Macroscopic friction

Friction on the macroscopic scale can take place either between dry contacts or between lubricated surfaces. An intermediate case called boundary lubrication is friction in which the surfaces are not separated by a thick layer of lubricant but just by surface layers such as oxide layers on metals or by a few molecular layers of adsorbed lubricants. 

In his original studies Amontons found a coefficient of friction of 0.3. Meanwhile it has become clear that friction coefficients can assume a whole range of values. With metals, a clear difference exists between clean metal surfaces, oxidized metal surfaces, and metal surfaces with adsorbed gas. Clean metals have coefficients of friction of 3-7. With oxidation, the value decreases to 0.6-1.0. A consequence is that the coefficient of friction can depend on the load. For small loads, friction is determined by the oxide coating. At high loads the microcontacts penetrate the oxide coating, the bare metals come into contact, and the coefficient of friction increases. 

There is a similar effect with diamond. In air we have //, = 0.1. In a vacuum, after heating to remove contamination and an oxide layer, //, rises approximately seven times. This can be explained again with an oxide coating. Many materials, in particular relatively isotropic, soft crystals such as krypton or sodium chloride, tie in the range 0.5-1.0. 

Values from tables of friction coefficients always have to be used with caution, since the experimental results not only depend on the materials but also on surface preparation, which is often not well characterized. In the case of plastic deformation, the static coefficient of friction may depend on contact time. Creeping motion due to thermally activated processes leads to an increase in the true contact area and hence the friction coefficient with time. This can often be described by a logarithmic time dependence 

---

Ultra-high vacuum (UHV) surface science methods allow preparation and characterization of perfectly clean, well ordered surfaces of single crystalline materials. By preparing pairs of such surfaces it is possible to fonn interfaces under highly controlled conditions. Furthennore, thin films of adsorbed species can be produced and characterized using a wide variety of methods. Surface science methods have been coupled with UHV measurements of macroscopic friction forces. Such measurements have demonstrated that adsorbate film thicknesses of a few monolayers are sufficient to lubricate metal surfaces [12, 181. 

Friction and Adhesion. The coefficient of friction p. is the constant of proportionality between the normal force P between two materials in contact and the perpendicular force F required to move one of the materials relative to the other. Macroscopic friction occurs from the contact of asperities on opposing surfaces as they sHde past each other. On the atomic level friction occurs from the formation of bonds between adjacent atoms as they sHde past one another. Friction coefficients are usually measured using a sliding pin on a disk arrangement. Friction coefficients for ceramic fibers in a matrix have been measured using fiber pushout tests (53). For various material combinations (43) ... 

An elegant explanation for the unusual viscosity dependence was provided by the non-Markovian rate theory (NMRT) of Grote and Hynes [149] which incorporates the idea of frequency dependence of the friction. According to this theory the friction experienced by the reactive motion is not the zero frequency macroscopic friction (related to viscosity) but the friction at a finite frequency which itself depends on the barrier curvature. The rate is obtained by a self-consistent calculation involving the frequency-dependent friction. 

Equation (320) predicts the TST result for very weak friction (Ar to ) and predicts the Kramers result for low barrier frequency (i.e., (ob —> 0) so that (2r) can be replaced by (0) in Eq. (322). If die barrier frequency is large (ia>b > 1013 s 1) and the friction is not negligible ( (0)/fi — cob), then the situation is not so straightforward. In this regime, which often turns out to be the relevant one experimentally, the effective friction (2r) can be quite small even if the zero frequency (i.e., the macroscopic) friction (proportional to viscosity) is very large. The non-Markovian effects can play a very important role in this regime. 

The second empirical law for dry, macroscopic friction is that of Coulomb Friction does not depend on the sliding velocity. 

MOLECULAR-LEVEL EXPLANATION OF MACROSCOPIC FRICTION LAWS ... 

The above results suggest that adsorbed molecules and other third bodies may prove key to understanding macroscopic friction measurements. It will be interesting to extend these studies to more realistic molecular potentials and to rough surfaces. Similar arguments apply to other so-called third bodies that are trapped between two surfaces, such as wear debris, grit, dust, and so on. 

To what extent does macroscopic friction , as indicated by solvent viscosity or by inverse self-diffusion coefficient, really reflects the microscopic friction experienced by the reaction coordinate ... 

Surprising results, compared with macroscopic friction, were obtained for PDMS 6, with a higher nanoscale friction before extraction of free chains. The influence of extraction was also more pronounced for lower normal loads. Macroscopic tests previously showed a sHght increase in friction after extraction. 

