KERNEL TRUNCATION/LTRUNCATE: Difference between revisions
No edit summary |
No edit summary |
||
| Line 3: | Line 3: | ||
Description: Truncates the Coulomb kernel to remove [[:Category:electrostatics|electrostatic interactions]] along non-periodic dimensions. | Description: Truncates the Coulomb kernel to remove [[:Category:electrostatics|electrostatic interactions]] along non-periodic dimensions. | ||
---- | ---- | ||
Setting {{TAG|KERNEL_TRUNCATION/LTRUNCATE}} = T switches on the Coulomb-kernel-truncation method{{cite|vijay:prb:2025}}. It effectively removes interactions with periodic replicas in non-periodic directions. In other words, the interactions are removed along the surface normal for [[2D materials]], and along all directions for 0D systems, i.e. for isolated atoms and molecules. | Setting {{TAG|KERNEL_TRUNCATION/LTRUNCATE}} = T switches on the Coulomb-kernel-truncation method{{cite|vijay:prb:2025}}{{cite|rozzi:prb:2006}}{{cite|sohier:prb:2017}}. It effectively removes interactions with periodic replicas in non-periodic directions. In other words, the interactions are removed along the surface normal for [[2D materials]], and along all directions for 0D systems, i.e. for isolated atoms and molecules. | ||
{{NB|important| Slabs must be centered in the unit cell.}} | {{NB|important| Slabs must be centered in the unit cell.}} | ||
In the simplest implementation of the Coulomb-kernel-truncation method ({{TAG|KERNEL_TRUNCATION/LCOARSEN|F}}), the computational cell provided in the {{FILE|POSCAR}} file is internally padded by an additional vacuum that is of the same extend as the computational cell in each direction. This implies increasing the [[Energy cutoff and FFT meshes|FFT-grid sizes]] by a certain factor (see {{TAG|KERNEL_TRUNCATION/IPAD}}) and thus leads to a significant increase in computational cost. | In the simplest implementation of the Coulomb-kernel-truncation method ({{TAG|KERNEL_TRUNCATION/LCOARSEN|F}}), the computational cell provided in the {{FILE|POSCAR}} file is internally padded by an additional vacuum that is of the same extend as the computational cell in each direction. This implies increasing the [[Energy cutoff and FFT meshes|FFT-grid sizes]] by a certain factor (see {{TAG|KERNEL_TRUNCATION/IPAD}}) and thus leads to a significant increase in computational cost. | ||
Revision as of 10:55, 17 March 2026
KERNEL_TRUNCATION/LTRUNCATE = .True. | .False.
Default: KERNEL_TRUNCATION/LTRUNCATE = .False.
Description: Truncates the Coulomb kernel to remove electrostatic interactions along non-periodic dimensions.
Setting KERNEL_TRUNCATION/LTRUNCATE = T switches on the Coulomb-kernel-truncation method[1][2][3]. It effectively removes interactions with periodic replicas in non-periodic directions. In other words, the interactions are removed along the surface normal for 2D materials, and along all directions for 0D systems, i.e. for isolated atoms and molecules.
| Important: Slabs must be centered in the unit cell. |
In the simplest implementation of the Coulomb-kernel-truncation method (KERNEL_TRUNCATION/LCOARSEN = F), the computational cell provided in the POSCAR file is internally padded by an additional vacuum that is of the same extend as the computational cell in each direction. This implies increasing the FFT-grid sizes by a certain factor (see KERNEL_TRUNCATION/IPAD) and thus leads to a significant increase in computational cost.
Tip: Use the KERNEL_TRUNCATION/LCOARSEN = T to avoid the increased FFT-grid sizes.
|
Mind:
|
Detailed information about the setting are documented on respective related tags.
Related tags and articles
KERNEL_TRUNCATION/LCOARSEN, KERNEL_TRUNCATION/IDIMENSIONALITY, KERNEL_TRUNCATION/ISURFACE, KERNEL_TRUNCATION/FACTOR, KERNEL_TRUNCATION/IPAD
References
- ↑ S. Vijay, M. Schlipf, H. Miranda, F. Karsai, M. Kaltak, M. Marsman, and G. Kresse, Efficient periodic density functional theory calculations of charged molecules and surfaces using Coulomb kernel truncation, Phys. Rev. B 112, 045409 (2025).
- ↑ C. A. Rozzi, D. Varsano, A. Marini, E. K. Gross, A. J. Rubio, Phys. Rev. B 73, 20511 (2006).
- ↑ T. Sohier, M. Calandra, and F. Mauri, Phys. Rev. B 96, 75448 (2017).