The roles of microstructural anisotropy in tribo-corrosion performance of one certain laser cladding Fe-based alloy

Because of the microstructural anisotropy for laser cladding materials, the tribo-corrosion performance can vary significantly with different directions. In this study, one certain Fe-based coating was fabricated by laser cladding. To study the effects of anisotropy, three working surfaces (0°, 45°, and 90° to the building direction) were machined from the laser cladding samples; as-cast samples with an approximately homogeneous structure were prepared as controls. The tribo-corrosion tests were conducted in a 3.5 wt% NaCl solution with varying normal loads (5, 10, and 15 N). The results demonstrated that the 45° surface has superior friction stability, corrosion resistance, and wear resistance. This was directly related to the crystal orientation and grain boundary density. In addition, a refined microstructure may enhance tribo-corrosion properties by increasing deformation resistance and decreasing surface activity.


Introduction
Owing to the excellent surface-degradation resistance, stainless steel has been widely used in many corrosive environments, including nuclear power plants, chemical reactors, marine-industrial settings, and deep-sea oil-drilling rigs [1,2]. However, its poor mechanical strength can result in early failure due to friction, wear, and contact fatigue, which seriously limits the applications in practice [3].
Wear-traditionally a major research topic in mechanical engineering-has been extensively studied over the past decades. Many significant progresses were achieved. For instance, a novel approach of single-grain scratching and sliding at a nanoscale depth of cut and 40.2 m/s, respectively, was proposed to investigate the fundamental wear mechanisms [4][5][6]. Bridging the gap between microscale phenomenon and macroscale applications, Zhang et al. [7] firstly achieved macroscopic superlubricity on the macroscale surfaces under ambient conditions at room temperature. The usual strategy adopted to combat wear failure of stainless-steel parts in corrosive surroundings involving restoring size, morphology, and function by fabricating a coating with optimal mechanical properties and corrosion resistance [8]. Among the many new materials (ceramic, Ni-based, and Co-based) developed for this purpose, Fe-based alloy is proven to be one of the best in terms of its cost advantages and the tunability of its properties [9].
Laser cladding is a versatile manufacturing route for fabrication of Fe-based coatings with enhanced microstructural features that are difficult to obtain by conventional processing [10,11]. It is because that the inherent localized heating and rapid cooling (>10 3 K/s) tend to refine the microstructure and suppress elemental segregation, which is beneficial to the improvement of mechanical properties [12][13][14]. References [15][16][17] have provided detailed information about the influence of processing parameters, such as laser power, scanning speed, and spot size, on the structure and mechanical properties. However, there are very few reports and limited understanding of the degradation mechanisms for tribo-corrosion.
Tribo-corrosion is the synergistic combination of wear and corrosion that leads to accelerated material degradation. References [18,19] have demonstrated that the dominant factors of tribo-corrosion resistance are more complex than those of purely mechanical wear or purely chemical corrosion; high compactness, effectively crack-controlled wear, and dynamic equilibrium between de-passivation and re-passivation on the worn surfaces are indispensable for good tribo-corrosion resistance. These factors depend heavily on the solidification process for laser cladding materials [20]. For example, the path of heat conductivity has obvious directionality owning to the existence of a cool substrate, and therefore causes the anisotropy of microstructure and properties.
Shittu et al. [21] studied the effect of sliding direction on the tribo-corrosion response of additively manufactured high-entropy alloy, selecting a constant cross-section as the working surface. They found that the wear volume loss and wear rate vary with sliding directions. However, the reasons still need to be investigated. Indeed, the microstructural anisotropy manifests itself not only along different sliding directions of a single surface but also on different working surfaces, although little research on this has yet been published. Additionally, the higher temperature gradient and thermal stress during rapid cooling process may seriously impact the crystal orientation. The corresponding mechanisms influencing tribocorrosion remain to be clarified.
In this study, a Fe60 alloy coating was prepared on stainless steel by laser cladding. Three different surfaces were designed and selected as working surfaces. The tribo-corrosion behavior and the associated structure-property-performance relationships were primarily investigated. As-cast Fe60 samples were simultaneously prepared as a control. The microstructure, composition, and properties were determined through standard techniques. The results can be used as reference data to determine reasonable processing schedule in practice.

