[gmx-users] SD integrator

[HANNIBAL LECTER]

Hi all,

I have been posting for a while with my problems in simulating peptides
inside CNTs. After a lot of trials it seems like simulating the CNT with
bonds and angles seems like a wise thing to do.

I am using the SD integrator + PR barostat. It seems like improper
thermostatting can blow up the system. For example, when I am trying to
simulate with tau-t for CNT as 1.0 the system blows up for 300K. The output
temperatures (from g_energy) for CNT is about 150K for a ref-t of 300K.

 However, if a stronger coupling is used tau-t = 0.005, the system works
fine. Are there any artifacts for coupling the CNT group with such a low
tau-t?