Dear All,
I am using LAMMPS (14 Feb 2014) and have some questions about shock simulation:
1.To generate planar shock waves in NEMD simulations via lammps, we usually use fix wall/piston command to drive a piston at a specific particle velocity onto a stationary simulation cell or impact a simulation cell onto a piston at a specified particle velocity.As we know, ‘fix wall/piston’ constructs an imaginary wall without thickness. However,some papers display another way. They construct a rigid piston wall with certain thickness.
1.Now,I wonder how they construct such a rigid wall via lammps?
2.Can these rigid piston walls have the same results as the imaginary wall via ‘fix wall/piston’ when driving a piston wall onto a stationary simulation cell or impacting a simulation cell onto a piston wall?
3.How do these atoms in the rigid wall interact with the atoms not in the rigid wall? Should we defind another internation between the atoms in the rigid wall and the atoms not in the rigid wall?Or still use the same potential as the initial setting which is used in the non-rigid region?
Dear All,
I am using LAMMPS (14 Feb 2014) and have some questions about shock simulation:
1.To generate planar shock waves in NEMD simulations via lammps, we usually use fix wall/piston command to drive a piston at a specific particle velocity onto a stationary simulation cell or impact a simulation cell onto a piston at a specified particle velocity.As we know, ‘fix wall/piston’ constructs an imaginary wall without thickness. However,some papers display another way. They construct a rigid piston wall with certain thickness.
1.Now,I wonder how they construct such a rigid wall via lamps?
Quite straightforward. Just layers of atoms without being time integrated.
2.Can these rigid piston walls have the same results as the imaginary wall via ‘fix wall/piston’ when driving a piston wall onto a stationary simulation cell or impacting a simulation cell onto a piston wall?
Why not test it yourself?
3.How do these atoms in the rigid wall interact with the atoms not in the rigid wall? Should we defind another internation between the atoms in the rigid wall and the atoms not in the rigid wall?Or still use the same potential as the initial setting which is used in the non-rigid region?
Same would be fine, if the repulsive core of the potential is well defined. If not, you can define a LJ repulsion between the wall and the atoms.
It is less important to think about the behaviour close to any piston; such effects are transient and relate to the setting up of the system state. This is why methods such as “fix msst” have been developed.
Alternatively run long (spatial in shock direction) calculations. Then watch as the shock wave evolves and develops.