magnetic crack testing does not react to cracks in the non-magnetic Cr coating. This required penetrant testing (right diagram), which revealed a network of cracks that differed from the patterns found by the magnetic examination.
If magnetic examination reveals cracks in the base material through a Cr coating that is several tenths of a millimeter thick, then the cracks are necessarily quite deep.
An additional confusing factor was that the cracking evidently increased with more intensive cooling. This behavior was explained by the shock-like cooling of the highly heated grinding surface. The grinding process apparently induced powerful tensile residual stresses in the base material, which caused spontaneous cracking. In this context, the burnishing process in an alkaline solution following grinding seems to be an important factor. Burnishing is known to cause SCC in unpeened base materials under sufficiently high tensile stress (caustic embrittlement, Fig. "Cracking in steels processed in burnishing baths"). This
also results in intergranular cracking. Therefore, an additional influence of the burnishing process on the cracking had to be considered in cases in which crack testing occurred after burnishing. The crack-inducing bath was able to reach the base material through the typical cracks in the Cr coating that were revealed by the penetrant testing.
The solution was shot peening of the base material (induction of protective compressive stresses) before coating and less intensive cooling (!) during grinding.

Figure "Testing bond strength of coatings": Serially implementable non-destructive testing of adhesive strength of coatings is extremely problematic. Specimens can be placed in a suitable position on the part and simultaneously sprayed, and then be tested later (Fig. "Measurement of coating bond strength"). This can be used to quantify the adhesive strength of the coating on the specimen. In addition, the crack pattern through the coating and/or along the bond layer is a qualitative criterion. Understandably, the significance of these specimens relative to the adhesive strength of the coating on the actual part is limited, at best. For this reason, these tests are generally just used to verify process stability.

In some cases, qualitative testing of adhesive strength is possible. On ductile metallic coatings, such as silver, peening with glass shot can be adjusted specifically to the coating properties. Naturally, it must be ensured that no damage occurs, such as unallowable material removal. The peening process creates compressive stresses in the coating, which will cause it to separate and blister in areas with insufficient adhesive strength (second diagram from top). The revealing of weak points depends on the peening intensity and coating properties.

If the base material has absorbed hydrogen during the coating process, embrittlement can be prevented through a subsequent heat treatment, which is usually prescribed as a method of disembrittlement. If the coating has suitable characteristics such as sufficient density, then the escaping hydrogen will result in blistering in zones with low adhesive strength (“pickle blisters”, third diagram from top). It must then be estimated whether the blistering is caused by unusually heavy gas absorption or insufficient adhesive strength of the coating.

If the thermal strain of the coating is considerably greater than that of the base material, compressive stresses will be created in the coating when it is heated. Brittle coatings such as ceramic thermal barriers and inorganic abradable coatings tend to separate if they have low bond strength (fourth diagram from top). In these cases, the gap created between the coating and the substrate will be minimal, and will not be visually recognizable as a blister. Detecting this type of separation requires a suitable non-destructive testing method. Thermography has proven effective in similar situations (Ref. 16.2.1.8-1). It is based on the poor heat transition in the separation gap, which becomes apparent when heat is applied from an outside source (thermal impulse).
The adhesive strength/adhesive behavior of sufficiently elastic elastomer coatings, such as abradable coatings made from filled silicon rubber, can be estimated with the aid of a technical test that can be classified as non-destructive (bottom frame). First, a hollow punch is used to cut a circle in the coating down to the base material. Next, a suitable load tester is affixed to the circular section. Then, flexural or tensile stress is placed on the circular section until it separates. If flexural stress is used, the tension-side edge of the coating will be placed under peel stress, which any adhesion flaws will react especially sensitively to (Ills. 16.2.1.4-18 and 16.2.1.5-6). The main assessment criterion is the fracture pattern within the coating, or the separation at the base material. After the test, the removed coating section can be taken off of the load tester and reaffixed to the part.

Figure "Hydrogen embrittlement by coatings": Hydrogen embrittlement can be caused by various processes (Volume 1, Chapter 5.4.4.2). The top diagram depicts the hydraulic piston of a variable thrust nozzle system, which experienced brittle cracking in several places following chromizing of the slide face. The cracks were primarily on edges and welds. It can be assumed that considerable tensile stresses at the edges promoted especially intensive hydrogen formation in these zones.
Baths used for cadmium coating have been found to be especially dangerous. They were used especially in older engine types, but their toxicity and environmental threat have led to their replacement by less hazardous anti-corrosion coatings. There are a large number of documented cases of damage through hydrogen embrittlement, especially on nuts and bolts.
During operation, there were several cases in which toothed gears that were case-hardened in the gaseous phase (bottom diagram) suffered tooth fractures after long run times. These were dynamic fatigue fractures that originated in hydrogen-induced cracks inside the teeth (bottom right detail).
In order to reduce the risk of hydrogen embrittlement to an acceptable level, it is necessary to conduct disembrittlement by heating the part to about 200 °C for several hours immediately after the coating process (see applicable regulations; Fig. "Hydrogen absorption during etching process").