Materials
Commercial self-fluxing Fe60 powder (Tianjin Cast Gold Technology & Development Co., Ltd., China) was selected as the target material for a strengthening coating on American Iron and Steel Institute (AISI) 304 stainless steel. The size of the powder particles is in the range of 45-80 μm. The chemical compositions of Fe60 are shown in Table 1. The Cr element facilitates the formation of a passive film, which is largely responsible for the strong corrosion resistance; the other elements in the alloy are primarily used to control mechanical properties [22].
A laser metal-deposition system (LDM8060) equipped with a 2,000-W fiber laser (Nanjing Zhongke Yuchen Laser Technology Corporation, Ltd., China) was employed to prepare the coating. A photograph and schematic diagram of the setup are shown in Figs. 1(a) and 1(b), respectively. The optimized processing parameters, chosen in advance on the basis of our research experience, are summarized in Table 2. Powders were fed to the melt pool by a co-axial nozzle. The standoff distance between the nozzles and workpiece was kept at ~9 mm to maintain deposition efficiency and stability. The processing environment was filled with high-purity argon, forming a protective atmosphere.
The as-received laser cladding sample (dimensions: 80 mm × 60 mm × 12 mm) is shown in Fig. 2


, defined as 0°, 45°, and 90°, respectively. The microstructural morphologies corresponding to different surfaces were obtained by metallographic examination (Fig. 2(b)). They exhibit an obvious anisotropy. For example, on the 90° surface, many coarse dendrites are observed, which are defined as primary dendrites. Their directional growth, towards the pool center, was strictly controlled by the heat conduction during solidification [20,23]. As for the 45° surface, many secondary dendrites are wholly or partly exposed, showing a poorer directivity. The 0° surface shows the transverse sections of primary dendrites and the longitudinal sections of secondary dendrites, whose directivity is almost eliminated.
To explore the relationships between microstructure and tribo-corrosion performance further, as-cast Fe60 samples with approximately homogeneous microstructures were prepared as experimental controls.
Prior to the foundry procedure, alloy powders were re-melted five times to ensure homogeneity [24].

Material characterization
Before tribo-corrosion testing, the microstructural characteristics of different specimens were determined by means of the electron backscatter diffraction (EBSD), scanning electron microscopy (SEM; SUPRA 55, Zeiss), transmission electron microscopy (TEM; ARM300F2, JEM), and optical microscopy (OM; Axiovert 40 MAT, Zeiss). The microhardness of samples was measured with nano-indentation (NanoIndenter XP, MTS) under a load of 1,000 g. The worn surfaces after tribo-corrosion testing were observed by the SEM, and their compositions were analyzed by using the energy-dispersive X-ray spectroscopy (EDS) and X-ray photoelectron spectroscopy (XPS; AXIS 165, Kratos). The volume losses after testing were evaluated by a three-dimensional (3D) optical interferometry (New view 5022, ZYGO). The phase analysis was conducted by the X-ray diffraction (XRD; PANalytical Empyrean).

Tribo-corrosion testing methods
The tribo-corrosion test was carried out on a   [25]. A schematic diagram of the test setup is shown in Fig. 3(a). Prior to friction, the testing surface was ground with SiC paper to achieve a roughness (R a ) of 0.6 μm. A Si 3 N 4 ball with a diameter of 6 mm was selected as the counterpart.
The tests were operated under various loads (5, 10, and 15 N) with a constant frequency of 2 Hz and a duration of 2,400 s. Figure 3(b) shows how the sliding tracks (length = 4 mm) was oriented on different testing surfaces. On each surface, the test was repeated three times to ensure the reliability of data. The sliding direction was 45° from OY  , which was known from Ref. [21] to be optimal. The electrochemical cell consists of the Fe60 sample as working electrode (WE), platinum wire as counter electrode (CE), and saturated calomel as reference electrode (RE). All the three electrodes were placed in a 3.5 wt% NaCl solution at room temperature. During test, the friction coefficient and open circuit potential (OCP) were monitored. After test, the wear performance was evaluated by the volume loss.

