during MD simulation, the atoms are not moving as per required. The problem (as per my observation) lies in energy minimization technique. upon application of tensile strain in x or y direction, the atoms start to dislocate in z direction.
Can you say something more about your simulations? Do you use a periodic boundary or not? Is it a crystal or glass structure? Did you use the nvt or npt ensemble?
Przemyslaw Dziegielewski sir, I have used periodic boundary conditions in all three directions, and been trying to simulate a crack propagation in aluminium alloy crystal structure. I tried using npt ensemble and also cg energy minimization. I have also tried simulating pure metal plates with central cracks. in all cases upon application of strain in y direction or velocity, the crack doesnt propagate after few steps, instead it colses and atoms pile up in z direction.
Vladimir S. Myasnichenko sir, I have done the basic example given in LAMMPS package. That works fine, but the problem is that i need to implement EAM potentials instead of L-J potentials since my material is a metal.
Subham Prasad Let's try to manipulate the timestep/number of timesteps. I am not sure that you can see the crack if the simulation time is too short. Moreover, if you use npt and 3d periodic boundary, I afraid that force goes to the change of the box shape and doesn't produce the crack. I recommend this work Article Molecular dynamics simulation of tensile deformation of nano...
with not MEAM but Morse potential, but it is not the most important thing.
Przemyslaw Dziegielewski , the timestep that I tried ranged from 0.001-0.005 i.e. 1fs to 5 fs. On further increasing the timestep gave bad termination of program. Also i have used nvt with 3d periodic conditions, nothing worked.
The article that you mentioned has also used similar timestep (1fs). So, it seems like there isnt any problem with timesteps. Regarding boundary conditions, could you please suggest what conditions should I use to get fair results.