Abstract
TiAlN and AlCrN coatings are widely used in applications that require high stress resistance. In this work, the tribological behaviour of physical vapour deposition (PVD) TiAlN and AlCrN commercial coatings deposited on AISI martensitic stainless steel is studied. Microstructure of the coatings was analysed by X-ray diffraction (XRD), optical microscopy (OM) and scanning electron microscopy (SEM) and nanohardness was measured. Pin-on-disk and abrasive wear tests were performed. Adhesion was evaluated using Rockwell C Indentation and scratch test. The thickness of both coatings was approximately 3 µm. AlCrN lost 30 and 10 times less volume than TiAlN in pin-on-disk tests, under low and high loads, respectively. The steady friction coefficient value was also lower. This indicated that the AlCrN coating had a better performance under sliding conditions. On the other hand, the mass loss was similar for both coatings under abrasive wear, even under severe conditions. In the scratch tests, TiAlN coating failed at 60 N load and AlCrN coating at 70 N, the latter showing a higher value of critical load. The deformation was similar for both coatings as it could be observed in the profiles obtained by a mechanical profilometer at 60 N, however, AlCrN did not show film delamination. This enhanced performance can be attributed to higher fracture toughness and load carrying capacity, which not only improved the mechanical properties of the coating but also its adhesion to the stainless-steel substrate.
Introduction
Martensitic stainless steels are used in numerous industrial applications, for instance, steam turbines, piston rings, blades, nozzles, where they are exposed to different wear and corrosion conditions which can cause failure or lead to frequent replacements of mechanical parts. 1
As a result of their high wear and corrosion resistance this type of steels is often employed for medical instruments, such as scalpels, scissors, valves and wearing parts of pumps, textile machine components and parts of machinery used in the food and packaging industry. In this case, steel parts are usually used with a small amount of lubrication or in dry running. Other applications include decorative and anti-abrasion coatings for household appliances.2,3 In order to extend the service life and improve surface properties of martensitic stainless-steel components, different treatments can be used, such as surface modification or coatings, for example electroplating, galvanising and nickel and chromium plating. These traditional treatments usually have a negative impact on the environment due to the waste and effluents produced. 4 Nevertheless, there are other techniques which are environmentally friendly, such as plasma assisted techniques. Physical vapour deposition (PVD) coatings can be applied by means of magnetron sputtering, cathodic arc evaporation and pulsed laser deposition. 5 This technology allows the formation of a uniform and dense coating layer of a few microns in thickness, and it enables the adjustment of the chemical composition of the coating as well as the deposition of multilayers of different materials. 6 Among the PVD protective nitride coatings, TiN and CrN have been commonly used for protection against wear of sliding components. Additionally, the introduction of Aluminium into the crystallographic structure of nitrides has proven to increase oxidation resistance and hardness.1,7,8
There are several publications about the tribological behaviour of AlCrN and TiAlN coatings.5–15 However, the behaviour of these coatings deposited on martensitic stainless steels has not been extensively studied. Since the tribological performance depends not only on the coating but also on the substrate microstructure and chemical composition, as well as the deposition method, it is important to consider the whole system: substrate-interphase-coating.1,16–18
In some previous publications about coated martensitic stainless steels, tests have been performed to evaluate their tribological behaviour under mild conditions (low load and pressure and short duration) or for specific applications like improving the durability of the cutting edge of certain tools.1,16–18 Taking into account that the industry is interested in estimating the service life of the treated parts or components, the aim of this work is to evaluate the tribological behaviour, adhesion and abrasion resistance of two commercial coatings under severe conditions, AlCrN and TiAlN, deposited by means of cathodic arc PVD by Oerlikon Balzers, Argentina.
