Introduction

Solid particle erosion (SPE) and water droplet erosion (WDE) represent two predominant degradation mechanisms threatening the integrity of turbine blades. SPE arises from high-velocity impingement of micro-particles (e.g., oxide spallations from boiler tubes) causing surface pitting and roughness escalation. WDE initiates from repetitive droplet impacts in wet steam zones1. The last stage blades of thermal power and nuclear power condensing steam turbines particularly suffer from severe WDE damage due to prolonged exposure to wet-steam zones. This degradation deteriorates aerodynamic efficiency and elevates safety risks in modern turbines employing lengthened blades with tip velocities exceeding 600 m/s2. Field evidence from power plants in China confirms this dual erosion mechanism. Each low-pressure cylinder accumulates up to 40 kg of oxide particles (Panshan plant) and severe SPE pitting occurs on low-pressure blades (Haimen plant), indicating the coexistence of SPE and WDE risks in actual operations of low-pressure blades, as shown in Fig. 1.This indicates that the entrained ferric oxides and impurities peeled from boiler superheater/reheater tubes within wet-steam flow contribute to secondary SPE damage in low-pressure cascades3.

Fig. 1
figure 1

The SPE damage morphologies in the blades of low pressure stage. (a) Oxide particles adsorption on the single condenser; (b) Oxide particles adsorption on several condenser; (c) Serious SPE damage of the first stage rotor blade in the low pressure cylinder (d) Serious SPE damage of the first stage stator blade in the low pressure cylinder.

Full size image

Recent research and microscopic observations have shown that WDE performance is highly sensitive to the regularity of surface structure4,5,6. And the effect of SPE will produce a large number of random pits on the surface of the material to form a special irregular surface structure. Therefore, the influence of the fine particles should be considered when studying the water droplet erosion process of the last stage blade substrates and surface strengthening layers7,8.

Studies by Mahdipoor et al.9, Mann et al.10, and Ahmad et al.11 demonstrated that initiation and expansion of crack initiates via stress waves and lateral jetting from droplet impacts, with pre-existing surface irregularities amplifying the damage. Concurrently, Scholars mainly focus on the theoretical model12,13,14,15,16, main influencing factors17, numerical methods18 and protective measures19,20 to study SPE behavior. Besides, the main failure mechanisms of SPE for plastic target are micro-cutting and extrusion-forging, or micro-ploughing and extrusion-forging corresponding to flaky or spherical particles, respectively21. SPE could may theoretically accelerate WDE progression through surface morphology modifications. However, conventional erosion models exclude this coupling effect, relying on the outdated premise of mutually exclusive particle and droplet impingement on blades. While surface roughness effects on WDE have been explored1,5, no systematic study quantifies how SPE-induced microstructures modulate subsequent WDE behavior7,11. The synergistic interaction of SPE and WDE on low-pressure blades exacerbates material degradation through coupled mechanical impacts, creating complex failure pathways validated by on-site situation. Current research remains fragmented, isolating these erosion mechanisms rather than addressing their interdependence—a critical knowledge gap for turbine reliability optimization in wet-steam environments.

On the basis of erosion rate, particle concentration measurements (0.001 mg/kg steam) by a typical erosion nozzle, support this dilute-phase premise (the particles within 0.1 kg/m3 as dilute phase)22,23. Therefore, the solid particles in the actual turbine cascade are dilute phase. The particles and water droplets simultaneously impinge on the blade surface is almost impossible. The sequential SPE-WDE interactions occur during the operational condition. With increase of last stage blade length, the maximum circumferential velocity on the tip of the blade has exceeded 600 m/s2. And the velocity of solid particles hitting the cascade is within the range of 150 to 360 m/s24,25,26.

Therefore, in this paper, a sequential SPE-WDE test was carried out using an upgraded SPE rig and high-speed rotating waterjet test platform to investigate the effect of SPE-generated surface microstructure induced by micro-particles impact on WDE progression for blade substrate and surface strengthening layers. The typical impact velocity of 200 m/s for solid particles and 650 m/s for water droplets were selected in our SPE tests and WDE tests. By correlating surface microstructure evolution with cumulative erosion weight loss, this work provides the mechanism framework and data foundation for SPE-WDE synergy in steam turbine blades under actual operating conditions, offering critical insights for novel erosion-resistant blade design.

Experimental procedures

Figure 2 presents the physical diagram of the SPE test system, which is mainly composed of the erosion test chamber, gas source system, particle feeding system and exhaust system. The brief introduction is presented in21. The working process of this test system is as follows:

Fig. 2
figure 2

The physical diagram of the SPE test system.

