博文
VASP学习之：保持收敛精度，提高优化速率的策略
||
关注：
如何保持收敛精度的情况下，提高优化速率
引子：具体见附注
优化四次以后的焓差，为提高优化速度，精度降低了点
由
EDIFF = 1E-6
EDIFFG = -2E-3
降低为：
EDIFF = 1E-5
EDIFFG = -3E-3
1. 解决措施建议：
在测试完ECUT和KPOINTS之后，准备用下面的参数再优化一遍试试
EDIFF = 1E-6
EDIFFG = -2E-3
POTIM = 0.05 若POTIM取得太大，如0.2会发生跳态现象！！！
NSW=500
1) POTIM代表步长，步长太大，收敛的不是极值；POTIM 太小是否会影响优化速度？如果已经接近收敛了，应该不会影响太大
在已经接近收敛的情况下， POTIM会影响收敛结果吗？ 一般 POTIM应该为多大。
2) ISMEAR = -5 保险起见设置成-5，如设置成1或0，有可能级占据数出错，从而多次优化后能量仍然相差很大。
ISMEAR=1时，面心立方结构XH3
OUTCAR_800-02: Total CPU time used (sec): 1385.830
OUTCAR_800-03: Total CPU time used (sec): 1759.621
774.22 100.48 -34.66049192 OUTCAR_800-01
774.27 100.47 -34.66053244 OUTCAR_800-04
800.07 100.47 -35.64301370 OUTCAR_800-03
800.35 100.62 -37.01398180 OUTCAR_800-02
CONTCAR_800-04
XH3
1.00000000000000
4.6366048767523642 -0.0000000000000000 0.0000000000000000
0.0000000000000000 4.6366048767523642 -0.0000000000000000
0.0000000000000000 -0.0000000000000000 4.6366048767523642
H X
12 4
Direct
0.5000000000000000 0.5000000000000000 0.5000000000000000
0.5000000000000000 0.0000000000000000 0.0000000000000000
0.0000000000000000 0.5000000000000000 0.0000000000000000
0.0000000000000000 0.0000000000000000 0.5000000000000000
0.7500000000000000 0.2500000000000000 0.2500000000000000
0.2500000000000000 0.7500000000000000 0.7500000000000000
0.2500000000000000 0.7500000000000000 0.2500000000000000
0.7500000000000000 0.2500000000000000 0.7500000000000000
0.2500000000000000 0.2500000000000000 0.7500000000000000
0.7500000000000000 0.7500000000000000 0.2500000000000000
0.7500000000000000 0.7500000000000000 0.7500000000000000
0.2500000000000000 0.2500000000000000 0.2500000000000000
0.0000000000000000 -0.0000000000000000 -0.0000000000000000
0.0000000000000000 0.5000000000000000 0.5000000000000000
0.5000000000000000 -0.0000000000000000 0.5000000000000000
0.5000000000000000 0.5000000000000000 -0.0000000000000000
ISMEAR=-5时，面心立方结构XH3
OUTCAR_800-01: Total CPU time used (sec): 1401.258
OUTCAR_800-02: Total CPU time used (sec): 340.458
OUTCAR_800-03: Total CPU time used (sec): 339.044
OUTCAR_800-04: Total CPU time used (sec): 339.415
799.92 100.45 -35.63488710 OUTCAR_800-01
800.08 100.45 -35.63699566 OUTCAR_800-02
800.08 100.45 -35.63699566 OUTCAR_800-03
800.08 100.45 -35.63699566 OUTCAR_800-04
CONTCAR_800-04
XH3
1.00000000000000
5.1862242792515811 0.0000000000000000 0.0000000000000000
0.0000000000000000 5.1862242792515802 0.0000000000000000
0.0000000000000000 0.0000000000000000 5.1862242792515838
H X
12 4
Direct
0.5000000000000000 0.5000000000000000 0.5000000000000000
0.5000000000000000 0.0000000000000000 0.0000000000000000
0.0000000000000000 0.5000000000000000 0.0000000000000000
0.0000000000000000 0.0000000000000000 0.5000000000000000
0.7500000000000000 0.2500000000000000 0.2500000000000000
0.2500000000000000 0.7500000000000000 0.7500000000000000
0.2500000000000000 0.7500000000000000 0.2500000000000000
0.7500000000000000 0.2500000000000000 0.7500000000000000
0.2500000000000000 0.2500000000000000 0.7500000000000000
0.7500000000000000 0.7500000000000000 0.2500000000000000
0.7500000000000000 0.7500000000000000 0.7500000000000000
0.2500000000000000 0.2500000000000000 0.2500000000000000
0.0000000000000000 0.0000000000000000 0.0000000000000000
0.0000000000000000 0.5000000000000000 0.5000000000000000
0.5000000000000000 0.0000000000000000 0.5000000000000000
0.5000000000000000 0.5000000000000000 0.0000000000000000
多次优化，越优化越糟糕的例子
cat pv_et.txt
压力、体积、 焓值
0.1 258.33 -172.73847059 OUTCAR_0-01
3.93 257.26 -172.68570617 OUTCAR_0-02
-2.66 258.70 -172.70586796 OUTCAR_0-03
耗时分析：
OUTCAR_0-01: Total CPU time used (sec): 284350.625
OUTCAR_0-02: Total CPU time used (sec): 203635.344
OUTCAR_0-03: Total CPU time used (sec): 392039.906
Handson IV: What can one do when SOC/magmon convergence is bad?
