[lammps-users] Periodicity

I am running a simulation on cellulose with triclinic boundary conditions the following input file. However, when I dump the x y z of the simulation the timestep 0 x y z the bonds are not intact
anymore and the x y z is different from the input file. I turned off the force field and the x y z is different. I did translate the x y z to the center of the triclinc box this imporve the x y z at timestep 0 where 11 atoms out of 46 are different from the input x y z. I wanted to know any suggestions why is this happening and how I can improve it

The following input x y z coordinates

  46 atoms
                       46 bonds
                       80 angles
                      121 dihedrals

                        5 atom types
                        5 bond types
                       10 angle types
                        8 dihedral types

           0.000 7.784 xlo xhi
           0.000 8.14746952 ylo yhi
           0.000 10.38 zlo zhi

          -0.93548922 0 0 xy xz yz

Atoms

  1 1 5 0.3500 0.146 -0.338 0.449
  2 1 1 0.1000 1.093 -0.153 0.621
  3 1 5 0.1500 -0.030 -1.502 -0.536
  4 1 1 0.1000 -0.970 -1.774 -0.567
  5 1 5 0.1500 0.442 -1.115 -1.918
  6 1 1 0.1000 1.421 -1.078 -1.910
  7 1 5 0.3500 -0.084 0.248 -2.336
  8 1 1 0.1000 -1.041 0.143 -2.517
  9 1 5 0.1000 0.052 1.299 -1.279
10 1 1 0.1000 1.003 1.484 -1.131
11 1 5 0.0500 -0.662 2.597 -1.587
12 1 1 0.1000 -1.578 2.411 -1.846
13 1 1 0.1000 -0.679 3.155 -0.794
14 1 4 -0.6500 0.768 -2.562 -0.003
15 1 2 0.4000 0.584 -2.602 0.955
16 1 4 -0.6600 0.035 -2.124 -2.864
17 1 2 0.4000 0.594 -1.955 -3.651
18 1 4 -0.5000 0.531 0.705 -3.554
19 1 3 -0.4000 -0.525 0.807 -0.055
20 1 3 -0.6500 -0.003 3.283 -2.637
21 1 2 0.4000 -0.573 4.012 -2.958
22 1 5 0.1500 -0.146 0.338 -4.741
23 1 4 -0.5000 -0.531 -0.705 1.636
24 2 5 0.3500 3.723 3.705 3.159
25 2 1 0.1000 4.688 3.761 3.322
26 2 5 0.1500 3.396 2.589 2.174
27 2 1 0.1000 2.423 2.491 2.130
28 2 5 0.1500 3.902 2.954 0.794
29 2 1 0.1000 4.877 3.058 0.823
30 2 5 0.1500 3.262 4.288 0.389
31 2 1 0.1000 2.291 4.187 0.300
32 2 5 0.1000 3.595 5.356 1.441
33 2 1 0.1000 4.553 5.565 1.422
34 2 5 0.0500 2.758 6.641 1.168
35 2 1 0.1000 1.867 6.390 0.879
36 2 1 0.1000 2.675 7.152 1.989
37 2 4 -0.6500 3.952 1.356 2.616
38 2 2 0.4000 3.770 1.298 3.579
39 2 3 -0.6500 3.569 1.895 -0.084
40 2 2 0.4000 3.717 2.227 -0.998
41 2 4 -0.5000 3.830 4.774 -0.837
42 2 3 -0.4000 3.207 4.946 2.739
43 2 4 -0.6500 3.364 7.444 0.175
44 2 2 0.4000 3.074 8.372 0.338
45 2 5 0.1500 3.126 4.442 -2.031
46 2 3 -0.5000 3.018 3.374 4.353

dump file
ITEM: TIMESTEP
0
ITEM: NUMBER OF ATOMS
46
ITEM: BOX BOUNDS
-0.935489 7.784
0 8.14747
0 10.38
ITEM: ATOMS
1 5 -0.789489 7.80947 0.449
2 1 0.157511 7.99447 0.621
3 5 6.81851 6.64547 9.844
4 1 5.87851 6.37347 9.813
5 5 -0.493489 7.03247 8.462
6 1 0.485511 7.06947 8.47
7 5 7.7 0.248 8.044
8 1 6.743 0.143 7.863
9 5 0.052 1.299 9.101
10 1 1.003 1.484 9.249
11 5 7.122 2.597 8.793
12 1 6.206 2.411 8.534
13 1 7.105 3.155 9.586
14 4 -0.167489 5.58547 10.377
15 2 -0.351489 5.54547 0.955
16 4 6.88351 6.02347 7.516
17 2 -0.341489 6.19247 6.729
18 4 0.531 0.705 6.826
19 3 7.259 0.807 10.325
20 3 -0.003 3.283 7.743
21 2 7.211 4.012 7.422
22 5 7.638 0.338 5.639
23 4 6.31751 7.44247 1.636
24 5 3.723 3.705 3.159
25 1 4.688 3.761 3.322
26 5 3.396 2.589 2.174
27 1 2.423 2.491 2.13
28 5 3.902 2.954 0.794
29 1 4.877 3.058 0.823
30 5 3.262 4.288 0.389
31 1 2.291 4.187 0.3
32 5 3.595 5.356 1.441
33 1 4.553 5.565 1.422
34 5 2.758 6.641 1.168
35 1 1.867 6.39 0.879
36 1 2.675 7.152 1.989
37 4 3.952 1.356 2.616
38 2 3.77 1.298 3.579
39 3 3.569 1.895 10.296
40 2 3.717 2.227 9.382
41 4 3.83 4.774 9.543
42 3 3.207 4.946 2.739
43 4 3.364 7.444 0.175
44 2 4.00949 0.22453 0.338
45 5 3.126 4.442 8.349
46 3 3.018 3.374 4.353

I think what you are saying is that the dump on timestep 0 lists the atoms
with different coords than the data file. This is b/c the atoms in the
data file are not inside the box you have defined. For example, atom 1
is outside the box in y. LAMMPS remaps the atoms inside the periodic
box when it reads the data file. In this case, the box is a tilted
triclinic, so
the mapping may be non-obvious. But it won't "break" any bonds that
overlap the periodic boundaries.

Steve