博文
GW计算的艰难算例尝试
||
关注：
1）XH3结构试验带隙的理论重现
2）尝试过程中遇到的错误摘录
工作目录：
/db/home/yexq/work/elec-dos/gw-cal/sch3-d194-0g/test2-noso/gw2
/db/home/yexq/work/elec-dos/gw-cal/yh3-p63mmc
遇到问题：
警告类：
| For optimal performance we recommend that you set |
| NPAR = 4 - approx SQRT( number of cores) |
| (number of cores/NPAR must be integer) |
| This setting will greatly improve the performance of VASP for DFT. |
| The default NPAR=number of cores might be grossly inefficient |
| on modern multi-core architectures or massively parallel machines. |
| Do your own testing. |
| Unfortunately you need to use the default for hybrid, GW and RPA |
| calculations.
| |
| You calculate the electron-hole interaction using AEXX=0. |
| This is hardly what you want to do. |
| Maybe you have forgotten to set AEXX in the INCAR file. |
| ( For LHFCALC=.TRUE. the default is AEXX=0.25, but if LHFCALC is not |
| set the default is AEXX=0.0 )
OUTCAR中出现NAN:
Xi from CALCULATE_LOCAL_FIELD_FOCK
-------------------------------------
w= 0.000 0.000
NaN NaN NaN NaN NaN NaN
NaN NaN NaN NaN NaN NaN
NaN NaN NaN NaN NaN NaN
0.000 NaN NaN dielectric constant
real part
NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN
NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN
NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN
NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN
NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN
NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN
NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN
NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN
NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN
NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN
imag part
NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN
摘录：http://blog.sina.com.cn/s/blog_5f15ead20100dxe0.html
I get the following information, instruction and experimental parameter in the yambo forum.
* NGsBlkXp~=1/10*EXXRLvcs
In general the size of the response function is 1/10 of the RL components you included in HF.
NGsBlk controls the size of the response function in Reciprocal Space. Note that when NGsBlk = 1 no Local Fields are considered. This corresponds to neglect the charge oscillations induced by the external potential.
* BndsRnXp~=1/3*GbndRnge
In general BndsRnXp (400 in your case) is 1/3 of the bands you include in the Green's function (GbndRnge).
* Regarding the GW0 approximation it is very important to remind that it is an approximation. Strictly speaking it is valid only in the limit where the Random Phase approximation (RPA) holds, homogeneous and dense system. Solids in general are not homogeneous, so GW0 may not work, especially in presence of localized s/p orbitals, if the metallic screening (like in noble and transition metals) does not make the RPA meaningful.
* In the GW method W is the screened electron-electron interaction. W is written in terms of a response function that contains an integral over the whole Brillouin zone. To do so you need to provide Yambo a uniform grid of k-points.
* QPerange
first k-point|last-kpoint|lower energy|higher energy|.
So the first two values define the k-point range (e.g 1|3|0.0|5.0| will take for the kpoits 1,2,3 all bands with energies between 0 and 5 eV)
* E_0 are the KS energies (shifted so that the 0.0 correspond to the top of the highest valence band in the set of input k-points). E are the QP energies, so E-E_0 is their difference (or QP correction) and Sc(E_0) self-energy matrix at the given kpoint,band at E0.
Note that you can set up the output format using QPreport.
To quickly find the QP-energy of the highest lying occupied state, try
grep " 4 " OUTCAR | sort -n -k 3 | tail -1 | awk '{print $3}'and for the lowest lying unoccupied state,
grep " 5 " OUTCAR | sort -n -k 3 | head -1 | awk '{print $3}'
https://blog.sciencenet.cn/blog-567091-818579.html
上一篇：XRD分析数据处理技巧-拾遗
下一篇：VASP电子结构计算错误-收集
