Article Cite This: Langmuir XXXX, XXX, XXX−XXX
pubs.acs.org/Langmuir
On Lubrication States after a Running-In Process in Aqueous Lubrication Wenpeng Jia,† Pengpeng Bai,† Wenling Zhang,‡ Liran Ma,*,† Yonggang Meng,† and Yu Tian*,† †
State Key Laboratory of Tribology, Department of Mechanical Engineering, Tsinghua University, Beijing 100084, China Department of Chemical and Materials Engineering, University of Alberta, Edmonton, Alberta T6G 1H9, Canada
‡
Downloaded via RUTGERS UNIV on August 8, 2019 at 13:43:59 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.
S Supporting Information *
ABSTRACT: Recently, many studies have reported the ultralow friction coefficient of sliding friction between rigid solid surfaces in aqueous lubrication. A running-in process that goes through high-friction and friction-decreasing regions to a stable ultralow friction region is often required. However, the role of the friction-decreasing region is often ascribed to tribofilm formation in which complexity hindered the quantitative description of the running-in process and the prediction of its subsequent lubrication state. In this work, the frictional energy (Ef) dissipated in the running-in process of a poly(oligo(ethylene glycol) methyl ether acrylate) aqueous lubrication was related to the wear of solid surfaces under different conditions and lubrication states. Experimental results indicated that the high-friction region was in a boundary lubrication state, contributed to most of the wear, and significantly reduced the contact pressure, whereas the friction-decreasing region was in a mixed lubrication state, contributed only to the slight and slow removal of materials, and slightly reduced the contact pressure. Therefore, by establishing relationships among the wear scar diameter, Ef, and the Stribeck curve of the tribological system, the subsequent lubrication state after a running-in process under various working loads and sliding speeds could be quantitatively predicted. The running-in experiments with different aqueous lubrication systems showed good agreement with the prediction of this method. This investigation provides an effective method for the wear and lubrication state prediction after a running-in process, further proving the importance of the Stribeck curve for a lubrication system. This study may also have important implications for the strategy design of the running-in process in various industrial applications.
1. INTRODUCTION Friction leads to an enormous loss of energy and materials, which inspires huge innovative efforts for reducing friction and wear.1−7 Ultralow friction, which has been researched for several years, is highly demanded in nature and industry. Because aqueous lubrication widely exists in biological systems, ranging from joints, eyeballs, and eyelids to the gastric mucosa, the ultralow friction achieved by aqueous solutions, whose excellent aqueous lubrication mechanism is crucial in the study of interfacial phenomena, has attracted a great amount of attention. The mechanism of the aqueous lubrication was always ascribed to (1) the hydration effect,8−12 (2) the adsorbed boundary layer established at the interface, (3) the hydrodynamic effect, and (4) the running-in process. The surface force apparatus (SFA) invented by Israelachvili et al. has a profound influence as an effective tool for characterizing the hydration force.13 Various studies worldwide have successfully measured the surface and interface force by using SFA, thereby greatly contributing to the introduction of aqueous lubrication and its mechanism.14−18 Meanwhile, the hydrodynamic effect makes an important contribution to aqueous lubrication, especially in the stable ultralow friction coefficient (μ) stage. In most macrotypical tribological © XXXX American Chemical Society
processes, a running-in process is necessary for the achievement of ultralow friction,12,19−32 which helps smooth the contacting surfaces and reduce the collision among asperities. The running-in process of most aqueous lubrications is divided into three periods: a high-friction region (the first region), a friction-decreasing region (the second region), and a stable, ultralow μ region.21,25−27,33 Polymer brushes, as materials that can exert aqueous ultralow friction, have attracted increasing interest in tribological performance with respect to providing excellent lubricity and wear protection in mechanical, medical, and biological applications. In our previous work, a brush-type poly(oligo(ethylene glycol) methyl ether acrylate) (P(OEGMA)) was synthesized, and its aqueous solution exhibited excellent aqueous lubrication properties.34 However, the role of the running-in process has not yet been thoroughly investigated. Previous studies mentioned that the mechanism Special Issue: Intermolecular Forces and Interfacial Science Received: April 15, 2019 Revised: May 23, 2019 Published: May 24, 2019 A
DOI: 10.1021/acs.langmuir.9b01105 Langmuir XXXX, XXX, XXX−XXX
Article
Langmuir
Figure 1. (a) Sketch of the system of tribological tests. The tribopairs consist of a Si3N4 ball (φ = 4 mm) and an Al2O3 ceramic disk. The evolution of μ vs time in (b) varying pc, (c) Fn, and (d) v tribological tests.
