Abstract: Objective To improve the wear and corrosion resistance of tubular inner wall parts in the petrochemical field and extend the service life of parts. Methods FeCoNiCrMo0.2 high entropy alloy (HEA) coatings at five different powder feeding rates were prepared on the inner wall of 316L steel pipe by laser cladding technology. The solid solution distribution, microstructure fineness, hardness and wear resistance of the five HEA coatings were evaluated by phase composition, grain structure, hardness test and wear test, and the wear mechanism was analyzed. Finally, the corrosion resistance of HEA coatings was characterized by potentiodynamic polarization curves. Results HEA coatings are composed of single-phase FCC, and the microstructure is mainly composed of columnar crystals and cellular crystals. The wear mechanism of the coatings includes adhesive wear and abrasive wear. The powder feeding rate of 15 g/min makes the coating form more fine cellular grain structure, increases the content of hard phase in the coating, has a greater solid solution strengthening effect on the coating, and reduces the friction coefficient of the coating to 0.58, which improves the tribological properties of HEA coatings. And in the electrochemical test, the 15 g/min HEA coating has the lowest Icorr value, and the in-situ generated passivation film on the surface of the coating has more anti-corrosion substances such as Cr2O3. Conclusion The powder feeding rate significantly affects the hardness and tribological properties of the coating by affecting the phase composition, organizational evolution and element distribution in the inner wall HEA coating. Optimizing the powder feeding rate can help reduce the friction coefficient and wear loss of the HEA coating, and also help to enhance the density of the passivation film formed in the coating during the corrosion process and increase the content of anti-corrosion substances such as Cr2O3 and MoO2.
Keywords: laser cladding; high entropy alloy; microstructure; microhardness; wear behavior; electrochemical corrosion

In fields such as machinery manufacturing and petrochemical industries, many tubular and inner wall parts usually operate under harsh conditions such as high pressure, high load and corrosive environment, and are very prone to wear and corrosion problems. And these parts play a key role in the field, such as the inner wall of hydraulic cylinders, injection molding machine barrels, and oil and gas pipelines. In this context, this highlights the importance of extending the life of metal parts and reducing economic losses. Therefore, it is critical to strengthen the processing process of these parts. In order to extend the service life of inner wall parts, surface modification of their internal structure is required.

High entropy alloy (HEA) is a new alloy system. Due to its four special effects of high entropy effect, lattice distortion effect, delayed diffusion effect, and cocktail effect, high entropy alloys exhibit stable high hardness, high wear resistance, high corrosion resistance, and high thermal fatigue resistance [2−3]. However, high entropy alloy systems are usually expensive. If the entire part is made of high entropy alloy, the manufacturing cost will be greatly increased. This limits the application of high entropy alloys in industry.

Laser cladding is a novel and efficient surface repair method with outstanding features such as fast processing speed, small thermal deformation of the substrate, and good metallurgical bonding with the substrate [4]. It can provide a new surface modification solution for parts such as curved surfaces and pipes, and can also be used to repair failed parts. Its effect is even comparable to that of newly manufactured products, thereby greatly reducing the waste of energy and resources [5-7]. Therefore, the use of laser cladding technology to prepare high-entropy alloy coatings can not only improve the performance of the surface of parts, but also greatly reduce the consumption and manufacturing cost of high-entropy alloys [8].

According to relevant research, the alloy coating prepared by laser cladding technology is greatly affected by the process parameters [9]. Therefore, in order to obtain alloy coatings with better comprehensive performance, it is often necessary to compare the performance of alloy coatings prepared under different process parameters and select the optimal parameter combination. In recent years, many experts and scholars have carried out a large number of studies on the influence of process parameters on the performance of alloy coatings. Wang et al. [9] used laser cladding technology to study the effect of scanning speed on the microstructure and performance of Fe-based cladding layers. The study showed that the amorphous content in Fe-based coatings is closely related to the scanning speed. The scanning speed affects the dilution rate and cooling rate by changing the heat input, and also affects the fine grain strengthening and dispersion strengthening effects of the coating. With the increase of scanning speed, the hardness of the coating increases and the corrosion resistance is enhanced. Jiao et al. [10] also studied the effect of scanning speed on the microstructure and wear performance of T15M cladding layer. The study showed that: with the increase of scanning speed, the carbides precipitated from the original austenite grain boundaries in the T15M coating increased and the local segregation in large areas decreased. Excessive scanning speed will cause defects such as holes, microcracks, and inclusions in the coating. When the scanning speed is 100 mm/min, the average hardness of the cladding area is the lowest. When the scanning speed reaches 300 mm/min, the wear performance of the coating deteriorates and the worn surface peels off in blocks. When the scanning speed is 200 mm/min, the coating has a fine grain size and high solid solubility, a high average hardness, and the best wear resistance. Li et al. [11] studied the effect of laser power on the microstructure and properties of laser cladding ZrB2/ZrC in-situ enhanced coatings. The results showed that the increase of laser power significantly increased the width and height of the coating, inhibited the secondary crystallization growth of dendrites, gradually transformed the grain shape into a blocky microstructure, and significantly improved the microhardness of the coating. When the laser power was 2.8 kW, the microhardness was the highest, about 720HV0.5. Liu et al. [12] studied the effect of powder feeding rate on the microstructure and wear and corrosion resistance of laser cladding Ni−60%WC coatings. The results showed that with the increase of powder feeding rate, the hardness of the coating increased first and then decreased. When the powder feeding rate was 12 g•min−1, the coating had the highest hardness and had the least friction weight loss and wear and corrosion resistance. However, the powder feeding rate did not change the wear mechanism of the Ni−60%WC coating. The wear mechanism of the coating still consisted of abrasive wear, adhesive wear and oxidative wear.

