- Email: [email protected]
Role of surface roughness and lubricant film thickness in nanolubrication of sliding components in adaptive optics
Journal of Colloid and Interface Science 353 (2011) 574–581
Contents lists available at ScienceDirect
Journal of Colloid and Interface Science www.elsevier.com/locate/jcis
Role of surface roughness and lubricant film thickness in nanolubrication of sliding components in adaptive optics Hyungoo Lee, Bharat Bhushan ⇑ Nanoprobe Laboratory for Bio- & Nanotechnology and Biomimetics, The Ohio State University, Columbus, OH 43210, United States
a r t i c l e
i n f o
Article history: Received 28 July 2010 Accepted 21 September 2010 Available online 27 September 2010 Keywords: Microprojector Lubricants Chemical bonding Chemical degradation Atomic force microscopy
a b s t r a c t Integrated microprojectors are being developed to project a large image on any surface chosen by users. For a laser-based microprojector, a piezo-electric based adaptive optics unit is adopted in the green laser architecture. Nanolubrication of adaptive optics sliding components is needed to reduce wear and for smooth operation. Mobile lubricant film thickness needs to be optimized for a given interface with a certain surface roughness to minimize stiction/friction and maximize durability. In this paper, the role of roughness and film thickness on adhesion, friction, and wear of the interface is studied. The results and associated mechanisms are presented. Ó 2010 Elsevier Inc. All rights reserved.
1. Introduction The development of microprojectors represents the miniaturization of the commercial projector. The microprojectors being sold commercially to date come as lightweight devices (about 100– 150 g) that can be attached to other hand-held devices (such as mobile phones) and can display images on any surface. One of these, laser-based microprojectors, uses two-dimensional pixelby-pixel scanning with modulated laser beams as the light source (Microvision, Redmond, Washington) [19,34]. It uses a single scan mirror to create a pixel array. It requires red, green, and blue lasers to provide a high contrast, high brightness, high resolution, and focus-free image and low power consumption. A compact green laser with high power is not commercially available. In order to produce a compact green laser, a frequency doubling technique is used with a distributed Bragg reflector (DBR) laser diode, where a 1060 nm wavelength light is passed through a second harmonic generating crystal to produce a 530 nm green laser [1,2]. To correct for any lens misalignment, the green laser module uses an adaptive optics component with a drive mechanism to align the optics and maintain a constant power output with time and temperature [2,3]. The drive mechanism is commonly referred to as the smooth impact drive mechanism (SIDM) (Fig. 1). The device consists of two SIDM units for x-axis and y-axis movement. The main components of the drive mechanism are the piezo element, driving rod, moving body, and friction plate. The moving body is a U-shaped frame upon which the lens is attached. The driving rod sits in the frame ⇑ Corresponding author. E-mail address: [email protected] (B. Bhushan). 0021-9797/$ - see front matter Ó 2010 Elsevier Inc. All rights reserved. doi:10.1016/j.jcis.2010.09.060
with friction plates on two sides and clamped with a leaf spring. Carbon fiber reinforced polymer (CFRP) is the material used as the driving rod in the drive mechanism [16,17]. It is used as a component for a wide variety of engineering applications. For the moving body, Zn alloy commonly used in automotive applications is used as the die-cast part. Wear is a significant concern for actuators such as the adaptive optics drive mechanism components described above, due to the high friction encountered during sliding [23,25,16,17]. The friction and wear behavior of materials is dependent on their mechanical properties, surface roughness, and the operating environment [7,8,10,11]. To improve the tribological performance of the sliding components, perfluoropolyether (PFPE) lubricants are applied to the surfaces. The ideal lubricant should be molecularly thick to protect the surface from wear, easily applied, able to chemically bond to the surface, and insensitive to the environment. PFPE lubricants are known to be most desirable [21,22,35,26,33,13,30,31]. They have low surface tension, high contact angle, and high adhesion to the substrate, allowing easy application and spreading onto the surface as well as providing hydrophobicity. Their chemical and thermal stability and low vapor pressure provide low degradation and low out-gassing. PFPEs have been extensively investigated, especially in the magnetic disk drive industry [4,14]. It has been shown that they reduce friction and wear, resulting in lower disk drive failure. A lubricant from Moresco (A20H), which has one hydroxyl group on one end and a PFPE backbone of a commonly used Z-DOL, is commonly used. At the opposite end of A20H, there is a cyclotriphosphazene group, giving A20H the characteristics of a phosphazene lubricant. A phosphazene lubricant renders the surface hydrophobic, minimizes stiction, and improves durability of a
575
H. Lee, B. Bhushan / Journal of Colloid and Interface Science 353 (2011) 574–581
Fig. 1. Schematic showing detailed construction of the drive mechanism (SIDM assembly) [2,3,29].
sample, especially in high humidity environments [33,13,15, 30,31]. The lubricant film can form menisci, which may result in high adhesion and stiction. In disk drive research, it has been shown that there is a critical hmobile/rcomposite below which the device should be operated for optimum tribological performance [20,36,37,14,7]. Here, hmobile is the total thickness of mobile liquid, and rcomposite is the composite standard deviation of surface roughness to the two mating surfaces. The objective of this research is to explore the role of surface roughness of CFRP and lubricant film thickness on adhesion, friction, and wear in order to obtain an optimum h/r. Results and associated mechanisms follow. 2. Experimental 2.1. Materials and sample preparation The CFRP rod, a carbon fiber reinforced polymer, is most often epoxy, but other polymers such as polyester, vinyl ester, and nylon are also used. In the fabrication of a CFRP rod, carbon fibers are aligned parallel to the long axis of the fiber and bound up by epoxy as a bundle. The dimension of a single fiber is about 2–5 lm in diameter and about 15 mm in length. It has a high strengthto-weight ratio due to its low density, and it has high strength. Moreover, it has a low coefficient of thermal expansion along the direction of the fiber. These are some of the reasons. The CFRP rod is chosen for an adaptive optics sliding component in microprojectors. The binder ratio measured using thermogravimetric analysis (TGA) is about 30%. The glass transition temperature (Tg) of the CFRP rod measured using a differential scanning calorimeter (DSC) is about 180 °C [16,17]. The CFRP rods were polished mechanically to obtain lower roughness (r) using a polishing wheel (EcoMet 3, Buehler, IL, US) at 50 rpm and a downward pressure of 2 N to obtain various roughnesses. Silicon carbide polishing pads (Buehler, IL, US) were used with grit sizes of p120, p1200, and p2400 and a nylon polishing cloth (40–7052) for 1 min, 5 min, 3 h and 3 h, respectively. The CFRP rods with 130 nm roughness were obtained by polishing using p120, p1200, and p2400 pads. For 55 nm roughness samples, all four polishing pads were used. During the polishing process, Al2O3 powder (0.3 lm, MicroPolish II, Buehler, IL, US) with water
was applied between the samples and the polishing pad. Material removal initially takes place on top of the asperities on carbon fibers, resulting in a flat surface on CFRP rods. For the nanolubrication studies, a lubricant with a hydroxyl group on one end and a cyclotriphosphazene group on the other end (Moresco A20H) was applied to a CFRP using a dip-coating technique. The method and the apparatus used have been described elsewhere [26,33]. Briefly, the dip-coater allows withdrawal of the samples from the lubricant reservoir at a constant velocity. The withdrawal speed ranged from 0.3 to 20 mm/s. The CFRP rod was submerged into a beaker containing a dilute solution of lubricant with a concentration of 0.4% lubricant in HFE 7100 (3 M, St. Paul, MN), which consists of isomers of methoxynonafluorobutane (C4F9OCH3). After 10 min, the CFRP rod was withdrawn from the solution. The thickness of the lubricant on the CFRP rod was measured by a thickness mapping technique using AFM [16]. The film thickness of lubricant on CFRP rod was controlled by the withdrawal speed of the dip-coating. The film thickness increases with the withdrawal speed as there is less time available for lubricant to drain at higher withdrawal speeds. Lubricant was deposited on both the unpolished samples (410 nm in RMS, roughness) and the polished (130 and 55 nm RMS) samples using a typical dip-coater [4,16,17]. The sample roughness and lubricant thickness used are shown in Table 1. The average film thicknesses of the lubricant as a function of the withdrawal speed of the dip-coater are shown in Fig. 2. It is observed that the lubricant film thickness decreases with decreasing surface roughness at the same withdrawal speed of the dip-coater. While withdrawing the CFRP rods from the lubricant reservoir, lubricant on smooth samples could drain easier than that from the rough surfaces. It is also shown that the lubricant thickness on epoxy is higher than that on carbon fibers. However, the thickness difference between epoxy and carbon fibers decreases with a decrease in sample roughness. This suggests that the uniformity of lubricant distribution on the CFRP rods is improved by decreasing the sample roughness. The length of error bars (standard deviation) becomes smaller with decreasing roughness of the CFRP rods. Bhushan et al. [16] reported that the lubricant on the CFRP rods is mobile and not chemically bonded. 