Help in fix_npt

Dear all

I have been trying to start off the simulation by initiating the temperature to 8000K before quenching it to 300k. However, I have obtained weird results as follow for my temperature. It drops drastically from 8000K to 3K to 8K then dropped to 2K again!

Step Temp mytemp Press Volume Enthalpy KinEng PotEng TotEng Lx Ly Lz

1 8000 8000 147985.85 45499.293 -47155.45 8271.6152 -59629.631 -51358.015 35.7 35.7 35.7

500 3.2476022 3.2476022 195740.8 45499.293 -18585.88 3.3578645 -24147.969 -24144.611 35.7 35.7 35.7

1000 8.209267 8.209267 164063.91 45499.293 -22146.158 8.4879872 -26813.803 -26805.315 35.7 35.7 35.7

1500 1.9333188 1.9333188 146585.27 45499.293 -23697.011 1.9989587 -27861.801 -27859.802 35.7 35.7 35.7

2000 2.8635126 2.8635126 140562.7 45499.293 -24062.913 2.9607343 -28057.633 -28054.672 35.7 35.7 35.7

2500 2.413017 2.413017 134935.82 45499.293 -24726.209 2.4949435 -28560.669 -28558.174 35.7 35.7 35.7

3000 2.3054034 2.3054034 124513.15 45499.293 -25402.911 2.3836762 -28941.272 -28938.889 35.7 35.7 35.7

3500 0.75438394 0.75438394 118059.3 45499.293 -25909.088 0.7799967 -29262.566 -29261.786 35.7 35.7 35.7

4000 4.1562003 4.1562003 113782.63 45499.293 -26278.981 4.2973112 -29514.526 -29510.228 35.7 35.7 35.7

4500 1.0812323 1.0812323 120325.21 45499.293 -26157.232 1.1179421 -29575.397 -29574.279 35.7 35.7 35.7

5000 2.2291664 2.2291664 114692.3 45499.293 -26529.938 2.3048508 -29789.324 -29787.019 35.7 35.7 35.7

5500 1.5746165 1.5746165 120730.51 45499.293 -26373.414 1.6280777 -29803.599 -29801.97 35.7 35.7 35.7

6000 0.89677911 0.89677911 116514.11 45499.293 -26485.288 0.92722647 -29795.033 -29794.105 35.7 35.7 35.7

6500 0.82060821 0.82060821 122853.67 45499.293 -26257.395 0.84846942 -29747.094 -29746.245 35.7 35.7 35.7

7000 0.9418315 0.9418315 112824.76 45499.293 -26738.675 0.97380847 -29943.695 -29942.721 35.7 35.7 35.7

7500 0.84125028 0.84125028 113635.15 45499.293 -26731.413 0.86981233 -29959.342 -29958.473 35.7 35.7 35.7

8000 1.1307306 1.1307306 108311.86 45499.293 -27036.491 1.1691211 -30113.547 -30112.377 35.7 35.7 35.7

8500 10872.179 10872.179 513852.99 45499.293 -2622.0308 11241.31 -28455.957 -17214.648 35.7 35.7 35.7

9000 0.13399614 0.13399614 109740.57 45499.293 -26929.072 0.13854556 -30045.67 -30045.532 35.7 35.7 35.7

9500 0.1347941 0.1347941 107783.32 45499.293 -27064.398 0.13937061 -30125.414 -30125.275 35.7 35.7 35.7

10000 0.32513566 0.32513566 104644.04 45499.293 -27218.198 0.33617463 -30190.26 -30189.924 35.7 35.7 35.7

10500 48.755549 48.755549 102550.61 45499.293 -27379.321 50.410893 -30342.008 -30291.597 35.7 35.7 35.7

11000 2.0365517 2.0365517 139822.1 45499.293 -24239.792 2.1056965 -28212.626 -28210.52 35.7 35.7 35.7

11500 1.4115211 1.4115211 123928.65 45499.293 -25673.814 1.4594449 -29194.652 -29193.192 35.7 35.7 35.7

12000 0.92354924 0.92354924 110936.52 45499.293 -26531.457 0.95490549 -29682.835 -29681.88 35.7 35.7 35.7

12500 0.19326419 0.19326419 108420.15 45499.293 -26718.875 0.19982587 -29798.037 -29797.837 35.7 35.7 35.7

13000 0.2824743 0.2824743 103538.62 45499.293 -26971.709 0.29206484 -29912.335 -29912.043 35.7 35.7 35.7

13500 0.52719964 0.52719964 99748.239 45499.293 -27228.166 0.54509907 -30061.404 -30060.859 35.7 35.7 35.7

14000 6.162192 6.162192 95414.448 45499.293 -27562.273 6.3714102 -30278.265 -30271.893 35.7 35.7 35.7

14500 1.3122217 1.3122217 124593.45 45499.293 -25658.971 1.3567741 -29198.585 -29197.229 35.7 35.7 35.7