The above studies show that the alloy coatings prepared using different process parameters still have great differences in performance. In addition, most scholars have explored the influence of process parameters on alloy coatings on the outer surface of flat or tubular parts [13−14], while there are few studies on high-entropy alloy coatings on the inner wall of tubular parts. However, many parts with concave inner walls, such as oil and gas pipelines, also have various failure modes such as wear and corrosion. Compared with flat plates or the outer surface of parts, when laser cladding technology is used to repair the inner wall of parts, due to its slow heat dissipation rate and concentrated thermal stress, the coating structure is often coarse and the mechanical properties are reduced. Based on this, this paper used laser cladding technology to prepare FeCoNiCrMo0.2 high entropy alloy coating on the inner wall of 316L stainless steel pipe (oil and gas transmission pipe simulation part) with different powder feeding rates as process parameters, and explored the influence of powder feeding rate on the macro morphology, phase structure, microstructure, hardness, wear and corrosion properties of FeCoNiCrMo0.2 high entropy alloy cladding layer, which can provide theoretical reference and experimental example for subsequent research on inner wall laser cladding process and comprehensive performance of high entropy alloy.

1 Experimental materials and methods
1.1 Materials
316L stainless steel is one of the commonly used materials in the petrochemical industry. Due to the addition of Cr, Mo, and Ni elements, 316L has been widely used in the industrial industry due to its excellent thermal fatigue resistance and corrosion resistance, such as oil and gas pipelines. As a key component in the petrochemical industry, the service life of the pipeline is related to the operation of the entire oil and gas transportation system. Therefore, in this study, 316L steel pipe (oil and gas pipeline simulation part) is selected as the substrate, and its chemical composition is shown in Table 1. The substrate size is 211 mm inner diameter, 200 mm long, and 8 mm thick. Before cladding, the surface to be clad is cut by 0.5 mm using a lathe to remove surface oxides and rust marks, which ensures the flatness of the cladding surface and reduces the reflection of the laser. Then, the surface oil is cleaned with anhydrous ethanol and set aside [15]. Spherical FeCoNiCrMo0.2 high entropy alloy powder with a purity higher than 99.95% prepared by atomization method is used as the cladding raw material. The specific chemical composition is shown in Table 2. Since the cladding powder contains Mo element, it is a refractory alloy powder. In order to help the alloy powder to be evenly heated and melted during cladding and reduce the powder sticking phenomenon of the coating, the powder particle size range is selected to be 15~53 μm. The SEM morphology of HEA powder is shown in Figure 1.

1.2 Laser cladding process
In the petrochemical field, the inner walls of a large number of inner-hole tubular parts are continuously subjected to wear and corrosion failure, such as oil and gas pipelines. However, due to the narrow internal structure of such parts, most existing lasers and cladding processes cannot achieve cladding on the inner wall of the pipe body. Therefore, a cladding head device for inner-hole laser cladding was used in the cladding process. The cladding head was controlled by a six-axis KUKA robot to carry out inner-wall laser cladding experiments. The ultra-high-speed GS-CW-6000 fiber laser produced by Shaanxi Guosheng Laser Technology Co., Ltd. was used to process single-pass and multi-pass samples of FeCoNiCrMo0.2 high-entropy alloy coatings. The coaxial powder feeding processing method was adopted, and the protective gas was argon [16]. The inner wall laser cladding processing diagram is shown in Figure 2. The specific parameters of the cladding process under different powder feeding rates are shown in Table 3.

1.3 Coating characterization
After the single-pass and multi-pass cladding samples are completed, the samples are cut into 8×8×5 mm in size by wire cutting, and then polished with 1 000# ~ 2 200# sandpaper, and then mechanically polished with a diamond polishing agent with a particle size of 0.5 μm. Aqua regia (volume ratio HCl:HNO3=3:1) was prepared to metallographically etch the multi-pass coating samples for 120 s.
The German Bruker D8 Advance X-ray diffractometer (XRD) was used to analyze the phase of the multi-pass coating, with a voltage of 40 kV, a current of 40 mA, a step size of 0.02, and a step speed of 0.2 sec/step. The JSM−7610F plus scanning electron microscope (SEM) and energy dispersive spectrometer (EDS) were used to observe the microstructure and element distribution of the multi-pass HEA coating. The samples to be tested were mechanically polished, and then the microhardness values ​​of single-pass HEA coatings prepared at five different powder feeding rates were measured using a Huayin HV-1000A digital microhardness tester. The applied force was 1 000 g and the dwell time was 15 s. To ensure the accuracy of the data, each sample was measured 3 times and the average value was calculated. The friction coefficient was tested using an MFT-5000 friction and wear tester. A Si3N4 ceramic ball with a diameter of 9 mm was used as the friction pair. The dry friction experiment was carried out under a load of 10N, the loading time was 30 min, and the sliding frequency was 2 Hz. After the dry sliding wear test, the wear weight loss of the sample was measured, and the wear morphology of the sample was observed by SEM, and its wear mechanism was explored. The corrosion resistance of the coating was analyzed using a CHI660D electrochemical workstation. The working electrode was the multi-pass cladding layer, the reference electrode was a saturated calomel electrode, and the platinum electrode was used as the auxiliary electrode. The electrolyte used was a neutral solution of NaCl with a mass fraction of 3.5wt%. The electrochemical polarization curve (LSV) test parameters included: starting voltage of −2 V, ending voltage of 2 V, and scanning rate of 4 mV/s. During the test, the remaining surfaces were sealed with resin glue, and the surface to be tested was exposed to the electrolyte.

