Pgfnf optimized molecule explode (sulphur polymorph)

GULP
1

Hi

A sulphur modification explode after optimization in GULP by pgfnff.
It is not a charged system = no ions, not the already discussed problem.
When optimized by DFT a large cut off must be used to descipbe week interaction.
Maybe similar option is missing in pgfn ff ?

Input data follows:


gfnff conp gwolf opti

vectors
10.46460000000000 0.00000000000000 0.00000000000000
0.00000000000000 12.86600000000000 0.00000000000000
0.00000000000000 0.00000000000000 24.48600000000000

spacegroup
F D D D

fractional
S 0.85584000000000 -0.04733000000000 -0.04854000000000
S 0.70733000000000 -0.02023000000000 0.00409000000000
S 0.78415000000000 0.03022000000000 0.07623000000000
S 0.78596000000000 -0.09232000000000 0.12947000000000

output cif FURHUV-out
2

The initial energy is a bit ugly, but I tried your input file with GULP-6.3 (latest version) and the optimisation converges. Attached below is the output file.
s.got (20.3 KB)