Artificially Intelligent Manufacturing Paradigm (AIM) for Composites
Researchers:
Program overview
This project is funded by DOE's Energy Frontier Research Center (EFRC) program.
As world demand for energy rapidly expands, transforming the way energy is collected, stored, and used has become a defining challenge of the 21st century. At its heart, this challenge is a scientific one, inspiring the U.S. Department of Energy’s (DOE) Office of Basic Energy Sciences (BES) to establish the Energy Frontier Research Center (EFRC) program in 2009. The EFRCs represent a unique approach, bringing together creative, multi-disciplinary scientific teams to tackle the toughest scientific challenges preventing advances in energy technologies. These centers take full advantage of powerful new tools for characterizing, understanding, modeling, and manipulating matter from atomic to macroscopic length scales. They also train the next-generation scientific workforce by attracting talented students and postdoctoral researchers interested in energy science.
Project summary
The mission of this EFRC, which consists of a multi-disciplinary team of experimentalists, computational researchers and computer scientists, is to build an AI-enabled inverse design approach for fundamental understanding and integrated material-manufacturing design of advanced polymer composites. While uncovering these fundamental insights, this EFRC also aims to build Inverse Design Software (InDeS) tools that accelerate the discovery of advanced polymer composites for improved performance, energy efficient manufacturing thereby enabling lower carbon footprint, lower structural weight, and lower cost.
| Research tasks | Challenges |
| 1. Material-process-microstructure-performance modeling pathway: Modeling and predicting (a) the physical interactions between the material constituents and manufacturing induced effects on the microstructures, and (b) correlations between material microstructure and property. | 1 and 3 |
| 2. Multiscale coupling: We propose a mathematical theory of coarse-graining and statistical methods applied to concurrent coupling of heterogeneous multiscale solvers for scale-bridging. | 1 and 3 |
| 3. Energy-economic-environment (E3)modeling: We integrate techno-economic analysis (TEA), life cy-cle assessment (LCA) and a Climate Social Model (CSM) [7] into a single analysis which will evaluate E3impacts associated with design choices related to materials and processes and incorporate human behavioral responses into carbon emission prediction based on the theory of planned behavior. | 1 |
| 4. Uncertainty quantification: A Bayesian network-based framework will be developed to calibrate and assess aleatory uncertainty contributions of simulation models in different stages. | 1 and 3 |
| 5. Experiments integration: In addition to the capabilitiesof CCC and U. of Florida, the team will seek assistance from national labs (SRNL and PNNL) to expand experimental condition envelops and generate new experimentaldata for validation of the DLC models. | 3 |
| 6. Data fusion and flow: We propose an integrated data container based on theFunctional Mock-Up Interface/Unit(FMI/FMU)framework to standardize the data format of the metadata, DLC and AI model input and results, experimental data as well as the binary solver codes. We will implement FAIR (Findability, Accessibility, Interoperability, and Reusability) principles for data fusion and flow. | 8 |