Dear All,
From the mail archive and also from text book knowledge, it seems to me that in atomic level there is no such significant difference, if I call a system of bulk water of pressure 1 bar with T= 300K and another system of bulk water of pressure 2 bar with T= 290 K. Most probably that’s why we accept the fluctuation of P and T during equlibriation and generally take the mean. Is there any wrong in my understanding ?
My question is that,
In some cases (suppose in case of sea), temperature will decrease and consequently pressure will increase. Normally it is well accepted that for each 10 m distance under water pressure will change only 1 bar and T approximately ~1 degree C. In case of underwater blast the behavior of water when a shock moves through it is different in different depth, even at the depth of 20m i.e P= 3 bar and T ~ 20 deg C. Specially Us-Up behavior is no longer linear as the depth increases.
So, I was trying to model that with TIP4P/2005 water model with 2 samples i.e P = 1 bar and 3 bar , T = 300K and 290 K and using MSST with the equlibriated sample, but it did not shows significant difference as far as the experimental observations and continuum models are concerned.
My system vol was 100^3 A^3.
Can anybody give me some idea/strategy ?
I don’t agree with you. Two systems with varying temperatures can be very different. Note that the temperature range you give crosses the freezing point! Just take a quick look at this paper below and you will see the effect of temperature.
http://dx.doi.org/10.1021/ja4028814
P and T are conceptually the mean value, or ensemble averages, which should not be confused with the “instantaneous” P and T. Large fluctuations are due to small system sizes. (Statistical Mechanics, McQuarrie, 4th ed., pp35, 57)
Your question is hard to understand for those unfamiliar with it. Could you list some references?
Your questions can be more easily answered if you provide answers to
my following questions:
1) After equilibrating at (P,T) of (1,300) and (3,290), were the
fluctuation in P and T less than 2 bar and 10 K, respectively?
2) After the samples were shocked using MSST, were the departures from
the Hugoniot and the Rayleigh line smaller than the fluctuations in T
and P, respectively? To provide more information to the questions,
you might want to mention what your Us and rise time were.
Ray
Your questions can be more easily answered if you provide answers to
my following questions:
1) After equilibrating at (P,T) of (1,300) and (3,290), were the
fluctuation in P and T less than 2 bar and 10 K, respectively?
Using TIP4P/2005, fix NPT and fix SHAKE with box size 100A^3 after ~1ns the
instantaneous thermo output fluctuation was high, but the output from
ave/time 1 1 10000 the fluctuation for 1 bar was -3 bar to ~3 bar and for 3
bar it was -6 bar to 5 bar. But the temperature fluctuation was lass than
~4 K.
2) After the samples were shocked using MSST, were the departures from
the Hugoniot and the Rayleigh line smaller than the fluctuations in T
and P, respectively?
No, after using different tscale values, in the most optimized case, the
departure from Hugoniot and Rayleigh line was greater than the fluctuation
in T and P during equlibration.
To provide more information to the questions,
you might want to mention what your Us and rise time were.
I was trying with Us= 3 Km/sec. In both cases the rise time was ~40 fs. In
Us-Up curve their was no significant difference.
Another question:
If I take into account that below 1 Km depth P=100 bar and T~ 280 and If I
equlibriate properly then will I get any difference regarding shock
response?
Does the incomprehensibility of water can affect the shock response ?
Comments and replies below.
Ray
Your questions can be more easily answered if you provide answers to
my following questions:
1) After equilibrating at (P,T) of (1,300) and (3,290), were the
fluctuation in P and T less than 2 bar and 10 K, respectively?
Using TIP4P/2005, fix NPT and fix SHAKE with box size 100A^3 after ~1ns the
instantaneous thermo output fluctuation was high, but the output from
ave/time 1 1 10000 the fluctuation for 1 bar was -3 bar to ~3 bar and for 3
bar it was -6 bar to 5 bar. But the temperature fluctuation was lass than ~4
K.
Your fix ave/time setting gave you instantaneous values, not time averaged.
But I suspect that if you have used the correct settings for the fix
ave/time command to obtain time averaged P and T, you will see that
their fluctuations are larger than their intended equilibrium values.
This means that these two systems are indistinguishable.
2) After the samples were shocked using MSST, were the departures from
the Hugoniot and the Rayleigh line smaller than the fluctuations in T
and P, respectively?
No, after using different tscale values, in the most optimized case, the
departure from Hugoniot and Rayleigh line was greater than the fluctuation
in T and P during equlibration.
Same as above. Not only equilibrated systems were indistinguishable,
shocked systems were also similar - which is expected.
To provide more information to the questions,
you might want to mention what your Us and rise time were.
I was trying with Us= 3 Km/sec. In both cases the rise time was ~40 fs. In
Us-Up curve their was no significant difference.
Expected, as the two systems are too similar to show any significant
differences.
Another question:
If I take into account that below 1 Km depth P=100 bar and T~ 280 and If I
equlibriate properly then will I get any difference regarding shock
response?
You will only find out after you try.
Does the incomprehensibility of water can affect the shock response ?
I don't know what "incomprehensibility" is.