计算化学公社

标题: ORCA寻找过渡态报错 [打印本页]

作者
Author: 闲尾Z 时间: 2021-5-13 10:40
标题: ORCA寻找过渡态报错
问题描述：用ORCA寻找过渡态（虚频）报错，请问各位老师应该如何解决，下面是out文件中的报错内容预计输入文件的信息
报错内容
Nuclear repulsion Hessian                      ... done (    0.0 sec)

----------------------------------------------
Forming right-hand sides of CP-SCF equations    ...
----------------------------------------------
One electron integral derivatives             ... done (  17605.2 sec)
Transforming the overlap derivative matrices    ... Adding tr(S(x)F(y) C-PCM term to the hessian    ...
job aborted:
[ranks] message

[0-39] process exited without calling finalize

---- error analysis -----

[0-39] on YHT-PC
D:\orca\orca\orca\orca_scfhess_mpi.exe ended prematurely and may have crashed. exit code 64

---- error analysis -----

ORCA finished by error termination in SCF Hessian
Calling Command: mpiexec -np 40  D:\orca\orca\orca\orca_scfhess_mpi.exe a1.scfhess.inp a1
[file orca_tools/qcmsg.cpp, line 458]:
  .... aborting the run

输入文件
!  BP def2-tzvp  TightSCF grid5 SlowConv  optTS freq
%Maxcore 3072
%pal  nprocs 40
end

%geom calc_hess true
optimizeHydrogens true # do not constrain any hydrogen position
end

!CPCM(water)

%scf
maxiter 500
end

服务器是164g,48核的

作者
Author: ms860309 时间: 2021-5-13 11:00
本帖最后由 ms860309 于 2021-5-13 11:02 编辑

CP-SCF要算很久可能是這個報錯記憶體不足
另外是我覺得freq ts 這部分不需要到tzvp
所以我覺得可以用 def2-svp
另外這個有支援rijcosx?  如果有可以考慮加 gridx4
!BP def2-svp tightscf grid4 optTS freq CPCM(water)
%geom
Calc_Hess true # Calculate Hessian in the beginning
MaxStep 0.08  # 0.08 a.u.
Convergence tight  # Same convergence criteria with tightopt
end
%scf
maxiter 300  # SCF interations.
ConvForced 1 # Make sure the scf is converge during calculation.
end

%pal
nprocs 40
end

%maxcore 5000 # memory per core

%method
Z_MaxIter 50 # The CPSCF interations while calculate actual hessian
end

作者
Author: sobereva 时间: 2021-5-13 18:49
换linux版再试，或者少用点核再试

欢迎光临计算化学公社 (http://bbs.keinsci.com/)