Microstructures of laser cladding and as-cast samples
The microstructural features of Fe60 prepared by laser cladding and foundry techniques are summarized in Figs. 4(a)-4(d). Two adjacent melt pools with an overlap rate of 45% are visible in Fig. 4(a). At the melt pool boundary, many columnar grains were formed due to the large thermal gradient along the radial direction [26]. With the decreasing solidification rate towards pool center, columnar grains began transforming into equiaxed ones. Some sub-grain cells near the pool boundary have an average size of 2-3 μm ( Fig. 4(b)). As reported in Refs. [12][13][14], this metastable structure was typically caused by rapid cooling. ). It can be observed that the laser cladding sample possesses a more obvious anisotropy as compared to the as-cast one. Namely, the [111] and [001] orientations of grains are approximately parallel to OX


. This was owed to the rotation of grains under high thermal stress during rapid cooling process [8,27]. In addition, the grains in the laser cladding sample were significantly refined with an average grain size of about 42 μm, while the corresponding value for the as-cast one was about 104 μm. The greater degree of refinement was responsible for the laser cladding sample's superior mechanical properties. On the one hand, the high density of grain boundaries could block dislocation slip and finally improve deformation resistance. On the other hand, grain boundaries acted as strong obstacles to propagation of cracks, which was beneficial to material strength [12][13][14].
The phase analysis results in Fig. 5(a) indicate the transition of composition and structure. Both of the  samples consist of an austenite matrix (γ) and a Cr-rich phase. The respective crystal structures of the phases (Fig. 5(b)) are face-centered cubic (Fcc) and body-centered tetragonal (Bct). The relative diffraction intensities of the Cr-rich phase (peaks (204) and (310)) in laser cladding Fe60 exhibit a decrease compared to those in the as-cast one. This may be attributed to the strong restriction on the precipitation during rapid solidification [28]. In addition, the diffraction peaks of  in the laser cladding specimen show a slight shift to the right, and the intensities of the (220) and (311) peaks are diminished. This also confirms the existence of strain associated with thermal stress.

Effects of anisotropy on tribo-corrosion performance
During test, the friction coefficients of laser cladding samples on different working surfaces are monitored and collected in Fig. 6(a). Initially, the friction coefficients for all samples increased gradually during the running-in stage. This resulted from the establishment of conformable contact between friction pairs, which would lead to the increase of real contact area [29,30]. As the friction continued, a steady state was finally reached for all samples. The 45° sample exhibited a better friction stability. Besides that, the friction coefficients of the 45° specimen corresponding to the steady stage are a little lower than those of the others. The friction performance is generally dominated by interactions at tribology interfaces, and well- connected with the morphologies of worn surfaces. Thus this part will be discussed together with wear performance and worn surfaces below. The OCP curves for different working surfaces are plotted in Fig. 6(b). It can be seen that the original values for different specimens are extremely similar. This means that the anisotropic grains had little direct influence on corrosion behavior. As wear begins, a sharp drop to more negative values is exhibited. This corresponds exactly to the increasing friction coefficient in the initial stage ( Fig. 6(a)). Similar results have been reported in Refs. [31,32], where the wear damage to the passive films outside the specimen surface was considered to be the crucial factor. Subsequently, as the friction tends to be stable, the OCP values begin to rise. This may be attributed to the re-passivation [33]. In addition, the 45° surface possesses more positive OCP values when the friction becomes more stable. The difference in OCP values for different specimens can be partly explained in terms of the competition between surface damage and re-passivation [19,21]. Namely, the higher mechanical wear resistance, the more positive OCP values.
