Lopez et al.⁴⁸⁾ studied the sliding wear behavior of LPS-SiC ceramics with different microstructures like elongated SiC grains (in situ toughened LPS SiC), coarse equiaxed SiC grains, and fine equiaxed SiC grains. Sliding wear resistance of elongated grain SiC ceramics was better than that of equiaxed SiC ceramics owing to a hard interlocking network of in situ toughened LPS SiC grains (see Fig. 1). The isolated nature of the yttrium aluminum garnet (YAG) second phase also played a role in enhancing wear resistance of in situ toughened LPS-SiC. This was in contrast to the equiaxed SiC ceramics, where the grains were embedded within a continuous YAG phase matrix.

In sliding of SiC ceramics prepared with YAG (Y₂O₃ : Al₂O₃ = 3 : 5), Lopez et al ⁵⁰⁾ suggested reduced content of the intergranular phase or grain size for improvement in wear resistance of LPS α-SiC ceramics. Reduced content of intergranular phase and the grain refinement resulted in reduction of rate of transition from deformation-controlled wear to fracture-controlled wear and hence in wear loss.

A simple model by Lawn et al.⁶⁹⁾ provided a framework for estimating the sliding wear properties in polycrystalline ceramics. The proposed model states that the plastic-deformation damage accumulates within the grains during deformation-controlled wear. This damage is in the form of dislocation pile-ups, confirmed using transmission electron microscopy.^69-71)

It was earlier concluded that SiC ceramics sintered using 3 wt% AlN-Y₂O₃ (SCY3) additives revealed mostly large equiaxed grains with amorphous grain boundary phase, while SiC ceramics prepared using 3 wt% AlN-Sc₂O₃ (SCSc3) additives exhibited duplex microstructure of elongated and fine equiaxed grains with clean grain boundary phase (see Figs. 2(a) and (b)). SiC ceramics prepared using 3 wt% AlN-Sc₂O₃ additives resulted ~ 18% less wear at 13 N load. Worn surfaces of SCY3 and SCSc3 are shown in Figs. 2(c), and (d), respectively.²⁾

Murthy et al.⁷²⁾ observed that SiC ceramics doped with different p-type (Al, Mg) or n-type (P) doping elements had influence on the wear behavior in two ways: (i) by directly affecting thermodynamics of tribochemical reactions or (ii) by modified frequency of grain boundaries having same orientation angle and change in the kinetics of tribochemical reactions. It was found that doping elements partly made a sloid solution with SiC and rest were segregated along grain boundaries. Segregated Al reduced solubility of silica largely in water, whereas P and Mg reduced solubility to a lesser extent.⁷³⁾ Kinetics of formation of hydrated silica varied with solubility and hence led to change in frictional characteristics of SiC ceramics.

For improvements in the sliding-wear resistance of pressureless LPS-SiC under diesel lubrication, Ciudad et al.⁷⁴⁾ suggested (i) reduction in intergranular phase (second phase) content, and (ii) decrease in grain size or increase in grain aspect ratio. They reported that SiC ceramics with similar grain size (of 1.1 μm) but having large amount of second phase exhibited high wear. Second phase led to decrease in hardness and increase in wear.

Ortiz et al.⁵⁴⁾ proposed a colloidal processing route to produce dense SiC ceramics via pressureless liquid-phase sintering (PLPS). In addition to uniform mixing, the colloidal processing route resulted in a nano film of ~ 5 nm thickness on the surface of SiC particles. Compared to conventional mixing and sintering, complete densification achieved in shorter sintering time and minimum defects found in ceramics sintered using colloidal processed powders. The time required for the transition from mild to severe wear was more than double for SiC ceramics prepared by colloidal processing as compared to that for the SiC ceramics prepared by conventional processing. It was advocated that, though the microstructures were similar in terms of the size and morphology of grains for ceramics prepared through both processing routes, the clear interphase with fully dense structure led to a superior wear resistance.

Zhou et al.⁷⁵⁾ reported that silicon carbide-graphite (SiC-C) composites exhibited lower coefficients of friction during sliding than the monolithic SiC. The flake like graphite particles present in SiC-C composites (see Fig. 3) were expected to provide solid lubrication with easy shear during sliding. However, in the high contact stress condition (pin-on-disk tests), wear rates enhanced sharply from 0.38 to 3.90 × 10⁻⁶ mm³/N.m for SiC-C composites which was attributed to the fracture of graphite flakes despite lowered coefficients of friction when compared to low contact stress condition (block-on-ring test) (See Table 1).