15000 1.4424984 1.4424984 117887.62 45499.293 -26157.821 1.491474 -29507.136 -29505.644 35.7 35.7 35.7

15500 1.7469319 1.7469319 111257.48 45499.293 -26411.474 1.8062435 -29572.818 -29571.011 35.7 35.7 35.7

16000 3.030266 3.030266 148686.84 45499.293 -24393.565 3.1331493 -28619.171 -28616.037 35.7 35.7 35.7

16500 115161.34 115161.34 2935036.2 45499.293 173317.26 119071.28 -29104.435 89966.848 35.7 35.7 35.7

17000 0.30762128 0.30762128 108913.06 45499.293 -26370.073 0.31806561 -29463.35 -29463.032 35.7 35.7 35.7

17500 0.9705821 0.9705821 103961.29 45499.293 -26661.395 1.0035352 -29614.736 -29613.732 35.7 35.7 35.7

18000 30.196232 30.196232 95284.665 45499.293 -27082.513 31.221451 -29819.669 -29788.447 35.7 35.7 35.7

18500 0.30440262 0.30440262 96730.446 45499.293 -27014.421 0.31473767 -29761.729 -29761.414 35.7 35.7 35.7

19000 0.5868431 0.5868431 93877.425 45499.293 -27206.595 0.60676754 -29873.173 -29872.567 35.7 35.7 35.7

19500 12.059545 12.059545 119159.81 45499.293 -23297.584 12.46899 -26694.005 -26681.536 35.7 35.7 35.7

20000 2.2120446 2.2120446 119698.52 45499.293 -25228.887 2.2871477 -28630.424 -28628.137 35.7 35.7 35.7

20001 2.0335475 2.0335475 120750.5 45499.293 -25201.332 2.1025903 -28632.559 -28630.456 35.7 35.7 35.7

Loop time of 1144.36 on 16 procs for 20000 steps with 4908 atoms

Is there some advice that could be given as such situation did not arise when fix npt were being used. The following is the input code used:

atom_style charge

boundary p p p

units metal

dimension 3

neighbor 2.5 bin

neigh_modify delay 1 every 1 check yes

Why dont you make a plot of timesteps vs temperature. Maybe the 3K to 8K to 2K is part of a fluctuation before the systems equlibrate. (fluctuares around the seted T value)…
Please look that all other values are : 2.413017 , 2.2291664 etc…

Thanks
Oscar G

fix 1 all nvt temp 8000 300 0.01

Does the system stay at 8000K if you
use NVE? If not, it is not at equilibrium
before you use NVT. And a damping constant
of 0.01 for metal units is 10 timesteps, which
is far too small. The doc page suggests 100
timesteps.

Steve

Hi Steve

Thank you for your response.

For my system, i used diamond lattice structure.
I have tried running it at 5000K to 300K using fix nve, and results look fine, without any lost atoms issues.
During equilibration of the system at 5000K, the values did fluctuates between 4000K to 5000K, but at the end it remains near to 5000K.

Hence i went on to use the same codes and included heating in the new system, from 5000K to 8000K. But before even reaching 8000K, the system went haywire again after hitting 6500K and lost atoms situation appeared again.
Hence, I am helpless in what to do with it.

Because I need to heat the diamond lattice structure high enough before I can quench it to ultimately calculate the sp3 ratio in the final state. Is there any advice?

I would really appreciate any help given. Thank you!

Best Regards
Ah Boy

Hence i went on to use the same codes and included heating in the new system, from 5000K to 8000K. But before even reaching >8000K, the system went haywire again after hitting 6500K and lost atoms situation appeared again.
Hence, I am helpless in what to do with it.

Then you're doing bad dynamics - it could be heating too fast, too big
a timestep, not reneighboring often enough, etc. You should
monitor the thermo output at high frequency to insure something
in your system isn't blowing up. What material and potential are
you using - maybe the potential breaks down if atoms get
too close to each other at high T.

Steve

Hi Steve

Can i clarify something about timestep? You mentioned that a damping constant of 0.01 for metal units is 10 timesteps. Actually how do you go about calculating timesteps from damping constant?

I am using AIREBO for my system. How do I check if the potential will break down if atoms get
too close to each other at high T?

Thanks a lot!

Best Regards
Ah Boy

Comments below.

Steve

Hi Steve

Can i clarify something about timestep? You mentioned that a damping
constant of 0.01 for metal units is 10 timesteps. Actually how do you go
about calculating timesteps from damping constant?

See the fix npt doc page. It explains it.

I am using AIREBO for my system. How do I check if the potential will break
down if atoms get
too close to each other at high T?

There is no simple answer to this Q. It also depends on T. Many
potentials, like LJ, do not "break down", as they go to infinity
as separation goes to 0. Some do not. I imagine AIREBO is fine
for physically realistic temperatures.