Solid particle erosion of composite materials: A critical review

Abstract

This article presents a critical review of papers dealing with solid particle erosion characteristics of polymer matrix composites, metal matrix composites, and ceramic matrix composites. In addition, the solid particle erosion characteristics of coatings for composite materials are also reviewed. Attention was paid to type of a reinforcement material (fiber/filler), amount of fiber/filler, fiber orientation, and interfacial strength between fiber/filler and matrix, which affect the solid particle erosion in addition to the variables affecting the solid particle erosion of monolithic materials, that is, impact angle, particle velocity, temperature, particle flux, and erodent properties such as shape, size, hardness, and so on. General characteristics of solid particle erosion for composites are extracted from the review of the papers.

Keywords

Solid particle erosion polymer matrix composites metal matrix composites ceramic matrix composites coatings

Introduction

Various materials such as metals, polymers, and ceramics are used as structural materials in various engineering fields. Composite materials have been developed to improve mechanical properties of monolithic materials. They are composed of matrix materials and reinforcement materials, and classified based on matrix materials as follows:

polymer matrix composite (PMC),

metal matrix composite (MMC), and

ceramic matrix composite (CMC).

Various materials such as continuous long fibers, woven cloths made of continuous long fibers, short fibers, whiskers, particulates, and so on have been used as reinforcement materials. Recently, carbon nanotube (CNT) has been utilized as a reinforcement material.

Because composite materials have good mechanical properties, they have been used as structural materials in various applications such as piping of hydraulic or pneumatic transportation, noses of high velocity vehicles, blades of water or wind turbines, internal liners of piping and vessels used at elevated temperatures, turbine blades of aircraft engines, helicopter blades, and so on. In such applications of composite materials, erosion, removal of material due to impingement of solid particles, is one of the most important failure modes. The first research paper on solid particle erosion goes back to 19th century by Reynolds¹ and was followed by Rayleigh.² The review articles concerning solid particle erosion of monolithic materials were published by Tilly³ and Finnie.⁴ As far as the author knows, the first paper on the erosion of composite materials was published in 1969 by Tilly.⁵ He performed solid particle erosion experiments of two kinds of PMCs, carbon fiber (CF) reinforced nylon resin and glass fiber (GF) reinforced epoxy resin (EP). Since then a lot of papers have been published on the solid particle erosion of PMCs composed of various kinds of thermosetting and thermoplastic resins including various kinds of reinforcement materials. The papers on the solid particle erosion of MMCs and CMCs have been also published in addition to the papers on PMCs. The erosion characteristics of composite materials are complicated, compared with those of monolithic materials, because composite materials are composed of matrix materials and reinforcement materials. In composite materials, type of a reinforcement material (fiber/filler), amount of fiber/filler, fiber orientation and interfacial strength between fiber/filler and matrix affect the solid particle erosion in addition to the variables affecting the solid particle erosion of monolithic materials, that is, impact angle, particle velocity, temperature, particle flux, and erodent properties such as shape, size, hardness, and so on.

Although several review articles were published on the solid particle erosion on PMCs,^6–9 there exists no review article dealing with all types of composite materials, that is, PMCs, MMCs, and CMCs. In this review article, erosion characteristics of all types of composite materials are discussed. Only the solid particle erosion caused by particles carried with air-flow is dealt with and slurry erosion caused by particles carried with liquid-flow is out of scope in the present article. In addition, the solid particle erosion characteristics of coatings for composite materials are also reviewed in the present article.

Method for measuring erosion rate

Figure 1 shows the schematic diagram of a test apparatus used in the present author’s research group, which is composed of an air compressor, a magnetic particle feeder which keeps the feed rate of particles constant, an ejector, a nozzle, and a specimen-holder which fixes a specimen at a prescribed angle α. Before performing erosion experiments, the average velocity of solid particles ejected from a nozzle is measured using a dual rotating circular disk¹⁰ and a calibration curve relating the average velocity of solid particles with the air velocity at the exit of the nozzle measured by a pitot tube can be obtained. The average velocity of solid particles in each experiment can be decided by measuring the air velocity and using the calibration curve.

Figure 1.

Schematic diagram of erosion test apparatus (cited from Miyazaki and Hamao¹⁶).

The weight loss of a test specimen due to erosion is measured by a precision balance. Figure 2 shows a typical erosion curve in the case of nylon 66 reinforced by short GF, the weight content of which is 30 wt% (N66/GF 30 wt%). The impact angle α and average SiC-particle velocity V_p are 90° and 35 m/s, respectively. The curve is composed of an incubation period, in which no erosion occurs and a steady state, in which weight loss is proportional to particle mass impacted. The weight erosion rate can be defined by the slope of the line in a steady state, that is, weight loss per particle mass impacted. This definition is not suitable for comparing various materials, which have different densities. Therefore, the volumetric erosion rate defined by the volume loss per particle mass impacted is often used as a measure of erosion rate. The volumetric erosion rate can be calculated from dividing the weight erosion rate by the density of a test specimen. The units of weight erosion rate and volumetric erosion rate are [mg/kg] and [mm³/kg], respectively.

Figure 2.

Typical erosion curve; nylon 66 reinforced by short glass fiber, V_p = 34.0 m/s, $α = 90 \circ$ (cited from Miyazaki and Takeda¹⁵).

Polymer matrix composites

Solid particle erosion characteristics of PMCs are reviewed here by classifying PMCs into conventional PMCs, natural filler reinforced PMCs, nanofiller reinforced PMCs, and hybrid PMCs.

Conventional PMCs

Tilly and Sage^5,11 first presented the solid particle erosion experimental data on PMCs. They obtained the experimental data that the reinforcement of nylon by either GF or CF reduced the erosion resistance and that steel particles in an EP improved the erosion resistance. These experimental data are, however, insufficient to generalize above findings, because they performed erosion experiments only at the particle impact angle of 90° and did not provide the detailed information on reinforcements.

Williams and Lau¹² performed systematic solid particle erosion experiments of unidirectional CF reinforced EP by varying erodent particle size d, erodent velocity V_p and impact angle α. They showed that the weight loss due to erosion increased with increasing d and V_p. According to the erosion rate versus α, the maximum erosion rate occurred around α = 65°. In general, ductile materials show the maximum erosion rate at a lower impact angle, α = 15–30°, while brittle materials show the maximum erosion rate at α = 90°. Williams and Lau’s experimental results indicated a semi-ductile erosion behavior.

Zahavi and Schmitt¹³ presented the erosion experimental data on quartz-reinforced polybutadiene, glass cloth-reinforced epoxy laminates, quartz-reinforced polyimide (PI), and glass-epoxy laminates. Each test material had 65% reinforcement and 35% resin. Erosion experiments were performed using natural sand with the size of 210–297 µm as erodent. The sand velocity was 42 m/s and the impact angle (α) was changed from 30° to 90°. They showed the ranking of erosion resistance (hereafter, the erosion resistance is defined as an inverse of erosion rate) for four kinds of composite materials as follows: glass-epoxy laminates > glass cloth-reinforced epoxy laminates > quartz-reinforced PI > quartz-reinforced polybutadiene. The maximum weight loss due to erosion reached at α = 45° for the glass-epoxy laminates and at α = 75° for other composites. So the composite materials tested by Zahavi and Schmitt showed a semi-ductile erosion behavior. Based on the SEM observations of eroded surfaces, they proposed the erosion processes that took place in these composites as follows:

erosion and local removal of material in the resin zones,

erosion in the fiber zones associated with breakage of fiber due to bending failure of unsupported sections where resin beneath these sections is removed, and

erosion of the interface zones between fibers and the adjacent matrix.

Pool et al.¹⁴ carried out solid particle erosion experiments of four composite materials; (a) a unidirectional continuous-carbon-fiber-reinforced PI laminate (CF/PI), (b) a woven [0/90] carbon-fiber-reinforced epoxy laminate (CF/EP), (c) a woven aramid-fiber-reinforced epoxy laminate in a quasi-isotropic [0/90/±45] symmetric lay-up (AF/EP), and (d) a chopped carbon-fiber-reinforced polyphenylene sulfide (Chopped CF/PPS). Hereafter, a composite material is denoted by A/B, in which A and B represent a reinforcement material and a matrix material, respectively. In the erosion experiments, silica sand of 155 µm diameter was used as erodent, and the particle velocity was 31 m/s. The ranking of erosion resistance obtained was as follows: Chopped CF/PPS > AF/EP > CF/EP > CF/PI. Erosion resistance varies with the amount of a reinforcement material as shown later of the present article. Because the amount of a reinforcement material is not given in Ref. 14, this ranking cannot be generalized. In the cases of CF/PI and CF/EP, the maximum erosion rate was located at the impact angle α of 90°. These composites were composed of CF as a reinforcement and thermosetting PI or EP as a matrix, both of which were brittle materials. So the CF/PI and CF/EP composites showed a brittle erosion behavior. On the other hand, either of a reinforcement or a matrix was a ductile material in the cases of the Chopped CF/PPS and AF/EP composites. That is, both thermoplastic PPS resin in the Chopped CF/PPS composite and aramid fiber in the AF/EP composite were ductile materials. Thus, the Chopped CF/PPS had a ductile erosion behavior with the maximum erosion rate at α = 25°, and the AF/EP composite had a semi-ductile erosion behavior with the maximum erosion rate at α = 90°–45°. They also provided the erosion data on fiber orientation for the CF/PI composite. They performed two cases of erosion experiments, in which the erodent stream and the fiber alignment direction were parallel and perpendicular. The erosion rate of the perpendicular case was larger than that of the parallel case, because of the reason that any bending stresses present were of greater magnitude in the perpendicular case than in the parallel case. Based on the above experimental results, Pool et al. pointed out the factors governing erosion rates in composites as follows: (a) the brittleness of the fibers, (b) whether the matrix is thermosetting or thermoplastic, and (c) the interfacial bond strength between the fibers and the matrix. The last item was not validated by Pool et al.’s experimental data, and later Miyazaki et al.^15,16 clarified the effect of interfacial bond strength between the fibers and the matrix on erosion rate.

Solid particle erosion experiments of unidirectional CF reinforced bismaleimide (BMI) composite, CF/BMI, were performed by Mathias et al.¹⁷ The volume content of CF was 65 to 70 vol%. The erodent particles were angular-shaped alumina abrasive with the mean diameters of 63, 130, and 390 µm. The particle velocities were 20, 40, and 60 m/s and the impact angles were 30° and 90°. Erosion experiment of neat BMI resin was performed for making a comparison with composite data. It was found that the erosion rate of composite was much higher than the matrix polymer. According to the short-beam strength and three-point flexural modulus of the CF/BMI composite, the fiber-matrix bond was supposed to be fairly weak, and the erosion of the composite material was dominated by removal of the CFs. As a consequence, the CF/BMI composite was eroded more rapidly than the matrix polymer. As for the effect of erodent size on the erosion rate, larger erodent showed higher erosion rate because of its higher kinetic energy. They also examined the effect of fiber orientations on the erosion rate at the impact angle of 30°. The perpendicular case, in which the erodent stream is perpendicular to the fiber alignment direction, showed larger erosion rate than the parallel case, in which the erodent stream is parallel to the fiber alignment direction, but the difference between both cases was small. Supplementary erosion data of the CF/BMI composite were provided by Karasek et al.¹⁸

Ballout et al.¹⁹ performed solid particle erosion experiments of unidirectional GF epoxy composites, GF/EP, the fiber and matrix weight ratios f/m (f: fiber weight, m: matrix weight) of which were 65/35, 50/50, and 35/65. Erosion experiments were performed at the impact angle α varying from 10° to 90°. The erodent particles used were alumina with 142 µm nominal diameter and the particle velocity was 66 m/s. The erosion rates of EP and S-glass, the constituents of the GF/EP composite, were respectively obtained in addition to those of the GF/EP composites. A semi-ductile erosion behavior with the maximum erosion rate at α = 45° was observed for the GF/EP composites with f/m = 50/50 and 35/65. On the other hand, the erosion rate of the GF/EP composite with f/m = 65/35 exhibited a brittle erosion behavior with the maximum erosion rate at α = 90°. Judging from the fact that the EP and S-glass showed a semi-ductile erosion behavior and a brittle erosion behavior, respectively, the shift of the location where the maximum erosion rate occurred to a higher impact angle for the GF/EP composite with f/m = 65/35 was due to the high inclusion of GF. Composites showed higher erosion rate than the neat EP, and the erosion rate increased with increasing reinforcement fiber content. It should be noted that the reinforcement fiber may reduce the erosion resistance. The linear rule of mixture (LROM) or the inverse rule of mixture (IROM) can be applied to predicting the erosion rate ER_c of composite materials. LROM and IROM are given as follows

LROM : {ER}_{c} = w_{f} \times {ER}_{f} + w_{m} \times {ER}_{m}

(1)

IROM : \frac{1}{{ER}_{c}} = \frac{w_{f}}{{ER}_{f}} + \frac{w_{m}}{{ER}_{m}}

(2)

where w and ER represent the weight fraction and erosion rate, respectively, and the subscripts f and m represent the quantities relevant to fiber and matrix. Ballout et al. calculated the erosion rates of the composite from the above equations using the experimental erosion rate data of the EP and S-glass, and found that the IROM was able to predict the erosion rate of composite better than LROM.

Miyazaki et al.^15,16,20 performed a series of solid particle erosion experiments of PMCs. They presented erosion data on short GF or CF reinforced thermoplastic resins. In Ref. 15, test materials were nylon 66 neat resin (N66), ABS neat resin (ABS), and their composite materials, that is, nylon 66 resin reinforced by short GF (GF/N66) and by short CF (CF/N66), and ABS resin reinforced by short GF (GF/ABS) and by short CF (CF/ABS). The detailed composition of these materials is summarized in Table 1. In Ref. 20, test materials were polyetheretherketone resin (PEEK) and new type of thermoplastic PI resin (New-TPI), and their composite materials. Table 2 provides the detailed composition of these materials. The erodent used was SiC particles with the diameter ranging from 100 to 150 µm. Figures 3(a) and (b) show the relation between the volumetric erosion rate and the impact angle α for the PEEK composites and the New-TPI composites respectively, in the case of particle velocity V_p = 34 m/s, together with the results of the corresponding neat resin.²⁰ As shown here, the erosion rates of composites are higher than those of the corresponding neat resins and increase with the increase in fiber content. The CF composites show the higher erosion rate than the GF composites, when the fiber weight contents are the same for both composites. The same results were obtained for N66 composites and ABS composites.¹⁵ Figure 4 shows the relation between the volumetric erosion rate and the fiber volume content for the PEEK composites and the New-TPI composites for V_p = 34 m/s and α = 60°.²⁰ The plotted points are almost on the same curves, independent of kinds of fiber, which indicates that the erosion rate of a composite is dominated only by the volume content of short fibers, if the matrix material is the same.

Table 1

Composition of tested materials; N66, GF/N66, CF/N66, ABS, GF/ABS and CF/ABS.

Designation	Fiber weight content (%)	Density (g/cm³)	Fiber diameter (µm)	Fiber length (µm)
N66	0	1.15	—	—
GF/N66	30	1.38	10–13	n.a
CF/N66	30	1.26	10–13	n.a
ABS	0	1.04	—	—
GF/ABS	20	1.19	10	n.a
CF/ABS	20	1.12	10	n.a

Note: n.a.: not available.

Figure 3.

Relation between volumetric erosion rate and impact angle; V_p = 34.0 m/s (cited from Miyazaki and Hamao²⁰). PEEK composites. (b) New-TPI composites.

Figure 4.

Relation between volumetric erosion rate and fiber volume content; V_p = 34.0 m/s (cited from Miyazaki and Hamao²⁰).

Table 2

Composition of tested materials; New-TPI, GF/New-TPI, CF/New-TPI, PEEK, GF/PEEK and CF/PEEK.

Designation	Fiber weight content (%)	Density (g/cm³)	Fiber volume content (%)	Average fiber diameter (µm)	Average fiber length (µm)
New-TPI Matrix	0	1.33	0	—	—
New-TPI GF15%	15	1.44	8.5	13	130
New-TPI GF30%	30	1.55	18.4	13	130
New-TPI CF15%	15	1.37	12.4	7	150
New-TPI CF30%	30	1.42	25.3	7	150
PEEK Matrix	0	1.30	0	—	—
PEEK GF15%	15	1.39	8.3	13	130
PEEK GF30%	30	1.51	18.0	13	130
PEEK CF15%	15	1.34	12.1	7	150
PEEK CF30%	30	1.38	25.1	7	150

In order to clarify the effect of the bond strength between reinforcement fibers and matrix resin on the erosion behavior of composites, Miyazaki et al.¹⁵ performed solid particle erosion experiments using unsaturated polyester (UP) resin reinforced by woven glass-cloths (GF/UP) and EP reinforced by unidirectional CFs (CF/EP).¹⁶ The solid particles used were SiC abrasive with the diameter ranging from 100 to 150 µm. Tables 3 and 4 show the material properties of the GF/UP and CF/EP composites. In the erosion experiments of the GF/UP composites, four kinds of composites designated by A, B, C, and D were used in addition to the UP neat resin designated by E. The materials A, B, and C were composites whose fiber weight contents were about 60 wt% but they had different fiber surface treatments. The fiber surface was treated by acryl silane in the material A, treated by chrome in the material B, and untreated in the material C. According to fiber surface treatments, they had different three-point bending strength as shown in Table 3. The three-point bending strength of the material C was only 67% of that of the material A. We can use the three-point bending strength as an indicator of the interface strength between fibers and a matrix resin. The material D had the same fiber treatment as the material A, but its fiber weight content was less than the material A. In the erosion experiments of the CF/EP composites, two kinds of composites designated by materials B and C were used in addition to the EP neat resin designated by the material A. CFs whose surfaces were treated to strengthen the bond between a matrix material and fibers were included in the material B, while CFs with untreated surfaces were included in the material C. As shown in Table 4, the tensile strength of the material B was nearly 1.6 times as large as that of the material C due to making the interface between a matrix material and fibers strong. Figure 5 shows the relation between the volumetric erosion rate R and the impact angle α in the case of particle velocity V_p of 35 m/s for the GF/UP composites.¹⁵ It is found from the figure that the composites A, B, C, and D have much larger erosion rate than the corresponding neat resin E. This can be confirmed by Figure 6 showing a cross-sectional view of the eroded surface of the composite C. We can see from the figure that the slope of the eroded surface is steeper in the GF regions than in the matrix regions. The fact indicates that the GF regions were selectively damaged and suffered greater damage than the matrix regions. This is an evidence for the fact that the erosion rate of a composite is larger than that of the corresponding neat resin. The peaks of the erosion rates of the GF/UP composites shift to higher impact angles. Comparing the results of the materials A, B, and C of the UP composites which have nearly the same fiber content but different fiber surface treatments, we can find that the erosion rate of the composites decreases with the increase in the interface strength between the matrix and fibers. That is, the material A, which has the strongest interface strength, shows the best erosion resistance except around high impact angles, and the material C, which has the weakest interface strength, shows the worst erosion resistance. Comparing the results of the materials A, D, and E, we can find that the erosion rate of the composite decreases with the decrease in fiber weight content. Figure 7 shows the relation between volumetric erosion rate R and impact angle α in the case of particle velocity V_p of 55.7 m/s for the CF/EP composites and the EP resin.¹⁶ The erosion experiments of the composites were performed in the cases, where the erodent stream and the fiber alignment direction were parallel and perpendicular. In Figure 7, β = 0° and β = 90° indicate parallel impact and perpendicular impact, respectively. The following are found from the figure: (a) The maximum erosion rate is located around α = 30° for the neat resin A and around α = 60° for the composites B and C, that is, the peak of erosion rate for the composites shifts to a higher impact angle compared with the corresponding neat resin, (b) the erosion rate of the composites is larger than that of the corresponding neat resin, (c) the erosion rate of the composite with untreated fibers is larger than that of the composite with treated fibers, and (d) difference in erosion rate is found between the parallel impact (β = 0°) and the perpendicular impact (β = 90°), that is, the perpendicular impact has larger erosion rate than the parallel impact at lower impact angles α, and vice versa at higher impact angles. The experimental results (a) to (c) are similar to those of the GF/UP composites shown in Figure 5. The experimental result (d) partly verifies the Pool et al.’s¹⁴ results, in which the erosion rate of the parallel impact was always higher than that of the perpendicular impact at the impact angle α ranging from 0° to 90°.

Table 3

Material properties of tested GF/UP composites.

Designation	Fiber surface treatment	Fiber weight content (%)	Density (g/cm³)	3-point bending strength (MPa)
A	Acryle silane	60.4	1.79	556
B	Chrome	60.3	1.80	446
C	None	58.2	1.75	374
D	Acryle silane	44.5	1.61	384
E	—	—	1.22	n.a.

Note: E: neat resin; n.a.: not available.

Figure 5.

Relation between volumetric erosion rate and impact angle for GF/UP composites; V_p = 35.0 m/s (cited from Miyazaki and Takeda¹⁵).

Figure 6.

Cross-sectional view of eroded surface of GF/UP composite C taken by optical microscope; V_p = 35 m/s and $α = 90 \circ$ (cited from Miyazaki and Takeda¹⁵).

Figure 7.

Relation between volumetric erosion rate and impact angle for CF/EP composites; V_p = 55.7 m/s (cited from Miyazaki and Hamao¹⁶).

Table 4

Material properties of tested CF/EP composites.

Designation	Fiber weight content (%)	Density (g/cm³)	Fiber volume content (%)	Average fiber diameter (µm)	Tensile strength (MPa)
A	0	1.20	0	—	—
B	65.0	1.67	52.7	9	38.7
C	61.4	1.62	48.8	9	23.9

Afterwards Miyazaki and Funakura²¹ performed solid particle erosion experiments of a composite degraded by hot water. The composite used was UP resin reinforced by chopped glass strand mats with the fiber volume content of 59 vol%. The test specimens were soaked in hot water of 80℃ during 50–1000 hrs to cause damage. The solid particles used were SiC abrasive with the diameter ranging from 100 to 150 µm. The Vickers hardness of a resin rich region was measured, and the interlaminar shear strength of the composite was obtained from a short beam three-point bending test. According to these experimental data, the resin became soft and the interface between the matrix and fibers was weakened due to degradation by hot water. The relation between the volumetric erosion rate and the impact angle are shown in Figure 8 for various immersion times in the case of particle velocity V_p of 48.9 m/s. In this figure, “Normal” means the data for a virgin composite and each numeral such as “100 hrs” means the data for a composite degraded during immersion time. Both the effects of decreasing hardness of the matrix resin which results in the decrease in the erosion rate²² and decreasing the interfacial strength which results in increasing in the erosion rate^15,16 are expected to appear in the erosion curves of the composites shown in Figure 8. Additional erosion experiments were performed in the cases of a low particle velocity V_p = 19.0 m/s and a high particle velocity V_p = 59.0 m/s to separate the above two effects. At the low particle velocity, the damage of the interface between the matrix rein and fibers may be small because of small kinetic energy of particles. In such a case, the effect of decreasing hardness of the matrix is relatively enhanced by suppressing the effect of decreasing interfacial strength between the matrix resin and fibers. The results of V_p = 19.0 m/s are shown in Figure 9, in which the volumetric erosion rate of the degraded composite is lower than that of the virgin composite because of the effect of decreasing hardness of the matrix resin. At the high particle velocity, the damage of the interface between the matrix resin and fibers cannot be ignored. In such a case, the decrease in interfacial strength due to degradation by hot water results in increase of erosion rate, which cancels the effect of decreasing hardness of the matrix resin. Therefore, in the case of V_p = 59.0 m/s shown in Figure 10, the volumetric erosion rate of the degraded composite approaches that of the virgin composite in the range of higher impact angles and is larger than that of the virgin composite in the range of lower impact angles.

Figure 8.

Relation between volumetric erosion rate and impact angle for unsaturated polyester resin reinforced by chopped glass strand mats; V_p = 48.9 m/s (cited from Miyazaki and Funakura²¹).

Figure 9.

Relation between volumetric erosion rate and impact angle for unsaturated polyester resin reinforced by chopped glass strand mats; V_p = 19.0 m/s (cited from Miyazaki and Funakura²¹).

Figure 10.

Relation between volumetric erosion rate and impact angle for unsaturated polyester resin reinforced by chopped glass strand mats; V_p = 59.0 m/s (cited from Miyazaki and Funakura²¹).