In other research,⁷⁶⁾ SiC ceramics with 0, 10, 30 or 50 wt% WC were slid against SiC ball. WC reinforcement led to a decrease in size of equiaxed grains from 835 μm to 578 μm of SiC ceramics and resulted improvement in wear resistance from 1.3 × 10⁻⁵ mm³/N.m to 4 × 10⁻⁶ mm³/N.m (see Fig. 4).

Mechanical properties such as fracture toughness, hardness, and elastic modulus have influence on sliding wear characteristics of SiC ceramics. Hardness and fracture toughness of SiC ceramics are reported to generally range from 16 GPa to 27 GPa, and 2.3 MPa.m^½ to 8.3 MPa.m^½, respectively based on types of sintering/additives, microstructure etc.^49,51,77-80) Particularly, the influence of hardness and fracture toughness on friction and wear behavior was debated in most of the reports.

A sharp indenter model and a blunt indenter model were proposed to estimate the sliding wear broadly in the tribocontact of brittle materials in absence of any layer formation. According to Marshall et al.⁸¹⁾ for the tribocontact of sharp indenters, wear in brittle materials occurs due to the formation and propagation of lateral cracks. For a given applied load (P), total sliding distance (S), hardness (H), fracture toughness (K_Ic) and elastic modulus (E), wear volume of the brittle solid (V_S) can be assessed using the following equation:⁸¹⁾

where α is a material constant. In case of blunt indenter pressed against a brittle solid, surface ring cracks are generated, which propagate towards downwards with repeated sliding contact and develop conical cracks to cause pull-out of material.⁸²⁾ Following Hertzian analysis, when there was no tribochemical layer formation, Tewari et al. estimated wear volume (V_B) for brittle materials in ball-on-disk fretting wear contacts as per the following relation.⁸³⁾

where R is radius of the ball, χ and β are constants, and s is the stroke length in fretting contact.

SiC ceramics sintered with small amount (0.2 wt%) of Y₂O₃ additives attributed to high strength (561 MPa) and high hardness (27.7 GPa) than the SiC ceramics sintered with large amount (3 wt%) of Y₂O₃ additives (542 MPa and 26.1 GPa), and exhibited superior wear resistance.⁸⁴⁾ Li et al.⁸⁵⁾ found that toughened SiC prepared by hot pressing showed high hardness and fracture toughness compared to the SiC ceramics prepared through pressureless sintering or reaction sintering. The same trend reflected for resistance against friction and wear in water-lubrication conditions against SiC counterbody. Lopez et al.⁵¹⁾ reported that elongated-grain LPS SiC, despite having lower hardness, showed almost a half of the wear rate than that obtained for equiaxed-grain LPS SiC, and reflected less extent of dislocation plasticity. They also differentiated the extent of wear as mild wear dominated by hardness and severe wear dominated by fracture toughness of LPS SiC ceramics. Kovalcikova et al.⁸⁶⁾ used brittleness index (H/K_1c) of material for qualitative estimation of specific wear rate of SiC against Si₃N₄. They observed that brittleness index had dominant role in the wear response. Their results showed that resistance to wear increases with increased index number. In other study on SiC-WC composites,⁷⁶⁾ the influence of hardness and fracture toughness on the sliding wear of the SiC-WC composites was realized. The maximum hardness of 26.3 GPa for SiC reinforced with 30 wt% WC particles resulted in homogeneous dispersion of WC particles in the SiC matrix, whereas, a higher fracture toughness obtained due to crack deflection and bridging by a large amount (30 or 50 wt%) of WC particles.⁶¹⁾ The minimum material removal for SiC ceramics sintered with 50 wt% WC attributed to the maximum fracture toughness. Thus, fracture toughness appeared to carry important role than hardness towards estimating the extent of sliding wear for the investigated SiC-WC composites.

Llorente et al.⁷⁹⁾ studied tribological nature of SiC composites prepared with different types and amounts of graphene-based sources. They demonstrated the importance of graphene fillers in resisting wear during sliding. According to them, the choice of the most suitable amount and type of graphene in SiC composites for tribological and mechanical applications is dependent on their specific working requirements. SiC composites with 20 vol% graphene nanoplatelets (GNPs) clearly exhibited 72% higher wear resistance compare to the monolithic SiC (see Fig. 5). It was demonstrated that an optimum combination of mechanical properties is required to provide the best wear resistance for SiC ceramics.

