Video analysis-based estimation of bearing friction factors

Abstract

Friction is an inherent factor in any real physical system, affecting, as it is natural, the dynamics of the entire system. In particular, the computation of friction losses are of special interest because friction factors can not be easily predicted theoretically in an accurate manner. In oscillatory systems friction is studied by two approaches, viscous and dry damping. Viscous damping is usually treated in physics courses due to its relatively easy-to-find analytical solution, whereas dry damping requires piece-wise solutions. This paper studies friction factors affecting a customised experimental platform for electric bikes. We seek to estimate the friction torque and then the coefficient of friction, where the damping of its amplitude will be evaluated. We study the oscillations of two different rolling systems. The first rolling system is a rowlock, and the second is a bicycle wheel axis. In both systems amplitude damping was measured and analysed. The experimental data for each rolling system was obtained by visual information, recording oscillations using a smartphone camera and then analysing it using Tracker, a freeware for video analysis. Results show that both systems behave very close to a dry damped oscillation. The model proposed matches properly the experimental data, and a value for the friction torque and coefficient of friction in both systems was estimated.