Intelligent Manufacturing
Workshop
May 2-3, 2024
Location
CAES
995 MK Simpson Blvd
Idaho Falls, Idaho 83401
Workshop Overview
The NSF EPSCoR Workshop on Intelligent Manufacturing for Extreme Environments will bring together experts to discuss challenges and solutions. This event will provide a venue for researchers to share their findings and foster new collaborations.
Workshop Thrusts
An important manufacturing challenge for nuclear reactors is the ability to manufacture electronic components capable of withstanding the extreme environments of the reactor core, including high temperatures and high radiation, for multi-year durations. Autonomous operation and supervised semi-autonomous operation of nuclear reactors for power generation is a future requirement for nuclear power to be cost competitive. Autonomous operation will require distributed and reliable sensors for continuous monitoring and load following to integrate with intermittent power generation.
Standardization is a significant challenge for the manufacture of electronic films and components in extreme environments. Standards and metrics for reliability in extreme environments are lacking. Variables and mitigations that impact performance are not well understood. A suitable printing modality, including atomic layer deposition, needs to be identified. Likewise, inks compatible with the print modality need to be developed, characterized, and qualified. This also means a mitigation of variability in ink composition batch to batch. A standardized library of material-process relationships would be highly valuable. Manufacture of electronic components for extreme environments requires a highly skilled workforce with specific training.
Future manufacturing will incorporate digital design with fabrication by robots in a workflow called a process. Controlling a process, and hence changing its behavior, requires first the characterization of the process itself in a suitable form. Identifying and integrating sensors is necessary to provide a feedback loop that feeds machine-based learning algorithms. These sensors also process parameters to minimize unwanted changes or to affect movement of the process in a desired direction.
Based on this description, and a control objective, the structure of the control algorithm can be formulated. For control of AM processes, identification algorithms that directly estimate the resulting feedback controller based on the observed process dynamics may also incorporate robustness and adaptive features. Methods and algorithms to extract process model parameters and process model structures need exploration. In this way, feedback control principles and extracted process models can address product quality and process variations.
Future Manufacturing will use artificial intelligence and integrated machine learning techniques in a materials-and-process-by-design approach. Designing new materials is a slow and laborious process, requiring significant investments in capital and labor. Recent advances in artificial intelligence provide opportunities to greatly improve and speed up this process. In particular, machine learning techniques can replace expensive and time-consuming laboratory processes and computational simulations to quickly and reliably predict how to create materials with desired properties and how materials will behave in specific circumstances. Machine-learning (ML) enabled computation and experiments can be used to identify processing conditions needed to obtain desired microstructure-property relationship. This approach can help develop physics-based understanding of materials at the atomic and electronic levels, explore the materials phase space, and optimize process variables in manufacturing.
Qualification is the key to manufacturing of the future. Performance of a component manufactured by AM techniques needs to be predictable and provide comparable results to a component manufactured by traditional methods. Characterization of the behavior of AM components in extreme environments is essential. Mechanisms and processes at high temperatures are significantly different than at low temperatures. Understanding that behavior is critical to developing processes and materials that meet performance criteria.
Development of integrated computational and experimental models to accurately predict materials behavior at extreme temperatures is critical to qualification of processes and components. Creep and creep-fatigue of additively manufactured components need to be identified before they fail. A cyber manufacturing approach will allow the elucidation of complex interactions of high temperature and other extreme conditions, such as irradiation, corrosion, and external fields. Machine learning techniques will be necessary to determine the dominant mechanism at extreme temperature. Innovative in-situ characterization techniques can inform computational models to explore multi-length-scale phenomena.
Workforce development is needed for local communities to provide the large number of highly trained workers to support the growth in the nuclear energy industry in eastern Idaho and Wyoming. Nuclear facilities have unique requirements, such as Nuclear Quality Assurance-1 (NQA-1), an industry consensus standard created and maintained by the Society of Mechanical Engineers. Community colleges play a unique role in the workforce development ecosystem and can provide the necessary technical training, such as NQA-1 orientation and radiation safety, for workers seeking on-ramps to the nuclear energy industry. Workforce development needs for the expected growth in advanced reactors in the region include a wide variety of different job areas including construction, trades, technology, health and human services, and business.
Workshop Sessions
Snake River Conference Room
CAES Auditorium
Teton Conference Room
CAES Gallery