Miyazaki²³ carried out experiments to study the solid particle erosion of composites with prior impact damage. Two kinds of EP composites reinforced by aramid fibers (AR/EP) and by poly p-phenylene-benzobisoxazole fibers (PBO/EP) were used. The AR/EP and PBO/EP laminates with the same layout structure had the volume contents of 55.8 and 53.8 vol%, respectively. Prior to erosion experiments, the test materials were damaged by the impact of a spherical projectile with the velocity of 130 m/s shot from an air gun. Solid particle erosion experiments were then performed for the damaged composites. The erosion experiments were performed at the particle velocity V_p = 57.8 m/s and SiC particles with the diameter ranging from 100 to 150 µm were used as erodent. The experimental results showed that the volumetric erosion rates of the composites with prior impact damage were less than those of the virgin materials. Less erosion rates of the composites with prior impact damage were probably because the composites with prior impact damage had a lot of transverse cracks and delaminations, which absorbed the kinetic energy of the solid particles used in the erosion tests.

Roy et al.²⁴ performed erosion experiments of four different types of PMCs reinforced by GF, that is, glass epoxy-resin, glass-phenolic resin (modified), glass-phenolic resin (unmodified), and glass-polyester resin. The glass-phenolic resin (modified) means that phenolic resin is modified so as to improve the mechanical properties of the composites. For the purpose of reinforcement, E-glass plain weave woven roving fabric was used. The eroding particles were silica sand with the size of 200 µm. Erosion rates of these composites were evaluated at two impact angles α = 30° and 90°, and two particle velocities V_p = 38 and 45 m/s. According to the erosion experimental data, the glass-EP exhibited the highest erosion resistance at all impact angles and impact velocities, while the glass-phenolic resin (modified) composite exhibited the least erosion resistance. They discussed the nature and mechanism of erosion based on the erosion efficiency parameter η proposed by Sundararajan et al.²⁵ This parameter is defined as the fraction of the volume that is actually removed as erosion debris out of that which is displaced, and given by the following equation

η = \frac{2 E_{r} H_{V}}{ρ V_{p}^{2}}

(3)

where E_r is the weight erosion rate, H_V is the hardness, ρ is the density of a target material and V_p is the particle velocity. This equation is valid for normal impact erosion (α: impact angle,

α = 90 \circ

). The extension of this equation for oblique impact erosion will be shown later. When ideal microplowing involving deformation of a target material from the crater without fracture is induced in a target material, the parameter η will be zero. On the other hand, in the case of ideal cutting, η will be unity. When erosion occurs by the formation of a lip and its subsequent fracture, η will be in the range from zero to unity. In the case of brittle materials, erosion occurs by spalling and interlinking of lateral or radial cracks that results in the removal of large chunks of material, the parameter η is expected to be even greater than unity. The values of the erosion efficiency parameter η for composite materials tested varied from 19 to 43% for

V_{p} = 45

m/s and from 12 to 28% for

V_{p} = 38

m/s. Thus, it was concluded that the erosion mechanism was the formation of a lip and its subsequent fracture. According to Ref. 24, the erosion efficiency values of the PMCs shown above are lower than those of polymer resins calculated from the literature data. Roy et al. thus stated that this was because hard reinforcement fibers resisted the crack propagation in a resin region and reduced the erosion efficiencies. Such a statement cannot be accepted because they did not perform erosion experiments of the corresponding neat polymers. Suresh Arjula and Harsha²⁶ evaluated the erosion efficiency parameter η defined by equation (3) for various polymers and PMCs by collecting available erosion rate data from literatures, and constructed the efficiency map, variation of erosion efficiency as a function of hardness. The erosion efficiency values of the short fiber reinforced composites were slightly larger than those of the corresponding neat polymers. This indicates that the same erosion mechanism as the neat polymer may govern the polymer composite. Moreover, the thermoplastic composites showed the erosion efficiency values in the range of 4–15%, a ductile erosion behavior, whereas the thermosetting composites showed η values in the range of 20–25%, a semi-ductile behavior. According to the erosion experiments performed by Mathias et al.¹⁷ and Miyazaki et al.,^15,16,20 composite materials had poor erosion resistance compared with the corresponding neat resin. Removal of reinforcement fibers caused by the damage between the matrix and reinforcement fibers during erosion process has an important effect on the erosion characteristics of composite materials. The erosion efficiency does not take account of such a physical phenomenon. So the erosion efficiency η cannot be always used as an indicator of erosion resistance for composite materials.

Saputra et al.²⁷ studied the solid particle erosion behavior of particulate-filled composite. The composite material used was EP filled with spherical glass particles having the mean diameter of 17 µm at the volume content of 30 vol%. In order to study the effect of adhesion between glass particle fillers and matrix on erosion behavior, four types of filler treatment were used as follows: (a) glass particles treated by γ-glycidoxypropyl trimesoxysiane (GPS), (b) as-received glass particles already treated by silane coupling agent, (c) glass particles with untreated surfaces, that is, silane coupling agent was removed from the surfaces of as-received glass particles by acetone, and (d) glass particles with surfaces coated with silicone oil. Solid particle erosion experiments were performed using angular shaped glass particles with the diameter of 178, 350, and 605 µm as erodent, and at the particle velocity of 15 m/s. The following results were obtained from the experimental data:

The ranking of erosion resistance was as follows: composite with GPS treated fillers > composite with as-received fillers > composite with acetone washed fillers > composite with silicone oil coated fillers. This ranking was the same as that of flexural strength, which represented the adhesion between glass particle fillers and matrix. It was therefore concluded that the stronger adhesion between glass particle fillers and matrix showed better erosion resistance or lower erosion rate. This fact was conformed to the preceding researches performed by Miyazaki et al.^15,16

The erosion rate of composite was larger than the corresponding neat resin, and the increase in filler content of composite causes the increase in erosion rate.

The maximum erosion rate of composites was located at a higher impact angle in comparison with the corresponding resin.

The results (b) and (c) were also already obtained in preceding researches.

Solid particle erosion characteristics of polyetherimide (PEI) composites were studied by Bijwe et al.²⁸ The test specimens used were neat PEI resin (PEI), PEI composite containing 20% short GF (GF20%/PEI), and PEI composite containing 25% short GF and three solid lubricants (GF25% + SL/PEI). The erosion experiments were performed at the impact angles of 15°, 30°, 45°, 60°_, and 75°, using silica sand of about 80 µm size as erodent. The ranking of erosion resistance was in the following order; PEI > GF20%/PEI > > GF25% + SL/PEI, so that the incorporation of fibers into resin deteriorated the erosion resistance, similar to other composites.

Huang et al. ²⁹ performed solid particle erosion experiments of silicone resin reinforced by SiC or Al₂O₃ particles (SiC/Silicone, Al₂O₃/Silicone), which can be used as heat-resistant coatings for elevated temperature situations. The mean particle sizes of reinforcement particles were 1.29, 5, and 20 µm. The erodent was Al₂O₃ particles. The erosion characteristics were given for the particle velocity of 14 m/s. The erosion experiments were performed by varying the weight content of the reinforcement particle and the impact angle. According to the erosion experiments, the best erosion resistance was obtained at the SiC weight content ranging from 60 to 75 wt% for the SiC/Silicone composite and at the Al₂O₃ weight content of 60 wt% for the Al₂O₃/Silicone composite. Although the authors stated that the erosion resistance of a silicone matrix was improved greatly through introducing evenly dispersed harder second phase ceramic particles, they did not provide the erosion data on neat silicone resin. The authors should provide such data in order to clarify how much SiC and Al₂O₃ particles improve the erosion resistance of composites.

Barkoula and Karger-Kocsis³⁰ studied the solid particle erosion behavior of EP reinforced by unidirectional GF with the volume content of 68 vol%. Two kinds of composites were used. One had GFs with EP-compatible sizing (GF/EP-M), and the other had GFs without compatible sizing, that is, EP-incompatible sizing (GF/EP). The interfacial shear strength of GF/EP-M was larger than that of GF/EP, so that GF/EP-M had better adhesion between fibers and matrix resin than GF/EP. Erosion experiments were performed at the particle velocity of 70 m/s and the impact angles of 30°, 60°, and 90°, using angular corundum particles with 60 and 120 µm as erodent. Erosion experiments were also performed for two kinds of relative fiber orientations. Fiber orientations designated by Pa and Pe were respectively parallel and perpendicular to the stream of erodent particles. The experimental results showed that the good fiber-matrix adhesion improved the erosion resistance. On the hand, the erosion rate of Pe showed larger erosion rate than that of Pa, but the difference was not so large. These results verified the Miyazaki et al.'s experimental results.¹⁶ Barkoula and Karger-Kocsis³⁰ performed not only solid particle erosion experiments of thermosetting resin composites (GF/EP-M and GF/EP) but also those of thermoplastic resin composites.³¹ The test specimens of thermoplastic resin composites were as follows: polypropylene (PP) resin reinforced by randomly oriented short GF (S-GF/PP, length of GF ≈ 2 mm, fiber content = 40 wt%), PP resin reinforced by randomly oriented long GF (L-GF/PP, length of GF ≈ 10 mm, fiber content = 40 wt%), and PP resin reinforced by unidirectional GF (UD-GF48/PP, UD-GF55/PP, and UD-GF60/PP, the numeral means fiber weight content of GF). They also performed erosion experiments for both relative fiber orientations Pa and Pe for the PP resin reinforced by unidirectional GF. Erosion data showed that the erosion rate reached at the maximum at the impact angle of 30°, which indicates a ductile erosion behavior and increased linearly with the increase of fiber content in the range from 0 to 60 wt%. As mentioned before, the LROM given by equation (1) or the IROM given by equation (2) can be applied to predicting the erosion rate ER_c of composite materials. It was found from the experimental erosion rate data and the theoretical predictions of LROM and IROM that the experimental erosion rate data lay between LROM and IROM but closer to IROM than to LROM. As for the effect of fiber orientations Pa and Pe on the erosion rate, the erosion rate of Pe was larger than that of Pa for both GF/EP-M and GF/EP especially at the impact angle of 30°,³⁰ and vice versa for UD-GF/PP.³¹ The former result is consistent with those of the preceding researches,^14–17 whereas the latter one is completely different from those of preceding researches. In the preceding researches, the matrix materials of composites used were thermosetting resins such as PI resin,¹⁴ EP,¹⁶ and BMI resin.¹⁷ On the other hand, the latter result of Barkoula and Karger-Kocsis³¹ was that of the composite made of PP, a thermoplastic resin. It may be deduced that the effect of fiber orientations on the erosion rate of PMCs depends on whether the matrix material of composite is thermosetting or thermoplastic. Barkoula and Karger-Kocsis³¹ discussed based on the SEM observations of eroded surfaces why the erosion rate of Pa was larger than that of Pe for UD-GF/PP. The effect of interleaves in the CF/EP laminates on the erosion behavior was studied by Barkoula and Karger-Kocsis³². An interleaf was an adhesive interlayer of a modified EP with a thickness of 0.125 mm. The CF/EP laminates with interleaves had higher erosion resistance than those without interleaves, because interleaves reduced the fiber bending due to impact, the fragment of fibers were not so easily removed due to the better adhesion between adjacent layer, and interleaves acted as a cushion of the impacted ply.

Tewari et al.³³ studied the solid particle erosion behavior of unidirectional CF reinforced PEEK composites (UD-CF/PEEK). The composites had 65 vol% CFs. In the erosion experiments, steel balls with the size of 300–500 µm were used as erodent, and the erodent velocity and the impact angle were 45 m/s and 85 m/s, and 15–90°, respectively. They carried out three cases of erosion experiments, in which erodent stream is 0°, 45°, and 90° inclined to the fiber orientation, and these cases are respectively designated as Pa (parallel), 45 and Pe (perpendicular). The erosion rate data showed the maximum around the impact angle of 60°. This means a semi-ductile erosion behavior. The order of erosion rate for Pa, 45 and Pe cases was Pe > 45 > Pa. The result Pe > Pa is in agreement with the previous ones of Refs. 14, 16, 17, and 30, but in disagreement with the result of Ref. 31. In addition to Pe and Pa, Tewari et al.’s provided the erosion rate data on 45, and showed that the erosion rate of 45 lay between those of Pe and Pa. Recently Drensky et al.,³⁴ and Pei and Friedrich³⁵ obtained the erosion rate data similar to Tewari et al.³³ Drensky et al.³⁴ conducted an experimental study to determine the erosion resistance of a composite material of unidirectional CF reinforced PEEK (UD-CF/PEEK) at temperatures up to 260℃. About 10 µm Arizona road dust particles and 100 µm sieved runway sand particles were used as erodent, and the particle velocity was 152.4 m/s. As reported in preceding researches, the erosion rate of Pe was higher than that of Pa at 21℃. The erosion rate was found to increase with temperature except at the impact angle of 90°. Drensky et al. stated in the conclusion of their paper that the erosion rates of 90° fiber orientation (Pe) were lower than those of 0° fiber orientation (Pa) at 260℃, but they did not provide the erosion rate data validating this fact in their paper. As for the effect of erodent size on the erosion rate, 100 µm sieved runway sand particles caused higher erosion rate than 10 µm Arizona road dust particle. Pei and Friedrich³⁵ performed solid particle erosion experiments for UD-CF/PEEK with the CF volume content of 60 vol% under the particle velocity of 25 m/s, but the effect of fiber orientations on the erosion rate was very small compared with the other effects such as particle velocity and impact angle. Afterwards Friedrich et al.³⁶ examined the effect of erodent. They used angular steel grit and angular sand grit. The results showed that no remarkable differences were seen for the fiber orientations, although the perpendicular orientation exhibited slightly higher erosion rate than the parallel orientation. Friedrich et al.³⁶ also performed solid particle erosion experiments of high density polyethylene (HDPE) reinforced by polyethyleneterephthalate (PET) microfibrillar composite (PET/HDPE). The weight content of PET microfibrillar was 30 wt%. Similar to brittle fibers such as GF and CF, the inclusion of ductile PET microfibrills in HDPE resin increased the erosion rate compared with the neat resin. That is, the erosion resistance of the composite PET/HDPE was worse than the HDPE resin. The erosion rates of 200 µm thick PEEK film and 300 µm thick thermoplastic polyurethane (TPU) film were evaluated, and they were much less than the erosion rate of the composite CF/PEEK. This fact indicates that the use of such films can be considered as a possible method for protecting composites surface from solid particle erosion as used in the wind energy rotor blade industry.

Tewari et al.³⁷ also performed the solid particle erosion experiments of unidirectional CF or GF reinforced epoxy composites (UD-CF/EP or UD-GF/EP), the experimental conditions of which was the same as those of UD-CF/PEEK performed by Tewari et al.³³ The volume contents of fibers were 56 vol% for UD-CF/EP and 53 vol% for UD-GF/EP, respectively. The comparison between UD-CF/EP and UD-GF/EP showed that the weight erosion rate of UD-GF/EP was higher than that of UD-CF/EP under the condition that both UD-CF/EP and UD-GF/EP composites have almost the same fiber volume contents. Although the composite materials used in the erosion experiments in both Refs. 33 and 37 were different, the same results were obtained for the effect of a fiber orientation to erodent flow on the erosion rate.

Harsha et al.³⁸ performed solid particle erosion experiments of short fiber reinforced composites. The fibers used were GF and CF, and the resins used were three types of polyaryletherketones (PAEKs) having different chemical structures and ketone/ether linkage ratios, that is, PEEK, polyetherketone (PEK), and polyetherketoneketone (PEKK), the ketone/ether linkage ratios of which were 33, 50, and 67%, respectively. They used four kinds of PEEK composites, the filler contents of which were 20 wt% GF, 30 wt% GF, 30 wt% CF and combination of three types of fillers, 10 wt% particulate polytetrafluoroethylene (PTFE), 10 wt% particulate graphite and 10 wt% CF, three kinds of PEK composites, the filler contents of which are 10, 20, and 30 wt% GF, and two kinds of PEKK composites, the filler contents of which were 30 wt% CF and combination of three types of fillers, 10 wt% particulate PTFE, 10 wt% of particulate graphite, and 10 wt% of CF. The erosion experiments of neat resins, PEEK, PEK, and PEKK, were conducted in addition to composite materials. In the erosion experiments, angular-shaped silica sand particles of the size ranging between 150 and 212 µm were used as erodent. The impact angles and the particle velocities were 15°, 30°, 60°, and 90°, and 39, 68, and 90 m/s, respectively. The following conclusions were obtained from the experimental results:

All composite materials had larger erosion rates than the corresponding neat resins.

The erosion rate increased with increase in filler content.

The peak of erosion rate for some composite materials, that is, PEEK with 30 wt% GF, 30 wt% CF and the combination of PTFE, graphite and CF, shifted to a higher impact angle than that of the neat PEEK resin, because of inclusion of brittle fillers in resins with ductile nature of erosion behavior.

The CF composites showed the higher weight erosion rate than the GF composites, when the fiber weight contents were the same in both composites.

Fillers such as PTFE and graphite were detrimental to erosive wear performance.

Among the conclusions mentioned above, the conclusions (a), (b), and (c) are very common findings that many researchers were already pointed out.^{11,15,16,19,20,27} Considering that the CF composite has lower density than the GF composite under the condition of the same fiber weight content, the conclusion (d) was the same as Miyazaki and Hamao²⁰ pointed out in their preceding research.

Harsha and Thakre³⁹ and Suresh Arjula et al.⁴⁰ performed solid particle erosion experiments of PEI matrix composites. Harsha et al.³⁹ used PEI resins reinforced by short E-GF, short CF, and combination of short E-GF and solid lubricants (PTFE, graphite, MoS₂). The erosion experiments given in Ref. 39 were almost the same as those given in Ref. 38 except that the composites used in the erosion experiments were different, that is, PEI matrix composites in Ref. 39 and PAEKs matrix composites in Ref. 38. There was no new finding in this research compared with their previous one.³⁸ Suresh Arjula et al.⁴⁰ used unidirectional CF reinforced PEI (UD-CF/PEI). In this case, the difference lay only in a matrix material compared with the erosion experiments performed in Ref. 33. That is, PEI resin was used in Ref. 40, whereas PEEK resin in Ref. 33. The conclusions obtained in Ref. 40 were essentially the same as those obtained in Ref. 33

Suresh Arjula et al.⁴¹ examined solid particle erosion characteristics of four kinds of composite materials, UD-GF/PEEK, UD-CF/PEEK, UD-GF/PEKK, and UD-CF/PEKK, the fiber volume contents of which were 60 vol%. The erosion experiment of PEEK neat resin was also performed in addition to the composites. Angular silica sand particles with the size of 150–250 µm were used as erodent. The erosion data obtained in this research verified the following solid particle erosion phenomena already obtained in the preceding researches:

The erosion rate of a composite is larger than that of a corresponding neat resin.

The peak of the erosion rate of a composite shift to a higher impact angle due to the inclusion of brittle fibers.

A GF reinforced composite has higher weight erosion rate than a CF reinforced composite, if the matrix material and the filler volume content are the same in both composites.

The erosion rate of a composite in the case of perpendicular impact to the fiber orientation is higher than that of a composite in the case of parallel impact, especially in the lower angle below 60°.

Suresh Arjula et al. provided the ranking of erosion resistance as follows: PEEK > UD-CF/PEEK > UD-CF/PEKK > UD-GF/PEKK > UD-GF/PEEK. Although a series of Harsha et al.’s researches^37–41 are useful for constructing erosion database, they provided almost no new finding for the solid particle erosion characteristics of PMCs.

As for PEEK-based composites, Sari et al.⁴² carried out solid particle erosion experiments of short GF (S-GF) or CF (S-CF) reinforced PEEK resin composites (S-GF/PEEK or S-CF/PEEK) at the low erodent particle velocity of 1.57 m/s. Sand-blast type or slinger type erosion test rigs were used for solid particle erosion experiments under a high particle velocity, while Sari et al. used a low velocity particle erosion rig consisting of a chamber full of erodent particles and a sample holder rotating at the fixed speed. Fiber weight contents of both composites, S-GF/PEEK and S-CF/PEEK, were 30 wt%. Brown fused Al₂O₃ particles with the size of 500–710 µm were used as erodent. Both composites showed a semi-ductile erosion behavior with the maximum erosion rate at the impact angle of 45°. The weight erosion rate of S-GF/PEEK was smaller than that of S-CF/PEEK under the condition of the same fiber weight content. Considering that S-GF/PEEK has larger density than S-CF/PEEK under the condition of the same fiber weight content, S-CF/PEEK has larger volumetric erosion rate than S-GF/PEEK. This fact was already pointed out by Miyazaki and Hamao.²⁰ They also pointed out in Ref. 20 as shown in Figure 4 of the present review article that the erosion rate of a composite was dominated only by the volume content of short fibers irrespective of a fiber material, CF or GF, if composites reinforced by short fibers were made of the same matrix material. The erosion data obtained by Sari et al.⁴² was insufficient to confirm this point.

Suresh Arjula et al.⁴³ performed sold particle erosion experiments of PPS reinforced by short GF (S-GF/PPS) with varying fiber content from 0 to 40 wt%, and proposed the method using an artificial neural network (ANN) technique for predicting erosion rate based on the experimentally measured data of S-GF/PPS. Angular silica sand particles with the size of 200 ± 50 µm were used as erodent, and the impact angles and the particle velocities were 15°, 30°, 60°, and 90° and 25, 50, and 66 m/s, respectively. Common erosion characteristics for PMCs were observed; the erosion rate of a composite was larger than that of corresponding neat resin, and increased with the increase in fiber content. An ANN technique was first applied to a solid particle erosion problem for polymers by Zhang et al.⁴⁴ Suresh Arjula et al. followed the ANN proposed by Zhang et al., and predicted the erosion rates within relative error of 25%, which was similar level of the wear measurements. They also applied the ANN to the erosion rate estimations of PEK reinforced by short GFs (S-GF/PEK) with varying fiber content (0–30 wt%).⁴⁵ Although Harsha et al.³⁸ carried out solid particle erosion experiments to obtain the erosion characteristics of PAEKs including PEK, Suresh Arjula et al. performed solid particle erosion experiments of S-GF/PEK and PEK neat resin under the similar experimental conditions to the previous erosion experiments³⁸ to build up an erosion rate database for S-GF/PEK and PEK neat resin. They applied the ANN to the erosion rate database to predict the erosion rate. The results showed that the predicted data were well agreed with the measured values.

In addition to Harsha et al.’s researches on PEI composites,^39,40 Rattan and Bijwe,^46,47 and Bijwe et al.⁴⁸ examined the solid particle erosion characteristics of carbon fabric reinforced PEI composites (CF/PEI). In Ref. 46 woven carbon fabric reinforced (55 vol%) PEI composites were fabricated using three types of weaves, that is, plain (P), twill (T), and satin-4 H (S) by impregnation technique in order to examine the effect of types of weaves for carbon fabric on the erosion resistance of composites. In addition to impregnation technique, film technique was used to fabricate CF/PEI composites. Silica sand particles with the size of 106–120 µm were used as erodent. The impact angle and the particle velocity were 30° and 40 m/s, respectively. The following results were obtained in Ref. 46:

The composites fabricated using impregnation technique showed higher erosion resistance than those using film technique.

In the case of impregnation technique, the plain weave reinforced composite showed the maximum erosion resistance followed by the satin and twill weave reinforced composites. In the case of film technique, the satin weave reinforced composite showed the maximum erosion resistance followed by the twill and plain weave reinforced composites.

In Ref. 47, solid particle erosion experiments of PEI composite reinforced with plain weave carbon fabric (CF/PEI) were performed at the impact angle of 30°, 45°, 60°, 75°, and 90° using silica sand particles with the size of 106–120 µm as erodent. The particle velocity was 26.88 m/s. The erosion data obtained showed that the erosion rate of composite CF/PEI was larger than that of the corresponding neat PEI resin. In other words, the reinforcement of composite adversely affected the erosion resistance of PEI. This fact is common to almost all PMCs. Bijwe et al.⁴⁸ performed solid particle erosion experiments of plain weave carbon fabric reinforced PEI composites (CF/PEI) to study the effect of amount of fabric on the erosion resistance. They used test specimens made by impregnation technique. The volume content of carbon fabric in the test specimens was changed from 40 to 85 vol%. The erosion experiment of neat PEI resin was also performed. Experimental conditions were the same as their previous research.⁴⁶ Their erosion experiments showed that the inclusion of reinforcement had adversely affected the erosion resistance of the composite, similar to most of PMCs.