Full size image

The primary airflow enters a long acceleration nozzle, where the particles produced by the screw feeder (Fig. 3a) are transported to the throat of the nozzle and accelerated to a specified speed. Particle feed rates are controlled via a stepper motor-driven system, ensuring precise gas–solid flow generation prior to target impingement in the test chamber. Post-impact particulates are evacuated through an ejector-cyclone separation system for continuous operation. Four 17-4PH stainless steel specimens are directly mounted on a servo-actuated T-groove fixture (Fig. 3b), enabling automated positional switching during testing. As fixture replacement necessitates recalibration of the air pressure and stepper motor speed for the feeding system, the outermost specimen (Position 1) serves as a pre-test calibration sample with subsequent data exclusion to guarantee the reliability of subsequent test results. Additionally, the integrated electric heater (max 800 °C ± 15 °C) facilitates thermal state SPE experiments.

Fig. 3
figure 3

The critical components of SPE test system.

Full size image

Schematic diagram of WDE test platform is shown in Fig. 4 (created by SOLIDWORKS 2018, https://www.solidworks.com). The system is mainly composed of ultra-high pressure pump (240–260 MPa), test chamber with impeller-mounted specimens connected to the rotating shaft (rotational speed controlled by the cabinet and the 1:3 gear speed increaser), lubrication system, and drainage system. As shown in Fig. 4, a waterjet with the maximum velocity of 650 m/s is generated by a nozzle embedded with diamond (Ø 0.15 mm). The detailed description of the test system could be seen in references4,27.

Fig. 4
figure 4

Schematic diagram of WDE test platform.

Full size image

Experimental sequence: a. SPE test: The martensitic substrates and stellite laser cladding specimens were subjected to iron oxide particle impacts (Sauter Mean Diameter ≈ 100 μm)21 at 200 m/s for various impact angles (15°–90°) and durations (1–5 min). b. WDE test: Subsequent testing employed a diamond-embedded nozzle with diameter of 0.15 mm generating 650 m/s waterjet (mass flow rate: 0.0115 kg/s with a rough estimation) for 90-min exposure period. The detailed test parameters of WDE are presented in Table 1.

Table 1 The test parameters of the WDE in the test.
Full size table

In order to study the effect of impact angles on the WDE characteristics, a stepped structure of specimen is designed (Fig. 5). Figure 5a depicts the martensite substrate sample after 2 min SPE test, with the top featuring a 60° sample and the bottom a 90° sample. Figure 5b shows the 30° stellite laser cladding sample after 2 min SPE test.

Fig. 5
figure 5

The samples after 2 min SPE test. (a) Martensite substrate samples after 2 min SPE test (the top is 60° sample and the bottom is 90° sample); (b) 30° sample of the stellite laser cladding material after 2 min SPE test.

Full size image

The specific test scheme is shown in Table 2. Table 3 lists the chemical composition and mechanical properties of martensite substrate and the stellite laser cladding materials.

Table 2 SPE test scheme.
Full size table
Table 3 Chemical composition and mechnical properties of blade materials.
Full size table

The installation of the specimen on the impeller turntable of high-speed waterjet test system is shown in Fig. 6a. Figure 6b displays the real-time image captured during the water droplet erosion test. The target samples after ultrasonic cleaning with anhydrous ethanol for 5 min and drying after SPE tests and 90 min WDE tests are presented in Fig. 7a and b, respectively.

Fig. 6
figure 6

The high-speed waterjet test rig. (a) The installation of the specimen on the impeller turntable; (b) The real-time image during the WDE test.

Full size image
Fig. 7
figure 7

The target samples after ultrasonic cleaning and drying. (a) The target samples after ultrasonic cleaning and drying after the SPE tests; (b) The target samples after ultrasonic cleaning and drying after 90 min WDE tests.

Full size image

Results and discussions

WDE mass loss and damage caused by SPE

Post-WDE mass loss quantification at 10-min, 30-min, and 90-min intervals revealed distinct erosion characteristics for substrates and cladding specimens, as shown in Fig. 8. 90° samples only by WDE test were used as the control group and labeled as 90°-WDE. The results indicate that the cumulative WDE mass loss of the substrate control group is about average 15 times than that of the cladding sample, which proved that the WDE resistance of the stellite cladding sample was far better than that of the martensite substrate.

Fig. 8
figure 8

The variation of mass loss of the substrate and the stellite laser cladding samples along the testing time.

