High-Speed Rubbing Behavior Between Nickel-Based Superalloy Blades and Ceramic-Based Abradable Coatings

The tribological behavior of nickel-based superalloy blades with YSZ ceramic-based abradable coatings was investigated using a self-developed high-speed rubbing tester, focusing on the effect of working parameters (linear velocity and incursion rate) on blade and coating wear mechanisms. A non-contact infrared thermometer and a high-speed camera were used to determine real-time changes in coating surface temperature and blade length. The wear morphology and elemental composition of the blade and coating were characterized by a scanning electron microscope, an ultra-depth three-dimensional microscope, and an energy spectrometer. The test results showed that at the same incursion rate, the variation of blade wear length exhibited a tendency of first increasing and then decreasing with the increase of linear velocity. The maximum degree of blade wear was reached at 200 m/s. There was no significant correlation between the degree of blade wear and the incursion rate at the same linear velocity. The linear velocity was the primary influence on driving friction, while the incursion rate became a secondary factor, contrary to the well-known law in metal-based abradable coatings. The main forms of damage to blades included grooving, overheating, and micro-rupture. Plastic removal was the main wear mechanism for blades. The wear mechanism of ceramic-based abradable coatings was mainly characterized by subsurface cracking. The brittle removal of the coating resulted in “stepped” wear of the blade and a short actual action time of the friction pair. In this paper, the wear mechanisms of nickel-based superalloy blades and ceramic-based abradable coatings were mapped for the first time.


