Bulk ice#

Fortran on ice Ih.

Two cells in this package are pure ice, and one mixed frame in seams-core is not. The engine walk-through of that mixed frame is Classify ice on the mixed TIP4P example.

Ice Ih#

input/traj/ice1h.lammpstrj is 128 waters. The program on the Fortran page reads it. CHILL+ labels every molecule hexagonal, the cage search finds 128 hexagonal cages and no double-diamond cage, and the ring histogram holds 256 six-rings. The mean local q6 is 0.62585547.

Cubic ice#

example/cubic.f90 builds a 64-atom cubic cell in memory, two repeats of the eight-site cell on an edge of 2.75 * 4 / sqrt(3) angstrom. fpm run --example cubic checks the classification and prints the census quoted on the API page. All 64 atoms are cubic ice. The cage census is 0 64 0 0: no hexagonal cage, 64 double-diamond cages, no water, no atom in both.

A mixed frame#

input/traj/exampleTraj.lammpstrj in a seams-core checkout is one frame of 750 atoms: 250 of type 2 and 500 of type 1. Read type 2 and pass typ=2 so CHILL+ and the cage search see those 250 atoms. The cutoff graph (the default) labels 9 interClathrate and 241 water. The mutual four-neighbour graph labels 1 clathrate, 7 interClathrate and 242 water. The seeded cage search labels all 250 water: cages returns 0 0 250 0.

frame = new_frame()
call frame%read("input/traj/exampleTraj.lammpstrj", type_filter=2_seams_int)
call frame%chill_plus(labels, counts, typ=2_seams_int)
call frame%chill_plus(labels, counts, typ=2_seams_int, graph=SEAMS_GRAPH_MUTUAL)
cages = frame%cages(typ=2_seams_int)

CHILL+ is a four-neighbour star. Cages ask whether the molecule belongs to a complete hexagonal or double-diamond cage. On this frame the star finds a few eclipsed shells and the cage search finds none. The same split, with the command-line names, is on the engine page linked above.