America Makes & AFRL Announce Winners of Metal 3D Printing Prediction Challenge

The National Additive Manufacturing Innovation Institute America Makes, managed and operated by the National Center for Defense Manufacturing and Machining (NCDMM) and headquartered in Youngstown, Ohio, was established in 2012 as the national accelerator for additive manufacturing in the US. Part of the Manufacturing USA network, this important organization is made up of member organizations from academia, government, industry, non-government agencies, and workforce and economic development resources.

Its mission is to increase the manufacturing competitiveness of the US on a global scale, by innovating the technology and increasing its rate of adoption around the country. This acceleration is achieved through projects, training, events and trips, apprenticeships, publishing useful technology standards and guides, and holding challenges and competitions. America Makes is just one of eight Manufacturing Innovation Institutes established and managed by the US Department of Defense (DoD) as public-private partnerships, and so works well with the military.

To that end, the organization recently teamed up with the Air Force Research Laboratory (AFRL) at Wright-Patterson Air Force Base, also located in Ohio, to help advance the AM industry by holding a modeling challenge series. More specifically, the AFRL’s Materials & Manufacturing Directorate, Structural Materials Division, Metals Branch (AFRL/RXCM), and America Makes held the challenge, which was launched in November of 2019 and closed submissions at the end of January 2020.

The challenge winners have finally been announced, and can divide the $235K prize between themselves. The AFRL AM Modeling Challenge Series was actually made up of four separate challenges:

1. Macro-scale Process-to-Structure Predictions
2. Micro-scale Process-to-Structure Predictions
3. Macro-scale Structure-to-Properties Predictions
4. Micro-scale Structure-to-Properties Predictions

The goal for this series was fairly distinct: participants were tasked with improving the accuracy of predicting the internal structure and resulting performance of models for metallic additive manufacturing, more specifically the nickel-chromium alloy INCONEL 625 (IN625). This material is often used in high-performance applications, such as for the aerospace, chemical, defense, oil, and power industries, that require excellent corrosion and temperature resistance, as well as thermal-fatigue properties. As the AFRL is in charge of the discovery, development and delivery efforts of weapons technologies for our country’s military forces, this material choice makes a lot of sense.

According to America Makes, the point of the challenge was to “account for material heterogeneity intelligently through geometry-sensitive property prediction at both the micro- and macro-structure level.”

“Models and simulations that can accurately account for this type of variability can be utilized by the AM design process and may be critical when designing complex parts with thin features,” the organization explained.

Participants were given the necessary calibration and validation data sets to create new models related to model prediction, and had to come up with algorithms and models that can fabricate dynamic material property prediction modules. For instance, Challenge 1 required teams to predict macroscale residual strain, while Challenge 2 asked for single layer predictions of deposit geometry and microstructural details at the single bead scale. Challenge 3 had teams predict aggregate stress-strain behavior using representative microstructure characteristics, and finally, Challenge 4 required a prediction of aggregate stress-strain behavior and strain in particular grains based on “an explicit microstructure representation.”

“Going into the AFRL AM Modeling Challenge Series, we knew that the outcomes would potentially lead to significantly improved predictability and accuracy of models and simulations, and the qualification of AM process and materials. The awardees of these four challenges certainly made solid contributions. They improved our understanding of the micro- and macro-structure level variability that was needed to advance the accuracy of modeling and simulation for AM metal,” stated America Makes Executive Director John Wilczynski. “We thank all those who participated and extend our congratulations to the awardees.”

The winning project teams of the AFRL AM Modeling Challenge Series are as follows:

• Challenge 1 — Dassault Systèmes Government Solutions Corp
• Challenge 2 — The Wing Kam Liu Group at Northwestern University
• Challenge 3 — QuesTek Innovations LLC
• Challenge 4 — University of Utah, Carnegie Mellon University, and Los Alamos National Laboratory

(Source: America Makes)