Sari and Sinmazcelik⁴⁹ performed solid particle erosion experiments of unidirectional CF reinforced PEI composite (UD-CF/PEI) under a low particle velocity. Silica sand particles with the size of 150–200 µm were used as erodent. The particle velocities were 1.96 and 2.88 m/s. The impact angle was changed from 15° to 90°. According to their erosion experiments, the UD-CF/PEI composite showed a semi-ductile erosion behavior in the relation between erosion rate and impact angle under low impact velocities. Such an erosion behavior was similar to that under a high particle velocity.³⁹

Sinmazcelik and Taskiran⁵⁰ also performed high velocity particle erosion experiments of polyphenylenesulphide (PPS) reinforced by short GFs (S-GF/PPS), the weight content of which was 40 wt%. In the PPS resin, CaCO₃ mineral particles were included with the weight content of 25 wt%. So the composite material used in the erosion experiments had three constituents, that is, PPS resin, short GF and CaCO₃ particle. As shown later, composites whose matrix materials are modified with fillers such as particulates, whiskers and so on are called hybrid composites. Silica sand particles with the size of 150–200 µm were used as erodent. The particle velocities were 20, 40, and 60 m/s, and the impact angle was changed from 15° to 90°. The experimental result showed that the erosion rate was the maximum at the impact angle of 60°, and thus a semi-ductile erosion behavior. As described previously, the erosion data obtained by Suresh Arjula et al.⁴³ showed that the erosion rate of the S-GF/PPS without CaCO₃ particle reached the maximum erosion rate at 30°. Such difference between the Sinmazcelik and Taskiran’s result⁵⁰ and Suresh Arjula et al.’s result⁴³ may be due to the inclusion of CaCO₃ particles in the matrix resin. Inclusion of CaCO₃ particles in the PPS resin induces the increase in hardness of the composite material, which results in the shift of erosion rate peak to higher angle.

Sari⁵¹ and Sinmazcelik and Sari,⁵² respectively examined the influence of erodent particle types and the erodent size on the solid particle erosion of PPS with CaCO₃ particles reinforced by short GF (S-GF/PPS) using a low speed particle erosion rig given in Ref. 42 In Ref. 51 the erodent particle types used in the erosion experiments were 150–212 µm brown fused aluminum oxide (Al₂O₃), 150–200 µm silica sand and 150–250 µm glass bead. The glass bead had a spherical shape, whereas the Al₂O₃ and silica sand had sharp and angular edges. In Ref. 52 Al₂O₃ particles with three different sizes, 300–425, 150–212, and 45–75 µm, were used. In both erosion experiments, the particle velocity was 1.57 m/s, and the impact angle was changed from 15° to 90°. The test results of Ref. 51 showed that the erosion rate caused by the glass bead was much smaller than those caused by the Al₂O₃ and silica sand. According to the test results of Ref. 52, small-sized eroding particles (45–75 µm) showed the smallest erosion rate because eroding particles had smallest kinetic energy whereas the large-sized eroding particles (300–425 µm) showed the largest erosion rate.

Sinmazcelik et al.⁵³ performed solid particle erosion experiments of cross-ply CF reinforced PPS composite (CF/PPS) using angular silica sand particles with the size of 150–200 µm as erodent. The particle velocities were 20, 40, and 60 m/s, and the impact angle was changed from 15° to 90°. The composite had the maximum erosion rate at the impact angle of 45°, showing a semi-ductile erosion behavior. They obtained the residual flexural strengths of composite test specimens after solid particle erosion. The residual flexural strengths after erosion at large impact angles and higher particle velocities were smaller compared with acute angles and lower particle velocities. This fact means that not only the erosion rates but also fiber fracture and subsurface deformations such as matrix cracks and deformations due to the impingement of particles affect the residual flexural strength.

Yilmaz⁵⁴ examined the effect of annealing period at the temperature of 180℃ on the solid particle erosion behavior of PPS hybrid composites, that is, short GF (40 wt%), CaCO₃ particulate (25 wt%) reinforced PPS (S-GF/PPS). In the erosion experiments, silica sand with 150–250 µm was used as erodent. The particle velocity was 60 m/s, and the impact angle was changed from 30° to 90°. Erosion rates were evaluated at four different annealing periods, 30, 60, 90, and 120 min. Except for annealing period of 120 min, the increase in annealing period resulted in the increase in the relative degree of crystallinity ζ defined by the improvement in crystallinity compared with an as-received sample. The ranking of degree of crystallinity was as follows:

ζ (90 min) > ζ (60 min) > ζ (120 min) > ζ (30 min) > ζ (as-received)

On the other hand, the ranking of the erosion rate of the composite ER was as follows:

ER (90 min) < ER (60 min) < ER (120 min) < ER (30 min) < ER (as-received)

Annealed composite showed lower erosion rate, thus higher erosion resistance than as-received composite, and the increase in the total crystallinity caused improvement of the erosion resistance of the composite. According to the previous research on the effect of crystallization of thermoplastic PI resin on the erosion resistance,²² the resin with higher crystallinity showed lower erosion resistance. As for the effect of crystallinity on the erosion resistance, the present research provided the result opposite to this previous research. In composite materials, crystallization formed at the fiber-matrix interface is expected to play a more important role than the crystallization in the matrix, and the pullout and debonding of fibers are prevented by the transcrystalline layer formation or strong interfacial bonding strength.

Recently, the erosion characteristics of PPS-based composites have been studied for the application of the composites to aircraft structure.^55,56 Ahmed et al.⁵⁵ proposed a new anti/de-icing concept for ice-prone composite structure based on multifunctional structures. They proposed a leading edge structure material composed of GF reinforced PPS (GF/PPS) and a PPS layer including metal fibers for conductive heater element. They performed solid particle erosion experiments of such a leading edge structure material. They prepared sample laminates for erosion experiments composed of two top layers of GF/PPS on the impact side, two top layers of CF/PPS on the side opposite to the impact side. On the top of the GF/PPS ply a metal mesh was embedded as a simulation of a conductive heater element layer. The experimental results using Australian (garnet) sand at the particle velocity of 70 m/s showed that the erosion rate of the GF/PPS laminate with a metal wire layer was almost the same magnitude of erosion rate as an aircraft-grade aluminum sheet, although the GF/PPS laminate without a metal wire layer showed much greater erosion rate than an aircraft-grade aluminum sheet. It was therefore concluded that the proposed leading edge structure material had both anti-erosion and anti/de-icing characteristics. Avcu et al.⁵⁶ performed solid particle erosion experiments of chopped glass mat reinforced PPS resin to examine the effects of erodent particle, the particle impact angle and velocity on the erosion behavior of GF/PPS composite. The volume content of glass mat was 25 vol%. Sharp-edge alumina (Al₂O₃) particles with the size of 212–300, 150–212 and 90–125 µm were used as erodent in the erosion experiments. The impact angle and the particle velocity were respectively changed from 30° to 90°, and from 56 to 100 m/s. There was no new finding except for the effect of particle size on erosion rate. According to the experimental results, the erosion rate of the composite increased with increase in particle size under a higher particle velocity, and vice versa under a lower particle velocity.

Tsuda et al.⁵⁷ obtained solid particle erosion characteristics of various types of GF reinforced UP resin (GF/UP). Three types of GFs were used for GF/UP, a chopped glass-mat (Mat-GF), a plain-woven glass cloth (Cloth-GF) and unidirectional glass bundles with no twisted strands (UD-GF). The composites containing these types of GFs are designated as Mat-GF/UP, Cloth-GF/UP and UD-GF/UP, respectively. The weight contents of GFs were 18.2–41.1 wt% for Mat-GF/UP, 22.8–52.9 wt% for Cloth-GF/UP and 47.9 wt% for UD-GF/UP, respectively. The crashed-glass-powder with the average diameter of 350 µm was used as erodent. The impact angle α was changed from 20° to 90°, and the particle velocity was 24.5 m/s. The angle between the solid particle stream and the fiber orientation, which is designated as β, for the UD-GF/UP was chosen at 0°(parallel impact), 30°, 45°, 60°, and 90°(perpendicular impact). They found the following erosion characteristics:

The peak of the erosion rate in the composite shifted to a higher impact angle with the increase in GF content. Thus the erosion behavior of the composite changed from a ductile manner to a brittle one with the increase in GF content.

For UD-GF/UP, the erosion rate increased with the increase in the angle β, and the effect of the angle β on the erosion rate was larger at the lower impact angle α.

The above erosion characteristics were already observed for other composite materials in the preceding researches. Tsuda et al. also proposed an estimation method for erosion rate of GF reinforced composite, E_c, by knowing only the values of the erosion rate of matrix resin, E_m, and GF volumetric content, V_f, in advance. They applied their proposed method to predicting the erosion rate of the experimental results obtained in the preceding researches.^15,30 The predicted results were in good agreement with experimental results.

As for the effect of the angle β between fiber orientation and particle flow on erosion rate in a unidirectional fiber reinforced composite, Kim and Kim⁵⁸ reconfirmed the results of preceding researches by performing solid particle erosion experiments of unidirectional CF reinforced EP composites (UD-CF/EP). They performed the erosion experiments for β = 0° (parallel impact: (0)), 45°((45)), and 90°(perpendicular impact: (90)). The ranking of erosion rate was as follows: (90) > (45) > (0). They also performed the erosion experiments of multidirectional laminated CF/EP composites, [0/90], [45/ − 45], [90/30/−30], and [0/60/−60]. It was found that the erosion of multidirectional laminated composites was not sensitive to the fiber direction of laminates with respect to particle flow direction. Biswas et al.⁵⁹ performed solid particle erosion experiments of GF reinforced EP (GF/EP) to examine the effect of fiber content and fiber orientation on the erosion characteristics of GF/EP composites. The composites were fabricated with three different fiber weight contents, 20, 30, and 40 wt%, and the GFs in the composites were arranged in four different fiber orientations, 15°, 30°, 45°, and 60°, with respect to the base fiber. For example, in a composite with 15° fiber orientation, woven glass cloths were piled up in such a way that the fiber orientations of two sequential cloths are inclined by 15°. According to the authors’ conclusions, the composite with 20 wt% fiber content showed the largest erosion rate for 30° fiber orientation, and the largest erosion rate was found in the case of 15° fiber orientation for other two fiber contents, 30 and 40 wt%. In their paper, the authors provided the relation between the erosion rate and the impact angle, indicating that the composites with all fiber contents, 20, 30 and 40 wt%, showed the largest erosion rate for 15° fiber orientation. This fact contradicts with the conclusion mentioned in the paper.

Mahapatra et al.⁶⁰ and Patnaik et al.^61,62 performed solid particle erosion experiments of cross-plied E-GFs reinforced UP resin (GF/UP). Both Refs. 61 and 62 have the same content, so that Ref. 62 is neglected in this article. In both erosion experiments,^60,61 different weight contents of GF were used as test specimens, and different sizes of dry silica sand were used as erodent. The weight contents of GF in test specimens used were 40, 50, and 60 wt% in the Mahapatra et al.’s experiment,⁶⁰ and 30, 40, and 50 wt% in the Patnaik et al.’s experiment.⁶¹ The erodent size used was 300, 500, and 800 µm in the Mahapatra et al.’s experiment,⁶⁰ and 450, 600, and 800 µm in the Patnaik et al.’s experiment.⁶¹ The particle velocities for both erosion experiments were 32, 45, and 58 m/s. The impact angle was varied from 15° to 90°. They used the design of experiments approach named Taguchi method to study the effect of various operational and material parameters on the solid particle erosion characteristics of these composites. The results obtained from these researches indicated a semi-ductile erosion behavior with the maximum erosion rate at the impact angle of 60°, and the increase in erosion rate with the increase in GF content. It was shown that Taguchi method provided a simple, systematic and efficient methodology for finding out the control factors affecting erosion rate and their interactions. They also successfully applied the generic algorithm (GA) and the ANN to obtaining the optimum value of factor settings for minimizing erosion rate. In Ref. 61 the erosion efficiency η was applied to qualitative characterization of erosion behavior of the GF/UP composites used in the experiments. The following erosion efficiency η considering the erodent impact angle α was used instead of equation (3)

η = \frac{2 E_{r} H_{V}}{ρ V_{p}^{2} \sin 2 α}

(4)

The erosion efficiency (η) values of the GF/UP composites used in the erosion experiments were 10–60% for a low particle velocity and less than 10% for a high particle velocity, so that the GF/UP composites exhibited a semi-ductile erosion response for a low particle velocity and a ductile erosion response for a high particle velocity. The erosion rate data given in Refs. 60 and 61 were greatly different from each other under almost the same experimental conditions. For example, in Ref. 60 the erosion rate is 6.164 mg/kg under the experimental conditions; particle velocity = 32 m/s, fiber content = 40 wt%, impact angle = 90°, erodent size = 800 µm, and stand-off distance = 160 mm. On the other hand, in Ref. 61 the erosion rate is 272.79 mg/kg under the experimental conditions; particle velocity = 32 m/s, fiber content = 40 wt%, impact angle = 90°, erodent size = 800 µm, and stand-off distance = 180 mm. Such difference in erosion rate is unbelievable.

Qian et al.⁶³ conducted solid particle erosion experiments of UP resin reinforced by unidirectional (UD) various fibers. The reinforcement fibers used were inorganic fiber, CF, and two kinds of organic high-polymer fibers, ultrahigh strength polyethylene fiber (Dyneema: DF) and poly (p-phenylene-2, 6-benzobisoxazole) fiber (Zylon: ZF). The volume contents of the composites were 49 vol% for UD-CF/UP, 46 vol% for UD-DF/UP, and 40 vol% for UD-ZF/UP, respectively. The erosion experiment of neat UP resin was also performed. Angular alumina abrasive with the average size of 11.5 µm was used as erodent. The particle velocity was 128 m/s and the impact angle α was changed from 15° to 90°. We can make a comparison of erosion rate among three kinds of UP composites, because difference in the volumetric fiber contents of the composite used in erosion experiments are relatively small. According to volumetric erosion rate data obtained from the experiments, the ranking of erosion resistance of the composites including the neat UP resin was as follows; UD-DF/UP ≈ UD-ZF/UP > neat UP ≫ UD-CF/UP. It was therefore concluded that the inclusion of reinforcement fiber enhanced erosion resistance of resin for organic or ductile fibers such as DF and ZF, while vice versa for inorganic or brittle fibers such as CF. The peak of erosion rate for neat UP resin took place at the impact angle α of 45°, while that for the UD-CF/UP composite took place at the impact angle α of 60°. Such a peak shift was caused by the inclusion of brittle fiber in a ductile resin. On the other hand, UD-DF/UP and UD-ZF/UP composites had the peak of erosion rate at the impact angle α of 45°, the same impact angle at which neat UP resin had the peak erosion rate. UD-DF/UP and UD-ZF/UP composites were composed of ductile fiber and ductile matrix, so that the composites had the same ductile erosion behavior as the neat resin. These erosion characteristics of the ductile organic-fiber reinforced composites are new findings.

Solid particle erosion experiments of composites including ductile organic fibers were performed by Kumar et al.,⁶⁴ Tejyan et al.,⁶⁵ and Patnaik and Tejyan.⁶⁶ Kumar et al.⁶⁴ performed erosion experiments of short aramid fiber reinforced vinyl ester resin. Four kinds of composite specimens VAF20, VAF30, VAF40, and VAF50 were prepared for erosion experiments. Each composite test specimen had the fiber weight content of 20 wt% for VAF20, 30 wt% for VAF30, 40 wt% for VAF40, and 50 wt% for VAF50. Erosion experiment was also performed for neat vinyl ester resin (VAF0). Dry silica sand was used as erodent. The erosion rate of each composite was evaluated by changing the experimental conditions, that is, impact angle (30–90°), particle velocity (43–76 m/s), erodent size (250–600 µm), and stand-off distance (55–85 mm), in order to clarify the effect of various operational and material parameters on solid particle erosion characteristics of these composites using Taguchi method. According to the erosion data, the ranking of erosion resistance was as follows; VAF30 > VAF20 > VAF40 > VAF50 > VAF0 (neat resin). This fact indicates that the inclusion of reinforcement fiber helps enhancement of erosion resistance compared with the neat resin, although nonlinear dependence on the fiber weight content was observed in the erosion resistance of composite. The decrease in erosion resistance or the increase in erosion rate of the composite at higher fiber contents, i.e. VAF40 and VAF50, may be attributed to the reason that the effect of fiber removal caused by fiber-matrix interfacial damage becomes large compared with that of resin removal. Kumar et al.⁶⁷ performed solid particle erosion experiments of short CF reinforced vinyl ester resin under the same test conditions as the erosion experiments for short aramid fiber reinforced vinyl ester resin,⁶⁴ and both composites showed quite similar erosion characteristics, although the former composite included brittle fiber while the latter one included ductile fiber. That is, the result of Ref. 67 showed that the short CF reinforced vinyl ester resin had lower erosion rate than the vinyl ester neat resin. In the preceding researches, the erosion rates of composites including brittle fiber such as GF and CF were higher than those of the corresponding neat resin. The result of Ref. 67 was contrary to that of preceding researches. Lower erosion rate of short CF reinforced vinyl ester resin in comparison with that of neat vinyl ester resin may be attributed to good adhesion between CFs and vinyl ester resin. Tejyan et al.⁶⁵ evaluated erosion rates of PP-based needle-punched nonwoven reinforced epoxy composites (NNC). The composites with three different nonwoven material weight contents, that is, 20, 30, and 40 wt% were fabricated, and they are respectively referred to as NNC20, NNC30, and NNC40. The composites included ductile reinforcement nonwoven material. Erosion experiments were performed using silica sand particles with the sizes of 250, 350, and 450 µm. The impact angles and the particle velocities were 30°, 60°, and 90°, and 35, 45, and 55 m/s, respectively. The composites showed a semi-ductile erosion behavior with the maximum erosion rate at the impact angle of 45°. The ranking of erosion resistance was as follows: NNC40 > NNC30 > NNC20. Patnaik and Tejyan⁶⁶ performed solid particle erosion experiments of viscose fiber-based needle punched nonwoven fabric mat reinforced EP composites. Three types of viscose fiber-based needle-punched nonwoven fabric mats VS200, VS400, and VS600 gsm (gsm: g/m²) were used for reinforcement material. The composite specimens designated as follows were used in solid particle erosion experiments: VS200-20, VS200-30, VS200-40: VS400-20, VS400-30, VS400-40: VS600-20, VS600-30, VS600-40. For example, VS200-20 indicates an EP composite reinforced by viscose fiber-based needle-punched nonwoven fabric VS200 gsm, whose weight content of fabric mat is 20 wt%. Other designations of composites are based on this rule. According to the erosion experiments performed at the particle velocity of 45 m/s using irregular shape silica sand with the particle size of 450 µm, the rankings of erosion resistance were as follows:

VS200-20 > VS200-30 > VS200-40: for VS200 gsm composites,

VS400-40 > VS400-30 > VS400-20: for VS400 gsm composites, and

VS600-40 > VS600-30 > VS600-20: for VS600 gsm composites.

The rankings of VS400 and VS600 gsm indicate that inclusion of ductile reinforcement material enhances the erosion resistance of composites, although such a physical phenomenon can be confirmed if the erosion rate of the neat resin is available and larger than those of the composites. As shown here,^63–66 the inclusion of ductile fibers/materials as reinforcement of a composite tends to improve the erosion resistance of neat resin.

Natural filler reinforced PMCs

In solid particle erosion studies on PMCs, much attention has been recently paid to natural filler reinforced resins. Chittaranjan Deo and Acharya,⁶⁸ Mishra and Acharya,⁶⁹ and Gupta et al.⁷⁰ performed solid particle erosion experiments of natural fiber polymer composites. Chittaranjan Deo and Acharya⁶⁸ used chopped lantana camara fiber reinforced EP composites as erosion test specimens. Four composites with different lantana camara fiber weight contents, 10, 20, 30, and 40 wt%, were fabricated by usual hand lay-up technique. The erosion rates of these composites were evaluated using angular shaped silica sand particles with the size of 200 µm at the particle velocities of 48, 70, 82, and 109 m/s, and the impact angles ranging from 30° from 90°. The erosion data showed a semi-ductile erosion behavior with the maximum erosion rate at the impact angle of 45°, and the increase in erosion rate or the decrease in erosion resistance with the increase in fiber content. Mishra and Acharya⁶⁹ performed solid particle erosion experiments of bagasse fiber reinforced EP composites with the fiber volume contents of 0 (neat EP), 10, 15, and 20 vol%. Silica sand with the size of 200 µm was used as erodent. The particle velocities were 48, 70, 82, and 109 m/s, and the impact angle was changed from 30° to 90°. Both the composites and neat resin showed a brittle erosion behavior with the maximum erosion rate at the impact angle of 90°. Although no relationship was found between the erosion rate or erosion resistance and fiber volume content, the composites had better erosion resistance than the neat EP. Gupta et al.⁷⁰ evaluated the erosion rates of roving bidirectional bamboo mat reinforced epoxy composites using silica sand particles with the size of 125–300 µm. Test specimens used were the composites with different fiber weight contents, 10, 20, 30, and 40 wt%, in addition to neat EP. The particle velocities were 30, 52, 60, and 88 m/s, and the impact angle was changed from 15° to 90°. Both the composites and neat resin showed a semi-brittle/semi-ductile erosion behavior with the peak erosion rate at the impact angle of 60–75°. The erosion rates of the composites are smaller than that of the neat resin, so that the inclusion of bamboo fibers improves the erosion resistance of the neat resin. Natural fibers such as lantana camara fiber, bagasse fiber, and bamboo fiber are organic materials. According to preceding researches,^63–66 organic reinforcement fibers improved the erosion resistance of neat resins. Both the bagasse fiber and bamboo fiber had the same effect on erosion resistance as the organic fibers used in the preceding researches,^63–66 whereas lantana camara fiber reduced the erosion resistance of neat resin. Satapathy et al.⁷¹ performed solid particle erosion experiments of fish (Labeo-rohita) scale filled epoxy matrix composites, whose filler weight contents were 5, 10, and 15 wt%. The main element of fish scale is hydroapatite, so that fish scale filler is a kind of inorganic fillers. Silica sand particles with the sizes of 300, 500, and 800 µm were used as erodent and the particle velocities were 32, 44, and 58 m/s. All the three composites showed almost the same erosion rate irrespective of filler contents and the maximum erosion rate at the impact angle α of 90°, which indicated a brittle erosion behavior. They concluded that the composites exhibited improved erosion wear performance compared with neat EP, but they did not provide any erosion data validating this conclusion. They calculated the erosion efficiency η defined by equation (3). The calculated η values ranged from 4% to 60%, which indicated a ductile or semi-ductile erosion behavior. On the other hand, an erosion rate versus impact angle curve for the composite showed a brittle behavior with the maximum erosion rate at $α = 90 \circ$ . In this study, the erosion efficiency was not able to predict the erosion behavior.

Nanofiller reinforced PMCs

Recently, several papers have been published on the erosion characteristics of nanofiller reinforced PMCs. Qian et al.⁷² fabricated an UP composite including vapor-grown CF (VGCF) as nanofiller. In addition to neat UP resin, the nanocomposite specimens were prepared with filler weight contents of 1.25, 2.5, and 5 wt% of VGCF. Angular shaped alumina solid particles with the size of 11.5 µm were used as erodent, and the particle velocities were 97.8, 128, and 152 m/s. The impact angle was changed from 15° to 90°. The results of erosion rate showed that the inclusion of nanofiller improved the erosion resistance of the matrix resin. Zhou et al.⁷³ evaluated solid particle erosion characteristics of the polyacrylate-based nanocomposite coatings. Nanosilica particle was chosen as filler of composites. For the composites with colloidal nanosilica particles, the weight content of filler was changed from 10 to 40 wt%. On the other hand, for the composites with pyrogenic nanosilica particles, it was changed from 5 to 20 wt%. Both sharp-edged and round steel balls were used as erodent. The particle velocity was 17 m/s. Both the neat resin polymer coating and nanocomposite coatings exhibited a brittle erosion behavior. For sharp-edged erodent, the erosion weight loss of the neat polymer coating was the maximum and those of nanocomposite coatings drastically decreased with the increase in filler content. On the other hand, the erosion loss was insensitive to the filler content for round erodent. Zhang et al.⁷⁴ performed solid particle erosion experiments of CF woven fabric (CF: 52 wt%) reinforced epoxy composite (CF/EP) and CF/EP composite coated by thin nanopaper made of carbon nanofibers (CNF) (CNF: 12 wt% and CF: 52 wt%). Silica sand particles with the size of 150 µm were used as erodent. The CNF-based nanopaper achieved much better erosion resistance than the CF/EP composite, and also better erosion resistance than the neat EP. Zhang et al.⁷⁵ also compared the erosion weight loss of GF reinforced epoxy composite (GF/EP; GF: 65 wt%) with GF/EP coated by CNF-based nanopaper (GF: 65 wt% and CNF: 2.5 wt%), and obtained the result that CNF-based nanopaper enhanced the erosion resistance compared with GF/EP without CNF-based nonopaper coating. These experimental results indicate that the CNF-based nanopaper is a good protective coating material for wind turbine blade made of composite materials. Chen et al.⁷⁶ performed solid particle erosion experiments of CNT reinforced epoxy (EP) composites (CNT/EP). Three kinds of CNT/EP composites were used in the erosion experiments. They were isotropic CNT/EP composites containing randomly dispersed CNTs (0.32 wt%), vertically aligned CNT/EP composites with nanotube tips exposed to the surface and horizontally aligned CNT/EP composites with side of CNTs exposed to the surface and impinging direction parallel to the CNT axes. Angular olivine sand with the size of 238 µm was used as erodent. The impact angles were 20°, 30°, 45°, and 90°, and the particle velocity was 40 m/s. The ranking of erosion rate of composite was as follows: isotropic CNT/EP > horizontally aligned CNT/EP > Neat EP > vertically aligned CNT/EP.