Materials and methods
AISI 420 martensitic stainless steel discs of 25 mm in diameter and 6 mm thickness were used as samples. Its chemical composition, determined by GDOES, is: 0.35 wt-% C, 12.4 wt-% Cr, 0.33 wt-% Mn, 0.40 wt-% Si, 0.03 wt-% P and Fe as balance. The samples were quenched and tempered at low temperature to attain maximal hardness in accordance to usual standards. They were then ground with SiC grit paper from #120 to #1000. The coatings, which are commercially known as Alcrona® and Futura Nano®, are basically AlCrN and TiAlN, respectively. They were deposited by cathodic arc PVD in Oerlikon Balzers Argentina.
Nanohardness and Young's modulus of the coatings were determined using a triboindenter Hysitron with a Berkovich indenter and a load of 9000 µN. The load value was selected to limit the indentation depth to less than 10% of the film thickness to obtain information only about the film properties. 19 In this case, the maximum acceptable indentation depth would be, approximately, 300 nm. The microhardness of the martensitic stainless steel was measured using a Shimadzu HV2 Vickers microhardness tester with a 0.49-N load. The ratios H/E and H3/E2 were calculated for both coatings and the substrate.
The film microstructure was analysed by X-ray diffraction (XRD), using Cu Kα radiation and normal incidence, optical microscopy (OM), and scanning electron microscopy (SEM). The coating thickness was measured by SEM analysis of the cross-section. In order to evaluate the wear behaviour, pin-on-disk tests and abrasive wear tests were performed according to ASTM G99 and G65 standards, respectively. The pin-on-disk tests were carried out with 5 N (Hertzian Pressure = 1.6 GPa) and 10 N (Hertzian pressure = 2.03 GPa) normal loads, 500 m sliding distance, a tangential velocity of 10 cm/s and a 6-mm diameter alumina ball as counterpart. The wear track radius was 5 mm and 7 mm. Friction coefficient was recorded during the entire test. Wear volume loss was calculated from the pin-on-disk track's section obtained using a mechanical profilometer. The abrasive wear tests were performed using AFS 70 sand (mainly SiO2) dried at 100 °C for an hour in the dry sand/rubber wheel apparatus under different conditions: 65 N, 10 min; 65 N, 30 min; 130 N, 10 min; 130 N, 30 min. The samples were weighed before and after the test using a 0.1-mg analytical scale to determine the mass loss. The wear tracks were observed by SEM and OM. Both wear tests were performed on the coated samples and on the non-coated ones to compare the wear resistance.
Scratch and Rockwell C Indentation tests were performed to evaluate the adhesion. The scratch tests were carried out with constant loads of 50 N, 60 N, 70 N, 80 N and 90 N. The tracks were observed by means of SEM and OM. The cutting efficiency factor (fcutting), which relates the valley area with lateral ridges areas of the cross-sectional profile of the scratch test track, was calculated. This factor determines the formation mechanism of the grooves: lateral ridges are formed when it is generated by ploughing with material displacement from the groove to the edges and fcutting = 0 (microploughing mechanism). On the other hand, if the material is completely removed from the groove, fcutting = 1 (microcutting mechanism).19,20 Rockwell C indentation test was performed using 150 kg load according to VDI 3198 standard.
Results and discussion
The coatings were (3.1 ± 0.2 µm) and (3.3 ± 0.3 µm) thick for TiAlN and AlCrN, respectively, as it can be seen in the micrographs shown in Figure 1. The coatings are regular and dense without any significant defects, and they have a regular interface with the substrate.
SEM images of the coatings: (a) AlCrN and (b) TiAlN.
In the coatings XRD analysis with normal incidence and CuKα radiation (Figure 2), the peaks corresponding to the (1 1 0), (2 0 0) and (2 2 0) planes of the substrate were detected as well in both cases because the X-ray penetration exceeds the coating thickness. The AlCrN coating presented the fcc CrN phase structure with a shift to higher angles due to the lattice contraction caused by the substitution of Cr atoms by small-diameter Al atoms (Figure 2(a)). 21 The TiAlN coatings presented the same cubic phase, which is generated by Al atoms displacing some Ti atoms in the TiN lattice (Figure 2(b)). 13 Moreover, in both coatings the AlN phase with hexagonal structure was detected. 17
X-ray diffractograms of: (a) AlCrN and (b) TiAlN coatings.
