A new approach to predicting precipitation sequences in rapidly solidifying metallic alloys

点击翻译按钮获取中文摘要

Numerous ways in which objects made of alloys are produced can roughly be divided into subtractive manufacturing, powder metallurgy, and additive manufacturing (3D printing).The first two praxes are mature; they undergo only gradual enhancements. The third one is different. Contents of peer-reviewed journal and magazines popularising science, engineering and technology attest that so much has been already achieved in additive manufacturing that 3D printing has transformed from an extravagantness to a useful production method. For high quality non-metallic parts, this conclusion reflects the reality, but mechanical properties of additively made metallic alloys are worse than of those made traditionally. Reasons are numerous; one of them stands apart in its importance. A mixture of metallic powders is placed on a surface of a partially fabricated object as a new layer, and this agglomeration is hit by a powerful laser beam. This impact causes an instantaneous melting of the layer followed by its solidification on a cold substrate's surface. This freezing is too fast to be treated within the Gulliver-Scheil paradigm; it is not too quick to result in an amorphous state. Rapid solidification accompanying repetitive steps in 3D printing is a distinctive process leading to supercooled melts, from which metastable phases may form, and therefore to byzantine microstructures whose crystallographic and compositional features can be ascertained using modern characterization tools. This information is vital for an intelligent design of post-printing properties-enhancing heat treatments.The ultimate goal, however, is not to describe post-printing microstructures, but to acquire an ability to predict them for Al, Ti and Cu alloys as well as for other multicomponent metallic systems with wide compositional modulations, which may eventually be tried in additive manufacturing. Such an ability cannot be based on the Edisonian approach; it must rest on a decent scientific foundation rather than on an artless empiricism.In 2010, the applicant conjectured that if compositions of thermodynamically possible phases are close to the composition of a remaining liquid, then the precipitation of these phases would be facilitated, because their nucleation within a supercooled melt will not require a slow long-range diffusion. That idea allowed to comprehend outcomes of rapid solidification of Al-Fe-Si melts, but its applicability to various multicomponent systems was never seriously tested.It is proposed to use an available melt spinner for rapid solidification of chemically and compositionally dissimilar metallic melts, and then to judge on the "compositional similarity" hypothesis feasibility. If its practicality is demonstrated, then the hypothesis will become a physically sound and computationally simple predictor of microstructures of metallic objects produced via additive manufacturing.