Dynamic Detection of Instability Defects in tube Rotary Draw Bending

Tube rotary draw bending is commonly used to manufacture complex shaped elements in a wide range of geometries and applications. When thin-walled tubes are formed with small bending radii, the components may suffer of wrinkling at the intrados due to the critical process parameters. Off-line optimization approaches, which aim at predicting the correct process parameters and tool geometries, often require accurate data of the workpieces properties or the application of numerical simulation analyses, that may fail due to the scatter of such properties along the production batches. Therefore, the availability of on-line approaches to provide an adaptive response of the machine is crucial for the efficiency of highly automated production lines.The paper presents the investigations carried out to assess the capability of dynamic analysis techniques to detect the instability defects that may affect the tubes in critical operating conditions. A new experimental approach, based on the analysis of dynamic data from the machine, is presented, as well as the evaluation of its performances when applied to the industrial process. The proposed approach appears promising for the detection of wrinkling, once they appear on the tube, and capable to give accurate feedback for the fast adjustment of the process parameters.