Dear LAMMPS users,
I am working on a flow problem which requires to create fix no. of particle in the simulation domain every time step for whole simulation run. The particle insertion pattern should be same as fix deposit which insert single particle in some time step. as you can see my problem is linked to fix deposit but not same. Please help me on this .
If I understand your problem correctly, the quick around to do this using fix deposit might be as follows.
Lets say if you want o deposit n atoms at once.
so, use fix deposit to deposit 1 atom at each timestep for n steps. When you do run for n steps. Don’t use any fix which update the velocities or coordinates. Once you are done, reset your timestep to the original timestep unfix that fix, run it for x (x could be 1) steps as normal. Create a loop do to this as long as you want. This is probably not the ideal solution but might work.
Regards,
Vikas
Please explain why you cannot use fix deposit? From your description it sounds to me as if you could.
Axel.
Hello sir,
According to lammps manual fix deposit insert single particle in domain every N time steps , so with this i can insert at the max one particle /time step . My requirement is to insert say M no. of particles every time step in the simulation domain, which is probably not possible with fix deposit. Please suggest me a solution or correct me if i understood fix deposit statement wrong.
Thank you,
Abhishek Verma
Hello sir,
According to lammps manual fix deposit insert single
particle in domain every N time steps , so with this i can insert at the max
one particle /time step . My requirement is to insert say M no. of
particles every time step in the simulation domain, which is probably not
possible with fix deposit. Please suggest me a solution or correct me if i
understood fix deposit statement wrong.
before discussing how to directly implement your requirement, do you
have proof that inserting a single particle in every time step is not
sufficient. that would be an awful lot of particles already.
axel.
Hello sir,
Yes , i have calculated the no. of particles insertion required. It will be around 5-10 varying with the other simulation parameters. basically, i am not simulating in periodic boundary frame so i have to maintain flux in one direction by generating atoms in the domain. I am thankful for your support.
Abhishek Verma
This is definitely an unusual way to run an MD simulation. Be that as
it may, a simple way to accomplish your goal is to use 5-10 fix
deposit commands. Another way is to reduce the timestep by a factor
of 5-10. Neither of these is particularly efficient, but I suspect
that will be the least of your worries.