Hi Zhongqiao,<br><br>I think it would work notwithstanding it's quite tricky to implement...<br><br>My current task is to test a sieving membrane. So some external action is required to force the mixture go towards the place where separation occurs. In papers, a rigid wall moving with constant velocity is sometimes used for this. If one just accelerates the mixture particles with acc-grps, it seems not realistically as for me, so I'm in guesses what's the best way to manage such tasks with gromacs.<br>
<br>Thank you for the useful remarks,<br>Vitaly<br><br><br>>I am not sure if it works. You can try to put your system into two<br>
>frozen layers which will exert a resisting force on the wall. You still<br>
>need to put an acceleration on the wall. But after a steady state is<br>
>reached, you will see the wall will move approximately at a constant<br>
>velocity.<br>
<br>
>Hi and thank you! This cleared the situation. I didn't care about<br>
comm-mode but only set comm-grps...<br>
<br>
>And let me one more question. Is there any possibility to define a<br>
constant velocity of my rigid wall, not an acceleration?<br>
<br>