Regenerative Instability of Impact-cutting Material Removal in the Grinding Process Performed by a Flexible Robot Arm

• Published in: Procedia CIRP Vol. 14; pp. 406 - 411
• Main Authors: Rafieian, Farzad, Hazel, Bruce, Liu, Zhaoheng
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 2014
• Subjects: Impact cutting; Regenerative self-excitation; Robotic machining chatter; Vibro-impact; Regenerative self-excitation; Impact cutting; Vibro-impact; Robotic machining chatter
• ISSN: 2212-8271; 2212-8271
• DOI: 10.1016/j.procir.2014.03.099

Robotic machining is emerging as a viable alternative for some conventional machining tasks. Nevertheless, a challenge to cutting performance remains: vibrational instability is reported to be more severe with robotic cutting due to robot compliance. In this paper, the regenerative vibrational instability of a robotic grinding operation is studied. The objective is to find the limit of stable operation when material removal is performed by a compliant robot. The traditional approach to regenerative chatter analysis is revisited for a robotic machining process developed at Hydro-Québec's research institute for tasks on hydropower equipment. Stability lobes are established with respect to system gain (ratio of cutting rigidity over robot stiffness) versus repeat frequency (ratio of spindle frequency over the robot's first natural frequency). Robotic machining is found to be far upper-right of the first stability lobe minimum, where system gain and repeat frequency are more than one order of magnitude larger than for conventional machining. The cyclic impacting dynamics of material removal in the operation under study is invoked to investigate instability in this region. A SDOF dynamic model for the robot is excited by an impact-cutting force and the stability of the simulated response is verified while increasing the cutting depth. The limit of stable impact cutting is thus determined for the process at typical rotational speeds, from which a stability boundary is plotted versus the repeat frequency. The boundary is found to be very close to the limit predicted using the traditional approach. It is concluded that the large gain is typical for robotic machining. The impacting dynamics of material removal due to robot compliance must be considered to understand such large gain values, never occurring in conventional machining.