Introduction
The efficiency of an aero-engine is highly dependent on the clearance between the rotating blades and the stationary case.In general, the working clearance between the blade tip and the case often needs to be kept as small as possible to reduce the loss of efficiency due to compressed air leakage [1,2].However, due to the presence of centrifugal forces, thermal expansion, machining, and installation errors, small clearances inevitably lead to undesired scratching of the rotor blades against the inner wall of the case [2,3].For this reason, in the engineering field, abradable sealing coatings are applied to the inner wall of the case to form a particular friction pair with the rotor blades so that the wear is borne Extended author information available on the last page of the article 128 Page 2 of 18 by the coating as much as possible when the scratch occurs between the two [4,5].While taking into account the small clearance, the blade damage is minimized, and blade safety is ensured.
Abradable sealing coatings must exhibit an excellent comprehensive performance (e.g., abradability, thermal shock resistance, erosion resistance, and oxidation resistance) to survive in harsh operating environments.Of these properties, abradability is undoubtedly the most important, as it has a direct impact on blade wear and, accordingly, on the efficiency and safety of the aero-engine.Therefore, the study and evaluation of the abradability of abradable sealing coatings have been a hot topic in engineering and academic fields.
By examining worn abradable sealing coatings, Borel et al. [6] summarized the primary wear mechanisms of seal coatings in aero-engine seals including cutting, smearing, material transfer, compaction, melting, and friction oxidation.They found a good correlation between the wear track roughness of the coating, blade mass variation, and the wear mechanism.Accordingly, they developed a "wear mechanism map" for aluminum-based seal coatings.
Bill et al. [7] studied the wear behavior of three seal coatings under different test conditions.They found that the temperature of the coating was inversely proportional to the incursion rate.A low incursion rate led to high coating temperature and severe blade wear.They pointed out that a continuous layer of plastic deformation formed on the worn surface of the coating during high-speed rubbing was the main cause of blade wear.Bill et al. developed heat transfer and particle escape models to understand the smearing behavior.In addition, they suggested that coating compaction was usually accompanied by rotor blade wear and high friction energy and gave two models for calculating friction energy.The theoretical predictions were consistent with the experimental results.
Ghasripoor et al. [8] studied the wear mechanisms of three aluminum-based sealing coatings.They found that, in most cases, blade damage was reduced or replaced by coating adhesion.Micro-fracture, dominated by solid lubricant phase fragmentation, was the primary wear mechanism of the abradable sealing coatings.Deformation and partial melting of the metallic phases in the coating occurred.
Laverty et al. [9,10] found that the friction energy, directly related to the friction interface temperature, was mainly influenced by the incursion rate, followed by the linear velocity and blade thickness.They concluded that the high incursion rate increased the temperature of the sealing coating, which then damaged the blade.The results are contrary to those obtained by Bill [7].The exact reasons have not been explored.
Bounazef et al. [3] used a high-speed rubbing test machine to simulate the high-speed rubbing behavior of aluminum-based sealing coatings and titanium alloy blades under different operating conditions, such as take-off, cruise, and landing.They found that high linear velocities and incursion rates usually resulted in the slight material transfer and smooth cut tracks.Low line velocities and low incursion rates usually led to severe material transfer and rough wear tracks.
Stringer et al. [11] carefully investigated the material transfer behavior between aluminum-based abradable coatings and titanium alloy blades and found the same wear mechanism as Bounazef.They monitored the length change of the blade using a high-speed camera and confirmed that the adhesion behavior of the coating to the blade was characterized by periodic adhesion and detachment.They considered that the mass and length changes of the blade after the test were insufficient to characterize the wear process that occurred during the test.
Based on the Stringer study, Fois et al. [12] recreated the wear mechanisms observed in compressors between the blade tip and the sealing coating.The authors found that the adhesion process also had a period of stable adhesion and that the adhesion rate was negatively correlated with the linear velocity and incursion rate.They also observed a mixed wear mechanism, whereby adhesion and heat-driven wear of the blade occurred in some tests at low incursion rates.
Taylor et al. [2] studied the wear mechanisms of nickelbased superalloy blades and nickel-based sealing coatings.They obtained three-dimensional plots of the specific wear of the blades for different linear velocities and incursion rates.They found that tangential forces and blade wear were high at high single pass depths.The coating was cut at low single pass depths, and the blade showed almost no wear.However, this phenomenon also occurred at the maximum single pass depth.They concluded that the wear mechanism between the friction pairs was complex and dynamic.
Sporer et al. [4] briefly evaluated the effect of porosity on the thermal shock life, abradability, and erosion resistance of YSZ ceramic-based abradable coatings.They found that the thermal shock life and abradability of the coatings increased with the increase in porosity, but their erosion resistance became progressively worse.In the high-speed rubbing test to evaluate the abradability of the coating, they only provided the wear degree of the blade and the macroscopic wear morphology of the coating, and did not develop an in-depth analysis.There is a lack of further research and interpretation of microscopic wear information, such as the Page 3 of 18 128 wear surface morphology, cross section, and composition of the blade or coating.
Gao et al. [13] and Zheng et al. [14] investigated the highspeed rubbing behavior of aluminum-and nickel-based sealing coatings with titanium alloy blades.They found that the thermal properties of the sealing coatings had an essential influence on the damage to the blades.If the metal phase of the coating had a low melting point and the thermal diffusivity was high, the coating tended to soften before the blade so that the blade was slightly worn or even replaced by the coating adhesion.Conversely, the blade was prone to softening before the coating, and thus the blade was severely damaged.
Xue et al. [15] found different material transfer behavior in two metal-based abradable coatings.They suggested that temperature differences at the interface, changes in the melting point of the coating, and plastic flow of the blade at high temperatures were the factors responsible for the different material transfer behavior.
Table 1 summarizes the high-speed rubbing test conditions for the references cited in this paper.The above studies have greatly enriched our understanding of the specific tribological behavior of seal coatings and their counterparts.However, there are some limitations to the existing research.
Firstly, most research has focused on the high-speed rubbing of metal-based sealing coatings and titanium alloy blades for compressors.Little research has been done on ceramic-based sealing coatings versus paired blades for turbines.The rsearch by Spore et al. is the only visible study reported for turbine ceramic-based sealing coatings [4].However, the work focuses on the degree of damage to the blades under different conditions and does not address the wear mechanism.
Secondly, the extremely high sliding velocities and the extremely low incursion rates are two of the most distinctive features that distinguish a sealing friction pair from a conventional friction pair, and are among the most critical factors affecting blade and coating damage.However, there is no consensus in existing research on the effect of both on blade damage, and there are even contradictions between the different studies.For example, Laverty et al. [9] suggested that high incursion rates led to more severe blade damage, but Bill et al. [7] concluded that low incursion rates resulted in more severe damage.Taylor et al. [2] considered that line velocities and incursion rates (single pass depth) all had a significant effect on blade damage.
In response to the above background, the tribological behavior of nickel-based superalloy blades with YSZ ceramic-based abradable coatings was investigated using a self-developed high-speed rubbing tester, focusing on the effect of working parameters (linear velocity 100-300 m/s and incursion rate 5-200 μm/s) on blade and coating wear mechanisms.By collecting and analyzing data on length, mass, temperature, and wear morphology, the wear behavior and wear mechanism of the turbine seal pair were thoroughly investigated.The wear mechanism of this system was mapped for the first time.The study can guide the design, optimization, and selection of abradable or blade tip strengthening coatings in turbines and provide data support for kinetic or finite element analysis [16][17][18].  2 shows the macroscopic morphology (before/after heat treatment) and microscopic morphology of the abradable coating used in the high-speed rubbing test.YSZ ceramicbased abradable coating has been successfully applied to the turbine seal.In order to relieve the thermal expansion coefficient difference between the ceramic surface layer and the metal substrate and to improve the bonding strength, NiCrAlY was applied as a transition layer.The substrate was also shot-peened before spraying.The sprayed coating samples were placed in a muffle furnace for heat treatment (900 °C, 1 h, cooling with the furnace) to remove the polyphenylene ester (PHB) and create porosity.YSZ abradable coatings were ground flat with sandpaper, and the roughness of the coatings was controlled to 6 μm.

