Tribological characteristics and micromilling performance of nanoparticle enhanced water based cutting fluids in minimum quantity lubrication

Abstract Friction and wear characteristics of water-based nanofluids were studied. Deionized water-based nanofluids were prepared with alumina (Al2O3), hBN, MoS2, and WS2 nanoparticles. Tribological study of the prepared nanofluids was undertaken on a ball-on-disk tribometer in minimum quantity lubrication (MQL) mode on Ti-6Al-4 V workpiece. The effect of flow rate on the coefficient of friction (μ) and wear of the workpiece was reported for different nanofluids. 3D profiles of the wear tracks were obtained to study the wear depth and wear profile. Alumina-based nanofluids have shown excellent performance in terms of friction and wear as compared to the other nanofluids. The coefficient of friction, wear track depth and specific wear were reduced by 53.89 %, 23.4 %, and 37.03 % respectively for alumina nanofluid in MQL mode as compared to the dry environment. A commercial cutting fluid (UNILUB 2032) was used for comparative studies. The commercial cutting fluid performed slightly better than the alumina nanofluid, both used in MQL mode. Even though hBN, MoS2, and WS2 are solid lubricants, their performance as nanofluids in MQL mode was not better than alumina nanofluids. Alumina particles, being spherical shaped, were able to provide better lubrication owing to the ball-bearing effect. Scanning electron microscopy (SEM) and energy dispersive X-ray spectroscopy (EDS) of the wear tracks were conducted to understand the mechanism of friction and wear reduction. Micromilling experiments were done on Ti-6Al-4 V using the prepared cutting fluids in MQL mode to evaluate their effect on machining forces and burr formation. Alumina nanofluids and the commercial lubricant have shown significant reduction in machining forces and burr formation as compared to the other nanofluids. This is in correlation to the tribological performance of these cutting fluids.