2.2. Nanoscale surface height, adhesion, friction and lubricant thickness measurements A commercial AFM (Nanoscope IIIa, Veeco, Santa Barbara, CA, USA) was used for this study [10,11]. Silicon nitride tips of nominal 50 nm radius attached to the end of a triangular cantilever beam (DNP, spring constant of 0.12 N/m) were used for surface height, adhesion, friction, and lubricant film thickness measurements. Adhesive force and lubricant film thickness on CFRP rod was calculated using the force distance curve technique [12,24,18,27,16]. The experiments were performed at room temperature (21 °C) and 45–55% relative humidity. The force distance curves were collected at the same maximum cantilever deflection of 70 nm (relative trigger mode). In order to obtain a map of adhesive forces and lubricant film thickness, a 64 64 force distance curve array (total of 4096 measurement points) was collected over a scan area of 20 lm 20 lm with a 3 Hz scan rate for a CFRP. For each force distance curve, there are
Table 1 Sample roughness (RMS) and lubricant thickness used in experiments. Roughness (RMS, nm) Approximate lubricant thickness (h, nm)
410 15, 30, 45, 65, 115
130 15, 30, 45, 60
55 5, 10, 30
576
H. Lee, B. Bhushan / Journal of Colloid and Interface Science 353 (2011) 574–581
Fig. 2. Average film thicknesses of lubricant coated on carbon fiber and epoxy regions in the CFRP rod (left column) as well as the average film thickness over the entire CFRP rod (right column) as a function of withdrawal speed of the dip-coater for the rod with three different roughnesses. Scatter bars represent one standard deviation limits.
128 sampling points. A custom program coded in Matlab was used to calculate the lubricant thickness and adhesive force. The quantitative measurement of friction force was made using the method described by Bhushan [5,6,8,10,11] and Palacio and Bhushan [32]. The normal load was varied (300–2500 nN), and a friction force measurement was taken at each increment. By plotting the friction force as a function of normal load, an average coefficient of friction was obtained from the slope of the best fit line of the data.
diamond tip is about 100 nm, and the spring constant (k) is 10 N/ m. For scanning in the transverse direction, cantilever damage may occur at high loads and high scan speed, therefore, this study was conducted by scanning in the longitudinal direction. The tip was scanned 10 times along the carbon fiber direction at 30 lN of normal load. The scan area and rate are 10 lm and 3 Hz, respectively. The surface morphology of each sample was measured before and after the wear tests. 2.4. Contact angle
2.3. Nanoscale wear For wear study on the nanoscale, an AFM diamond tip was repeatedly scanned on the CFRP rod surfaces. The radius of the
The wetting properties of CFRP rods are important in understanding the tribological properties of the samples as a function of lubricant thickness and sample roughness, because water is
H. Lee, B. Bhushan / Journal of Colloid and Interface Science 353 (2011) 574–581
577
Fig. 3. Representative AFM maps for surface height, friction force, adhesive force, and film thickness distribution for unpolished (410 nm in RMS) CFRPs unlubricated and lubricated (65 nm thickness), and polished (130 nm in RMS) CFRPs unlubricated and lubricated (15 nm thickness). Shown above each image is a cross-section taken at a position denoted by the corresponding arrows.
prevalent in the air. The experiments were carried out at room temperature (21 °C) and in 45–55% relative humidity. The static contact angle was measured by placing a DI water droplet of
about 5 lL on the CFRP rods using a microsyringe. After placement, the droplet was imaged and analyzed using ScionImage software.
578
H. Lee, B. Bhushan / Journal of Colloid and Interface Science 353 (2011) 574–581
Fig. 4. Static contact angle as a function of lubricant film thickness and h/r for unpolished and polished CFRP rods.
3. Results and discussion 3.1. Nanoscale height, friction, adhesion, lubricant thickness distribution, and wear The surface height, friction force, adhesive force, and wear of samples with various roughnesses and lubricant thicknesses were obtained. The representative maps of the unpolished and one of the polished CFRP rods are shown in Fig. 3. In the surface height images, the bright long columns correspond to the individual carbon fibers, and the dark ones correspond to epoxy regions. In the unpolished samples, as compared to the lubricant thickness on the carbon fibers, a larger fraction of the lubricant resides in the recessed epoxy region acting as a lubricant reservoir. However, the lubricant on polished samples is distributed evenly all around. Contact angle is a property of interest for wetting phenomena. A higher contact angle is desirable for repulsion of condensed water from the environment. Fig. 4 shows the static contact angle data as
a function of lubricant film thickness for unpolished and polished samples. The static contact angle increases with an increase in film thickness and a decrease in surface roughness at a given film thickness. The contact angle data are also plotted as a function of h/r. Contact angle increases with h/r. Contact angle hysteresis values are of interest for drag reduction during fluid flow [28]. For both unlubricated and lubricated unpolished samples, contact angle hysteresis values are shown in Table 2 [17]). The contact angle hysteresis is believed to occur due to chemical heterogeneity and surface roughness. In a lubricated sample, the lubricant resides first in the recessed epoxy region, because this region acts as a lubricant reservoir. In the case of lubricant thickness lower than surface roughness (low h/r), lubricated and unlubricated areas on the sample surface are expected to result in a high degree of heterogeneity on the surface. However, with lubricant film thickness higher than roughness (high h/r), the lubricant resides not only in the recessed epoxy region but also on the exposed carbon fibers. It means that the heterogeneity is decreasing with increase in the lubricant thickness. Therefore, with the improved uniformity of the lubricant distribution observed at higher h and lower r (higher h/r), contact angle hysteresis is expected to decrease. Accordingly, it is expected that the actual microprojectors with the smoother CFRP rods operate in more stable and reliable conditions with better lubricant uniformity observed at higher h/r. Coefficient of friction (top, left) and adhesive force (bottom, right) data are shown in Fig. 5 (left column) as a function of lubricant film thickness on the unpolished and polished CFRP rods. With an increase in the film thickness, the coefficient of friction decreases for all three sample roughnesses, whereas adhesive force increases with an increase in lubricant film thickness for all samples. It is believed that increasing the lubricant film thickness leads to lower coefficient of friction due to lower overall shear strength, and adhesive force increases because of larger meniscus formation in a thick film [6,7,8,9,10]. With a decrease in roughness, the coefficient of friction decreases probably because of an increase in the contact area between the AFM tip and the sample surface, leading to more lubricant involved in the shearing. However, the enlarged contact area results in higher adhesive force. Even at the same lubricant film thickness, adhesive force on the low roughness surface is higher than that on high roughness. Wear depth at 30 lN normal load on carbon fibers and epoxy regions are shown in Fig. 5 (right column) as a function of lubricant film thickness on the unpolished and polished CFRP rods. Wear depth on carbon fibers and epoxy for all three sample roughnesses decreases with an increase in lubricant film thickness. The wear depth on epoxy is five times larger than that on carbon fibers, and also deeper than the thickness of lubricant film on epoxy. It indicates that the normal load of 30 lN used in the experiments was beyond the point of the plastic deformation of epoxy. With thicker lubricant on carbon fibers and epoxy, the contact between the AFM diamond tip and the CFRP rod surfaces is mitigated by the mobile fraction of lubricant molecules on the CFRP rods, resulting in a decrease in the wear depth. With the decrease in sample
Table 2 Measured contact angles of unlubricated and lubricated CFRP rods (unpolished, 410 nm RMS). Sample type Unlubricated Lubricated
*
Lubricant thickness (nm) 0 15 30 45 65 110
± values represent ± one standard deviation limits.