High-Speed Rubbing Test
The high-speed rubbing tests were carried out using the selfdeveloped high-speed rubbing tester (Fig. 3a).The highspeed rubbing test process can be outlined as follows: The electric spindle drives a metal disk equipped with a simulated blade and a balanced blade to a specified speed.The abradable coating mounted on the stepper stage approaches the rotating blade at a given incursion rate.When the abradable coating reaches the target incursion depth, the high-speed rubbing between the coating and the blade was finished, and the coating would be driven back to its original position by the stepper stage.The operating principle of the tester is summarized in Fig. 3b.The parameters of the tester and detailed operation procedures can be found in reference [19][20][21].
In addition, the high-speed rubbing tester is equipped with an infrared thermometer and high-speed camera to measure the coating surface temperature and blade tip length (variation) during high-speed rubbing, which helps to better understand the high-speed rubbing behavior between the blade and the coating.The high-speed camera system consists of an RT200 LED lighting controller, an OPB916 photologic slotted optical switch, and a PL-B741U digital camera with a macro-zoom lens.A pulse that triggers the camera is generated when the blade passes over the photologic slotted optical switch.With the LED flash, the digital camera will capture the image of the rotating blades.
In this study, we chose a test matrix to cover various conditions encountered in aero engines, such as thermal expansion, sudden loading, and engine vibration.Based on our preliminary study and the results of other researchers [19,21,22], it is known that line velocity and incursion rate are important factors that affect the high-speed rubbing test.Therefore, in this study, line velocity and incursion rate were varied.Compared with compressors, the incursion of rotor blades in turbines is much smaller, and the friction interaction usually occurs at relatively low linear velocities.Therefore, three incursion rates (5, 50, and 200 μm/s) and four linear velocities (100, 150, 200, and 300 m/s) were chosen.A total of 12 sets of  experiments were performed in this paper.The incursion depth for all experiments was kept constant at 400 μm.The incursion depth was determined based on observations of abradable coatings in service and a number of references consulted.The wide range of tests helped us study the effects of line velocity and incursion rate in more detail.These parameters were chosen based on the operating conditions of the turbine engine and have also been used by other researchers [4].They have been shown to reproduce the wear mechanisms in turbine engines.The single pass depth used in this study was calculated to vary in the range of 0.02 to 2.5 μm/pass, which was reasonable.The formula for calculating the single pass depth can be found in the literature [21].The specific experimental parameters are shown in Table 3.All experiments were done at room temperature, and each experiment was repeated more than twice to ensure the accuracy of the experiments.

Measurement and Characterization
The masses of the blades and coatings before and after the test were determined using a balance with an accuracy of 0. Republic).The element distribution on the blade surface and cross section was investigated using an Ultim MaxN silicon drift type energy spectrometer (Oxford, UK).The macroscopic morphology of the blade and coating was examined using a VHX-6000 ultra-depth three-dimensional microscope (Keyence, Japan), and the roughness of the blade or coating, the scratch depth, and the percentage of metal in the coating wear scar were analyzed.A D/Max-2500PC X-ray diffractometer (RIKEN, Japan) was used to determine the phase composition of the tribo-film on the blade surface (CuKα, 10°/min, 10° ~ 90°).

Variation in Blade Length and Mass
Figure 4 summarizes the variation in blade length and mass under different experimental conditions.From Fig. 4a and b, it can be seen that the variation in blade length and mass has a similar trend.At any incursion rate, the variation of both shows a trend of increasing and then decreasing with the increase of linear velocity.At 200 m/s, the variation of both blade length and mass reaches a maximum value, which indicates that the blade undergoes severe wear.From the values of both changes, it can be found that the degree of wear of the bare blade is unacceptable under any conditions.There is no significant correlation between the variation of blade length or mass and the incursion rate at the same line velocity.This is surprising.This is because of no accordance with the law that is observed and well known in compressors, where the incursion rate is the main influencing factor for driving friction [23,24].

