1 "LAMMPS WWW Site"_lws - "LAMMPS Documentation"_ld - "LAMMPS Commands"_lc :c
3 :link(lws,http://lammps.sandia.gov)
5 :link(lc,Section_commands.html#comm)
9 fix nve/limit command :h3
13 fix ID group-ID nve/limit xmax :pre
15 ID, group-ID are documented in "fix"_fix.html command
16 nve = style name of this fix command
17 xmax = maximum distance an atom can move in one timestep (distance units) :ul
21 fix 1 all nve/limit 0.1 :pre
25 Perform constant NVE updates of position and velocity for atoms in the
26 group each timestep. A limit is imposed on the maximum distance an
27 atom can move in one timestep. This is useful when starting a
28 simulation with a configuration containing highly overlapped atoms.
29 Normally this would generate huge forces which would blow atoms out of
30 the simulation box, causing LAMMPS to stop with an error.
32 Using this fix can overcome that problem. Forces on atoms must still
33 be computable (which typically means 2 atoms must have a separation
34 distance > 0.0). But large velocities generated by large forces are
35 reset to a value that corresponds to a displacement of length {xmax}
36 in a single timestep. {Xmax} is specified in distance units; see the
37 "units"_units.html command for details. The value of {xmax} should be
38 consistent with the neighbor skin distance and the frequency of
39 neighbor list re-building, so that pairwise interactions are not
40 missed on successive timesteps as atoms move. See the
41 "neighbor"_neighbor.html and "neigh_modify"_neigh_modify.html commands
44 Note that if a velocity reset occurs the integrator will not conserve
45 energy. On steps where no velocity resets occur, this integrator is
46 exactly like the "fix nve"_fix_nve.html command. Since forces are
47 unaltered, pressures computed by thermodynamic output will still be
48 very large for overlapped configurations.
50 NOTE: You should not use "fix shake"_fix_shake.html in conjunction
51 with this fix. That is because fix shake applies contraint forces
52 based on the predicted postitions of atoms after the next timestep.
53 It has no way of knowing the timestep may change due to this fix,
54 which will cause the constraint forces to be invalid. A better
55 strategy is to turn off fix shake when performing initial dynamics
56 that need this fix, then turn fix shake on when doing normal dynamics
57 with a fixed-size timestep.
59 [Restart, fix_modify, output, run start/stop, minimize info:]
61 No information about this fix is written to "binary restart
62 files"_restart.html. None of the "fix_modify"_fix_modify.html options
63 are relevant to this fix.
65 This fix computes a global scalar which can be accessed by various
66 "output commands"_Section_howto.html#howto_15. The scalar is the
67 count of how many updates of atom's velocity/position were limited by
68 the maximum distance criterion. This should be roughly the number of
69 atoms so affected, except that updates occur at both the beginning and
70 end of a timestep in a velocity Verlet timestepping algorithm. This
71 is a cumulative quantity for the current run, but is re-initialized to
72 zero each time a run is performed. The scalar value calculated by
73 this fix is "extensive".
75 No parameter of this fix can be used with the {start/stop} keywords of
76 the "run"_run.html command. This fix is not invoked during "energy
77 minimization"_minimize.html.
83 "fix nve"_fix_nve.html, "fix nve/noforce"_fix_nve_noforce.html,
84 "pair_style soft"_pair_soft.html