Hello everyone
Recently, i want to do bending and torsion simulation for one nano-composite materials, but , unfortunately, i cannot find a good way to define the boundary of modeling. So i want to ask someone for help:
Here is my questions:
-
for bulk materials, i need apply the period boundary for xyz direction,but, if i use non-period boundary on one direction, is that means my bulk model changing to like beam model?
-
the bending experiment, we can see that there always have two clamps on each side, then applied the force on the middle of model, therefore, it leads to bending appearance. So, which command can be used for this situation?
-
If possible, could you show some example for these situation?
Thankyou so much
Cong
Hello everyone
Recently, i want to do bending and torsion simulation for one
nano-composite materials, but , unfortunately, i cannot find a good way to
define the boundary of modeling. So i want to ask someone for help:
Here is my questions:
1. for bulk materials, i need apply the period boundary for xyz
direction,but, if i use non-period boundary on one direction, is that means
my bulk model changing to like beam model?
well, you can easily rationalize with a back-of-the-envelope
consideration, that either operation cannot be handled with periodic
boundary conditions: consider a very large non-periodic system and
decompose it into sub-domains of the size that you would like to simulate
with PBC. it should be obvious that each sub-domain will experience
different strain, particularly those close to the surface. you can also
argue the other way around: neither torsion nor bending motions are
possible at all with an infinitely large system as this is what is
represented by a PBC system.
2. the bending experiment, we can see that there always have two clamps on
each side, then applied the force on the middle of model, therefore, it
leads to bending appearance. So, which command can be used for this
situation?
having a look at the fix setforce and fix addforce might help. another
option to immobilize atoms is to do time integration only on the mobile
atoms.
3. If possible, could you show some example for these situation?
have you looked at relevant publications? they would have to have a
detailed description of how the model was constructed.
axel.