Macroscopic Morphology of Blade and Coating
Figure 5 shows the macroscopic morphology of the blade tip and wear scars of the abradable coating at different test conditions.From Fig. 5, it can be found that the blade tip is worn over the whole width.The surface of the blade tip is covered with a large number of grooves of varying depths, which are parallel to the direction of blade motion.This is a typical feature of abrasive wear.The blade tip shows different degrees of heat-to-discoloration, which is particularly severe at 200 m/s.The degree of heat-to-discoloration reflects the temperature of the blade tip.Shear lip features are also found at the trailing edge of the blade.In addition, at the highest linear velocity (300 m/s), both sides of the blade tip are severely worn, and the blade profile is curved.
A comparison of the test samples at different incursion rates reveals that there does not appear to be a significant correlation between blade wear and incursion rate.
As seen in the wear tracks of the abradable coatings, the worn areas seem to be characterized by metal smearing in all tested conditions.However, the samples produced under different conditions have different metal-smearing behavior, which seems to be significantly correlated only with the linear velocity.At 100 m/s (Fig. 5a1, b1, and c1), "islandlike" metal smearing (dark areas) is uniformly distributed in the wear scar of the coating.Some pit-like features (lightcolored areas) consistent with the spalling mechanism are also found.At 150 m/s (Fig. 5a2, b2, and c2), the abradable coating surface shows large areas of metal smears and spalling pits, which are mainly located at the leading and trailing edges of the wear scar.At 200 m/s (Fig. 5a3, b3, and c3), the severe friction heat induces the formation of a tribofilm on the surface of the wear scar, accompanied by intense densification.Only a few spalling pits are distributed in it.At 300 m/s (Fig. 5a4, b4, and c4), the surface of the wear scar shows some fine metal adhesion to the blade tip, and the size of the spalling pits increases at this time.The severe metal smearing at the edge of the wear track corresponds to the wear of the blade tip.At 200 m/s (Fig. 5a3, b3, and c3), compaction of the abradable coating is found, resulting in much greater blade wear and temperature generated at this linear velocity than at other linear velocities.

Microstructure of Blades with Different Linear Velocities at the Same Incursion Rate
From the wear degree of the blade, the macroscopic wear morphology of the blade or coating, and the statistical results, it can be found that the linear velocity is the main influencing factor driving the friction, and the incursion rate becomes a secondary factor.Therefore, we keep the incursion rate constant (50 μm/s) and analyze the microscopic morphology and element composition of the blade tip at different linear velocities.The results are shown in Fig. 6.At 100 m/s (Fig. 6a), the blade tip is mainly dominated by grooves.The strip-like dark tribo-film is distributed on the blade tip surface.EDS analysis shows that the tribo-film is mainly composed of Al, Zr, and O elements.The friction heat leads to the oxidation of active elements (such as Al).Zirconium element is derived from the abradable coating.This striped tribo-film corresponds to "island-like" metal smearing in the wear scar of the abradable coating.A number of circular spots covered with cracks are found on the top of the tribo-film.These spots are enriched with Zr, Al, Mo, Ti, and S. The presence of these elements contributes to the formation of thermal cracks.In addition, penetrating cracks between the blade and the shear lip are also found at the edge of the blade.
At 150 m/s (Fig. 6b), the area of the tribo-film increases significantly.From the local magnification, it is found that the tribo-film surface is covered with transverse cracks perpendicular to the direction of blade motion caused by thermal stress.Some "tadpole-like" features caused by the melting of the metal and large cracks are also found on the trailing edge of these blades.It is clear that the high linear velocity generates high friction heat.The temperature generated in the high-speed rubbing is sufficient to melt the blade metal.
At 200 m/s (Fig. 6c), the tribo-film covers almost the entire blade tip, and a large number of transverse cracks induced by thermal stresses are also found on its surface.The grooves on the surface of the tribo-film are significantly reduced.The thickness of the tribo-film formed at this linear velocity is larger, and the distribution of elements is more uniform compared with 150 m/s.At 200 m/s, the intense friction heat triggers severe oxidation of the metal elements.
At 300 m/s (Fig. 6d), the fine wear debris appears to partially melt and adhere to the blade tip surface as they are removed.This makes the blade tip surface rough.At this point, the friction heat effect does not seem to be significant, but a thin tribo-film is still formed.Similar to Fig. 6a, cracks are produced in the Zr-, Mo-, Ti-, and S-enriched areas.Large cracks are also found at the trailing edge of the blade.
Figure 7 shows the cross-sectional morphology of the blade tip with different linear velocities at a 50 μm/s incursion rate and the corresponding EDS results.The black area at the top of the picture is epoxy resin, which is used to prevent the destructive behavior generated during the grinding process.From Fig. 7a, it can be found that the tribo-film formed on the blade tip surface at 100 m/s is characterized by a discontinuous distribution with a thickness of about 250 nm.The EDS results show that the tribo-film is mainly composed of Al, Zr, Y, and O.This is consistent with the tribofilm shown in Fig. 6a.At 150 m/s (Fig. 7b), the tribo-film becomes continuous, and the thickness increases to 370 nm.A mixture of metal debris and ceramic material, which is the primary material forming the tribo-film, is also found above the tribo-film.The thickness of the tribo-film increases to 540 nm at 200 m/s (Fig. 7c).The tribo-film is fractured by thermal stress and mechanical shear.In addition, thermal cracking through carbide extension to the substrate is also found.Under high friction heat, the metal elements of the blade substrate are oxidized severely, and aluminum-and titanium-depleted zones appear below the tribo-film.At 300 m/s (Fig. 7d), the tribo-film almost disappears (about 45 nm).The grooves on the blade tip surface at this linear velocity have a significant fluctuation compared with the low linear velocity.This is consistent with the statistical results of blade roughness.
The tribo-film observed in Fig. 7 was examined by X-ray diffraction technique, and the results are shown in Fig. 8. From Fig. 8, it can be found that, in addition to the diffraction peak of the substrate, a typical diffuse scattering peak appears at a low angle, which indicates that the tribo-film consists mainly of amorphous material.The amorphous material is formed during the rapid cooling of the molten metal, and its internal atoms do not have time to make an orderly arrangement, which in turn forms the amorphous state.This again proves that the friction heat generated during the high-speed rubbing process is sufficient to melt the blade metal.The generated friction heat can be rapidly consumed or absorbed through conduction, convection, and radiation.In addition, the high friction heat leads to the formation of Al 2 O 3 at 200 m/s.

