MSST

Dear all,

Sorry for such frequent mails regarding this MSST simulation.

I have some precise questions about MSST,

  1. After equilibrating my sample at 300K temperature I start MSST calculation but at the first step the temperature showing 196K and also the first P value also differs a lot. Does it have physical significance or this is due to any mistake ?

( I have equilibrated my sample (water) for a long time and the time averaged value of P and T was fluctuating around the intending T and P. But the fluctuation was not so high, for T, it was +/- 5 and for P, it was +/- 15 from ave/time result, ave/time 1 10000 10000, tstep = 1fs , box size 24 Angs cube )

  1. In many reference regarding MSST said that it requires small system size and in case of different system it differs, basically it differs depending on the atomic corelation function. So in case of water how can I decide that this box size would be sufficient to satisfy the equations ?

In short to say how can I assume that this system size should be proper for MSST ?

  1. If the departure of dHugoniot and dReleigh is small with respect to time and can it be proper to decide that the shock arrangement and related dynamics of my sample would be perfect ?

Please make your comments, it would be a great help for me.

Thanks in advance

Ravi.

Dear all,

Sorry for such frequent mails regarding this MSST simulation.

I have some precise questions about MSST,

1. After equilibrating my sample at 300K temperature I start MSST
calculation but at the first step the temperature showing 196K and also the
first P value also differs a lot. Does it have physical significance or this
is due to any mistake ?

( I have equilibrated my sample (water) for a long time and the time
averaged value of P and T was fluctuating around the intending T and P. But
the fluctuation was not so high, for T, it was +/- 5 and for P, it was +/-
15 from ave/time result, ave/time 1 10000 10000, tstep = 1fs , box size
24 Angs cube )

difficult to say, but the 200K/300K hints at fix shake being involved
somehow. not sure if using shake is a good idea in this context and
whether using a flexible water potential wouldn't be better.

2. In many reference regarding MSST said that it requires small system size
and in case of different system it differs, basically it differs depending
on the atomic corelation function. So in case of water how can I decide that
this box size would be sufficient to satisfy the equations ?

i don't know. this is not my area of expertise.

In short to say how can I assume that this system size should be proper for
MSST ?

this is always difficult to answer. the best strategy is to first do
some simulations on systems with known results and experiment to find
out what signifies a bad choice and then translate that experience to
your system of interest.

axel.

Dear sir,

To test if there any difference, I test also with TIP4P water model with SHAKE , but the result in both cases were nearly same.
Here is the input ( courtesy to L C Liu and Moultos, Othon ) I used–

units real
dimension 3
boundary p p p
atom_style full
newton on
#atom_modify sort 1 0

you say you start from an equilibrated restart, but then you
initialize the temperature to 1K . why?

axel.

Dear Axel Sir,

This is the script I used during eqilibration, not for MSST calculation. The output restart file getting from this script was utilized and was the starting point of my MSST calculation.

Dear Axel Sir,

This is the script I used during eqilibration, not for MSST calculation. The
output restart file getting from this script was utilized and was the
starting point of my MSST calculation.

then i don't understand what you want to prove with that. in any case,
it doesn't affect my original assessment.

axel.

Dear Axel sir,

From your assessment, if I don’t misunderstand, you are suggesting to not use SHAKE and MSST together ?

So, is it non-physical to use fix SHAKE and fix MSST together in a simulation ?

Even in the example script of the MSST, the temperature have mentioned 600 in velocity command.

“# Initial velocities correspond to around 300K.
velocity all create 600.0 9999”

Here I wonder, so to get the intended temperature, one should eqilibriate a system at twice of the intended temperature ?

But as per my experience in case of shock/NEMD simulation, I equilibriated my sample at that temperature what I wise and the results were fine wrt the experimental findings.

Thanks in advance.

Dear Axel sir,

From your assessment, if I don't misunderstand, you are suggesting to not
use SHAKE and MSST together ?
So, is it non-physical to use fix SHAKE and fix MSST together in a
simulation ?

i don't know. as i already stated, i am not an expert in these
matters, but it seems an odd choice to me.

Even in the example script of the MSST, the temperature have mentioned 600
in velocity command.

"# Initial velocities correspond to around 300K.
velocity all create 600.0 9999"

Here I wonder, so to get the *intended* temperature, one should eqilibriate
a system at twice of the intended temperature ?

of course not.

But as per my experience in case of shock/NEMD simulation, I equilibriated
my sample at that temperature what I wise and the results were fine wrt the
experimental findings.

well, you don't have to convince me, but the people reviewing your
work (that are hopefully experts). everybody can say "those results
look fine to me".

axel.