The variety of damage mechanisms in the turbine is matched by the number of different remedies (Chapters 184.108.40.206 and 220.127.116.11). The fundamental rule for these problems is: without a reliable damage analysis, including the identification and assessment of causes, it is not possible to develop a sufficiently safe and successful remedy. The following text only discusses remedies for turbine-specific damages that are caused by overly high blade temperatures. These are primarily creep damage, oxidation, and thermal fatigue. Other mechanisms such as mechanically induced dynamic fatigue (LCF, HCF), hot gas corrosion, or rubbing problems (Volume 2, Chapter 7.1.4) are covered in separate chapters on these topics (also see the index).
Overtemperatures in blades can have very different causes:
The following questions must be answered within the framework of the required damage analysis:
Tips for answering these questions are as follows:
Was there even a damage-causing overtemperature?
Typical externally observable characteristics ( ) indicate the occurrence of damage-causing thermal loads.
Despite this, it is not necessarily a matter of course that overtemperatures will be recognized as damage or a cause of damage. Overtemperatures that do not create any outwardly observable damage symptoms must also be identified. If the engine parts are made from materials that depend on the gamma` phase (hardening phase) for their creep resistance, such as most used Ni-based materials, then the amount, size, form, and composition of this phase can be used to determine thermal influences ( ). However, a complicating factor is the fact that long subsequent periods of operation at normal temperature alter the analyzable symptoms. This is due to further hardening. Another possibility for later analysis of thermal influences is a metallographic inspection of protective coatings (diffusion coatings, applied coatings) for distinctive changes (Fig. "Damaging temperatures and environmental Effects " -6).
Hot parts subject to high creep loads can exhibit so-called creep pores in their structure or on the fractured surface. These pores provide clues as to the acting damage mechanism, direction of loads, and degree of damage (Chapter 12.4). Unlike the forged turbine blades of older engine types, the cast alloys typically used today do not exhibit pronounced creep pore formation. The lack of creep pores does not allow one to conclude that creep was <U>not</U> the cause of damage.
How high was the overtemperature?
The minimum attained temperatures can be determined by structural characteristics (Chapter 12.4), fracture surface characteristics, and signs of melting. This assumes that these effects only occur at specific characteristic temperatures (Fig. "Typical duration and temperature level of overheating"). If there are pronounced signs of diffusion, then the temporal influence must be considered.
Which parts and part zones are affected?
The geometric distributon (e.g. on the rotor, Fig. "Overheating of turbine blades during stand stillS" or relative to the gas flow) of the overheated engine parts and their simultaneously altered operating behavior (e.g. combustion chamber, cooling air supply, or compressor) can provide important clues regarding the progress and cause of the damage (Fig. "Typical duration and temperature level of overheating"). Traces of hot streaks in the combustion chamber are a typical indicator (Fig. "Hot gas streaks as combustion chamber problem").
When did the overtemperature occur?
Answering this question is important not only for the damage mechanism, but also for more exact estimation of the part temperatures based on structural changes. Information for this can be found in testing protocols or data documented online, etc. Records that are not directly relevant to temperatures can also be useful (e.g. parameter changes that indicate a stall, or the vibration indicator for an imbalance).
What was the duration of the overtemperature?
If the time of the process, especially the creation of the overtemperature, is sufficiently limited, it can reveal the duration of the incident ( ). Tarnishing can be expected only
on fresh metal surfaces that were created during the damage process. The tarnishing probably occurred at the very end of the damage process, possibly during the cooling phase. Due to the additional atmospheric influence, extreme care must be taken when drawing conclusions. The thickness of oxide coatings only allows a very rough estimation of time, if any. One must consider that fresh fractured surfaces are very oxidation-friendly at first, and that cracks oxidize differently than surfaces that are more exposed to the atmosphere. The estimation of a damage-causing operating temperature becomes increasingly realistic with every known additional source of information regarding the temporal progress.
Which characteristics or events can plausibly explain the overtemperature?
Answering this question demands a comprehensive view of the damage process. For this reason, a prerequisite for sufficiently accurate conclusions is that all affected components, and preferably the entire engine, are available for analysis. No satisfactory conclusions can be expected if there was a “pre-selection” of parts or if only visibly damaged parts are available.
