Erosion wear of centrifugal pumps widely exists in the transportation, aerospace, chemical, mining, environmental protection, and other industries. Erosion wear generally refers to the hard sand with a certain kinetic energy that repeatedly impacts and cuts the surface of flow-through parts of the hydraulic machinery and results in mass loss, which is a very important cause of equipment failure or damage. The erosion wear results in the decrease of centrifugal pump efficiency, which will cause direct economic loss and energy waste.

Many scholars have studied erosion wear caused by solid–liquid two-phase flow. Tarodiya et al. (Tarodiya and Gandhi, 2019) carried out experiments on two kinds of sand mud of equal size at two pump speeds and two flows and examined the SCANNING electron microscopy (SEM) images of wear samples, and the results showed that the wear at the volute tongue was caused by both cutting and deformation. Bozzini et al. (2003) simulated the particle’s motion trajectory and collision velocity in the pipeline and deduced the erosion model parameters based on the experiment. Azimian et al. (Azimian and Bart, 2015) conducted experiments on the centrifugal rotary wear equipment and found that the size of erosion pits increases with the increase in sediment concentration. Walker et al. (Walker and Bodkin, 2000) summarized the influence of solid particle size, solid–liquid two-phase fluid concentration, and the rotation speed of the pump on wear and explained the wear rule. Amaro et al. (2020) studied the effect of different impact angles, impact distances, and particle sizes on the impact fatigue life. Peng et al. (Guangjie et al., 2013) studied the relationship between blade surface wear rate and sediment concentration and obtained the distribution rule of the blade surface wear rate. Islam et al. (Islam and Farhat, 2014) carried out an erosion test on APL-X42 steel to study its erosion wear performance under different erosion angles. Feng et al. (2020) tested and analyzed WC-CO cemented carbide and found that the anti-erosion performance of cemented carbide decreased with the increase in WC grain size and Co content. Liu et al. (2004) studied the erosion wear of the centrifugal mud pump and the abrasion resistance of a new kind of anti-erosive wear material (Al₂O₃) through experiments. The results showed that wear of the pressure surface is more serious than that of the suction surface. Compared with cast iron (FC20) and AISI 1316L stainless steel, Al₂O₃ had excellent anti-wear performance. Walker et al. (Walker and Robbie, 2013) measured the wear resistance of natural rubber, eutectic, and hypereutectic white cast iron under the conditions of erosion wear and compared the results with the wear of the materials in a centrifugal slurry pump in a factory. In laboratory experiments, the wear rate of hypereutectic white cast iron is higher than that of eutectic white cast iron, and the wear rate of natural rubber is much lower than that of eutectic white cast iron. The aforementioned laboratory results are the opposite of the field trial results. Walker believed that the aforementioned difference is largely due to particle impact energy or impingement angle and particle shape or size. Xiong et al. (2013) carried out experimental studies on erosion wear of Ti(C, N)-Mo2C-Co cermet in artificial seawater containing SiO₂ solid particles. The results showed that the adhesive phase began to wear at first and gradually expanded and fractured. Nguyen (Nguyen et al., 2014) conducted slurry erosion tests on SUS-304 stainless steel. The results showed that the corrosion rate is high at the beginning and gradually decreases as the test time goes on. Arabnejad et al. (2015) studied the impact of particle hardness on the erosion of stainless steel at low impact velocity and found that the erosion rate increased with the increase of particle hardness. Shah and Jain (Shah and Jain, 2008) used computational fluid dynamics (CFD) and studied the relationship between erosion rate and flow velocity, slurry concentration, proppant (solid particle) size, and density. The results showed that the slurry flow rate was the most important factor affecting erosion, and the slurry concentration and fluid viscosity had a greater influence on the erosion rate. Shen et al. (2016) used computational fluid dynamics (CFD) to study the wear of the screw centrifugal pump when the transmission medium is solid–liquid two-phase flow with large-size particles. The research results showed that the volume fraction of the solid phase has a greater impact on the trajectory when transporting large particles, while for small particles, it has little effect. Prasanna et al. (2018) successfully sprayed Stellite-6, 10%Al₂O₃+90%CoCrTaY, and 25%Cr₃C₂-20(Ni-Cr)+75%NiCrAlY on three kinds of turbine alloys: Ti-6Al-4V, Co-based superalloy (Super Co 605), and Fe-based special steel (MDN121). The microstructure and mechanical properties of the coating were characterized. The research results showed that the relative corrosion resistance of various coatings can be arranged in the following order: Stellite-6> Cr₃C₂-20(Ni-Cr)+75%NiCrAlY > Al₂O₃+90%CoCrTaY. Iwai et al. (Iwai and Nambu, 1997) conducted tests on 13 kinds of materials, including polymers, metals, and ceramics, and found that among these materials, rubber had the best wear resistance and polyurethane had the worst. Bansal et al. (2020) used different influencing factors, namely, average erodent particle size, slurry concentration, impact angle, and impact velocity to test the anti-erosion performance of coated and uncoated SS410 steel and analyzed the effect of the thickness of the hydrophobic coating on the anti-erosion performance. The results showed that the PTFE coating with the smallest thickness had the lowest corrosion rate compared with similar coatings. Medvedovski et al. (Medvedovski and Antonov, 2020) tested the iron boride coatings and carbon steel commonly used in industry and observed the erosion surface using a scanning electron microscope. All test results showed that the performance of iron boride coatings is better than that of carbon steel. The erosion mechanism of carbon steel is determined by plastic deformation and ploughing, and the erosion mechanism of iron boride coatings is related to microcracks such as in many advanced ceramics. Singh et al. (2019) studied the erosion performance of ceramic-reinforced WC–10Co4Cr cermet coatings under different parameters. The microstructure analysis showed that there were interlocked lamellae splats, pores, and unmelted particles on the sprayed surface, and it has good anti-erosion performance in a weakly acidic environment. Grant et al. (Grant and Tabakoff, 1975) verified the erosion wear prediction model that was established on the basis of considering the aerodynamic resistance of solid particles, rebound dynamics of wall surface, and material removal process. The predicted erosion wear amount and erosion wear location of blades are almost consistent with those of the tested. Dong et al. (Xing et al., 2009) analyzed the position and process of pump wear through simulation and experiment and found that the impact of pulverized coal with a certain diameter, shape, impact velocity, and invasion angle on the flowing parts would produce pits on the surface of the flowing parts. Lai et al. (2019) combined the rebound model and erosion model proposed by Grant and Tabakoff to study the erosion wear problem of solid particles in a solid–liquid two-phase centrifugal pump. The results showed that the maximum erosion rate occurred at the impeller hub, and the most serious erosion area of the blade was located at the leading edge between 55° and 60° curvature angle. Liu et al. (Liu and Zhu, 2011) analyzed the wear characteristics of the impeller of the centrifugal sewage pump and proposed measures to reduce the wear of the impeller. The research results can provide theoretical reference for the optimal design of the centrifugal sewage pump. Ojala et al. (2016) revealed the behavior differences of different materials under mud erosion conditions. The wear surface analysis shows that relative wear properties of steel increase with the increase of abrasive size, while the relative wear properties of elastomer decrease with the increase of abrasive size. Marshall et al. (Marshall, 2009) reviewed the forces and torques applied to particles by fluid motion, particle–particle collision, and van der Waals forces. The effects of different parameters on the normal collision and adhesion of two particles are studied in detail. Ovchinnikov et al. (2019) proposed a calculation method for the maximum metal consumption loss of the impeller of a single-stage centrifugal pump, and the results show that this method can effectively reduce the residual deformation risk of the pump shaft. Oka et al. (2005) proposed an equation that can predict the impact wear of solid particles, which is applicable to any impact condition and any type of material. Through erosion experiments on several materials, the results show that the erosion damage is related to the initial hardness, plastic deformation characteristics, or brittleness of materials. Boon et al. (Ng et al., 2008) conducted a numerical simulation on dense solid–liquid two-phase flow and found that particle friction is an important factor in establishing the constitutive equation of dense particle flow. Aslfattahi et al. (2020) used nanomaterials (MXene) with the chemical formula Ti3C2 together with silicone oil for the first time to improve the thermophysical properties of Mxene-based silicone oil. Das et al. (2021) and Rubbi et al. (2021) reviewed the applications of ionic liquids and ionic fluids in different solar systems and the latest progress of nanoparticles in solar systems. Samylingam et al. (2020) synthesized nanomaterials (MXene) with the chemical formula Ti3C2 by the wet chemistry method, suspended its suspension in pure oil palm oil (OPO), and prepared a new type of heat transfer fluid. Kadirgama et al. (2021) used nanomaterials (MXene) and ethylene glycol (EG) as the base liquid to prepare a homogeneous mixture of nanofluids with good optical properties, which might enable high optical performance applications in the solar system.

