[gmx-users] Structure blowing up after increased the temperature

[Justin Lemkul]

On 3/22/17 12:55 AM, Sameer Edirisinghe wrote:
> Thank you for the help Dr. Jsutin
>
> On 3/21/17 12:39 PM, Sameer Edirisinghe wrote:
>> Dear users,
>>
>> I'm trying to find the melting point and Tg of my polymer system using
>> Density vs temperature curve. I'm using all oplsaa forcefield and  I'm
> going
>> for NVT and followed by NPT equilibration ( Both at 523K) of 400ps each.
>> then after the production run (at 523K for 5ns) the system is blowing up
>> and can be observed broken molecules of the trajectory from VMD. But no
>> blowing up occurring at 300K temperature (200ns).
>>
>
> "Broken" molecules are a consequence of PBC, not instability.
>
> *I used -pbc nojump flag in trajconv command but still molecules seems to
> be broken*
>

Because -pbc nojump does not make molecules whole.  There's a different option 
for that.  Please look at 
http://www.gromacs.org/Documentation/Terminology/Periodic_Boundary_Conditions#Suggested_trjconv_workflow 
and the trjconv help information.

>> But if I only did the production run (at 523K) up to 200ps it seems no
>> blowing up occurring until 200ps.
>>
>
> Systems simulated at higher temperature will naturally diffuse faster,
> therefore
> cross periodic boundaries more frequently or more quickly.
>
>> Note. used .mdp files has attached with the mail
>>
>
> The mailing list does not accept attachments.
>
>> 1) I'm not sure if the .mdp files caused the blowing up or the topology
>> caused the blow up of the system.
>>
>> 2) Can use same .mdp files by only changing coupling temperature for any
>> temperature?
>>
>
> High-temperature simulations often require shorter time steps and perhaps a
> bit
> tighter regulation of temperature and pressure to be stable, but as of yet
> we
> have no evidence that your simulation was actually unstable.
>
> *Here i used 0.001ps as time step. Do i need to reduce it more ?*
>

Not likely, but again you haven't presented any evidence of unstable physics, 
anyway.

> *Pressure and temperature have fluctuations after the equilibration while
> pressure seems to have higher fluctuations.*
>

Naturally.

http://www.gromacs.org/Documentation/Terminology/Pressure

> *see my temperature and pressure coupling options below. I used them for
> both equilibrium and production run.*
>
> *tcoupl = Nose-Hoover        *
> *tc-grps = MD ; *
> *tau_t = 8        ; *
> *ref_t = 523          ;*
>
> *; Pressure coupling is on*
> *pcoupl        = Parrinello-Rahman     *
> *pcoupltype        = isotropic            ; uniform scaling of box vectors*
> *tau_p        = 5.0            ; time constant, in ps*
> *ref_p        = 1.0            ; reference pressure, in bar*
> *compressibility     = 4.46e-5            ; *
>

Those are fairly permissive tau-t and tau-p values, but if they get the job done 
and the distributions are OK, then I suppose it's alright.

-Justin

-- 
==================================================

Justin A. Lemkul, Ph.D.
Ruth L. Kirschstein NRSA Postdoctoral Fellow

Department of Pharmaceutical Sciences
School of Pharmacy
Health Sciences Facility II, Room 629
University of Maryland, Baltimore
20 Penn St.
Baltimore, MD 21201

jalemkul at outerbanks.umaryland.edu | (410) 706-7441
http://mackerell.umaryland.edu/~jalemkul

==================================================