Full size image

Substrate erosion rates increased exponentially with SPE exposure time for 1 min, 2 min, 3 min, and 5 min (0.308 mg/min → 0.317 mg/min → 0.324 mg/min → 0.334 mg/min at 90°). The 90°-WDE control group exhibited stable erosion rates (0.056 mg/min), approximately 17.47% of the mass loss rates observed in all 90° samples subjected to solid particle erosion. while SPE-pretreated samples demonstrated angle- and duration-dependent WDE damage acceleration.

For the stellite cladding samples exposed to solid particle erosion, the cumulative WDE mass loss rates during the initial 10 min WDE test, were 0.02 mg/min, 0.03 mg/min, 0.05 mg/min, 0.07 mg/min, and 0.09 mg/min, respectively. Subsequent 20-min and the final 60-min exposure showed approximately 75.92% and 88.34% reductions in mass loss rates, respectively, contrasting with the minimal rate of the control group (0.0033 mg/min), which was significantly lower than all SPE-treated samples.

In conclusion, increasing the impact angles and test duration of SPE significantly exacerbate WDE damage. Within the duration of this experiment, the effect of solid particle erosion on WDE rate of cladding samples notably weakens with the increasing WDE duration, while the erosion rate of substrate samples remains relatively constant, and the influence of WDE duration can be ignored.

The clearly visible deep groove was formed on the surface of 17-4PH substrate samples, while only a white trace zone was formed in the WDE region, as shown in Fig. 9. In contrast, only an etched trace band with a width of about 430 μm with an extremely thin hair-like trace groove in the center was observed under the scanning electron microscope (SEM) after 90-min WDE test due to the slow progression of WDE damage for cladding samples, under SEM. The depth of the groove was very shallow and almost negligible, as shown in Fig. 10.

Fig. 9
figure 9

The surface damage marks of the substrates and the stellite laser cladding samples.

Full size image
Fig. 10
figure 10

A water-etched trace band after 90 min WDE test for the stellite cladding sample.

Full size image

Figure 11 shows the distribution of the width and depth of the erosion zone of the substrate along testing time. The 17-4PH substrate exhibited obvious visible surface delamination (122–156.6 μm width, 72.1–114.3 μm depth) after 10-min WDE exposure, progressing to catastrophic erosion (813.5–901.6 μm width, 262.7–307.9 μm depth post-90-min) due to lateral flow constraints concentrating impact forces. It can be seen that the erosion width and depth of the substrate increased with the increasing erosion time. In addition, as the water erosion time increased, the growth rate of erosion width of the substrate decreased with the test time. And the growth rate of erosion depth gradually increased with test time. This is because in the initial stage, the waterjet can be freely expanded, so the width expansion was relatively easy. As the erosion continued, the water column impacted in the WDE groove, and will be subjected to a larger lateral constraint, making the impact force more concentrated. So the width expansion becomes slow, and the depth expansion speeds up.

Fig. 11
figure 11

The distribution of the width and depth of the erosion zone of the substrate along testing time.

Full size image

The typical microscopic morphology of SPE and WDE

The micro WDE and SPE morphology of the martensite substrate after 5 min SPE and 10 min WDE are shown in Fig. 12. It can be seen from the figures that there is an obvious difference in the microscopic morphology of the SPE and WDE zone for the substrate samples. The SPE zone has a typical plate-like structure. And there are many cutting scratches, grooves and lips disorderly distributed on the surface. And the WDE region is a multi-fibrous terracing step-shaped structure with river-like patterns. It can be seen from Fig. 13 that the SPE morphologies of the martensite substrate and the stellite cladding sample are basically similar both with typical lamellar morphology. The microscopic morphology of the stellite laser cladding specimen after 5 min SPE and 10 min WDE is shown in Fig. 14. It can be seen that the surface of the stellite laser cladding sample exhibits a very obvious plate-like SPE morphology. While the water droplet erosion marks are still almost invisible.

Fig. 12
figure 12

The micro morphology WDE and SPE of the martensite substrate after 5 min SPE and 10 min WDE. (a) × 100; (b) × 500.

Full size image
Fig. 13
figure 13

SPE morphologies of the martensite substrate and the stellite cladding sample. (a) Substrate; (b) Stellite laser cladding sample.

Full size image
Fig. 14
figure 14

The microscopic plate-like SPE morphology and WDE morphology of the stellite laser cladding specimen after 5 min SPE and 10 min WDE.

Full size image

The morphologies of the stellite laser cladding sample and the martensite substrate after 30 min WDE test are shown in Fig. 15 (2 min SPE) and Fig. 16 (5 min SPE), respectively. The obvious damage morphology formed in the WDE zone of the stellite cladding specimen.

