R. Crawford, G. E. Cook, Welding Automation Laboratory Vanderbilt University, Nashville, TN; A. M. Strauss, Vanderbilt University, Nashville, TN; D. A. Hartman, Beyond6 Sigma, Santa Fe, NM

Summary:

Vanderbilt researchers have sought to quantify the relationship between axial force (Fz), rotational speed, travel speed, and other process parameters for friction stir welding (FSW). Experimental data for Al 6061-T6 welded at high rotational (1500-5000 rpm) and travel speeds ranging from 7-60 ipm will be presented. The results will be discussed and will be compared to a two-dimensional fluid flow model implemented using the computational fluid mechanics package FLUENT. In addition, the results will be compared to results predicted by other researchers, and where applicable, correlated with existing three dimensional mechanical models. The forces and torques associated with friction stir welding will be discussed relative to enhancing the feasibility for robotic implementation of FSW.

The major requirements for a robot capable of FSW is two fold; simple positioning of the weld tool and providing the axial force (Fz) necessary for FSW, which have been shown to be quite substantial. Current FSW applications employ heavy duty machinery. The rigidity and precision of the machine tool allows the axial force to be maintained over the extent of the weld, however they are only capable of following twodimensional contours. With a robotic implementation capable of following three dimensional contours, the contact forces associated with FSW using current process parameters, leads to compliance of the manipulator. For completeness, future implementation of force feedback control is discussed as well as optimum parameters for robotic FSW.

In order to build an efficient control system for robotic FSW, a valid force prediction model needs to be established, and many of the basic physical mechanics of this process need to be understood. Current friction stir welding modeling uses either a computational fluids mechanics (CFD) or finite element method (FEM) approach [1-2] [3-4]. In order to determine which methodology is more appropriate, Vanderbilt researchers are developing models using both methods. A presentation of Vanderbilt researchers FSW process modeling efforts using the CFD and the FEM approach will be given.