MD melting problem

[Matt W]

Your input file has

STRESS_TENSOR NUMERICAL

this is probably very expensive. You don't need the full stress tensor for 
NPT_I, and analytical should be faster and more accurate than numerical (if 
available for your forcefield)


On Thursday, May 3, 2018 at 1:18:49 PM UTC+1, jie wu wrote:
>
> Hello, Matt, thanks for the answer, but I am a little confused. Should I 
> use the STRESS_TENSOR DIAGONAL_ANALYTICAL instead of the MM force field or 
> to analysis the stress changing during the simulation?
>
> 在 2018年5月3日星期四 UTC+8下午6:04:44，Matt W写道：
>>
>> Probably unrelated to you actual question, but changing to
>>
>> STRESS_TENSOR DIAGONAL_ANALYTICAL
>>
>> should give you a big speed up.
>>
>> Matt
>>
>> On Thursday, May 3, 2018 at 1:37:55 AM UTC+1, 498... at qq.com wrote:
>>>
>>> Hello, everyone, I want to simulate the melting structure of molten 
>>> salts. The simulation began with  crystal structure and temperature was set 
>>> at 3000 K in NPT ensemble， which is far more higher than the melting point. 
>>> However, the result is frustrating. After 100 ps runs, the ions still in 
>>> crystal structure. The parameters in the force field is in published paper 
>>> and widely used, so I wonder somewhere in the input file is wrong. The file 
>>> is listed below, could anyone tell me the possible problems? Thanks very 
>>> much.
>>>
>>>
>>> &GLOBAL                         ! Section with general information 
>>> regarding which kind of simulation to perform an parameters for the whole 
>>> PROGRAM
>>>   PROJECT C6                    ! Name of the project. This word will 
>>> appear as part of a name of all ouput files (except main ouput file, 
>>> specified with -o option)
>>>   RUN_TYPE MD           ! Molecular Dynamics
>>>   PRINT_LEVEL low
>>> &END GLOBAL
>>>
>>>
>>> &FORCE_EVAL ! This section defines method for calculating energy and 
>>> forces
>>>   METHOD FIST ! Using Molecular Mechanics
>>>   STRESS_TENSOR NUMERICAL
>>>   &MM
>>>     &FORCEFIELD ! This section specifies forcefield parameters
>>>       ..........
>>>     &END FORCEFIELD
>>>     &POISSON ! This section specifies parameters for the Poisson solver
>>>       &EWALD ! This section specifies parameters for the EWALD 
>>> summation method (for the electrostatics)
>>>         EWALD_TYPE none ! Standard real-space coulomb potential is 
>>> computed together with the non-bonded contributions.
>>>       &END EWALD
>>>     &END POISSON
>>> &NEIGHBOR_LISTS
>>>   GEO_CHECK TRUE
>>>   NEIGHBOR_LISTS_FROM_SCRATCH TRUE
>>>   VERLET_SKIN 1
>>>     &END NEIGHBOR_LISTS 
>>>   &END MM
>>>   &SUBSYS ! This section defines the system
>>>     &CELL ! Unit cell set up
>>>       ABC 22.56 22.56 22.56 ! Lengths of the cell vectors A, B, and C
>>>       ALPHA_BETA_GAMMA 90.0 90.0 90.0
>>>       PERIODIC  XYZ
>>>     &END CELL
>>>     &TOPOLOGY 
>>>      COORD_FILE_NAME  NaCl.xyz 
>>>      COORDINATE xyz ! The type of the file  is XYZ
>>>     &END TOPOLOGY
>>>   &END SUBSYS
>>> &END FORCE_EVAL
>>>
>>> &MOTION ! This section defines a set of tool connected with the motion 
>>> of the nuclei.
>>>  &PRINT
>>>  &TRAJECTORY
>>>     LOG_PRINT_KEY T
>>>     FORMAT XMOL
>>>     &EACH
>>>      MD 10000
>>>     &END EACH
>>>     ADD_LAST NUMERIC   
>>>  &END TRAJECTORY
>>>  &END PRINT
>>>  &MD 
>>>     ENSEMBLE NPT_I
>>>     TEMPERATURE 3000
>>> &BAROSTAT
>>>   PRESSURE 0
>>>   TEMPERATURE 3000
>>> &END BAROSTAT
>>> &THERMOSTAT
>>>         TYPE NOSE
>>>          REGION GLOBAL
>>>          &NOSE
>>>            TIMECON 100
>>>            LENGTH 3
>>>            MTS 2
>>>          &END NOSE
>>>      &END THERMOSTAT
>>> TIMESTEP 1.0
>>> STEPS 100000
>>>  &END MD
>>>
>>>   &PRINT
>>>     &RESTART
>>>       LOG_PRINT_KEY T
>>>       &EACH
>>>         MD 10000
>>>       &END EACH
>>>       ADD_LAST NUMERIC
>>>     &END RESTART
>>>   &END PRINT
>>> &END MOTION
>>>  &END
>>> &END MOTION
>>>
>>>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20180503/2bd9e13b/attachment.htm>