After the whole test, the volume losses of different specimens were measured. Figure 6(c) shows clearly that the 45° specimen has the highest tribo-corrosion resistance. To understand why, the longitudinal cross sections beneath the contact surface were observed by the SEM.  www.Springer.com/journal/40544 | Friction incline is directional and controlled by the sliding direction. This reminds us that mechanical plowing may play an important role in material wear. Therefore, the etched cross-sectional microstructures were further analyzed by the focused ion-beam SEM (FIB-SEM). In Fig. 7(d), plastic deformation caused by friction force is obviously demonstrated. As higher deformation resistance means greater wear performance [34], tanα is employed in this study. (Here, α is the angle between the tangent line of the metal-flow curve and the normal of the tribological interface, as shown in Fig. 7(e)). Based on the corresponding FIB-SEM images, the values of tanα for the 0°, 45°, and 90° specimens were measured as 3.3, 0.9, and 2.9, respectively. Namely, the 45° specimen has the highest deformation resistance. This can be explained by the corresponding microstructural features. On the one hand, the high density of grain boundaries ( Fig. 2(b)) could prevent plastic deformation through a pinning mechanism for dislocations [28,35]. On the other hand, it can be deduced from Fig. 8 (which gives the crystal orientation of the γ phase in the sample) and 4(c) (which gives the EBSD results) that the 45° working surface is approximately parallel to (101), and the atomic density of it is ~0.554. This is much lower than that of (001), which is parallel to the 0° and 90° surfaces. Lower atomic density resulted in higher plastic-deformation resistance under shearing stress during relative sliding [36]. The highest deformation resistance of the 45° surface was also responsible for the lower friction coefficient (Fig. 6(a)) by reducing complex mechanical interactions (e.g., three-body abrasion) at tribological interface.
Moreover, Fig. 7(c) exhibits more cracks in the subsurface zone compared to Figs. 7(a) and 7(b). This reveals a wear mechanism transformation, which was responsible for the poor tribo-corrosion performance of the 90° specimen. This may be closely related to the Cr-rich phase that could cause stress concentration under cyclic loads, and thus developed into a crack initiation site [37]. Additionally, the Cr-rich phase with Bct structure has cleavage planes [38,39], along which cracks were easily propagated, enhancing wear. The fracture morphologies of the Cr-rich phase in Figs. 9(a)-9(c) support that the fractured surface parallel to the 90° surface are characterized by obvious brittle features, which can accelerate the horizontal propagation of cracks under shearing stress.
Corrosion behavior is another factor influencing tribo-corrosion performance. Figure 6(b) shows a more positive OCP value for the 45° working surface, i.e., it possesses greater corrosion resistance, which could reduce the interactive effects between wear and corrosion [40]. To verify this, an approximately non-corrosive environment-deionized water (0 wt% NaCl) was used for the same test (Fig. 6(c)). An obvious decrease in volume loss can be observed for all specimens, revealing the importance of corrosion behavior in the stronger tribo-corrosion resistance of 45°. It is claimed that corrosion can accelerate  initiation and propagation of cracks in a corrosive and abrasive environment [19,40,41]. Figure 6(d) exhibits the v during different time periods calculated by Eq. (1): where V is the volume loss during this period, and t is the corresponding time duration. It is evident that v decreased gradually over time. On the one hand, this confirms the re-passive behavior mentioned above, which was beneficial for corrosion resistance.
On the other hand, it was related to the establishment of friction stability, which could reduce the contact stress and consequently the complex mechanical interactions at tribological interface [42].

