Automatic Weld Planning by Finite Element Simulation and Iterative Learning : An automatic off-line planning system was developed to predict the optimum control variables in gas metal arc welding

This paper presents a system for automatic planning of gas metal arc welding (GMAW) operations. The system automatically generates optimal dynamic trajectories of welding control variables (power input, current, and voltage; and travel speed) such that a desired weld quality is obtained despite known changes in process conditions, for example, geometric changes. The system is based upon a finite element simulation of the GMAW process. The finite element model developed for the purpose consists of two parts: a heat conduction model simulating the temperature distribution within the workpiece and a model simulating the weld pool surface shape. To improve the accuracy of the finite element model, a calibration method has been developed that enables precise simulation with varying welding control variables. The finite element simulation is coupled to an iterative learning controller, which exploits the capability of simulations to be repeated. The iterative learning controller identifies the best possible dynamic trajectories of the welding control variables in an iterative process. The results of open loop execution of automatically planned GMA welding on butt joints with constant and varying thermal regions are presented. A comparison with industrially applied data showed that a significant reduction of the heat input and workpiece distortion could be achieved while maintaining a satisfactory weld quality. The results demonstrate the superiority of process performance based on planning by iterative learning control compared to manual execution of welding task and compared to process planning based on traditional methods.

[1]  Hans Christian Elm Kjeldsen.  Sensor Based Welding Automation Modelling System , 2007 .

[2]  Mikael Ericsson.  Simulation of robot paths and heat effects in welding , 2006 .

[3]  Hans Holm,et al.  Welding Process Control Based on a Numerical Model and State-Space Techniques , 1999 .

[4]  J. Mackerle.  Finite element analysis and simulation of polymers?an addendum: a bibliography (1996?2002) , 2002 .

[5]  S. Baque,et al.  Learning control of cyclic production processes , 1997, 1997 IEEE 6th International Conference on Emerging Technologies and Factory Automation Proceedings, EFTA '97.

[6]  Carsten Bro,et al.  Application of Numerical Model for Off-line Programming of Posture Robot Welding , 2002 .

[8]  C. Doumanidis.  Thermal modeling and adaptive control of scan welding , 1998 .

[9]  J. Goldak,et al.  A new finite element model for welding heat sources , 1984 .

[10]  Kevin L. Moore,et al.  Iterative Learning Control: An Expository Overview , 1999 .

[11]  R. Cook,et al.  Concepts and Applications of Finite Element Analysis , 1974 .

[12]  Hans Holm,et al.  Method for Calibration of Full Penetration GMA I-Joint Welding Model to Single and Multiple Experiments , 2004 .

[13]  Xing Cai,et al.  A Flexible Architecture for Welding Simulators Used in Weld Planning , 2003 .

[14]  Hans Holm,et al.  A Reference Architecture for the Design of an Industrial Temperature Feedback Welding Control System , 2003 .

[15]  Hans Petter Langtangen,et al.  Computational Partial Differential Equations - Numerical Methods and Diffpack Programming , 1999, Lecture Notes in Computational Science and Engineering.