Optimization and Modeling of Weld Bead Area in Electron Beam Welding (EBW) of Aerospace Alloy (Inconel 825)
Keywords:
Electron beam welding, neuro-regression modeling, optimization, weld bead areaAbstract
This research examines the ideal weld area on Inconel 825, a common aerospace alloy produced using
the electron beam welding method. To investigate the weld bead area (WA) of the weldments, the
following process factors are taken into consideration: welding speed (S), beam current (I), accelerating
voltage (V), and beam oscillation (O). As a fundamental introduction to neuro regression modeling using
artificial neural network (ANN) philosophy, an instructional research on various non-linear neural
regression analyses has been conducted. In order to assess the accuracy of their predictions, 14 predictive
functional structures were used to represent the experimental prediction using basic regression modal
types. To train the program using the selected model R2, test it, confirm that R2 is employed accurately,
and determine if the results fall within the engineering bounds. Three distinct situations have been used
in optimization algorithms. Out of the fourteen models, only one produced findings that were realistic. It
has been observed that the models used for design variables, the sorts of scenarios, and the choice of
constraints all have an impact on the optimization outcomes.