Effect of Laser Surface Texturing on Coating Adherence and Tribological Properties of CuNiIn and MoS2 Coating

Copper nickel indium (CuNiIn) and Molybdenum disulphide (MoS2) duplex coating is used in aero engine compressor blades of Ti6Al4V to prevent fretting failure. Surface before coating is roughened by grit blasting process for coating adhesion. In present study, in place of grit blasting, laser surface texturing (LST) process is used for surface preparation and compared with grit blasted surface. Four different elliptical and square patterns are created by LST on Ti6Al4V. Difference of surface topography between LST and grit blasted surface is analyzed by SEM and white light interferometery. CuNiIn is deposited on Ti6Al4V by atmospheric plasma spray. MoS2 lubricating coating is deposited on top of CuNiIn by spray painting and curing. Effect of difference in surface preparation on coating microstructure as well as on tribological properties is studied. The results showed that geometry and dimensions of LST pattern influence the coating adherence and wear performance. LST process can be optimized for better performance and explored as an alternative surface preparation process in industry for thermal spray coatings.


Introduction
High strength-to-weight ratio, excellent corrosion resistance and its mechanical properties make titanium alloy a very good choice for several critical applications in aerospace, energy and chemical industries. Titanium alloys are used in many applications in gas turbine engine. Many critical and highly stressed airframe parts of civil and military aircrafts are also made of titanium alloys [1].
Ti6Al4V, Grade 5 alloy is the widely used a ? b titanium alloy. This high-strength alloy is being used from cryogenic temperatures up to 427°C. Some of aero engine applications are: gas turbine aero engine components, rocket engine cases, helicopter rotor hubs, fasteners, critical forgings requiring high strength-to-weight ratio [2]. Critical and wide applications in aero gas turbine engine are high speed compressor rotating disk, aerofoil blades, casings and vanes.
But these titanium alloys are very poor in fretting and galling conditions, which may lead to fretting failure. Use of unlubricated tribological system in titanium and its alloys should be avoided in practical application [3]. Figure 1 shows a fan rotor of Ti6Al4V which is bladed disk configuration with total 130 blades in three stages of low pressure compressor rotor. Blade root is assembled in disk dovetail slot. Fan disk and all fan blades are manufactured from Ti6Al4V alloy and create direct metal to metal contact.
All 130 joints between aerofoil and rotating disk are subjected to sliding motion with small amplitude and high frequency vibrations during the engine operation and may lead to fretting failure if not protected properly. In order to prevent this kind of failure, CuNiIn ? MoS 2 coating is provided on the blade dovetail. Figure 2 shows first stage low pressure compressor blade of aero gas turbine engine.
The dovetail of blade is coated with CuNiIn by thermal spray process. This CuNiIn coating is sprayed on compressor blade dovetail as anti-fretting coating for preventing direct metal to metal contact. In order to roughen the surface for CuNiIn coating adhesion, the dovetail surfaces are cleaned and then grit blasted with alumina grit. In thermal spray industry, grit blasting is used as a standard process to prepare and roughen the surface of blade dovetail. Roughening with blasting increases the contact area for mechanical anchoring of coating with blade dovetail.
CuNiIn thickness of 13-51 lm is applied on one of the mating parts usually on compressor blades in case of bladed disk configuration [4]. Interface wear is further improved using dry film lubrication with molybdenum disulphide coating on top of copper based coating. MoS 2 lubricating coating up to a thickness of 13-18 lm is applied on top of CuNiIn coating before assembling the blade with disk. MoS 2 coating can withstand temperature up to 300°C. This duplex configuration coating of CuNiIn and MoS 2 is used in engine to improve the tribological properties and prevent fretting failure between two contacting surface of titanium alloy.
Freimanis, Segall [5] investigated CuNiIn coating systems between titanium blades and rotor of jet engine. Study shows that at 221°C, titanium from disk surface gets transferred to CuNiIn coating in the absence of molybdenum disulphide lubricant. This creates metal to metal contact between titanium between disk and blade, damages the coating and increases the chance of fretting wear and fretting failure. Fayeulle, Blanchard [6] studied the fretting behavior of Ti6Al4V alloy. It is reported that wear debris after detachment are oxidized into TiO and TiO 2. Hager Jr, Sanders [7] investigated the effect of fretting wear of CuNiIn plasma spray coatings in Ti6Al4V alloy in unlubricated condition. Since the contact between thermal spray coating and Ti6Al4V surface is unlubricated, all the sprayed coating tested is not able to protect the Ti6Al4V surface. Hager Jr, Sanders [8] reported that nickel graphite coating reduces the interface wear on mated Ti6Al4V surface by formation of lubricating graphite and nickel oxide film.
Bonding of any coating with Ti6Al4V substrate depends on pretreatment of substrate surface. There are different methods for pre-cleaning, intermediate cleaning like alkaline, emulsion, solvent cleaning etc. [9]. Pretreatment does surface cleaning and removes the outer surface impurities like oil, grease and oxides. After proper cleaning, surface roughness needs to be created. Grit blasting process is widely followed by industries for creating surface roughness thereby increasing the contact surface area [10]. Figure 4 shows 2D topography of grit blasted Ti6Al4V surface. Grit blasting process does not create any fixed and repeated pattern of roughness. Peaks and troughs on blasted surface allow molten or semi molten powder particles to flow in to gaps and solidify around the asperities on the surface. This results in mechanical interlocking between coating and substrate. Mechanical interlocking is the main mechanism for adhesion of thermal spray coating with substrate [11].