Hybrid PMCs

Recently, a lot of papers have been published on solid particle erosion studies of hybrid PMCs including two or more reinforcements. One of the objectives for developing hybrid composites is to enhance erosion resistance of conventional PMCs. For this purpose, a low weight percentage of fillers are usually added to conventional PMCs as modification of a matrix resin in addition to a high weight percentage of fiber or filler as reinforcement. Inorganic fillers, organic fillers, and metal fillers are used to modify matrix resins. Among them, inorganic fillers are most commonly used. Flyash, SiC particles, Al₂O₃ particles, CaCO₃ particles, boric acid particles, and so on are typical fillers used for hybrid PMCs. Solid particle erosion characteristics of hybrid PMCs containing flyash particles were studied by several researchers.^77–79 The first solid particle erosion study of a hybrid PMC was performed by Srivastava and Pawar.⁷⁷ They used GF reinforced flyash filled EP composites as test specimens. The contents of GF and flyash were not given in the paper. Silica sand particles with the size of 150–250 µm were used as erodent, and the particle velocities were 24, 35, and 52 m/s. The impact angle was changed from 30° to 90°. The GF reinforced EP (GF/EP) with flyash filler had lower erosion rate than the GF/EP without flyash filler, and the addition of flyash filler improved the erosion resistance of the GF/EP composite, although it reduced the hardness, the tensile strength as well as the density. Patnaik et al.⁷⁸ and Bishoyee et al.⁷⁹ published the papers on the solid particle erosion study of GF reinforced UP resin (GF/UP) with flyash filler as test specimens, but these two papers are based on the same experimental data. The weight content of flyash filler in the composites was change as 0, 10, and 20 wt%, the weight content of GF being kept at 50 wt%. Dry silica sand of different particle sizes, 300, 500, and 800 µm, were used as erodent. The particle velocities were 32, 45, and 58 m/s. The impact angle was changed from 15° to 90°. All the test specimens showed a semi-ductile erosion behavior with the maximum erosion rate at the impact angle of 60°. It was also found that the presence of flyash in the composite reduced the erosion rate. The composite with filler weight content of 20 wt% showed the better erosion resistance than that of 10 wt%. They applied Taguchi experimental design method and proposed a predictive equation for erosion wear of these fiber reinforced composites. They also employed the ANN technique to predict the erosion rate of the composites in a parameter space larger than the experimental domain. The calculated results of the erosion efficiency η defined by equation (4) for the flyash-filled GF/UP composites exhibited a semi-ductile erosion response (η = 10–60%) for a low particle velocity and a ductile erosion response (η < 10%) for a relatively high particle velocity. Before presenting the paper on the solid particle erosion study on the GF/flyash-filled UP composites, Patnaik et al.⁸⁰ performed the solid particle erosion studies on GF/Al₂O₃-filled UP composites and GF/SiC-filled UP composites⁸¹ under the same experimental conditions as GF/flyash-filled UP composites,^78,79 and obtained the same results as GF/flyash-filled UP composites. Afterwards Patnaik et al.⁸² compared the erosion resistance among three kinds of hybrid PMCs, that is, GF/flyash-filled UP composites, GF/Al₂O₃-filled UP composites, and GF/SiC-filled UP composites using the experimental data shown in Refs. 78, 80 and 81. As a result, the Al₂O₃-filled composites showed the best erosion resistance, the SiC-filled composites showed the intermediate erosion resistance, and the flyash-filled composites showed the worst one. Mahapatra et al.⁸³ published the paper by replacing the erosion data of GF/flyash-filled UP composites with those of GF/cement by-pass dust(CBPD)-filled UP composites and provided completely the same conclusions as their previous research⁸² except for replacing flyash with CBPD.

Hereafter, erosion results of various kinds of hybrid PMCs including inorganic fillers were shown. Yilmaz et al.⁸⁴ studied solid particle erosion characteristics of GF reinforced UP resin filled with CaCO₃ particles (GF/CaCO₃-filled UP). Different CaCO₃ particle sizes of 1, 2, 3, 5, and 10 µm were used as fillers. As for the effect of CaCO₃ filler on the erosion characteristics of GF/CaCO₃-filled UP, they provided no useful result except that the GF/CaCO₃-filled UP showed a brittle erosion behavior with the maximum erosion rate at the impact angle of 90°, and that the erosion rate became higher with the increase in the CaCO₃ particle size. It is not clear whether or not the inclusion of CaCO₃ particles improves the erosion resistance of GF/UP composites. Karsli et al.⁸⁵ performed solid particle erosion experiments of CaCO₃ particle-filled short GF reinforced ABS/PA6 resin composites (S-GF/CaCO₃-filled (ABS/A6)). Hereafter the notation a/b/c are used to represent the weight contents of the respective constituents of S-GF/CaCO₃-filled (ABS/A6) composites; a: ABS/A6 matrix wt%, b: S-GF wt%, and c: CaCO₃ particle wt%. Silica sand particles with the size of 150–250 µm were used as erodent, and the particle velocity was 60 m/s. The impact angle was changed from 15° to 90°. Various kinds of test specimens were used to clarify the effects of CaCO₃ particle, S-GF, and both CaCO₃ particle and S-GF on the erosion characteristics of the composites. The rankings of erosion rate for various composites were as follows:

Effect of CaCO₃ particle: 100/0/0 ≈ 95/0/5 < 90/0/10 < 80/0/20 < 70/0/30. This result indicates that the increase in the amount of CaCO₃ particle leads to the increase in erosion rate of the composites.

Effect of S-GF: 90/10/0 < 80/20/0 < 70/30/0 < 60/40/0. This result indicates that the increase in the amount of S-GF leads to the increase in erosion rate of the composites.

Effect of both CaCO₃ particle and S-GF:

80/10/10 < 70/15/15 < 60/20/20. This result indicates that the increase in total amount of CaCO₃ and S-GF increases the erosion rate of the composites.

80/20/0 < 80/10/10 < 80/0/20, 70/30/0 < 70/15/15 < 70/0/30. These results indicate that the addition of CaCO₃ particle in the conventional PMC (S-GF/(ABS/A6)) leads to the increase in erosion rate, thus does not improve the erosion resistance of the composites.

In conclusion, the hybrid PMCs, S-GF/CaCO₃-filled (ABS/A6), are not suitable under erosive environment condition.

Biswas and Satapathy⁸⁶ studied solid particle erosion characteristics of GF reinforced EP composites (GF/EP) filled with copper slag particles. Copper slag is a by-product obtained during smelting and refining of copper, and composed of various oxides such as Fe₂O₃, SiO₂, CaO, Al₂O₃, and so on. Copper slag particles with the size of about 70 µm were used as filler of the composites. Three kinds of the composites were used in solid particle erosion experiments. They were GF/EP composite without copper slag particles, GF/EP composite with 10 wt% copper slag and GF/EP composite with 20 wt% copper slag. Among the three kinds of composites, the GF/EP composite with 20 wt% copper slag showed the best erosion resistance, the GF/EP composite with 10 wt% copper slag showed intermediate erosion resistance, and the GF/EP composite without copper slag showed the worst erosion resistance. Therefore the copper slag filler improved the erosion resistance of the GF/EP composites.

Patnaik et al.⁸⁷ evaluated solid particle erosion characteristics of SiC particulate filled short GF reinforced UP resin composites (S-GF/SiC filled UP). The weight contents of short GF and SiC particulate were changed from 10 to 50 wt% and from 0 to 20 wt%, respectively. Silica particles with the size of 350 µm were used as erodent. The composite S-GF/UP without SiC particulate had the peak of erosion rate at the impact angle of 60°, whereas the composites S-GF/UP with SiC particulate (S-GF/SiC filled UP) had the peak of erosion rate at the impact angle of 75°. Such a shift of erosion rate peak to a higher impact angle was attributed to the addition of brittle SiC particulate. In addition, SiC particulate improved the erosion resistance of the S-GF/UP composite. They also performed a finite element analysis using explicit dynamic code ANSYS/LS-DYNA to predict the erosion rate of the S-GF/SiC filled PE composite.

Mahapatra⁸⁸ performed solid particle erosion experiments of two kinds of cross-plied woven E-GF reinforced UP hybrid composites. CBPD particulate with the size of 27 µm and alumina particulate with the size of 44 µm were used as fillers of the hybrid composites. Alumina is harder than CBPD. They are named as GF/CBPD filled UP and GF/Alumina filled UP. Erosion experiments were performed for these hybrid composites and GF/UP composite without particulate filler. Particulate filler weight contents were 10 and 20 wt%. All the test specimens included 50 wt% E-GF. Spherical shaped dry silica sand of different particle sizes, 300, 500, and 800 µm, was used as erodent. Experimental results showed that the composites with particulate filler had better erosion resistance than the composites without particulate filler. He supposed that the presence of hard particle fillers in the matrix helped in absorbing a good fraction of this kinetic energy and therefore energy available for the plastic deformation of thermoplastic polyester became less, which led to delay of the initiation of fiber exposure compared with the composite without any filler. The hybrid composites with higher filler content (20 wt%) showed better erosion resistance than those with lower filler content (10 wt%). All the composites tested showed a semi-ductile erosion behavior with the peak erosion rate at the impact angle of 60°. The hybrid composites with alumina filler showed better erosion resistance than those with CBPD filler. He applied Taguchi method and the genetic algorithm to find out the optimal parameter settings for the minimum erosion rate. Biswas and Satapathy⁸⁹ performed solid particle erosion experiments similar to Mahapatra’s study.⁸⁸ They used E-GF reinforced EP filled with alumina particles. The alumina particle size was 70–90 µm. Erosion experiments were performed under almost the same conditions as Mahapatra.⁸⁸ The results obtained are similar to Mahapatra.⁸⁸ That is, the hybrid composites including alumina filler had better erosion resistance than the composites without the filler, and the erosion resistance increased with the filler content. All the composites tested showed a semi-ductile erosion behavior with the peak erosion rate at the impact angle of 60°.

Jha et al.⁹⁰ evaluated solid particle erosion characteristics of jute fiber reinforced EP composite (Jute/EP) filled with SiC particles. They compared the erosion resistance of Jute/SiC filled EP hybrid composites with that of neat Jute/EP composite. They used two kinds of Jute/SiC filled EP hybrid composites with different filler weight contents, 10 and 20 wt%. The Jute/SiC filled EP hybrid composites and neat Jute/EP composite had the same jute weight content, 40 wt%. According to their experimental results, the inclusion of SiC particles in Jute/EP composite improved the erosion resistance, and the hybrid composite with 20 wt% SiC filler had better erosion resistance than that with 10 wt% SiC filler. It was also found that the peak of the erosion rate for the hybrid composites shifted to a higher angle, compared with that for the neat composite, due to the inclusion of brittle SiC.

Bagci and Imrek^91,92 examined solid particle erosion characteristics of GF reinforced boric acid filled EP composite (GF/Boric acid filled EP). The weight content of GF was 50 wt%. The test specimens used were neat GF/EP composite without boric acid filler, GF/EP composite with 15 wt% boric acid filler⁹¹ and GF/EP composite with 30 wt% boric acid filler.⁹² Angular alumina particles were used as erodent. According to the erosion test data, addition of 15 wt% boric acid filler sharply reduced the erosion resistance of the neat GF/EP composite, while addition of 30 wt% boric acid filler recovered the erosion resistance at the level of the neat EP/GF composite. The authors explained these phenomena as follows; In the case of 15 wt% boric acid filler, the boric acid did not form a strong bond with EP, but in the case of 30 wt% boric acid filler, it forms a strong bond with EP. As for bond strength between boric acid and EP, there was no evidence that validated the authors’ explanation.

Kumar et al.⁹³ used vinyl ester resin reinforced by both short GF and short CF as test specimens for solid particle erosion experiments. The weight contents of GF and CF were same, and total fiber content was changed from 0 to 50 wt%. They obtained the steady state responses of these composites with respect to impact angle, particle velocity, and erodent size, but could not give how the inclusion of both GF and CF influenced the erosion behavior of the hybrid composites.

Recently, various inorganic fillers have been used in hybrid composites. The fillers in the hybrid composites used in the solid particle erosion experiments were granite particles,⁹⁴ organo-modified montmorillonite nanoparticles,⁹⁵ WC particles,⁹⁶ blast furnace slag (BFS) particles,⁹⁷ and AlN particles.⁹⁸ These fillers improved the erosion resistance of PMCs.

Metal fillers and organic fillers can be used in hybrid PMCs. Aktas⁹⁹ performed solid particle erosion experiments of bidirectional woven GF reinforced epoxy (GF/EP) composite strengthened by metal powder, the main composition of which was Ni. Four different specimens were tested. One was GF/EP composite without metal powder, and the others were GF/EP composites with metal powder, the weight contents of which were 5, 10, and 20 wt%, respectively. The results showed that the erosion quantities measured by weight loss of four different composites were nearly the same. Aktas therefore concluded that the percentage content of metal powder did not influence erosive wear considerably. This conclusion is incorrect. Considering that the densities of the GF/EP composites with metal powder are larger than that of the GF/EP composite without metal powder and increase with the increase in percentage content of metal powder, we can conclude that the erosion quantities measured by volumetric loss of the GF/EP composites with metal powder are smaller than the GF/EP composite without metal powder and reduce with the increase in percentage content of metal powder, if the erosion quantities measured by weight loss of four different composites are nearly the same. The present author is of opinion that the quantity of erosion should be measured not by weight loss or weight erosion rate but by volumetric loss or volumetric erosion rate. Mohan et al.^100,101 performed solid particle erosion experiments of aramid fabric reinforced EP (AF/EP) composite filled with ultrahigh molecular weight polyethylene particles (AF/UHMWPE filled EP). In Ref. 101 they used AF/EP composite without UHMWPE filler and two kinds of UHMWPE filled AF/EP composites: U1-A-E and U2-A-E, as test specimens. The weight contents of EP, aramid fabric and UHMWPE filler were 60, 38, and 2 wt% for U1-A-E, and 60, 36, and 4 wt% for U2-A-E. The ranking of the erosion resistance was as follows; U2-A-E > U1-A-E > neat AF/EP. So the inclusion of UHMWPE particle improved the erosion resistance of neat AF/EP composite. In addition, the inclusion of UHMWPE particle improved the mechanical properties such as tensile strength, while the inclusion of inorganic fillers in hybrid composites often deteriorate mechanical properties.

Metal matrix composites

Solid particle erosion characteristics of MMCs are reviewed here by classifying MMCs into fiber/particle reinforced metals (FRMs/PRMs) and cermets.

FRMs/PRMs

Sreeran Srinivasan¹⁰² performed solid particle erosion experiments of Al-4pct Cu metal reinforced by δ-alumina fibers, the dimensions of which were approximately 2 to 3 µm in diameter with length up to 20 to 30 µm. The fiber volume contents were 0, 5, 20, and 30 vol%, in which 0 vol% represents a neat Al-Cu matrix material. The erodent particles used were E-17 Norton Alundum. According to the experimental data, the erosion rates of the composites were larger than that of the neat matrix material and the erosion rate increased with the increase in fiber volume content. Hence the erosion resistance decreased with the inclusion of fibers.

Morita et al.¹⁰³ studied solid particle erosion behavior of alumina particle reinforced cast iron composite. The diameter and volume content of alumina particles were 3 mm and 60 vol%, respectively. Three kinds of erodent particles were used; silica sand with the size of 0.65 µm, sinter steel grid with the size of 0.2 µm and steel grit with the size of 0.3 µm. Experimental results showed that the composite material had lower erosion rate than the neat cast iron matrix material. According to the study on the erosion behavior of multiphase systems performed by Hovis et al.,¹⁰⁴ the IROM given by equation (2) exactly holds if the filler size of a composite is larger than an impact event size. When the erosion rate of filler ER_f is larger than that of a matrix material ER_m, the IROM gives the smaller erosion rate of a composite material ER_c than that of a matrix material ER_m. In Morita et al.’s erosion experiments, the size of alumina particle was much larger than the erodent size, so that the erosion rate of the composite could be evaluated by the IROM. Moreover, alumina particle had smaller erosion rate than cast iron; that is, ER_f < ER_m. That is why the composite materials had lower erosion rate than the neat cast iron matrix material.

Morrison et al.¹⁰⁵ measured the erosion rates of an aligned alumina-stainless steel composite. Alumina rods with 500 µm in diameter were fully aligned and formed a regular two-dimensional array. The volume contents of the alumina rods and stainless steel were approximately equal. Alumina erodent particles with the mean diameters between 37 and 390 µm were used in the erosion experiments. So the size and spacing of the alumina rods were large compared with erosion impact events. It was therefore possible to describe the erosion rate of the composite in terms of the erosion rate of the individual components.¹⁰⁴ Morrison et al. proposed a model, in which the erosion rate was constrained in such a way that the erosion rate of the stainless steel determines a stationary erosion surface profile, and showed the reasonable agreement between the predictions by this model and the erosion test data.

Wu et al.¹⁰⁶ performed solid particle erosion experiments of 2014 Al reinforced by SiC or Al₂O₃ particles (SiC/Al or Al₂O₃/Al). There was no information on particle content and the particle diameter except that Al₂O₃ particles were about the same size as SiC particle. As erodent, angular Al₂O₃ particles with the average sizes of 23, 42, 63, 143, and 390 µm were used for most erosion experiments, and angular SiC particles were used for some experiments. The particle velocities were 50, 75, and 100 m/s. The impact angle ranged from 10° to 90°. The MMCs had larger erosion rates than the neat Al matrix metal. This may be due to the reason that brittle reinforcement particles are fractured by the impact of erodent, and easily removed from the eroded surface. It was also found that the maximum erosion rate occurred at the impact angle between 15° and 30° for the MMCs, while it occurred at the impact angle of 15° for the neat Al matrix metal. The inclusion of brittle fillers such as Al₂O₃ and SiC particles in MMCs was supposed to induce the shift of the location where the maximum erosion occurred to a higher impact angle, compared with the neat Al matrix metal. On the other hand, Saravanan et al.¹⁰⁷ performed solid particle erosion experiments of Al alloy reinforced by SiC particles with the average size of 40 µm, the volume content of which was 10 vol%, and showed that the erosion rate of the MMC was comparable to that of the unreinforced alloy.

Alahelisten et al.¹⁰⁸ studied erosion characteristics of alumina short fiber reinforced Al, Mg and Mg alloy (Mg-9Al-1Zn). The alumina short fibers had the mean diameter of 3 µm and the length of 500 µm, and the fiber volume contents were 0, 10, and 30 vol% for the Al-based MMC, and 0, 10, and 20 vol% for the Mg-based and the Mg alloy-based MMCs, respectively. SiC particles with the sizes of 100–150 and 300–500 µm were used as erodent. The particle velocities were 46 m/s for smaller erodent and 43 m/s for larger erodent. The impact angle was 45°. The Al-based and Mg-based MMCs displayed decreasing erosion resistance with increasing fiber content. On the other hand, the Mg alloy-based MMC showed the optimum erosion resistance at the fiber content of 10 vol%. In general, brittle fibers tend to crack and their fragments are easily removed from the eroded surface, so that the MMCs have worse erosion resistance than the neat matrix metals, as shown in the erosion study by Wu et al.¹⁰⁶ In the case of Mg alloy-based MMC, the erosion behavior was supposed to be affected not only by the inclusion of fibers but also the microstructure of the alloy.

Miyazaki and Funakura¹⁰⁹ carried out solid particle erosion experiments of MMCs to clarify the effects of fiber volume content, type of reinforcement fibers and impact angle on the erosion characteristics of MMCs. Three kinds of MMCs were used for the erosion experiments. They were aluminum 6061 alloy reinforced by βSi₃N₄ whisker (βSi₃N₄/6061Al), aluminum 1070 alloy unidirectionally reinforced by Si-Ti-C-O fiber (Si-Ti-C-O/1070Al), and pure aluminum reinforced by SiC plain-woven cloth (SiC/Al). Table 5 shows the fiber volume content, fiber length, fiber average diameter, and density of the respective MMCs. The solid particles used as erodent were SiC abrasive with the size ranging from 100 to 150 µm. Erosion experiments were performed under the several particle velocities V_p and the impact angles α of 15°, 30°, 45°, 60°, and 90°. For the case of the unidirectionally reinforced MMC, Si-Ti-C-O/1070Al, the angle between the solid particle flow and fiber alignment direction β was chosen as 0° (parallel impact) and 90°(perpendicular impact). The relations between the volumetric erosion rate and the impact angle of the respective MMCs are respectively shown in Figure 11 for βSi₃N₄/6061Al, Figure 12 for Si-Ti-C-O/1070Al, and Figure 13 for SiC/Al, together with those of the corresponding neat matrix metals. In the case of βSi₃N₄/6061Al shown in Figure 11, the erosion rate of MMC with the fiber volume content (V_f) of 12 vol% is nearly the same as that of the corresponding neat matrix metal except around α = 30°. On the other hand, the MMC of V_f = 30 vol% has larger erosion rate than the corresponding neat matrix metal, but the difference between two materials are small. In the case of Si-Ti-C-O/1070Al shown in Figure 12, the MMC has much larger erosion rate than the corresponding neat matrix metal. As for the effect of the fiber orientation to erodent flow, the erosion rate of the perpendicular impact (β = 90°) is larger than that of the parallel impact (β = 0°). This phenomenon is also observed in the erosion of unidirectional PMCs as described in the preceding chapter. As shown in Figure 12, the maximum value of erosion rate is located around α = 45–60° for the MMC and around α = 30° for the corresponding neat matrix metal; that is, the peak of erosion rate for the MMC shifts to a higher impact angle compared with the corresponding neat matrix metal. This is because brittle nature of reinforcement fibers appears due to the high volume content of fiber. In the case of SiC/Al composite shown in Figure 13, the MMC has much larger erosion rate than the corresponding neat matrix metal, and the peak of the erosion rate for the MMC shifts to a little bit higher impact angle than for the corresponding neat matrix metal because of small amount of brittle fibers, compared with Si-Ti-C-O/1070Al. The following important conclusions are derived:

For all MMCs including brittle fillers, the erosion rates of MMCs are larger than those of the corresponding neat matrix metals.

In the case where the sizes of reinforcement fibers are much smaller than those of the particles impacted, almost all the fibers removed from the MMC are included in a matrix material removed by the impact of particles. In this case, the difference in erosion rate is small between the MMC and the corresponding matrix metal.

In the case where the reinforcement fibers are included in form of unidirectionally long fibers or cloths, fiber-rich parts are selectively eroded as shown in Figure 14 and a lot of fibers are removed from an eroded surface of the MMC as shown in Figure 15. In this case, the MMC has much larger erosion rate than the corresponding matrix metal.

Table 5.

Composition of tested materials; βSi₃N₄/6061Al, Si-Ti-C-O/1070Al and SiC/Al.

Designation	Fiber volume content (%)	Fiber length (µm)	Average fiber diameter (µm)	Density (g/cm³)
βSi₃N₄/6061Al A	12	10–30	0.5–1.5	2.75
βSi₃N₄/6061Al B	30	10–30	0.5–1.5	2.86
6061Al	—	—	—	2.70
Si-Ti-C-O/1070Al	55	n.a.	8–12	2.51
1070Al	—	—	—	2.71
SiC/Al	33	n.a.	10–15	2.53
Al	—	—	—	2.63

Figure 11.

