Frictional aging, de-aging, and re-aging in a monolayer-coated micromachined interface

Abstract

Measurements on monolayer-coated polycrystalline silicon surfaces have shown that the static friction coefficient ${\ensuremath{\mu}}_{s}$ strongly depends on loading parameters including hold time and normal hold force by Corwin and de Boer [J. Microelectromech. Syst. 18, 250 (2009)]. In that work, ${\ensuremath{\mu}}_{s}$ was measured by keeping the tangential force constant and lowering normal force until motion occurred. Results indicated that ${\ensuremath{\mu}}_{s}$ also depends strongly on normal force ramp-down rate. Here, we postulate that if the normal load is lowered instantaneously, the time for the block to begin moving, the ``release time'' ${t}_{r}$, will be greater than the inertial response time, which is on the order of $5\text{ }\ensuremath{\mu}\text{s}$. We measure the release time and find that it spans nearly six decades from less than $100\text{ }\ensuremath{\mu}\text{s}$ to almost 50 s. Release time depends on the loading and unloading history through all three of the parameters varied: hold time, hold force, and release force. An empirical model incorporating all three of these parameters fits the release time data over the full range. Release time decreases after the contacting surfaces are held together at increasing hold force levels and this qualitatively explains a previous observation that static friction aging is suppressed with increasing normal force at a fixed tangential load in this interfacial system. We further quantitatively relate the previous ${\ensuremath{\mu}}_{s}$ loading dependence on all three parameters to the release time model established here by introducing a ``re-aging'' parameter. This work firmly establishes that release time is a more fundamental parameter than the static friction coefficient and is the origin of static friction coefficient dependencies in this micromachined interface.