Data-Driven Modelling of Complex Current–Voltage Waveform Controlled Gas Metal Arc-Wire DED Processes

Gas metal arc-based wire arc additive manufacturing (GMA-WAAM) is an attractive process for producing geometrically simple and large-volume components. In this process, the arc energy and droplet transfer behaviour are primarily controlled by the current–voltage waveforms. Finite element analysis based on 3D transient gas metal arc heat source models has been extensively used to derive mean process parameters and predict the thermal fields during GMA-WAAM. These models often simplify a complex current–voltage waveform by averaging it over a defined time period and use heat source efficiency, heat source parameters, and heat transfer co-efficient as fitting parameters. This simplification leads to inaccuracy in the predictions of thermal fields. Therefore, a data-driven-based approach is proposed in this work to develop a physically-based instantaneous arc heat source model. This model is aimed to effectively describe a complex current–voltage waveform used for controlled dip short-circuiting transfer by adapting goldak’s double ellipsoidal arc heat source model. Instantaneous heat source parameters are derived from constant current waveform experiments corresponding to individual instances in a short-circuiting current–voltage waveform. Arc energies are calculated as a function of instantaneous heat source parameters while depositing 1.2 mm diameter Al-Si12 (ER4047) wires on a pure aluminium substrate. Thermal fields are measured using these instantaneous parameters and validated with experimental observations. Results show that this data-driven approach predicts the thermal fields with less than 1% relative error in the peak temperatures using the physically accurate heat source efficiency, heat source, and heat transfer parameters for a GMA-WAAM process. Graphical abstract: [Figure not available: see fulltext.]