Experimental stiffness identification in the joints of a lightweight robot
Document typeBachelor thesis
Rights accessOpen Access
The estimation of stiffness parameters of a robot is of paramount importance for the precision of movement, in fact in the last year the application of robots is justified by many factors, one of these factors is the high precision of movement. The aim of this thesis is to develop a simple experimental methodology for the joint stiffness identification and its application to any robot and also to prove a methods already present in bibliography. The case study in this work is a lightweight robot, namely the UR5 manipulators. Usually these robots are slender and low-weight, and provided with flexible joints, these characteristics shall make a modal analysis of fundamental importance, because the robots that present a high flexibility may generate vibratory phenomenon, that would affect its performance. Therefore the stiffness analysis is the basis and preliminary of modal analysis. The stiffness analysis was made necessary by the lack of knowledge of some UR5 robot parameters as the stiffness, as this thesis’ case, or even damping. In the first three chapters there are, respectively, an introduction on the factors that influence the stiffness in the robotic structures; then are provided all the physical and mathematical knowledges, which underlie the study of stiffness, and finally is illustrated the state of art in stiffness evaluation methods. In the last three chapters there are, respectively, an overview of all components that have been used in laboratory for the evaluation of stiffness, subsequently the value of stiffness are presented respectively through the adopted methodology and the developed methodology, as final step an analysis of results was conducted to compare and to analyze the two methods. Some fundamental conclusions of this work are, respectively, that for conduct an analysis of stiffness is necessary a measurement system with high accuracy, subsequently both the adopted method and developed method may be applied in particular configurations, where the Jacobian matrix don’t change following the load application.