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 pair_style dpd/fdt command :h3
10 pair_style dpd/fdt/energy command :h3
14 pair_style style args :pre
16 style = {dpd/fdt} or {dpd/fdt/energy}
17 args = list of arguments for a particular style :ul
18 {dpd/fdt} args = T cutoff seed
19 T = temperature (temperature units)
20 cutoff = global cutoff for DPD interactions (distance units)
21 seed = random # seed (positive integer)
22 {dpd/fdt/energy} args = cutoff seed
23 cutoff = global cutoff for DPD interactions (distance units)
24 seed = random # seed (positive integer) :pre
28 pair_style dpd/fdt 300.0 2.5 34387
29 pair_coeff * * 3.0 1.0 2.5 :pre
31 pair_style dpd/fdt/energy 2.5 34387
32 pair_coeff * * 3.0 1.0 0.1 2.5 :pre
36 Styles {dpd/fdt} and {dpd/fdt/energy} compute the force for dissipative
37 particle dynamics (DPD) simulations. The {dpd/fdt} style is used to
38 perform DPD simulations under isothermal and isobaric conditions,
39 while the {dpd/fdt/energy} style is used to perform DPD simulations
40 under isoenergetic and isoenthalpic conditions (see "(Lisal)"_#Lisal).
41 For DPD simulations in general, the force on atom I due to atom J is
42 given as a sum of 3 terms
44 :c,image(Eqs/pair_dpd.jpg)
46 where Fc is a conservative force, Fd is a dissipative force, and Fr is
47 a random force. Rij is a unit vector in the direction Ri - Rj, Vij is
48 the vector difference in velocities of the two atoms = Vi - Vj, alpha
49 is a Gaussian random number with zero mean and unit variance, dt is
50 the timestep size, and w(r) is a weighting factor that varies between
51 0 and 1. Rc is the cutoff. The weighting factor, omega_ij, varies
52 between 0 and 1, and is chosen to have the following functional form:
54 :c,image(Eqs/pair_dpd_omega.jpg)
56 Note that alternative definitions of the weighting function exist, but
57 would have to be implemented as a separate pair style command.
59 For style {dpd/fdt}, the fluctuation-dissipation theorem defines gamma
60 to be set equal to sigma*sigma/(2 T), where T is the set point
61 temperature specified as a pair style parameter in the above examples.
62 The following coefficients must be defined for each pair of atoms types
63 via the "pair_coeff"_pair_coeff.html command as in the examples above,
64 or in the data file or restart files read by the
65 "read_data"_read_data.html or "read_restart"_read_restart.html commands:
68 sigma (force*time^(1/2) units)
69 cutoff (distance units) :ul
71 The last coefficient is optional. If not specified, the global DPD
74 Style {dpd/fdt/energy} is used to perform DPD simulations
75 under isoenergetic and isoenthalpic conditions. The fluctuation-dissipation
76 theorem defines gamma to be set equal to sigma*sigma/(2 dpdTheta), where
77 dpdTheta is the average internal temperature for the pair. The particle
78 internal temperature is related to the particle internal energy through
79 a mesoparticle equation of state (see "fix eos"_fix.html). The
80 differential internal conductive and mechanical energies are computed
81 within style {dpd/fdt/energy} as:
83 :c,image(Eqs/pair_dpd_energy.jpg)
87 :c,image(Eqs/pair_dpd_energy_terms.jpg)
89 Zeta_ij^q is a second Gaussian random number with zero mean and unit
90 variance that is used to compute the internal conductive energy. The
91 fluctuation-dissipation theorem defines alpha*alpha to be set
92 equal to 2*kB*kappa, where kappa is the mesoparticle thermal
93 conductivity parameter. The following coefficients must be defined for
94 each pair of atoms types via the "pair_coeff"_pair_coeff.html
95 command as in the examples above, or in the data file or restart files
96 read by the "read_data"_read_data.html or "read_restart"_read_restart.html
100 sigma (force*time^(1/2) units)
101 kappa (energy*temperature/time units)
102 cutoff (distance units) :ul
104 The last coefficient is optional. If not specified, the global DPD
107 The pairwise energy associated with styles {dpd/fdt} and
108 {dpd/fdt/energy} is only due to the conservative force term Fc, and is
109 shifted to be zero at the cutoff distance Rc. The pairwise virial is
110 calculated using only the conservative term.
112 The forces computed through the {dpd/fdt} and {dpd/fdt/energy} styles
113 can be integrated with the velocity-Verlet integration scheme or the
114 Shardlow splitting integration scheme described by "(Lisal)"_#Lisal.
115 In the cases when these pair styles are combined with the
116 "fix shardlow"_fix_shardlow.html, these pair styles differ from the
117 other dpd styles in that the dissipative and random forces are split
118 from the force calculation and are not computed within the pair style.
119 Thus, only the conservative force is computed by the pair style,
120 while the stochastic integration of the dissipative and random forces
121 are handled through the Shardlow splitting algorithm approach. The
122 Shardlow splitting algorithm is advantageous, especially when
123 performing DPD under isoenergetic conditions, as it allows
124 significantly larger timesteps to be taken.
130 These commands are part of the USER-DPD package. They are only
131 enabled if LAMMPS was built with that package. See the "Making
132 LAMMPS"_Section_start.html#start_3 section for more info.
134 Pair styles {dpd/fdt} and {dpd/fdt/energy} require use of the
135 "comm_modify vel yes"_comm_modify.html option so that velocites are
136 stored by ghost atoms.
138 Pair style {dpd/fdt/energy} requires "atom_style dpd"_atom_style.html
139 to be used in order to properly account for the particle internal
140 energies and temperatures.
144 "pair_coeff"_pair_coeff.html, "fix shardlow"_fix_shardlow.html
151 [(Lisal)] M. Lisal, J.K. Brennan, J. Bonet Avalos, "Dissipative
152 particle dynamics at isothermal, isobaric, isoenergetic, and
153 isoenthalpic conditions using Shardlow-like splitting algorithms.",
154 J. Chem. Phys., 135, 204105 (2011).