||
关注:
1) 精致计算,需测试截断能和K点
2)截断能与总能计算结果的关系
截断能选取对总能影响较大,是因为没有选择让总能收敛的截断能?
3) 截断能与压力的关系
不同压力下,结构的总能比较,应选择满足最高压力下总能收敛的截断能测试
4) 截断能大小与计算耗时的关系
5) 截断能测试的收敛标准
http://muchong.com/html/201308/6193321.html
ENCUT(eV) time(s)
200 24.438
400 40.815
600 76.253
800 113.419
MS软件可以查看你所计算体系的截断能大概是多少,然后在这附近每隔50eV 左右取点优化,你这样优化跨度是不是太大了?不同的截断能会严重影响计算速度,而且与你所取的赝势有关,US系列势截断能较低,计算速度快,paw系列较高,计算速度较慢,不同的赝势要搭配 “合适” 的ENCUT。我优化晶格常数时取PAW势一个点要跑十几分钟,一次要跑1-2小时。
不知道对你有没有帮助。
http://muchong.com/html/201212/5264940.html
总能量相差小于0.001eV时,就说是已经收敛了,但我最近做的NaAlH4的超胞,96个原子,很难收敛到0.001eV,因此,我想是不是候博士那个地方说的不够清楚,应该是平均每个原子的能量相差0.001eV,而不是总能量。后来,我问导师,导师说是按照总能量的,但我觉得这样不合理啊~如何一个大的体系,要求总能量的变化在0.001eV,那不是很困难吗?本来体系大的话就难以计算,还哪能要求这么严格?因此,带着这个疑问,我想在这里跟大家交流一下,到底是总能量还是平均每个原子的能量,希望有经验的虫友勇于发言啊~
这是我计算结果曲线,虽然都计算到了800eV,但总能量还没有达到0.001eV。
To relax the structures and calculate the electronic structures, we employ the planewave
basis projector augmented wave (PAW) formalism [36] in the framework of DFT within the generalized gradient
approximation as implemented in the VASP code [37]. The use of energy cutoff of 850 eV and Monkhorst-Pack k meshes with grid spacing of 2π × 0.02 °A−1 are then adopted to ensure that the enthalpy converges to an accuracy better than 1 meV/atom. The convergence threshold was set as 10−6 eV
Once we have the predicted structures at hand, we then optimized a number of low-enthalpy
structures at a higher accuracy as a function of pressure using either VASP or CASTEP method.
The plane wave basis set was constructed using the energy cutoff of 800 eV in all calculations, and
the Brillouin zone was sampled with a k-point resolution of 2π × 0.03 Å−1. This usually gives total enthalpies well converged within ~ 1 meV/atom
in energy and 10−3 eV in force. Sc
Archiver|手机版|科学网 ( 京ICP备07017567号-12 )
GMT+8, 2024-12-23 00:50
Powered by ScienceNet.cn
Copyright © 2007- 中国科学报社