Static contact angle* (degrees) 69 (±0.5) 90 93 95 102 104
(±0.5) (±0.5) (±0.5) (±1) (±1)
Advancing angle* (degrees) 68 (±1) 99 101 103 106 108
(±1) (±1) (±1) (±2) (±3)
Receding angle* (degrees)
Contact angle hysteresis* (degrees)
50 (±1)
18
52 65 69 69 73
47 36 34 37 35
(±1) (±1) (±1) (±1) (±1)
H. Lee, B. Bhushan / Journal of Colloid and Interface Science 353 (2011) 574–581
579
Fig. 5. Coefficient of friction, adhesive force and wear depth on carbon fibers and epoxy regions at 30 lN of normal load as a function of lubricant thickness on the unpolished and polished CFRP rods.
surface roughness, it is observed that wear depth on the CFRP rods decreases, and the trendline slopes of wear depth on low roughness samples are high. As discussed earlier, the decrease in wear depth may be attributed to an increase in contact area and more lubricant involved in shearing. Therefore, decreasing surface roughness is beneficial for wear protection, but it leads to higher adhesive force. There is a trade-off between the coefficient of friction, adhesive force, and wear with lubricant film thickness and surface roughness of the CFRP rods. 3.2. Optimized lubricant thickness and roughness Lubricant film thickness and surface roughness should be optimized for low coefficient of friction, adhesive force, and wear. For this, the effect of hmobile/r on friction mechanism and wear (durability) is investigated. All data for the CFRP rods with three roughnesses with all lubricant films are collected and used to study the dependence of h/r on coefficient of friction, adhesive force, and wear. The results are shown in Fig. 6. The dotted lines in each plot are the trendlines for the data. With an increase in h/r, coefficient of friction and wear depth on epoxy and carbon fibers decrease, but adhesive force increases. To more easily compare all plots, the trendlines are also plotted in Fig. 7a. At high h/r, coefficient of friction and wear is low, but adhesive force is high. For the actual operation of microprojectors, it is expected that high h/r will lead to large adhesion and rest stiction [16]). With low h/r, to obtain a small adhesive force, the coefficient of friction and wear would be large, resulting in the potential failure of the drive operation. A high h/r is desirable for low coefficient of friction and wear. Therefore, the optimized value of h/r should be used for a given application. The dotted vertical line in the plot is the h/r value (0.16) of
the samples with 410 nm of RMS and 65 nm of lubricant thickness used in some commercial microprojectors. A schematic of the multiple contacting asperities region showing the effect of h/r on the interface of CFRP rods and a moving part of the microprojector is shown in Fig. 7b. Three distinct regimes of h/r are shown; low, intermediate, and high h/r. In each regime, the gray area between the CFRP rods and the moving part represent lubricant. The recessed epoxy areas of the CFRP rods act as a reservoir so that lubricant first fills in the recessed areas [16,17]). At low h/r, there are some contacting asperities of the moving part to CFRP rods, but there are no asperities immersed into the lubricant. In that case, lubricant has little effect on friction and adhesion. Hence, it results in low adhesive force (absence of meniscus contribution), but high coefficient of friction and wear. As lubricant film thickness increases, menisci start to form around the asperities coming in contact with the lubricant, which has the potential of generating adhesive force. At intermediate h/r, some asperities are immersed in lubricant, which initially has an attractive pressure higher than the disjoining pressure in the lubricant film. It results in an increase of adhesive force and a decrease of coefficient of friction and wear. At high h/r, saturation occurs. In the low h/r regime equivalent to ‘toe-dipping’ regime in Bhushan [9], adhesive force (Fad) between a single asperity and a surface due to meniscus contribution in an elastic contact can be modeled as [4,9],
F ad
W ½E rp ðrp =Rp Þ1=2 =½16:6cl ðcos h1 þ cos h2 Þ 1
ð1Þ
where W is normal load, E is the composite elastic modulus, rp is the standard deviation of the peak heights, Rp is the mean asperity
580
H. Lee, B. Bhushan / Journal of Colloid and Interface Science 353 (2011) 574–581
Fig. 6. The effect of h/r on coefficient of friction, adhesive force and wear depth on carbon fibers and epoxy regions at 30 lN of normal load. The dotted lines are the trendlines for the data.
Fig. 7. (a) The trend of the effect of h/r on coefficient of friction, adhesive force and wear depth. The dotted vertical line is the h/r value (0.16) of the samples with 410 nm of RMS and 65 nm of lubricant thickness used in some commercial microprojectors. (b) A schematic showing the rough interface of CFRP rods and a moving part of microprojector. The gray colored area represents lubricant.
