Effect of temperature variation on repeatability positioning of a robot when assembling parts with cylindrical surfaces

Abstract

In this paper, the effect of the errors induced by temperature changes on the repeatability positioning error of an industrial robot is analysed. It has been shown that after the stabilization of the thermal conditions, these errors can be identified with the systematic errors. It has also been shown that if the ambient temperature cannot be sufficiently stabilized, the temperature errors can be described using a normal or uniform probability distribution. Depending on the choice of a point in the robot’s workspace and temperature fluctuations, these errors can comprise a small share of the total error of the robot. Then the total repeatability positioning error can be approximated with sufficient accuracy by a normal probability distribution or it can comprise the dominant component of this error. In this case, the total error can be approximated using a flat normal distribution. It has been shown that, depending on the choice of location in the workspace in which the assembly operation is carried out, it is possible to obtain both different probabilities of assembling the parts correctly and a different effect of errors caused by slight temperature changes on the value of those probabilities. The results found indicate the potential possibility of increasing the reliability of the process by proposing the selection of the location in the robot workspace which has the lowest sensitivity to errors ascribed to changes in temperature