[gmx-users] relative constraint deviation after LINCS

[Pallavi Banerjee]

Hello Users,

I am carrying out simulated annealing of Polyacrylamide melt (NPT ensemble)
consisting of polymer chains of length 60; taking the temperature up from
400 K to 950 K and beating it down again. I do this in intervals of 50 K,
alternating between heating over 2 ns and equilibration for 3 ns, for a
total run time of 68 ns. What confounds me is that if I use OPLS as the
force field, it's a smooth error-free run but on using GROMOS, the LINCS
warning: "relative constraint deviation after LINCS" shows up after a
significant amount of run, (20-30 ns). I do not really understand the
force-field dependency here. Could someone throw some light on this?

Thanks,

Pallavi Banerjee.