H. Lee, B. Bhushan / Journal of Colloid and Interface Science 353 (2011) 574–581
radius, cl is surface tension of liquid, h1 and h2 are the contact angles of the liquid with the two mating surfaces. Adhesive force for a single asperity contact is mainly dominated by rp and Rp. It increases with a decrease in rp and an increase in Rp. For a multiple-asperity contact, total adhesive force can be obtained by multiplying the number of asperity contacts with adhesive force for an asperity contact given by Eq. (1). Higher rp and lower h reduce the number of contacting asperities. Thus, at low or intermediate h/r regimes, an increase in roughness (rp) of the CFRP rods or a decrease in h results in a decrease of adhesive force, however with an increase in coefficient of friction and wear. With the extreme regime of high h/r equivalent to ‘flooded’ regime in Bhushan [9], Fad can be modeled as [4,9],
F ad
Aa cl ðcos h1 þ cos h2 Þ h
ð2Þ
where Aa is the apparent surface area. From Eq. (2), it is observed that a decrease in the film thickness results in higher adhesive force. In the data presented, this is not observed. In the flooded regime, most asperities are immersed into the lubricant, leading to a large amount of meniscus formation and consequently high adhesive force with low coefficient of friction and wear. Any further decrease in r or in increase in h does not affect the number of menisci and adhesive force is expected to level off. 4. Conclusions The interplay of surface roughness and film thickness of lubricant needs to be optimized for low adhesion, friction, and wear. The effect of h/r on friction, adhesion, and durability is studied. With an increase in h/r, the uniformity of the lubricant distribution on the CFRP rods is improved, resulting in stable and reliable operation of the actual devices. Coefficient of friction and wear depth decrease with a decrease in h/r, but adhesive force increases due to an increase in the amount of lubricant involved in surface interface and in the contact area between the two mating materials, and vice versa. Therefore, there is a trade-off in lubricant film thickness and surface roughness for optimum values off coefficient of friction, adhesive force, and durability. The ratio of the mobile fraction of lubricant film thickness to sample roughness should be decided by the requirement and functionality of the actual devices. Acknowledgments We would like to thank Dr. Vikram Bhatia and Dr. Satish C. Chaparala from Corning Inc. for scientific discussions and financial support of this study. We also would like to acknowledge Dr. Yukihiro Ozeki and Mr. Yoshiaki Hata from Konica Minolta Opto, Japan for providing the CFRP rods. Dr. Manuel Palacio provided support in some of the data interpretation.
581
References [1] V. Bhatia, S.J. Gregorski, D. Pikula, S.C. Chaparala, D.A.S. Loeber, J. Gollier, Y. Ozeki, Y. Hata, K. Shibatani, F. Nagai, N. Mori, Y. Nakabayashi, N. Mitsugi, S. Nakano, Efficient and compact green laser incorporating adaptive optics for wide operating temperature range, in: SID International Symposium on Digest of Technical Papers, vol. 39, 2008, pp. 962–965. [2] V. Bhatia, M. Hempstead, J. Grochosinski, N. Sekiguchi, A. Okada, D. Loeber, J. Soc. Info. Display 17 (1) (2009) 1–6. [3] V. Bhatia, S.J. Gregorski, D. Pikula, S.C. Chaparala, D.A.S. Loeber, J. Gollier, J.D. Gregorski, M. Hempstead, Y. Ozeki, Y. Hata, K. Shibatani, F. Nagai, N. Mori, Y. Nakabayashi, N. Mitsugi, S. Nakano, J. Soc. Info. Display 17 (3) (2009) 1–6. [4] B. Bhushan, Tribology and Mechanics of Magnetic Storage Devices, Springer, New York, 1996. [5] B. Bhushan, Handbook of Micro/Nanotribology, second ed., CRC Press, Boca Raton, FL, 1999. [6] B. Bhushan, Principles and Applications of Tribology, Wiley, NY, 1999. [7] B. Bhushan, Macro- and microtribology of magnetic storage devices, in: B. Bhushan (Ed.), Modern Tribology Handbook, vol. 2, CRC Press, Boca Raton, Florida, 2001, pp. 1413–1513. [8] B. Bhushan, Introduction to Tribology, Wiley, NY, 2002. [9] B. Bhushan, J. Vac. Sci. Technol. B 21 (2003) 2262–2296. [10] B. Bhushan (Ed.), Nanotribology and Nanomechanics, second ed., Springer, Heidelberg, Germany, 2008. [11] B. Bhushan (Ed.), Springer Handbook of Nanotechnology, third ed., Springer, Heidelberg, Germany, 2010. [12] B. Bhushan, G.S. Blackman, ASME J. Tribol. 113 (1991) 452–457. [13] B. Bhushan, Z. Tao, Microsyst. Technol. 12 (2006) 579–587. [14] B. Bhushan, Z. Zhao, J. Info. Storage Proc. Syst. 1 (1999) 1–21. [15] B. Bhushan, M. Cichomski, Z. Tao, N.T. Tran, T. Ethen, C. Merton, R.E. Jewett, ASME J. Tribol. 129 (2007) 621–627. [16] B. Bhushan, H. Lee, S.C. Chaparala, V. Bhatia, Appl. Surface Sci. 256 (2010) 7545–7558. [17] B. Bhushan, H. Lee, S.C. Chaparala, V. Bhatia, Chemical bonding and degradation of ultrathin liquid films for nanolubrication of sliding components in adaptive optics, Colloids Surf. A 369 (2010) 39–52. [18] N. Chen, B. Bhushan, Ultramicroscopy 106 (2006) 755–764. [19] M. Freeman, M. Champion, S. Madhavan, Opt. Photonics News 20 (2009) 29–34. [20] C. Gao, B. Bhushan, Wear 190 (1995) 60–75. [21] V.N. Koinkar, B. Bhushan, J. Vac. Sci. Technol. A 14 (1996) 2378–2391. [22] V.N. Koinkar, B. Bhushan, J. Appl. Phys. 79 (1996) 8071–8075. [23] K. Kosaka, T. Iwabuchi, T. Baba, T. endo, H. Hashiguchi, H. Furukawa, et al., Jpn. J. Appl. Phys. 45 (2006) 1005–1011. [24] C. LaTorre, B. Bhushan, Ultramicroscopy 105 (2005) 155–175. [25] C.P. Lee, C.Y., Huang, T.H., Shih, Lens actuating device, US Patent 7, 446, 963, 2008. [26] H. Liu, B. Bhushan, Ultramicroscopy 97 (2003) 321–340. [27] R.A. Lodge, B. Bhushan, J. Vac. Sci. Technol. A 24 (2006) 1258–1269. [28] M. Nosonovsky, B. Bhushan, Multiscale Dissipative Mechanisms and Hierarchical Surfaces: Friction, Superhydrophobicity, and Biomimetics, Springer-Verlag, Heidelberg, Germany, 2008. [29] Y. Ozeki, T. Hata, K. Shibatani, F. Nagai, N. Mori, Y. Nakabayashi, et al., Konica Minolta Tech. Report 6 (2009) 82–87. [30] M. Palacio, B. Bhushan, J. Colloid Interface Sci. 315 (2007) 261–269. [31] M. Palacio, B. Bhushan, Ultramicroscopy 109 (2009) 980–990. [32] M. Palacio, B. Bhushan, Crit. Rev. Solid State Mater. Sci. 35 (2010) 73–104. [33] Z. Tao, B. Bhushan, Wear 259 (2005) 1352–1361. [34] J. Tauscher, W.O. Davis, D. Brown, M. Ellis, Y. Ma, M.E. Sherwood, D. Bowman, M.P. Helsel, S. Lee, J.W. Coy, Evolution of MEMS scanning mirrors for laser projection in compact consumer electronics, in: H. Schenk, W. Piyawattanametha (Eds.), Proceedings of the SPIE, vol. 7594, SPIE, Bellingham, Washington, 2010. [35] X. Zhao, B. Bhushan, Proc. Instn. Mech. Eng., Part J: J. Eng. Tribol. 215 (2001) 173–181. [36] Z. Zhao, B. Bhushan, Wear 202 (1996) 50–59. [37] Z. Zhao, B. Bhushan, IEEE Trans. Magn. 34 (1998) 1708–1710.
