PE/substrate discontinuity upon restart

Hello – I’m simulating a droplet of water resting on a substrate (wall/lj93) and I have noticed that each time I restart, there is a large discontinuity at the potential energy at the water/substrate interface (re: “substr” column in the output file). It’s almost as if the droplet/surface interaction needs to re-equilibrate from scratch. There is a lesser discontinuity in the total potential energy, although in that case the blip seems to be only momentary; overall potential energy seems to continuously settle down even restarting. I have attached the file containing the thermodynamic quantities, comprising three simulations i.e. restarted twice. I have also attached copy of the input file.

Advance thanks for your assisstance,

Masako Yamada, Ph.D

GE Global Research

hemisphere_mini.08.in (1.8 KB)

restart.txt (62.5 KB)

Hello -- I’m simulating a droplet of water resting on a substrate
(wall/lj93) and I have noticed that each time I restart, there is a large
discontinuity at the potential energy at the water/substrate interface (re:
“substr” column in the output file). It’s almost as if the droplet/surface
interaction needs to re-equilibrate from scratch. There is a lesser
discontinuity in the total potential energy, although in that case the blip
seems to be only momentary; overall potential energy seems to continuously
settle down even restarting. I have attached the file containing the
thermodynamic quantities, comprising three simulations i.e. restarted twice.
I have also attached copy of the input file.

i would suspect the reason for that is the use of shrinkwrap boundary
conditions in combination with placing the wall based on the value of ylo.

please try using a fixed boundary where you place the wall
if you don't want the wall to be bounced around.

cheers,
    axel.

Thanks Axel -- indeed, changing the definition of ylo from ${ylo} to a fixed number seems to have done the trick.

Masako