In addition to grit blasting, there are different surface treatment processes like chemical reaction, conventional and nonconventional machining processes which can also create roughness. Recently, ablating by a laser source called laser surface texturing has gained much interest among researchers. Creation of predefined surface texture by LST is found to be an innovative and effective method to improve the adhesion strength of thermal spray coatings with component surface. Surface texturing creates predefined patterns and increases Ra by up to 4 times. Shear strength value is also increased up to 3.48 times by selecting a suitable pattern in laser texturing process [12]. Surface texturing on aluminum alloy AA5052-H32 sheet created by desktop micro rolling has increased total effective surface area for bonding. This helped in improving the mechanical anchoring and in turn adhesive strength of joint [13]. Two different laser sources CO 2 laser and fiber laser have been used to create texturing on 316L stainless steel and significant increase in adhesive strength is observed with both lasers [14].
Surface texturing helps to improve adhesion strength of Ti6Al4V and carbon fiber reinforced plastic (CFRP) three times as compared to non-textured surface [15]. Different types of patterns can be created by laser texturing on Ti6Al4V alloy to get the desired roughness [16]. Pattern topography can be optimized for range of applications with an aim of improving adhesive strength up to two to three times [17]. Laser micro textures on Ti6Al4V also promote lubrication absorption and retention which improves the wear and frictional properties of the surface as compared to non-textured surface [18]. Ahmed, Ahmad [19] studied laser milling of Ti6Al4V with different set of laser parameters. Microstructure analysis of sub-layer underneath the laser treated area does not reflect any changes in microstructure of Ti-6Al-4V substrate.
High temperature durability of bond coat less thermal barrier coating system on single crystal Ni-based AM1 single crystal alloy is improved after laser texturing [20].
Laser texturing with grooves in stainless steel increases the adhesion strength of atmospheric plasma sprayed molybdenum coating by 49.7% as compared to grit blasted surface [21]. Triangular surface texturing by LST enhances tribological performance and adhesive bond strength of TiN coating [22]. Zhang, Deng [23] reported that nanoscale textures on TiAlN coating surface improves the effective life of WS2 solid lubricating film layer for a longer period as compared to non-textured tool. Li, He [24] concluded that the coefficient of friction of silver coating on the textured surface is better as compared to the un-textured one. Fan, Su [25] reported that LST improves the hydrophobicity of PTFE coating up to 2 times. This has resulted in increase in wear life of coating up to 5 times that of the un-textured surface coating. Investigations have shown that texturing improves the tribological properties of brass and aluminum alloy [26]. Texturing improves machinability of Ti6Al4V both with uncoated and coated textured tool. Perpendicular texture on TiAlN coated tool shows better machinability than TiN. Friction force reduction was also noticed in textured tool [27].
As reported in earlier study by Bagade, Duraiselvam [28], copper nickel indium plating CuNiIn is sprayed by atmospheric plasma spray process on Ti6Al4V substrate. Two different processes are used to prepare surface for CuNiIn coating application. One surface is prepared by grit blasting process and other by laser texturing. Two types of square texture patterns are created by varying texturing parameters. Effect on coating adherence due to variation in surface preparation process is compared. Coating adhesive strength values for laser textured pattern and a grit blasted samples are evaluated. Interface quality of the coating is improved and coating adhesive strength of 18.67 MPa is achieved on laser textured surface which is 9.82% greater than coating adhesive strength with grit blasted surface. It is concluded that a texturing pattern can be optimized for improved adhesive strength.
From the literature survey, it is observed that there are very few researchers actually implementing LST in surface preparation for thermal spray coating. Even though LST has greater repeatability in creating texture on surface, there is no inclination shown in research for industrial application of LST for surface preparation in aero gas turbine engine. Surface preparation by LST on small aero engine components like compressor blade will lead to greater repeatability and less rejection in surface preparation for anti-fretting coating application. Very few researchers have carried out LST on Ti-6Al-4V alloy. There is no report or literature related to tribological behavior of CuNiIn and MoS 2 duplex coating with LST. There is a scope for LST process to be used as a surface preparation process for small components. LST will overcome the limitations of grit blasting process such as nonuniformity of blasting pattern in different directions, induced stresses, prevention of deformation in thin components and practical difficulties in inspection of blasted component surfaces.
In earlier study, square geometry of laser texture pattern was fixed and coating adhesive strength of 9.82% greater than grit blasting was achieved in laser texturing by varying texturing parameters [28]. In the present a work, Ti6Al4V surface prepared by grit blasting is compared with elliptical and square textures prepared by LST. Four different textures are created by laser texturing on different Ti6Al4V samples by varying texture dimensions and laser scanning parameters. CuNiIn coating is sprayed by atmospheric plasma spray process on grit blasted as well as on all textured samples. Molybdenum disulphide lubricating coating on top of CuNiIn coating is applied by paint spray and drying process. Effect of different laser textured patterns on coating substrate interface quality is reported. Pin on disk wear test is conducted on Ti6Al4V samples with MoS 2 and without MoS 2 lubricating coating. Effect of different texturing patterns on tribological properties of coating is also reported.

