56
56
velocity profile created by fix nvt/sllod. LAMMPS will give an
57
57
error if this setting is not consistent.
59
<P>The SLLOD equations of motion coupled to a Nose/Hoover thermostat are
60
discussed in <A HREF = "#Tuckerman">(Tuckerman)</A> (eqs 4 and 5), which is what is
61
implemented in LAMMPS in a velocity Verlet formulation.
59
<P>The SLLOD equations of motion, originally proposed by Hoover and Ladd
60
(see <A HREF = "#Evans">(Evans and Morriss)</A>), were proven to be equivalent to
61
Newton's equations of motion for shear flow by <A HREF = "#Evans">(Evans and
62
Morriss)</A>. They were later shown to generate the desired
63
velocity gradient and the correct production of work by stresses for
64
all forms of homogeneous flow by <A HREF = "#Daivis">(Daivis and Todd)</A>. As
65
implemented in LAMMPS, they are coupled to a Nose/Hoover chain
66
thermostat in a velocity Verlet formulation, closely following the
67
implementation used for the <A HREF = "fix_nh.html">fix nvt</A> command.
63
69
<P>Additional parameters affecting the thermostat are specified by
64
70
keywords and values documented with the <A HREF = "fix_nh.html">fix nvt</A>
106
<P>Styles with a <I>cuda</I>, <I>gpu</I>, <I>omp</I>, or <I>opt</I> suffix are functionally
107
the same as the corresponding style without the suffix. They have
108
been optimized to run faster, depending on your available hardware, as
109
discussed in <A HREF = "Section_accelerate.html">Section_accelerate</A> of the
110
manual. The accelerated styles take the same arguments and should
111
produce the same results, except for round-off and precision issues.
112
<P>Styles with a <I>cuda</I>, <I>gpu</I>, <I>intel</I>, <I>kk</I>, <I>omp</I>, or <I>opt</I> suffix are
113
functionally the same as the corresponding style without the suffix.
114
They have been optimized to run faster, depending on your available
115
hardware, as discussed in <A HREF = "Section_accelerate.html">Section_accelerate</A>
116
of the manual. The accelerated styles take the same arguments and
117
should produce the same results, except for round-off and precision
113
<P>These accelerated styles are part of the USER-CUDA, GPU, USER-OMP and OPT
114
packages, respectively. They are only enabled if LAMMPS was built with
115
those packages. See the <A HREF = "Section_start.html#start_3">Making LAMMPS</A>
116
section for more info.
120
<P>These accelerated styles are part of the USER-CUDA, GPU, USER-INTEL,
121
KOKKOS, USER-OMP and OPT packages, respectively. They are only
122
enabled if LAMMPS was built with those packages. See the <A HREF = "Section_start.html#start_3">Making
123
LAMMPS</A> section for more info.
118
125
<P>You can specify the accelerated styles explicitly in your input script
119
126
by including their suffix, or you can use the <A HREF = "Section_start.html#start_7">-suffix command-line
172
<A NAME = "Tuckerman"></A>
174
<P><B>(Tuckerman)</B> Tuckerman, Mundy, Balasubramanian, Klein, J Chem Phys,
179
<A NAME = "Evans"></A>
181
<P><B>(Evans and Morriss)</B> Evans and Morriss, Phys Rev A, 30, 1528 (1984).
183
<A NAME = "Daivis"></A>
185
<P><B>(Daivis and Todd)</B> Daivis and Todd, J Chem Phys, 124, 194103 (2006).