Why the TotEng is positive ?

Dear lammps-users
I want to ask you a question.
Are all systems of total energy is negative?Can someone explain why the system of the total energy is negative?However, when I simuliated the methanol,I find a strange phenomenon,In the energy minimization,the total energy is negative, but when npt simulation,the total energy is positive,I dont know the reason.

Looking forwarding your reply. Thanks. Jack zhuang

PPPM initialization …
G vector (1/distance)= 0.279756
grid = 15 15 15
stencil order = 5
estimated absolute RMS force accuracy = 0.0144811
estimated relative force accuracy = 4.36094e-05
using double precision FFTs
3d grid and FFT values/proc = 4116 960
Setting up minimization …
Memory usage per processor = 6.93392 Mbytes
Step Temp E_pair E_mol TotEng Press
0 0 3432.9879 983.1922 4416.1801 30347.694
1000 0 -2167.4834 72.727137 -2094.7562 -1097.0266
1003 0 -2167.1703 72.227869 -2094.9425 -1034.1686
Loop time of 34.7178 on 4 procs for 1003 steps with 1500 atoms

Minimization stats:
Stopping criterion = max force evaluations
Energy initial, next-to-last, final =
4416.18012358 -2094.90288435 -2094.94247056
Force two-norm initial, final = 16381.1 7.639
Force max component initial, final = 8937.28 1.54507
Final line search alpha, max atom move = 0.00135656 0.00209597
Iterations, force evaluations = 1003 5001

Pair time () = 18.1113 (52.1671) Bond time () = 0.645308 (1.85872)
Kspce time () = 12.1217 (34.9151) Neigh time () = 0.0615801 (0.177373)
Comm time () = 3.36572 (9.69453) Outpt time () = 7.27177e-05 (0.000209454)
Other time (%) = 0.412091 (1.18697)

FFT time (% of Kspce) = 0.232376 (1.91702)
FFT Gflps 3d (1d only) = 2.36696 8.14115

Nlocal: 375 ave 407 max 326 min
Histogram: 1 0 0 0 0 1 0 0 1 1
Nghost: 3656.5 ave 3763 max 3513 min
Histogram: 1 0 0 0 0 1 0 1 0 1
Neighs: 59689.8 ave 66303 max 47371 min
Histogram: 1 0 0 0 0 0 0 1 1 1

Total # of neighbors = 238759
Ave neighs/atom = 159.173
Ave special neighs/atom = 5
Neighbor list builds = 15
Dangerous builds = 0
PPPM initialization …
G vector (1/distance)= 0.279756
grid = 15 15 15
stencil order = 5
estimated absolute RMS force accuracy = 0.0144811
estimated relative force accuracy = 4.36094e-05
using double precision FFTs
3d grid and FFT values/proc = 4116 960
Setting up run …
Memory usage per processor = 6.21898 Mbytes
Step Temp TotEng Volume den_nvt
1003 298.15 -762.73842 27000 0.49267501
2000 267.51582 391.20822 26776.45 0.49678823
3000 282.28435 486.60513 25652.685 0.518551
4000 287.02643 450.48722 24877.362 0.53471204
5000 297.13156 513.93055 24318.34 0.54700384
6000 300.1999 505.5098 23219.203 0.57289759
7000 297.96665 459.84821 22200.261 0.59919228
8000 305.20581 425.08797 21639.964 0.61470644
9000 295.64187 393.72675 20722.411 0.6419246
10000 299.78943 358.67459 19998.905 0.66514769
11000 309.90264 451.57076 19855.346 0.66995686
12000 299.88925 347.04874 19818.162 0.67121386
13000 293.51759 348.77915 19910.263 0.66810896

Dear lammps-users
    I want to ask you a question.
   Are all systems of total energy is negative?

​no. the absolute value of the energy is arbitrary. MD requires forces and
thus what matters for MD are the derivatives. but with typical potentials
and parameters it is often negative under normal circumstances.

Can someone explain why the system of the total energy is negative?However,
when I simuliated the methanol,I find a strange phenomenon,In the energy
minimization,the total energy is negative, but when npt simulation,the
total energy is positive,I dont know the reason.

​your system is collapsing extremely fast. check out the volume. that is​
usually a sign for bad parameters.
you have to carefully visualize your trajectory to see what is happening.
also you should first test, whether your simulation runs stable and
conserves energy when running with fix nve only, i.e. without a thermostat
and barostat.

axel.