Comparison of Different Weld Bead Area Prediction Models for Overlay Deposition by Robot-Based Cold Metal Transfer Process

Arc welding, a predominant joining process, is increasingly being utilized in new applications, particularly metal 3D printing. Before 3D printing a part, its 3D CAD model is sliced into thin layers. Accurate identification of the geometry of individual layers and proper hatch spacing between these layers are equally important. Both factors significantly influence the quality of the final metal part printed by a robot-based Cold metal transfer wire arc additive manufacturing (CMT-WAAM) process. Experiments were conducted based on a two-factor central composite rotatable design. The bead area measured from the bead profile is compared with the area predicted by different bead models and, it was found that at lower wire feed speed to torch ratio (A < 8), the cosine model best predicts the weld bead profile area whereas at higher wire feed speed to torch ratio (A > 8), the circular model best predicts the weld profile area. Further, these models have been utilized to identify the optimum hatch spacing, so that proper area filling could be achieved during the CMT-W AAM process. Based on this prediction, multi-bead horizontal overlay experiments were conducted that were found free from any kind of lack of fusion voids and produced minimal surface waviness. Extended inter-layer dwell time reduces heat accumulation but increases surface waviness and lack of fusion voids due to arc deflection and misalignment, resulting in uneven bead heights.