*** Start from charge density of non-spin-polarized calculation, using
ISTART = 0 (or remove WAVECAR)
ICHARG = 1
***Linear mixing
BMIX = 0.0001 ; BMIX MAG = 0.0001
*****Mix slowly, i.e., reduce AMIX and AMIX MAG
**** Reduce MAXMIX, the number of steps stored in the Broyden mixer
(default = 45)
*****Restart from partly converged results
(stop a calculation after say 20 steps and restart from the WAVECAR)
****
Use constraints to stabilize the magetic configuration
**** Pray
附：
各结构压力、体积、焓值如下
Reference Fm3m
899.82 97.80 -29.45703134 OUTCAR_900-01
900.14 97.81 -30.82635860 OUTCAR_900-02
900.03 97.82 -30.82636284 OUTCAR_900-03
900.3 97.83 -30.82652255 OUTCAR_900-04
P63/mmc
801.51 54.55 -13.36070286 OUTCAR_900-01
899.13 52.80 -13.41211260 OUTCAR_900-03
900.27 52.76 -13.41201011 OUTCAR_900-04
902.01 52.72 -13.41212669 OUTCAR_900-02
cmcm
898.48 91.11 -30.19358510 OUTCAR_900-01
900.09 91.10 -30.19350168 OUTCAR_900-02
897.42 91.08 -30.19335751 OUTCAR_900-03
900.13 91.05 -30.19346612 OUTCAR_900-04
I4/mmm
399.59 58.32 -28.02171600 OUTCAR_500-01
499.62 55.93 -28.09219441 OUTCAR_500-02
499.64 55.94 -28.09214407 OUTCAR_500-03
499.75 55.94 -28.09216114 OUTCAR_500-04
499.07 55.94 -24.60062156 OUTCAR_600-01
599.76 53.67 -24.66736113 OUTCAR_600-02
599.72 54.61 -25.54481911 OUTCAR_600-03
600.41 54.36 -25.75783663 OUTCAR_600-04【显著没有收敛？！】
802.09 50.81 -15.78792093 OUTCAR_900-01
899.8 49.35 -15.83013190 OUTCAR_900-02
896.47 49.35 -15.82991320 OUTCAR_900-03
900.46 49.34 -15.82987949 OUTCAR_900-04
更多check
Reference Fm3m
-0.18 152.17 -96.51603492 OUTCAR_0-01
0.16 152.17 -96.51831796 OUTCAR_0-02
0.16 152.17 -96.51831796 OUTCAR_0-03
0.16 152.17 -96.51831796 OUTCAR_0-04
10.03 150.43 -95.57375897 OUTCAR_10-01
10.08 150.42 -95.57057636 OUTCAR_10-02
9.85 150.42 -95.57057566 OUTCAR_10-03
9.89 150.42 -95.57057531 OUTCAR_10-04
29.95 147.26 -93.70988490 OUTCAR_30-01
30.27 145.99 -92.10728795 OUTCAR_30-02
30.18 145.99 -92.10518443 OUTCAR_30-03
30.18 145.99 -92.10518443 OUTCAR_30-04
40.03 144.59 -91.19804959 OUTCAR_40-01
40.08 144.59 -91.19543583 OUTCAR_40-02
40.15 144.58 -91.19543519 OUTCAR_40-03
40.19 144.58 -91.19543456 OUTCAR_40-04
49.69 143.29 -90.29700978 OUTCAR_50-01
49.82 143.28 -90.29453601 OUTCAR_50-02
49.88 143.27 -90.29453665 OUTCAR_50-03
49.94 143.27 -90.29453700 OUTCAR_50-04
99.75 137.35 -85.90960018 OUTCAR_100-02
99.75 137.38 -85.91966193 OUTCAR_100-01
99.97 137.33 -85.90934886 OUTCAR_100-03
100.13 137.31 -85.90935543 OUTCAR_100-04
P63/mmc
-2.5 86.83 -49.68412788 OUTCAR_0-01
0.24 86.51 -49.68401035 OUTCAR_0-02
-0.06 86.54 -49.68391695 OUTCAR_0-03
-0.04 86.54 -49.68399149 OUTCAR_0-04
10.17 85.42 -49.14743841 OUTCAR_10-01
9.86 85.42 -49.14581458 OUTCAR_10-02
9.89 85.42 -49.14583803 OUTCAR_10-03
9.89 85.42 -49.14583803 OUTCAR_10-04
10.01 85.42 -48.61270673 OUTCAR_20-01
20.27 84.33 -48.61447260 OUTCAR_20-02
20 84.34 -48.61449787 OUTCAR_20-03
20 84.34 -48.61449787 OUTCAR_20-04
20.07 84.34 -48.08808145 OUTCAR_30-01
29.19 83.37 -48.08968064 OUTCAR_30-02
28.11 83.42 -48.08980680 OUTCAR_30-03
28.91 83.38 -48.08977812 OUTCAR_30-04
30.16 83.35 -47.56947207 OUTCAR_40-01
40.08 82.37 -47.57111081 OUTCAR_40-02
39.95 82.39 -47.57108153 OUTCAR_40-03
40.05 82.39 -47.57109595 OUTCAR_40-04
40.24 82.39 -47.05684724 OUTCAR_50-01
50.37 81.47 -47.05825923 OUTCAR_50-02
49.94 81.46 -47.05826749 OUTCAR_50-03
49.87 81.46 -47.05830007 OUTCAR_50-04
100.08 77.66 -44.57765305 OUTCAR_100-01
100.03 77.63 -44.57172277 OUTCAR_100-02
100.15 77.63 -44.57167357 OUTCAR_100-03
100.18 77.63 -44.57167355 OUTCAR_100-04
https://blog.sciencenet.cn/blog-567091-887473.html
上一篇：磁性设置再谈：VASP自旋轨道耦合-磁性计算之百家说法
下一篇：vdW、ZPE、SOC vs. Quantum effects
