计算化学公社
标题: gamess不收敛 [打印本页]
作者Author: 413 时间: 2022-11-1 10:29
标题: gamess不收敛
本帖最后由 413 于 2022-11-1 10:34 编辑
请问下面这个错误咋解决呢？我搜了gamess forum上，那个有个建议，但是没解决呢
我还是了guess=hcore，这个错误是避开了，但是200圈的时候scf没收敛
“gamess forum的建议http://classic.chem.msu.su/cgi-b ... HW-8870-1309+00.htm
As a workaround, try to use: $SCF DIRSCF=.TRUE. FDIFF=.FALSE. NCONV=6 EXTRAP=.TRUE. DAMP=.TRUE. SHIFT=.TRUE. RSTRCT=.FALSE. DIIS=.TRUE. SOSCF=.FALSE. $END”
* * * * * WARNING * * * * *
POSSIBLE LINEAR DEPENDENCE PROBLEMS DETECTED, INPUT QMTTOL= 1.0E-06,
WITH 7 OVERLAP EIGENVALUES LYING BETWEEN 1.0E-05 AND QMTTOL.
IF YOU EXPERIENCE SCF CONVERGENCE DIFFICULTIES WITH THE DENSITY ALMOST BUT
NOT QUITE CONVERGED, YOU MAY NEED TO LOOSEN -QMTTOL- BY A FACTOR 5 TO 10.
**** ERROR, ORBITAL 1 IS MISSING FROM REORDERING INSTRUCTIONS
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
SOMETIMES THE ABOVE ERROR IS CASED BY AN INABILITY TO
PROJECT THE HUCKEL GUESS ONTO THE CURRENT BASIS SET,
E.G. WHEN THE 1S HUCKEL ORBITAL OF A H ATOM CANNOT
BE PROJECTED ONTO A PURELY P BASIS SET BECAUSE YOU
ARE TRYING TO DO EXCITED STATE H ATOMS. IN SUCH A
CASE, PLEASE USE GUESS=HCORE OR GUESS=MOREAD INSTEAD.
EXECUTION OF GAMESS TERMINATED -ABNORMALLY- AT 10:23:14 01-NOV-2022
作者Author: zjxitcc 时间: 2022-11-1 11:22
本帖最后由 zjxitcc 于 2022-11-1 11:24 编辑
如果你是做常规HF/DFT计算，推荐使用MOKIT中的fch2inp小程序从Gaussian向GAMESS传轨道；
如果你是做EDA计算，推荐使用GKS-EDA方法，并且使用frag_guess_wfn小程序产生带片段和整体轨道信息的.inp文件。建议阅读《GKS-EDA计算简介》
不建议尝试所列举的英文建议。GAMESS的SCF收敛性远不如Gaussian，所以不建议在GAMESS中换关键词试来试去（关键词加多了 自己也没明白是啥意思），而且即使好不容易收敛了 也没法检验波函数稳定性。
使用上述两种技巧，论坛上几乎所有GAMESS单点SCF不收敛、xxx-EDA过程中SCF不收敛的帖子都可得到解决，彻底终结此类问题。
作者Author: 413 时间: 2022-11-1 12:09
谢谢提醒哈
其实我就是用的fch2inp传的轨道，后来自己手动改了命令行，把MOREAD抹去了，结果搞的不收敛
现在已经重新提交上去了，谢谢谢谢。有了结果我会及时反馈
作者Author: 413 时间: 2022-11-1 21:21
收敛了，加上了moread这个关键词
| 欢迎光临 计算化学公社 (http://bbs.keinsci.com/) |
Powered by Discuz! X3.3 |