Table 1 shows the results of the nanohardness and Young's modulus measurements for the coatings and the microhardness for the substrate expressed in GPa. The maximum displacement in both cases is less than 150 nm, half of the required indentation depth to avoid the influence of the substrate, confirming that the properties shown in Table 1 are those of the coatings.
Nanohardness and microhardness results.
The hardness of the coatings was approximately nine times higher than the substrate. The AlCrN coating showed a slightly higher hardness than TiAlN which could be related to the microstructure and density. These nanohardness values are in the range of those reported by other authors for this kind of coatings. 22–24
The H/E and H3/E2 factors were calculated. The H/E ratio is the elastic strain to failure which considered as coating plasticity and the H3/E2 is considered as the resistance to plastic deformation. 17 H/E is considered a measure of the energy dissipation during contact and coating resilience. A higher value of this ratio could indicate dominance of elastic deformation, which could lead to superior wear resistance. H3/E2 is an indicator of the coating's resistance to plastic deformation or load bearing capacity. A coating with a higher value of this ratio deforms less under load so the bending stresses are lower.25,26
AlCrN showed higher values of these factors than TiAlN and this fact is related to the coatings’ responses to tribological tests and adhesion as it will be mentioned later. The coatings showed a better tribological behaviour than the bare steel, which lost a volume of (70 ± 10). 10-3 mm3 at 5 N (1.6 GPa). This value is 111 times higher in comparison with AlCrN. Considering that this difference is already large, the substrate was not evaluated under more severe conditions.
The wear volume loss was thirty times lower for the AlCrN than TiAlN at a 5-N load and 10 times lower applying 10 N, as it can be seen in Figure 3. It could be assessed that the wear track depth did not surpass the coating thickness in any of the cases. The difference in the wear behaviour at increasing loads may be because the influence of the substrate increased as the load increased, being this more noticeable for the TiAlN coating, where the maximum wear track depth was 2.1 µm. However, for the AlCrN coating, it was 0.3 µm, which corresponds to about 10% of the coating thickness, and consequently the substrate influence was reduced.
Wear volume loss for the coatings in the pin-on-disk tests under the two evaluated wear conditions.
In the SEM images of Figure 4, smooth scratches are observed for AlCrN and some black areas, which could correspond to oxides as it was stated in the literature. 27 Different chemical reactions may occur due to the high local pressure on the contact surface during sliding, since Hertzian stresses are about 2 GPa, and flash temperature. They can alter the chemical and phase composition of the outer surface layer, as well as its mechanical properties. On hard coatings, one of the primary processes is the oxides formation.14,28 According to the literature, it was found that the tribo-chemical products which were formed in AlCrN contain chrome oxide and alumina and they were stable and providing effective protection for the coatings from wear. 29 In the TiAlN coating the grooves were found to be deeper and scratches are clearly observed. This plastic deformation could indicate an abrasive wear mechanism. Additionally, the presence of numerous craters inside the wear tracks could be observed. Some small craters are visible, reaching a maximum diameter of 15 µm. These may have resulted from the removal of macroparticles during the wear process. 28
SEM images of pin-on-disk wear tracks for both coatings under 10 N load: (a) AlCrN and (b) TiAlN.
The friction coefficient was measured during the pin-on-disk test, and both coatings showed a CoF steady value lower than the substrate. The TiAlN coating registered a 10% reduction and the AlCrN, a 25% compared to the untreated steel, as it can be seen in Figure 5. It can be considered that the friction coefficient is influenced by three phenomena: abrasion (ploughing or scratching of the surface by asperities or debris), shearing (interfacial material or transfer layer), and adhesion (breaking of adhesive junction under sliding conditions). 30 Taking this into consideration, it can be suggested that the contribution of abrasion was reduced for AlCrN compared to TiAlN, as evidenced by smoother scratches observed in Figure 4.
Graph of friction coefficients as a function of distance for both coatings under both loads and for the heat-treated steel at 5 N and 10 N load.
