CDI: Additive Manufacturing
Additive manufacturing (AM) has evolved in recent years from primarily a rapid prototyping tool into a more widespread manufacturing approach for functional polymer- or metal-based parts. Increased part quality with faster production rates and the adoption of high-performance feedstock materials for printing have all contributed to this transformation and growth in demand for these technologies. These advances in AM part properties coupled with decreases in equipment costs have decreased the barrier to entry for would-be producers.
An emerging challenge due to this proliferation of AM is the need for producers and end-users to verify the authenticity of the parts produced. Furthermore, with current technology and understanding, AM-produced parts may often function differently, possess more defects, and have less reproducibility and assurance in their structure than analogous traditionally manufactured parts. These limitations with AM goods arise from the difficulty of correlating both feedstock and specific hardware or build conditions to the final structures printed.
As part of CDI, this use case aims to establish testbeds to:
- create data-driven performance models for AM-produced components with uncertainties well understood
- streamline the correlation of material or process signatures characteristic of AM processes for provenance, increased fundamental understanding, and to support QA/QC in the field
Examples of specific areas of focus include:
- printing structures with high-performance polymers and related blends or composites on low-cost machines
- data assimilation for microstructure prediction by coupling models and in situ experiments
- evolution of micro- and -chemical structure present in printed parts due to post-processing and during application in harsh real-world environments
- the use of data and workflow management infrastructure to accelerate analysis
- application of deep learning tools on large data sets collected for classification and property prediction