Résumé:
The imperfections at the origin of the lack of repeatability are of random nature, caused by the adjustments of the servo-control and especially by non-geometric phenomena such as the play and friction existing at the joints level. The occurring friction constitutes one of the main causes of performance loss. The friction modeling isn’t an easy task, and therefore, its correction in advance remains difficult. The adopted method consisted of employing mechanical approaches in order to theoretically identify the contribution of friction on the pose repeatability. The tribological experimental study was carried out on the alternative tribometer and is based on the steel grade 42CD4 with a steel pin 100C6 used on the robot joints. Due to the variability of the friction values coefficient, the main asperities on isotropic surfaces were analyzed by considering their characteristics, namely the top altitude of each asperity, the distance between two consecutive views, and its radius. Each of these geometric characteristics is described using statistical analysis with probability density functions such as Gaussian laws. Finally, these developments resulted in a model that was applied by using measurements of three different types of microgeometry, ranging from quasi-periodic to quasi-random. The obtained results are interesting because they served to improve the analysis of repeatability by taking into account
