[lammps-users] Need suggestion regarding shear simulations


Continuous shear usually use something like

fix 1 all deform 1 xz erate 1e-3 remap v
fix 2 all nvt/sllod 1.0 1.0 10.0

in LJ units, or

fix 2 all nvt 300.0 300.0 100.0

for real units. Remapping velocity allows periodic mapping of the flow profile.


Dear Peter, Steve,

I tried what you suggested for shearing my box but it seems that it doesn’t like to shear an orthagonal box…
What I tried is as follows.

fix 1 all deform 1 xy erate 0.001 remap v units box
fix 2 all nvt 300.0 300.0 100.0

The error I got was
Error: Fix deform tilt factors require triclinic box

Steve, displace_box command (to deform or shear) also gives same error.

Am I missing something here? Any help will be appreciated.
Is the above command applicable for shearing the box along xy plane?

Thanks and regards,

I haven’t tried recently but as far as i remember earlier in a similar case i added the three triclinic factors as 0.0 in the data file which made the box “triclinic” and worked as a workaround. Hope this helps.


Hi Vikas,
My answer is the same as Arnab’s.

How is your initial simulation box defined? For Planar Couette flow (deform xy, as in your input), you need to tell LAMMPS to use a general box, i.e., triclinic. You could add the following line in the file from which you read_data.

0 0 0 xy xz yz

xlo_value xhi_value xlo xhi
ylo_value yhi_value ylo yhi
zlo_value zhi_value zlo zhi

followed by Atoms section etc…

The first line above is the key. See if the log file prints that the box has been built to be triclinic (even though at time t=0, the off diagonal components of the cell matrix are zero)


Dear lammps users…
Thanks for all the responses. I should have read the read_data page before posting the problem. This was my first shear simulation and was not aware that there could be additional lines (tilt related) in the data file.