Simulation of rolling process of polycrystalline Fe at room temperature

Dear lammps developer,
Recently, I wanted to do a simulation of the rolling process of polycrystalline Fe at room temperature, but I still couldn’t solve the problem that the atomic system temperature is more than 100000 k after I wrote the in file,The contents of the in file of my rolling process simulation are as follows:
##定义变量##
variable a equal 1.0 ##定义原子模型截断半径
variable T_start equal 300
variable T_end equal 300
##模拟环境初始化##
units metal
atom_style atomic
boundary s p s
neighbor ${a} bin
neigh_modify delay 0 every 1 check yes

##工件和轧辊的组合原子模型

read_data Fe-C.data(2)

###对刚性滚子和被压工件进行划分区域
region rollera cylinder y 140.04 -69.016 69.984 EDGE EDGE units box
region rollerb cylinder y 140.04 62.996 14.004 EDGE EDGE units box
region rollerc cylinder y 349.992 61.484 14.004 EDGE EDGE units box
region rollerd cylinder y 349.992 -67.504 69.984 EDGE EDGE units box
region gongjian block 0 100 0 66.7 0 50 units box
##对刚性棍子和被压工件进行分组
group rollera region rollera
group rollerb region rollerb
group rollerc region rollerc
group rollerd region rollerd
group gongjian region gongjian

##定义轧辊和工件之间的相互作用势##
pair_style hybrid eam/fs morse 4.5
pair_coeff * * eam/fs Fe_Earth_core.eam.fs Fe NULL
pair_coeff 1 2 morse 0.72140 1.59380 2.49010 4.0
pair_coeff 2 2 morse 0.38540 1.27180 2.81780 4.0

##对各组尽心固定
#fix rollera rollera setforce 0.0 0.0 0.0
#fix rollerb rollerb setforce 0.0 0.0 0.0
#fix rollerc rollerc setforce 0.0 0.0 0.0
#fix rollerd rollerd setforce 0.0 0.0 0.0

compute myTemp gongjian temp/partial 0 0 1 # 在计算温度的时候不考虑x轴,y轴的速度
#compute myTemp gongjian temp/com

#输出原子热力学信息和位置
thermo 100
thermo_style custom step lx ly lz press pxx pyy pzz pe c_myTemp temp vol dt time
dump 1 all custom 200 Fe-C.xyz id type x y z
##thermo_modify temp myTemp

##在NVE系综下对轧辊和工件的组合模型进行保温500ps
fix 2 all nvt temp {T_start} {T_end} 0.05
fix_modify 2 temp myTemp # 将控温对象改为temp2

##对各组尽心固定
#fix rollera rollera setforce 0.0 0.0 0.0
#fix rollerb rollerb setforce 0.0 0.0 0.0
#fix rollerc rollerc setforce 0.0 0.0 0.0
#fix rollerd rollerd setforce 0.0 0.0 0.0

##对刚体组的速度进行固定
velocity rollera set 0.0 0.0 0.0
velocity rollerb set 0.0 0.0 0.0
velocity rollerc set 0.0 0.0 0.0
velocity rollerd set 0.0 0.0 0.0

##让刚体进行旋转
fix 3 rollera move rotate 140.04 0.0 -69.016 0 1 0 1 units box
fix 4 rollerb move rotate 140.04 0.0 62.996 0 -1 0 1 units box
fix 5 rollerc move rotate 349.992 0.0 61.484 0 -1 0 1 units box
fix 6 rollerd move rotate 349.992 0.0 -67.504 0 1 0 1 units box
##fix 7 gongjian move linear 6.0 0 0 units box
velocity gongjian set 6.0 NULL NULL sum yes units box

##设置运行时间
timestep 0.002
run 50000

###输出原子模型
write_data Fe-C.data(3)
I still don’t know how to solve this problem?
Xue bao shuai

Hi Xue,

A large temperature could be consequence of several issues, and it is difficult to help you without your data file. On the top of my mind, some of the most probable causes are :
_ the timestep is too large (I am no expert of eam potential but 2 fs seems like a large value)
_ initial overlap between your atoms causing crazy acceleration (you can make sure of this one by looking at your trajectory with VMD)
_ the potential is wrong (I would double check all the values)
_ the dynamics is wrong (for example using two fixes that update the atom position, but in your case this is not the issue)

Best luck,
Simon

Btw you should post your LAMMPS questions here : LAMMPS - Materials Science Community Discourse