Laboratory experimentation remains the only practical method available for the accurate identification of friction coefficients for arbitrary material pairs. However, accurate and repeatable friction coefficient measurement remains challenging due to the dependence of friction coefficients on the material, surface, environment, and measuring equipment. The purpose of this paper is to critically examine the experimental uncertainty associated with the instrumentation used in dynamic friction coefficient measurements. In order to enable the confident use of experimental data, it is necessary to provide a quantitative, defensible statement regarding its reliability. The topic of data uncertainty has been addressed by researchers for several individual tribological measurement conditions [1-6]. In this paper, we focus on establishing a systematic framework for the evaluation of instrument-related uncertainty in tribological testing. Our example application is the determination of low dynamic coefficients of friction.

We consider friction coefficient measurements carried out using a traditional pin-on-disk tribometer, where a pin is pressed against a reciprocating counterface. This tribometer uses a pneumatic cylinder and a multi-channel load cell located directly above the pin to continuously monitor the contact force vector, which is then decomposed into friction force and normal force vectors. The uncertainty analysis follows the guidelines provided in references [7, 8]. Uncertainty contributors include load cell calibration, where both the applied load and voltage measurement uncertainties have been considered, and instrument geometry.

Adobe Acrobat File

Journal of Tribology
Vol. 127 (2005) pp. 673-678