Relation between volumetric erosion rate and impact angle for βSi₃N₄/6061Al composites; V_p = 54.5 m/s (cited from Miyazaki and Funakura¹⁰⁹).

Figure 12.

Relation between volumetric erosion rate and impact angle for Si-Ti-C-O/1070Al composite; V_p = 54.5 m/s (cited from Miyazaki and Funakura¹⁰⁹).

Figure 13.

Relation between volumetric erosion rate and impact angle for SiC/Al composite; V_p = 33.1 m/s (cited from Miyazaki and Funakura¹⁰⁹).

Figure 14.

Cross-sectional view of eroded surface of SiC/Al taken by optical microscope; V_p = 48.9 m/s and α = 90° (cited from Miyazaki and Funakura¹⁰⁹).

Figure 15.

A SEM micrograph of eroded surface of SiC/Al; V_p = 48.9 m/s and α = 30° (cited from Miyazaki and Funakura¹⁰⁹).

Tu et al.^110,111 studied solid particle erosion behavior of alumina borate (Al₁₈B₄O₃₃) whisker reinforced AC4C Al alloy (Al₁₈B₄O₃₃/Al). The volume content of whisker was 19.5 vol%. Erosion experiments were performed using angular silica sand with the size of 106–149 µm under the particle velocities of 55, 140, 198, and 243 m/s. The impact angle was changed from 15° to 90°. Experimental data showed that the erosion rate of the MMC was larger than that of the neat matrix metal except at a very shallow impact angle α = 15°.¹¹⁰ They conducted two reaction treatments in order to discuss the effect of interface interaction between the Al₁₈B₄O₃₃ whiskers and AC4C Al matrix.¹¹¹ In the reaction treatments, the specimens were treated at 500℃ for 4 hrs and 525℃ for 8 hrs. Erosion experiments were performed not only for heat-treated specimens but also for untreated specimens. The ranking of the erosion resistance was as follows: Treated specimen (500℃ for 4 hrs) > Untreated specimen > Treated specimen (525℃ for 8 hrs). In the specimens treated at 500℃ for 4 hrs, formation of a strong bond between the whiskers and matrix metal resulted in higher erosion resistance. On the other hand, in the test specimen treated at 525℃ for 8 hrs, a weak and brittle layer was maybe formed at the whisker-matrix interface due to excessive treatment. Fracture of such a weak bond between the whiskers and matrix metal reduced the erosion resistance of the test specimen treated at 525℃ for 8 hrs, compared with that of the untreated specimen.

Levin et al.¹¹² performed solid particle erosion experiments of Al₂O₃/Ni composites fabricated using a hot isostatic pressing (HIP) and an electrodeposition technique. The composites fabricated by the HIP contained 0–45 vol% of Al₂O₃ particles with the size of 12 µm. On the other hand, the electrodeposited composites contained 0–39 vol% of much smaller Al₂O₃ particles with the size of about 1 µm. In the latter case, erosion test specimens were made by depositing an Al₂O₃/Ni coating layer with the thickness of about 100 µm on a pure Ni substrate. The erosion experiments were performed using angular alumina particles with the size of 355–425 µm as erodent, the impact angle and the particle velocity of which were respectively 90° and 40 m/s. It was found that for both types of composites the erosion resistance of each composite was lower than that of the neat matrix metal and decreased with the increase in Al₂O₃ content. The electrodeposition composites with small Al₂O₃ particles (1 µm) had better erosion resistance than the composites fabricated by the HIP with large Al₂O₃ particles (12 µm).

Gui et al.¹¹³ studied solid particle erosion characteristics of 15 vol% TiC particle reinforced Al-5Cu-based composite (TiC/Al-5Cu) as well as the monolithic Al-5Cu alloy. The size of TiC particles was in the range of 1–3 µm. The erodent particles used were silica sand. The particle velocity was 83 m/s and three impact angles, 15°, 35°, and 90°, were chosen. The erosion rate of the TiC/Al-5Cu composite was higher than that of the monolithic Al-5Cu alloy by 15–20%.

Das et al.¹¹⁴ performed solid particle erosion experiments of SiC particle reinforced Al alloy composites (SiC/Al). The average size of SiC particle was 75 µm, and its weight content was 10 wt%. Two kinds of SiC/Al composites, that is, as-cast composite ((SiC/Al)_AC) and heat-treated composite ((SiC/Al)_HT), were used as erosion test specimens. Erosion experiments of as-cast Al alloy (Al_AC) and heat-treated Al alloy (Al_HT) were carried out in addition to the composites. In the erosion experiments, Al₂O₃ erodent with the average size of 75 µm was used, and the particle velocity was 100 m/s. The impact angle was changed from 15° to 90°. The following results were obtained for erosion resistance:

The erosion resistance of each composite, (SiC/Al)_AC or (SiC/Al)_HT, was higher than that of the matrix metal, Al_AC or Al_HT.

The erosion resistance of the heat treated material, (SiC/Al)_HT or alloy Al_HT, was higher than that of the as-cast material, (SiC/Al)_AC or Al_AC.

The result (a) is different from the preceding researches,^{102,106,108,109–113} in which the erosion resistance of composite was lower than that of corresponding matrix metal. Erosion resistance of composites depends on the bonding between the metallic matrix and the fillers. The SEM observation showed the good bonding between the Al alloy matrix and the SiC particles, which led to high erosion resistance.

Kumar et al.¹¹⁵ performed solid particle erosion experiments of A356 alloy reinforced by in situ TiB₂ particles which were fabricated by the reaction of halide salts with aluminum melt, and formation of brittle Al₃Ti phase was completely suppressed. The particles were less than 2 µm. The weight content of TiB₂ was 10 wt%. SiC particles with the average sizes of 50 and 110 µm were used as erodent. The particle velocities were 20 and 50 m/s, and the impact angles were 30° and 90°. They used Taguchi’s design of experiment technique to obtain the erosion experimental data effectively, and developed a mathematical model for predicting erosion rate by using regression analysis. Their results showed that the inclusion of in situ TiB₂ particles improved the erosion resistance of unreinforced A356 alloy as in the case of Das et al.’s study.¹¹⁴ They supposed that the higher hardness of TiB₂ than that of SiC erodent particle and good bonding strength between TiB₂ particles and Al¹¹⁶ led to the improvement of erosion resistance compared with the neat A356 alloy. Similar result was obtained in solid particle erosion experiments of aluminum reinforced with redmud composite¹¹⁷ and bronze reinforced with diamond particulate composite.¹¹⁸ Redmud is a major waste material during production of aluminum from bauxite by the Bayer’s process. In both researches,^117,118 solid particle erosion experiments were carried out by changing filler content of composites, and the results indicated that the erosion resistance of the composites increased with the filler content.

Mamatha et al.¹¹⁹ performed solid particle erosion experiments of Zn-Al alloy reinforced by SiC particles (SiC/Zn-Al), and finite element simulations of erosion characteristics of the composites to validate experimental results. It was observed that the SiC/Zn-Al composites exhibited the better erosion resistance than unreinforced Zn-Al alloy. The finite element simulations agreed well with the test data.

Paul et al.¹²⁰ examined solid particle erosion characteristics of Ni-based composite clad layer with WC particles deposited on austenitic stainless steel using a CO₂ laser-based cladding system. The WC particle contents were 5, 10, and 15 wt%. In the erosion experiments, alumina particles with the size of 50 µm were used as erodent. The impact angles and the particle velocities were 30°, 60°, and 90°, and 30, 50, and 70 m/s, respectively. The erosion resistance of the composites increased with the increase in WC content. The erosion resistance of the Ni-clad layer with WC was found to be at least four times higher than that without WC particles. The observation of the eroded surface indicated that the erosion was primarily governed by ductile erosion of Ni matrix followed by the removal of WC particulates from the matrix.

Cermets

Cermet is named after combining a part of words ceramic (cer) and metal (met), and is defined as a composite material composed of ceramic and metallic materials. Although cermet is classified into MMC, the volume content of a metal in a cermet is usually less than 20 vol%, and the metal is used as a binder for a ceramic such as carbide, nitride, oxide, and boride. Solid particle erosion characteristics of cermets are reviewed here.

Ninham and Levy¹²¹ performed solid particle erosion experiments of a number of alloys composed of carbide particles in a ductile metal. They used the alloys with carbide volume content less than 30 vol% (spheroidized plain carbon steels and cast high carbon alloys) and greater than 60 vol% (cermets). Erosion experiments were carried out using 75–200 µm angular quartz particles as erodent at the particle velocity of 60 m/s and impact angles of 30°, 60°, and 90°. For low carbide content alloys, the carbides promoted the erosion. This may be attributed to the loss of ductility for small size carbides (1–2 µm). For larger carbides, carbide fracture and void formation at carbides resulted in increased erosion rate in addition to the loss of ductility. On the other hand, high carbide content cermets showed increasing erosion resistance with the increase in carbide content. This may be explained as follows. In the cermets, volume content of carbide is so high compared with that of metal that the erosion behavior is mainly dominated by the ceramic phase (carbide). In general, a ceramic has much higher erosion resistance than a metal used as binder of cermet.

Anand and Conrad¹²² performed solid particle erosion experiments of WC-6 wt%Co cermet to study the effect of carbide grain size and influence of scaling between the impact damage size and microstructure on the erosion characteristics. Hereafter a cermet is denoted by A-B, in which A and B represent a ceramic phase and a metal binder, respectively. Experimental conditions were changed as follows: WC grain size = 0.57–3.25 µm, Al₂O₃ erodent size = 63–405 µm, particle velocity = 35–93 m/s and impact angle = 20–90°. The erosion data showed that the impact angle, at which the maximum erosion rate occurred, decreased from 90° as the erodent particle size and the particle velocity increased and the WC grain size decreased. SEM observations on the eroded surface revealed that the erosion mode was of a brittle type when the impact crater encompassed less than 10 WC grains, whereas the erosion mode sifted to a ductile mode when it encompassed more than 100 WC grains. So larger WC grain size and smaller erodent size tended to show a brittle erosion behavior and smaller WC grain size and larger erodent size tended to show a ductile one. The main material removal mechanism in the brittle erosion mode was cracking and crushing of WC grains, while that in the ductile erosion mode was plastic deformation and fracture of metal binder phase, WC grains remaining relatively intact, which resulted in smaller erosion rate compared with the brittle erosion mode.

Wang and Luer¹²³ studied erosion-oxidation behavior of a hypersonic velocity oxygen fuel (HVOF) Cr₃C₂-NiCr cermet coating using elevated temperature service environments including fluidized bed boiler, coal-fired boiler, municipal waste incinerator, and so on. The material used in the erosion experiments was HVOF Cr₃C₂-NiCr cermet containing 75% Cr₃C₂ and 25% NiCr as a binder coated on a mild steel. The erosion behavior of HVOF Cr₃C₂-NiCr coating was compared with those of 1018 steel and coatings produced by thermal spray methods other than HVOF. The test temperature was changed from 25℃ to 750℃. Two kinds of boiler bed ash with the sizes of 421 and 649 µm were used as erodent. The impact angle and the particle velocity were changed from 15° to 90° and from 20 to 80 m/s, respectively. HVOF Cr₃C₂-NiCr test specimens showed a brittle erosion-oxidation behavior, in which the maximum thickness loss of the test specimens occurred at the impact angle of 90°. The HVOF Cr₃C₂-NiC had better erosion-oxidation resistance than 1018 steel and other coating produced by thermal spray methods. This is because the HVOF Cr₃C₂-NiC had low porosity, fine grain structure and homogeneous distribution of hard carbides/oxides which formed a skeletal network within a ductile and corrosive-resistant metal binder. As for the temperature dependence of erosion-oxidation resistance, it increased from 25 to 300℃ and then decreased from 300 to 750℃. Afterwards Wang and Verstak¹²⁴ studied the solid particle erosion behavior of HVOF Cr₃C₂/TiC-NiCrMo cermet coating and compared the elevated temperature erosion behavior of this coating with those of 1018 low carbon steel and other thermal spray coating such as HVOF Cr₃C₂-NiCr, HVOF WC-Co cermet coatings and so on. The HVOF Cr₃C₂/TiC-NiCrMo coating exhibited excellent erosion resistance with 2–3 times less thickness loss due to erosion than the HVOF Cr₃C₂-NiCr coating and the same level of erosion resistance as the HVOF WC-Co coating. The temperature dependence of erosion resistance for the HVOF Cr₃C₂/TiC-NiCrMo coating was similar to that for the HVOF Cr₃C₂-NiCr. The erosion resistance of the HVOF Cr₃C₂/TiC-NiCrMo coating was higher than that of the HVOF Cr₃C₂-NiCr below 600℃, but vice versa over 600℃. Thus, the HVOF Cr₃C₂/TiC-NiCrMo coating was not suitable for application over 600℃ during service. Wang and Shui¹²⁵ studied the influence of powder type for spray coating on the hot erosion behavior of HVOF chromium carbide-metal cermet coatings. Moreover, they performed solid particle erosion experiments of HVOF WC-17CrCo cermet coating in addition to HVOF chromium carbide-metal cermet coatings tested in Ref. 125, and obtained the result that among the coatings tested the HVOF WC-17CrCo coating exhibited the highest erosion resistance.¹²⁶

Reshetnyak and Kuybarsepp¹²⁷ studied solid particle erosion behavior of the WC-Co cermets, the Co content of which was 9–20 wt% and the TiC-based cermets, the TiC content of which was 40–80 wt%. In the TiC-based cermets, steels with considerably different composition and structure were used as binders of cermets. Erosion test conditions were as follows: the size of abrasive quartz sand = 0.2–0.3 mm, the particle velocity = 80 m/s and the impact angle = 30°. They found from the erosion data that the erosion resistance of cermets depended on the combined factor defined as the product of hardness and compressive strength or proof stress, and proposed an equation for evaluating the erosion resistance as a function of modulus of elasticity and 0.1% proof stress in compression.

D’Errico et al.¹²⁸ studied the erosion characteristics of cermets composed of TiCN as the primary ceramic phase, TaC, NbC, WC, and Mo₂C as the secondary ceramic phase, and Ni and Co as a binder phase. They found that hardness was the most important controlling factor for solid particle erosion in addition to toughness.

Hussainova et al. performed a series of researches on the solid particle characteristics of cermets.^129–137 In Refs. 129 and 130, the solid particle erosion resistance of cermets with different composition was investigated. Completely the same erosion data were used in both papers. The test specimens used were made of WC-Co cermets, TiC-based cermets and Cr₃C₂-Ni cermets. Solid particle erosion experiments were performed using SiC and SiO₂ abrasive particles with the size of 0.1–0.3 mm, and the particle velocity and the impact angle were changed from 30 to 80 m/s and from 30° to 90°, respectively. According to the erosion data, all tested materials showed a significantly lower erosion rate when eroded by softer SiO₂ particles than by harder SiC particles. According to the erosion data for three types of cermets, the maximum erosion rate occurred at the impact angle α of 60° for the WC-Co cermets, at α = 75° for the TiC-based cermets and at α = 90° for the Cr₃C₂-Ni cermets. So the TiC-based cermets showed a brittle erosion behavior, while the WC-Co cermets showed rather a ductile erosion behavior. Among the cermets tested, the WC-Co cermets had the highest erosion resistance. It was also found that the erosion resistance of the tested cermets mainly depended on modulus of elasticity and fracture toughness values, while the hardness seemed to be of minor importance. In Ref. 131, the solid particle erosion characteristics of TiC-based cermets, that is, TiC-NiMo and TiC-FeNi, were studied by varying the weight content of TiC phase from 50 to 80 wt%. The erosion resistance of TiC-based cermets increased with decreasing binder metal content. In the TiC-NiMo cermets, the erosion resistance was improved by increasing Mo content. Such improvement in erosion resistance is believed to be mainly due to the increase in interphase bond strength and the decrease in the number of microstructural flows, which are stress concentrators needed for nucleation and propagation cracks resulting in material removal. Hussainova concluded that microstructural factors including binder chemistry, grain boundary structures, and residual stresses created during sintering process which were induced by thermal expansion mismatch in different phases in cermets had a primarily important effect on the erosion behavior of the Ti-based cermets rather than the hardness and fracture toughness. Afterwards Hussainova et al.¹³² performed a residual stress analysis of TiC-NiMo cermets. They used an analysis model composed of TiC core, Mo₂C rim and Ni matrix. Such a core-rim-matrix model was based on the SEM images of TiC-NiMo cermets. They found that the erosion rates of TiC-NiMo cermets obtained from the erosion experiments were well correlated with the radial residual stress at the rim-matrix interface after cooling down from high sintering process temperature to the room temperature. The effect of mechanically mixed layer (MML) on the solid particle erosion characteristics of cermets were examined in Ref. 132. The MML is the mixture of metal phase, ceramic phase, wear debris, crushed abrasive, parts of oxidation scales, etc., which can be observed on the surface and subsurface regions. In Ref. 133, the measurements of MML formed by solid particle erosion were presented for three kinds of cermets, WC-12vol%Co, Cr₃C₂-12vol%Ni, and TiC-12vol%NiMo. The thickness of MML layer was determined from the hardness measurements along the depth direction from the eroded surface. Experimental results showed that erosion resistance depended on thickness and hardness of MML as well as on applied test conditions such as impact angle, particle velocity, erodent size, and so on. Among the three kinds of cermets tested, the WC-12vol%Co had the lowest erosion resistance and the thickest MML consisted of oxides and damaged layer of bulk material. The TiC-12vol%NiMo material showed the highest erosion resistance and MML formed on its surface was less pronounced as compared with other cermets tested in this study. In Ref. 134 solid particle erosion characteristics of Cr₃C₂-Ni cermets with three different Ni binder content of 10, 20, and 30 wt% were investigated. The cermets were prepared by two different methods, that is, conventional powder metallurgy sintering and reactive carburizing sintering. The erosion experiments were performed using SiO₂ particles with the size of 0.1–0.3 mm, and the particle velocity and the impact angle were respectively changed from 30 to 80 m/s and from 30° to 90°. The maximum erosion rate occurred at the impact angle of 90° for the Ni binder content of 10 and 20 wt%, while it occurred at 75° for the Ni binder content of 30 wt%. Such a shift of erosion peak to a lower impact angle for the Ni binder content of 30 wt% may be due to the effect of ductile metal binder. The erosion rate increased with the increase in Ni binder content. Moreover, the reactive carburizing sintered cermets had moderately better erosion resistance than conventionally produced ones. Hussainova and Schade¹³⁵ examined the correlation between erosion rate and energy absorbed by a target material during particle-target collision. They performed solid particle erosion experiments of WC-Co, Cr₃C₂-Ni, Cr₃C₂-NiMo, and Ti-NiMo cermets to obtain erosion rate. They also calculated the energy values absorbed by the respective target materials using a formula derived by themselves. They found the common tendency of increase in erosion rate with the increase in the amount of energy absorbed by the target material. Hussainova et al.¹³⁶ performed solid particle erosion experiments of Cr₃C₂-Ni cermets with different additives for Ni binder, and examined the effect of the additives on the erosion characteristic of Cr₃C₂-Ni cermets. Mo and Cu were used as the additives. As a result, the additive Mo improved the erosion resistance of the Cr₃C₂-Ni cermets, but the additive Cu did not. Addition of Mo in Cr₃C₂-Ni cermets resulted in lowering of residual stress and improving phase bonding, which led to the increase in erosion resistance. In Ref. 137 the erosion resistance of WC-based and yttria-stabilized zirconia doped (13 vol%) cermets with different metal binders (Co, Ni, or Fe) were obtained by solid particle erosion experiments using silica abrasive particles with the size of 0.1–0.3 mm at the particle velocity of 80 m/s and the impact angles of 30° and 90°. Generally, the WC-based and yttria-stabilized zirconia doped cermets with Co or Ni binder had better erosion resistance than that with Ni and Fe binder and the conventional WC-Co cermet. The erosion resistance of the WC-based and yttria-stabilized zirconia doped cermets with Co or Ni binder was highly dependent on sintering conditions.

Alman et al.¹³⁸ performed solid particle erosion experiments for three kinds of cermets, that is, TiB₂(40vol%)-FeAl, TiC(80vol%)-FeAl, and WC(80vol%)-FeAl cermets, and compared with the erosion behavior of WC(90vol%)-Co. Alumina particles with the size of 50 µm were used as erodent, and the particle velocity and the impact angle were 40 m/s and 75°, respectively. The test temperature was changed from the room temperature to 700℃. The ranking of erosion resistance for the three-kinds of cermets was as follows; WC(80vol%)-FeAl > TiB₂(40vol%)-FeAl > TiC(80vol%)-FeAl. Erosion resistance was related to both cermet microstructure and material removal mechanism. As for the erosion characteristics at elevated temperatures, the erosion rates of the three kinds of cermets tested in this study were roughly of the same magnitude as the WC(90vol%)-Co at 700℃, even though they had lower contents of hard ceramic phases than the WC(90vol%)-Co cermet. The erosion resistance was constant for the TiB₂(40vol%)-FeAl or increased for the WC(80vol%)-FeAl and TiC(80vol%)-FeAl cermets with the increase in test temperature, whereas it decreased with the increase in test temperature for the WC(90vol%)-Co cermets. Such erosion behavior was related to the oxidation resistance of Co binder and FeAl binder. The FeAl binder of the WC(80vol%)-FeAl and TiC(80vol%)-FeAl cermets provided more oxidation protection than the Co binder of the WC(90vol%)-Co cermets. During erosion at elevated temperatures, the formation of oxide products on the surface could significantly affect erosion resistance. A brittle oxide coating was more easily removed during erosion.

Rateick et al.¹³⁹ conducted solid particle erosion experiments of WC-Co cermet consisted of 90vol% of submicrometer WC embedded in 10vol% Co binder and hardened 440C stainless steel. Angular alumina abrasive was used as the erodent. Experimental variables were as follows: impact angle = 20°, 50° and 90°; particle velocity = 60 m/s and 120 m/s; erodent nominal diameter = 63 and 143 µm. Erosion data showed that the cermet had better erosion resistance than the stainless steel. The erosion mechanism of the stainless steel was more classically ductile, whereas that of the cermet was a combination of both ductile and brittle fracture.

Lin et al.¹⁴⁰ performed solid particle erosion experiments of TiC-Fe cermets using SiO₂ and Al₂O₃ particles as erodent. The particle velocity was 60 m/s and the test temperatures were the room temperature and 350℃. The erosion rate was higher in the case of Al₂O₃ erodent than in the case of SiO₂ erodent, because the hardness of Al₂O₃ was higher than that of SiO₂. As for the temperature effect on the erosion rate, the erosion rate at 350℃ was higher than that at the room temperature.

Sun et al.¹⁴¹ examined the erosion resistance of ultrafine grained WC-Co cermets containing different amounts of ceria nanoparticles up to 0.6 wt% prepared by spark plasma sintering. Addition of ceria nanoparticles effectively suppressed the abnormal grain growth of WC, leading to uniform and fine microstructures. Such ultrafine grained WC-Co cermets improved both hardness and fracture toughness and led to enhancement of erosion resistance. When the content of ceria nanoparticles was more than 1 wt%, Co pools started to form, which resulted in lowering material density, hardness and fracture toughness, and consequently the erosion resistance.

Ceramic matrix composites

Honeycutt et al.¹⁴² performed a pioneering research on solid particle erosion of CMCs. They examined the erosion behavior of tantalum carbide composites containing excess free carbon and rhenium alloyed tungsten fibers. Individual specimens containing 0–3 wt% excess carbon and 0–5 wt% tungsten-rhenium fiber were used in the erosion experiments. Alumina particles were used as erodent. The test temperature was 1800℃. It was shown that the tantalum carbide composites had higher erosion resistance at a high temperature than graphite and hafnia-tungusten composite.

Sykes et al.¹⁴³ performed solid particle erosion experiments of SiC fiber reinforced alumina ceramic composites. The composites contained SiC fibers contents of 5, 15, and 25 wt% (6.1, 17.7, and 28.8 vol%, respectively). The SiC fibers were approximately 1 µm in diameter and 30 µm in length. Alumina abrasive particles with the mean diameter of 37, 130, and 270 µm were used as erodent, and the particle velocities were 75, 100, and 125 m/s. The impact angle was changed from 15° to 90°. Addition of SiC fibers improved the erosion resistance, compared with neat alumina ceramic, and the erosion resistance increased with increasing fiber content. The erosion resistance of alumina ceramic increased by a factor of five or more by the addition of SiC fibers. Such improvement of erosion resistance may be due to the reason that the fibers offer resistance to crack propagation because of a closing-pressure effect.

