0
TECHNICAL PAPERS

# Modeling the Tribochemical Aspects of Friction and Gradual Wear of Diamond-Like Carbon Films

[+] Author and Article Information
Feodor M. Borodich

School of Engineering, Cardiff University, P.O. Box 925, Queen’s Buildings, Cardiff CF24 3AT, UK

Department of Mechanical Engineering,  State University of New York, Stony Brook, NY 11794-2300

Leon M. Keer

Center for Surface Engineering and Tribology, Technological Institute, Northwestern University, Evanston, IL 60208-3109l-keer@northwestern.edu

J. Appl. Mech 74(1), 23-30 (Dec 02, 2005) (8 pages) doi:10.1115/1.2172267 History: Received July 28, 2004; Revised December 02, 2005

## Abstract

Problems in nanomechanics often need to combine mechanical approaches together with methods of physics and chemistry that are outside of the traditional mechanics scope. Recent experimental studies of dry sliding between two hydrogenated DLC (diamond-like carbon) coated counterparts in low oxygen environment showed that adsorbates have considerable influence on friction and the friction coefficient increases with the increasing of the time interval between contacts. The observed friction phenomena are assumed caused by a reaction between the adsorbate and carbon atoms of the coatings, and when the slider passes a point on the track, it removes mechanically some adsorbate from the surface. The mechanical action leads to reexposure of the surface to gases in the environment. This paper focuses on physical and tribochemical processes that occur in sliding contact between the DLC coated slider and the counterpart. We develop further our recently presented model of the process and assume that there is a transient short-life high temperature field at the vicinities of contacting protuberances that may cause various transformations of the surface. In particular, the $sp3$ phase of DLC films may transform to graphite-like $sp2$ carbon. Our model does not depend directly on the assumption that the adsorbate is oxygen. However, due to the prevalence of oxygen in atmospheric gas it is assumed that the adsorbate is oxygen in the model presented. We suppose that first an oxygen molecule becomes physically adsorbed to the surface and then due to rubbing the molecule dissociates into two chemically active oxygen atoms. This process leads to chemisorbtion between the carbon atoms of the coating and the “sticky” oxygen atoms. The latter atoms can interact with the counterpart. Our modeling established a direct connection between this kind of molecular friction and gradual wear. In particular, it is shown that the initial roughness of the DLC surface may have a considerable influence on the probability of breaking bonds during mechanical removal of adsorbate. Ab initio calculations of the bond dissociation energies between carbon atoms and carbon-oxygen atoms were performed using GAUSSIAN98 at the Møller-Plesset level of model chemistry. The bond dissociation energy found for the carbon-carbon bonds is $523kJ∕mol$, while for the carbon-oxygen bonds it is $1447kJ∕mol$. It is assumed that carbon wear particles will not be formed during gradual degradation since the coating carbon molecules are dissolved within the environment gases. The model helps to explain how microscopic processes, such as the breaking and forming of interatomic bonds, may affect macroscopic phenomena, such as friction and wear.

<>

## Figures

Figure 1

Schematic of the pin-on-disk tribometer numerical simulation performed along a linear track with nm∕2 number of measurement points, Δs spacing between points, track length L, and distances dl and dr from the left and right, respectively, of the k-th point to the track endpoints. The relative speed between the fixed pin and oscillating disk is νo. The pin has a spherical end cap.

Figure 2

Plot of the friction coefficient (μ) versus cycle number (i) for the entire experimental range. Data points (x) are the reciprocating pin-on-disk experimental data (from Heimberg (23)) for DLC on DLC contact. Solid lines represent data obtained using numerical simulation of the friction coefficient based on the integral of the Elovich Eq. 11, with B=4.1∙10−4s−1, μc=0.0065, α=0.6, p=0.14, tp=L∕νo, c=0.81, to=0, and θ(0)=0.001, and relative speed between pin and disk (νo) of 10μm∕s, 30μm∕s, 50μm∕s, 75μm∕s, 100μm∕s, 303μm∕s, and 513μm∕s.

Figure 3

Plot of the friction coefficient (μ) versus cycle number (i) for the slow speed ranges only. Data points (x) are the reciprocating pin-on-disk experimental data (from Heimberg (23)) for DLC on DLC contact. Data points (o) represent data obtained using the numerical simulation of the friction coefficient based on the integral of the Elovich Eq. 11, with B=4.1∙10−4s−1, μc=0.0065, α=0.6, p=0.14, tp=L∕νo, c=0.81, to=0, and θ(0)=0.001, and relative speed between pin and disk (νo) of 10μm∕s, 30μm∕s, 50μm∕s, 75μm∕s, 303μm∕s, and 513μm∕s.

Figure 4

Plot of the friction coefficient (μ) versus cycle number (i) for the time-delay tests. Data points (x) are the reciprocating pin-on-disk experimental data (from Heimberg (23)) for DLC on DLC contact. Data points (o) represent data obtained using the numerical simulation of the friction coefficient based on the integral of the Elovich equation 11, with B=2.3∙10−4s−1, μc=0.003, α=0.6, p=0.3, tp=L∕νo+td, c=0.81, and relative speed between pin and disk νo=1053μm∕s. Time delays at the track endpoints (td) were 5μs, 12μs, 45μs, 95μs, and 162μs.

## Discussions

Some tools below are only available to our subscribers or users with an online account.

### Related Content

Customize your page view by dragging and repositioning the boxes below.

Related Journal Articles
Related Proceedings Articles
Related eBook Content
Topic Collections