Amirthan et al. ⁸⁷⁾ studied the reciprocating sliding friction and wear characteristics of cotton fabric/teak wood/jute fiber based Si/SiC. Among the investigated composites, fine teak wood and cotton based samples exhibited less COF of 0.25 and 0.23 and less wear rate of 3.0 × 10⁻⁶ and 1.24 × 10⁻⁶ mm³/N.m respectively in dry sliding condition. This was attributed to the carbon film formation that acted as lubricating media on the wear track. Alumina particles, generated as debris from alumina counterbody balls acted as third body for harder surfaces. Third body abrasion of the surfaces led to high wear of harder chemical vapor infiltration treated cotton based sample. On the other hand, porous nature of the coarse wood particle and jute fiber based sample surfaces exhibited high wear.

Candelario et al.⁸⁸⁾ introduced graphite nanodispersoids into the microstructure of a fine-grained LPS SiC ceramics fabricated by spark plasma sintering and studied the sliding wear characteristics in regular diesel fuel (viscosity of ~ 3.8 cSt) lubricating conditions. Addition of graphite nanodispersoids was detrimental to mild wear and transition from mild to severe wear occurred in shorter time due to softer graphite particles. On other side, the addition of graphite nanodispersoids led to increased fracture toughness and provided external lubrication after grain pull-out. This was resulted in effectively decreasing wear damage for a prolonged sliding. Important findings from sliding wear studies of SiC based ceramics are listed in Table 2.

In this section, studies relating to material removal mechanisms during sliding wear are discussed in terms of material characteristics. Different results for frictional behavior between varieties of pairs of SiC ceramics during wear indicate complexity in frictional mechanism. Kato⁹²⁾ divided wear mechanisms of SiC ceramics in sliding wear conditions into two modes: (a) mechanical wear, where various types of cracks like lateral, median and radial cracks are generated by friction around the Hertzian contact zone (see Fig. 6), and (b) tribo-chemical wear, where silicon oxide was found on the surface in unlubricated sliding wear.^93,94)

Wang et al.⁴³⁾ found that tribological contact stresses predominantly affect mechanisms of material removal. They found that SiC ceramics removed via plastic deformation induced microfracture for less stress on contact surface. As the contact stress increased to a critical level, varieties of cracks like partial cone cracks, lateral/shallow cracks, and radial cracks etc. are generated. Further increase in contact stress propagates cracks to intersect each other and detach chunks of material. The detached material crushed into fine particles by further tribological contact and evolved as wear debris. The dominant damage patterns with contact stress severity for indentation, scratch, and wear tests are synchronized in Table 3.

Lopez et al.⁵¹⁾ reported that for equiaxed-grained SiC ceramics, sliding wear initiated with plastic dislocation as mild wear and led to sever wear, with fracture and pull-out of material.^{37,69,85,95-96)} For coarser microstructures, transition in wear mechanism occurred sooner and exhibited poor wear resistance.^{43,47,50,69-70,96-99)} However, these mechanisms were not observed for highly heterogeneous ceramics such as elongated-grain LPS-SiC.⁵¹⁾ For elongated grain LPS SiC ceramics, despite the grain-boundary fracture, interlocking network was responsible for the surface integrity of worn surface which delayed transition in wear mechanism.

Kovalcikova et al.⁸⁶⁾ studied wear mechanisms for SiC ceramics slid against different counterparts Si₃N₄, Al₂O₃, WC-Co and ZrO₂ materials. They found microfracture and removal of oxides against any counterpart. But the extent of material removal was different. The steady state coefficient of friction and specific wear rate were affected with different tribological partners. The SiC/Si₃N₄ couple exhibited the highest friction probably due to chemical similarity of counterparts.

Andersson et al.⁴⁵⁾ reported two types of oxidation in sliding at elevated contact temperatures: passive oxidation and active oxidation. At elevated temperature (usually above 575°C), SiC oxidizes to SiO₂ according to the following reactions:100-101)

Above reactions occur at an appropriate pressure in oxygen environment. If temperature is above 1000 °C, with low partial pressures of oxygen, the following reactions might occur on SiC ceramic surface.101)

The SiO powder formed was brownish to yellowish in color and often formed with β-SiC powders in an argon thermal plasma jet.102)

