Normal Vibrations

A reduction in friction due to normal contact motions during sliding has been observed by many researchers. In the first works related to this subject, Tolstoi et al. (1973) modeled the contact region between two surfaces as a non-linear spring without damping, using an

empirical stiffness relation which are based on the experiment data given by static loading.

When the normal vibration is applied in the contact region, the average compression of the non-linear spring is reduced. The reduced average compression can be calculated by his empirical stiffness relation from energy considerations. Then the effective value of normal load can be calculated by substituting the reduced average distance into the empirical stiffness relation. Finally, the friction reduction by normal vibration was calculated by the reduction of the effective normal load. Both the model and measurement showed that the friction reduction due to normal vibration could be as large as 30% for various steel surfaces. No attempt was made to analyze the system dynamics.

In the case of the kinetic friction between plastics and ice or snow, Lehtovaara (1987) presented experimental results. In his apparatus, vibrations were induced in the test specimen by an exciter while the test specimen was sliding on smooth ice. The results showed that the frictional force reduction was greatest at the first natural frequency of the test specimen where the vibration amplitude was large compared with the amplitudes of the antiresonance frequencies. At temperature below –1 ºC, where dry friction is dominant, the kinetic friction was reduced even when the acceleration of the vibrating body was much lower than the acceleration of gravity (no loss of contact). The reduction in kinetic friction is almost linearly dependent on the normal acceleration. At air temperatures above zero, where wet friction (viscous shear of water) dominates, vibration has no effect on friction.

Tworzydlo and Becker (1991) investigated the influence of forced vibration on the static coefficient of friction. They applied a model of frictional interface, assuming the existence of non-linear normal compliance of the interface, to the transient analysis of vibration of the system. Their analysis revealed that in the presence of normal vibrations the tangential motion of the slider consisted of microscopic sticks and slips, which in the macroscopic scale are perceived as a “creep”-type motion. The experimental results showed that the rate of

decrease of the friction was very sensitive to the amount of damping applied on the interface.

The maximum reduction of the static friction for steels is about 84% for clean surfaces (lower damping) and is about 30% for contaminated surfaces (higher damping). The presence of interface damping weakens the effect of friction reduction, especially in the vicinity of a resonance zone.

Hess and Soom (1991a, 1991b, 1993) analyzed resonant non-linear normal vibrations as well as the associated instantaneous contact area and friction force under harmonic loads applied to both Hertzian contacts and rough planar contacts using a non-linear mass-spring-damper model. A decrease in the average contact deflection under dynamic loading was predicted in each case. This resulted in an 11% reduction in the average friction force for Hertzian contacts when the normal vibration amplitude was just below that required to produce momentary loss of contact. However, the average contact area and friction force for the rough planar contacts were hardly affected. Further, the normal vibrations and friction at Hertzian contact under random excitation were analyzed. Both external excitation and internal excitation that arose from surface roughness were considered. It was found that for a 5% probability of contact loss, a reduction in the mean friction force of approximately 9% is expected for both cases. The reduction in average friction arose due to the non-linear relationship between the normal contact load and the area of contact, under the assumption that the instantaneous friction force was proportional to the instantaneous area of contact.

Average friction measurements taken during continuous sliding were in agreement with the analysis. However, their analysis was restricted to the conditions without loss of contact, and the tangential compliance was not considered.

To analyze the mechanism of friction drive with ultrasonic wave, Adachi et al. (1996) developed an apparatus that can measure the friction force at the interface between a rotational disk and an oscillatory pin induced by ultrasonic wave. The experimental results

showed that the friction force decreased with the decrease of the rotational speed of disk and the increase of the amplitude of pin motion (the pin-disk contact is broken for part of the vibration cycle). They introduced a simple relationship between tangential coefficient and micro-displacement at the contact region to explain the friction reduction phenomenon.

The friction force microscope (FFM) has opened a way for the study of the friction in micro-/nanoscopic mechanical contacts. Dinelli et al. (1997) studied the dynamic friction dependence on out-of-plane ultrasonic vibration, using friction force microscopy and a scanning probe technique, the ultrasonic force microscope (UFM), which can probe the dynamics of the tip-sample elastic contact as a submicrosecond scale. The results showed that friction fell sharply when the tip-surface contact broke for part of the out-of-plane vibration cycle. Moreover, the friction force reduced well before such a break, and this reduction does not depend on the normal load. They suggested that the contact was solid-liquid-solid. However, the mechanism of the friction reduction was not studied in detail.

Similarly, Hesjedal and Behme (2000, 2002) experimentally studied the friction reduction phenomena in microscopic mechanical contacts using a scanning force microscope in the lateral force mode (LFM) and a scanning acoustic force microscope (SAFM). The data suggested that the lateral oscillation component has no influence on the reduction of friction. They concluded that friction reduction effect is only due to the vertical oscillation component that leads to an effective shift of the cantilever away from the surface.

The general conclusion of these researches is that average friction falls sharply when the contact broke for part of the out-of-plane vibration cycle and the presence of interface damping weakens the effect of friction reduction. Analytical models that assumed the non-linear normal compliance of the interface showed that without loss of contact the vibration has little effect on the average friction. However, these analytical models were

restricted to the conditions without lost of contact, and the tangential compliance was not considered.