Troubleshooting electronic convergence: Difference between revisions

Revision as of 13:05, 26 March 2021

There can be many reasons why the SCF convergence is not reached. Naturally, there are also many different possible solutions.

Try to create a minimal INCAR file with as few tags as possible. If the calculation converges then gradually add them back until you find which one was causing the system not to converge.

Check if you have enough bands. You can do this by looking at the OUTCAR file and checking that there are enough 'empty' states (i.e. states with zero occupation). When using an iterative solver the last states might not be accurately described, if these are occupied them convergence is more difficult.

Systems with f-orbitals are in particular difficult to converge. Here are a couple of tips for a magnetic calculation with LDA+U:

give initial magnetization only to the magnetic atoms
use spin-polarized calculation
perform the calculation in 3 steps (always starting from the previous WAVECAR):
1. step 1 with ICHARG=12 and ALGO=Normal
2. step 2 with ALGO=All (Conjugate gradient) and a small TIME step 0.05 instead of the default 0.4 (this is crucial)
3. step 3 LDA+U keeping ALGO=All and small TIME

@@ Line 5: / Line 5: @@
 If the calculation converges then gradually add them back until you find which one was causing the system not to converge.
-Check if you have enough bands. You can do this by looking at the OUTCAR file and checking that there are enough 'empty' states (i.e. states with zero occupation). When using an iterative solver the last states might not be accurately described, if these are occupied them convergence is more difficult.
+Check if you have enough bands. You can do this by looking at the [[OUTCAR]] file and checking that there are enough 'empty' states (i.e. states with zero occupation). When using an iterative solver the last states might not be accurately described, if these are occupied them convergence is more difficult.
 Systems with f-orbitals are in particular difficult to converge.
@@ Line 14: / Line 14: @@
 # perform the calculation in 3 steps (always starting from the previous WAVECAR):
 ## step 1 with {{TAG|ICHARG}}=12 and {{TAG|ALGO}}=Normal
-## step 2 with {{TAG|ALGO}}=All (Conjugate gradient) and a small TIME step 0.05 instead of the default 0.4 (this is crucial)
+## step 2 with {{TAG|ALGO}}=All (Conjugate gradient) and a small {{TAG|TIME}} step 0.05 instead of the default 0.4 (this is crucial)
-## step 3 LDA+U keeping {{TAG|ALGO}}=All and small TIME
+## step 3 LDA+U keeping {{TAG|ALGO}}=All and small {{TAG|TIME}}