In addition, the lower friction coefficient of the AlCrN coating is related to the formation of a transfer layer that probably contains chromium and aluminium oxides according to the literature. 31 This transfer layer reduces the contact area and provides a lubricating effect, thereby diminishing the shearing component and resulting in a decreased friction coefficient. 30
The surface damage in the counterpart was lesser for the balls used with AlCrN than the ones used with TiAlN coatings. The contact area diameter was about 565 µm and 588 µm, respectively, for the example showed in Figure 6. A higher amount of adhered material could be observed in the counterpart used with TiAlN (Figure 6(b)).
Optical micrographs of the contact area in the counterparts for both coatings under 10 N load: (a) AlCrN coating and (b) TiAlN coating.
The better performance of the AlCrN coating under sliding wear tests can be related to H/E and H3/E2 ratios (Table 1) which are higher for this coating than for TiAlN. A high H/E value corresponds to a high wear resistance, due to a higher resistance to plastic deformation. 25 The coatings with high H/E and H3/E2 ratios usually have excellent tribological properties as it was reported in the literature. 26 The H3/E2 is usually seen as an indicator of the coating's resistance to plastic deformation or load bearing capacity25,26 as it was mentioned above, is about 30% higher for AlCrN than TiAlN. This is corresponded with the good wear resistance of this coating.
It is worth mentioning that the hexagonal phase detected in XRD in this work might lead to a reduction in hardness and in Young's modulus and an increase in the adhesive wear rate according to some authors 32 However, other researchers have indicated that it is important to note that a reduction in hardness does not necessarily imply a decrease in wear resistance. 33
On the other hand, as it is known, hexagonal crystals have better tribological properties than cubic crystals. Therefore, the presence of hex-AlN phases could enhance the sliding wear resistance of coatings. 34 However, in the studied coatings, there was no significant reduction in the hardness of both coatings that could be attributed to the hex-AlN phase. Additionally, the presence of this phase is not so detectable in the diffractograms and consequently it could be concluded that the effect of this phase was not noticeable in these experiments.
The abrasive wear tests were performed under different conditions of load and time. In all cases, the wear mass loss for coated samples was undetectable or within the method error, even under the most severe condition. On the contrary, the non-coated steel sample showed increasing values for wear mass loss with a 248-mg mass loss when tested at 130 N and 30 min.
Both coatings improved the wear resistance considerably and it can be explained as follows. In this type of tests, both the surface hardness and the relation between the abrasive hardness (Ha) and the surface hardness (Hs) influence the wear resistance.35,36 If this ratio (Ha/Hs) is higher than 1.2, the abrasion is considered ‘hard’. 35 In this case, the abrasive medium was sand, which is supposed to be composed mainly of quartz, with an average hardness of 820 HV. For both coatings, this ratio is much lower, about 0.2, which corresponds to a high abrasive wear resistance. However, for the untreated sample, this ratio is greater, about 1.6, which is consistent with the high levels of abrasion and the considerable wear mass loss.
Regarding the abrasive wear mechanism, the removal of material can be produced by plastic deformation or brittle fracture. 37 In this work, no coating fracture was detected in the wear tracks, only a soft plastic deformation. Grooves in the sliding direction can be observed (Figure 7). These were practically undetectable for AlCrN and they were more marked for TiAlN coatings. In this abrasion test, high shear stresses on the surface material are generated, and they decrease with depth. The deformation is proportional to the indentation depth of abrasive particles. Since the penetration depth of the abrasive particles was very shallow, there was minimal deformation, which corresponds to the undetectable mass loss and the barely marked tracks registered in this test. 37 Moreover, it is known that aluminium-based coatings provide a high degree of protection from abrasive wear and are oxidation-resistant. Other authors have pointed out that the low chemical reactivity and crystal structure stability of AlCrN compared to TiAlN could be responsible for the higher abrasion wear resistance. 38
SEM images of abrasive wear tests for both coatings under 130 N and 30 min: (a) AlCrN and (b) TiAlN.