Wada et al.¹⁴⁴ performed solid particle erosion experiments of Al₂O₃-based composites containing up to 30 wt% SiC particle (SiC_p/Al₂O₃) or whisker (SiC_w/Al₂O₃). Their erosion characteristics were examined using SiC and Al₂O₃ abrasives with the average particle diameter of 500 µm. The particle velocity was 250–300 m/s. The erosion rates of the composites decreased with the increase in SiC content regardless of abrasive. Thus, the addition of SiC particle or whisker improved the erosion resistance of Al₂O₃ ceramic. Especially, the erosion rate of Al₂O₃ with 30 wt% SiC whisker was four times smaller than that of monolithic Al₂O₃ ceramic. The decrease in erosion rate was related to the increase of hardness and fracture toughness of composites due to SiC addition. The erosion rates of whisker composites were smaller than those of particle composites, when the SiC content was the same. As for the effect of abrasive on erosion rate, the erosion rates by SiC abrasive were one order larger than those by Al₂O₃ abrasive. Similar results were obtained by Kamiya et al.¹⁴⁵ for tetragonal zirconia (t-ZrO₂ with 3 mol% Y₂O₃) toughened Al₂O₃ composites. The phenomenon that erosion rate depends on the erodent can be explained as follows. According to a diametral compression test under quasi-static loading, the mean fracture load of SiC erodent particles was 2.3 times higher than for Al₂O₃ particles.¹⁴⁵ Almost all Al₂O₃ erodent particles broke at collision with target materials, but SiC erodent particles were only slightly damaged. Thus, the mean impact force due to Al₂O₃ erodent particles was smaller than that due to SiC erodent particles and did not reach the threshold force for lateral crack fracture of composites.

Liu et al.¹⁴⁶ performed solid particle erosion experiments of Cr₃C₂/Al₂O₃ ceramic composites with Cr₃C₂ particles ranging from 10 to 40 vol% using SiC erodent particles with the size of 150 µm at the particle velocity of 95 m/s. Experimental results showed that the erosion rate of the composite was smaller than that of monolithic Al₂O₃ ceramic and decreased with increasing Cr₃C₂ content. That is, addition of Cr₃C₂ particles into Al₂O₃ ceramic improved erosion resistance. Based on the erosion data, they gave a formula for the measured erosion rate of the composite as a function of hardness, fracture toughness, and microstructure parameter representing the average spacing between Cr₃C₂ particles.

Gochnour et al.¹⁴⁷ examined the erosion characteristics of Al₂OC/SiC ceramic composites with Al₂OC content varying from 5 to 75 wt% using alumina erodent particles. The particle velocity and the impact angle were 120 m/s and 90°, respectively. The erosion rate of the ceramic composite increased with increasing Al₂OC content. According to the experimental results, the Al₂OC/SiC ceramic composites became harder with decreasing Al₂OC content, because the SiC phase was harder than the Al₂OC, and fracture toughness was almost constant irrespective of Al₂OC content. The erosion experimental results showed that the volumetric erosion rate of SiC/Al₂OC ceramic composites was well correlated with the hardness of the composite ceramics.

Kim and Park^148,149 performed solid particle erosion experiments of SiC ceramic composite with TiB₂ particles (TiB₂/SiC). TiB₂ particles were added to SiC to enhance the fracture toughness. Al₂O₃ or Y₂O₃ was added into monolithic SiC and TiB₂/SiC as sintering aid for hot-pressed sintered materials. In Ref. 148 the monolithic SiC containing 5 vol% Al₂O₃ and the composite TiB₂/SiC with 5 vol% Al₂O₃ and 14 vol% TiB₂ were used as erosion test specimens, while the monolithic SiC containing 5 vol% Y₂O₃ and the composite TiB₂/SiC with 5 vol% Y₂O₃ and 14 vol% TiB₂ were used in Ref. 149. The erodent used was SiC with the average sizes of 50, 100, and 150 µm. The particle velocities were 40, 70, and 100 m/s. The impact angles were 30°, 60°, and 90°. In Ref. 148 the authors concluded that the erosion rate of the monolithic SiC was higher than that of the composite TiB₂/SiC in the case where Al₂O₃ was used as sintering aid, but such a conclusion seems to be incorrect. According to erosion rate versus impact angle curves given in Ref. 148, the general trend may indicate that the erosion rate of the monolithic SiC is higher than the composite TiB₂/SiC for a high impact velocity (100 m/s), and vice versa for a low impact velocity (40 m/s). On the other hand, the erosion rate was higher for the composite TiB₂/SiC than for the monolithic SiC in the case where Y₂O₃ was used as sintering aid.¹⁴⁹ Kim and Park did not provide the definite reason why the erosion rate was dependent on the kind of sintering aid. The solid particle erosion behavior of the same composite was studied by Colclough and Yoemans.¹⁵⁰ The test materials used were a pressureless sintered monolithic SiC and the composite material with 16 vol% TiB₂ particles, TiB₂/SiC. The erosion experiments were performed both at the room temperature and at elevated temperatures from 400 to 1000℃. The average sizes of SiC particles used as erodent were 105, 250, and 420 µm for the erosion experiments at the room temperature, and 105 µm for the erosion experiments at elevated temperatures. The impact angle was 90°. Although the particle velocity was not measured, the gas stream was 200 m/s for the erosion experiments at the room temperature, and 70 m/s for the erosion experiments at the elevated temperatures. The composite TiB₂/SiC showed a higher erosion rate than the monolithic SiC due to the enhancement of lateral cracking for small scale contact events (low velocities/small particle sizes), and vice versa for large scale impact events and at elevated temperatures. Increase in erosion rate with temperature was due to the reason that the materials became softer, leading to larger plastic zone sizes and thus increased residual stresses to drive lateral crack propagation.

Sharma et al.¹⁵¹ performed solid particle erosion experiments of SiC ceramic with 0, 10, 30, and 50 wt% WC particle (WC/SiC). Erodent particles used were Al₂O₃ and SiC particles with the size of 40–70 µm. The particles velocity was 47 m/s, and the impact angles were 30°, 60°, and 90°. As for the influence of the erodent particle, the erosion rate using Al₂O₃ erodent particles was two orders of magnitude smaller than that using SiC erodent particles because Al₂O₃ erodent particles were softer than SiC erodent particles. According to the erosion experiments using SiC erodent particles, the composites WC/SiC with 10 and 30 wt% WC particle showed reduced erosion rates compared with the monolithic SiC ceramic. The composites WC/SiC with 30 wt% WC particle had the minimum erosion rate, and the erosion rate increased, when WC particles increased from 30 to 50 wt%.

Wang and Levy¹⁵² performed solid particle erosion experiments of SiC fiber reinforced SiC ceramic composite (SiC/SiC). The test specimens were fabricated by chemical vapor deposition (CVD) infiltration of a fiber material. Angular SiO₂ particles of average diameter 130 µm were used as erodent. The particle velocity was 30 m/s, and the impact angles were 30°, 60°, and 90°. The erosion rate of the composite was determined at 25 and 850℃. The erosion rate of the composite at 850℃ was an order of magnitude lower than that at 25℃ due to the increased ductility of the composite. Shu et al.^153,154 also performed solid particle erosion studies of SiC/SiC composite. The composites were fabricated by the nanopowder infiltration and transient eutectoid (NITE) process, a highly optimized liquid phase sintering (LPS) process. Based on the experimental results, they pointed out that erosion resistance increased with decreasing the porosity of the composite.

Park et al.¹⁵⁵ conducted solid particle erosion experiments of sialon-based ceramic composite materials containing 0–20 wt% Si₃N₄ whiskers. Erosion experiments were performed at the room temperature and 773 K using SiC particles with the average diameter of 90–130 µm. The particle velocity and the impact angle were 40 m/s and 90°, respectively. Si₃N₄ whiskers acted as seeds for grain growth and caused formation of large elongated grains in the microstructure. Thus, the composite materials with higher content of Si₃N₄ whiskers showed higher fracture toughness. SEM observations of eroded surfaces revealed that the erosion mechanism was not brittle fracture but grain pullout and plastic deformation. The higher erosion rate was observed for higher temperature and more inclusion of Si₃N₄ whiskers. Liu et al.¹⁵⁶ performed solid particle erosion experiments of dual-phase sialon composites composed of different ratios of α- and β-phase. In the composites, the content of α-phase was changed from 0 to 40 wt%. SiC particles with the average size of 150 µm were used as erodent, and the particle velocity and the impact angle were 80 m/s and 90°, respectively. They obtained the experimental data indicating that the erosion rate decreased with increasing hardness (H) and fracture toughness (K_c), which were varied with ratio of α- and β-phase, and proposed the expression for the erosion rate of the dual-phase sialon composite as a function of H and K_c.

Ham et al.^157,158 performed solid particle erosion experiments of calcium almino-silicate (CAS) glass reinforced with unidirectional SiC fibers (SiC/CAS) with the average fiber diameter of 15 µm. The fiber volume content was 34 vol%. The erodent used was silica sand with the size of 100–150 µm. The particle velocity was 24 m/s. The test temperature was changed from 20 to 726℃. The eroded surface was perpendicular to fiber direction and the impact angle was 90°. The erosion rate shown in Ref. 157 increased significantly in the range of test temperature from the room temperature to 400℃, but decreased slightly or remains roughly constant above the temperature of 400℃. In the SiC/CAS composites, residual tensile stress was induced in the matrix due to CTE (coefficient of thermal expansion) mismatch of fibers and matrix. Lateral cracking was the main mechanism of material removal in the erosion of the SiC/CAS composite. Residual tensile stress in the matrix was a driving force for lateral cracking. Relaxation of such residual stress occurred when the test temperature increased, and it suppressed the material removal due to erosion. That is why the erosion rate of the SiC/CAS composite decreased above the temperature of 400℃. On the other hand, the reason for the increase in erosion rate from the room temperature to 400℃ was given in Ref. 158. The erosion experiments in Ref. 157 were performed at the constant flow rate irrespective of test temperatures. As shown in Ref. 158, air and particle velocity increases with temperature under a constant flow rate. To obtain the relation between erosion rate and temperature under a constant particle velocity, the effect of increasing particle velocity caused by increasing test temperature should be removed from the original erosion rate. As a result of such consideration, the erosion rate of the SiC/CAS composite was shown to decrease from room temperature to 300–400℃ under a constant particle velocity as shown in Ref. 158, which was consistent with the decrease in the driving force for lateral cracking as a result of the relaxation of residual tensile stress in the matrix.

Alman et al.¹⁵⁹ studied solid particle characteristics of MoSi₂/Si₃N₄ composite at the room temperature and elevated temperatures, 180, 500, 700, and 900℃. The MoSi₂/Si₃N₄ composite was fabricated from Si₃N₄ and 50 vol% MoSi₂ powders using a combination of hot-pressing and hot-isostatic pressing. The erosion characteristics of the composite were compared with those of a monolithic β-Si₃N₄, WC-6%Co, 304 SS, IN-800 (Ni-Fe-Cr alloy), and Stellite-6B (Co-Cr-W-Mo alloy). About 50 µm angular Al₂O₃ particles were used as erodent. The particle velocity was 40 m/s and the impact angles were 60°, 75°, and 90°. The erosion rate of the MoSi₂/Si₃N₄ composite was lower than those of β-Si₃N₄, 304 SS, IN-800, and Stellite-6B at the temperatures below 700℃, but higher than that of WC-6%Co. As for the influence of temperature on the erosion rate of the Si₃N₄-MoSi₂ composite, the erosion rate remained relatively constants from room temperature to 700℃, and decreases at 900℃. This decrease may be a result of softening of the MoSi₂/Si₃N₄ composite, especially of the MoSi₂ phase at 900℃. Iizuka et al.¹⁶⁰ performed solid particle erosion experiments of Si₃N₄ matrix composites including 1.5–8.1 wt% Mo₅Si₃ particles (Mo₅Si₃/Si₃N₄). Silica sand with the average diameter of 600 µm was used as erodent. The particle velocity and the impact angle were 70 m/s and 45°, respectively. The composites containing Mo₅Si₃ particles less than 2 wt% showed better erosion resistance compared with the monolithic Si₃N₄ ceramic, whereas the composites containing Mo₅Si₃ particles more than 2 wt% had lower erosion resistance than the monolithic Si₃N₄ ceramic and the erosion resistance decreased with increasing Mo₅Si₃ content. Such a reduction in erosion resistance of the composites was due to the lower values of both fracture toughness and hardness of the composites.

Sato and Kamiya¹⁶¹ performed solid particle erosion experiments of pressureless sintered Al₂O₃ CMCs including TiC particles of 10, 20, 25, and 30 wt% (TiC/Al₂O₃). SiC and Al₂O₃ particles were used as erodent. The impact angle and the particle velocity were changed from 30° to 80° and from 70 to 240 m/s, respectively. The erosion rate of composite slightly decreased up to 30 wt% of the TiC particles for the SiC erodent, while it greatly decreased with increasing TiC content for Al₂O₃ erodent. The SiC erodent induced higher erosion rate of the composite than the Al₂O₃ erodent. The impact force produced by the SiC erodent was much larger than that of the Al₂O₃ erodent, because the SiC particles had larger fracture strength than the Al₂O₃ particles. This is why the erosion rate yielded by the SiC particles was larger than that of the Al₂O₃ particles. Afterwards Sato et al.¹⁶² compared the erosion behavior of pressureless sintered TiC/Al₂O₃ composite including 30 wt% TiC particles with those of various commercial pressurized sintered ceramic composites, ZnO₂, SiC, Si₃N₄, Al₂O₃ monolithic ceramics and WC-Co cermet. For the Al₂O₃ erodent, the pressureless sintered TiC/Al₂O₃ composite has the highest erosion resistance compared with other materials. On the other hand, for the SiC erodent, the WC-Co cermet had the highest erosion resistance, and the pressureless sintered TiC/Al₂O₃ composite showed almost the same erosion resistance as commercial pressurized sintered ceramic composites, ZnO₂ and Al₂O₃.

Saewong and Rawlings¹⁶³ performed solid particle erosion experiments of borosilicate glass laminated composites containing 33 vol% CF (CF/Glass). For comparison, the erosion experiment of monolithic borosilicate glass was also performed. Both tough and brittle test specimens containing different types of CFs were prepared for the erosion experiments. Sand particles of 150–212 µm were used as erodent. The particle velocity was 15.1 m/s. The impact angles were 45°, 60°, and 90°. The tough composite, brittle composite, and monolithic glass behaved a typical brittle erosion behavior in that the erosion rate reached the maximum at the impact angle of 90°. The ranking of erosion resistance was as follows: monolithic glass > brittle composite > tough composite. The hardness of monolithic glass was 2–3 times higher than those of the composites. That is why the monolithic glass had higher erosion resistance than the composites. The fiber pullout length was greater for the tough composite than for the brittle composite. This fact indicates that the brittle composite had stronger bonding between fibers and matrix material. Thus, the erosion resistance of the brittle composite was higher than that of the tough composite.

Yamada et al.¹⁶⁴ fabricated monolithic B₄C and B₄C-20 mol%CrB₂ (CrB₂/B₄C) ceramic composites by hot-pressing at 1900 and 2050℃, and performed solid particle erosion experiments of these materials in addition to a commercial B₄C ceramic. Erodent particle used was Al₂O₃, and the impact angle was 90°. A lower erosion rate was obtained for the monolithic B₄C and B₄C/CrB₂ composite compared with that of a commercial B₄C ceramic. The B₄C/CrB₂ composite hot-pressed at 1900℃ had lower erosion rate than the composite hot-pressed at 2050℃, and almost the same erosion rate as the monolithic B₄C. The erosion characteristics of these materials were discussed from the viewpoint of microstructures and fracture toughness.

Liu and Sun¹⁶⁵ examined the erosion characteristics of B₄C-based ceramic composites made from B₄C and TiO₂ particles by uniaxial hot pressing and pure B₄C ceramic. Three kinds of composites were prepared for erosion experiments. They were NT1, NT2, and NT3 composed of B₄C phase, TiB₂ phase, and C phase. The composition of each composite was as follows:

NT1: B₄C phase = 95.7 wt%, TiB₂ phase = 3.5 wt%, C phase = 0.8 wt%

NT2: B₄C phase = 91.8 wt%, TiB₂ phase = 5.2 wt%, C phase = 3.0 wt%

NT3: B₄C phase = 87.4 wt%, TiB₂ phase = 7.0 wt%, C phase = 5.6 wt%

SiC particles with the size of 150–180 µm were used as erodent. The impact angle was changed from 15° to 90°. The particle velocity was changed from 30 to 75 m/s. The composites had higher erosion resistance than the pure B₄C ceramic and the ranking of erosion resistance was as follows; NT2 > NT1 > NT3 > pure B₄C. This ranking was related with the hardness and fracture toughness of materials. That is, the composite NT2 had highest hardness and fracture toughness and the pure B₄C ceramic had lowest hardness and fracture toughness.

Ouyang et al.¹⁶⁶ performed solid particle erosion experiments of WC ceramic including 4.3 wt% MgO particle (MgO/WC). They examined erosion characteristics of MgO/WC composites with and without grain-growth inhibitor using silica sand with the size of 200 µm as erodent. The particle velocity was 85 m/s. The inhibitors reduced both grain size and MgO particle size, and increased both hardness and fracture toughness. Thus, the MgO/WC composite with grain-growth inhibitor had higher erosion resistance than that without the inhibitor.

Amirthan et al.¹⁶⁷ examined the solid particle erosion characteristics of SiC composites fabricated using natural materials. They used four different kinds of SiC composites as erosion test specimens. Test materials used were cotton fabric-based Si/SiC with and without chemical vapor infiltration (CVI), fine teak wood powder based Si/SiC, and coarse teak wood powder based Si/SiC. These composites composed of SiC, Si, and C phases. They showed the ranking of erosion resistance for these composites.

Li et al.¹⁶⁸ performed solid particle erosion experiments of mullite (alminium silicate)–SiC composite refractories including Si powder from 0 to 12 wt%. The main phase of mullite-SiC composite refractories by nitriding reaction sintering was mullite, α-SiC, and β-sialon. SiC particles with the size of 325–830 µm were used as erodent. The particle velocity and the impact angle were 50 m/s and 90°, respectively. Test temperature was varied from 25 to 1400℃. The erosion resistance of mullite-SiC composite refractories increased with the increase in Si powder addition and test temperature. The content of β-sialon increased with the increase in Si powder addition, and β-sialon enhanced the erosion resistance of mullite-SiC composite refractories.

Carbon is not classified into ceramic, but the erosion characteristics of CF reinforced carbon (C/C) composites are shown here. Baxter et al.^169,170 performed solid particle erosion experiments of carbon-bonded CF (CBCF), which consisted of CF network bonded together at the intersections of the fibers by discrete regions of carbon matrix. Hence, the vast majority of the volume (70–90%) consisted of interconnected pores. The CBCF is a class of highly porous carbon-carbon composites with low density, and utilized as thermal insulation in vacuum or inert gas furnaces at the temperatures up to 2800℃. The CFs used in the CBCF were virgin fibers less than 2 mm in length and recycled fibrous material classified through either a 2 mm or 50 µm sieve. In Ref. 169 the CBCF materials used in the erosion experiments were fabricated by varying the relative proportion of the virgin fibers and the recycled fibrous material, and the relationship between the erosion rate and the proportion of recycled fibrous material was examined. In Ref. 170 CBCF with 66% recycled fibrous material and 33% virgin fibers was densified by CVI. Erosion experiments were performed using 230 µm angular silica as erodent. The particle velocity and the impact angle were 6 m/s and 90°, respectively. The erosion rate decreased exponentially with increase in the bulk density of CBCF.

Sarkar et al.¹⁷¹ examined the erosion characteristics of two and four dimensionally reinforced carbon/carbon (C/C) composites as well as that of CF reinforced phenolic composite (C/P). The two and four dimensionally reinforced C/C are respectively, denoted as C-2D/C and C-4D/C. The volume content of reinforcement was 50 vol% for C-2D/C, 40 vol% C-4D/C, and 50 vol% for C/P. Silica sand with the average size of 200 µm was used as erodent. Erosion experiments were performed for two different impact angles, 30° and 90°, and for two different particle velocities, 45 and 60 m/s. The ranking of erosion resistance was as follows; C/P > > C-2D/C > C-4D/C. According to SEM observations of eroded surfaces, the C/P composite had stronger bonding between the CF and the matrix than the C-2D/C and C-4D/C composites, and thus, the erosion resistance of the C/P composite was higher than those of the C-2D/C and C-4D/C composites. The erosion rate data at the impact angles of 30° and 90° indicated that all the composites used in the erosion experiments showed a semi-ductile erosion response. They calculated the erosion efficiency defined by equation (3). The erosion efficiencies of these composites vary in the range of 23–143 and 40–206% for particle velocities of 45 and 60 m/s, respectively. This fact suggests that the mechanism of erosion involved microploughing as well as microcutting for lower erosion efficiencies and cracking followed by spallation for higher erosion efficiencies.

Erosion-resistant coatings

Composite materials, especially PMCs, are usually inferior to the corresponding monolithic materials in erosion resistance. Therefore, composite materials often need erosion-resistant coatings when they are used in erosive environments. Various kinds of MMCs and CMCs are used as erosion-resistant coatings, some of which are described in the chapters “Metal Matrix Composites” and “Ceramics Matrix Composites”.

First of all, the papers on the erosion-resistant coatings for PMCs are reviewed. Alonso et al.¹⁷² dealt with plasma-splayed erosion-resistant coatings on CF reinforced EP composites (CF/EP). They showed that the spraying of alumina-glass mixture onto CF/EP composite as a bondcoat prior to forming erosion protection coatings of WC-12Co and Al₂O₃ as topcoats was effective for avoiding degradation of composite substrate because of heat sensitivity of polymeric substrate. Solid particle erosion experiments were carried out for the WC-12Co and Al₂O₃ coatings deposited onto CF/EP composite using silica particle and corundum. According to the erosion data, the Al₂O₃ coating showed a brittle erosion behavior with the maximum erosion rate at the impact angle of 90°, while the WC-12Co coating presented some degree of a ductile erosion behavior. They did not provide the results on the degree of enhancing the erosion resistance due to the WC-12Co and Al₂O₃ coatings, compared with uncoated CF/EP composite.

Sutter et al.¹⁷³ and Miyoshi et al.¹⁷⁴ studied solid particle erosion characteristics of coated and uncoated carbon-Kevlar fiber reinforced EP composite. Such a composite was used fan bypass vane in jet engine (AE 3700). The erosion characteristics of two erosion-resistant coating systems were examined in these studies. These coatings systems included a bondcoat and a hard erosion-resistant topcoat. For both coating systems, WC-Co is a primary topcoat constituent. The difference between the two systems was that one coating system contained a bondcoat with no polymer and the other contained a bondcoat with a polymer to improve adhesion to the PMC substrate. Erosion experiments were performed using Arizona road dust as erodent at the temperatures of 294 and 366 K. The impact angles were 20°, 60°, and 90°, and the particle velocity was 229 m/s. The coatings with a bondcoat and a WC-Co topcoat enhanced erosion resistance and reduced the volumetric erosion loss by a factor of nearly two. The coating with a bondcoat containing polymer had better erosion resistance than the coating with a bondcoat containing no polymer.

Ivosevic et al.¹⁷⁵ studied solid particle erosion characteristics of thermally sprayed functionally graded material (FGM) coatings based on PI matrix filled with varying volume contents of WC–Co to improve the erosion and oxidation resistance of PMC. The PMC substrate used was PI matrix (PMR-15) reinforced by a carbon-fiber weave. An FGM coating used in this work had a graded composition going from pure PMR-II PI similar to the PMC substrate matrix material to a 100% wear resistant WC–Co outer layer. Such an FGM coating was useful to prevent thermomechanical fatigue failure caused by large differences in coefficient of thermal expansion between the substrate and coating materials. The erosion experiments were performed for two kinds of coating systems and an uncoated PMC as a control sample. Coating system I was composed of an FGM layer grading from pure PMR-II PI to a 100% erosion resistant WC–Co outer layer. In this system, an FGM layer and a WC–Co outer layer were respectively formed using high velocity oxy-fuel (HVOF) combustion spray process and flame arc process. Coating system II included a metallic Zn binding layer between an FGM layer and a WC-Co topcoat layer. In this system, a metallic Zn layer and a WC-Co topcoat layer were respectively formed using wire-arc process and HVOF process. Angular alumina particles with the nominal diameter of 50 µm were used as erodent. The particle velocity was 100 m/s and the impact angles were 20° and 90°. The test temperatures were the room temperature and 250℃. The Coating system I provided only a small amount of additional erosion protection to the uncoated PMC due to the porous and poorly cohesive WC-Co topcoat layer formed by flame arc process. On the other hand, the Coating system II had very good erosion resistance both at the room temperature and 250℃, compared with the Coating system I and the uncoated PMC.

