Hi All,
I am attempting to run a cell and geometry optimization of a simple orthogonal unit cell, but am having issues with the final cell parameters.
I’ve set up a simple 4x4x4 lattice of regularly arranged atoms connected by bonds (see photo). In LAMMPS, I have turned off all pair/angle/dihedral/improper interactions, such that all of the energy in the system is from the bond stretch. The default bond length is set to 5, such that the optimal cell should be [20, 20, 20]. As a test case, I am trying to minimize a [20, 20, 24] cell, such that the bonds along the z-axis are strained, but the x-, y-axis bonds are not. To minimize the cell, I use ‘fix box/relax’ during a ‘minimize’ run.
The problem I’ve noticed is as follows: the box lengths maintain their relative lengths, rather than varying indepe
ndently. As a result, the optimized cell size is approx. [18.4, 18.4, 22.1]. The full command I use is “fix 1 all box/relax iso 0.0”. I’m assuming that this behavior is related to the fact that I am using a uniform pressure, but my understanding from the docs was that all three cell parameters should be able to vary independently.
I’ve attached the key input and output files. If anyone could provide any input, that would be a huge help! (This is using stable_16Mar2018).
Regards,
Brian
in.test (454 Bytes)
log.lammps (3.83 KB)
data.test (5.05 KB)
dump.min (17.6 KB)
dump.bonds (37.1 KB)
Hi All,
I am attempting to run a cell and geometry optimization of a simple orthogonal unit cell, but am having issues with the final cell parameters.
I’ve set up a simple 4x4x4 lattice of regularly arranged atoms connected by bonds (see photo). In LAMMPS, I have turned off all pair/angle/dihedral/improper interactions, such that all of the energy in the system is from the bond stretch. The default bond length is set to 5, such that the optimal cell should be [20, 20, 20]. As a test case, I am trying to minimize a [20, 20, 24] cell, such that the bonds along the z-axis are strained, but the x-, y-axis bonds are not. To minimize the cell, I use ‘fix box/relax’ during a ‘minimize’ run.
The problem I’ve noticed is as follows: the box lengths maintain their relative lengths, rather than varying indepe
ndently. As a result, the optimized cell size is approx. [18.4, 18.4, 22.1]. The full command I use is “fix 1 all box/relax iso 0.0”. I’m assuming that this behavior is related to the fact that I am using a uniform pressure, but my understanding from the docs was that all three cell parameters should be able to vary independently.
yes, that is possible, but that is not what you are asking for. if you have a closer look at the documentation, then you should see that the option “iso 0.0” is a shortcut for
“x 0.0 y 0.0 z 0.0 couple xyz” which means that the scaling is isotropic, i.e. maintaining the ratio of box dimensions, so what you are expecting to see is therefore impossible. The setting you should have used is “x 0.0 y 0.0 z 0.0 couple none” or short “ansio 0.0”.
I’ve attached the key input and output files. If anyone could provide any input, that would be a huge help! (This is using stable_16Mar2018).
it is all in the documentation.
axel.
For some reason I was thinking that was only maintaining the box tilt, but the documentation does make what you said clear. Thank you so much for your reply!
-Brian