| 查看: 514 | 回复: 0 | ||
| 当前主题已经存档。 | ||
[资源]
【分享】Trust radius
|
||
|
Today, I try to opitimize carbon atom chain with 4 atoms per cell. When I got the following atom position, it failed. ATOMIC_POSITIONS (crystal) C 0.001769728 0.000000000 0.000000000 C 0.251731600 0.000000000 0.000000000 C 0.501768886 0.000000000 0.000000000 C 0.751729786 0.000000000 0.000000000 From the above data, it is easy to see that these four atoms almost equally divide the primitive vector of x axis. Besides, I find more information in .rx.out file: trust_radius < trust_radius_min Total force = 0.000080 Total SCF correction = 0.000222 SCF correction compared to forces is too large, reduce conv_thr history already reset at previous step Furthermore, I find useful guide at http://www.democritos.it/piperma ... ctober/005249.html: "My advise is that you should not reduce trust_radius_min. Instead, increase the accuracy of your forces by reducing conv_thr even more (or by increasing upscale). You can monitor the error on the forces due to lack of self-consistency by checking the size of total scf correction (it must be much smaller than the total force). However this is not the only source of numerical errors. Since some terms are computed in real space on a equispaced grid, the energy of the system is not invariant for arbitrary translations. This can easily confuse BFGS algorithm when the convergence criteria are too tight.To reduce this error you should increase ecutrho. However I don't see the reason to waste cpu-time by using such a tight criteria. Also, consider that the self-consistency itself is going to fail if you ask conv_thr < 1.D-12 because the iterative diagonalization schemes we use can suffer numerical instabilities when the required accuracy for eigenvalues/eigenvectors is to high." From the expert's explanation, I know the accuracy of forces is close related with the conv_thr of scf. By the way,if atom positions are almost equal,trust_radius_ini should be reduced. 以上内容来自我的博客：blog.sina.com.cn/nkasir |
回复此楼» 猜你喜欢
越来越担心硕士毕业后没有竞争力了
已经有14人回复
-
论文的挂名怎么评判贡献？基金申请中，排名第一的挂名文章
已经有3人回复
-
申博自荐 2篇一区TOP+1篇EI+2公开发明专利 2027博士申请
已经有3人回复
-
闲聊
已经有6人回复
-
申博求助
已经有4人回复