About whether the addforce command can be canceled

LAMMPS
1

I have some doubts about the fixaddforce command. After I applied a constant force of 4ns to the atom, I canceled the addforce to carry out the statistical results without speed. Applying 2ns to the atom, there is a speed to count without canceling the force. Why does the same force 4ns have different results? Periodic boundary

#------1----------------------------------------------------------------------#
fix NVE2 h2o nve
#----------------------------------------------------------------------------#
fix add h2o addforce 0.001 0 0 region add1
run 4000_ps
unfix add
#----------------------------------------------------------------------------#
compute cc1 h2o chunk/atom bin/2d z 0.0 3.923 x 0.0 3.923 units box
fix ave1 h2o ave/chunk 10 100000 1000000 cc1 vx norm sample file vx.txt
run 2000_ps
unfix NVE2

In this case, there is no normal speed in the x direction.

#--------2--------------------------------------------------------------------#
fix NVE2 h2o nve
fix add h2o addforce 0.001 0 0 region add1
run 2000_ps
#----------------------------------------------------------------------------#
compute cc1 h2o chunk/atom bin/2d z 0.0 3.923 x 0.0 3.923 units box
fix ave1 h2o ave/chunk 10 100000 1000000 cc1 vx norm sample file vx.txt
run 2000_ps
unfix NVE2
#In this case, normal speed can be obtained.

2

Please take a look at the rules here and format your post properly. Both your question and script are almost impossible to read.

This, in particular, makes no sense.

Also, I don’t see any unfixing of the addforce command in the script you copy-pasted, so I really can’t make any sense of what you are asking.

3

This input cannot run, because this is incorrect syntax.

Which atom where?

Which statistical results of what and related to what speed?

This sentence makes not sense grammatically.

This is not even a sentence.

@chenmo If you want help, we first need to understand what your problem is. This cannot work without providing sufficient context. There is no context at all in your post. Either you improve your questions and provide sufficient context in whole sentences that make sense to people that know nothing about you and your research, or you need to ask in a different place.

6

inmodel22555.lmp (7.5 KB)

I’m really sorry. I uploaded my in file. My doubt is that when I add the command “unfix”, the calculated speed is incorrect. It’s right not to add this command.
In order to effectively simulate convective heat transfer, the thermal pump method is introduced to the fluid domain. The thermal pump method divides the fluid domain into three separate regions: a forcing region, a temperature reset region, and a data collection region.

7

Your post is not really helping. It looks like some randomly copies sentences and neither explains what you are doing in a way accessible to an outsider nor really explains what your concern is and why you are worrying about that specific part, nor explains what kind of speed you are computing and how you would determine what is “correct” or not. For as long as you are unwilling (or unable?) to explain this properly, you are probably best helped by discussing with someone that is familiar with your research and has a vested interest in your work. Please also note that this forum is about discussing LAMMPS and not whether turning a feature of LAMMPS needs to be turned on or of for your research. In other words, we would discuss how to turn fix addforce on or off, but not whether it is wise to do so for your application.