[gmx-users] constant pressure in NPT simulation

[Alex]

The generally accepted protocol for simulations that end up running in 
NPT is EM > relaxation NPT (possibly a series of relaxations) > 
production NPT. You will almost certainly see high pressure fluctuations 
right after NVT, especially given your extremely short timescales. Also, 
your tau_t is rather short for a piston thermostat (e.g. Nose-Hoover), 
especially in combination with a tau_p of 1 ps for your NPT run.

Alex


On 8/6/2018 7:07 AM, Lucia Mydlova wrote:
> Thank you very much for your helpful advice.
>
> I simulated polymer PMMA with two organic chromophores in NVT ensemble for
> 2ns and, the same system one more time for 1.5ns
> with parameters:
> constraints         =  none
> integrator          =  md
> dt                  =  0.001
> nsteps              =  1500000
> comm_mode           =  Linear
> nstcomm             =  1
> nstxout             =  1000
> nstvout             =  1000
> nstfout             =  1000
> nstlog              =  1000
> nstenergy           =  1000
> nstxtcout           =  1000
> xtc_precision       =  1000
> pbc                 =  xyz
> nstlist             =  1
> cutoff-scheme       =  group
> ns_type             =  grid
> coulombtype         =  Cut-off
> vdwtype             =  Cut-off
> rlist               =  1.10
> rcoulomb            =  1.10
> rvdw                =  1.10
> tcoupl              =  nose-hoover
> tc_grps             =  system
> tau_t               =  0.02
> ref_t               =  300
> nsttcouple          =  1
>
> and now I want to simulate the system in NPT ensemble for different
> temperature from 100K up to 700K
> integrator          =  md
> dt                  =  0.001 ; ps !
> nsteps              =  2000000
> comm_mode           =  Linear
> nstcomm             =  1
> nstxout             =  1000
> nstvout             =  1000
> nstfout             =  1000
> nstlog              =  1000
> nstenergy           =  1000
> nstxtcout           =  1000
> xtc_precision       =  1000
> nstlist             =  1
> cutoff-scheme       =  group
> ns_type             =  grid
> pbc                 =  xyz
> coulombtype         =  Cut-off
> vdwtype             =  Cut-off
> rlist               =  1.1
> rcoulomb            =  1.1
> rvdw                =  1.1
> tcoupl              =  nose-hoover
> tc_grps             =  system
> tau_t               =  0.02
> ref_t               =  100
> Pcoupl              =  Parrinello-Rahman
> pcoupltype          =  isotropic
> tau_p               =  1.0
> compressibility     =  4.5e-5
> ref_p               =  1.0
>
> The fluctuation of pressure is high in simulations. Should I use the same
> barostat for my system?
>
> Lucia
>
>
> 2018-08-03 9:14 GMT+02:00 Lucia Mydlova <lucia.mydlova at gmail.com>:
>
>> Dear all customers.
>>
>> I have a problem, in my NPT simulations the pressure is not constant and
>> the temperature is. Earlier simulation in NVT conditions was ok. I use
>> these parameters:
>>   tcoupl              =  nose-hoover
>> tc_grps             =  system
>> tau_t               =  0.002
>> ref_t               =  500
>> Pcoupl              =  parrinello-rahman
>> pcoupltype          =  isotropic
>> tau_p               =  1.0
>> compressibility     =  4.5e-5
>> ref_p               =  1.0
>>
>> What should I change in the .mpd file?
>> Thank you very much.
>> Lucia
>>