Most of the abovementioned research studies are based on the condition of large particles and low concentration, but a few research studies on the wear of high-concentration fine particles on the centrifugal pump are carried out experimentally. However, the collision frequency of dense fine particles with a wall surface is much higher than that of large particles, resulting in different wear characteristics. In order to enrich the wear research of centrifugal pumps under the condition of dense fine particles, the wear tests of the solid–liquid two-phase centrifugal pump impeller under different high mass concentrations, fine particle diameters, and impeller materials were carried out in this study.

Figure 1 shows the schematic diagram of the wear experimental system, while parameters of the main components related to the experimental system are shown in Table 1. The pressure sensor and electromagnetic flow sensor are used to monitor the centrifugal pump operation during the test. The running time is 72 h per test, and the rotation speed r = 1450 rpm. The valve is adjusted to ensure that the inlet flow reaches the rated flow of 16 m3/h. The impellers of the tested pump were made of three materials, namely, carbon structural steel (Q235), gray cast iron (HT200), and low-alloy high-strength steel (16 Mn). The major properties of the three materials are shown in Table 2. Considering economy, universality, and no pollution, SiO₂ particles were chosen as solid components with 2650 kg/m³ density and 475 MPa Brinell hardness, and the particle diameter ranges are 0.125–0.212 mm, 0.212–0.425 mm, and 0.425–0.710 mm, as shown in Figure 2.

An ultrasonic thickness gage (Figure 3A) was used to measure the thickness of the blade. The wear loss of the blade is expressed by the variation of blade thickness at the measuring position before and after the wear experiment, namely, Δ= T 1 − T 2 , where Δ is the thickness variation; T 1 is the thickness of the blade at the measuring position before the experiment; T 2 is the thickness of the blade at the measuring position after the experiment. For every measuring position, measurements are taken three times and their average value is calculated. The measurement uncertainty of the thickness is estimated as ±0.27%. A digital microscope (Figure 3B) was used to observe the wear morphology of the blade surface. The observation sites were the head and tail of the pressure surface of impellers of different materials.

As shown in Figure 4, the blades of the experiment impeller (Figure 4A) are numbered 1 to 5, the seventeen measuring points are evenly arranged on the pressure surface of each blade, and the longitudinal distance of each measuring point is 5 mm; the transverse distance is 10 mm (Figure 4B).

Many scholars (Walker and Bodkin, 2000; Liu et al., 2004; Guangjie et al., 2013) have found that the wear of the impeller pressure surface is much more serious than that of the suction surface. This is because the density of particles is higher than that of water, and they tend to gather at the pressure surface end under the action of centrifugal force. Therefore, this work mainly analyzes the wear of impeller pressure surface by measuring point data, wear morphology, and electron microscope observation.

In this study, the erosion wear of centrifugal pump blades was studied by experiment and the following conclusions were obtained:

(1) The wear of solid particles on the blade is mainly concentrated in the middle and trailing edge of the blade. In the axial direction of the impeller, the wear on the top of the pressure surface is more serious.