Substrate Preparation
Ti6Al4V is used as a substrate material in this study. Chemical composition of Ti6Al4V alloy is shown in Table 1. 25 mm diameter 9 6 mm thick cylindrical sample is used for microstructure studies. Pins of 8 mm diameter 9 50 mm long are used for wear studies. All samples are machined and alkaline cleaning in ultrasonic cleaner is carried out before proceeding either for LST or grit blasting. The surface roughness of all the samples after machining is achieved within Ra 0.8 lm.

Grit Blasting
Grit blasting is carried out with 18-20 mesh alumina grit. Pressure type grit blasting machine with 10 mm diameter nozzle and 150 mm standoff distance is used during blasting. 25 mm dia 9 6 mm thick flat sample is grit blasted on one face. Similarly, 8 mm diameter pin 9 50 mm long pins are grit blasted on one face. After grit blasting these samples are subjected to coating.

Laser Surface Texturing
Nd: YAG nanosecond pulsed laser with wavelength of 1064 nm and pulse frequency of 3 kHz is used as a source for LST. Laser source having Gaussian distribution profile with beam focus diameter of 7 lm is adopted. Laser movement is controlled by CNC with X-Y manipulator. Square and elliptical geometry is selected as texture patterns. Two types of square and two types of elliptical patterns are textured. Width and pitch of the pattern are kept much higher than CuNiIn coating particle size distribution range. Particle size of CuNiIn coating is in the range of 45-75 lm. Pattern dimensions, laser power and scanning speed are varied. This will provide variation in depth of pattern. Table 2 shows the geometrical dimension and LST parameters of square and elliptical textures. Figure 3 shows the schematic of LST set up. LST is carried out on single face of both types of samples before coating application. 25 mm diameter 9 6 mm thick round sample is laser textured entirely on one face. Similarly, pin of 8 mm diameter 9 50 mm long is textured fully on single face. Coating is applied on entire area where LST is carried out.

