Ensemble Predictions of Material Behavior for ICMSE for Additive Structures
Abstract
Integrated Computational Materials Engineering (ICME) is an engineering approach whereby model linkages as well as experimental and computational integration are exploited in order to efficiently explore materials processing-performance relationships. Many materials models and simulations are deterministic in nature. Achieving a statistical confidence in a simulation output requires, first, the identification of the various sources of error and uncertainty affecting the simulation results. Statistical inference can then be used to recover information about unknown model parameters by conditioning on available data while taking into account the various sources of uncertainty. A statistical random effects hierarchical framework was developed and demonstrated. First for parameter estimation and response prediction in a phenomenological viscoplastic self-consistent (VPSC) crystal plasticity model [1]. Inference is performed under two different scenarios: 1) with the consideration of model discrepancy, modeled through a Gaussian process and 2) without the consideration of model discrepancy. Second, uncertainty quantification and propagation is demonstrated for calculations of phase equilibrium for the binary AgCu system. A Metropolis-Hastings Markov chain Monte Carlo algorithm is used to estimate model parameters with a quantified level of uncertainty, which induces a probability distribution in the simulation output, and then is quantified using posterior summaries.
Document Details
- Document Type
- Technical Report
- Publication Date
- Aug 24, 2021
- Accession Number
- AD1162752
Entities
People
- Denielle Ricciardi
- Stephen R Niezgoda
Organizations
- Ohio State University