Effects of normal load on tribo-corrosion performance
The tribo-corrosion responses of laser cladding (45°) and as-cast Fe60 specimens under loads of 5, 10, and 15 N are shown in Figs. 10(a)-10(d). As shown in Figs. 10(a) and 10(b), the friction coefficients corresponding to the steady stage diminish with increasing normal load. This agrees with Refs. [28,29,42] concerning dry sliding friction. According to Coulomb's law, an increase in normal load should increase the friction force [28,29]. However, if the compressive stress simultaneously increases the real contact area between friction pairs disproportionately, the final friction force per unit area will decrease instead. For a given load, the friction coefficient of the laser cladding specimen is lower than that of the as-cast one. As is well known, friction performance is dependent on mechanical properties, chemical composition, operating conditions, and other factors [28,42,43]. In the present study, possible variations in mechanical properties induced by different fabricating techniques were considered as the most important factor. For this, the microhardness (HV) of different samples were measured. Four times were repeated to examine the data reproducibility. The indentation positions and results for the laser cladding specimen are given in Fig. 11(a). The hardness first decreases, and then increases with the coating depth. The average value for the laser cladding specimen (HV = 740) is much higher than that for the as-cast one (HV = 530). As reported in Refs. [28,44], the reduction of friction coefficient at higher hardness is mainly due to the strong deformation resistance, which can reduce drastic abrasion behavior at the tribological interface. In addition, the indentation morphologies of different specimen surfaces are also observed ( Fig. 11(b)). In the as-cast specimen, there is an obvious spring-back after unloading. This implies elastic-plastic hysteresis during friction, which could also consume frictional energy [42]. Figures 10(c) and 10(d) show the OCP curves during test. Because of the destruction of passive film, a decrease with the increasing friction coefficient is revealed. As the frictional stability is gradually attained, the OCP curves rise slightly because of the re-passivation behavior. The original OCPs of the laser cladding specimens (−0.12 V) are higher than those of the as-cast ones (−0.18 V), suggesting a greater corrosion resistance. Because corrosion behavior depends heavily on the surface energy [45,46], the distributions of misorientation angle for different specimens were determined by means of the EBSD (Figs. 12(a) and  12(b)). In the laser cladding sample, the number fraction of 0°-5° low-angle boundaries with lower surface activity is much higher than that in the as-cast sample. This should result in better corrosion performance. Figure 13(a) shows the volume losses of the laser cladding and as-cast samples. An increase with the applied normal load (F) is presented. On the one hand, this can be explained by the decrease of the corresponding OCPs in Figs. 10(c) and 10(d). The poorer corrosion resistance under larger loads could enhance the mechanical damage during friction [18,19]. On the other hand, the severe mechanical plowing, adhesive wear, and even oxidation wear induced by high-energy density may contribute [42,47,48].
To evaluate the load sensitivity, the wear rate (ω) was calculated by Eq.   where L is the total sliding distance. The results are provided in Fig. 13(b). It can be found that the ω of the laser cladding specimen (45°) changes very little with F, while that of the as-cast one increases gradually. Thus, the laser cladding specimen has a lower sensitivity to normal load. This can be explained by the mechanical stability during relative sliding. For instance, Figs. 14(a) and 14(b) show the TEM microstructures in the subsurface zone after tribocorrosion test for different samples. The refined microstructure for laser cladding specimen has a higher dislocation density due to the pinning mechanism of higher grain boundary density, which implies a better work hardening capacity [49] and is useful for maintaining surface integrity.
Above all, the tribo-corrosion performance depends on mechanical properties, corrosion behavior, and testing conditions. Similar results have been reported in Refs. [21,40,[50][51][52] concerning metal materials in a corrosive and abrasive environment. As shown in Table 3, the wear rate of SUS304 (with better corrosion resistance) is lower than that of Q235 under the same testing conditions [52]; SUS304 in distilled water performs better than that in artificial seawater or 0.5 M NaCl solution [40,50,52]; high-entropy alloy CoCrFeMnNi and AlCrFeNiW 0.2 Ti 0.5 with higher strength has a relatively low wear rate [21,52]. It was concluded that the mutual activities of corrosion and wear can accelerate the damage to materials. On the one hand, a metal starts to have its passivation film destroyed or undergo severe plastic deformation in corrosive media under the condition of mechanical wear, and the increased defect density promotes a higher corrosive activity, thereby promoting corrosion [40]. On the other hand, corrosion has a certain role in promoting wear. During sliding contact, the corroded surface first experiences plastic deformation; then deformation accumulation brings low-cycle fatigue to the material, and the surface of the material starts to spall [53], thus forming a large number of cracks on the worn surface.

