Digital Engineering and Advanced Manufacturing

Within a model-based framework, results from simulations, models, and experiments are linked.  It is infeasible to probe the entire space of possible microstructures in a component, so it is necessary to capture this information in a probabilistic framework, specifically addressing how uncertainties are quantified in the modeling/experimental approach and propagate throughout the framework.  In such projects, models are used with appropriate verification, validation, and uncertainy quantification activities necessary to inform decision making.  

AFRL Conceptual Figure V5

Such projects within our group have used data-driven approaches and fused together data from experiments and models to employ machine learning to uncover the underpinning physics.  For instance, we have made inferences between the results of HEDM experiments and crystal plasticity modeling within a Bayesian network to understand how a short crack is moving through the local  microstructure, thus using machine learning to identify the driving force for crack advancement.

Research Themes:

  • Advanced Manufacturing
  • Digital Twin
  • Decision Making
  • Machine Learning / Data-Driven Approaches
  • Uncertainty Quantification

Design, analysis, manufacturing, inspection, maintenance, and repair are all part of the produce life cycle, albeit each individual role typically uses disparate technology.  When passing information from one job role to another, industry relies on human-to-human interactions or paper drawings.  Based on a digital manufacturing project in collaboration with Prof. Hartman of Purdue, we are working to replace drawings by completing the model-based definition to store, archive, and retrieve process data, materials information, and variable as-built data created during the manufacturing life cycle within computer aided design models.  The goal of this project is to make the flow of information across the complete life cycle seamless (figure below) and thereby eliminate the need for paper drawings.  The project is part of the Digital Manufacturing and Design Innovations Instutitue and is led by Lockheed Martin and team members.  This is necessary to fully realize a digital twin framework, thereby enabling the use of manufacturing process and materials performance modeling.

Screen Shot 2018-01-20 at 9.45.54 AM

© msangid 2022