Complexity-reduction using automatic level grouping for atomic collisional-radiative models
| Autor: | David Bilyeu, Richard June Abrantes, Robert Martin, Eder Sousa |
|---|---|
| Rok vydání: | 2020 |
| Předmět: |
Numerical Analysis
Work (thermodynamics) Physics and Astronomy (miscellaneous) Computer science Applied Mathematics Construct (python library) Interference (wave propagation) Computer Science Applications Reduction (complexity) Computational Mathematics symbols.namesake Modeling and Simulation Boltzmann constant Radiative transfer symbols Cluster analysis Algorithm |
| Zdroj: | Journal of Computational Physics. 407:109213 |
| ISSN: | 0021-9991 |
| DOI: | 10.1016/j.jcp.2019.109213 |
| Popis: | In a previous work that investigated atomic collisional-radiative models, Boltzmann grouping was implemented to reduce and accelerate detailed-configuration-accounting or DCA-based CR simulations, exhibiting distinct advantages over other CR reduction techniques. However, the selection of level groups under this reduction technique was manually performed and required several iterations to construct the appropriate groups that properly showed the advantages associated with Boltzmann grouping. Therefore, a clustering technique was implemented to automatically group detailed (LS-coupled) atomic states with limited user interference. Clustering in conjunction with the Boltzmann grouping technique were applied to collisional and coronal simulations in this work to demonstrate the feasibility of automatic level grouping techniques to produce accurate reduced simulations under various plasma conditions. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|
Full Text from ScienceDirect