Microstructure of Coatings with Different Linear Velocities at the Same Incursion Rate
Figure 9 shows the microscopic morphology of the abradable coating wear scar for different linear velocities.From Fig. 9, it can be observed that with the increase of linear velocity, the coating is gradually compacted (Pores are significantly reduced), and a dense and smooth surface appears.The tear-like spalling feature on the surface of the coating wear scar also becomes progressively more pronounced.At 100 m/s (Fig. 9a), scattered "island-like" metal smears are distributed on the surface of the wear scar.Large areas of metal smears are found at 150 m/s (Fig. 9b).At 200 m/s and 300 m/s (Fig. 9a, c), the original porosity of the coating is significantly reduced, which indicates that the coating is compacted.At the edge of compaction, tear-like spalling pits are formed.
Figure 10 shows the cross-sectional morphology of the abradable coating wear scar for different linear velocities.As can be seen from Fig. 10, cracks are produced on the subsurface of the abradable coating at all linear velocities.At low linear velocities (Fig. 10a), cracks are characterized by high number and big size.Most of the cracks are parallel to the abradable coating surface.The cracks cause large pieces of material to peel off from the substrate, but the abradable material is still attached and not completely separated.At moderate linear velocities (Fig. 10b, c), the cracks are characterized by low numbers and small size.In contrast to the low linear velocity, these cracks do not show connectivity.In addition, the surface of the abradable coating becomes relatively flat.At high linear velocities (Fig. 10d), the abradable coating appears to be torn, and the coating surface becomes rough.

Diagram of Wear Mechanism
The above analysis shows a systematic correlation between the wear mechanism and the test parameters, which allows us to map the wear mechanism of the system.We record the ratio of blade wear to total wear (blade wear and rubbing depth of the abradable coating wear scar) as the degree of blade wear.The results of the temperature change of the abradable coating, the roughness of the wear track, and the macroscopic or microscopic morphology of the blade or coating are used to determine the wear mechanism.Figure 11 shows a diagram of the wear mechanism for this system under all tested conditions.In this case, the wear mechanisms occurring under a set of conditions are labeled with the initials of the relevant mechanism.From Fig. 11, it can be observed that severe wear occurs on the bare blade under all tested conditions.The results obtained in this test are more severe than those obtained by Sporer et al. [4].This may be related to the differences in the properties of the blade materials or the abradable coatings.The wear of the blades is caused by shear removal and melt wear.Blade tip surfaces are mainly dominated by grooves.At 200 m/s, the blade tips are overheated at any incursion rate.The blade wear is heat driven.At this point, the blade wear reaches a maximum (80%).In addition, some micro-rupture phenomena occur in the test matrix.This wear mechanism map is allowed to predict the degree of blade wear and the wear mechanism.