2 Results and Analysis
2.1 Forming Quality
Figure 3 shows the forming quality of multiple inner wall HEA coatings at different powder feeding rates and the cross-sectional morphology of a single coating. It can be observed from the figure that when the powder feeding rate is 6~15g/min, the surface of the cladding layer is bright, smooth and continuous as a whole, without obvious defects such as cracks and pores. However, the HEA coating prepared at 6~9 g/min is thinner and the cross-sectional appearance is relatively poor. The coating prepared at 12~15 g/min has excellent forming quality and forms a good metallurgical bond with the substrate, and the cross-sectional morphology is close to a semi-ellipse. As the powder feeding rate continued to increase to 18 g/min, the phenomenon of unmelted powder in the coating intensified, and the defects such as pores formed after the coating and the molten pool solidified increased. The coating was not formed and the forming quality was poor. This is because when the powder feeding rate is too large, the amount of alloy powder transmitted to the action area per unit time is too large, which leads to the inability of the laser energy to completely melt the input powder during the inner wall laser cladding process, thereby reducing the forming quality of the coating. In addition, the cross-sectional morphology of the single-pass coating is still close to a semi-ellipse, and the melt width is also larger than that of 15 g/min. This is due to the fast heating and cooling characteristics of laser cladding. During the cladding process, the powder accumulated on the top of the coating flows to the surroundings and solidifies, which slightly increases the melt width.

The dilution rate reflects the combination of the coating and the substrate, and also reflects the thermal effect of the laser beam on the substrate during the cladding process. In order to more clearly explore the influence of different powder feeding rates on the dilution rate of HEA coatings, the height of the five coatings and the depth of the molten pool were measured, and the schematic diagram of the morphological parameters is shown in Figure 4. Combining Figures 3a2~e2 and Figure 4, and according to the calculation formula of the dilution rate η in formula (1), the dilution rates of the five HEA coatings were calculated, and the results are shown in Table 4. Among them, η is the dilution rate, H1 is the molten pool depth (melting depth), H2 is the cladding layer thickness (melting height), S1 is the heat affected zone area, and S2 is the cladding layer area.

As shown in Table 4, the HEA coating at 6 g/min has the largest dilution rate of 52%. This is because at a powder feeding rate of 6 g/min, the number of high entropy alloy powder particles irradiated by laser energy per unit area of ​​the substrate is relatively reduced, and more laser energy acts on the substrate to form a deeper molten pool, which makes the coating melt height relatively low, resulting in a relatively large dilution rate. In addition, its overall area is relatively small, and the cross-sectional appearance is relatively poor. When the powder feeding rate increases to 9 g/min, the coating’s melt height increases, the melt depth decreases, and the dilution rate decreases, indicating that compared with 6 g/min, the powder feeding rate of 9 g/min has less thermal effect on the substrate, reducing the probability of micro-thermal deformation of the substrate during the cladding process. When the powder feeding rate is 18 g/min, the HEA coating’s melt height reaches the maximum value, the melt depth reaches the minimum value, and the dilution rate reaches the minimum value of 44%. This is because the powder feeding rate is too high, the amount of molten powder accumulation in the coating increases, which increases the melt height. At the same time, the powder accumulation is too large, causing a large amount of laser energy to act on the powder to form a coating, and less laser energy to act on the substrate, causing the melt depth to decrease. In general, under these 5 different powder feeding rates, the HEA coating prepared at a powder feeding rate of 12~15 g/min has the best forming quality and a relatively moderate dilution rate.

2.2 Phase composition
Figure 5 shows the XRD patterns of the FeCoNiCrMo0.2 high entropy alloy coatings prepared at powder feeding rates of 6 g/min, 12 g/min, and 18 g/min. It can be seen from the figure that the lattice structures of the HEA coatings prepared at different powder feeding rates are basically the same, all of which are face-centered cubic (FCC) single-phase solid solution structures, and no new phases are generated [17−19]. The PDF standard card number of the FCC phase structure in the coating is PDF#33−0397 in the ICDD database. By comparing the PDF card, the diffraction peak of the FCC phase is determined to be the Fm−3m space group. The number of diffraction peaks of the HEA coatings at different powder feeding rates is the same, but the intensity of the diffraction peaks has changed strongly, indicating that the crystal orientation of the coating surface has changed [20]. And it can be clearly seen from Figure 5b that in the XRD pattern at high resolution, the increase in powder feeding rate causes the (200) crystal plane diffraction peak to shift slightly to the right. This shows that the difference in powder feeding rate will not only affect the diffraction peak intensity, but also cause lattice distortion inside the HEA coating, but will not have much effect on the phase type of the coating [21−23]. At the same time, the difference in atomic radius will also cause lattice distortion and affect the size of the lattice parameter. In order to more clearly understand the effect of powder feeding rate on lattice parameters, according to the Bragg diffraction law of formula (2), the average lattice parameters of the three main peaks in the HEA coating prepared with powder feeding rates of 6 g/min, 12 g/min and 18 g/min are calculated to be 3.5923 Å, 3.5948 Å and 3.5953 Å respectively. The change in lattice parameters further proves that the change in powder feeding rate causes local lattice distortion in the coating, affecting the lattice parameters.

In formula (2), λ is the wavelength of the incident wave, θ is the angle between the incident light and the crystal plane, a is the lattice constant, and h, k, and l are the crystal plane indices.
Moreover, it can be seen from Figure 5b that with the increase of powder feeding rate, the diffraction peak becomes wider near 2θ=50.79°, and all elements are dissolved in the same lattice to form a single disordered solid solution phase. This shows that the change of powder feeding rate can affect the dynamic hysteresis diffusion effect of high entropy alloys, thereby inhibiting the diffusion in the coating alloy and making it difficult for atoms to diffuse. The increase of powder feeding rate also increases the solid solubility between different alloying elements, thereby increasing the generated FCC solid solution phase and the lattice parameter. This shows that the process parameters can play the role of FCC stabilizer in the composition design of HEA coating. In addition, in the coating with a powder feeding rate of 18 g/min, the diffraction peak of the coating increases, and the intensity of the FCC phase diffraction peak (200) increases, indicating that the coating at this time contains more intermetallic compounds such as Co3Fe7 and FeNi3.

