Investigation for an Automated Avoidance of Ring Climbing in Radial-Axial Ring Rolling

Abstract

Abstract Radial-axial ring rolling is the most important production process for seamless rings. One of the most occurring process errors is the climbing of a ring in the radial rolling gap. With the conventional sensors of a radial-axial ring rolling machine, it has not yet been possible to detect a ring’s climbing in any way. For this purpose, an image processing system was installed at the radial-axial ring rolling machine of the LPS (Chair of Production Systems) which particularly records the side of a ring in the radial rolling gap. This paper proves the basic measuring capability of this intelligent image processing system in the rather difficult setting and presents different steps and experiments chosen to give reliable information on the system’s measurement accuracy. Furthermore the used algorithm to detect the climbing height of rings is presented and the degree of repeatability gets examined and evaluated. Once it is proven that the system delivers reliable results, this work uses different parameters based on the state of the art to affect the climbing of a ring. The intensity with which these parameters affect a ring’s climbing is tested in a fractional factorial design. This knowledge is used to design rolling strategies which avoid a climbing of the ring. Based on these strategies, this work introduces first experiments with static parameters during the entire rolling process. This paper shows the necessary steps and the ways these are realized to implement an image processing system to determine a ring’s climbing height. This finally serves to implement an automated system to avoid a ring’s climbing at an early stage in radial-axial ring rolling.