Scratch tests were performed with different constant loads, as it was mentioned in the experimental part. According to ASTM C1624, critical load Lc1 is used to identify the point at which the first cohesive failure occurs in the coating and Lc2 corresponds to the load value at which the adhesive failure is produced between the coating and the substrate. 39 In these coatings Lc1 could not be determined because no cohesive failure was detected. The first failure for these coatings was observed at 60 N load for the TiAlN coating and at 70 N load for AlCrN 40 (Figure 8). As it can be seen in Figure 8(a), the damage observed for a load of 70 N on AlCrN can be identified as ‘buckling spallation’ because delamination occurs within the scratch groove. Otherwise, the damage around the scratch scars with a load of 60 N for TiAlN (Figure 8(b)) can be classified as ‘recovery spallation’ according to ASTM C1624 since it only occurs outside the track. The scratch test tracks were also analysed by SEM for both coatings at 60 N, and a coating detachment at the edge of the track was confirmed for the TiAlN coating (Figure 9). This can be classified as Lc2. These critical load values correspond to values mentioned in the literature for AlCrN coatings. 12
Microscopic images of the track after the scratch test: (a) at 70 N for CrAlN coating and (b) at 60 N for TiAlN.
SEM images of scratch tracks at 60 N load for: (a) AlCrN and (b) TiAlN.
The better adhesion of the AlCrN is also related with its higher hardness and Young's modulus. The latter indicates that forces holding atoms together are stronger, which corresponds with the higher adhesion strength of this coating in comparison with the TiAlN coating. Also, a higher Young's modulus is related to the coating's greater resistance to plastic deformation, a phenomenon that the sample is subjected to during scratch testing. 31
Scratch profiles were obtained using a mechanical profilometer in the case of 60 N load for both coatings and they are shown in Figure 10. The deformation was similar for both coatings as can be seen in Figure 10. Nevertheless, it can be said that the AlCrN showed a tougher behaviour because it supported the same deformation as TiAlN without failure and it did not show the chipping observed in Figure 10 for the TiAlN coating. The toughness is related to H/E ratio which is higher for this coating than for TiAlN as it was mentioned above (Table 1). Furthermore, coatings with high H3/E2 can be resistant to fracture initiation, which has been used as a measure of toughness. Additionally, a high H3/E2 rate corresponds to a higher load carrying capacity as it was reported by other authors. 26
Scratch test tracks profiles for both coatings at 60 N.
Moreover, the cutting factor (fcutting) was calculated, and the values were very close for both coatings, 0.50 for AlCrN and 0.57 for TiAlN under 60 N load. These values indicate that microcutting and microploughing mechanisms coexist in both coatings and there is no predominance of one of them. Furthermore, when the geometry of the sliding indenter and the load remain constant, the value of fcutting depends on the ductility and toughness of the material. With an increase in these properties, there is a corresponding decrease in the fcutting factor. Therefore, the slightly lower fcutting value for AlCrN suggests a minimal tendency for this coating to exhibit a more ductile behaviour compared to the TiAlN coating. 19
Scratch tests were performed with higher loads, and it was clear that at 90 N, both coatings reached a noticeable damage (Figure 11). Part of the coating was detached, and the tracks showed plastic deformation as it was reported by other authors. 34 Considering the classification given by ASTM C1624, chipping and spallation can be seen for both coatings. The removal of the coating in certain areas of the track was confirmed by EDS spectra (not shown) within and outside of the track. This load can be corresponded to Lc3 where the coating peeling (or total delamination) begins. 41 It can be added that the scratch test simulates the situation of one body abrasion, and the results can be analyzed as abrasion wear resistance.
SEM images of scratch tracks at 90 N load for: (a) AlCrN and (b) TiAlN.
Adhesion was evaluated by means of the Rockwell C Indentation test and it was acceptable for both coatings according to VDI 3198. 42 Considering this standard, the indentations could be classified as HF1, HF2 and HF3 43 as it is shown in Figure 13. The AlCrN coating had better adhesion because a greater number of indentations could be classified as HF1 (Figure 12).
