Date:
Monday, March 11, 2019
Time:
2:30 p.m. to 5:30 p.m.
Location:
Henry B. Gonzalez Convention Center, Lila Cockrell Theater
Sponsored by:
TMS Additive Manufacturing Committee
Organizer:
Ryan Dehoff, Oak Ridge National Laboratory
Tying together the seven additive manufacturing symposia held at TMS2019, this special keynote session features five invited speakers discussing topics of broad interest to the additive manufacturing community.
Featured Speakers:
Tresa Pollock, University of California, Santa Barbara
Presentation Title: "Solidification of Superalloys: From Single Crystals to Additive Manufacturing"
About the Presentation
Careful control of the solidification process to achieve desired material properties has long been of critical importance to the production of a wide variety of superalloy components, ranging from large-scale ingots to highly tailored single crystals. To fully leverage emerging additive manufacturing approaches for superalloys, a full understanding of the solidification processes will be required at the melt pool scale, including the unique features that arise during remelting in the characteristic layer-by-layer approach. Three-dimensional characterization of melt pools using the TriBeam tomography approach will be reviewed. Solidification features unique to the additive process have been revealed. The implications for properties and for design of future alloys will be discussed.
Hamish Fraser, The Ohio State University
Presentation Title: "Optimizing the Performance of Additively Manufactured Ti Alloy Components"
About the Presentation
There are a number of defects associated with the additive manufacturing (AM) of Ti alloys. These include the formation of coarse columnar microstructures, generally parallel to the growth/deposition direction in builds produced by blown powder, the presence of porosity, and residual stresses. This talk focuses on the first of these, coarse microstructures. Emphasis has been on the use of alloying to effect an equiaxed microstructure through modification of the solidification mechanism, by inducing a columnar to equiaxed transition (CET). The experiments have involved using a combinatorial approach, where a LENS™ (blown powder) AM device with two powder hoppers has been used to produce variations in the minor alloying elements of interest, such that modifications to microstructure can be directly related to changes in composition. The mechanical properties of new Ti alloys, with alloying additions that result in CETs, have been assessed and will be reported.
Gregory Olson, Northwestern University & QuesTek Innovations LLC
Presentation Title: "Printable Alloys by Design"
About the Presentation
Thirty years of commercial development by a network of small businesses has delivered a mature technology of computational materials design and accelerated qualification grounded in a system of fundamental CALPHAD databases now known as the Materials Genome. A major focus of current application is the rapid development of the new alloys enabling the much-desired technology of additive manufacturing, with adaptation of the AIM methodology to accelerate qualification of printed components. Processability constraints address deposition hot-tearing resistance and quench-suppressible precipitation hardening. Designed steels and titanium alloys exploit reversion of nonthermoelastic martensitic transformation for recrystallization to equiaxed grain structures. ICME coupling of deposition process simulations to PrecipiCalc simulations allows predictive control of novel submicron oxide dispersions providing grain refinement and reduced fatigue nucleation potency. Exploitation of rapid solidification has enabled high-entropy-stabilized multicomponent non-Sc L12 dispersions in novel high-temperature Al alloys.
Elizabeth Holm, Carnegie Mellon University
Presentation Title: "Opportunities in Machine Learning for Additive Manufacturing"
About the Presentation
Capitalizing on advances in data science, machine learning (ML) is revolutionizing technologies from self-driving vehicles to medical imaging. With its ability to monitor a rich data stream of process variables, it is no surprise that additive manufacturing (AM) offers promising opportunities for ML applications. This talk will discuss ML approaches throughout the AM workflow. For example, computer vision can be used with ML to characterize feedstock materials for correlation to build outcomes. During the build, a variety of visual, audible, and electronic data can inform build quality and process control. Finally, especially for low volume and custom parts, ML can support qualification by relating part characteristics to properties. Using examples from each of these areas, we will compare ML methods including deep learning and discuss issues such as interpretability, data size and quality, and model supervision.
Sudarsanam Babu, The University of Tennessee, Knoxville
Presentation Title: "Solidification and Solid-state Transformations during Metal Additive Manufacturing under Thermo-mechanical-chemical Transients"
About the Presentation
Additive manufacturing (AM) of metal provides the unique ability to produce components with complex geometries that can lead to a combination of both mechanical and functional performances that cannot be achieved by traditional manufacturing. Currently, AM technologies rely on either melting and solidification or solid-state joining of material feedstocks. The interactions between boundary conditions imposed by these processes lead to wide variations of thermal-mechanical-chemical signatures, which in turn affects the reproducibility and qualification of parts made by AM. Interestingly, this challenge is quite similar to the welding and joining of metals experienced by the materials and manufacturing community. In this talk, validity of extending the welding metallurgy principles to AM will be discussed with examples to demonstrate engineering solutions for minimization of defects and cracking, control of site-specific crystallographic texture and microstructure, and development of hybrid materials for extreme environments. Finally, some of the unresolved scientific questions will also be outlined.