EFrans - 2020-08-06
A few different molecular dynamics (MD) simulations of H₂O molecules where the temperature is slowly decreasing (simulated annealing) in order to obtain liquid and then solid states (water and ice). Some thermodynamic properties are observed along the simulations. Neighbor count is the average number of oxygen neighbors (within 5Å) each oxygen has, RDF (radial distribution function) is computed for oxygen-oxygen. Video attempted to be rendered in 4K UHD 60FPS. Interaction model for H20 used is SPC/E, for more info see https://en.wikipedia.org/wiki/Water_model MD simulations carried out with LAMMPS, https://lammps.sandia.gov/ Visualization with OVITO, https://www.ovito.org/ Music: Pure Potentiality by Benjamin Martins Vitality by Benjamin Martins
Is this the kind of behavior that would be seen in zero g? A vacuum? Or both?
P.S. I'm loving the videos you produce. These are exceptional, and make me think about the world around me in new ways.
Yes both. There is no gravity here, though gravity wouldn't have much effect on this type of simulation.
A small rain drop contains about 10^23 water molecules, whereas these simulations are done with a few thousands molecules.
Thank you so much for the kind words!
@EFrans Thank you for the replies. I certainly hadn't considered the scale. There is such a large disconnect when we are so large versus the molecules.
If it's not too personal, are all of these patterns (mathematical as they are) related to your profession or just an interesting hobby?
@Blake Carson Ive studied math and physics, but the simulation videos Ive made are just for fun :)
can you provide the lammps input script for rdf? Great work!!
Thank you!
I computed the RDF in Ovito, but if you like to do it directly in lammps you can maybe find some help here https://lammps.sandia.gov/doc/compute_rdf.html
This is so cool. How can I play around with something like this?
Can you provide the lammps input script for this model?
Lammps have examples/benchmark scripts for water models, check out for instance https://github.com/lammps/lammps/blob/master/bench/POTENTIALS/in.spce
What soft did you use for this simulation?
LAMMPS, see video description.
Would be nice if you put the units on the y-axis
Yes that's true.
For what it is worth temperature and energy are given in Kelvin and eV/atom.
Nice
Thank you
water lore
Simulating Physics - 2020-08-13
Love it.
Have you tried cooling it slow enough to crystallize in a hexagonal lattice?
EFrans - 2020-08-13
Yes I tried slower cooling rates but it seemed difficult to crystallize a perfect lattice. It did work somewhat better in the last simulation of the video where the space in z-direction was reduced, but I would guess one needs much longer time scale to get the perfect lattice.