本文所有内容来自互联网，本人使用vasp时间有限，仅供参考，如有不适，感谢留言
参考
pdf:Tutorial 6:Vasp Calcula1ons for Ab Ini’o Molecular Dynamics
Molecular – Dynamics
pdf:VASP Tutorial:Atoms,molecules,and bulk systems
VASP md 模拟 INCAR 例子
苏长荣-VASP安装和使用说明
一份被四处转载的【原创】vasp的分子动力学模拟
参数
推荐参数，增加计算速度
PREC = Normal # chose Low only after tests
ISYM = 0 #Strong recommend for MD 0 ; 0-nonsym 1-usesym 2-fastsym
ALGO=Very Fast #
LREAL=A #the projection operators in real space
LWAVE = F #do not write WAVECAR
LCHARG = F #do not write CHGCAR
根据Molecular – Dynamics的建议
Use ALGO=Very Fast (RMM-DIIS for electrons) for large molecular dynamics runs. One should also evaluate the projection operators in real space (LREAL=A) to speed up the calculations, and it is recommended to use at least 4 electronic iterations per ionic step (NELMIN = 4). For surface or difficult systems, you might need to increase this value to NELMIN = 8.
核心参数
下面的参数跟体系不同而不同
IBRION = 0 # ionic relax: 0-MD 1-quasi-New 2-CG 5,6 zhendong
POTIM=20 #每一步的时间步长，单位fs
NSW=2000 #模拟步数，总time=NSW*POTIM
NELMIN= 4; # 电子自洽最小迭代次数，表面或其他的特殊系统可能需要设为8 min NELM of ELM steps
NBLOCK = 1; #每过NBLOCK个离子步写一次CONCAR
KBLOCK = 1 #每过KBLOCK*NBLOCK离子步输出平均对关联函数和DOS
TEIN = 0.0 # initial temperature，默认0K，目前我还没有设置过，用了再补充
TEBEG = 2000 #MD开始温度，程序会按照TEBEG按照Maxwell-Boltzman分布随机产生原子的初始速度
TEEND = 3000 # MD结束温度temperature during run，默认TEBEG
SMASS = 0 # -3 NVE; -1 模拟退火; >=0 NVT Nose mass-parameter (am)
具体解释，此处引用苏长荣老师的说明
其他
精度
EDIFF=1E-4或1E-5 #对于长时间的分子动力学的模拟，精度小一点也无所谓，但不能太小，恒温找能量最低结构时，精度太小算的不准
ENCUT= #其他人的计算普遍取得比较低，如https://www.doi.org/10.1103/PhysRevB.75.235405计算Pt13团簇使用191eV
KPOINTS
原子数较多或元胞较大时常取一个Γ点
原子数
考察空位的运动时，晶格要足够大，原子数较少时，运动比较困难
结果分析
CONTCAR是最后一步的位置和速度
任务终止后，可以cp CONTCAR POSCAR继续计算
After performing one MD run, it is possible to continue to run, by copying the CONTCAR to POSCAR file and restarting VASP. Since the CONTCAR file is written after every single step, it is also possible to restart the molecular dynamics from a crashed run.
XDATCAR MD每一步的原子坐标
可使用xdat2xyz.pl 转变为xyz动画
xdat2xyz.pl 来自VTST-scripts
OSZICAR
- T前面的数字代表MD的步数
- T=当前温度
- E=总能量(离子动能和电子总能)
E to the total free energy (including the kinetic energy of the ions and the energy of the Nosé thermostat). - F = 是体系的总能量(电子的能量，不包括离子动能), 与OUTCAR中 free energy TOTEN 后面的值相等
- E0 后面的能量对应OUTCAR中 energy (sigma->0)后面的能量(电子的能量，不包括离子动能)
实例:输入文件和具体过程见 密度泛函习题解答中第9章的MD部分
举其中一例如下
对体相Si空位，使用能量恒定的MD方法，生成体系总能严格守恒的轨迹
使用NVE系综
输入文件DFT-EXERCISES
INCAR
# MD
IBRION=0 #0MD
POTIM=10 #fs
NSW=30000 #time=NSW*POTIM
NELMIN= 6; # min NELM # NELMDL= 0 # of ELM steps
NBLOCK = 5; #step to write CONCAR
KBLOCK = 10 #NBLOCK*KBLOCK step to write XDATCAR inner block; outer block
# TEIN = 0.0 initial temperature
TEBEG = 3000
# TEEND = 300 # temperature during run
SMASS = -3 # -3 NVE; -1 tui huo; >=0 NVT Nose mass-parameter (am)
# estimated Nose-frequenzy (Omega) = 0.10E-29 period in steps =****** mass= -0.143E-26a.u.
ISYM = 0 # 0-nonsym 1-usesym 2-fastsym
#
#
# MAIN
ISTART = 0 job : 0-new 1-cont 2-samecut
ICHARG = 2 charge: 1-file 2-atom 10-const
# ISPIN = 1 spin polarized calculation?
ENCUT = 300.0 eV # 36.75 Ry 6.06 a.u. 14.43 14.43 43.64*2*pi/ulx,y,z
NELM = 400; #
EDIFF = 1E-4 stopping-criterion for ELM
ISMEAR = 0;
SIGMA = 0.1 broadening in eV -4-tet -1-fermi 0-gaus
EDIFFG = -0.03 # NEB must < 0 stopping-criterion for IOM
# NSW = 500 number of steps for IOM
# IBRION = 2 # ionic relax: 0-MD 1-quasi-New 2-CG 5,6 zhendong
# NFREE = 2 #IBRION=5,6 #zhendong steps in history (QN), initial steepest desc. (CG)
ISIF = 2 # 2 surface 3 bulk stress and relaxation
# POTIM = 0.1000 time-step for ionic-motion
LWAVE = F write WAVECAR
LCHARG = F write CHGCAR
提取计算结果
grep T OSZICAR |awk '{print $1"\t"$5}'
本文首发于我的博客@cndaqiang.
本博客所有文章除特别声明外，均采用 CC BY-SA 4.0 协议 ，转载请注明出处！