Wear Mechanisms of Blades and Abradable Coatings
The above experimental results indicate that the blades appear to have a similar wear mechanism at any linear velocity and incursion rate.The wear of the blade is all dominated by shear removal (plastic removal).In this case, when the force required to shear the blade tip is less than the force at the contact, the shear point will arise on the blade tip side.Blade wear will then occur.The shear strength of the blade tip is a function of the blade tip temperature.Under the action of the tangential force and friction heat, the sheared blade material will be coated on the abradable coating surface.As shown in Figs. 5  and 6, the abradable coating is characterized by metal smearing within the wear area and the blade tip is covered with grooves at all test conditions.This indicates that the main wear mechanisms of the blade are adhesive wear and abrasive wear.This is confirmed by the presence of a shear lip on the trailing edge of the blade.The shear lip is the result of the combined effect of the shear mixed layer and the softening of the blade substrate, both of which are the result of high friction heat [22].The effect of friction heat is also not negligible.
The "tadpole-like" metal solidification particles found on the blade tip surface indicate that the friction heat is sufficient to melt the blade.This indicates that at certain linear velocities, both shearing and melting of the blade have occurred.The metal debris and ceramic materials are gradually crushed, mixed, and sintered under the action of forces and friction heat, resulting in the formation of an amorphous tribo-film.
There is no doubt that the "tadpole-like" metal particles are the main material for the formation of the tribo-film.In addition, the high friction heat induces tribo-oxidation.γʹ-Ni 3 (Al, Ti) phase is the main strengthening phase of the blade.The loss of the element makes the blade tip lose some of its mechanical properties, thus making the blade more susceptible to shear.This is also one of the reasons for the severe blade wear under high friction heat.
For the abradable coatings, varying degrees of metal smearing and spalling characteristics are observed in the wear tracks.However, the wear mechanism of the abradable coatings is worn by subsurface rupture at all linear velocities and incursion rates.This wear mechanism is similar to that of NiCrAlbentonite [24].At low linear velocities, repeated impacts of the blade will mechanically weaken the structural integrity of the coating and cause damage to the abradable coating.However, each strike will not completely remove the damaged abradable material (Fig. 10a).At high linear velocities, subsurface layers damaged by the repeated accumulation of friction heat and mechanical stress are removed by successive strikes (Fig. 10d).As a result, fewer cracks are found at high linear velocities.Similar results were found by Fois in metal-based abradable coatings [25].In addition, the reduction of porosity on the coating surface indicates that compaction of the abradable coating occurs.This compaction phenomenon is particularly evident at high linear velocities.The compaction phenomenon may be related to factors such as the smearing of the blade metal, incursion stress, and friction heat.

Effect of Linear Velocity and Incursion Rate on Blade Wear
As can be seen from the degree of blade wear (Fig. 4), the damage patterns of nickel-based superalloy blades rubbed against ceramic-based abradable coatings are clearly different from those previously obtained in friction pairs with metal-based abradable coatings.When the linear velocity is below 200 m/s, the degree of blade wear increases significantly with increasing linear velocity, regardless of the variation in incursion rate; however, when the linear velocity exceeds 200 m/s (i.e., 300 m/s), a reduction in the degree of blade wear occurs.At a high incursion rate (i.e., 200 μm/s), the degree of blade wear is even lower than that at 100 m/s.The reasons for these unusual damage patterns in the blades may be related to the following factors: (1) The temperature at the blade tip, (2) The wear mechanisms of the blade, and (3) The contact state of the blade with the ceramic-based abradable coating.These three possible factors are described below: 1.The temperature at the blade tip High-speed rubbing tests differ from conventional friction and wear tests in that the intense friction heat and its effects influence the mechanical properties of both sides of the friction pair, thus the degree of wear of both.Compared to ceramic-based abradable coatings, the effect of temperature on the mechanical properties of the blade is even more significant.Obviously, the higher the blade temperature, the greater the loss of mechanical properties of the blade, and the more likely it is to cause wear.
The importance of friction heat and its effects in highspeed rubbing has been recognized, and some attempts have been made to characterize or predict the temperature on friction surfaces.The surface temperature of a stationary abradable coating is more easily measured than that of a rotating blade.Figure 12 illustrates the temperature variation of the abradable coating surface at different experimental parameters.It can be seen from Fig. 12 that the temperature of the abradable coating surface tends to increase and then decrease with increasing linear velocity at the same incursion rate.This is similar to the trend in blade length and blade mass (Fig. 4).It seems to confirm that the wear of the blade is driven by friction heat.The temperature of the abradable coating surface reaches a maximum value (371.26℃) at a line velocity of 200 m/s and an incursion rate of 200 μm/s.
Although the measured temperature of the abradable coating is relatively low, the temperature of the blade may be much higher, supported by the heat-to-discoloration characteristics of the blade tip surface and the metal solidification particles (Figs. 5 and 6).Blade tip strength is a function of blade tip temperature.Accurate calculation of the blade tip temperature helps understand high-speed rubbing behavior and is also extremely important.However, the temperature of the blade tip surface is usually difficult to measure directly.Researchers can obtain relatively coarse temperature values through finite element simulations and theoretical derivations.The temperature of the blade tip is estimated based on our previously proposed one-dimensional heat transfer equations [13,20]: where T b and T c are the temperatures of the blade and the abradable coating, respectively; θ is the ratio of the temperature rise rate between the blade and the abradable coating; L is the rubbing length; b is the thickness of the blade, 2 mm; a b is the thermal diffusion coefficient of the blade, ~ 2.60 × 10 -6 m 2 /s; a c is the thermal diffusion coefficient of the abradable coating, ~ 8.70 × 10 -7 m 2 /s; N is the total number of rubbing; i is the number of rubbing; R is the rubbing radius, 0.205 m; d is the depth of one rubbing; n is the rotation speed (rpm); D is the rubbing depth; V i is the incursion rate.Obviously, the value of L is proportional to the depth of rubbing.Considering that the damage of ( 1) the abradable coating usually exceeds the incursion rate, we choose the L value at a rubbing depth of 0.2 mm to represent the actual average rubbing length of the blade.Table 4 shows the parameters related to the estimation of the blade tip temperature based on the one-dimensional heat transfer equation.
From Table 4, it can be seen that the temperature of the blade is approximately four times the temperature of the abradable coating surface at all test conditions.At 100-200 m/s, the temperature at the blade tip is positively correlated with the linear velocity.The temperatures generated under certain tests are already close to or above the melting point of the blade (~ 1340 °C).At this point, the blade loses almost any mechanical strength and is extremely susceptible to shear.In addition, the high friction heat induces tribo-oxidation.The loss of strengthening elements also reduces the mechanical properties of the blade tip (Fig. 7).As a result, at 100-200 m/s, the degree of blade wear increases with increasing linear velocity.At 300 m/s, the temperature at the blade tip drops significantly.The low temperature allows the blade to exhibit good mechanical properties and thus causes minor blade wear.Figure 13 summarizes the degree of blade wear as a function of blade tip temperature at different experimental parameters.It can be seen from Fig. 13 that the degree of blade wear is almost proportional to the blade tip temperature, which proves that blade tip wear is driven by friction heat (temperature).Therefore, the unusual damage patterns that occur in the blade are significantly correlated with the temperature variation at the blade tip.In addition, it can also be found from Fig. 13 that the blade tip temperature increases with the increase of the incursion rate at 100-200 m/s.This is consistent with the findings obtained by Watson in metal-based abradable coatings [24].However, the blade tip temperature is inversely proportional to the incursion rate as the linear velocity increases up to 300 m/s.This may be related to the fact that high linear velocities change the contact state between the blade and the abradable coating.
However, it should be noted that the temperature of the abradable coating surface is below 300 ℃ in some tests.Since the reading range of the infrared thermometer is limited to 300 to 1100 ℃, the values outside these ranges are recorded as boundary values.This leads to the possibility that the recorded and estimated temperatures are large.