Thermal Spray and Paint Spray Coating
Thermal spray coating of samples is carried out on Mul-tiCoat thermal spray coating facility (M/S Oerlikon Metco, Switzerland). Atmospheric plasma spray (APS) with 80 kW, 9 MB plasma spray torch is used for thermal spray coating of samples. Test samples are mounted on turn table with fixture. Samples of 25 mm diameter 9 6 mm thick and pins of 8 mm diameter 9 50 mm long are sprayed in a single set up with 6 axes robot integrated with 2 axis turn table. Cu36.5Ni5In (Metco 58NS) powder is alloy powder, produces very dense coatings of porosity less than 0.5% along with low oxide content. This coating is best suited to resist wear by fretting [4]. Powder is stirred well and pre-heated to 50°C to remove any moisture before spraying. Thermal spray coating process parameters for coating of Metco 58NS is listed in Table 3. These spray parameters are established spray parameters to meet CuNiIn coating quality requirements of specified hardness, low oxide and less porosity [4]. All the samples are coated

Pin on Disk Wear Test
Pin on disk wear test is performed on 8 mm diameter and 50 mm length pin against Ti6Al4V disk. All pins are coated on 8 mm diameter face with CuNiIn coating of 50 lm thickness. MoS 2 lubricating coating of 15-20 lm is applied on CuNiIn. The tribological performance of CuNiIn coating with MoS 2 and without MoS 2 has been studied. Based on initial trials on bare Ti6AL4V pin, load is fixed at 75 N for testing on coated pins. Sliding speed is kept at 2 m/s and sliding distance is set as 2000 m.

Surface Topography of Grit Blasted and Laser Textured Samples
Comparative study of surface prepared by grit blasting process and laser texturing process is performed. Grit blasting is carried out on the sample by 18-20 mesh alumina grit with pressure of 1.8 bar and standoff distance of 200 mm (Mecshot pressure blasting machine). Samples are  then ultrasonically cleaned in acetone for 5 min before analysis. Figure 4a, b shows WLI and SEM images of samples which are grit blasted. Roughness value Ra obtained is 5.76 lm. SEM image of grit blasted sample shows entrapment of grit on the surface which is already cleaned with ultrasonic process. Grit entrapment on blasted surfaces is one of the main concerns of grit blasting process. If grit entrapment is more, it may cause delamination or detachment of the coating from surface. Figure 5 shows 2D topography of grit blasted surface showing randomness in peaks and valleys. There is no fixed pattern and spacing between asperities.
Varieties of patterns can be created on Ti6Al4V by variation in LST parameters [16]. 25 mm diameter and 8 mm thick flat samples are textured with four different patterns as per the parameters listed in Table 2. Samples are then ultrasonically cleaned in acetone for 5 min before analysis. Figure 6a, b shows WLI 3D topography and 2D topography of ellipse-2 pattern. Figure 7a, b shows surface morphology of ellipse-2 pattern. Figure 8a, b shows WLI 3D topography and 2D topography of square-1 pattern. Figure 9a, b shows surface morphology of square-1 pattern. Intense and concentrated heat input by laser source ablates the metal. Resolidified metal is observed at the periphery of elliptical pattern (Fig. 7a, b) and in square pattern (Fig. 9a, b). Surface roughness for any surface will be understood by amplitude parameter which is called roughness. But spacing between the roughness profile is also very important to understand the surface [29]. There is a fixed spacing between square to square and ellipse to ellipse in laser texturing. As compared to 2D topography of grit blasted surface (Fig. 5), 2D topography of laser textured surface (Figs. 6b and 8b) shows pillars, asperities and valleys. 3D WLI images of laser textured surface ( Fig. 6a and 8a) also show uniformity in texturing. LST creates fixed pattern for mechanical anchoring of the coating. This kind of surface texture which is fixed in pattern will never be achieved by grit blasting process.
Components of aero gas turbine like dovetail of aerofoil blades having 3 mm to 4 mm width and 100 mm to 150 mm length can be easily processed by LST to create fixed repeated patterns at faster rate. LST process will provide high repeatability and productivity if process is used on small components with high volume. LST becomes typical case of mass production of small components. Titanium alloy is used under different operating condition in aero gas turbine and needs different types of thermal spray coatings like abradable, wear resistance, anti-fretting etc. [30][31][32].