Under passive oxidation layer, loose wear debris consisting of SiO₂ formed a protective and less shear strength tribofilms which further favored to reduced wear rates. Against passive oxidation layer, wear rates varied from 10⁻⁶ to 10⁻⁵ mm³/N.m and the coefficients of friction was approximately 0.5. As SiO did not form tribofilms and favored to active oxidation to some extent, wear rates increased to 10⁻⁵ -10⁻⁴ mm³/N.m and the coefficients of friction to 0.8. It was found that the additive composition had minimal effect on frictional behavior of SiC ceramics under dry sliding conditions when compared to sliding load.²⁾ Wear mechanisms changed from surface grooving and micro-cracking at lower load (1 N) to dominant tribochemical wear at higher (6 or 13 N) load for all the ceramics. In other study,⁸⁴⁾ sliding wear behavior of SiC ceramics sintered with (0.2 wt.% or 3 wt.%) Y₂O₃ additive, microcracks induced fracture and pull-out were found responsible for the material removal.

In self-mated sliding of SiC ceramics, Gahr et al.⁴¹⁾ reported that running-in period was much longer in vacuum than in humid air and would favor tribochemical smoothening of surface roughness. At a tribocontact of relatively smooth surface, formation of thin water film led to increased friction owing to menisci formed at asperities. On the other hand, thick water films separated the mated bodies and resulted for reduced friction due to low viscosity of H₂O and assumed a proper wettability of tribosurfaces with water film, i.e. hydrophilic surfaces.

Available data on the effect of humidity on coefficient of friction and wear rate of self-mated SiC ceramics is compiled and presented as Fig. 7.^{72, 85, 91, 93)} Murthy et al.⁷²⁾ found that at lower humidity, major wear mechanism was fracture of tribosurface followed by attrition, while tribochemical reaction (oxidation or hydrolysis) was a leading wear mechanism at higher humidity. Owing to the availability of more water molecules, oxidation (SiC + 2H₂O → SiO₂ + CH₄) and hydrolysis (SiO₂ + H₂O → Si(OH)₄) reactions were accelerated at higher humidity. The faster kinetics of tribochemical reactions resulted in decreased wear with a narrow difference in friction.

Li et al.⁸⁵⁾ observed difference in wear mechanism during sliding in water lubrication conditions for SiC ceramics prepared through different processing routes. The wear occurred by pull-out of grains, which was mainly controlled by free Si along grain boundary for the reaction sintered SiC. Tribochemical oxidation in water and local grain fracture occurred due to its low toughness for the pressureless sintered SiC. Excellent tribological properties and low friction were attributed for the hydrodynamic lubrication for the toughened SiC. It was believed that tribochemical oxidation took place and an ultraflat surface consisting of SiO₂ and (SiO₂.nH₂O) formed on the worn track.

Candelaria et al.⁸⁸⁾ found less wear of SiC with graphite nanodispersoids compared to SiC ceramics, due to the combination of (i) higher toughness, which hinders the crack propagation and coalescence responsible for the grain pullout, (ii) lower hardness of the wear debris due to its lower SiC concentration, and (iii) external lubrication imposed by pull-out of graphite nanoparticles, which reduced abrasion by third bodies. Zhu et al.103) studied polishing of SiC ceramics in oxidant solution of 3 wt% CrO₃. The removal of material by a chemical dissolution, stimulated by friction was referred as tribochemical.

Ciudad et al.⁷⁴⁾ studied wear mechanisms in diesel for pressureless LPS-SiC. They found a two-step wear process, initially controlled by plastic deformation (corresponding to mild wear regime) followed by fracture of materials (corresponding to severe wear regime). Major characteristics of surfaces in mild wear and severe wear regimes are schematically illustrated in Fig. 8. Erickson et al.104) observed chemical wear on mated surface as major deterioration mechanism in oil/water lubricated sliding conditions along with several mechanisms like plastic deformation, subsurface microcrack formation, microfracture and microabrasion of the surface of worn SiC. The absence of tribofilm on the contact surface indicates that wear debris transported out of the system or directly dissolved by the oil/water lubricant as water is able to dissolve SiO₂ easily.105-107)

Zhou et al.⁷⁵⁾ reported that the tribochemical reaction led to material removal for monolithic SiC and SiC-C (graphite) composite. For SiC, they found rough worn surfaces resulted in fracture and three body abrasion, whereas graphite particles exerted a lubricating effect and resulted lower coefficients of friction for the SiC-C composites. Micele et al.⁹⁰⁾ studied the tribological characteristics of SiC-MoSi₂ composites against alumina in dry sliding. The material loss was caused by ploughing effects, while no significant effect of brittle fracture observed on the worn surface of the ceramic composite.

