Publications Details

SOAR: Smartweld optimization and analysis routines

Eisler, G.R.; Fuerschbach, P.W.

A suite of MATLAB-based code modules has been developed to provide optimal weld schedules, regulating weld process parameters for C0[sub 2] and pulse ND:YAG laser welding methods, and arc welding in support of the Smartweld manufacturing initiative. The optimization methodology consists of mixed genetic and gradient-based algorithms to query semi-empirical, nonlinear algebraic models. The optimization output provides heat-input-efficient welds for user-specified weld dimensions. User querying of all weld models is available to examine sub-optimal schedules. In addition, a heat conduction equation solver for 2-D heat flow is available to provide the user with an additional check of weld thermal effects. The inclusion of thermodynamic properties allows the extension of the empirical models to include materials other than those tested. All solution methods are provided with graphical user interfaces and display pertinent results in two and three-dimensional form. The code architecture provides an extensible framework to add an arbitrary number of modules.