Introduction
The demand for miniature sized components has increased rapidly in the last few decades in the fields of electronics, automotive, medicine, aerospace, communications, etc. Mechanical micro machining processes like micromilling have distinct advantages over lithography and laser-based techniques like the ability to machine complex 3D features on both metallic and non-metallic materials with high precision [1], and the ability to machine high aspect ratio structures like thin walls [2]. The formation of burrs in mechanical machining processes is a well-known phenomenon and is a challenge in micromilling since the relative size of the burrs to the dimensions of the component is large [3,4], and these burrs negatively affect the quality and functionality of the final product. This is mainly due to the dominance of ploughing, rubbing, and plastic deformation in the material removal process [5] arising as a result of size effects at the micro-scale [6].
Minimisation and control of burr formation in micromachining has gained increased importance due to the wide range of applications of these processes. Lee and Dornfeld [3] have reported that in micromilling, built-up edge tends to remain on the tool for longer and the rate of wear decreases as the cutting velocity increases, thereby decreasing the burr size. Lee and Dornfeld [7] have further studied the burr formation in micromilling of aluminium and copper and reported that most of the burrs formed are top burrs. Cheng et al. [8] have studied the top burr formation in micromilling of Ti-6Al-4V and have observed that the downmilling burrs were larger than the upmilling burrs. They have also reported that the burr size can be significantly reduced when the feed rate is sufficiently larger than the maximum uncut chip thickness.
The use of metalworking fluids has been integral to the metal cutting processes since these fluids aid in improving product quality and increased productivity. Flood cooling has several advantages like reducing tool wear, chip-tool interface temperature, work-tool interface friction, and surface damage on the workpiece. However, these fluids also pose problems related to the environment, human health, higher cost of handling and disposal of cutting fluids, and a need for sustainable manufacturing [9,10]. Hence, researchers have turned to sustainable alternatives like minimum quantity lubrication (MQL). In MQL, a small amount of cutting fluid is atomized, mixed with compressed air, and directly applied to the machining zone effectively [11]. Sun et al. [12] have compared MQL and flood cooling using soluble oil while end milling Ti-6Al-4V. They have reported that MQL achieves the best lubricating effect, thereby resulting in a reduction of machining forces and tool wear.
Titanium alloys have certain unique properties like high strength to weight ratio, hightemperature strength, fracture toughness, and good corrosion resistance which has resulted in the being widely used in the fields of aerospace, bio-medical, defence, etc [13]. But these alloys are considered difficult-to-machine due to high chemical reactivity with the tool materials [14].
This has led to increased research to improve the machinability of titanium and its alloys.
Vazquez et al. [15] have studied the effect of different cooling and lubrication conditions during the micromilling of Ti-6Al-4V. They have reported that MQL has offered the best dimensional accuracy of the micro-channels whereas flood cooling produced a dimensional error of 20%.
They have also observed that MQL has resulted in better surface finish as compared to flood cooling. Ziberov et al. [16] have studied the effect of the application of cutting fluid on the tool wear, burr formation, and machined surface during micromilling of Ti-6Al-4V. They have reported that MQL has led to an absence of built-up edge which resulted in a reduction in burr formation and surface roughness while increasing the tool wear.
The addition of nanometer-sized particles to the traditional metalworking fluids has several advantages. These "nanofluids" are known to be better than the pure fluids in terms of heat transfer coefficient [17,18], coefficient of friction [19][20][21][22], viscosity, density [23], and load-bearing capacity [20,22,24]. A lot of research is being carried out to assess the role of nanofluids in lubrication. Nanofluids have shown superior anti-friction and anti-wear properties as compared to conventional lubricants. Spherical nanoparticles like alumina (Al2O3), CuO, and diamond act as nano ball-bearings, thereby reducing the friction between sliding components [19,21,22]. These nanoparticles also get entrapped in the surface asperities of the workpiece, thereby strengthening the lubrication film created by the base fluid [20][21][22].
Solid lubricant based nanoparticles like hBN, MoS2, and WS2 impart lubricating properties to the base fluids as these particles have a graphite-like lamellar structure with layers able to slide over one another, thereby reducing friction [25,26]. These nanofluids, when used as cutting fluids in MQL mode, have shown superior performance as compared to conventional cutting fluids in terms of machining forces, temperature, surface finish, and tool wear. Liao et al. [27] have studied the effect of nanofluid MQL and flood cooling in the grinding of Ti-6Al-4V. They have reported that the grinding forces, coefficient of friction, and wheel loading have reduced in the case of nanofluid MQL. Setti et al. [28] have studied the performance of MQL using water-based Al2O3 and CuO nanofluids during the grinding of Ti-6Al-4V. They have reported that the addition of nanoparticles has improved the lubricity of the cutting fluids, reduced the grinding temperature, and improved the grindability of the work material. Kim et al. [29] have studied the performance of nanodiamond MQL during micro end milling of Ti-6Al-4V. They have reported that the addition of nanodiamond particles has enhanced the lubricating effect of the cutting fluid which resulted in significant reduction in burr formation and surface roughness.
Past literature has discussed the mechanism of lubrication using different lubricants like term exposure [30][31][32][33]. A commercial lubricant (UNILUB 2032) has been tested for comparison purposes. Further, the effectiveness of these nanofluids in reducing the machining forces and burr formation in micromilling of Ti-6Al-4V alloy has been studied.

Preparation of nanofluids
Alumina (Al2O3), hBN, MoS2, and WS2 nanoparticles (Nanoshel, India) of average particle size less than 100 nm were dispersed in deionized water using an ultrasonic probe sonicator (PCI Analytics, India). Transmission electron microscope (TEM) images of the nanoparticles are shown in Fig. 1. Alumina particles are spherical in shape and hBN, MoS2, and WS2 have a layered structure. The properties of the nanoparticles are shown in Table 1.
Dispersion was carried out for one hour at a sonication frequency of 20 kHz and a power of 300 W at a constant temperature of 25 °C. Several surfactants were tested to stabilise the dispersions like sodium dodecyl sulphate (SDS), polyvinylpyrrolidone (PVP), cetrimonium bromide (CTAB), and sodium dodecylbenzene sulfonate (SDBS). Out of the tested surfactants, SDS was found to be the best for alumina, MoS2, and WS2 nanofluids, and PVP was found to be the most suitable for hBN nanofluids. Stability of the dispersions was assessed visually by observing the nanofluids 24 hours after preparation as shown in Fig. 2. A synthetic ester-based commercial lubricant (UNILUB 2032) has also been used for studying its suitability as a cutting fluid in comparison to the prepared nanofluids. UNILUB 2032 is a non-water soluble lubricant consisting of a mixture of esters, manufactured from natural fatty acids (> C16) and sterically hindered alcohol. Composition of the prepared nanofluids is shown in Table 2.

