Mathematical analyses and experimental verification of elimination of measurement error in UMT-2 rotating friction system

Abstract

This paper aims to investigate and eliminate the measurement errors of coefficient of friction (COF) in the rotating friction system of UMT-2 and the corresponding solutions are proposed. Three main factors leading to measurement errors were discussed: the displacement of the friction center (Δx), the tilt of the sample holder, and the tilt angle of the friction disc (angle α). Analysis based on Abbé error/offset shows that a smaller displacement Δx leads to a smaller measurement error, and ensuring that Δx is equal to the spacing between the axis of rotation and the trajectory of the friction center can reduce the measurement error. In addition, the verification tests show that the measurement error was greater than 5 % when the friction force was greater than 0.87 N for the selected sensor and the measurement error due to the tilt of sample holder is proportional to the friction force. Therefore, this verification method is recommended to determine the applicable measurement range of selected sensors, avoiding friction force exceeding the reliable measurement range of selected sensor. Pin-on-disc friction tests carried out in clockwise and counterclockwise were conducted under load of 9 N and speed of 7.85 mm/s to examine the effectiveness of the leveling device proposed in this work. The results show that the tilt angle α of the disc was reduced from 0.135° to 0.005°, a reduction of 96.3 % with the help of the leveling device. The relative error due to the angle α was as low as 0.8 % at a COF of 0.001 magnitude, which fundamentally improved the measurement accuracy of the friction test at ultra-low COF. In addition, future research is proposed on how to eliminate the measurement error of friction, which can help researchers and instrument designers to improve the measurement accuracy.