BUG: spin-polarized CG cell-relax blast for Fe3C bulk #5433
Labels
Bugs
Bugs that only solvable with sufficient knowledge of DFT
GeometryRelaxation
Issues related to geometry relaxation
Describe the bug
During relaxation, the LARGEST GRAD will stuck around 0.08 eV/A ng, and after 5 steps the gradient will blast to 25 eV/Ang
This blast is due to magnetic moment unstable during relaxation process
Expected behavior
This relaxation process should be normal
To Reproduce
Fe3C-relax-blast.tar.gz
Environment
ABACUS 3.7.5 commit 732c2b8 by intel toolchain
Additional Context
No response
Task list for Issue attackers (only for developers)
The text was updated successfully, but these errors were encountered: