Skip to Main content Skip to Navigation
Conference papers

Dry friction modeling in dynamic identification for robot manipulators: Theory and experiments

Abstract : Nowadays, many robotic applications require an accurate model to perform tasks where dynamics is significant. The friction model discussed in this paper aims at improving the existing rigid robot model. The losses in joint transmission originate in friction between moving parts in contact or between moving parts and the ambient fluid. Commonly, robotic identification models represent joint transmission friction force as a viscous friction force, depending on the velocity, added to a constant dry friction force. However, the tribology science field teaches that friction in general depends on load (reaction force normal to the contact surface). It is important to consider this dependence when variable loads are applied on the joint transmission (external payloads, inertial load and gravity forces). Since these mechanisms are lubricated, it is appropriate to refer to the Stribeck curve (rather than Coulomb). This curve describes the friction coefficient as being dependent on a parameter (Hershey) combining the velocity and the load. This paper proposes a new expression of the load-velocity friction model, in order to identify a serial n degrees of freedom (DOF) robot. The friction force of this new inverse dynamic identification model is a linear function of both inertial and external forces. An experimental validation on an industrial manipulator used as a force feedback telerobot in nuclear plant concludes this paper.
Document type :
Conference papers
Complete list of metadatas
Contributor : Bruno Savelli <>
Submitted on : Thursday, July 23, 2015 - 2:13:26 PM
Last modification on : Monday, February 10, 2020 - 6:12:33 PM




N. Kammerer, Philippe Garrec. Dry friction modeling in dynamic identification for robot manipulators: Theory and experiments. 2013 IEEE International Conference on Mechatronics, Feb 2013, Vicenza, Italy. pp.422-429, ⟨10.1109/ICMECH.2013.6518574⟩. ⟨cea-01179832⟩



Record views