Damage prevention through coatings:
Often it is not possible to constructively change the blading (e.g. cooling configuration) or to prevent damage causes outside of the turbine. For example, if the temperature profile in the gas flow following the combustion chamber cannot be improved or the cooling air supply cannot be increased, then an alternative may be a thermal barrier coating. Ceramic thermal barrier coatings (TBC; ) can reduce the heat transfer into the blade and lower material temperature. These coatings can be applied as additional security. For example, they can protect the area near a hot spot (Fig. "High pressure turbine rotor blades overhaul intervals"). If, however, the coating must guarantee the life span of the engine part, i.e. it is included in the design configuration, then spalling is a high risk occurrence (see Chapters 13 and 14).
The correct selection of suitable coating systems to prevent oxidation and hot gas corrosion is very important for success in the specific application. Dense, firmly bonding, chemically and mechanically stable (against erosion and thermal fatigue) oxide coatings with slow growth rates are desirable. The protective effect against oxidation of the most frequently used coating system relies on the creation of Cr2O3 or Al2O3. These oxide protection coatings are produced with the aid of diffusion annealing in special media (powders, gases) or as applied coatings, usually by thermal spraying. The coating thicknesses are usually between 0.050 and 0.100 mm. The thicker the coating, the more it can affect the substrate strength. This necessitates compromises. Protection against hot gas corrosion (sulfidation, Volume 1, Chapter 5.4.5) is offered primarily by Cr2O3 coatings. However, there are experiences that do not confirm that Cr diffusion coatings (chromized) offer a sufficient protective effect from sulfidation. Aluminizing coatings are the most commonly used coating type. An intermediate Pt layer on the substrate creates so-called PtAl coatings. Pt acts as a diffusion inhibitor. In addition, it improves protection against type I sulfidation, since it also concentrates in the outer coating zone during operation.
Diffusion coatings are relatively brittle at low temperatures. Large heat strain is created during startup and shutdown. This, combined with relatively low part temperatures, promotes coating cracks (Fig. "Coatings sensitive to thermal fatigue"). Plasma spray coatings of the type MCrAlY (M stands for metal, e.g. Fe, Co, Ni, etc.) behave better than diffusion coatings. The transition from brittle to ductile behavior occurs at relatively low temperatures, which is an improvement over diffusion coatings.
If overheating is suspected, the combustion chamber must also be inspected for indications of a poor temperature profile. Signs of this include deformations, cracking, fractures, unusual coke buildup, damaged or blocked nozzles, etc.
Figure "Typical duration and temperature level of overheating": This diagram attempts to show probable temporal relationships between the
These estimates are subjective and based on experience. The estimation of overtemperatures and their duration should be made easier by giving clues as to their likely causes. Experience has shown that titanium fires or compressor stalls only last a few seconds. A creeping cooling air shortage can take a long time before causing externally detectable damage. A fouled pyrometer, caused by insufficient maintenance, can raise the temperature levels over several hundred hours and shorten engine part life spans. Increased leakages and deterioration of efficiency require increased gas temperatures to provide the same power levels. This may manifest itself as damage only after years of operation.
Figure "Influence of the tip shroud at the turbine blade 'live'": If damage is suspected to have been caused by overtemperatures, then the (most likely) cause must first be determined. For this, it is helpful to assign typical damage symptoms to the damage causes.
Various influences can compromise the effectiveness of turbine blade cooling systems. Worsened cooling increases the danger of overheating and reduces part life considerably ( ).
“A”: A foreign object strike constricts cooling edge bores near the edge (e.g. coke particles from the combustion chamber = carbon impact).
“B”: High operating temperatures create an insulating oxide layer in the cooling air bores. This causes the less efficiently cooled, and therefore hotter, blade wall area to oxidize more rapidly, making this a self-increasing process.
“C”: Dusts in the cooling air block the cooling air ducts. Typical blockages can be traced back to wear products from labyrinths and housing coatings.
“D”: Clogging of dust removal bores (Fig. "Blockages in turbine blade cooling system").
“E”: Thermal insulating effect of oxide coatings at fouled areas in the cooling air duct. Remnants from insufficient flushing of the blade after aggressive cleaning during an overhaul can result in this type of oxide coating. Compare with “B”.
“F”: Cooling air leaks at cracks in the blade.
