信手请教关于Gaussian 优化的问题

43 views
Skip to first unread message

sophier

unread,
Sep 4, 2008, 1:32:29 AM9/4/08
to Gaussian
请教一个关于优化的问题:
我在计算分子结构优化时,先用Gaussiview 生成*.com 文件,然后进行结构优化,
文件如下:
%chk=bmimo.chk
%mem=400MW
%nprocs=2
# opt hf/3-21g

bmimo+

1 1
C -6.77387084 2.84359468 0.90395800
C -5.23814584 2.84359468 0.90395800
.....................(省略)
.......................(省略)

--Link1--

%chk=bmimo.chk
%mem=400MW
%nprocs=2
# hf/6-31g* opt test guess=read geom=check

bmimo+

1 1

--Link1--

%chk=bmimo.chk
%mem=400MW
%nprocs=2
# mp2/6-31g* opt=tight test guess=read geom=check

bmimo+

1 1

--Link1--

%chk=bmimo.chk
%mem=400MW
%nprocs=2
# mp2/6-31g* test guess=read geom=check pop=mk iop(6/33=2)
density=current

bmimo+

1 1

当运行到第三个link(# mp2/6-31g* opt=tight test guess=read geom=check)时,运算结构不收
敛,重新启动运行多次还是不收敛
Item Value Threshold Converged?
Maximum Force 0.000000 0.000015 YES
RMS Force 0.000000 0.000010 YES
Maximum Displacement 0.001495 0.000060 NO
RMS Displacement 0.000256 0.000040 NO
Predicted change in Energy=-6.057542D-10)
,请教是什么问题??????

请问我可不可以修改收敛准则??或者简单的将上面的opt=tight 改为opt??????

最近被这个问题搞得头昏脑胀,希望各位高手指点!!!!3x


shenyun

unread,
Sep 4, 2008, 3:14:19 AM9/4/08
to Gaus...@googlegroups.com
Dear sophier:
       程序最后在哪一个程序上停止了啊?L502?还是L9999啊?还是其他的。。。
一般如果是L9999的话可以用增加循环次数就可以收敛了,如果是L502的话可以试试加大SCF的循环次数看看,如果不行你在看看你的分子结构是不是可以再调整一下。我觉得收敛准则没有必要改吧。


我也接触高斯不久。

祝好!


2008/9/4 sophier <sophi...@gmail.com>



--
Sincerely,
ShenYun
沈赟,
The Department of Chemistry
Xiamen University, Fujian, P.R.China

Jianming Wu

unread,
Sep 4, 2008, 4:14:14 AM9/4/08
to Gaus...@googlegroups.com
试试
# mp2/6-31g* opt(int=ultrafine) scf=tight iop(5/13=1) guess=read
geom=checkpoint

祝你好运!

sophier

unread,
Sep 4, 2008, 1:23:07 PM9/4/08
to Gaussian
Thanks so much!!!!
this problem has been solved. I use a long time, then it converges.

sophier

unread,
Sep 4, 2008, 1:26:41 PM9/4/08
to Gaussian
thank you so much! I have a new problem. my program stops after it
finish the third link. but for som synatx error, it doesn't continue
do the
las link # mp2/6-31g* test guess=read geom=check pop=mk iop(6/33=2)
density=current .
Do you know how can I restart the program from the last link?
Thanks !!!!

On Sep 4, 1:14 am, shenyun <antibond...@gmail.com> wrote:
> Dear sophier:
> 程序最后在哪一个程序上停止了啊?L502?还是L9999啊?还是其他的。。。
> 一般如果是L9999的话可以用增加循环次数就可以收敛了,如果是L502的话可以试试加大SCF的循环次数看看,如果不行你在看看你的分子结构是不是可以再调整一下。我觉得收敛准则没有必要改吧。
>
> http://59.77.33.35/cgi-bin/leobbs.cgi
> 这个是厦门大学理论化学研究中心的论坛,有问题可以去里面提问的。 <http://59.77.33.35/cgi-bin/leobbs.cgi>
>
> 我也接触高斯不久。
>
> 祝好!
>
> 2008/9/4 sophier <sophier1...@gmail.com>

sophier

unread,
Sep 4, 2008, 1:27:09 PM9/4/08
to Gaussian
Thanks!!!
Reply all
Reply to author
Forward
0 new messages