Contents lists available at ScienceDirect
Journal of Colloid and Interface Science www.elsevier.com/locate/jcis
Role of surface roughness and lubricant film thickness in nanolubrication of sliding components in adaptive optics Hyungoo Lee, Bharat Bhushan ⇑ Nanoprobe Laboratory for Bio- & Nanotechnology and Biomimetics, The Ohio State University, Columbus, OH 43210, United States
a r t i c l e
i n f o
Article history: Received 28 July 2010 Accepted 21 September 2010 Available online 27 September 2010 Keywords: Microprojector Lubricants Chemical bonding Chemical degradation Atomic force microscopy
a b s t r a c t Integrated microprojectors are being developed to project a large image on any surface chosen by users. For a laser-based microprojector, a piezo-electric based adaptive optics unit is adopted in the green laser architecture. Nanolubrication of adaptive optics sliding components is needed to reduce wear and for smooth operation. Mobile lubricant film thickness needs to be optimized for a given interface with a certain surface roughness to minimize stiction/friction and maximize durability. In this paper, the role of roughness and film thickness on adhesion, friction, and wear of the interface is studied. The results and associated mechanisms are presented. Ó 2010 Elsevier Inc. All rights reserved.
1. Introduction The development of microprojectors represents the miniaturization of the commercial projector. The microprojectors being sold commercially to date come as lightweight devices (about 100– 150 g) that can be attached to other hand-held devices (such as mobile phones) and can display images on any surface. One of these, laser-based microprojectors, uses two-dimensional pixelby-pixel scanning with modulated laser beams as the light source (Microvision, Redmond, Washington) [19,34]. It uses a single scan mirror to create a pixel array. It requires red, green, and blue lasers to provide a high contrast, high brightness, high resolution, and focus-free image and low power consumption. A compact green laser with high power is not commercially available. In order to produce a compact green laser, a frequency doubling technique is used with a distributed Bragg reflector (DBR) laser diode, where a 1060 nm wavelength light is passed through a second harmonic generating crystal to produce a 530 nm green laser [1,2]. To correct for any lens misalignment, the green laser module uses an adaptive optics component with a drive mechanism to align the optics and maintain a constant power output with time and temperature [2,3]. The drive mechanism is commonly referred to as the smooth impact drive mechanism (SIDM) (Fig. 1). The device consists of two SIDM units for x-axis and y-axis movement. The main components of the drive mechanism are the piezo element, driving rod, moving body, and friction plate. The moving body is a U-shaped frame upon which the lens is attached. The driving rod sits in the frame ⇑ Corresponding author. E-mail address: [email protected] (B. Bhushan). 0021-9797/$ - see front matter Ó 2010 Elsevier Inc. All rights reserved. doi:10.1016/j.jcis.2010.09.060
with friction plates on two sides and clamped with a leaf spring. Carbon fiber reinforced polymer (CFRP) is the material used as the driving rod in the drive mechanism [16,17]. It is used as a component for a wide variety of engineering applications. For the moving body, Zn alloy commonly used in automotive applications is used as the die-cast part. Wear is a significant concern for actuators such as the adaptive optics drive mechanism components described above, due to the high friction encountered during sliding [23,25,16,17]. The friction and wear behavior of materials is dependent on their mechanical properties, surface roughness, and the operating environment [7,8,10,11]. To improve the tribological performance of the sliding components, perfluoropolyether (PFPE) lubricants are applied to the surfaces. The ideal lubricant should be molecularly thick to protect the surface from wear, easily applied, able to chemically bond to the surface, and insensitive to the environment. PFPE lubricants are known to be most desirable [21,22,35,26,33,13,30,31]. They have low surface tension, high contact angle, and high adhesion to the substrate, allowing easy application and spreading onto the surface as well as providing hydrophobicity. Their chemical and thermal stability and low vapor pressure provide low degradation and low out-gassing. PFPEs have been extensively investigated, especially in the magnetic disk drive industry [4,14]. It has been shown that they reduce friction and wear, resulting in lower disk drive failure. A lubricant from Moresco (A20H), which has one hydroxyl group on one end and a PFPE backbone of a commonly used Z-DOL, is commonly used. At the opposite end of A20H, there is a cyclotriphosphazene group, giving A20H the characteristics of a phosphazene lubricant. A phosphazene lubricant renders the surface hydrophobic, minimizes stiction, and improves durability of a
575
H. Lee, B. Bhushan / Journal of Colloid and Interface Science 353 (2011) 574–581
Fig. 1. Schematic showing detailed construction of the drive mechanism (SIDM assembly) [2,3,29].
sample, especially in high humidity environments [33,13,15, 30,31]. The lubricant film can form menisci, which may result in high adhesion and stiction. In disk drive research, it has been shown that there is a critical hmobile/rcomposite below which the device should be operated for optimum tribological performance [20,36,37,14,7]. Here, hmobile is the total thickness of mobile liquid, and rcomposite is the composite standard deviation of surface roughness to the two mating surfaces. The objective of this research is to explore the role of surface roughness of CFRP and lubricant film thickness on adhesion, friction, and wear in order to obtain an optimum h/r. Results and associated mechanisms follow. 2. Experimental 2.1. Materials and sample preparation The CFRP rod, a carbon fiber reinforced polymer, is most often epoxy, but other polymers such as polyester, vinyl ester, and nylon are also used. In the fabrication of a CFRP rod, carbon fibers are aligned parallel to the long axis of the fiber and bound up by epoxy as a bundle. The dimension of a single fiber is about 2–5 lm in diameter and about 15 mm in length. It has a high strengthto-weight ratio due to its low density, and it has high strength. Moreover, it has a low coefficient of thermal expansion along the direction of the fiber. These are some of the reasons. The CFRP rod is chosen for an adaptive optics sliding component in microprojectors. The binder ratio measured using thermogravimetric analysis (TGA) is about 30%. The glass transition temperature (Tg) of the CFRP rod measured using a differential scanning calorimeter (DSC) is about 180 °C [16,17]. The CFRP rods were polished mechanically to obtain lower roughness (r) using a polishing wheel (EcoMet 3, Buehler, IL, US) at 50 rpm and a downward pressure of 2 N to obtain various roughnesses. Silicon carbide polishing pads (Buehler, IL, US) were used with grit sizes of p120, p1200, and p2400 and a nylon polishing cloth (40–7052) for 1 min, 5 min, 3 h and 3 h, respectively. The CFRP rods with 130 nm roughness were obtained by polishing using p120, p1200, and p2400 pads. For 55 nm roughness samples, all four polishing pads were used. During the polishing process, Al2O3 powder (0.3 lm, MicroPolish II, Buehler, IL, US) with water
was applied between the samples and the polishing pad. Material removal initially takes place on top of the asperities on carbon fibers, resulting in a flat surface on CFRP rods. For the nanolubrication studies, a lubricant with a hydroxyl group on one end and a cyclotriphosphazene group on the other end (Moresco A20H) was applied to a CFRP using a dip-coating technique. The method and the apparatus used have been described elsewhere [26,33]. Briefly, the dip-coater allows withdrawal of the samples from the lubricant reservoir at a constant velocity. The withdrawal speed ranged from 0.3 to 20 mm/s. The CFRP rod was submerged into a beaker containing a dilute solution of lubricant with a concentration of 0.4% lubricant in HFE 7100 (3 M, St. Paul, MN), which consists of isomers of methoxynonafluorobutane (C4F9OCH3). After 10 min, the CFRP rod was withdrawn from the solution. The thickness of the lubricant on the CFRP rod was measured by a thickness mapping technique using AFM [16]. The film thickness of lubricant on CFRP rod was controlled by the withdrawal speed of the dip-coating. The film thickness increases with the withdrawal speed as there is less time available for lubricant to drain at higher withdrawal speeds. Lubricant was deposited on both the unpolished samples (410 nm in RMS, roughness) and the polished (130 and 55 nm RMS) samples using a typical dip-coater [4,16,17]. The sample roughness and lubricant thickness used are shown in Table 1. The average film thicknesses of the lubricant as a function of the withdrawal speed of the dip-coater are shown in Fig. 2. It is observed that the lubricant film thickness decreases with decreasing surface roughness at the same withdrawal speed of the dip-coater. While withdrawing the CFRP rods from the lubricant reservoir, lubricant on smooth samples could drain easier than that from the rough surfaces. It is also shown that the lubricant thickness on epoxy is higher than that on carbon fibers. However, the thickness difference between epoxy and carbon fibers decreases with a decrease in sample roughness. This suggests that the uniformity of lubricant distribution on the CFRP rods is improved by decreasing the sample roughness. The length of error bars (standard deviation) becomes smaller with decreasing roughness of the CFRP rods. Bhushan et al. [16] reported that the lubricant on the CFRP rods is mobile and not chemically bonded. 