Dear LAMMPS users and developers,
I want a group of atoms to move linearly regardless of its interaction with other atoms.
When I use fix move linear, it still may rotate under impact of other atoms.
To prevent it, I have to use fix move transrot with a huge period.
How to set period to infinity? Should I just keep using a huge number?
Is there other way to prohibit rotation while setting fixed linear move ?
Hello,
Have you you tried to use:
fix myfix mygroup setforce 0 0 0 # erase all forces applied on the atoms from mygroup
velocity mygroup set 0 1 0 # set the velocity of the atoms from mygroup
Simon
Please provide a reproducer input example for your claim. There is no reason why applying fix move linear should induce a rotation. Fix move moves atoms with complete disregard of the forces on them. So there must be an error in your input. Please check the log for warnings.
I am on Ubuntu 22.04.4 LTS, LAMMPS (2 Aug 2023 - Update 2) running lmp -in in.script
in.script:
units lj
boundary f f f
comm_modify vel yes
atom_style body rounded/polyhedron 1 8
atom_modify map array
region reg block 0 60 0 60 0 30 units box
create_box 4 reg
read_data data.2polyhedrons add append
variable cut_inner equal 0
variable k_n equal 1000
variable k_na equal 0
variable c_n equal 10
variable c_t equal 10
variable mu equal 0
variable A_ua equal 1
pair_style body/rounded/polyhedron ${c_n} ${c_t} ${mu} ${A_ua} ${cut_inner}
pair_coeff * * ${k_n} ${k_na}
neighbor 0.5 bin
neigh_modify every 1 delay 0 check yes
timestep 0.001
group g1 id 1
group g2 id 2
fix g1l g1 move linear 0 1 0
fix g all gravity 1.0 vector 0 0 -1
fix 1 all nve/body
thermo_style custom step atoms ke pe etotal press
thermo 1000
thermo_modify lost ignore flush yes
dump 2 all image 500 *.gif type type &
zoom 1.5 adiam 1.5 body type 0 0 view 75 15
dump_modify 2 pad 6
run 10000
with
data.2polyhedrons:
Two polyhedrons
2 atoms
2 bodies
1 atom types
0 50 xlo xhi
0 50 ylo yhi
0 10 zlo zhi
Atoms
1 1 1 1 30 30 5
2 1 1 1 35 30 10
Velocities
1 0 0 0 0 0 0
2 0 0 0 0 0 0
Bodies
1 3 79
8 12 6
1 1 1 0 0 0
4.75 4.75 0.25
4.75 -4.75 0.25
-4.75 -4.75 0.25
-4.75 4.75 0.25
4.75 4.75 -0.25
4.75 -4.75 -0.25
-4.75 -4.75 -0.25
-4.75 4.75 -0.25
0 1
1 2
2 3
3 0
4 5
5 6
6 7
7 4
0 4
1 5
2 6
3 7
0 1 2 3
4 5 6 7
0 1 5 4
1 2 6 5
2 3 7 6
3 0 4 7
0.5
2 3 79
8 12 6
1 1 1 0 0 0
0.75 0.75 0.75
0.75 -0.75 0.75
-0.75 -0.75 0.75
-0.75 0.75 0.75
0.75 0.75 -0.75
0.75 -0.75 -0.75
-0.75 -0.75 -0.75
-0.75 0.75 -0.75
0 1
1 2
2 3
3 0
4 5
5 6
6 7
7 4
0 4
1 5
2 6
3 7
0 1 2 3
4 5 6 7
0 1 5 4
1 2 6 5
2 3 7 6
3 0 4 7
0.5
yields
I would expect the small cube bounce off the plate, and the plate keep moving without rotation. That is not what actually happening.
Thank you, Simon. I’ve tried it, but the result was the same: collision causes rotation (please, see the animation above).
Don’t include the particle from g1 group in the fix nve/body group. With the setup you have now, this particle is first moved by fix move and then its trajectory is further modified by fix nve/body.
The simulation is executed correctly. If you want polyhedron 1 to be unaffected by its interaction with id 2, you need to remove their interactions with
neigh_modify exclude group g1 g2
No.
That will make one object “pass through” the other, not bounce off it.
To me, this can only happen if the two particles do not interact. I missed the part about them bouncing 
This (unphysical) situation is also indicated by a warning in the output:
WARNING: One or more atoms are time integrated more than once (src/modify.cpp:289)
LAMMPS warnings should not be ignored unless you know exactly what you are doing.
LAMMPS does not support writing out images in GIF format. Those files will be in Netpbm “raw” format:
$ file *.gif
000000.gif: Netpbm image data, size = 512 x 512, rawbits, pixmap
000500.gif: Netpbm image data, size = 512 x 512, rawbits, pixmap
001000.gif: Netpbm image data, size = 512 x 512, rawbits, pixmap
001500.gif: Netpbm image data, size = 512 x 512, rawbits, pixmap
002000.gif: Netpbm image data, size = 512 x 512, rawbits, pixmap
I’ve changed the script to
group g1 id 1 # moving linear
group g2 id 2 # falling
fix 1 g2 nve/body
fix g1l g1 move linear 0 1 0
as @mkanski and @akohlmey suggested.
There are no warnings now, but the g1 plane is still rotating after collision.
[False. Works fine, see my post below]
Not for me.
Sorry, yes, works fine now, looked at the old images.
Thank you very much!