| Abstract Scope |
Anisotropic mechanical properties, reduction in ductility, and endurance limit of the printed material with Ti6Al4V in laser powder bed additive manufacturing result from forming undesired microstructure such as columnar grains. The columnar grains formation stems from a steep temperature gradient and high cooling rate in the melt pool, which is a characteristic of this manufacturing technique. One promising approach to control the microstructure is beam shaping or using synchronized multi-beam additive manufacturing to tailor the temperature field and subsequently alter the microstructure. Here, a machine learning (ML) algorithm was employed to predict the temperature field for various synchronized multi-beam configurations such as circular, linear (horizontal and vertical), circular, and square configurations. The ML approach is promising compared to other existing computational approaches considering the wider processing window of these synchronized beams, i.e., scanning speed, laser power, number of lasers, laser configuration, and internal spacing among lasers inside a specific configuration.
|