Top Page > Browsing

Non-Collinear DFT Euler Angles Date: 2022/09/14 00:56 Name: Max Fust&#233; Costa   <max.fuste.costa@gmail.com> Hi, I'm currently doing a study on a CoPt grid using non-collinear DFT. I've run into a problem while modifying the values of the Euler angles. When the values of either theta or phi are over 40, the code does not converge. I've been told that there's a way to use previous results in new runs (so I could start from 30 and go to 40, instead of going from 0 to 40), but I have no idea how to do so. Would it be possible for any of you to give me a hand on it? (I think that increasing computation time would work, but I really don't want to send jobs of 2hrs, as I need to do a lot of them and it would take forever). Thanks in advance!

Page: [1]

Re: Non-Collinear DFT Euler Angles ( No.1 ) Date: 2022/09/14 12:56 Name: Naoya Yamaguchi Hi, You can try another charge mixing scheme. The manual (http://www.openmx-square.org/openmx_man3.9/node40.html) says "'RMM-DIISH', mixes Kohn-Sham Hamiltonian matrices, which may be suitable for the plus U method and the constraint schemes." Regards, Naoya Yamaguchi Re: Non-Collinear DFT Euler Angles ( No.2 ) Date: 2022/09/14 22:51 Name: T. Ozaki Hi, If the angle you want to set is far from an easy axis, it is natural to see that the convergence is difficult. For such a case, one can use the constraint scheme as explained at http://www.openmx-square.org/openmx_man3.9/node113.html Also, RMM-DIISH may improve the convergence. Regards, TO

Page: [1]