After the preparation of the PMAA brushes, their lubrication ability under low contact pressures was tested in a neutral aqueous solution. It was shown that the macroscopic friction between polyelectrolyle bmshes of different molecular weights and soft, hydrophilic ox-PDMS pins was below the detection limit of the employed pin-on-disk tribometer over the entire speed range tested. While the PMAA bmshes could not be distinguished with n vs sliding speed plots, the long-term stability of short 15 nm PMAA bmshes was shown to be inferior to long bmshes (240 nm dry thickness). A further comparison between... 

Pin-on-disk tribometers (CSEM, Neuchatel, Switzerland) were used to measure macroscopic frictional forces under pure sliding conditions. Two tribometers operating in different speed ranges were employed to enable the sliding speed to be varied over a wide range. The slower tribom-eter measures frictional forces in the speed range of 0.1-20 mm/s and the faster tribometer from 25 to 400 mm/s. A fixed pin that holds the steel ball (diameter = 6 mm, DIN 5401-20 G20, Hydrel AG, Romanshom, Switzerland) was... 

An early description of friction, Amontons law, which describes the macroscopic friction between rough surfaces, gives a relation between friction force l Mction and applied loadL (eq. 6). 

Early work on the macroscopic friction of SAM films was carried out by DePahna and others using a pin-on-disk type friction tester, and they observed a low friction coefficient [1,13] however, their frictional property containing a wear resistance was demonstrated under a limited condition. For example, the self-assembled octadecyltrichlorosilane monolayer revealed a good wear resistance at low load [14], but failed to lubricate at a normal load of 1.0 N due to wear [15]. To overcome the poor wear resistance of thin films, a dual-layer film was prepared by a combination of aminoalkylsilane and stearic acid, which improved the wear resistance of the SAM film [16]. In this work, the authors also proposed a defect-free double layer on a flat substrate by chemical adsorption of alkylsilane compound. 

We have stndied the macroscopic frictional properties of high-density polymer brushes prepared by surface-initialed ATRP of methyl methacrylate (MM A) [31] and hydrophilic methacrylates [32, 33] from silicon substrates. Friction tests were carried out using a stainless steel or glass ball as the sliding probe under a normal load of 100 MPa from the viewpoint of practical engineering applications. This chapter reviews the macroscopic frictional properties of polymer brushes under a high normal load, the dependence of solvent qnaUty, the effect of humidity on hydrophilic brnsh, and wear resistance, and we compare these with alkylsilane monolayers. 

Macroscopic friction tests on the NTS monolayer and bilayer in air were performed by sliding a stainless steel ball along a distance of 20 mm at a sliding velocity of 90 mm/min under a load of 0.2 N. Both thin films showed similar dynamic friction coefficients of 0.080-0.083, while a large difference was observed in wear resistance. The friction coefficient of the NTS monolayer increased after the 150 friction cycles, as shown in fig. 5.8a. On the other hand, the bilayer demonstrated a stable friction coefficient until 2200 friction cycles (fig. 5.8b). A wear track was clearly observed on the surface of the monolayer after 300 friction cycles (fig. 5.8c), while only a slight wear track was present on the surface of the bilayer, even after 1000 slidings, as shown in fig. 5.8d. Generally, a wear track should be formed by direct contact of the stainless steel probe with the silicon substrate at the point where... 

Macroscopic friction tests on PMMA brushes were carried out by sliding a stainless steel ball on the substrates at a rate of 90 mm/min in air under the normal load of 0.49 N at room temperature. In the case of a nonmodifled silicon wafer under a normal... 

The field of nanotribology evolves around the attempts to understand the relationship between macroscopic frictional forces and the microscopic properties of the embedded system. Recent revival of interest in friction [/, 2, 3, 4, 5, 6, 7, 8] has unraveled a broad range of phenomena and new behaviors which help shed light on some fundamental concepts which are already considered textbook material. These include the static and kinetic friction forces, transition to sliding, thinning, and memory effects, which have been widely discussed but whose microscopic origins are still lacking. 

Macroscopic friction measurements between untreated SiOa balls and ODTS treated Si(lOO) flats show that the friction and durability of these surfaces are directly related to the quantity of ODTS chemisorbed on the surface. Friction decreases and the number of sliding cycles required for the friction coefficient to exceed 0.2 increases as the coverage of ODTS increases. 

Yet another theory has emerged recently from the field of nanotribology. Surface force apparatus studies, combined with molecular dynamics simulations, of simplified model systems, such as molecularly flat mica separated by a few molecules thick lubricant layers, have identified a solid-melt transition as the cause for stick-slip motion to occur for such confined liquids (18, 106, 111-113, 144-149). A similarly confined liquid can be found in macroscopic friction systems in the boundary lubrication regime wherein thin lubricant layers are trapped between surface asperities in very close proximity. 

---