Towards a more robust automated MAG welding

Ultra-high-strength steels like 22MnB5, increasingly used in crash relevant body-in-white components, are very sensitive against too much heat input in industrial robot welding. Hence, in this work, the highly resolved visualisation of the dynamical transient temperature fields on the joint is investigated in order to obtain the real cooling rates of the melt pool as well as in the HAZ during the running welding procedure. For this purpose, a 2-colour HDRC pyro camera was used, enabling an absolute in-situ temperature measuring without the knowledge of the specific emission coefficients. A concurrently measuring of the overlapped weld joint via Laser triangulation enables a compensation of the actual gaps. In this work, an adaptive real-time control has been developed by means of adapting the welding process parameters to real gap and heat input conditions in the joint in order to preserve the maximal joint strength for automated robot welding under rough industrial production environments.