The key factors that affect the phase composition in HEA coatings are mixing entropy (ΔSmix), mixing enthalpy (ΔHmix), atomic size difference (δ), relative contribution of ΔSmix and ΔHmix to Gibbs free energy (Ω), and valence electron concentration (VEC). ΔSmix represents the degree of disorder in the HEA system. The larger the value of mixing entropy, the more disordered the system is and the easier it is to form a solid solution phase. ΔHmix reflects the affinity of elements to a certain extent and represents the probability of forming compounds between different elements. The more negative the mixing enthalpy value, the easier it is to form compounds. According to the Stefan-Boltzman law and thermodynamic principles, the calculation formulas for ΔSmix, ΔHmix, δ, Ω and VEC valence electron concentration are shown in equations (3) to (7) [8].

When 11≤ΔSmix≤19.5 J/(K•mol), −22≤ΔHmix≤7 kJ/mol, 0≤δ≤8.5, Ω≥1.1, these conditions are met at the same time, the HEA system usually tends to form a simple solid solution phase. In addition, the type of solid solution can be predicted by the valence electron concentration (VEC): when VEC<6.87, HEA tends to form a single-phase body-centered cubic (BCC) structure; when the VEC value is in the range of 6.87~8, the FCC+BCC two-phase solid solution structure is dominant; when VEC>8, HEA tends to form a single-phase face-centered cubic (FCC) solid solution structure [8].
The values ​​of atomic radius (r) and valence electron concentration (VEC) corresponding to each element, as well as the mixing enthalpy ( ) between different elements, are shown in Tables 5 and 6. According to the above formulas (3) to (7), the values ​​of ΔSmix, ΔHmix, δ and VEC of the FeCoNiCrMo0.2 high entropy alloy coating were calculated respectively, and the calculation results are shown in Table 7. It can be clearly seen from Table 7 that the VEC value of the HEA coating is greater than 8, and the coating only tends to form a single-phase FCC structure. This further proves that there is only a single FCC solid solution phase in the coating, which is consistent with the results obtained from the XRD spectrum. This shows that the formation process of the solid solution phase composition in the FeCoNiCrMo0.2 coating is accurate in judging and predicting by ΔSmix, ΔHmix, δ, Ω and VEC. Other process variables such as powder feeding rate have little effect on the formation of the solid solution phase type.

2.3 Microstructure
The SEM microstructures of the FeCoNiCrMo0.2 coatings prepared at different powder feeding rates are shown in Figure 6. As can be seen from the figure, different grain structures appear in the same coating. This is because during the cladding process, the five main elements in the HEA coating react with each other, and after cooling, different trace metal compounds are formed, such as Fe-Ni, Cr-Ni, and Cr-Mo series compounds. The SEM microstructures in Figure 6 show that no obvious defects such as pores and cracks appear in the HEA coatings prepared at powder feeding rates of 6 g/min, 12 g/min, 15 g/min, and 18 g/min, and they have good metallurgical bonding with the 316L substrate [24]. As can be seen from Figures 6a and 6b, in the FeCoNiCrMo0.2 coating, the microstructure presents a columnar grain structure that grows perpendicular to the interface, the grain distribution is relatively tight, and the unit cell grows in a directional manner. This shows that at a powder feeding rate of 6 g/min, the substrate absorbs more laser energy, making the temperature distribution in the molten pool uniform, which is beneficial to improving the uniformity of the HEA coating.

It is clear from Figure 6c~d that when the powder feeding rate rises to 12 g/min, the volume of a single columnar crystal in the coating area gradually decreases, most of the grains are transforming from columnar crystals to cellular crystals, the grains become denser and more disordered, and grain boundary strengthening occurs. This shows that there is a large temperature difference between the 12 g/min coating and the 316L substrate, and finer cellular crystals have been formed after directional solidification along the temperature gradient direction. In Figure 6d, it can be observed that the eutectic structure state appears in the microstructure, and the columnar crystals gradually transform into new fine dendrites. From Figure 6e~f, it can be seen that when the powder feeding rate is 15 g/min, the microstructure of the coating is mainly composed of fine cellular crystals, the distribution of coarse columnar crystals is relatively small, and a larger number of columnar crystals have been transformed into fine dendrite structures, the grain boundaries are denser, and the growth direction is chaotic and disordered. This shows that the powder feeding rate of 15 g/min increases the proportion of high entropy alloy particles in the molten state in the coating, which can bring more intermetallic compound nucleation to the coating, such as Cr-Fe and Cr-Ni series. The driving force of this nucleation increases the number of nuclei of strengthening phases such as chromide and promotes their continued expansion. This enhances the solid solution strengthening and fine grain strengthening effects of the coating, and improves the hardness and strength of the coating.

From Figure 6g~h, it can be seen that when the powder feeding rate increases to 18 g/min, the coating still has fine cellular crystals, but the coarse columnar crystals increase, the grain structure is a coexistence of columnar crystals and cellular crystals, and the microstructure is in a multi-directional and disordered state [25]. The grain boundaries are more obvious and the organizational structure is relatively sparse, indicating that the grain structure of the coating is not dense enough. This shows that the microstructure of the HEA coating is closely related to the laser process parameters. The powder feeding rate can make the coating organization finer, promote the uniform distribution of the strength and toughness of the coating, and improve the hardness and wear resistance of the coating. In addition, for the high cooling rate laser cladding forming process, the microstructure of the coating during solidification is not only related to the temperature gradient (G) between the solid-liquid interface, but also to the rapid solidification rate (R) during the cladding process [26]. There is a large temperature gradient (G) between the 316L steel substrate and the HEA coating, and the heat of the molten pool is mainly dissipated through the substrate during solidification. The larger the G/R value, the more unstable the solid-liquid interface becomes, the coarser the grain structure becomes, and many columnar crystals are formed; conversely, the finer the grain structure becomes, forming a cellular shape. In addition, during the solidification of the cladding layer, the diffusion of high entropy alloy elements often occurs at the boundary between the cladding layer and the substrate, and runs through the entire solidification process. The inhomogeneity of the elemental composition in the cladding layer will also lead to the inhomogeneity of the grain structure in the molten pool.