Fig. 15
figure 15

The WDE morphologies of the stellite laser cladding sample and the martensite substrate after 30 min WDE test (2 min SPE + 30 min WDE).

Full size image
Fig. 16
figure 16

The WDE morphologies of the stellite laser cladding sample and the martensite substrate after 30 min WDE test (5 min SPE + 30 min WDE).

Full size image

In addition, it can be clearly seen from the figures that there is a significant difference in the micro structure of the WDE region for the substrate and the stellite cladding samples. The WDE area of the substrate sample has a very pronounced stepped fibrous structure. While the micro morphology of the stellite cladding sample presents a honeycomb stalactite peak cluster structure which is similar to the natural stalactite morphology. The formation process of the natural stalactite morphology usually takes thousands or millions of years. Due to such a lengthy forming process, it is speculated further that such a microstructure is probably the underlying reason for the superior WDE resistance of the stellite laser cladding sample.

Besides, the size of WDE microstructures of the two materials is also worth emphasizing. The size of each stalactite-like petal in WDE zone of the stellite cladding samples is about 2 microns, while the stepped structure of martensite substrate is obvious larger within the range of 10–20 microns.

The microscopic morphologies of the two kinds of materials after 2 min SPE and 90 min WDE test are shown in Fig. 17. It is obvious that the microstructures in the WDE zone for the two materials are more distinct.

Fig. 17
figure 17

The microscopic morphologies of the stellite laser cladding specimen and the martensite substrate after 2 min SPE and 90 min WDE test.

Full size image

The quantitative analysis of WDE damage by SPE-induced surface microstructure

The variation of cumulative mass loss curve along erosion time of wire cutting 17-4PH and stellite laser cladding sample for only conventional WDE test is shown in Fig. 18. The corresponding data are used as the control group, as shown in Tables 4 and 7.

Fig. 18
figure 18

The variation of cumulative mass loss curve along testing time of wire cutting 17-4PH and stellite laser cladding sample along erosion time for the conventional WDE test.

Full size image

The effect of solid particle erosion on WDE damage behaviour for 17-4PH substrate was analysed base on the 30 min test data. It can be seen from Tables 4 and 5 that the solid particle erosion of micro particles had a significant influence on the WDE damage characteristics of the blade substrate, which accelerated WDE damage of the wire cutting 17-4PH substrate samples by about 5.87 times. The WDE damage surfaces of 17-4PH with prior SPE and without SPE under 30 min and 155 min WDE test are presented in Fig. 19, respectively. The eroded average width and corresponding mass loss of WDE zone was 704 μm/9.5 mg and 760 μm/8.7 mg, respectively. The results also provided powerful support for the fact that SPE could lead to more severe WDE damage.

Table 4 Statistical table of mass loss for conventional WDE test of wire-cut 17-4PH substrate samples.
Full size table
Table 5 Statistical table of mass loss of martensite 17-4PH substrate samples exposed to SPE.
Full size table
Fig. 19
figure 19

The WDE damage surfaces of 17-4PH with prior SPE and without SPE under 30 min and 155 min WDE test. (a) 30 min WDE with prior SPE corresponding mass loss: 9.5 mg; (b) 155 min WDE without SPE corresponding mass loss: 8.7 mg.

Full size image

The surface roughness of the samples measured by portable surface roughometer TIME 3100. It was found that the roughness after SPE only increased by 0.0445–0.1295 μm by measuring the surface roughness before and after solid particle erosion. The average variation was 5.67%, as shown in Table 6. From the above results, it was indicated that such a little change in surface roughness before and after SPE still significantly aggravated the WDE damage.

Table 6 Statistical table of surface roughness before and after SPE for the wire-cut substrate samples.
Full size table

Furthermore, the surface morphologies of the samples were studied by microscope observation. It was worth noticing that the microstructures were significantly different between the wire cutting and SPE samples under SEM observation. The surface of the conventional wire cutting specimen had a regular metal texture or mesh structure (see Fig. 20a). While the target surface after SPE formed an intricate irregular morphology. Various grooves, accumulated lips and scratches were randomly distributed on the surface as shown in Fig. 20b.

Fig. 20
figure 20

Surface microscopic morphology of martensite substrates. (a) Surface morphology of the wire-cut 17-4PH substrate sample; (b) Surface morphology with SPE.