Liu et al.¹⁷⁶ performed solid particle erosion experiments of coated PMC. The PMC substrate was a CF reinforced thermosetting PI. A cored wire composed of steel skin and Ni–Cr–B–Si as a filler material was used as a coating material, and an erosion-resistant coating was produced on the PMC substrate by arc spray. Erosion experiments were carried out using corundum as erodent at the impact angle of 90°. The erosive weight loss of the coated PMC was half of that of the uncoated one, and the arc-sprayed coatings can provide good protection of the PMC from erosion.

Maurer and Schulz¹⁷⁷ examined solid particle erosion characteristics of Ti-based coatings deposited on CF reinforced EP (CF/EP) or PEEK resin (CF/PEEK). Three kinds of Ti-based coatings, that is, pure Ti, TiN, and Ti-TiN multilayer, were prepared for the erosion experiments. A Ti coating had ductility, moderate hardness and a rather low elastic modulus mismatch compared with a substrate PMC. A TiN coating was expected to offer excellent erosion protection due to its high hardness. A Ti-TiN multilayer coating combined hardness with ductility in one coating system. Erosion experiments were performed by varying coating thickness using alumina particle as erodent at the particle velocity of 65 m/s and the impact angle of 90°. These coatings were produced by physical vapor deposition (PVD) using magnetron sputtering, which is suitable for polymeric substrate because of low process temperature. Coatings with a thickness of 10 µm or less were not able to improve erosion behavior of PMCs. For the coated CF/EP, the 5 µm thin TiN coating did not improve the erosion behavior, while the 30 µm thick Ti coating was able to reduce erosive weight loss by a factor of 5 in comparison with the bare substrate, and the 30 µm thick Ti-TiN multilayer coating by a factor of 4, respectively. On the other hand, for coated CF/PEEK, no reduction of erosive weight loss was observed because PEEK had a high erosion resistance at normal impact angle, but improvement in erosion resistance of around 30% was attained when considering the volumetric erosion rate. Conclusively, pure Ti with a thickness of 30 µm was the most promising coating system for both CF/EP and CF/PEEK substrates for reduction of the erosive wear under normal impact.

Keegan et al.¹⁷⁸ published a review paper on erosion issues associated with the leading edge of wind turbine blades. Most of the wind turbine blades are made of thermosetting PMCs such as epoxy or UP reinforced with GF or CF. Wind turbines are used in erosive environment such as rain, hailstone, sand, dust, and so forth. Erosion resistant coatings for the leading edge of wind turbine fan-blades are one of important problems. Keegan et al. dealt with this problem in their review paper.

Baxter and Rawlings¹⁷⁹ performed solid particle erosion experiments of candidate coatings (colloidal graphite paints) and claddings (dense carbon-carbon composites and graphite foil) employed to protect highly porous CF reinforced carbon (C/C) composites known as carbon-bonded CF (CBCF) composites.^169,170 The erosion experiments were performed using angular silica sand as erodent with the average diameter of 150–300 µm at the particle velocity of 6 m/s. The impact angle was varied from 30° to 90°. The materials used for coatings and claddings on porous CBCF substrate showed a significantly greater erosion resistance than the bared substrate. In general, the coating and cladding materials showed a brittle erosion behavior with the maximum erosion rate at the impact angle of 90° except for the graphite foil, which had the maximum erosion rate at the impact angle of 60°. Smeacetto et al.¹⁸⁰ also studied erosion protective coatings for highly porous CBCF composite. Three coatings were investigated, that is, SiC, TiC, and a glass-ceramic. All coatings were deposited on CVI densified CBCF composite by a low-cost slurry based technique. Although all the coatings exhibited much better erosion resistance than the bared CBCF composite, the glass-ceramic coating had best erosion resistance.

Various types of composite coatings are used to protect substrate materials from erosive wear. Yan et al.¹⁸¹ examined the solid particle erosion characteristics of polymer-ceramic composite coatings manufactured from calcium carbonate and ethylene-methacrylic acid copolymer (EMAA), which was deposited on substrate materials via a thermal spray process employing different CaCO₃ filler sizes, 2.8, 9, and 36 µm, and loading levels from 2.5 to 7 wt%. Solid particle erosion experiments were performed both for pure EMAA coating and for CaCO₃/EMAA composite coatings including different contents of CaCO₃ filler. The erodent used was garnet with the size of 200 to 600 µm. The particle velocity was 20 m/s and the impact angle was changed from 0° to 90°. The erosion experimental data showed that the erosion resistance decreased with the increase in filler content. It was also found that the erosion resistance of CaCO₃/EMAA composite coating containing above 5 wt% CaCO₃ filler was lower than that of the pure EMMA coating.

Liu et al.¹⁸² investigated the effect of the content of Ni on solid particle erosion behavior of the TiC-xNi coating at elevated temperatures. The TiC-xNi composite coatings were in-situ synthesized on nickel-based alloy substrate by the electro-thermal explosion ultra-high speed spraying (EEUSS) device. The Ni content x was varied as 0, 5, 10, 15, and 20 wt%. Solid quartz particles with the average diameter of 250 µm were used as erodent. The erosion experiments were performed under the temperature of 650℃, the particle velocity of 62 m/s and the impact angle of 90°. The erosion resistance of the TiC-xNi coatings increased first and decreased afterwards with the increase in the content of Ni. The coating with 10 wt% Ni addition had the highest erosion resistance, which was about two times higher than that with 20 wt% Ni addition. It was found from the SEM observations of eroded surfaces that the pure TiC coating showed a brittle erosion behavior, whereas the addition of the binder phase of Ni had the effect of mitigating a brittle behavior of TiC, which resulted in improving the erosion resistance compared with the pure TiC coating.

Sharma et al.¹⁸³ studied solid particle erosion characteristics of a coating made from high-velocity oxy-fuel (HVOF) composite powder (MEC 1031C) modified by adding 0.4 wt% La₂O₃ in the composite powder. Solid particle erosion experiments were performed for both modified and unmodified (without La₂O₃) HVOF-sprayed coatings on a carbon steel substrate. Flyash was used as erodent at the particle velocities of 30, 45, and 60 m/s and the impact angles of 30°, 60°, and 90°. Test temperatures were 150, 300, and 450℃. The modified HVOF-sprayed coating showed high erosive resistance compared with unmodified coating. This is due to the increase in hardness of the modified coating. Sharma et al. developed an empirical equation for predicting the erosive wear of coatings in terms of particle velocity, impact angle, test temperature, and erodent feed rate using the response surface methodology based on erosion tests data.

Keller et al.¹⁸⁴ conducted a study on self-healing performance of a coating subjected to erosion damage. They used two self-healing coating systems. One was an elastomeric material with a two-part poly (dimethyl siloxane) healing chemistry, and the other was an epoxy coating with a one-part isocyanate-based healing chemistry. Coatings were microcapsule-based self-healing materials. Both coating systems were evaluated by solid particle erosion experiments. It was found from the erosion experiments that the elastomeric coating system did not successfully self-heal because of a combination of high viscosity and slow reaction kinetics, while the isocyanate-based system successfully self-healed compared with a nonhealing control. Weight loss due to erosion was reduced by almost 300% for this material when compared with a microcapsule-filled nonhealing coating.

Pasha and Kaleemulla¹⁸⁵ published a review paper on solid particle erosion studies of coated MMCs.

Concluding remarks

Solid particle erosion studies on composite materials are critically reviewed in the present article. This review article covers PMCs, MMCs, CMCs, and erosion-resistant coatings from 1960s up to the present. Brief summaries are presented here for each item.

Polymer matrix composites

General characteristics of erosion for PMCs are summarized as follows:

Most PMCs, especially PMCs with brittle or inorganic fillers, show larger erosion rates than the corresponding neat resins, and the increase in filler content of a composite causes the increase in erosion rate. Possible exceptions are the PMCs with organic or ductile fillers, which enhance the erosion resistance of PMCs compared with the corresponding neat resins.

In most PMCs with brittle or inorganic fillers, the location where the maximum erosion rate takes place shifts to a higher impact angle compared with that of the corresponding neat resin.

Stronger adhesion between fillers and a matrix material shows better erosion resistance or smaller erosion rate. Therefore, the surface treatments of filler surfaces to improve adhesion between fillers and a matrix material are effective to enhance erosion resistance of PMCs.

CF and GF are typical fillers for PMCs. In the case of short CF and GF, a CF composite shows higher volumetric erosion rate²⁰ or higher weight erosion rate^38,42 than a GF composite, if the same matrix resin and the same fiber weight content are used in both composites. The results of Refs. 38 and 42 indicate the same result as that of Ref. 20, because a CF composite has smaller density than a GF composite under the condition of the same matrix material and the same fiber weight content. Conclusively, a GF composite has better volumetric erosion resistance than a CF composite, if the same matrix resin and the same fiber weight content are used in both composites. Moreover, the volumetric erosion rate of a composite with short fibers seems to be dominated only by the volume content of short fibers irrespective of kind of fiber, CF or GF, if the same matrix material is used in both composites.²⁰ In the case of continuous long CF and GF, a GF composite tends to show a larger weight erosion rate than a CF composite, if the same matrix resin and the same fiber volume content are used in both composites.^37,41 Considering that the density of a composite with continuous long GFs is larger than that of a composite with continuous long CFs, it cannot be denied that the erosion volume loss of a composite with continuous long fibers may be dominated only by the volumetric content of fibers irrespective of kind of fibers, as in the case of composites with short fibers. A further study is needed to validate this presumption.

In the case of unidirectional fiber reinforced plastics, the erosion resistance of parallel impact, in which the fiber orientation is parallel to the erodent flow direction, tends to be higher than that of perpendicular impact, in which the fiber orientation is perpendicular to the erodent flow direction. In elevated temperatures or in the case of the composites with thermoplastic resins, the parallel impact has sometimes lower erosion resistance than the perpendicular impact.

Almost all the general characteristics of erosion for PMCs described above were already obtained in the researches performed before 2000, especially by Pool et al., Mathias et al., Miyazaki et al., and Saputra et al. A lot of papers were published on the erosion characteristics of various PMCs after 2000, but many of them reconfirmed the general characteristics of erosion described above for various PMCs including new kinds of PMCs such as natural filler reinforced resins and nanofiller reinforced resins. One of the features of the erosion studies after 2000 is to deal with hybrid PMCs including two or more reinforcements. In hybrid PMCs, a low weight content of fillers are usually added to conventional PMCs as modification of a matrix resin in order to enhance erosion resistance of conventional PMCs in addition to high weight content of fillers as reinforcement. Inorganic fillers, organic fillers and metal fillers are used to modify matrix resins of hybrid PMCs. The erosion characteristics of hybrid PMCs are summarized as follows:

Most hybrid PMCs have better erosion resistance than the conventional PMCs without fillers to modify matrix resins.

In most hybrid PMCs, there exists trade-off between the erosion performance and the tensile strength, one of the most important mechanical properties for structural materials. That is, enhancement of erosion resistance by adding fillers to modify a matrix resin may lead to the reduction in tensile strength of a hybrid PMC.

Metal matrix composites

FRMs and PRMs. General characteristics of erosion for FRMs and PRMs are summarized as follows:

The erosion rates of most FRMs and PRMs are larger than those of the corresponding matrix materials and the erosion rate increases with the increase in fiber or particle volume content. Hence, the erosion resistance decreases with the inclusion of fibers or particles.

In the case where the sizes of reinforcement fibers or particles are much smaller than those of the particles impacted, almost all the fibers or particles removed from a MMC are included in a matrix metal removed by the impact of particles. In this case, the difference in erosion rate is small between the MMC and the corresponding matrix metal. On the other hand, in the case where the reinforcement fibers are included in the form of unidirectionally long fiber or cloths, fiber-rich parts are selectively eroded and a lot of fibers are removed from an eroded surface of the FRM. In such a case, the MMC has much larger erosion rate than the corresponding matrix metal.

The inclusion of brittle fillers in a FRP or a PRM induces shift of the location where the maximum erosion occurs to a higher impact angle compared with the neat matrix metal.

In the case of unidirectional FRM, the erosion resistance of parallel impact, in which the fiber orientation is parallel to the erodent flow direction, tends to be higher than that of perpendicular impact, in which the fiber orientation is perpendicular to the erodent flow direction.

If the bond between fibers or particles and a matrix metal is strengthen, erosion resistance of a FRM or a PRM is enhanced. Heat-treatment may be effective to strengthen the bond between fibers or particles and a matrix metal.

The general characteristics of erosion for FRMs and PRMs mentioned above are common to those of PMCs.

Cermets. General characteristics of erosion for cermets are shown hereafter.

In the cermets, volume fraction of ceramic is so high compared with that of metal that the erosion behavior is mainly dominated by the ceramic phase and a ceramic has much higher erosion resistance than a metal used as binder of cermet. Thus, cermets generally show high erosion resistance.

The erosion resistance of most cermets increases with decreasing binder metal content.

Microstructural factors including binder chemistry, grain boundary structures, and residual stresses created by sintering process which are induced by thermal expansion mismatch in different phases in cermets have an important effect on the erosion behavior of cermets.

Ceramic matrix composites

General characteristics of erosion for CMCs are summarized as follows:

Some CMCs have higher erosion resistance than monolithic ceramics, and some CMCs have lower erosion resistance than monolithic ceramics. If the hardness and fracture toughness increases by the addition of fillers, CMCs tends to have larger erosion resistance than monolithic ceramics. In general, the fracture toughness of a CMC tends to be larger than that of a monolithic ceramic because fillers in a CMC offer resistance to crack propagation. Thus, the hardness would be primarily important for the erosion resistance of a CMC, and the addition of fillers harder than a matrix ceramic would improve the erosion resistance compared with a monolithic ceramic.

The erosion rate yielded by harder erodent such as SiC is larger than that of softer one such as Al₂O₃. This is because the harder erodent has larger fracture strength than the softer one, and the impact force produced by the harder erodent is much larger than that of the softer one.

Erosion-resistant coatings

Most PMCs are usually inferior to the corresponding neat resins in erosion resistance. Therefore, composite materials often need erosion-resistant coatings when they are used in erosive environments. The formation of bondcoat prior to forming erosion protection coatings is effective for avoiding degradation of a PMC substrate because of heat sensitivity of a polymeric material, for improving adhesion to a PMC substrate, and for preventing thermomechanical fatigue failure caused by large differences in coefficient of thermal expansion between the substrate and coating material.

Comments on solid particle erosion studies

Both erosion weight loss and erosion volume loss of a target material have been used to represent the erosion characteristics of a target material. As for erosion rate of a target material, both weight erosion rate and volumetric erosion rate have been used so far. The erosion volume loss and the volumetric erosion rate can be respectively calculated from the erosion weight loss and the weight erosion rate using the density of a target material. Let us consider the comparison between the erosion characteristics of two composites, which have the same erosion weight loss or the same weight erosion rate but different densities. In such a case, the erosion weight loss or the weight erosion rate is not suitable to represent the erosion characteristics of composites, because the erosion volume loss or the volumetric erosion rate of a composite with smaller density are lager than that of a composite with larger density, even though both composites have the same erosion weight loss or weight erosion rate. In other words, the erosion weight loss or the weight erosion rate is not equivalent to the erosion volume loss or the volumetric erosion rate when target materials have the different densities, and it can be concluded that a target material with smaller erosion volume loss or the volumetric erosion rate has better erosion characteristics than that with larger erosion volume loss or the volumetric erosion rate. Although the erosion weight loss or the weight erosion rate was used in a lot of papers on erosion studies of composites, that is why the present author exclusively used the volumetric erosion rate in his papers.^{15,16,20,21,23,109}

It should be noted that reducing the erosion efficiency η of a composite material does not always mean the decrease in erosion rate or improvement of erosion resistance of a composite materials. Removal of reinforcement fibers caused by the damage between the matrix and reinforcement fillers during erosion process has an important effect on the erosion characteristics of composite materials. The erosion efficiency does not take account of such a physical phenomenon. So the erosion efficiency η cannot be always used as an indicator of erosion resistance for composite materials.

Suggestions on future studies

The followings remain for future studies:

Most PMCs have the combination of a soft matrix material and a hard filler material. On the other hand, a combination of a soft matrix material and a hard filler material and its reverse combination are possible in MMCs and CMCs. It has not been clarified how such combinations of a matrix material and a filler material affect the erosion characteristics of MMCs and CMCs.

There has been no solid particle erosion data on newly developed advanced composite materials such as carbon based nano-reinforcements, CNTs, graphenes, and so on.

Simulations studies on solid particle erosion of composites have been not yet enough done.^87, ^119, ^186–188 Further studies will be needed to quantitative simulations.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

References

Reynolds

. On the action of a blast of sand in cutting hard materials. Philos Mag 1873; 46: 337–343.

Rayleigh

JWS

. The sand blast. Nature 1912; 93: 188–188.

Tilly

. Sand erosion of metals and plastics: a brief review. Wear 1969; 14: 241–248.

Finnie

. Some reflections on the past and future of erosion. Wear 1995; 186–187: 1–10.

Tilly

. Erosion caused by airborne particles. Wear 1969; 14: 63–79.

Barkoula

Karger-Kocsis

. Review: processes and influencing parameters of the solid particle erosion of polymers and their composites. J Mater Sci 2002; 37: 3807–3820.

Patnaik

Satapathy

Chand

. Solid particle erosion wear characteristics of fiber and particulate filled polymer composites: a review. Wear 2010; 268: 249–263.

Biswas

Satapathy

Patnaik

. Erosion wear behavior of polymer composites: a review. J Reinf Plast Compos 2010; 29: 2898–2924.

Kaundal

. Role of process variables on the solid particle erosion of polymer composites: a critical review. Silicon 2014; 6: 5–20.

10.

Ruff

Ives

. Measurement of solid particle velocity in erosive wear. Wear 1975; 35: 195–199.

11.

Tilly

Sage

. The interaction of particle and material behavior in erosion processes. Wear 1970; 16: 447–465.

12.

Williams

Jr Lau

. Solid particle erosion of graphite-epoxy composites. Wear 1974; 29: 219–230.

13.

Zahavi

Schmitt

Jr . Solid particle erosion of reinforced composite materials. Wear 1981; 71: 179–190.

14.

Pool

Dharan

CKH

Finnie

. Erosion wear of composite materials. Wear 1986; 107: 1–12.

15.

Miyazaki

Takeda

. Solid particle erosion of fiber reinforced plastics. J Compos Mater 1993; 27: 21–31.

16.

Miyazaki

Hamao

. Effect of interfacial strength on erosion behavior of FRPs. J Compos Mater 1996; 30: 35–50.

17.

Mathias

Goretta

. Solid particle erosion of a graphite-fiber-reinforced bismaleimide polymer composite. Wear 1989; 135: 161–169.

18.

Karasek

Goretta

Helberg

. Erosion in bismaleimide polymers and bismaleimide-polymer composites. J Mater Sci Lett 1992; 11: 1143–1144.

19.

Ballout

Hovis

Talia

. Erosion in glass-fiber reinforced epoxy composite. Scripta Metall Mater 1990; 24: 195–200.

20.

Miyazaki

Hamao

. Solid particle erosion of thermoplastic resins reinforced by short fibers. J Compos Mater 1994; 28: 871–883.

21.

Miyazaki

Funakura

. Solid particle erosion behavior of FRP degraded by hot water. J Compos Mater 1998; 32: 1295–1305.

22.

Miyazaki

. Effect of crystallization of thermoplastic polyimide resin on erosion rate. J Mater Sci Lett 1996; 15: 1422–1423.

23.

Miyazaki

. Solid particle erosion behavior of FRPs with prior impact damage. J Compos Mater 2007; 41: 703–712.

24.

Roy

Vishwanathan

Sundararajan

. The solid particle erosion of polymer matrix composites. Wear 1994; 171: 149–161.

25.

Sundararajan

Roy

Venkataraman

. Erosion efficiency – a new parameter to characterize the dominant erosion micromechanism. Wear 1990; 140: 369–381.

26.

Arjula

Suresh

Harsha

. Study of erosion efficiency of polymers and polymer composites. Polym Test 2006; 25: 188–196.

27.

Saputra

Hasrin

Kubouchi

. Effect of surface treatment on sand erosion behavior of glass beads-filled epoxy resin. Compos Interface 1999; 6: 531–541.

28.

Bijwe

Indumathi

Rajesh

. Friction and wear behavior of polyetherimide composites in various wear modes. Wear 2001; 249: 715–726.

29.

Huang

Yuan

. The effect of reinforcing particles on the erosive wear behavior of particles reinforced silicone matrix composite coating. Wear 2001; 249: 1046–1050.

30.

Barkoula

Karger-Kocsis

. Solid particle erosion of unidirectional GF reinforced EP composites with different fiber/matrix adhesion. J Reinf Plast Compos 2002; 21: 1377–1388.

31.

Barkoula

Karger-Kocsis

. Effects of fibre content and relative fibre-orientation on the solid particle erosion of GF/PP composites. Wear 2002; 252: 80–87.

32.

Barkoula

Papanicolaou

Karger-Kocsis

. Prediction of the residual tensile strengths of carbon-fiber/epoxy laminates with and without interleaves after solid particle erosion. Compos Sci Technol 2002; 62: 121–130.

33.

Tewari

Harsha

Häger

. Solid particle erosion of unidirectional carbon fibre reinforced polyetheretherketone composites. Wear 2002; 252: 992–1000.

34.

Drensky

Hamed

Tabakoff

. Experimental investigation of polymer matrix reinforced composite erosion characteristics. Wear 2011; 270: 146–151.

35.

Pei

Friedrich

. Erosive wear properties of unidirectional carbon fiber reinforced PEEK composites. Tribol Int 2012; 25: 949–963.

36.

Friedrich

Pei

Almajid

. Specific erosive wear rate of neat polymer films and various polymer composites. J Reinf Plast Compos 2013; 32: 631–643.

37.

Tewari

Harsha

Häger

. Solid particle erosion of carbon fibre- and glass fibre-epoxy composites. Compos Sci Technol 2003; 63: 549–557.

38.

Harsha

Tewari

Venkatraman

. Solid particle erosion behavior of various polyaryletherketone composites. Wear 2003; 254: 693–712.

39.

Harsha

Thakre

. Investigation on solid particle erosion behaviour of polyetherimide and its composites. Wear 2007; 262: 807–818.

40.

Arjula

Suresh

Harsha

Ghosh

. Erosive wear of unidirectional carbon fibre reinforced polyetherimide composite. Mater Lett 2008; 6: 3246–3249.

41.

Arjula

Suresh

Harsha

Ghosh

. Solid particle erosion of unidirectional fibre reinforced thermoplastic composites. Wear 2009; 267: 1516–1524.

42.

Sari

Sinmazcelik

Yilmaz

. Erosive wear studies of glass fiber- and carbon fiber-reinforced polyetheretherketone composites at low particle speed. J Thermoplast Compos Mater 2011; 24: 333–350.

43.

Arjula

Suresh

Harsha

Ghosh

. Solid particle erosion studies on polyphenylene sulfide composites and prediction on erosion data using artificial neural networks. Wear 2009; 266: 184–193.

44.

Zhang

Barkoula

Karger-Kocsis

. Artificial neural network predictions on erosive wear of polymers. Wear 2003; 255: 708–713.

45.

Arjula

Suresh

Harsha

Ghosh

. Erosion studies of short glass fiber-reinforced thermoplastic composites and prediction of erosion rate using ANNs. J Reinf Plast Compos 2010; 29: 1641–1652.

46.

Rattan

Bijwe

. Carbon fabric reinforced polyetherimide composites: influence of weave of fabric and processing parameters on performance properties and erosive wear. Mater Sci Eng A 2006; 420: 342–350.

47.

Rattan

Bijw

. Influence of impingement angle on solid particle erosion of carbon fabric reinforced polyetherimide composite. Wear 2007; 262: 568–574.

48.

