A1 Vertaisarvioitu alkuperäisartikkeli tieteellisessä lehdessä
An automated algorithm for reliable equation of state fitting of magnetic systems
Tekijät: H.Levämäki, L.-Y. Tian, L.Vitos, M.Ropo
Kustantaja: ELSEVIER SCIENCE BV
Julkaisuvuosi: 2019
Journal: Computational Materials Science
Tietokannassa oleva lehden nimi: COMPUTATIONAL MATERIALS SCIENCE
Lehden akronyymi: COMP MATER SCI
Vuosikerta: 156
Aloitussivu: 121
Lopetussivu: 128
Sivujen määrä: 8
ISSN: 0927-0256
DOI: https://doi.org/10.1016/j.commatsci.2018.09.026
Tiivistelmä
In computational physics and materials science ground-state properties are often extracted from an equation of state fit to energy-volume data. Magnetic systems often have multiple magnetic phases present in the energy-volume data, which poses a challenge for the fitting approach because the results are sensitive to the selection of included fitting points. This is because practically all popular equation of state fitting functions, such as Murnaghan and Birch-Murnaghan, assume just one phase and therefore cannot correctly fit magnetic energy-volume data that contains multiple phases. When fitting magnetic energy-volume data it is therefore important to select the range of fitting points in such a way that only points from the one relevant phase are included. We present a simple algorithm that makes the point selection automatically. Selecting fitting points automatically removes human bias and should also be useful for large-scale projects where selecting all fitting points by hand is not feasible.
In computational physics and materials science ground-state properties are often extracted from an equation of state fit to energy-volume data. Magnetic systems often have multiple magnetic phases present in the energy-volume data, which poses a challenge for the fitting approach because the results are sensitive to the selection of included fitting points. This is because practically all popular equation of state fitting functions, such as Murnaghan and Birch-Murnaghan, assume just one phase and therefore cannot correctly fit magnetic energy-volume data that contains multiple phases. When fitting magnetic energy-volume data it is therefore important to select the range of fitting points in such a way that only points from the one relevant phase are included. We present a simple algorithm that makes the point selection automatically. Selecting fitting points automatically removes human bias and should also be useful for large-scale projects where selecting all fitting points by hand is not feasible.
Turun yliopiston Tutkimustietojärjestelmän portaali