The wear mechanisms of the blade
As can be seen from the analysis in Sect.3.6, the main wear mechanisms for blades are abrasive wear and adhesive wear (blade material smeared on the coating surface).As two forms of material wear, adhesive wear is generally considered to cause more severe material damage than abrasive wear in high-speed rubbing experiments [2,22,24].Therefore the greater the degree of adhesive wear, the more severe the blade damage.
In this system, the area percentage of the metal phase in the coating wear scars is used to characterize the degree of adhesive wear.The percentage of the metal phase in the coating wear scars was calculated using the "automatic Fig. 13 The degree of blade wear as a function of blade tip temperature at different experimental parameters area measurement" function of the VHX-6000 ultra-deep three-dimensional microscope and compared to the degree of blade wear, and the results are shown in Fig. 14.From Fig. 14, it can be observed that the degree of adhesive wear of the blade increases with the linear velocity at 100-200 m/s.This results in the degree of blade wear being proportional to the linear velocity at 100-200 m/s.At 300 m/s, a decrease in the degree of adhesive wear of the blade occurs, when the degree of blade wear is less.Figure 14 shows that there is a good correspondence between the degree of adhesive wear of the blade and the degree of blade wear.Therefore, the degree of adhesive wear of the blade under different operating parameters is considered to be one of the reasons for the unusual damage patterns.In particular, the degree of adhesive wear of the blade is related to the temperature variation at the blade tip.