Evaluation of Surface Roughness
Surface roughening is mandatory before coating. Earlier work reports surface roughening by single type of square pattern on titanium alloy with CuNiIn anti-fretting coating [28]. For the present work, the surface texture parameters for square and elliptical pattern inspected by 3D WLI profilometer are indicated in Table 4. Random or repetitive deviation of the surface from nominal surface is called surface texture. Surface texture includes roughness, waviness, lay and flaws. Roughness is surface irregularity of smaller wavelength. Waviness is surface irregularities of longer wavelength greater than roughness sampling length and shorter than waviness sampling length [29]. In the present study, Average roughness (R a ), Maximum height of the roughness (R t ), Average maximum height of the profile (R z ), Waviness average (W a ) and maximum height of waviness (W max ) is taken for more clarity and understanding the texture of surface. Average surface roughness   (R a ) value of grit blasted sample is 5.76 lm, and R a of four textured patterns are Square S1-5.423 lm, Square-S2 7.491 lm, Ellipse-E1 5.775 lm and Ellipse-E14.755 lm. These average roughness values are not having considerable variation from grit blasted surface to textured surface. But there is a considerable variation in other roughness parameters like R t , R z , W a and W max . The pattern square-S1 and ellipse-E2 has higher pitch of 400 lm as compared to 300 lm pitch in sample square-S2 and ellipse-E1. Higher W a & W max values are achieved in square-S1 and elliptical-E2 texturing where pattern pitch is higher.    Figure 10 shows CuNiIn and MoS2 coating microstructure on grit blasted samples. There is no fixed pattern of anchoring on grit blasted surface. Many up and downs with pits of the order of 50-100 lm on grit blasted surface are observed. But adherence of CuNiIn coating with aerofoil blade substrate mainly depends on surface preparation. Any variation in surface preparation can lead to detachment or removal of the coating during service life and can lead to fretting failure. Experience shows that during blasting, if surfaces less than 3 mm thickness are not supported properly by fixturing, it can lead to distortion due to induced residual stresses in blasting process. Distorted surfaces require repair in pre or post coating operations. CuNiIn coating microstructure on the four different types of laser textured pattern Square-S1, Square-S2, Ellipse-E1 and Ellipse-E2 is shown in Fig. 11a-d, respectively. All the four microstructure revels that there is a repeatability in surface created by laser surface texturing as compared to unevenness of grit blasted surface (Fig. 10). Mechanical interlocking is the main mechanism in bonding of coating with the substrate [11]. Out of four LST patterns, coating is mechanically found interlocked with the laser surface textured pattern in two types of pattern square S1 (Fig. 11a) and ellipse E2 (Fig. 11d). Coating is found adhered to laser textured Ti6Al4V substrate. Whereas, there is a complete detachment of coating with the substrate in pattern square S2 (Fig. 11b) and ellipse E1 (Fig. 11d). It can be concluded that samples with higher pitch have good coating adherence as compared to sample with lower pitch. Further investigations on width 'W' and depth 'D' (Fig. 12) of the profile created by LST process will provide more information about mechanical interlocking of the coating with the surface textures.