Figure "Thermal barrier coatings at hot parts": Thermal barrier coatings serve to reduce hot part temperatures and/or minimize the cooling air consumption. These are ceramic coatings that consist primarily of zirconium oxide (ZrO2). They are produced with thermal spraying processes or physical vapor deposition (PVD; ). These coatings were initially used in combustion chambers (A). They were later used on the inner sides of shrouds (D) and then on the blades (B) of high-pressure turbine stator vanes. Today, pre-coated turbine running blades (C) are already in serial use. These primarily use PVD coatings, which have a columnal structure that makes them resistant to thermal fatigue (Fig. "Thermal barrier coatings of turbine rotor blades").
Figure "What splashed foreign material can can tell us": The combustion chamber melts passing particles. These then strike the following cooled, and therefore relatively cold, turbine blades. This creates deposits that have an appearance and structure similar to thermal spray coatings. These deposits can have damaging effects, but their properties can also allow conclusions as to the damage processes and mechanisms.
Number of coatings: If there are several coatings, it indicates that the processes that created the particles occurred at different times. In combination with events during operation and a bit of luck, conclusions can be made regarding the primary damage and consequential damages, or a repeating operating state that creates the particles.
Coating sequence and thickness: If the coatings have different compositions, structures, or thicknesses, then these differences are further clues regarding the damage process.
The coating thickness can indicate the intensity of the processes (e.g. rubbing, rotor deflection, imbalances, etc.) or the dust levels in the surrounding areas (i.e. the location of damage).
Composition of the coating: An analysis of the coating can yield important information concerning the origin of the dusts. The origin of metallic deposits should usually be searched for in the engine itself (e.g. abradable coatings and blade material). Non-metallic coatings with a composition typical of dusts in ingested air indicate that this is their source. For example, increased Al content in deposits can indicate that the source is ingested felspar dust, as opposed to SiO2 deposits without aluminum. This kind of knowledge allows one to reconstruct the process by which the coating was created, i.e. the operating area where it originated.
Differences in the composition of coatings can also be seen in the wear products from different abradable coatings in the compressor. NiC coatings of the type commonly used as abradable coatings in modern engine types are characteristically different from alternatively used coatings made from NiCrAl bentonite. In this way, the Al content can reveal whether a deposit was created during an earlier engine run in the development stage, in which a different coating type was used, and simply remained unnoticed.
Coating structure: The structure depends on the type of particles (melting point, composition), their size, impact speed, impact angle, reaction/diffusion properties, and the temperatures. This makes it possible to gain information regarding the time and temperature levels when the deposits were created. This, in turn, allows conclusions regarding the state of operation when the particles were created. Lower temperatures indicate that the engine was being started up or shut down (e.g. heavy rubbing due to rotor bow, Volume 2, Ill. 7.1.2-9). High temperatures indicate clearance gap bridging at high power levels.
Influence on the substrate: Reactions with diffusion in the substrate and/or coatings (e.g. diffusion coatings for oxidation protection) can irreparably damage blades. This can result in embrittlement and cracking. These effects can be expected especially from metallic deposits that are low-melting and alloy with metallic surfaces. Typical damage is caused by unusually heavy material removal from metallic Al abradable coatings. Surprisingly, ingested dusts can react with and damage ceramic thermal barrier coatings (Volume 1, Ill. 18.104.22.168-5). If dusts have a corrosive effect, then increased hot gas corrosion, e.g. sulfidation in the case of gypsum dusts and volcanic ash (Volume 1, Chapters 5.3.2 and 5.4.5), is to be expected.
Coating cracks: Brittle coatings can crack due to thermal strain inside them or in the substrate. These notched areas accelerate fatigue of the engine part.
Clogging of cooling air bores: This can occur if a large amount of dust is ingested. A typical example is the ingestion of volcanic ash during flight (Volume 1, Chapter 5.3.2). Deterioration of the cooling air system creates immediate danger of intolerably high part temperatures. Depending on the overtemperature, the part life may gradually be decreased, or the part may even spontaneously fail after a brief period.
22.214.171.124-1 P.König, A.Rossmann, “Ratgeber für Gasturbinenbetreiber”, ISBN 3-8027-2545-X, ASUE-Schriftreihe, Vulkan - Verlag Essen, 1999.
126.96.36.199-2 K. Steffens, “Technik der Luftfahrtantriebe”, lecture at the TU Aachen, 2002/2003.