Minimum quantity lubrication (MQL) setup
Tribology experiments and subsequent micromilling experiments were conducted in dry and MQL environments. An MQL setup developed in-house has been used for this purpose.     Table 3. The counter-body was slid against the disk under a constant normal load of 10 N (nominal contact pressure of 1.37 GPa between the sliding surfaces as calculated using Hertzian contact theory [34]) while the disk was rotating at a constant speed of 320 rpm.
Diameter of the wear track was 10 mm. MQL nozzle was positioned against the direction of motion of the ball relative to the disk. Distance between the nozzle and the sliding zone was 20 mm and the angle between the nozzle and the workpiece was 15°. Experiments were conducted for a sliding time of 20 min at ambient temperature. Two repetitions were conducted for each experimental condition. Friction force was continuously measured throughout the tribo test with the help of an inbuilt load cell. The coefficient of friction, which is a ratio of the friction force to the normal force, was determined using the measured friction force values.    Table 4. Micro-channels of width 500 µm were milled on the prepared samples with a high precision CNC micromachining centre (Kern Evo). The prepared nanofluids were supplied to the machining zone using the in-house made MQL setup described earlier.
Machining setup is shown in Fig. 6. MQL nozzle was positioned against the direction of the workpiece feed as shown in Fig. 6. Distance between the nozzle and the sliding zone was 20 mm and the angle between the nozzle and the workpiece was 15°. Two sets of parameter combinations were used for machining consisting of two levels of cutting velocities (10 m/min and 60 m/min). The feed used was 6 µm/flute, the depth of cut was 25 µm and the channel length was 5 mm. Two such channels were machined for each parametric combination to ensure test repetition.

Results and discussions
Tribology experiments coefficient was very close to that of the commercial lubricant (10.2% at 6 ml/hr, 5.49% at 50 ml/hr, and 11.48% at 100 ml/hr higher than the commercial lubricant). An increase in the flow rate leads to a reduction in the coefficient of friction for all the fluids tested. This can be attributed to the increased tribofilm stability at higher flow rates and also the availability of more nanoparticles at higher flow rates, both of which tend to reduce the sliding friction [19,20,22]. At 100 ml/hr, reduction in the coefficient of friction for alumina nanofluid and the commercial lubricant are 53.89% and 59.19% respectively as compared to dry friction.
Alumina nanoparticles, due to their spherical shape, cause a nano ball-bearing effect between the sliding surfaces, thereby converting the friction mechanism from sliding friction to rolling friction and reducing friction [19,22]. Further, alumina nanoparticles get trapped in the asperities of the workpiece surface, thereby strengthening the lubrication film and lowering the coefficient of friction [20][21][22]. This can be confirmed from the elemental composition of the wear surfaces by EDS as shown in Fig. 10. Another aspect of the nanoparticles is that they act as a third-body [20,22,24], meaning that the nanoparticles act as load-bearing bodies, thereby reducing direct contact between the sliding surfaces and reducing friction.    and 28.37% respectively as compared to dry friction. It can be inferred from Fig. 9 that the commercial lubricant has the best anti-wear properties, closely followed by alumina nanofluids.
Alumina nanofluids, even though are effective in reducing the coefficient of friction, are not as effective as the commercial lubricant in controlling wear. This can be attributed to abrasion between the nanoparticles and the sliding surface [38]. hBN, MoS2, and WS2 nanofluids performed worse than alumina nanofluids, with wear at 6 ml/hr similar to that of dry sliding.    Fig. 10. hBN, MoS2, and WS2 nanofluids have failed to significantly decrease friction and wear of the surface. The structure of these materials is lamellar in which the layers are held together by weak Van der Waals force. The lubricating mechanism is by the sliding of the layers [40] which has provided some amount of lubrication in the present study, but not as adequately as alumina nanofluids. The findings are in agreement with the measured coefficient of friction, wear depth and specific wear as shown in Fig. 8 and Fig. 9.