2.2. Nanoscale surface height, adhesion, friction and lubricant thickness measurements A commercial AFM (Nanoscope IIIa, Veeco, Santa Barbara, CA, USA) was used for this study [10,11]. Silicon nitride tips of nominal 50 nm radius attached to the end of a triangular cantilever beam (DNP, spring constant of 0.12 N/m) were used for surface height, adhesion, friction, and lubricant film thickness measurements. Adhesive force and lubricant film thickness on CFRP rod was calculated using the force distance curve technique [12,24,18,27,16]. The experiments were performed at room temperature (21 °C) and 45–55% relative humidity. The force distance curves were collected at the same maximum cantilever deflection of 70 nm (relative trigger mode). In order to obtain a map of adhesive forces and lubricant film thickness, a 64 64 force distance curve array (total of 4096 measurement points) was collected over a scan area of 20 lm 20 lm with a 3 Hz scan rate for a CFRP. For each force distance curve, there are
Table 1 Sample roughness (RMS) and lubricant thickness used in experiments. Roughness (RMS, nm) Approximate lubricant thickness (h, nm)
410 15, 30, 45, 65, 115
130 15, 30, 45, 60
55 5, 10, 30
576
H. Lee, B. Bhushan / Journal of Colloid and Interface Science 353 (2011) 574–581
Fig. 2. Average film thicknesses of lubricant coated on carbon fiber and epoxy regions in the CFRP rod (left column) as well as the average film thickness over the entire CFRP rod (right column) as a function of withdrawal speed of the dip-coater for the rod with three different roughnesses. Scatter bars represent one standard deviation limits.
128 sampling points. A custom program coded in Matlab was used to calculate the lubricant thickness and adhesive force. The quantitative measurement of friction force was made using the method described by Bhushan [5,6,8,10,11] and Palacio and Bhushan [32]. The normal load was varied (300–2500 nN), and a friction force measurement was taken at each increment. By plotting the friction force as a function of normal load, an average coefficient of friction was obtained from the slope of the best fit line of the data.
diamond tip is about 100 nm, and the spring constant (k) is 10 N/ m. For scanning in the transverse direction, cantilever damage may occur at high loads and high scan speed, therefore, this study was conducted by scanning in the longitudinal direction. The tip was scanned 10 times along the carbon fiber direction at 30 lN of normal load. The scan area and rate are 10 lm and 3 Hz, respectively. The surface morphology of each sample was measured before and after the wear tests. 2.4. Contact angle
2.3. Nanoscale wear For wear study on the nanoscale, an AFM diamond tip was repeatedly scanned on the CFRP rod surfaces. The radius of the
The wetting properties of CFRP rods are important in understanding the tribological properties of the samples as a function of lubricant thickness and sample roughness, because water is
H. Lee, B. Bhushan / Journal of Colloid and Interface Science 353 (2011) 574–581
577
Fig. 3. Representative AFM maps for surface height, friction force, adhesive force, and film thickness distribution for unpolished (410 nm in RMS) CFRPs unlubricated and lubricated (65 nm thickness), and polished (130 nm in RMS) CFRPs unlubricated and lubricated (15 nm thickness). Shown above each image is a cross-section taken at a position denoted by the corresponding arrows.
prevalent in the air. The experiments were carried out at room temperature (21 °C) and in 45–55% relative humidity. The static contact angle was measured by placing a DI water droplet of
about 5 lL on the CFRP rods using a microsyringe. After placement, the droplet was imaged and analyzed using ScionImage software.
578
H. Lee, B. Bhushan / Journal of Colloid and Interface Science 353 (2011) 574–581
Fig. 4. Static contact angle as a function of lubricant film thickness and h/r for unpolished and polished CFRP rods.
3. Results and discussion 3.1. Nanoscale height, friction, adhesion, lubricant thickness distribution, and wear The surface height, friction force, adhesive force, and wear of samples with various roughnesses and lubricant thicknesses were obtained. The representative maps of the unpolished and one of the polished CFRP rods are shown in Fig. 3. In the surface height images, the bright long columns correspond to the individual carbon fibers, and the dark ones correspond to epoxy regions. In the unpolished samples, as compared to the lubricant thickness on the carbon fibers, a larger fraction of the lubricant resides in the recessed epoxy region acting as a lubricant reservoir. However, the lubricant on polished samples is distributed evenly all around. Contact angle is a property of interest for wetting phenomena. A higher contact angle is desirable for repulsion of condensed water from the environment. Fig. 4 shows the static contact angle data as
a function of lubricant film thickness for unpolished and polished samples. The static contact angle increases with an increase in film thickness and a decrease in surface roughness at a given film thickness. The contact angle data are also plotted as a function of h/r. Contact angle increases with h/r. Contact angle hysteresis values are of interest for drag reduction during fluid flow [28]. For both unlubricated and lubricated unpolished samples, contact angle hysteresis values are shown in Table 2 [17]). The contact angle hysteresis is believed to occur due to chemical heterogeneity and surface roughness. In a lubricated sample, the lubricant resides first in the recessed epoxy region, because this region acts as a lubricant reservoir. In the case of lubricant thickness lower than surface roughness (low h/r), lubricated and unlubricated areas on the sample surface are expected to result in a high degree of heterogeneity on the surface. However, with lubricant film thickness higher than roughness (high h/r), the lubricant resides not only in the recessed epoxy region but also on the exposed carbon fibers. It means that the heterogeneity is decreasing with increase in the lubricant thickness. Therefore, with the improved uniformity of the lubricant distribution observed at higher h and lower r (higher h/r), contact angle hysteresis is expected to decrease. Accordingly, it is expected that the actual microprojectors with the smoother CFRP rods operate in more stable and reliable conditions with better lubricant uniformity observed at higher h/r. Coefficient of friction (top, left) and adhesive force (bottom, right) data are shown in Fig. 5 (left column) as a function of lubricant film thickness on the unpolished and polished CFRP rods. With an increase in the film thickness, the coefficient of friction decreases for all three sample roughnesses, whereas adhesive force increases with an increase in lubricant film thickness for all samples. It is believed that increasing the lubricant film thickness leads to lower coefficient of friction due to lower overall shear strength, and adhesive force increases because of larger meniscus formation in a thick film [6,7,8,9,10]. With a decrease in roughness, the coefficient of friction decreases probably because of an increase in the contact area between the AFM tip and the sample surface, leading to more lubricant involved in the shearing. However, the enlarged contact area results in higher adhesive force. Even at the same lubricant film thickness, adhesive force on the low roughness surface is higher than that on high roughness. Wear depth at 30 lN normal load on carbon fibers and epoxy regions are shown in Fig. 5 (right column) as a function of lubricant film thickness on the unpolished and polished CFRP rods. Wear depth on carbon fibers and epoxy for all three sample roughnesses decreases with an increase in lubricant film thickness. The wear depth on epoxy is five times larger than that on carbon fibers, and also deeper than the thickness of lubricant film on epoxy. It indicates that the normal load of 30 lN used in the experiments was beyond the point of the plastic deformation of epoxy. With thicker lubricant on carbon fibers and epoxy, the contact between the AFM diamond tip and the CFRP rod surfaces is mitigated by the mobile fraction of lubricant molecules on the CFRP rods, resulting in a decrease in the wear depth. With the decrease in sample
Table 2 Measured contact angles of unlubricated and lubricated CFRP rods (unpolished, 410 nm RMS). Sample type Unlubricated Lubricated
*
Lubricant thickness (nm) 0 15 30 45 65 110
± values represent ± one standard deviation limits.