Bijwe

Rattan

Fahim

. Erosive wear of carbon fabric reinforced polyetherimide composites: role of amount of fabric and processing technique. Polym Compos 2008; 29: 337–344.

49.

Sari

Sinmazcelik

. Erosive wear behaviour of carbon fibre/polyetherimide composites under low particle speed. Mater Des 2007; 28: 351–355.

50.

Sinmazcelik

Taskiran

. Erosive wear behaviour of polyphenylenesulphide (PPS) composites. Mater Des 2007; 28: 2471–2477.

51.

Sari

. Influence of erodent particle types on solid particle erosion of polyphenylene sulphide composite under low particle speed. Polym Compos 2009; 30: 1442–1449.

52.

Sinmazcelik

Sari

. Erodent size effect on the erosion of polyphenylene sulfide composite. Polym Compos 2010; 31: 985–994.

53.

Sinmazcelik

Fidan

Gunnay

. Residual mechanical properties of carbon/polyphenylenesulphide composites after solid particle erosion. Mater Des 2008; 29: 1419–1426.

54.

Yilmaz

. Influence of annealing duration on the erosive wear behavior of polyphenylenesulphide composites. J Mater Sci 2010; 45: 2381–2389.

55.

Ahmed

Nino

Bersee

HEN

. Improving erosion resistance of polymer reinforced composites. J Thermoplast Compos Mater 2009; 22: 703–725.

56.

Avcu

Fiden

Bora

. Solid particle erosive wear behavior of glass mat reinforced PPS composites: influence of erodent particle size, pressure, particle impingement angle, and velocity. Adv Polym Technol 2013; 32: E387–E398.

57.

Tsuda

Kubouchi

Sakai

. General method for predicting the sand erosion rate of GFRP. Wear 2006; 260: 1045–1052.

58.

Kim

. Solid particle erosion of CFRP composite with different laminate orientations. Wear 2009; 267: 1922–1926.

59.

Biswas

Deo

Patnaik

. Effect of fiber loading and orientation on mechanical and erosion wear behaviors of glass–epoxy composites. Polym Compos 2011; 32: 665–674.

60.

Mahapatra

Patnaik

Satapathy

. Taguchi method applied to parametric appraisal of erosion behavior of GF-reinforced polyester composites. Wear 2008; 265: 214–222.

61.

Patnaik

Satapathy

Mahapatra

. A modeling approach for prediction of erosion behavior of glass fiber-polyester composites. J Polym Res 2008; 15: 147–160.

62.

Patnaik

Satapathy

Mahapatra

. A Taguchi approach for investigation of erosion of glass fiber-polyester composites. J Reinf Plast Compos 2008; 27: 871–888.

63.

Qian

Bao

Takatera

. Fiber-reinforced polymer composite materials with high specific strength and excellent solid particle erosion resistance. Wear 2010; 268: 637–642.

64.

Kumar

Satapathy

Patnaik

. Viscoelastic interpretations of erosion performance of short aramid fibre reinforced vinyl ester resin composites. J Mater Sci 2011; 46: 7489–7500.

65.

Tejyan

Patnaik

Rawal

. Structural and mechanical properties of needle-punched nonwoven reinforced composites in erosive environment. J Appl Polym Sci 2012; 123: 1698–1707.

66.

Patnaik

Tejyan

. A Taguchi approach for investigation of solid particle erosion response of needle-punched nonwoven reinforced polymer composites: part II. J Ind Text 2014; 43: 458–480.

67.

Kumar

Satapathy

Patnaik

. Thermo-mechanical correlations to erosion performance of short carbon fibre reinforced vinyl ester resin composites. Mater Des 2011; 32: 2260–2268.

68.

Deo

Chittaranjan

Acharya

. Solid particle erosion of lantana camara fiber-reinforced polymer matrix composite. Polym-Plast Technol Eng 2009; 48: 1084–1087.

69.

Mishra

Acharya

. Solid particle erosion of bagasse fiber reinforced epoxy composite. Int J Phys Sci 2010; 5: Article No. 5A0E50822834–Article No. 5A0E50822834.

70.

Gupta

Kumar

Patnaik

. Effect of different parameters on mechanical and erosion wear behavior of bamboo fiber reinforced epoxy composites. Int J Polym Sci 2011; 2011: Article ID 592906–Article ID 592906.

71.

Satapathy

Patnaik

Pradhan

. A study on processing, characterization and erosion behavior of fish (labeo-rohita) scale filled epoxy matrix composites. Mater Des 2009; 30: 2359–2371.

72.

Qian

Bao

. Improvement of erosion behavior and mechanical strength of UP resin composites filled with VGCF. Appl Mech Mater 2011; 43: 712–716.

73.

Zhou

Zhang

Pei

. Erosive wear of transparent nanocomposite coatings. Tribol Int 2013; 61: 62–69.

74.

Zhang

Yang

Shen

. Particle erosion on carbon nanofiber paper coated carbon fiber/epoxy composites. Compos B 2013; 54: 209–214.

75.

Zhang

Yang

Guerra

. Enhancing particle erosion resistance of glass-reinforced polymeric composites using carbon nanofiber-based nanopaper coatings. J Appl Polym Sci 2013; 129: 1875–1881.

76.

Chen

Hutchings

Deng

. The effect of carbon nanotube orientation on erosive wear resistance of CNT-epoxy based composites. Carbon 2014; 73: 421–431.

77.

Srivastava

Pawar

. Solid particle erosion of glass fibre reinforced flyash filled epoxy resin composites. Compos Sci Technol 2006; 66: 3021–3028.

78.

Patnaik

Satapathy

Mahapatra

. Modeling and prediction of erosion response of glass reinforced polyester-flyash composites. J Reinf Plast Compos 2009; 28: 513–536.

79.

Bishoyee

Dash

Mishra

. A grey-based Taguchi approach for characterization of erosive wear phenomenon of glass-polyester fly ash filled composites. J Polym Environ 2010; 18: 177–187.

80.

Patnaik

Satapathy

Mahapatra

. Parametric optimization erosion wear of polyester-GF-alumina hybrid composites using the Taguchi method. J Reinf Plast Compos 2008; 27: 1039–1058.

81.

Patnaik

Satapathy

Mahapatra

. Implementation of Taguchi design for erosion of fiber-reinforced polyester composite systems with SiC filler. J Reinf Plast Compos 2008; 27: 1093–1111.

82.

Patnaik

Satapathy

Mahapatra

. Study on erosion response of hybrid composites using Taguchi experimental design. J Eng Mater Technol 2009; 131. : Paper No. 031011.

83.

Mahapatra

Patnaik

. Study on mechanical and erosion wear behavior of hybrid composites using Taguchi experimental design. Mater Des 2009; 30: 2791–2801.

84.

Yilmaz

Unal

Mimaroglu

. Study of the strength and erosive behavior of CaCO₃/glass fiber reinforced polyester composite. eXPRESS Polym Lett 2008; 2: 890–895.

85.

Karsli

Yilmaz

Aytac

. Investigation of erosive wear behavior and physical properties of SGF and/or calcite reinforced ABS/PA6 composites. Compos B 2013; 44: 385–393.

86.

Biswas

Satapathy

. Use of copper slag in glass-epoxy composites for improved wear resistance. Waste Manage Res 2010; 28: 615–625.

87.

Patnaik

Kaundal

Satapathy

. Solid particle erosion of particulate filled short glass fiber reinforced polyester resin composites. Adv Mater Res 2010; 123–125: 213–216.

88.

Mahapatra

. Modelling and analysis of erosion wear behaviour of hybrid composites using Taguchi experimental design. Proc Inst Mech Eng J, J Eng Tribol 2010; 224: 157–167.

89.

Biswas

Satapathy

. A study on tribological behavior of alumina-filled glass-epoxy composites using Taguchi experimental design. Tribol Trans 2010; 53: 520–532.

90.

Jha

Mantry

Satapathy

. Erosive wear performance analysis of jute-epoxy-SiC hybrid composites. J Compos Mater 2010; 44: 1623–1641.

91.

Bagci

Imrek

. Solid particle erosion behaviour of glass fibre reinforced boric acid filled epoxy resin composites. Tribol Int 2011; 44: 1704–1710.

92.

Bagci

Imrek

. Application of Taguchi method on optimization of testing parameters for erosion of glass fiber reinforced epoxy composite materials. Mater Des 2013; 46: 706–712.

93.

Kumar

Satapathy

Patnaik

. Thermo-mechanical correlations to erosion performance of short glass/carbon fiber reinforced vinyl ester resin hybrid composites. Comput Mater Sci 2012; 60: 250–260.

94.

Rout

Satapathy

Mantry

. Erosion wear performance analysis of polyester-GF-granite hybrid composites using Taguchi method. Procedia Eng 2012; 38: 1863–1882.

95.

Ismail

Suresha

Ananthaiah

. Parametric appraisal of erosion behaviour of organo-modified montmorillonite-filled epoxy hybrid composites by Taguchi approach. J Reinf Plast Compos 2013; 32: 1586–1602.

96.

Mohan

Mahesha

Rajaprakash

. Erosive wear behaviour of WC filled glass epoxy composites. Procedia Eng 2013; 68: 694–702.

97.

Padhi

Satapathy

. Solid particle erosion behavior of BFS-filled epoxy–SGF composites using Taguchi’s experimental design and ANN. Tribol Trans 2014; 57: 396–407.

98.

Panda

Mantry

Mohapatra

. A study on erosive wear analysis of glass fiber-epoxy-AlN hybrid composites. J Compos Mater 2014; 48: 107–118.

99.

Aktas

. Solid particle erosion of glass epoxy composite strengthened by metal powder. J Thermoplast Compos Mater 2012; 25: 949–963.

100.

Mohan

Natarajan

Kumaresh Babu

. Solid particle erosion of UHMWPE filled aramid fabric-epoxy hybrid composites. Adv Mater Res 2010; 123–125: 1051–1054.

101.

Mohan

Natarajan

Kumaresh Babu

. Studies on erosive wear behavior of UHMWPE-filled aramid-epoxy hybrid composites. Mater Manuf Process 2012; 27: 430–435.

102.

Srinivasan

Sreeran

Scattergood

Warren

. Erosion of fiber reinforced Al-4pc Cu composites. Metal Trans A 1988; 19A: 1785–1793.

103.

Morita

Kawashima

Seguchi

. Erosion of ceramic particle products for pipe transportation. Tetsu-to-Hagane 1989; 75: 297–304.

104.

Hovis

Talia

Scattergood

. Erosion in multiphase systems. Wear 1986; 108: 139–155.

105.

Morrison

Routbort

Scattergood

. Erosion of an alumina-stainless steel composite. Wear 1993; 160: 345–350.

106.

Goretta

Routbort

. Erosion of 2014 Al reinforced with SiC or Al₂O₃ particles. Mater Sci Eng 1992; A151: 85–95.

107.

Saravanan

Surappa

Pramila Bai

. Erosion of A356 Al-SiC_p composites due to multiple particle impact. Wear 1997; 202: 154–164.

108.

Alahelisten

Bergman

Olsson

. On the wear of aluminium and magnesium metal matrix composites. Wear 1993; 165: 221–226.

109.

Miyazaki

Funakura

. Solid particle erosion behavior of metal matrix composites. J Compos Mater 1996; 30: 1670–1682.

110.

Pan

Matsumura

. The solid particle erosion behavior of Al₁₈B₄O₃₃ whisker-reinforced AC4C Al alloy matrix composites. Wear 1998; 223: 22–30.

111.

Matsumura

. Interfacial reaction effects on erosion of aluminium matrix composites. Scr Mater 1999; 40: 645–656.

112.

Levin

DuPont

Marder

. The effect of second phase volume fraction on the erosion resistance of metal-matrix composites. Wear 2000; 238: 160–167.

113.

Gui

Wang

. Erosion of in-situ TiC particle reinforced Al–5Cu composite. Mater Res Bull 2001; 36: 1573–1585.

114.

Das

Mondal

Sawla

. Solid particle erosion of Al alloy and Al-alloy composites: effect of heat treatment and angle of impingement. Metal Mater Trans A 2004; 35A: 1369–1379.

115.

Kumar

Subramanya Sarma

Murty

. A statistical analysis on erosion wear behaviour of A356 alloy reinforced with in situ formed TiB₂ particles. Mater Sci Eng A 2008; 476: 333–340.

116.

Mitra

Chiou

Fine

. Interface in as-extruded XD Al/TiC and Al/TiB₂ metal matrix composites. J Mater Res 1993; 8: 2380–2392.

117.

Acharya

Dikshit

Mishra

. Erosive wear behaviour of redmud filled metal matrix composite. J Reinf Plast Compos 2008; 27: 145–152.

118.

MacMillin

Roll

Funkenbusch

. Erosion and surface structure development of metal-diamond particulate composites. Wear 2010; 269: 875–883.

119.

Mamatha

Patnaik

Biswas

. Finite element modelling and development of SiC-filled ZA-27 alloy composites in erosive wear environment: a comparative analysis. Proc Inst Mech Eng J, J Eng Tribol 2011; 225: 1106–1120.

120.

Paul

Mishra

Tiwari

. Solid-particle erosion behaviour of WC/Ni composite clad layers with different contents of WC particles. Opt Laser Technol 2013; 50: 155–162.

121.

Ninham

Levy

. The erosion of carbide-metal composites. Wear 1988; 121: 347–361.

122.

Anand

Conrad

. Local impact damage and erosion mechanisms in WC-6wt.%Co alloys. Mater Sci Eng 1988; A105/106: 411–421.

123.

Wang

Luer

. The erosion-oxidation behavior of HVOF Cr₃C₂-NiCr cermet coating. Wear 1994; 174: 177–185.

124.

Wang

Verstak

. Elevated temperature erosion of HVOF Cr₂C₃/TiC-NiCrMo cermet. Wear 1999; 233–235: 342–351.

125.

Wang

Shui

. The hot erosion behavior of HVOF chromium carbide-metal cermet coatings sprayed with different powders. Wear 2002; 253: 550–557.

126.

Wang

Shui

. Hot erosion behavior of carbide-metal composite coatings. J Mater Process Technol 2003; 143–144: 87–92.

127.

Reshetnyak

Kuybarsepp

. Mechanical properties of hard metals and their erosive wear resistance. Wear 1994; 177: 185–193.

128.

D’Errico

Bugliosi

Cuppiui

. A study of cermets’ wear behavior. Wear 1997; 203–204: 242–246.

129.

Hussainova

Kubarsepp

Pirso

. Mechanical properties and features of erosion of cermets. Wear 2001; 250: 818–825.

130.

Hussainova

. Some aspects of solid particle erosion of cermets. Tribol Int 2001; 34: 89–93.

131.

Hussainova

. Effect of microsturucture on the erosive wear of titanium carbide-based cermets. Wear 2003; 255: 121–128.

132.

Hussainova

Kolesnikova

Hussainov

. Effect of thermo-elastic residual stresses on erosive performance of cermets with core-rim structured ceramic grains. Wear 2009; 267: 177–185.

133.

Antonov

Hussainova

Pirso

. Assessment of mechanically mixed layer developed during high temperature erosion of cermets. Wear 2007; 263: 878–886.

134.

Hussainova

Pirso

Antonov

. Erosion and abrasion of chromium carbide based cermets produced by different methods. Wear 2007; 263: 905–911.

135.

Hussainova

Schade

. Correlation between solid particle erosion of cermets and particle impact dynamics. Tribol Int 2008; 41: 323–330.

136.

Hussainova

Jasiuk

Sardela

. Micromechanical properties and erosive wear performance of chromium carbide based cermets. Wear 2009; 267: 152–159.

137.

Hussainova

Antonov

Zikin

. Erosive wear of advanced composites based on WC. Tribol Int 2012; 46: 254–260.

138.

Alman

Tylczak

Hawk

. An assessment of the erosion resistance of iron-aluminide cermets at room and elevated temperatures. Mater Sci Eng A 2002; 329–331: 602–609.

139.

Rateick

Jr Karasek

Cunningham

. Solid-particle erosion of tungsten carbide/cobalt cermet and hardened 440C stainless steel – a comparison. Wear 2006; 261: 773–778.

140.

Lin

Liu

Chen

. Erosion resistance of TiC/Fe composite at high temperatures. Adv Mater Res 2009; 79–82: 1087–1090.

141.

Sun

Wang

. Mechanical properties and erosion resistance of ceria nano-particle-doped ultrafine WC-12Co composite prepared by spark plasma sintering. Wear 2013; 301: 406–414.

142.

Honeycutt

III Jennings

Manning

Jr . High temperature erosion of tantalum carbide composites. Wear 1976; 37: 209–216.

143.

Sykes

Scattergood

Routbort

. Erosion of SiC–reinforced alumina ceramic composites. Composites 1987; 18: 153–163.

144.

Wada

Watanabe

Tani

. Solid particle erosion of brittle materials (Part 6) –the erosive wear of Al₂O₃-SiC composites–. J Ceram Soc Jpn 1988; 96: 111–118.

145.

Kamiya

Sakakibara

Sakurai

. Erosion wear properties of tetragonal ZrO₂(Y₂O₃)-toughened Al₂O₃ composites. J Am Ceram Soc 1994; 77: 666–672.

146.

Liu

Lin

. Erosion wear in Al₂O₃-Cr₃C₂ composites. Ceram Int 1996; 22: 415–420–415–420.

147.

Gochnour

Bright

Shetty

. Solid particle erosion of SiC-Al₂OC ceramics. J Mater Sci Lett 1990; 25: 3229–3235.

148.

Kim

Park

. Solid-particle erosion of hot-pressed silicon carbide and SiC-TiB₂ composite. J Mater Sci Lett 1997; 16: 821–823.

149.

Kim

Park

. Solid particle erosion of SiC and SiC-TiB₂ composite hot-pressed with Y₂O₃. Wear 1998; 222: 114–119.

150.

Colclough

Yoemans

. Hard particle erosion of silicon carbide and silicon carbide-titanium deboride from room temperature to 1000℃. Wear 1997; 209: 229–236.

151.

Sharma

Manoj Kumar

. Erosion behavior of SiC-WC composites. Ceram Int 2014; l40: 6829–6839.

152.

Wang

Levy

. Erosion behavior of SiC fiber-SiC matrix composites. Wear 1990; 138: 125–136.

153.

Suh

Kohyama

. Effect of porosity on particle erosion wear behavior of lab. scale SiC_F/SiC composites. Int J Mod Phys B 2010; 24: 2934–2939.

154.

Suh

Hinoki

Kohyama

. Erosive wear mechanism of new SiC/SiC composites by solid particles. Tribol Lett 2011; 41: 503–513.

155.

Park

Han

Lim

. A study on erosion wear and erosion of sialon-Si₃N₄ whisker ceramic composites. Wear 1997; 203–204: 284–290.

156.

Liu

Lin

Lee

. Erosive wear behaviour in duophase SiAlON composites. Ceram Int 1998; 24: 217–221.

157.

Ham

Yoemans

Watts

. Elevated temperature solid particle erosion of silicon carbide continuous fibre reinforced calcium aluminosilicate glass-ceramic matrix composite. Wear 1997; 203–204: 387–392.

158.

Ham

Yeomans

Watts

. Effect of temperature and particle velocity on the erosion of a silicon carbide continuous fibre reinforced calcium aluminosilicate. Wear 1999; 233–235: 237–245.

159.

Alman

Tylczak

Hawk

. Solid particle erosion behavior of an Si₃N₄-MoSi₂ composite at room and elevated temperatures. Mater Sci Eng A 1999; 261: 245–251.

160.

Iizuka

Kita

Hirai

. Sliding wear and solid particle erosion behaviors of Mo₅Si₃-Si₃N₄ composites. J Ceram Soc Jpn 2003; 111: 919–923.

161.

Sato

Kamiya

. Erosion wear properties of TiC-Al₂O₃ composites. J Ceram Soc Jpn 2003; 111: 485–490.

162.

Sato

Tsukada

Kamiya

. Erosion wear properties of pressureless sintered TiC/Al₂O₃ composites –comparison between pressureless sintered commercial cutting tool materials–. J Ceram Soc Jpn 2004; 112: 117–120.

163.

Saewong

Rawlings

. Erosion of carbon fiber reinforced glass matrix composites. J Mater Sci Lett 1999; 18: 1915–1919.

164.

Yamada

Sakaguchi

Hirao

. Effect of microstructure on mechanical properties and erosion wear of B₄C and B₄C-CrB₂ composites. J Ceram Soc Jpn 2003; 111: 53–58.

165.

Liu

Sun

. Erosion behaviour of B₄C-based ceramic composites. Ceram Int 2010; 36: 1297–1302.

166.

Ouyang

Zhu

. Experimental and simulation studies on the solid-particle erosion of WC-MgO composites. Tribol Lett 2013; 52: 501–510.

167.

Amirthan

Udayakumar

Bhanu Prasad

. Solid particle erosion studies on biomorphic Si/SiC ceramic composites. Wear 2010; 268: 145–152.

168.

Chen

Huang

. Effect of silicon addition on high-temperature solid particle erosion-wear behaviour of mullite-SiC composite refractories prepared by nitriding reactive. Adv Mater Sci Eng 2014; 2014: Article ID 943542.

169.

Baxter

Rawlings

. Sold particle erosion of highly porous carbon-carbon composites. Mater Sci Technol 1998; 14: 161–169.

170.

Baxter

Rawlings

Iwashita

. Effect of chemical vapor infiltration on erosion and thermal properties of porous carbon/carbon composite thermal insulation. Carbon 2000; 38: 441–449.

171.

Sarkar

Sekharan

Mitra

. The solid particle erosion behavior of carbon/carbon and carbon/phenolic composite used in re-entry vehicles. Tribol Trans 2009; 52: 777–787.

172.

Alonso

Fagoaga

Oregui

. Erosion protection of carbon-epoxy composites by plasma-sprayed coatings. Surf Coat Technol 1991; 49: 482–488.

173.

Sutter

Miyoshi

Bowman

. Erosion coatings for polymer matrix composites in propulsion applications. High Perform Polym 2003; 15: 421–440.

174.

Miyoshi

Sutter

Horan

. Assessment of erosion resistance of coated polymer matrix composites for propulsion applications. Tribol Lett 2004; 17: 377–387.

175.

Ivosevic

Knight

Kalidindi

. Solid particle erosion resistance of thermally sprayed functionally graded coatings for polymer matrix composites. Surf Coat Technol 2006; 200: 5145–5151.

176.

Liu

Guo

Zhao

. Arc sprayed erosion-resistant coating for carbon fiber reinforced polymer matrix composite substrates. Surf Coat Technol 2006; 200: 3073–3077.

177.

Maurer

Schulz

. Erosion resistant titanium based PVD coatings on CFRP. Wear 2013; 302: 937–945.

178.

Keegan

Nash

Stack

. On erosion issues associated with the leading edge of wind turbine blades. J Phys D, Appl Phys 2013; 46: Paper No. 383001.

179.

Baxter

Rawlings

. Microstructure and solid particle erosion of carbon based materials used for the protection of highly porous carbon-carbon composite thermal insulation. J Mater Sci 1997; 32: 4485–4492.

180.

Smeacetto

Salvo

Ferraris

. Erosion protective coatings for low density, highly porous carbon/carbon composites. Carbon 2009; 47: 1511–1519.

181.

Yan

Gross

Simon

. Mechanical and erosion properties of CaCO₃-EMAA thermal spray coatings. Polym Eng Sci 2004; 44: 1448–1459.

182.

Liu

Zhao

. Erosion behavior of TiC-xNi composite coatings in-situ synthesized by electro-thermal explosion ultra-high speed spraying. Adv Mater Res 2008; 33–37: 1277–1282.

183.

Sharma

Dwivedi

Jain

. Study and model development of erosive wear by RSM. Proc Inst Mech Eng J, J Eng Tribol 2012; 226: 57–70.

184.

Keller

Hampton

McLaury

. Self-healing of erosion damage in a polymer coating. Wear 2013; 307: 218–225.

185.

Pasha

Kaleemulla

. Erosive wear behavior of sprayed metal matrix composites: an overview. Proc Inst Mech Eng J, J Eng Tribol 2013; 227: 1063–1075.

186.

Chen

. Computer simulation of solid-particle erosion composite materials. Wear 2003; 255: 78–84.

187.

Chen

. Is porosity always detrimental to the wear resistance of materials? – A computational study on the effect of porosity on erosive wear of TiC/Cu composites. Wear 2009; 267: 1153–1159.

188.

Chen

Cook

. A further simulation study on the dual role of porosity in solid-particle erosion of materials. Wear 2011; 271: 1325–1330.