In order to further understand the distribution of the five main elements in the coating, the HEA coating at four different powder feeding rates was analyzed by energy spectrum (EDS). Point A~E were selected in the SEM area of ​​Figure 6 for point composition analysis. At the same time, EDS mapping analysis was performed on a part of Figure 6d and named Area B. The test results are shown in Figure 7 and Table 8. It can be clearly seen from the test results that the percentage of Fe content in the four coatings increased compared with the original powder. The smaller the powder feeding rate, the higher the percentage of Fe element. As shown in Table 6, on the one hand, this is because the absolute value of the mixing enthalpy (ij ΔHmix) of Fe and the other four metal elements is small, and it is not easy to form compounds with other metal elements, but it is easy to occur segregation between dendrites, so its mass percentage is relatively high. On the other hand, it is also related to the 316L substrate. Due to the high content of Fe in the 316L stainless steel matrix, under the action of the high-energy laser beam, the molten pool has greater fluidity, and the metal elements and their compounds undergo convection in the molten pool, so that part of the Fe element diffuses into the coating, resulting in a high Fe content in the coating [27]. The smaller the powder feeding rate, the thinner the HEA coating thickness. As the Fe element in the substrate participates in the molten pool convection, a large amount of Fe elements are more easily diffused into the entire coating, resulting in a higher Fe content. On the contrary, the larger the powder feeding rate, the thicker the coating thickness, and the less susceptible to the molten pool convection. Combining Figure 7 and Table 8, it can be seen that the main element elements of the five HEA powders at different powder feeding rates can achieve uniform distribution in the HEA coating. Therefore, laser cladding technology can achieve uniform distribution of elements in FeCoNiCrMo0.2 high entropy alloy. In addition, the above analysis shows that during the laser cladding process, the HEA elements and the substrate elements have been fully mixed in the molten pool, which forms a good metallurgical bond between the HEA coating and the 316L steel substrate. In general, the powder feeding rate can transform the internal structure of the coating from regular, coarse columnar crystals to disordered, fine cellular crystals, and the uniformity of element distribution also changes. Among them, the HEA coating prepared at 15 g/min has the densest microstructure and relatively uniform element distribution.

2.4 Microhardness
The hardness of HEA coatings with five different powder feeding rates was measured to estimate the tribological properties of the coatings[28]. Figure 8a shows the microhardness distribution curves of HEA coatings prepared at different powder feeding rates along the depth direction. As can be seen from the figure, the microhardness is distributed in a gradient from the coating to the substrate. This is because the rapid melting and cooling during the laser cladding process prevents the grains from growing rapidly, resulting in obvious differences in the microstructure from the top to the bottom of the coating[29]. Combined with the XRD in Figure 5 and the SEM in Figure 6, it can be concluded that the different distribution of FCC phase structure and microstructure in the HEA coating has a significant effect on the hardness of the coating[30−31]. In the coating area, the highest cross-sectional microhardness of the HEA coatings with a rate of 6 g/min, 9 g/min, 12 g/min, 15 g/min, and 18 g/min reached 393.95 HV, 383.25 HV, 402.06 HV, 439.99 HV, and 387.04 HV, respectively, which were 1.97 times, 1.91 times, 2.01 times, 2.20 times, and 1.93 times the hardness of the 316L substrate (about 200 HV), respectively. This shows that the FeCoNiCrMo0.2 coating can significantly improve the microhardness of 316L steel and achieve the purpose of strengthening the inner surface of the part. This is mainly due to the formation of some hard phases between elements under the high-energy laser beam, such as Cr-Ni, Mo-Ni and other compounds; on the other hand, it is because the Cr element is dissolved in the FCC phase of the coating. The addition of the Cr element can cause a large lattice distortion in the HEA coating, which significantly improves the hardness of the FCC phase in the coating [32]. However, different powder feeding rates also lead to differences in the hardness of each coating.

