| Autor: |
Andrew E. Slaughter, Zachary M. Prince, Peter German, Ian Halvic, Wen Jiang, Benjamin W. Spencer, Somayajulu L.N. Dhulipala, Derek R. Gaston |
| Jazyk: |
angličtina |
| Rok vydání: |
2023 |
| Předmět: |
|
| Zdroj: |
SoftwareX, Vol 22, Iss , Pp 101345- (2023) |
| Druh dokumentu: |
article |
| ISSN: |
2352-7110 |
| DOI: |
10.1016/j.softx.2023.101345 |
| Popis: |
Stochastic simulations are ubiquitous across scientific disciplines. The Multiphysics Object-Oriented Simulation Environment (MOOSE) includes an optional module – stochastic tools – for implementing stochastic simulations. It implements an efficient and scalable scheme for performing stochastic analysis in memory. It can be used for building meta models to reduce the computational expense of multiphysics problems as well as perform analyses requiring up to millions of stochastic simulations. To illustrate, we have provided an example that trains a proper orthogonal decomposition reduced-basis model. The impact of the module is detailed by explaining how it is being used for failure analysis in nuclear fuel and reducing the computational burden via dynamic meta model training. The module is unique in that it provides the ability to use a single framework for simulations and stochastic analysis, especially for memory intensive problems and intrusive meta modeling methods. |
| Databáze: |
Directory of Open Access Journals |
| Externí odkaz: |
|
