Ni-Fe ADP potential , ERROR: Lost atoms: original 4000 current 3998 (../thermo.cpp:442)

LAMMPS LAMMPS Mailing List Mirror
1

Respected Sir,

I am working on adp potential, and created Fe-Ni.adp file myself.

While run, I have noticed that box size in decreasing substantially, and temperature is far from desired value 300.

How to resolve this issue?

Please advice.

Output.
Step Temp PotEng TotEng Press Volume Lx Ly Lz Xlo Xhi Ylo Yhi Zlo Zhi
0 300 -1.064 -1.026 -2.145e+05 4.444e+04 35.42 35.42 35.42 0 35.42 0 35.42 0 35.42
100 2647 -1.708 -1.366 -2.646e+05 3.655e+04 33.19 33.18 33.19 1.116 34.31 1.119 34.3 1.116 34.31
200 3470 -2.543 -2.095 -1.766e+05 3.228e+04 31.84 31.82 31.86 1.793 33.63 1.799 33.62 1.781 33.64
300 2781 -3.109 -2.749 -1.449e+05 2.893e+04 30.71 30.66 30.72 2.355 33.07 2.38 33.04 2.353 33.07
400 2118 -3.532 -3.258 -1.4e+05 2.633e+04 29.78 29.69 29.78 2.82 32.6 2.868 32.55 2.82 32.6
500 1637 -3.844 -3.632 -1.415e+05 2.415e+04 28.94 28.82 28.95 3.243 32.18 3.299 32.12 3.235 32.19
600 1322 -4.08 -3.909 -1.439e+05 2.227e+04 28.16 28.06 28.19 3.633 31.79 3.682 31.74 3.618 31.8
700 1144 -4.276 -4.128 -1.586e+05 2.054e+04 27.39 27.33 27.44 4.015 31.41 4.045 31.38 3.99 31.43
800 1045 -4.448 -4.313 -1.575e+05 1.901e+04 26.66 26.66 26.74 4.379 31.04 4.38 31.04 4.34 31.08
900 939.8 -4.597 -4.476 -1.711e+05 1.761e+04 25.96 26.02 26.08 4.733 30.69 4.702 30.72 4.671 30.75
1000 823.3 -4.724 -4.617 -1.937e+05 1.631e+04 25.26 25.4 25.42 5.079 30.34 5.011 30.41 4.999 30.42
1100 777.3 -4.846 -4.746 -2.058e+05 1.508e+04 24.58 24.78 24.76 5.421 30 5.323 30.1 5.329 30.09
1200 722.7 -4.955 -4.861 -2.314e+05 1.396e+04 23.94 24.17 24.13 5.74 29.68 5.626 29.8 5.647 29.77
1300 707.7 -5.062 -4.971 -2.533e+05 1.291e+04 23.32 23.55 23.5 6.049 29.37 5.936 29.49 5.96 29.46
1400 703.2 -5.169 -5.078 -2.712e+05 1.194e+04 22.74 22.93 22.9 6.34 29.08 6.246 29.18 6.262 29.16
1500 672.4 -5.268 -5.182 -2.938e+05 1.105e+04 22.17 22.31 22.33 6.626 28.8 6.554 28.87 6.547 28.87
1600 676.9 -5.37 -5.283 -3.406e+05 1.021e+04 21.59 21.7 21.79 6.914 28.51 6.862 28.56 6.818 28.6
1700 643.8 -5.463 -5.38 -3.413e+05 9445 21.03 21.1 21.29 7.198 28.22 7.161 28.26 7.066 28.36
1800 617.3 -5.553 -5.473 -3.935e+05 8736 20.47 20.51 20.81 7.477 27.95 7.457 27.96 7.305 28.12
1900 623.6 -5.646 -5.566 -4.377e+05 8067 19.91 19.92 20.34 7.755 27.67 7.751 27.67 7.542 27.88
2000 617.2 -5.74 -5.66 -4.621e+05 7452 19.37 19.35 19.88 8.024 27.4 8.037 27.39 7.772 27.65
2100 589.6 -5.829 -5.753 -5.224e+05 6881 18.85 18.79 19.42 8.286 27.14 8.314 27.11 8 27.42
2200 547.6 -5.91 -5.839 -5.867e+05 6346 18.34 18.25 18.96 8.54 26.88 8.586 26.84 8.232 27.19
2300 602.4 -6.006 -5.928 -7.232e+05 5832 17.83 17.7 18.47 8.795 26.63 8.859 26.56 8.474 26.95
2400 679.2 -6.113 -6.025 -7.475e+05 5348 17.32 17.17 17.99 9.049 26.37 9.128 26.29 8.718 26.7
ERROR: Lost atoms: original 4000 current 3998 (…/thermo.cpp:442)
Last command: run 3000

output.dat (5.08 KB)

input1 (1.52 KB)

2

Respected Sir,

I am working on adp potential, and created Fe-Ni.adp file myself.
While run, I have noticed that box size in decreasing substantially, and
temperature is far from desired value 300.
How to resolve this issue?

​this sounds like you have bad force field parameters or not the correct
units setting for those parameters.
​
axel.

3

Sir,

Parameter files is taken from NIST data base (https://www.ctcms.nist.gov/potentials/Ni.html#Fe-Ni) and was modified by interpolating the data ,to have values at required data points.

Units for Metal style are , angstroms for distance and eV for energy, which are same to units, mentioned in force filed data points.

Now, if we consider, units is not a problem, then what should I do, to solve the problem of bad force filed parameters.

4

Sir,
Parameter files is taken from NIST data base (https://www.ctcms.nist.gov/
potentials/Ni.html#Fe-Ni) and was modified by interpolating the data ,to
have values at required data points.

Units for Metal style are , angstroms for distance and eV for energy,
which are same to units, mentioned in force filed data points.

Now, if we consider, units is not a problem, then what should I do, to
solve the problem of bad force filed parameters.

​look up the publication describing the potential parameters and try to
reproduce the calculations in the paper.
please note, that what you are asking about is less an issue of LAMMPS, but
mostly an issue of correctly setting up MD simulations and validating them.
you should discuss this with your adviser/supervisor. after all, it is
quite difficult to have a look over your shoulder from an e-mail.

axel.​