UC Berkeley

Over the last decade, critical advancements have allowed metal additive manufacturing to graduate from a mere prototyping tool to a full-scale manufacturing process. Substitution of load-bearing components with their lighter, stronger, and less expensive additively manufactured counterparts is already becoming commonplace in the aerospace, automotive, and energy industries. With qualification underway, research efforts can now be directed towards bringing the more revolutionary capabilities of additive manufacturing to fruition. One such area is the ability to incorporate multiple alloys within a single build to fabricate complex geometry functionally graded structures, thus eliminating the need for extraneous welds and joints. As it turns out, local tailoring of composition is within the means of commercial laser deposition systems. However, its implementation is wholeheartedly a materials problem, driven by the difficulty of establishing strong metallurgical bonds between incompatible alloys.

This dissertation provides a comprehensive treatment of the subject, from initial consideration of its challenges, to the development of specific gradient components. Based on a review of advances in dissimilar metal welding and additive-based functional grading, a pragmatic approach is conceived of for developing individual gradients from start to finish. Incompatibilities between dissimilar alloys are generalized as brittle intermetallic formation, thermal property mismatch, and other metallurgical issues. Building upon the efforts of others, a series of strategies is then defined for exploiting the unique capabilities of additive manufacturing to overcome these. Among the most compelling is the use of multi-powder feeder laser deposition to carry out nonlinear paths in composition space that circumvent deleterious phases, as predicted by CALPHAD thermodynamic modeling. Also proposed is the elimination of sharp interfaces between alloys of mismatching properties using smooth compositional grading.

Upon implementing these strategies in the course of this thesis, it quickly became clear that each gradient system still requires some level of independent development to address alloy-specific concerns. In response, an approach is introduced that makes use of both modeling and experimental efforts to iteratively improve upon a given gradient design. Prototype gradients are fabricated with laser deposition and subjected to microstructural characterization to identify the critical issues that impact the gradient’s integrity. Thermodynamic modeling efforts are then directed towards identifying design modifications to overcome these issues. The process is repeated until successful, repeatable gradients are achieved.

The latter part of the thesis is dedicated to demonstrating this approach in two distinct experimental case studies, concerning the development of gradients from (1) titanium to stainless steel, and (2) maraging steel to austenitic stainless steel. In both cases, microstructural characterization proved to be instrumental in identifying the particular issues affecting the gradients, such that they could be addressed directly with modeling. For the case of joining titanium with steel, severe brittle intermetallic formation was identified as the primary cause of failure. Modeling was thus focused on predicting phase equilibria along the gradient, such to identify a nonlinear composition path free of intermetallic formation. The maraging to stainless steel gradients, by contrast, exhibited greater inherent chemical compatibility. Their development rather required close attention to the alloys’ impurity contents, hardening mechanisms, and microstructural features to determine a means of reducing undesirable secondary phases and defects. Critical analysis of the results for both gradients imparted a number of “lessons learned”, providing guidance for future development of this capability.