Recently, sliding wear mechanisms were studied for monolithic SiC and SiC-WC composites against SiC counterbody. ⁷⁶⁾ For monolithic SiC ceramics, blunt indenter-induced mechanical fracture with micro-cracking was found as dominant material removal mechanism, while WC addition reduced grain pull-out for SiC-WC composites leading to reduced wear rates. Analysis of debris collected from worn surfaces indicated increase in compaction or oxidation with WC addition. Wasche and Klaffke108) studied the wear mechanism for SiC-TiC-TiB₂ in water when slid against SiC or Al₂O₃. In case of SiC/SiC and Al₂O₃/SiC, no films were formed. An oxide film was formed during wear of SiC-TiC and SiC-TiC-TiB₂ materials. Reduction in wear of materials was affected by formation of titanium oxide at the tribocontact. In case of Al₂O₃ ball and composite disk system, the tribooxidatively formed layer was partly adhered on to the ball. The wear of SiC-TiC or SiC-TiC-TiB₂ disk against SiC or Al₂O₃ ball was reduced by a factor of 10 compared to monolithic SiC disk system.

Solid particle erosion of SiC ceramics and their composites is reported to occur generally by brittle fracture as a result of lateral and radial cracking.^62,67,110-113) Bell and Rogers118) noted fracture toughness as predominant factor in estimating erosion wear characteristics of brittle materials at normal impact of erodent,^65,119) whereas hardness was found important at low impingement angles. Shetty et al.120) also reported hardness as predominant property of reaction-bonded SiC ceramics to estimate erosion behavior of ceramics. Routbort and Matzke114) studied the erosion for SiC ceramics prepared through hot-pressing and reaction-bonded sintering techniques. The Vickers hardness (H) and fracture toughness (K_Ic) of these ceramics were used to estimate the erosion rate (ΔW) of ceramics using following equation:

where, C is proportionality constant and the exponents a and b valued according to details of the used erosion model 114) (see Table 4). However, the wear rates estimated using existing wear models were not in agreement with the experimentally measured wear rates. This can be primarily attributed to the fact that these models neglected important contributions from microstructure of the target material in estimating erosion loss.^65-67,112-114)

Wang et al.⁶⁵⁾ studied the effect of microstructure on erosive wear of SiC ceramics, prepared by pressureless sintering, hot pressing and hot isostatic pressing techniques. Microcracking with some extent of plastic deformation was observed as main material removal mechanism at higher impact angles.121-122) Grain refinement and reduction in the amount and size of pores of hot isostatically pressed SiC resulted in superior wear resistance.

SiC ceramics sintered with 10, 30 or 50 wt% WC particles via hot press technique were subjected to erosion against SiC and alumina erodent.⁶¹⁾ Compared to erosion rates against alumina erodent, erosion rates against SiC erodent increased by two orders (from 10 to 10³ mm³/kg) for normal impingement (see Table 5). SiC ceramics achieved steady state erosion after 5 min of erosion against SiC erodent which delayed to 20 min against alumina. Maximum erosion resistance was obtained for SiC ceramics sintered with 30 wt% WC which possessed maximum hardness. SiC ceramics sintered with 50 wt% WC ceramics having highest fracture toughness showed lower wear resistance than SiC ceramics sintered with 30 wt% WC (see Fig. 9). This implies that hardness played important role in erosion of SiC-WC composites.⁶¹⁾

Wang and Levy121) studied the erosion behavior of SiC fiber reinforced SiC composite at 25°C and 850°C and found an inverse relation of density and hardness to the erosion rate of the composite. The erosion rate reduced by one order of magnitude at 850°C (0.10 mg/g) compared to that at 25°C (2.20 mg/g). At 850°C, the difference between erosion rates at 90° and 30° (0.02 mg/g) was much less than that found at 25°C (1 mg/g), which attributed to increased ductility of composite at 850°C. Similarly in a recent study,⁶³⁾ solid particle erosion rate of SiC-Si₃N₄ composite at elevated temperature increased firstly up to 800°C, and then decreased with increase in temperature to 1400°C. At lower temperature (up to 800°C), wear occurred by brittle fracture where stripping of aggregates of composite resulted in loss of protection from the matrix. On the other hand, plastic deformation and oxidation protection dominated at higher temperature (beyond 800°C). The highest erosion rate at 800°C for SiC-Si₃N₄ is contradictory to the lowest erosion at 850°C for SiC ceramics.⁶⁴⁾

Amirthan et al.124) found cleavage like brittle fracture on biomorphic Si/SiC composite surface exposed to erosion. Fine teak wood particle based Si/SiC composite showed least erosion compared to other Si/SiC composite. Further, coarse teak wood particle based Si/SiC composite with large pores caused the highest erosion rate. Lopez et al.125) studied the erosion behavior of biomorphic (eucalyptus and pine reinforced) SiC, reaction-bonded SiC, and hot-pressed SiC and found that erosion resistance was highest for biomorphic based SiC ceramics. Owing to the highest hardness, hot-pressed SiC exhibited least erosion rate.