Static contact angle* (degrees) 69 (±0.5) 90 93 95 102 104
(±0.5) (±0.5) (±0.5) (±1) (±1)
Advancing angle* (degrees) 68 (±1) 99 101 103 106 108
(±1) (±1) (±1) (±2) (±3)
Receding angle* (degrees)
Contact angle hysteresis* (degrees)
50 (±1)
18
52 65 69 69 73
47 36 34 37 35
(±1) (±1) (±1) (±1) (±1)
H. Lee, B. Bhushan / Journal of Colloid and Interface Science 353 (2011) 574–581
579
Fig. 5. Coefficient of friction, adhesive force and wear depth on carbon fibers and epoxy regions at 30 lN of normal load as a function of lubricant thickness on the unpolished and polished CFRP rods.
surface roughness, it is observed that wear depth on the CFRP rods decreases, and the trendline slopes of wear depth on low roughness samples are high. As discussed earlier, the decrease in wear depth may be attributed to an increase in contact area and more lubricant involved in shearing. Therefore, decreasing surface roughness is beneficial for wear protection, but it leads to higher adhesive force. There is a trade-off between the coefficient of friction, adhesive force, and wear with lubricant film thickness and surface roughness of the CFRP rods. 3.2. Optimized lubricant thickness and roughness Lubricant film thickness and surface roughness should be optimized for low coefficient of friction, adhesive force, and wear. For this, the effect of hmobile/r on friction mechanism and wear (durability) is investigated. All data for the CFRP rods with three roughnesses with all lubricant films are collected and used to study the dependence of h/r on coefficient of friction, adhesive force, and wear. The results are shown in Fig. 6. The dotted lines in each plot are the trendlines for the data. With an increase in h/r, coefficient of friction and wear depth on epoxy and carbon fibers decrease, but adhesive force increases. To more easily compare all plots, the trendlines are also plotted in Fig. 7a. At high h/r, coefficient of friction and wear is low, but adhesive force is high. For the actual operation of microprojectors, it is expected that high h/r will lead to large adhesion and rest stiction [16]). With low h/r, to obtain a small adhesive force, the coefficient of friction and wear would be large, resulting in the potential failure of the drive operation. A high h/r is desirable for low coefficient of friction and wear. Therefore, the optimized value of h/r should be used for a given application. The dotted vertical line in the plot is the h/r value (0.16) of
the samples with 410 nm of RMS and 65 nm of lubricant thickness used in some commercial microprojectors. A schematic of the multiple contacting asperities region showing the effect of h/r on the interface of CFRP rods and a moving part of the microprojector is shown in Fig. 7b. Three distinct regimes of h/r are shown; low, intermediate, and high h/r. In each regime, the gray area between the CFRP rods and the moving part represent lubricant. The recessed epoxy areas of the CFRP rods act as a reservoir so that lubricant first fills in the recessed areas [16,17]). At low h/r, there are some contacting asperities of the moving part to CFRP rods, but there are no asperities immersed into the lubricant. In that case, lubricant has little effect on friction and adhesion. Hence, it results in low adhesive force (absence of meniscus contribution), but high coefficient of friction and wear. As lubricant film thickness increases, menisci start to form around the asperities coming in contact with the lubricant, which has the potential of generating adhesive force. At intermediate h/r, some asperities are immersed in lubricant, which initially has an attractive pressure higher than the disjoining pressure in the lubricant film. It results in an increase of adhesive force and a decrease of coefficient of friction and wear. At high h/r, saturation occurs. In the low h/r regime equivalent to ‘toe-dipping’ regime in Bhushan [9], adhesive force (Fad) between a single asperity and a surface due to meniscus contribution in an elastic contact can be modeled as [4,9],
F ad
W ½E rp ðrp =Rp Þ1=2 =½16:6cl ðcos h1 þ cos h2 Þ 1
ð1Þ
where W is normal load, E is the composite elastic modulus, rp is the standard deviation of the peak heights, Rp is the mean asperity
580
H. Lee, B. Bhushan / Journal of Colloid and Interface Science 353 (2011) 574–581
Fig. 6. The effect of h/r on coefficient of friction, adhesive force and wear depth on carbon fibers and epoxy regions at 30 lN of normal load. The dotted lines are the trendlines for the data.
Fig. 7. (a) The trend of the effect of h/r on coefficient of friction, adhesive force and wear depth. The dotted vertical line is the h/r value (0.16) of the samples with 410 nm of RMS and 65 nm of lubricant thickness used in some commercial microprojectors. (b) A schematic showing the rough interface of CFRP rods and a moving part of microprojector. The gray colored area represents lubricant.