In addition, the fineness of the microstructure in the coating has a direct effect on the hardness. Combined with the SEM image in Figure 6, the hardness curve distribution in Figure 8a can be explained. When the grains are finer and the grain boundaries are wider, the dislocation slip phenomenon in the crystal will be hindered, and the hardness of the coating will be improved. The relationship between hardness and grain size can be well expressed by the classic Hall-Petch equation (8) [32]: See formula (8) in the figure. In the formula, H is the hardness value of the sample, D is the average grain size of the sample, H0 is the inherent hardness of the sample, and KHP is the Hall-Petch coefficient. According to other research results [32], the two constants (H0 and KHP) of the HEA coating at room temperature are approximately 125HV and 494, respectively. However, related studies have shown that it is inaccurate to estimate the hardness of the sample using only the grain size, and it is also necessary to combine the XRD phase structure and other microstructural characteristics for evaluation [31]. When the powder feeding rate is 6 g/min, the microstructure of the coating is mainly composed of fine cellular crystals and relatively fine columnar crystals, which makes the coating have a higher hardness. However, due to the inconsistent grain size of the columnar crystals, the hard phases such as Cr-Fe are unevenly distributed in the coating, resulting in fluctuations in the hardness under this parameter, which reduces the mechanical stability of the coating. When the powder feeding rate increases to 15 g/min, the grain structure in the coating is mostly dominated by fine cellular crystals, the organizational structure is refined, and there are more fine grain strengthening effects, which increases the hardness. When the powder feeding rate continues to increase to 18 g/min, due to the increase in columnar crystals in the coating. There are a large number of columnar crystals and cellular crystals in the coating, which makes the hardness distribution of the coating unstable and the hardness decreases. Combined with the XRD analysis results, the reason for the high hardness of the FeCoNiCrMo0.2 high entropy alloy coating is that the five main elements in the alloy HEA coating react to form trace metal compounds, such as Fe-Cr and Cr-Ni series compounds; at the same time, the lattice distortion caused by the different atomic radii between the elements will also change the hardness. In the heat affected zone (HAZ), that is, the transition zone between the substrate and the HEA coating, the microhardness of the coating keeps steadily decreasing along the depth of the heat affected zone, and finally tends to the hardness value of the substrate steadily. This is because the laser irradiation causes a small amount of melting of the substrate metal, and the elements in the HEA coating are diluted into the 316L substrate, so that the interface between the coating and the substrate presents a good metallurgical bonding. At the junction, the diluted HEA elements form a good solid solution phase with the substrate elements. The closer to the substrate, the more elements are diluted in HEA. In addition, the high entropy alloy at the bottom of the cladding layer has good compatibility with the Fe element in the substrate, which makes the microhardness transition of the bonding zone smooth.

The average microhardness of the cladding layer area at different powder feeding rates at each test point on the cross section was calculated, and the results are shown in Figure 8b. The average microhardness of the HEA coating also reflects the distribution of the hard phase content to a certain extent [32]. The results show that when the powder feeding rate is 6 g/min, the average hardness of the coating area is 369.88 HV. As the powder feeding rate increases, the content of solid solution compounds in the coating gradually increases, and the microhardness gradually increases. When the powder feeding rate is 15 g/min, the average microhardness of the coating reaches the highest value, with a value of 414.70 HV, indicating that the coating contains a large amount of hardness. However, when the powder feeding rate continues to increase to 18 g/min, the average hardness of the coating decreases to 364.71 HV. This is because unmelted powder appears on the surface of the coating, and all HEA powders do not fully achieve the effect of enhancing the hardness of the coating, reducing the solid solution strengthening effect in the coating, resulting in a decrease in the average hardness of the coating. Combined with XRD spectrum analysis, this phenomenon also shows that the powder feeding rate helps to increase the energy of lattice distortion in the FCC solid solution phase, and also enhances the solid solution strengthening effect of the coating during laser cladding.

2.5 Tribological behavior
In order to clearly understand the degree of improvement of the wear performance of the 316L substrate by the FeCoNiCrMo0.2 coating prepared at different powder feeding rates, the wear performance tests were carried out on 6 different samples [33−35]. Figures 9a~b show the friction coefficient curves of the HEA coating and substrate prepared at different powder feeding rates as a function of wear time, as well as the wear weight loss bar graphs of different samples before and after the test. The schematic diagram of the wear process is shown in Figure 9c. It can be seen that under the room temperature dry sliding friction test, the five coatings and substrates all have two stages: initial running-in and stable running-in. From Figure 9a, we can notice that the HEA coating and 316L substrate prepared at 6~18 g/min showed a sharp increase in friction coefficient in a very short time at the beginning of the friction test, and then stabilized. This is because in the initial wear stage, the contact area between the friction pair Si3N4 ceramic ball and the smoother coating surface is small, making it easy for the friction pair to embed into the coating surface. As the wear test continued, the contact area between the friction pair and the coating continued to increase, changing from point contact to surface contact, resulting in severe uneven wear. This led to a sudden increase in the friction coefficient. During this period, the friction pair and the sample performed reciprocating friction motion, and the friction pair and the hard phase in the coating and the micro-convex bodies on the surface of the sample continued to collide and rub. The friction pair and the wear surface adapted to each other, generating different contact stresses, which led to a certain degree of friction coefficient fluctuation in the initial wear stage. As the test continued, the hard phases such as Cr-Ni and Cr-Fe in the coating were gradually worn flat, and abrasive wear was prone to occur at this time. When the actual contact area between the friction pair and the coating continued to increase to a certain extent, the wear surface of the sample and the friction pair reached a dynamic equilibrium state, and the friction coefficient tended to stabilize and entered the stable running-in stage. The running-in time of the 6 g/min coating was about 510 s, indicating that the hard phase content in the coating was unevenly distributed at this time, which reduced the mechanical stability of the coating. The running-in time of the 9~18 g/min coating was about 150 s. The results show that the coating with a speed of 9-18 g/min has good tribological stability. In the stable running-in stage, the friction coefficient curves of 316L substrate, 6 g/min, 9 g/min, 12 g/min, 15 g/min, and 18 g/min fluctuate around 0.83, 0.68, 0.65, 0.61, 0.58, and 0.63, respectively. It can be seen that the friction coefficient of the HEA coating is significantly lower than that of 316L steel, and the coating with a speed of 15 g/min has the smallest friction coefficient. This is because the HEA coating is rich in FCC phase, and the solid solutions such as Cr-Ni and Cr-Co in the FCC phase have the characteristics of high hardness. In addition, the microstructure of the coating with a speed of 15 g/min is fine and dense, and has good toughness. It also shows that the FCC phase metal compounds are relatively evenly distributed in the coating, and the grain boundaries are strengthened by rapid cooling, which reduces the friction coefficient and can effectively resist the reciprocating friction of Si3N4 ceramic balls.