mixed lubrication, and hydrodynamic lubrication, depending on the Sommerfeld number (ηv/P) based on the lubricant viscosity (η), v, and the contact pressure (P). In a running-in process, the contact pressure will generally gradually decrease because of the wear of friction pairs and affect the lubrication state. In this work, the running-in process of aqueous lubrication was investigated from the aspects of Ef and wear by using P(OEGMA) as a typical example. The effects of polymer concentration (pc), Fn, and v on the running-in process of P(OEGMA) aqueous solutions were investigated to establish a quantitative relationship between Ef and the wear of solid surfaces. The lubrication state could be predicted in accordance with the Stribeck curve of the tribological system based on the achieved relationship. The role of the decreasing period of the running-in process was also discussed.
of the second region in the running-in process was considered to be the establishing process of the boundary film.12,19,23 Nevertheless, the establishment of a boundary film should exist in the entire friction process, not only in the second region. Therefore, the mechanical behaviors of materials in the running-in process must be examined further. The energy concept has been used as an effective tool for analyzing the running-in process. From the energy perspective, the input friction force multiplied by the sliding distance (S) corresponds to the frictional energy (Ef), which is often dissipated in several ways: (1) frictional heat with increases in temperature and mechanical vibrations, (2) wear generation (worn surfaces and abrasive particles), and (3) changes in the physical properties of solid surfaces, such as plastic deformation.35−37 Among them, wear is a progressive loss of solid surface materials depending on various parameters, including the normal load (Fn), sliding velocity (v), S, and the hardness of tribopairs (H).35,37 Several classic models have been proposed to characterize material wear.35,37−41 The wear volume (WV) is proportional to Fn × S based on the Coulomb friction model. The Holm−Archard equation presented a proportional relation between WV and Fn × S/H. The friction energy intensity (Qf) was introduced to show the energy dissipated per unit of contact area. The energy pulse considered Qf with the time of sliding movement, but it works only for the contact where the contact point moves relative to both surfaces. Thus, the energy concept is able to analyze the procedure of wear and influencing factors and to predict some unknown parameters, such as the wear and working life of materials and the temperature increase. However, for the tribological process of aqueous lubrication, few studies have reported the running-in process from the aspect of energy and the quantitative relationship between wear and Ef, which will be significant. The Stribeck curve can efficiently describe the lubrication states of a tribological system, namely, boundary lubrication,
2. MATERIALS AND METHODS 2.1. Synthesis of P(OEGMA). The raw materials used for the synthesis include 2,2′-azo-bis (isobutyronitrile) (AIBN, Shanghai Macklin Biochemical Co., Ltd.), P(OEGMA) (Mn = 480 g/mol, Sigma-Aldrich), and RAFT agent (synthesized in accordance with the literature42). P(OEGMA) was synthesized via RAFT polymerization:. Initially, we dispersed 0.0167 mol of EG-MA and 0.4 mM RAFT chain-transfer agent into 10 mL of methylbenzene. Then, 0.133 mM AIBN and 10 mL of methylbenzene were dissolved in the mixture. Nitrogen was used to deoxygenate the resulting solution for 30 min before it was transferred into a sealed round-bottomed flask in a preheated oil bath at 70 °C for 5 h. Finally, the product was purified with petroleum ether several times and dried under vacuum overnight. 2.2. Tribological Tests. The P(OEGMA) aqueous solutions were prepared by dispersing the obtained P(OEGMA) with a numberaverage molecular weight (Mn) of ∼20K Daltons into ultrapure water. Tribopairs used in this study consisted of an Al2O3 ceramic substrate and the stationary Si3N4 ceramic ball (φ = 4 mm). Ball-on-disk tribological tests were performed on a Universal Micro Tribotester-3 (UMT-3, CETR, U.S.) in reciprocating mode with a stroke of 3 mm. After the tribopairs were mounted, the liquid lubricant was placed in B
