[lammps-users] A problem in Peridynamics EPS modeling

Dear LAMMPS users,

The version of my LAMMPS is 30Nov2020 and my workstation possesses Ubuntu16.04. I am doing a uniaxial tensile simulation using Peridynamics EPS model and my input file is enclosed for your reference.

PD’s sudden bonding effect would cause a platform phase in the beginning of its stress-strain curve so that I am considering lowering down the loading speed. It is found that when close to the yield point, lower loading speed (v=0.03, units lattice) will lead the model to a strange out-of-plane vibration and then explodes, while higher loading speed (v=0.3, units lattice) will not. I am wondering why such circumstance happens and what am I supposed to do to circumvent this problem.

Input File:
#=================Give file name here=============#
variable filename string pd01

#=============Initialization======================#
dimension 3
log ${filename}.log
units metal
boundary s s s
atom_style peri
atom_modify map array
neighbor 4.046 bin

#=================Atomic Structure====================#

create box

lattice sc 4.046
region box block -15 34 -15 34 -4 4 units lattice
create_box 1 box

create atoms

create_atoms 1 region box

#===================atom property=====================#

Set mass of atoms

set group all density 2.7
set group all volume 66.430125

#=================Atomic Potential====================#
pair_style peri/eps
pair_coeff * * 0.4393375 0.175 12.25 0.8 0.1 0.125

#=================region and group====================#
region right block 33 INF INF INF INF INF units lattice
region top block INF INF 33 INF INF INF units lattice
region bottom block INF INF INF -14 INF INF units lattice