The wear weight loss of the HEA coating and substrate before and after the wear test was measured, and the measurement results are shown in Figure 9b. The wear weight of the HEA coating and 316L substrate samples prepared at 6 g/min, 9 g/min, 12 g/min, 15 g/min, and 18 g/min was 2.861 mg, 2.777 mg, 2.587 mg, 2.456 mg, 2.713 mg, and 3.442 mg, respectively. The mass loss of the coating is significantly less than that of the 316L substrate, which indicates that the solid solution strengthening phase in the coating does enhance the coating’s ability to resist deformation and damage in dry sliding wear tests. The wear test results show that due to the differences in the phase composition and microstructure of the HEA cladding layer prepared at different powder feeding rates, the friction coefficient and wear mass loss of the coating are very different. The formula for calculating the wear rate K is shown in formula (9) [35]: See formula (9) in the figure. In formula (9), V represents the wear volume (mm3), N represents the applied load force (N), and d represents the relative sliding distance (m). The size of the wear rate K reflects the excellent degree of the sample’s wear resistance. The smaller K is, the smaller the wear volume of the sample is, the smaller the wear loss is, and the sample is more wear-resistant. According to Figure 9b, the order of the wear rate K is: 316L > 6 g/min > 9 g/min > 18 g/min > 12 g/min > 15 g/min. Therefore, the wear resistance of the sample is: 316L < 6 g/min < 9 g/min < 18 g/min < 12 g/min < 15 g/min. This proves that the relationship between wear volume and hardness conforms to the Archard wear law, indicating that hardness can be used to a certain extent for the wear performance of HEA coatings. Of course, hardness distribution, elastic modulus, toughness, organization and phase composition also affect the tribological behavior of alloy materials.

2.6 Wear mechanism
Figure 10 shows the SEM microstructures of the HEA coating samples prepared with 316L substrate, 6 g/min, 12 g/min, 15 g/min, and 18 g/min after wear at different magnifications. It can be seen that the main wear mechanisms in the coating include adhesive wear and abrasive wear. The presence of abrasive wear indicates that a small amount of material transfer occurred on the surface of the HEA coating during the wear test, and some wear debris with higher hardness acted as abrasives between the sample and the friction pair, causing scratches and micro-cutting on the coating surface, resulting in material loss. The plowing effect on the wear track can prove the occurrence of abrasive wear. At the same time, the HEA coating that fell off due to wear subsequently formed wear debris. Among these wear debris, some wear debris with higher toughness and lower hardness were repeatedly rolled by Si3N4 balls and adhered to the coating surface, forming adhesive wear. In fact, adhesive wear has a certain protective effect on the wear resistance of the coating, reducing the friction coefficient and mass loss of the coating. However, excessive adhesive wear will cause a large amount of peeling on the sample surface, which will also have a negative impact on the wear performance of the coating.

It is clear from Figure 10a~b that the wear surface of the 316L substrate is relatively rough, and the wear characteristics mainly include a large amount of peeling, adhesive wear and strong plastic deformation. There are a large number of irregular pits and deep furrows with attached wear debris on the wear marks on the substrate surface. This is because the substrate has a low hardness and high viscosity. The substrate is easy to contact with the friction pair and produce plastic deformation. The tearing of the adhesion point produces a large number of pits and wear debris, and the wear resistance is poor [36−38]. At the same time, cracks also appear on the surface morphology. This is because the substrate is subjected to cyclic contact stress in the local area during the wear process, and repeated deformation leads to crack initiation and expansion. This also proves that the thermal fatigue resistance of the substrate is poor during the repeated friction and heat generation process of the friction pair. As shown in Figure 10c~h, compared with the 316L substrate, the wear track surface of the HEA coating is smoother and the morphology is improved after wear. It shows that the hard phases such as Cr-Ni, Cr-Fe, and Mo-Ni in the coating enhance the solid solution strengthening effect of the coating, play an important role in the coating’s resistance to plastic deformation and damage, and make the local deformation of the coating under the action of contact force and shear stress very small, which significantly improves the tribological properties of the material. In addition, the lattice distortion caused by different atomic radii in the HEA coating also plays a positive role in improving the wear resistance of the coating [39-40]. When the powder feeding rate is 6 g/min, the peeling in the coating is reduced, and a large number of deeper furrows and plastic deformation are replaced. It shows that the coating has experienced more severe abrasive wear and adhesive wear. Under the action of the Si3N4 friction pair, the wear debris and oxides produce considerable friction with the coating during the friction process. This is because the 6 g/min HEA coating has a lower hardness and is more susceptible to the influence of abrasive particles, debris, etc. in the coating, forming severe furrows. Deeper furrows are generally distributed on materials with lower hardness, which is consistent with the theory of Li et al. [25,41]. It is worth mentioning that when the powder feeding rate is in the range of 12~15 g/min, the wear mechanism is still composed of adhesive wear and abrasive wear. However, the surface of the HEA coating after wear is flat and the wear depth is shallow. Wear debris and a small amount of peeling appear on the worn surface, accompanied by shallow plowing. The wear characteristics in the coating are mainly slight scratches and a small amount of plastic deformation caused by abrasive wear. Compared with the HEA coating at 6 g/min, the wear morphology of 12~15 g/min is smoother, indicating that the fine cellular crystals and harder Cr-Fe, Cr-Ni and other compounds in the coating jointly improve the wear performance of the FeCoNiCrMo0.2 coating, making the coating have better anti-ploughing performance. Among them, the HEA coating of 15 g/min has the best resistance to wear and plastic deformation. When the powder feeding rate continues to increase to 18 g/min, the failure form of the coating after wear is still mainly peeling, abrasive wear and a small amount of plowing. Among them, there are more peeling features on the worn surface, and there are also some residual particles. This shows that the coating material undergoes plastic deformation during the sliding process of the friction pair and adheres to the surface of the friction pair [42]. As the wear experiment progresses, these adhesive substances gradually form and remain on the sample surface. And due to the poor plowing resistance of the coating, the sample cannot withstand the shear force caused by the repeated rolling of the Si3N4 ball. Therefore, more peeling is formed on the surface. This once again shows that the main wear mechanism of the 18 g/min HEA coating is adhesive wear, and the wear resistance is relatively poor.

