Autor: |
Jaylen R. James, Meet Sanghvi, Austin R. C. Gerlt, Douglas Allaire, Raymundo Arroyave, Manny Gonzales |
Jazyk: |
angličtina |
Rok vydání: |
2022 |
Předmět: |
|
Zdroj: |
Frontiers in Materials, Vol 9 (2022) |
Druh dokumentu: |
article |
ISSN: |
2296-8016 |
DOI: |
10.3389/fmats.2022.932574 |
Popis: |
In computational materials research, uncertainty analysis (more specifically, uncertainty propagation, UP) in the outcomes of model predictions is essential in order to establish confidence in the models as well as to validate them against the ground truth (experiments or higher fidelity simulations). Unfortunately, conventional UP models relying on exhaustive sampling from the distributions of input parameters may be impractical, particularly when the models are computationally expensive. In these cases, investigators must sacrifice accuracy in the propagated uncertainty by down-sampling the input distribution. Recently, a method was developed to correct for these inaccuracies by re-weighing the input distributions to create more statistically representative samples. In this work, the method is applied to computational models for the response of materials under high strain rates. The method is shown to effectively approximate converged output distributions at a lower cost than using conventional sampling approaches. |
Databáze: |
Directory of Open Access Journals |
Externí odkaz: |
|