Full size image

Hence, it can be speculated that the orderly texture or grid pattern created on the machined surface facilitates the diversion and dispersion of liquid flow, resulting in relatively minor damage. Conversely, the irregular surface structure induced by SPE impedes the unhindered expansion of liquid flow, significantly reducing the latency period of water droplet erosion, thereby leading to more severe WDE damage. The above studies fully demonstrate that the significant effect of solid particle erosion on the acceleration of WDE damage of the substrate is mainly determined by the surface microstructure.

Similarly, the influence of solid particle erosion on the water droplet erosion performance of the stellite cladding sample is studied. It is also found that the solid particle erosion has a significant effect on the WDE damage. The solid particle erosion accelerates by an average of about 2.52 times WDE damage compared with the conventional wire cutting stellite laser cladding sample. The detailed data are shown in Tables 7 and 8.

Table 7 Statistical table of mass loss for conventional WDE test of the wire-cut cladding samples.
Full size table
Table 8 Statistical table of mass loss of cladding sample exposed to SPE.
Full size table

The WDE damage surfaces of stellite laser cladding samples with prior SPE and without SPE under 90 min WDE test is presented in Fig. 21, respectively. The average mass loss of WDE zone is 1.0 mg and 0.4 mg, respectively. The visible damage morphologies and mass loss of the samples with prior SPE are significantly more severe than those without SPE in the same 90 min WDE test, which fully indicated that SPE damaged surface greatly accelerated the water droplet erosion rate.

Fig. 21
figure 21

The WDE damage surfaces of stellite laser cladding samples with prior SPE and without SPE under 90 min WDE test. (a) 90 min WDE with prior SPE corresponding mass loss: 1.0 mg; (b) 90 min WDE without SPE corresponding mass loss: 0.4 mg.

Full size image

It is found that the surface roughness after SPE increases 0.0425–0.318 μm compared with the wire-cut ones. The average variation is 19.35%, as shown in Table 9. Likewise, the WDE damage of cladding material is also aggravated by SPE despite of the little change of the surface roughness.

Table 9 Statistical table of surface roughness before and after SPE of the wire-cut cladding samples.
Full size table

Microstructural analysis reveals conventional wire cutting specimen exhibits a regular metal texture or mesh structure, while the surface after SPE is in wild disorder as shown in Fig. 22. The above studies show that the water droplet erosion performance is particularly sensitive to the surface structure. Namely, the solid particle erosion accelerates the water droplet erosion damage of the blade substrate and the strengthened stellite laser cladding material, which is mainly determined by the surface irregular structure. The irregular and uneven microstructure of surface formed by SPE is not conducive to the drainage and expansion of the liquid flow, resulting in greatly shortening the latency and accelerating the erosion damage.

Fig. 22
figure 22

Surface microstructures of the stellite laser cladding sample. (a) Surface morphology of the wire-cut stellite laser cladding sample; (b) Surface morphology with SPE.

Full size image

Conclusions

This study investigates the synergistic effects of solid particle erosion (SPE) on water droplet erosion (WDE) performance through a self-developed SPE test system and the high-speed rotating waterjet test platform. Key findings are summarized as follows:

  1. 1.

    SPE generates irregular surface structures that shorten WDE latency and accelerate damage progression. Regular machined textures is more conducive to the drainage and expansion of the liquid flow than that of the irregular structure surface formed by the solid particle erosion, which makes more serious WDE damage of the latter.

  2. 2.

    Solid particle erosion significantly affects the water droplet erosion performance of 17-4PH substrate and the Stellite laser cladding samples, accelerating WDE damage by 5.87 times and 2.52 times, respectively.

  3. 3.

    It is worth noting that the SPE microstructure of the blade substrate is similar to the stellite cladding sample in the SEM observation. But the WDE morphology is significantly different. a. WDE morphology for Substrate surface exhibits a terracing step-shaped fibrous structure, and the size is significantly larger in the range of 10–20 microns; b. the WDE microscopic morphology of the stellite cladding sample shows a honeycomb stalactite peak cluster structure, which is very close to the natural stalactite morphology, and the size of each stalactite petal is about 2 microns. The microstructures may be the underlying reason for determining its superior water droplet erosion resistance.

  4. 4.

    The hierarchical honeycomb stalactite clusters in Stellite laser cladding, analogous to millennia-formed natural speleothems, enhance WDE resistance through optimized energy dissipation mechanisms associated with their micron-scale stalactite petal morphology.

The findings enable predictive modeling of multi-phase erosion in turbine blades and oil/gas pipelines by correlating SPE-induced surface topographies with accelerated WDE damage. This indicates the potential advantages of regular textured surfaces for applications in aerospace and wind turbine components exposed to rain and sand erosion.