The dominant mechanisms of material removal in solid particle erosion wear conditions of brittle ceramics are elastic-plastic deformation based micro-fracture, and crack formation below the plastic zone of subsurface.^62,67,110-112) Major findings from studies on erosion wear mechanisms for SiC ceramics are listed in Table 6.

A detailed investigation of impact sites of worn surface in brittle materials exhibits an intense plastic deformation beneath the immediate area of the contact.126-129) The elasticplastic zone at the impact site is the major driving force for the surface fracture that led to material loss on erosion.130-132) The residual stresses generated under plastic zone on impact of erodent to the brittle materials originate small lateral cracks to grow beneath the impact sites. Initially these cracks propagate parallel to the impacted zone of target surface and further interact to the surface resulting in material loss. Wang and Mao133) observed that size and type of cracks generated on the eroded surface are dependent on several factors like erodent shape and size, mass flow rate, velocity and angle of impingement of erodent, temperature and hardness, toughness and microstructure of target material. ^{61, 65,} 113-114)

Two types of crack systems are generally found in erosion at low velocities; (a) cone cracks generated due to impact of blunt (rounded) erodent113,134-135) and (b) lateral and median cracks generated due to impact of sharp (angular) erodent angular.113,126,136-137) In case of high velocities of erodent, material is removed by plowing, severe cracking and chipping. 110,113,130,132,134) In other study,⁶⁷⁾ precise mechanisms for the erosion of reaction bonded-SiC were not clear, but reduced erosion wear rate of mixed phase SiC-Si region was related to either by arresting of cracks around the small SiC grains or by increased apparent volume fraction of uncovered erosion resistant fine SiC grains of the two-phase region, generated beneath the impacted sites.

Suh et al.137) studied the effect of SiC fiber reinforcement on erosion wear of SiC ceramics. During erosion of the composite, initially SiC fibers were detached from the interface and matrix of SiC subsequently eroded out. Pores presented in matrix with aperture around the fibers led to easy removal of material from the impact sites. The erosion wear was also attributed to issues regarding durability and reliability that depend on geometric discordance, irregular performance, and related energy dissipation.

Gochnour et al.138) studied erosion behavior of SiC-Al₂OC composites against alumina erodent. They found surface appeared smooth and polished, with grain pull-out regions in erosion of SiC-5wt % A1₂OC composites. Softer Al₂OC phases resulted in grain pull-out regions and smooth zones were corresponded to wear of harder SiC phase. However, severe grain pull-out occurred in erosion of SiC-50 wt % A1₂OC composite and resulted in isolated smooth regions only. The steady-state erosion rate (ΔE) for brittle materials was proportional to erodent size and found to be related as

where R is the erodent radius.113,139) This was predicted by both the dynamic113) and the quasi-static139) theories of erosion for brittle materials. However, Kim and Park140) found that erosion rates of SiC and SiC-TiB₂ were not increased monotonically with the increased erodent size, in contrast to the prediction by theories. The eroded surface of SiC exhibited deformed lips and cracks, whereas intergranular fracture observed together with deformed lips for SiC-TiB₂ composite.

Limited reports are available on erosion behavior of SiC ceramics and composites as function of temperature. Wang and Levy⁶⁴⁾ reported that erosion for SiC fiber reinforced SiC composite at 25°C occurred by cracking and chipping of the matrix as well as the fibers, while erosion rates considerably decreased at 850 °C owing to the increased ductility. Li et al.⁶³⁾ studied the erosion mechanism for SiC-Si₃N₄ composites at different temperatures (25 °C-1400 °C). At low temperature, material removal occurred mainly by brittle fracture and stripping of aggregates as the protection from the matrix was lost. At higher temperature, erosion mechanisms were primarily ascribed to plastic deformation as well as oxidation protection. Oxidation of Si₃N₄ and SiC on the material surface at 1400 °C generated a thin dense layer of SiO₂ glassy film, which further reduced direct contact of erodent with the material surface at a certain extent and resulted in less loss of the material (illustrated schematically as Fig. 10).141)