AVS 60th International Symposium and Exhibition
    Tribology Focus Topic Monday Sessions
       Session TR+AS-MoM

Paper TR+AS-MoM11
Speed-Dependence of Atomic-Scale Friction

Monday, October 28, 2013, 11:40 am, Room 203 C

Session: Bridging Scales and Characterization
Presenter: Z. Ye, University of California Merced
Authors: A. Martini, University of California Merced
Z. Ye, University of California Merced
Y. Dong, Purdue University
P. Egberts, University of Pennsylvania
XZ. Liu, University of Pennsylvania
R.W. Carpick, University of Pennsylvania
Correspondent: Click to Email
Atomistic simulations and experimental atomic force microscopy measurements on a variety of different materials have shown that atomic-scale single asperity friction can be significantly affected by sliding speed. However, physical insights into how and why sliding speed affects friction are limited because the speeds accessible to most simulations are several orders of magnitude faster than those in the corresponding experiments. Typical simulations must be run at fast sliding speeds due to their necessarily short time scale, and accurate experimental nanoscale asperity friction measurements are limited to slow speeds because of difficulties in measuring high-speed forces with picoNewton resolution. Here we present friction results from molecular dynamics simulations where the sliding speeds are greatly reduced by using parallel replica dynamics. Parallel replica dynamics is an accelerated simulation technique that distributes simulation time across multiple processors and therefore adequately samples the various possible state-to-state pathways accessible to the system, as would a standard, single-processor simulation run for a very long time. This technique, accompanied by experiments where data is obtained using a novel high-speed data acquisition method, enables measurements and simulations to be quantitatively compared within the same physical regime; specifically, at the same sliding speed. Furthermore, the materials, load, contact size and orientation, system compliance, and temperature are identical within experimental uncertainty so as to minimize differences between experiments and simulations, allowing robust comparisons and interpretations. These coordinated studies enable us to understand the dependence of atomic-scale friction on sliding speed, and to determine the limits of validity of the Tomlinson-Prandtl model, a reduced-order model widely-used to describe atomic-scale sliding.