Number of indentations classified as HF1, HF2 and HF3 for both coatings.
SEM images of Rockwell C indentations for both coatings: (a) AlCrN and (b) TiAlN.
The AlCrN coating showed fewer radial cracks, smaller circular cracks and smaller detachment zones (Figure 13). Nevertheless, the TiAlN coating presented radial and circular cracks, in addition to areas where the coating had peeled off. Other authors reported similar damage morphology for coatings applied over AISI H11 steel without any pretreatment. 25
Contrary to what occurs in the scratch test, where a shear stress is applied as the stylus moves linearly, the applied force during the indentation test is normal to the coating surface. 34 The substrate's load bearing capacity is important to the structural integrity of the coating-substrate system. In this case, the substrate was the same for both coatings, however, the adhesion of the AlCrN coating to the substrate was higher than that of the TiAlN. This response is probably due to the stress at the interface, which is related to the coating and its bonding to the substrate. In addition, adhesion is related to the residual stresses in the coating, that is lower compressive residual stresses promote less delamination. 34 The stresses can be generated in differences in the thermal expansion coefficients between the coatings and the steel substrate, which is higher for TiAlN than for AlCrN as it was reported by other authors.25,38
One significant drawback of laboratory tests is their inability to accurately replicate real-world conditions, where various wear mechanisms may act simultaneously, and corrosion may also be a contributing factor. The presented results indicate a 90% superior performance of AlCrN over TiAlN in the pin-on-disk test. It can be assumed this relationship would persist in an adhesive wear scenario with a different geometry, such as a conformal system setup. Additionally, the relationship between mechanical properties and adhesion values in the Scratch test is expected to remain valid in practical situations, including one body abrasion or abrasion induced by hard particles.
Furthermore, manufacturers typically do not provide results from abrasive tests, and while the 90% improvement may not directly translate to real-world scenarios, one coating may still be recommended over the other for certain applications involving abrasion with hard particles. The ASTM G65 test, which is specifically designed to evaluate and rank wear-resistant materials, supports this understanding.
Conclusions
According to the results, the following conclusions can be drawn:
The thickness reached approximately 3 µm in both coatings. The AlCrN coating presented better performance under sliding wear conditions. The wear volume loss was 20 times lower for the AlCrN than the TiAlN in pin-on-disk tests under high loads, and the steady friction coefficient value was also lower. The mass loss was undetectable for both coatings under abrasive wear ASTM G65 even under the most severe conditions. The AlCrN coating failed under a higher load than the TiAlN in the dynamic adhesion test, which would indicate a higher critical load, Lc. Moreover, this coating showed better adhesion in the adhesion test under static conditions.
In general, the AlCrN coating showed a better tribological performance in comparison to the TiAlN coating under severe conditions.
The AlCrN can be proposed as a good protective coating for AISI 420 martensitic stainless steel under different wear mechanisms since it presents good adhesion and higher resistance to plastic deformation. The proposed applications include die casting, plastic processing, tool reconditioning, forming, and punching tools, tribological coatings in motor vehicles (piston pins, pumps, injectors, among others) and aircrafts components, where high level stresses are present. Considering that these parts could be operating in highly corrosive environments, future lines of research include the analysis of the behaviour of these films under tribocorrosion or erosion corrosion, where corrosion and wear mechanisms not only act together but synergically.
Footnotes
Acknowledgements
The authors would like to thank Oerlikon Balzers Argentina for the coatings deposition, Bernardo Daga for nanohardness measurements (INTEMA, Argentina), María de las Mercedes Pianetti (INTI) for SEM images. Moreover, the authors acknowledge the assistance in the experimental part of this work of the students from the Surface Engineering Group of UTN. They also thank the National University of Technology (Faculty of Concepción del Uruguay) Argentina for the financial support.
Declaration of conflicting interests
The authors declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.
Funding
The authors disclosed receipt of the following financial support for the research, authorship and/or publication of this article: This work was supported by the Universidad Tecnológica Nacional.