The contact state of the blade with the ceramic-based abradable coating
Despite the presence of a certain amount of porosity in the ceramic-based abradable coating, the high hardness of the ceramic-based abradable coating is likely to cause damage or wear to the mating material (blades).And the longer the wear distance, the greater the damage to the blade.When the incursion depth (D) is fixed, the blade and the coating interaction when the theoretical wear distance can be obtained by the formula (3)(4)(5).As can be seen from the formula, the theoretical wear distance is proportional to the linear velocity (n) and inversely proportional to the incursion rate (V i ).
As a brittle material, the high-speed impact of the blade on the ceramic-based abradable coating during the rubbing process often leads to large pieces of the coating material peeling off (subsurface cracking), which results in a single cutting depth that is often less than the peeling depth.In this case, the blade and the abradable coating may be in a contact-free state, that is, the actual interaction time between the blade and the abradable coating (actual wear distance) is less than the theoretical interaction time (theoretical wear distance).
The appearance of the contact-free state can be reflected in a variation of blade length that may not be uniform.To test this idea, we recorded the real-time variation of blade length during high-speed rubbing at different linear velocities at an incursion rate of 50 μm/s using a high-speed camera, and the results are shown in Fig. 15.The results at other incursion rates are similar to those at 50 μm/s and are therefore not presented.As can be seen from Fig. 15, the variation in blade length is characterized by a slight "stepped" variation at 100-200 m/s, but decreases almost uniformly.This indicates that there is a slight absence of contact between the blade and the abradable coating at 100-200 m/s.At 300 m/s the variation in blade length is clearly characterized by "stepped" wear, that is, the blade does not interact with the abradable coating for a considerable period of time (corresponding to the platform in Fig. 15).The appearance of this phenomenon directly supports our inference.The intermittent rubbing sound produced during the experiment confirms the no-contact state of the blade and the abradable coating.Therefore, even at high linear velocities or high incursion rates (e.g., 300 m/s, 200 μm/s), the state of intermittent contact between the blade and the ceramic-based abradable coating may result in low blade tip temperatures or an inverse ratio of blade tip temperatures to incursion rates as found in Fig. 13.
As a result, the damage to the abradable coating caused by blade impact is compatible with the blade incursion at 100-200 m/s.The high linear velocity increases the friction heat power and the actual friction length, which then leads to high temperatures and considerable blade wear.The blade tip wear is driven by friction heat.At 300 m/s, the damage to the abradable coating caused by blade impact goes beyond the blade incursion.This leaves the friction pair in a contact-free state for a long time, which results in a significant reduction in the interaction time between the blade and the abradable coating (i.e., a reduction in the wear distance).Therefore, this linear velocity results in low temperatures and slight blade wear.

Conclusion
The high-speed rubbing behavior between nickel-based superalloy blades and ceramic-based abradable coatings was investigated using a high-speed rubbing tester.Two main influencing factors (linear velocity and incursion rate) were varied to examine their effect on the wear mechanism.

Fig. 1 Fig. 2
Fig. 1 Physical drawing of the simulated blade, balanced blade, and their dimensions

Fig. 3
Fig. 3 High-speed rubbing tester (a) and high-speed rubbing experimental principle (b) 1 mg.The length change of the blade was recorded with a micrometer with an accuracy of 1 μm.The changes in mass or height of the blade and coating before and after the test are obtained by subtracting the original values of the samples from the values of the tested samples.Positive values represent wear, and higher values indicate greater wear.Each value recorded is the average of the results of at least two tests.A non-contact infrared thermometer (RAYM-M2MLSF2L, USA) was used to determine the temperature change of the abradable coating surface.The temperature range of the infrared thermometer was 300 to 1100 ℃.If the temperature was below or above this temperature range, the temperature was recorded as a boundary value.A highspeed camera was used to monitor real-time changes in blade length during the test.The wear morphology of the blades and coatings was observed using a TESCAN MIRA3 field emission scanning electron microscope (TESCAN, Czech

Fig. 4 Fig. 5
Fig. 4 Variation in blade length (a) and mass (b) under different experimental conditions

Fig. 6
Fig. 6 Microscopic morphology of blade tips with different linear velocities at 50 μm/s incursion rate and the corresponding EDS results.a 100 m/s; b 150 m/s; c 200 m/s; d 300 m/s

Fig. 7
Fig. 7 Cross-sectional morphology of blade tips with different linear velocities at a 50 μm/s incursion rate and the corresponding EDS results.a 100 m/s; b 150 m/s; c 200 m/s; d 300 m/s

Fig. 8 Fig. 9
Fig. 8 Phase composition of the tribo-film on the blade tip at different linear velocities at a 50 μm/s incursion rate

Fig. 10 Fig. 11
Fig. 10 Cross-sectional morphology of abradable coating wear scars for different linear velocities at a 50 μm incursion rate.a 100 m/s; b 150 m/s; c 200 m/s; d 300 m/s

Fig. 12
Fig.12 The temperature variation of the abradable coating surface at different experimental parameters

Fig. 14
Fig. 14 Area percentage of metal phase in wear scars of abradable coatings

Fig. 15
Fig. 15 Real-time variation of the blade length at different linear velocities (50 μm/s)

128 Page 4 of 18 2 Experiment 2.1 Materials K417G
nickel-based superalloy was selected as the blade material, and its chemical composition is shown in Table2.K417G superalloy was cut into simulated blades and balanced blades according to the dimensions shown in Fig.1.The friction surface area is 2 × 4 mm 2 .The blade tips were ground with sandpaper to 0.5 ± 0.1 μm to simulate the roughness of the real blade tips.

Table 3
Test parameters used in this work