Ionic relaxation calculations with constrained magnetic moments

[liang_zhang]

Dear all,

Is it possible to do the ionic relaxation calculations with constrained magnetic moments? I know that the when do the non-collinear constrained magnetic calculations at a specific ionic geometry, it should increase LAMBDA stepwise so one can bring E_p down by reading the CHGCAR. However, how can I do the ionic relaxation calculations (IBRION = 2) at the same time, if I need to increase LAMBDA at each ionic step? Or can I simply fix LAMBDA to a specific value (e.g. 10) when doing the ionic relaxation calculations?

Thanks and looking forward to your response.

Best wishes,
Liang

[michael_wolloch]

Dear Liang,

you can use the constrained magnetic moments approach in relaxations and keep LAMBDA constant during the relaxation.

One feasible approach would be to increase LAMBDA stepwise in the starting configuration until your constraint is as small as possible while still giving you the intended result, and then relax with that value.

Monitor the magnetic moments during the relaxation to verify that the constraint is strong enough.

You can then repeat that process for the fully relaxed structure.

Let me know if that explanation was sufficient, Michael

[liang_zhang]

Thanks for the reply! I will try that.