Coating Substrate Interface Evaluation
Surface roughening on titanium alloys improves the adhesion [33]. Figure 12 shows the analysis of profiles created by laser surface texturing on Ti6Al4V surface. The textured surface with square S1 pattern has higher pitch of 400 lm, higher Wa and Wmax. The dotted rectangles shown in Fig. 12 are having dimensions 'D' depth of the hole and 'W' width (diameter) of the hole in which CuNiIn coating splat is deposited. This analysis is done for all the four texture patterns. Table 5 shows the variation in depth and width of coating splat in laser textured patterns.
Pattern square-S1 and ellipse-E2 have higher depth of texturing 'D' as compared to square-S2 and ellipse-S1. Maximum depth of texturing 'D' is found in ellipse-2 as compared to square pattern S1. Pattern square-S2 and ellipse-E1 has low depth which affect the bonding of coating with Ti6Al4V substrate. Width of texture 'W' in ellipse E1 varies from 117 lm to 62 lm which is on the higher side and same pattern has less depth of texture between 14 lm and 22 lm that causes delamination in ellipse E1 texture. Hence depth of texture 'D' has more influence on mechanical interlocking of coating than width 'W' of texturing.

Pin on disk wear test
The test is done at ambient temperature. The density of Ti6Al4V is taken as 4.47 g/cm 3 . For wear track diameter of 60 mm, speed of the disk is kept at 637 rpm. During the test, samples are weighed before and after the test, and mass loss is calculated. The wear rate and coefficient of friction are calculated as per Eq. 1 and 2 Wear rate mm 3 Coefficient of frictionðlÞ ¼ FrictionForceðFÞ NormalLoadðNÞ ð2Þ Using pin on disk wear test, the tribological properties of single CuNiIn coating and duplex CuNiIn with MoS 2 are studied at 75 N load with 2000 m sliding distance. This study is done mainly to analyze and compare influence of surface preparation process and geometry of surface texture on coefficient of friction and wear rate. The coating is applied on grit blasted as well as on laser textured samples. Figure 13 shows the wear test results on 10 different samples with and without MoS 2 coating. The coefficient of friction tends to decrease almost in all surface textures after application of MoS 2 lubricant coating on top of CuNiIn coating. In elliptical pattern, wear rate decreases after application of MoS 2 coating. But the wear rate increases in case of grit blasted and square pattern samples after application of MoS 2 coating. This is due to the removal of asperities in sliding from the top surface especially with high Rt and high Rz values (Table 4) of square pattern as compared to elliptical pattern. From the results, it is interpreted that elliptical pattern gives better anchoring and interlocking effect than square pattern where filling of sprayed powders in the sharp corners could be difficult. Hence, geometry of pattern has influence on wear rate and coefficient of friction.

Conclusion
In the present study, for application of CuNiIn and MoS 2 coating on Ti6Al4V, surface preparation is carried out with two different processes. First process is grit blasting, a conventional method followed by industry and second process is LST with varied patterns. Different types of elliptical and square patterns are created by varying laser surface texturing parameters. CuNiIn coating by APS and MoS 2 lubricant by painting and curing is applied on the samples. Analysis of different surfaces prepared by LST on coating substrate interface is studied. The effect of grit blasting surface and laser textured surface on tribological properties CuNiIn and MoS 2 coating is evaluated. The conclusions drawn from this study are: • Entrapment of grit on the grit blasted surface even after ultrasonic cleaning post grit blasting is a major concern of grit blasting process. This will cause grit inclusion between coating and surface which has undesirable effect. • Laser surface texturing is a viable option to create a controlled surface modification. LST will completely eliminate the effect of grit entrapment. • Geometry of pattern and spacing of the pattern (pitch) has influence on roughness created by LST. LST with higher pitch shows higher Wmax. • Depth of texturing created by laser has more influence than width of texturing on mechanical interlocking of coating with substrate. • Elliptical patterns have resulted in reduction in wear rate after application of MoS 2 lubrication coating on CuNiIn. Square pattern and grit blasting show increase in wear rate.
• Pattern geometries and laser surface texturing parameters can be created and optimized to improve the coating adherence and tribological properties of the CuNiIn and MoS 2 coating. • Optimized LST will lead to higher productivity on small components like low pressure and high-pressure compressor blades.