Micromilling experiments
Micromilling experiments were conducted on Ti-6Al-4V to evaluate the machining performance of the prepared nanofluids in minimum quantity lubrication (MQL) mode at different flow rates. The schematic of the micromilling operation showing the direction of forces is shown in Fig. 11a. Fig. 11b   Cutting forces for alumina nanofluids are close to those for the commercial lubricant with feed force (FX) being up to 8% higher, transverse force (FY) being up to 12.8% higher, and thrust force (FZ) being up to 9.5% higher than the commercial lubricants at a cutting velocity of 60 m/min. hBN, MoS2, and WS2 nanofluids were not as effective as alumina nanofluid in reducing the cutting forces. This significant reduction in cutting forces using alumina nanofluids can be attributed to both the nano ball-bearing effect of the spherical alumina nanoparticles and the entrapment of the nanoparticles in the asperities of the workpiece, thereby reducing friction [19][20][21][22]24]. Further, the reduction in the forces was observed to be higher when the flow rate was increased from 50 to 100 ml/hr than when the flow rate was increased from 6 to 50 ml/hr for all the lubricants used. This can be attributed to the increase in tribofilm stability and the higher availability of nanoparticles at 100 ml/hr as explained from the tribology experiments.
Further, the cutting forces are observed to increase with an increase in the cutting velocity from 10 m/min to 60 m/min. This can be attributed to the strain hardening effect at higher cutting velocities. Johnson-Cook material model states the flow stress for a material as a function of strain, strain rate, and temperature as given in Eq. (1).
where is the material yield stress constant, is the hardening module of the material, is the strain rate susceptibility, is the hardening coefficient, and is the thermal softening coefficient of the material. Other parameters are: is the plastic strain, is the strain rate (s −1 ), is the reference strain rate (s −1 ), is the workpiece temperature, is the room temperature, and is the workpiece melting temperature. J-C constants for Ti-6Al-4V are given in Table 5 [41][42][43], and material properties are given in Table 6 [44].
where is the cutting velocity, is the orthogonal rake angle of the tool (5°), is the shear plane angle, and ∆ is the adiabatic shear band thickness. It can be seen that the strain rate increases by 6 fold as the cutting velocity increases from 10 m/min to 60 m/min. The temperature rise is not high enough in micro-scale machining to cause significant strainsoftening effects as compared to meso-scale machining for the above range of cutting velocities [45], and hence, the temperature softening effects can be safely neglected. From Eq. (1), an increase in strain rate can contribute to the rise in flow stress, thereby increasing the cutting forces at 60 m/min cutting velocity.  ploughing and rubbing of the tool edge against the workpiece surface have a higher contribution than that in meso-scale machining. This phenomenon, known as size effect, has been studied extensively [7,[46][47][48][49][50]. More ploughing and rubbing are known to cause higher burr formation [50]. The large top burrs in micromilling due to size effect has been explained by Bissacco et.al [4].

Conclusions
The lubrication ability of different water-based nanofluids in minimum quantity lubrication (MQL) mode has been studied. Water-based alumina, hBN, MoS2, and WS2 nanofluids have been prepared and their anti-friction and anti-wear properties have been studied using ball-on-disk tribometry at different flow rates in MQL mode. The results are compared with a commercial lubricant (UNILUB 2032). Micromilling experiments were performed on Ti-6Al-4V using the prepared nanofluids in MQL mode. The following conclusions can be drawn based on the experimental results.
 Alumina nanofluids have shown the best anti-friction and anti-wear properties among all the nanofluids tested due to their spherical shape, which causes a ball bearing effect and reduces friction.
 Alumina nanoparticles also fill in the asperities in the workpiece surface due to their spherical shape, further reducing the contact between the sliding surfaces, thereby reducing friction and wear. This can be confirmed from the EDS of the wear tracks.
 The commercial lubricant has shown the best anti-friction and anti-wear performance due to the formation of a stable tribofilm by the commercial lubricant as compared to the nanofluids.
 An increase in flow rate further reduced friction and wear. This is due to the increase in the tribofilm stability and the increased availability of the nanoparticles, both of which tend to reduce friction and wear.  Alumina nanofluids and the commercial lubricant have shown significant reduction in machining forces and burr formation as compared to the other nanofluids. This is in correlation to the tribological performance of these cutting fluids.
 From the experimental results, it can be concluded that water-based alumina nanofluids can be a viable alternative for conventional cutting fluids in micromachining of Ti-6Al-4V.