DOI: 10.1021/acs.langmuir.9b01105 Langmuir XXXX, XXX, XXX−XXX
Article
Langmuir
Figure 2. Evolution of (a) μ vs time and (b) μ vs the equivalent Sommerfeld number (ηv/Fn) in a tribological test. The stable μ values (b) obtained in all tribological experiments vs ηv/Fn and for (c) a P(OEGMA) 50 wt % aqueous solution with varying Fn and v tests. Estimated Ef and corresponding μ as a function of (d) pc, (e) Fn, and (f) v. the contact area with a volume of 100 μL. In varying pc tests, P(OEGMA) aqueous solution with pc = 10−70 wt % was lubricated in the contact area under Fn = 4 N and v = 24 mm/s. In varying Fn tests, a P(OEGMA) 50 wt % aqueous solution was selected as the lubricant to use at v = 24 mm/s. Fn was set from 0.5 to 9 N, corresponding to the maximum contact pressure from 0.9 to 2.3 GPa, respectively. In varying v tests, a P(OEGMA) 50 wt % aqueous solution worked under Fn = 4 N, and the reciprocating frequency (f) between the ball and disk was set from 0.5 to 8 Hz, corresponding to v = 3 to 48 mm/s, respectively. Verified tribological tests were conducted with a P(OEGMA) 50 wt % aqueous-solution-lubricated Si3N4/Al2O3 interface, and the working condition was set with Fn = 3 N and v of variable values. Three different running-in processes were applied to perform the first step, where Fn and v were set to perform the running-in process several times. Here, three different running-in processes were carried out: 6 N and 24 mm/s for 800 s, 8 N and 24 mm/s for 650 s, and 9 N and 24 mm/s for 630 s. After the completion of the first step, the tribological test continued under the working conditions in the second step. Every test was conducted three times to reduce experimental errors. All tests were performed at room temperature at about 25 °C and ended when μ reached and maintained a stable value for 120 s. Here, μ was collected as a function of time (t). 2.3. Characterization Methods. The surface roughness of balls and disks was measured using a white light interferometer (Zygo
Nexview, Ametek, U.S.) before tribological tests. After tribological tests, the surfaces of balls and disks were cleaned with ultrapure water and ethanol and then dried with a nitrogen gun. We measured the wear scar diameter (WSD) on the surfaces of upper balls by using an optical microscope, and we observed the wear sizes and morphologies of wear tracks on disks with a white light interferometer.
3. RESULTS AND DISCUSSION 3.1. Lubricating Behaviors of P(OEGMA) Aqueous Solutions under Different Conditions. The lubrication state of P(OEGMA) aqueous solutions is affected by several factors, such as pc, Fn, and v. Tribological performance under different conditions was characterized by a ball-on-disk tribometer as shown in Figure 1a, wherein a contact and relative movement occur between the upper ball and the lower disk. Tribological tests using the same experimental parameters (Fn = 4 N and v = 24 mm/s) were conducted by adding P(OEGMA) aqueous solutions with different pc values to the contact area of the Si3N4/Al2O3 ceramic interface. Figure 1b and Figure S1 present the evolution of μ as a function of time. Samples with different pc values experienced different evolution processes for μ. For P(OEGMA) 10−60 wt % aqueous solutions, μ gradually reached ultralow values (
