lijiext
/
lammps
mirror of https://github.com/lammps/lammps.git
1
0
Fork 1
Code Issues Packages Projects Releases Wiki Activity

Update minimization documentation to reflect line quadratic default 

git-svn-id: svn://svn.icms.temple.edu/lammps-ro/trunk@12530 f3b2605a-c512-4ea7-a41b-209d697bcdaa
This commit is contained in:
athomps 2014-09-24 00:43:55 +00:00
parent f45b388d12
commit edd0d2dea7
3 changed files with 25 additions and 15 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

7
doc/fix_box_relax.html
View File

@ -261,8 +261,11 @@ volume, it can be difficult for the minimizer to equilibrate the
system the desired pressure with high precision, particularly for system the desired pressure with high precision, particularly for
solids. Some techniques that seem to help are (a) use the solids. Some techniques that seem to help are (a) use the
"min_modify line quadratic" option when minimizing with box "min_modify line quadratic" option when minimizing with box
relaxations, and (b) minimize several times in succession if need be, relaxations, (b) minimize several times in succession if need be,
to drive the pressure closer to the target pressure. Also note that to drive the pressure closer to the target pressure, (c) relax
the atom positions before relaxing the box, and (d) relax the box
to the target hydrostatic pressure before relaxing to a target
shear stress state. Also note that
some systems (e.g. liquids) will not sustain a non-hydrostatic applied some systems (e.g. liquids) will not sustain a non-hydrostatic applied
pressure, which means the minimizer will not converge. pressure, which means the minimizer will not converge.
</P> </P>

7
doc/fix_box_relax.txt
View File

@ -254,8 +254,11 @@ volume, it can be difficult for the minimizer to equilibrate the
system the desired pressure with high precision, particularly for system the desired pressure with high precision, particularly for
solids. Some techniques that seem to help are (a) use the solids. Some techniques that seem to help are (a) use the
"min_modify line quadratic" option when minimizing with box "min_modify line quadratic" option when minimizing with box
relaxations, and (b) minimize several times in succession if need be, relaxations, (b) minimize several times in succession if need be,
to drive the pressure closer to the target pressure. Also note that to drive the pressure closer to the target pressure, (c) relax
the atom positions before relaxing the box, and (d) relax the box
to the target hydrostatic pressure before relaxing to a target
shear stress state. Also note that
some systems (e.g. liquids) will not sustain a non-hydrostatic applied some systems (e.g. liquids) will not sustain a non-hydrostatic applied
pressure, which means the minimizer will not converge. pressure, which means the minimizer will not converge.

26
doc/min_modify.txt
View File

@ -44,8 +44,20 @@ highly overlapped atoms from being moved long distances (e.g. through
another atom) due to large forces. another atom) due to large forces.
The choice of line search algorithm for the {cg} and {sd} minimization The choice of line search algorithm for the {cg} and {sd} minimization
styles can be selected via the {line} keyword. The default styles can be selected via the {line} keyword.
backtracking search is robust and should always find a local energy The default {quadratic} line search algorithm starts out using
the robust backtracking method described below. However, once
the system gets close to a local
minimum and the linesearch steps get small, so that the energy
is approximately quadratic in the step length, it uses the
estimated location of zero gradient as the linesearch step,
provided the energy change is downhill.
This becomes more efficient than backtracking
for highly-converged relaxations. The {forcezero}
line search algorithm is similar to {quadratic}.
It may be more efficient than {quadratic} on some systems.
The backtracking search is robust and should always find a local energy
minimum. However, it will "converge" when it can no longer reduce the minimum. However, it will "converge" when it can no longer reduce the
energy of the system. Individual atom forces may still be larger than energy of the system. Individual atom forces may still be larger than
desired at this point, because the energy change is measured as the desired at this point, because the energy change is measured as the
@ -53,14 +65,6 @@ difference of two large values (energy before and energy after) and
that difference may be smaller than machine epsilon even if atoms that difference may be smaller than machine epsilon even if atoms
could move in the gradient direction to reduce forces further. could move in the gradient direction to reduce forces further.
By contrast, the {quadratic} line search algorithm tries to
reduce the forces to zero, while guaranteeing that the energy
changes is not positive (uphill). For some systems, it may also
be more efficient than the backtracking algorithm by
requiring fewer energy/force evaluations. The {forcezero}
line search algorithm is similar to {quadratic}.
It may be more efficient than {quadratic} on some systems.
[Restrictions:] none [Restrictions:] none
[Related commands:] [Related commands:]
@ -69,4 +73,4 @@ It may be more efficient than {quadratic} on some systems.
[Default:] [Default:]
The option defaults are dmax = 0.1 and line = backtrack. The option defaults are dmax = 0.1 and line = quadratic.