Please keep the reply in the list, thanks.
thanks, but my question is just about system falling apart.
I successfully run my simulation with real wall molecules but i can not
I ran your script, with corrections you mentioned, and it still fell
apart without the fix wall/region (also removed addforce) command. It
seems your structure is just not stable.
model it with "wall model". because i have a wall with sharp corners and
atoms fell very large amount of force near these corners and scape from the
simulation and will lost.
An alternative (that may or may not work) to what you are trying to do
is to pre-process a data file to be read in by read_data. You can
easily write a program to carve a 3d crystal into a shape you want.
Group the whole structure into one single region, then apply fix
wall/region to that region. In that case, the wall will only exist on
the surface of the geometric region.