2.7 Electrochemical corrosion test
Figure 11 shows the potentiodynamic polarization curves of the FeCoNiCrMo0.2 high entropy alloy coating prepared at different powder feeding rates. The polarization curves characterize the pitting corrosion resistance of the HEA coating with different powder feeding rates in 3.5wt% NaCl electrolyte at pH=7. The five polarization curves in the figure show that the five HEA coatings with different powder feeding rates can form a stable passivation zone in the electrolyte. The occurrence of passivation can increase the corrosion resistance of the coating. The self-corrosion potential (Ecorr) and self-corrosion current density (Icorr) after Tafel fitting are displayed in the horizontal and vertical directions of the polarization curve, respectively. Ecorr refers to the probability of the coating being corroded. The closer its value is to a positive value, the lower the possibility of corrosion. The self-corrosion current density of Icorr can reveal the corrosion process of the sample. The lower the value of Icorr, the slower the corrosion rate will become[43]. Icorr is more representative of the anti-pitting performance of the coating, followed by Ecorr. Table 9 shows the parameters of the potentiodynamic polarization curves of the five HEA coatings during the corrosion process. Combining Figure 11 and Table 9, it can be seen that when the powder feeding rate is 6 g/min, the passivation film density of the HEA coating is small, and the self-corrosion current density Icorr in the passivation zone is large, which leads to a relatively weak protective effect of the passivation film. This is because at a lower powder feeding rate, the HEA powder absorbs more laser energy, causing a small amount of powder to burn, reducing the solid solution strengthening effect on the coating. And from the previous analysis, the coating prepared at a lower powder feeding rate is easily affected by the molten pool convection, which also reduces the corrosion resistance of the coating. When the powder feeding rate is 9~15 g/min, the Icorr of the coating is significantly reduced, and the passivation film becomes denser. The passivation area of ​​the polarization curve of the HEA coating shows a rapid increase and decrease, which reflects the situation that the local passivation film is damaged and quickly recovered, and also reflects that the passivation film has excellent self-healing properties. When the powder feeding rate is 18 g/min, due to the increase of unmelted powder in the coating, some powders are not completely molten, which reduces the content of reinforcing phase in the coating, and finally increases Icorr.
In general, compared with 6 g/min and 18 g/min, the Icorr of HEA coating prepared at 9~15 g/min is reduced by 1~2 orders of magnitude, showing excellent anti-pitting performance.
In addition, it can be clearly seen from the opening size of the polarization curve that the larger the opening, the lower the Icorr, indicating that the passivation interval generated by the coating is larger, which is more beneficial to the anti-pitting performance of the coating. Although the opening of the polarization curve of the HEA coating reaches the maximum when the powder feeding rate reaches 6 g/min, due to its large Icorr of 3.87×10−7 A•cm−2, it shows that the degree of passivation of the coating is small during the corrosion process, and the rate of Clˉ erosion of the coating is large, which is not conducive to the pitting resistance of the coating. The polarization curve shows that the HEA coating with a surface temperature of 15 g/min has the lowest Icorr value and a relatively large open passivation interval, indicating that the coating has more anti-corrosion substances, such as Cr2O3 and MoO2, in the in-situ generated passivation film on the sample surface in 3.5% NaCl solvent, and has relatively excellent corrosion resistance.

3 Conclusions
In this study, FeCoNiCrMo0.2 high entropy alloy coating was prepared by inner wall laser cladding equipment. The effects of different powder feeding rates on the macroscopic morphology, phase structure, microstructure, hardness, tribological behavior and electrochemical corrosion performance of HEA coating on the inner surface of 316L steel pipe were investigated. The main conclusions are as follows:

1) The powder feeding rate will greatly affect the thickness of HEA coating, and will also have a certain effect on the dilution rate and the convection of alloy elements in the molten pool. HEA coating is composed of FCC phase. The powder feeding rate will not change the phase type of HEA coating, but will affect the content of solid solution phase in the coating, change the intensity of diffraction peaks, and even cause lattice distortion of the coating, affecting the lattice constant.

2) A powder feeding rate of 12~15 g/min can refine the crystalline structure of the coating, change the preferred orientation of the grains, and thus improve the toughness and strength of the coating. In the coating with a feed rate of 15 g/min, the number of coarse columnar crystals decreased, and was replaced by fine cellular crystals, and the grains became denser and more disordered.

3) The powder feeding rate can improve the microhardness and tribological properties of the HEA coating to a certain extent. The HEA coating prepared at 15 g/min has a high content of solid solution hard phase, and the average microhardness is as high as 414.70HV, which is about 2.07 times the hardness of the substrate. And the friction coefficient of the coating is stable at about 0.58, with the best tribological properties. The powder feeding rate did not change the wear mechanism of the HEA coating, which was still dominated by abrasive wear and adhesive wear, but the powder feeding rate of 12~15 g/min would make the wear morphology of the coating smoother, with less peeling, and reduce the loss of sample quality.

4) Electrochemical corrosion tests show that the powder feeding rate can reduce the self-corrosion current density Icorr of the HEA coating. In 3.5wt% NaCl solvent, the Icorr of HEA coatings prepared at 9-15 g/min decreased by 1-2 orders of magnitude compared to 6 g/min and 18 g/min, showing excellent anti-pitting performance. The HEA coating at 15 g/min has the lowest Icorr value and a relatively large passivation interval, indicating that during the corrosion process, the in-situ generated passivation film on the surface of the coating contains more anti-corrosion substances, such as Cr2O3 and MoO2, which makes it have relatively excellent corrosion resistance.