Figure 1

(a) A schematic view of the experimental system. (top) Transparent slider and base blocks are loaded through a soft 40-spring array. Displacement is measured at both the leading and trailing edges (where $F_S$ is applied) of the slider. (bottom) A laser sheet illuminates the interface though the base. Both the transmitted and reflected beams are imaged onto a fast camera. (b) The actual area of contact, $A$ (in % of nominal contact area), vs $F_N$ for 4 successive loading or unloading cycles. The samples were not separated between cycles and a minimum load of 270 N was retained. (inset) $A$, as a function of time at the lowest values of $F_N$, shows negligible contact area accumulation between consecutive cycles. (c) Static friction coefficient, ${\mu }_s$ directly measured upon loading (▪) and unloading (●) over an $F_N$ cycle, as in (b). The dashed line indicates the difference in contact area between loading and unloading, $\delta A$, as derived from (b). In (b) and (c) the slider and base were of PMMA, with a nominal contact area of $150\times 6\mathrm{mm}$.Reuse & Permissions

Figure 2

$A$ (in % of nominal contact area) as a function of $F_N$ (top) and loading histories (bottom) in successive loading or unloading cycles for a PMMA/PMMA interface. (a) $F_N$ is cycled from 270 N to decreasing maximal loads. Between cycles 1 mm slip was generated. inset of (a)—When $F_N$ is normalized by $\Delta F_N$, all curves collapse to a single curve. (b) The maximal $F_N$ is fixed and successively lower minimal loads are applied. No slip was imposed between cycles and long-term memory is retained, as seen by the reloading curve axes orientation towards the maximal point achieved in the initial loading cycle. Here, [inset of (b)] data collapse upon normalization by $\Delta F_N$ occurs only after rotation to offset the axis orientation induced by long-term memory. Slider and base were of PMMA with a $30\times 6\mathrm{mm}$ nominal contact area.Reuse & Permissions

Figure 3

$A$ as a function of $F_N$ is independent of humidity, material contrast, or viscoelasticity. (a) $A-F_N$ curves for a PMMA/PMMA interface at ambient humidities of 50% (line and squares) and 2% in dry nitrogen (circles) and ($\nabla$) a PMMA/BK7 interface (diamonds). Similar results where obtained for a silanated BK7 base. (b) $A-F_N$ curves for a PMMA/PMMA interface where loading rates were altered by over 2 orders of magnitude. Besides upward shifts due to aging, no apparent rate effects are observed.Reuse & Permissions

Figure 4

Poisson effect and contact renewal. (a) Small amplitude ($\sim 400\mathrm{N}$) oscillatory loading cycles, imposed on a full loading cycle yield negligible amounts of hysteresis for $F_N>2500\mathrm{N}$, when the base material is PMMA. The inset shows that this region of negligible hysteresis is extended to a $\sim 1200\mathrm{N}$ oscillation amplitude when using a base of float BK7 glass and roughened ($\sim 15\mu \mathrm{m}$ rms) PMMA slider. (b) Total slider expansion, $\Delta L$, as a function of $F_N$ during one loading or unloading cycle.Reuse & Permissions

Figure 5

Aging and de-aging of the system. (a) Measurement of slider expansion, $\Delta L$, for two consecutive loading cycles where in the second cycle (circles) the load was held at 4000 N upon loading, and 1000 N upon unloading. Increase at 4000 N (upper inset) and decrease at 1000 N (lower inset) of slider length with time. The decrease is observed subsequent to a rapid reduction of $F_N$ and continues until arriving at the steady-state (cross) curve. (b) Change in $A$ (% of nominal contact area) as a function of $\Delta L$ where the normal load was held at $F_N=2000\mathrm{N}$ during loading (light line) and at $F_N=1500\mathrm{N}$ during unloading (dark line). The nearly linear dependence demonstrates both the strong coupling between $\Delta L$ and $A$ and the equivalence of the increase in $\Delta L$ with aging and decrease with de-aging.Reuse & Permissions