The objective of the AMNOW Metals Challenge was to evaluate the repeatability and predictability of the Laser Powder Bed Fusion process and therefore maximize the insights from the rich AM build and post-processing data produced during the AMNOW program. The challenge participants were asked to apply artificial intelligence and data analytics to the data set, correlating material and process input data to material property data. The associated AM digital thread data analyzed during the challenge included thirty-nine (39) 316L AM builds, across four (4) machine models, six (6) machines, three (3) suppliers, two (2) heat treatments, and over eight hundred (800) test coupons. The challenge was divided into two phases:
The focus of Phase 1 was to organize, characterize and analyze the full set of data, identifying trends, anomalies, insights, and correlations. The participants were asked to create algorithms to predict key results such as coupon test results and compare them to actual results. Finally, the participants were asked to identify opportunities for improvement for the final two builds at one of the three suppliers.
The focus of Phase 2 was the actual collaboration between Applied Optimization/Innovative3D Manufacturing, Waterloo University/Penn United Technologies, and The Roux Institute at Northeastern University/ATI to validate the predictions or improve the outcome of the final two (2) builds and test results. Additionally, Addiguru organized all of the 316L data into a prototype data schema and applied basic analyses for future researchers.

1) AMNOW Metal Challenge Background and Introduction (15 min)

2) University of Waterloo (15 min)

3) Applied Optimization (15 min)

4) The Roux Institute at Northeastern University (15 min)

5) Addiguru (15 min)

6) Questions (15 min)

Each Presentation will cover the following:
• Approach
• Tools
• Challenges
• Results
• Recommendations

Sep 29, 2022

2:00 pm
Back to Events