Morphologies of worn surfaces
The worn surface morphologies of the 0°, 45°, and 90° specimens are presented in Figs. 15(a), 15(b), and 15(c), respectively. Many grooves were generated on all surfaces by mechanical plowing, which were caused by hard particles inserting into the soft material surface during relative sliding (abrasive wear) [29,42]. By contrast, the 45° worn surface exhibits slighter  Artificial seawater [52] www.Springer.com/journal/40544 | Friction abrasive damage, strongly confirming the inference mentioned above that higher deformation resistance is responsible for lower wear rate, better friction stability, and reduced mechanical interactions between hard debris and contact surfaces. In addition, a wear mechanism transformation for the 90° specimen is also revealed in Fig. 15(c). Some cracks were formed under cycling load. This agrees with the cross-sectional morphology in Fig. 7(c). The chemical compositions of different specimens after test were analyzed by the XPS; the high-resolution spectra of Fe 2p 3/2 are given in Figs. 16(a)-16(c). The Fe 2p 3/2 spectra show the presence of Fe 0 (metallic), Fe 2+ , and Fe 3+ oxidation states. According to Refs. [51,54], the Fe 2+ and Fe 3+ oxidation states were estimated to be present as Fe(OH) 3 and Fe 3 O 4 based on the peak position and the splitting energy between the two energy peaks. This can be proven by the O 1s spectra for the three worn surfaces (Figs. 16(d)-16(f)), consisting of three peaks (O 2− , OH − , and adsorbed water), which suggests the formation of metal oxide or hydroxide and contributes to the corrosion characteristics. Moreover, the quantitative results of the XPS spectra reveal that the metal oxide or hydroxide contents for the 0° and 90° surfaces are higher than those for the 45° surface, which led to poorer mechanical properties of contact surfaces, and thus could accelerate the mechanical damage [40,53].
The morphologies of worn surfaces for laser cladding (45°) and as-cast specimens under various normal loads are shown in Figs. 17(a)-17(f). Figure 17(a) presents the topographic characteristics of the 45° specimen under 5 N, where a light polishing behavior was responsible for the smooth surface. As the normal load increases to 10 N, the 45° worn surface in Fig. 17(b) is characterized by obvious grooves. This   Fig. 17(c)), revealing a typical feature of adhesive wear [29,55]. Such deformation indicates that the shearing stress at tribological interface (whether developed by the intermolecular force or by other bonding forces in the real contact area) surpassed the material's yield strength [42]. The chemical compositions of the 45° worn surfaces under different normal loads were analyzed by the EDS; the results are listed in Table 4. The particles embedded in the surface (e.g., spot S1 in Fig. 17(a)) have much higher oxygen contents than the rest of the zone (e.g., spot S2 in Fig. 17(a)), suggesting the existence of oxidation corrosion [21]. However, as the normal load increases from 10 to 15 N, the oxygen contents decrease sharply. Figures 18(a) and 18(b) exhibit the EDS results for the 45° specimens under 10 N (e.g., spot S3 in Fig. 17(b)) and 15 N (e.g., spot S4 in Fig. 17(c)), respectively. To know about the detailed oxidation corrosion, the Fe 2p XPS peaks were also obtained for various normal loads. As shown in Figs. 19(a) and 19(b), at 10 N, Fe exists as Fe 0 , Fe 2+ , and Fe 3+ on the worn surfaces; whereas at 15 N, it is mainly Fe 0 on the worn surfaces. In light of the O 1s spectra in Fig. 16(e), Fe(OH) 3 and Fe 3 O 4 were the main oxidation products at 10 N. By contrast, very little oxidation wear occurred at 15 N. This implies a wear transformation from oxidation and abrasive wear to adhesive wear as the normal load increases. Figures 17(d)-17(f) show the morphologies of worn surfaces for as-cast specimens under various normal loads. It can be seen that the trend with increasing load is similar to that in the laser cladding samples, except that more serious tribo-corrosion took place under a given load. For example, more cracks occurred on the as-cast surface under 5 N (Fig. 17(d)), as compared to Fig. 17(a)); severer corrosion had made the material's surface brittle. Moreover, many oxide particles are observed to have peeled off, leaving pits on the surface. These hard particles could be arrested at the tribological interface and played critical roles in the abrasive wear that followed. In the 10 N case,   Fig. 17(e) imply severer abrasive wear than those of Fig. 17(b). References [28,29,40] attributed this to poor deformation resistance. With a load of 15 N, the worn surface in Fig. 17(f) is unsurprisingly characterized by more serious plastic deformation. This severer adhesive wear also resulted from the lower mechanical strength of as-cast sample.
In summary, the tribo-corrosion mechanisms changed as a function of applied load. Under lower loads, the wear of all the specimens mainly took the forms of abrasive wear and corrosion. As the loads increased, the wear was enhanced by the increasing mechanical damage and oxidation corrosion. Beyond some critical value, the wear mechanisms gradually transformed to an adhesive wear mode. Thus, the worn surfaces exactly support the results of the OCP, volume loss, and friction coefficient measurements discussed above.

Conclusions
This study examined the effects of microstructure anisotropy on the tribo-corrosion performance of laser cladding Fe60 coating. Three working surfaces with different orientations (0°, 45°, and 90° to the building direction) were selected. We arrived at the following conclusions: 1) The microstructural features dominate the mechanical properties of Fe60, and further influence the tribo-corrosion properties through complex mechanical interactions at the tribological interface. The refined microstructure improves the resistance of coating materials to mechanical damage. In addition, the higher density of low-angle boundaries reduces the surface activity, which improves the corrosion resistance.
2) The applied normal load significantly affects the wear mechanism. Under lower loads, material removal is jointly dominated by abrasion and oxidation corrosion. As the load increases, adhesive wear starts to become dominant.
3) The working-surface orientation strongly influences tribo-corrosion properties. This is related to the difference in crystal orientation caused by thermal stress during directional solidification. A working surface at 45° to the building direction of the coating possesses better friction stability and tribo-corrosion resistance.