Abstract:
Vibrational analysis performed on a common design of pin-on-disk tribometer
showed that high frequency oscillatory friction forces were generated between the pin
and the disk during a test. These resulted in the tribometer exhibiting friction-induced
vibration, during which it oscillated at a natural frequency. This oscillation had two
effects; firstly the friction data generated between the pin and the disk was modified
such that the recorded friction data was not an accurate representation of 'real' data.
The second effect was that the tribometer modified the sliding conditions being tested,
leading to a disparity in the recorded friction data, when compared to data obtained
with the sliding conditions held at a constant value. For meaningful friction data to be
recorded the influence of the tribometer must be kept to a minimum.
A dynamic model of a tribometer was constructed to investigate the coupling between
oscillatory friction data and the tribometer's dynamics, and the influence it had on the
accuracy of the recorded friction data. The model quantified the influence that the
tribometer had on the recorded friction data and the deviation in sliding conditions.
Using the model a stiffer design of the tribometer was designed and constructed with
the aim of minimising the influence of the dynamic characteristics on the friction data.
Vibrational analysis of the modified stiffer tribometer set-up showed that it was still
exhibiting friction-induced vibration, and as such the dynamics of the tribometer still
had a significant impact on the accuracy of the friction data produced. There were
significant differences in the way the tribometer's dynamics influenced the friction
data for the two set-ups. Friction testing with the two set-ups showed that both set-
ups produced repeatable friction data, but with very different average friction
coefficients and variations in the friction data during a test.