H. Lee, B. Bhushan / Journal of Colloid and Interface Science 353 (2011) 574–581
radius, cl is surface tension of liquid, h1 and h2 are the contact angles of the liquid with the two mating surfaces. Adhesive force for a single asperity contact is mainly dominated by rp and Rp. It increases with a decrease in rp and an increase in Rp. For a multiple-asperity contact, total adhesive force can be obtained by multiplying the number of asperity contacts with adhesive force for an asperity contact given by Eq. (1). Higher rp and lower h reduce the number of contacting asperities. Thus, at low or intermediate h/r regimes, an increase in roughness (rp) of the CFRP rods or a decrease in h results in a decrease of adhesive force, however with an increase in coefficient of friction and wear. With the extreme regime of high h/r equivalent to ‘flooded’ regime in Bhushan [9], Fad can be modeled as [4,9],
F ad
Aa cl ðcos h1 þ cos h2 Þ h
ð2Þ
where Aa is the apparent surface area. From Eq. (2), it is observed that a decrease in the film thickness results in higher adhesive force. In the data presented, this is not observed. In the flooded regime, most asperities are immersed into the lubricant, leading to a large amount of meniscus formation and consequently high adhesive force with low coefficient of friction and wear. Any further decrease in r or in increase in h does not affect the number of menisci and adhesive force is expected to level off. 4. Conclusions The interplay of surface roughness and film thickness of lubricant needs to be optimized for low adhesion, friction, and wear. The effect of h/r on friction, adhesion, and durability is studied. With an increase in h/r, the uniformity of the lubricant distribution on the CFRP rods is improved, resulting in stable and reliable operation of the actual devices. Coefficient of friction and wear depth decrease with a decrease in h/r, but adhesive force increases due to an increase in the amount of lubricant involved in surface interface and in the contact area between the two mating materials, and vice versa. Therefore, there is a trade-off in lubricant film thickness and surface roughness for optimum values off coefficient of friction, adhesive force, and durability. The ratio of the mobile fraction of lubricant film thickness to sample roughness should be decided by the requirement and functionality of the actual devices. Acknowledgments We would like to thank Dr. Vikram Bhatia and Dr. Satish C. Chaparala from Corning Inc. for scientific discussions and financial support of this study. We also would like to acknowledge Dr. Yukihiro Ozeki and Mr. Yoshiaki Hata from Konica Minolta Opto, Japan for providing the CFRP rods. Dr. Manuel Palacio provided support in some of the data interpretation.
581
References [1] V. Bhatia, S.J. Gregorski, D. Pikula, S.C. Chaparala, D.A.S. Loeber, J. Gollier, Y. Ozeki, Y. Hata, K. Shibatani, F. Nagai, N. Mori, Y. Nakabayashi, N. Mitsugi, S. Nakano, Efficient and compact green laser incorporating adaptive optics for wide operating temperature range, in: SID International Symposium on Digest of Technical Papers, vol. 39, 2008, pp. 962–965. [2] V. Bhatia, M. Hempstead, J. Grochosinski, N. Sekiguchi, A. Okada, D. Loeber, J. Soc. Info. Display 17 (1) (2009) 1–6. [3] V. Bhatia, S.J. Gregorski, D. Pikula, S.C. Chaparala, D.A.S. Loeber, J. Gollier, J.D. Gregorski, M. Hempstead, Y. Ozeki, Y. Hata, K. Shibatani, F. Nagai, N. Mori, Y. Nakabayashi, N. Mitsugi, S. Nakano, J. Soc. Info. Display 17 (3) (2009) 1–6. [4] B. Bhushan, Tribology and Mechanics of Magnetic Storage Devices, Springer, New York, 1996. [5] B. Bhushan, Handbook of Micro/Nanotribology, second ed., CRC Press, Boca Raton, FL, 1999. [6] B. Bhushan, Principles and Applications of Tribology, Wiley, NY, 1999. [7] B. Bhushan, Macro- and microtribology of magnetic storage devices, in: B. Bhushan (Ed.), Modern Tribology Handbook, vol. 2, CRC Press, Boca Raton, Florida, 2001, pp. 1413–1513. [8] B. Bhushan, Introduction to Tribology, Wiley, NY, 2002. [9] B. Bhushan, J. Vac. Sci. Technol. B 21 (2003) 2262–2296. [10] B. Bhushan (Ed.), Nanotribology and Nanomechanics, second ed., Springer, Heidelberg, Germany, 2008. [11] B. Bhushan (Ed.), Springer Handbook of Nanotechnology, third ed., Springer, Heidelberg, Germany, 2010. [12] B. Bhushan, G.S. Blackman, ASME J. Tribol. 113 (1991) 452–457. [13] B. Bhushan, Z. Tao, Microsyst. Technol. 12 (2006) 579–587. [14] B. Bhushan, Z. Zhao, J. Info. Storage Proc. Syst. 1 (1999) 1–21. [15] B. Bhushan, M. Cichomski, Z. Tao, N.T. Tran, T. Ethen, C. Merton, R.E. Jewett, ASME J. Tribol. 129 (2007) 621–627. [16] B. Bhushan, H. Lee, S.C. Chaparala, V. Bhatia, Appl. Surface Sci. 256 (2010) 7545–7558. [17] B. Bhushan, H. Lee, S.C. Chaparala, V. Bhatia, Chemical bonding and degradation of ultrathin liquid films for nanolubrication of sliding components in adaptive optics, Colloids Surf. A 369 (2010) 39–52. [18] N. Chen, B. Bhushan, Ultramicroscopy 106 (2006) 755–764. [19] M. Freeman, M. Champion, S. Madhavan, Opt. Photonics News 20 (2009) 29–34. [20] C. Gao, B. Bhushan, Wear 190 (1995) 60–75. [21] V.N. Koinkar, B. Bhushan, J. Vac. Sci. Technol. A 14 (1996) 2378–2391. [22] V.N. Koinkar, B. Bhushan, J. Appl. Phys. 79 (1996) 8071–8075. [23] K. Kosaka, T. Iwabuchi, T. Baba, T. endo, H. Hashiguchi, H. Furukawa, et al., Jpn. J. Appl. Phys. 45 (2006) 1005–1011. [24] C. LaTorre, B. Bhushan, Ultramicroscopy 105 (2005) 155–175. [25] C.P. Lee, C.Y., Huang, T.H., Shih, Lens actuating device, US Patent 7, 446, 963, 2008. [26] H. Liu, B. Bhushan, Ultramicroscopy 97 (2003) 321–340. [27] R.A. Lodge, B. Bhushan, J. Vac. Sci. Technol. A 24 (2006) 1258–1269. [28] M. Nosonovsky, B. Bhushan, Multiscale Dissipative Mechanisms and Hierarchical Surfaces: Friction, Superhydrophobicity, and Biomimetics, Springer-Verlag, Heidelberg, Germany, 2008. [29] Y. Ozeki, T. Hata, K. Shibatani, F. Nagai, N. Mori, Y. Nakabayashi, et al., Konica Minolta Tech. Report 6 (2009) 82–87. [30] M. Palacio, B. Bhushan, J. Colloid Interface Sci. 315 (2007) 261–269. [31] M. Palacio, B. Bhushan, Ultramicroscopy 109 (2009) 980–990. [32] M. Palacio, B. Bhushan, Crit. Rev. Solid State Mater. Sci. 35 (2010) 73–104. [33] Z. Tao, B. Bhushan, Wear 259 (2005) 1352–1361. [34] J. Tauscher, W.O. Davis, D. Brown, M. Ellis, Y. Ma, M.E. Sherwood, D. Bowman, M.P. Helsel, S. Lee, J.W. Coy, Evolution of MEMS scanning mirrors for laser projection in compact consumer electronics, in: H. Schenk, W. Piyawattanametha (Eds.), Proceedings of the SPIE, vol. 7594, SPIE, Bellingham, Washington, 2010. [35] X. Zhao, B. Bhushan, Proc. Instn. Mech. Eng., Part J: J. Eng. Tribol. 215 (2001) 173–181. [36] Z. Zhao, B. Bhushan, Wear 202 (1996) 50–59. [37] Z. Zhao, B. Bhushan, IEEE Trans. Magn. 34 (1998) 1708–1710.