A dynamic group of researchers at the University of California, Irvine (UCI) has been hard at work on two groundbreaking initiatives that will forever impact material manufacturing. The first research project focuses on advancing a new and evolving additive-manufacturing (AM) process: wire-arc additive manufacturing (WAAM). The second project is identifying solutions for resolving process-induced porosity of casting aluminum alloys through finite-element modeling.
Both projects are solving the need for more cost-effective solutions to real-world problems.
The work is being performed under the direction of Diran Apelian, director of the Advanced Casting Research Center (ACRC). ACRC is one of the largest industry-university centers in North America dedicated to collaborative research in metal processing and manufacturing. Its focus is metal casting and digital manufacturing, and its work has far-reaching impact on the industrial sector. It was first established in the 1980s and has a long-standing track record for carrying out fundamental research that is impactful and of utility to the business sector. It is located within the Samueli School of Engineering at UCI.
Read on for more detail on these two exciting, new projects.
Wire-Arc Additive Manufacturing
WAAM is a variation of direct energy deposition (DED) technology and uses an arc-welding process to 3D print metal parts. It combines automated metal inert-gas (MIG) welding, or laser hot-wire welding, with direct-deposition 3D printing. It is a cost-effective AM process and makes use of bead manipula-tion to obtain the desired component shape. The starting material – the feedstock in WAAM – is wire, so it precludes the need to produce powders via atomization. It also addresses the storage issues that one normally needs to attend to with powder processes. It is a near-net-shape manufacturing process that lends itself to making large components.
This bridge is a good example of a large part manufactured by WAAM technology. The bridge was made by MX3D, all via WAAM, in record time. The stainless steel bridge is fully functional and crosses one of the oldest and most famous canals in the center of Amsterdam, the Oudezijds Achterburgwal. (Image courtesy of MX3D)
The ACRC team working on this project includes Carl Söderhjelm, associate director of ACRC, Apelian and Lorenzo Valdevit, director of the Institute for Design and Manufacturing Innovation (IDMI). The team is focused on taking WAAM to the next level by addressing residual stress and ther-mal-management issues.
Finite-Element Modeling of Casting Materials with Tomography-Infused Porosity
The other project being worked on by ACRC researchers is identifying solutions for resolving the ex-istence of process-induced porosity in the casting of aluminum alloys. These pores considerably impair alloys’ properties in high-performance applications. To address this issue, ACRC is focusing on using finite-element models and tomography reconstruction techniques to quantify the impacts of pores on cast components. Its research is also helping industry have a better understanding of pores with dif-ferent characteristics.
Since process-induced pores are generally small in size and complex in morphology, integrating their actual geometry into a traditional finite-element model often leads to a large number of small-sized and ill-shaped elements. Solving such a model is computationally challenging because it requires not only a significant amount of computer memory storage but also prohibitively long simulation time. To resolve the computational challenges, ACRC’s research team aims to extend the existing finite-element method by developing a mechanistic reduced-order model and a machine-learning-based surrogate model.
Reduced-order models lessen computational costs while maintaining high accuracy and versatility by minimizing the number of unknown variables. With fewer variables in a system, the model can be executed with lower memory storage and shorter simulation time.
Machine-learning-based surrogate models further improve computational efficiency by conducting ex-pensive simulations in an offline step, which dramatically accelerates online calculations. In this process, a database of porosity-oriented material properties is generated via the offline simulation and is applicable to various cast components with distinct pore characteristics. Surrogate models coupled with microstructure characterization and reconstruction algorithms can mimic a more realistic pore representation, where the importance of different pore geometrical descriptors is identified via sensitivity analyses.
Three recent papers give further details of this work.
- Deng, S., Söderhjelm, C., Apelian, D. & Bostanabad, R. (2021), “Reduced-Order Multiscale Mod-eling of Plastic Deformations in 3D Cast Metallic Alloys with Spatially Varying Microstructures,” arXiv preprint arXiv:2108.03742
- Deng, S., Söderhjelm, C., Apelian, D. & Suresh, K. (2021), “Second Order Defeaturing Error Esti-mator of Porosity on Structural Elastic Performance in Manufactured Metallic Components,” arXiv preprint arXiv:2108.03740
- Deng, S., Söderhjelm, C., Apelian, D., & Suresh, K. (2021), “Estimation of elastic behaviors of metal components containing process induced porosity,” Computers & Structures, 254, 106558
Diran Apelian is founding director of ACRC. He was recently appointed Distinguished Professor at UCI’s Department of Materials Science and Engineering after retiring as Alcoa-Howmet Professor of Mechanical Engineering at WPI. Apelian is widely recognized for his innovative work in metal pro-cessing and for his leadership as a researcher and educator. He is a member of the National Academy of Engineering, National Academy of Inventors, European Academy of Sciences and the Armenian Academy of Sciences. He has received numerous honors and awards; has over 700 publications; and serves on several technical, corporate and editorial boards.
All graphics provided by the author, except where noted.