tx258/lammps-gran-kokkos
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

small updates to the QEQ package for better testing

Browse Source 
- add a "warn no/yes" keyword/value pair to allow turning of the convergence warning
- add a scalar compute to retrieve the number of QEq itration from the fix
- update the buck example input to run all QEq methods from a common restart
- document changes
This commit is contained in:
Axel Kohlmeyer
2021-04-20 14:00:51 -04:00
parent 6c4ad8ad35
commit 799fb284c8
14 changed files with 1446 additions and 272 deletions
Download Patch File Download Diff File Expand all files Collapse all files

19
doc/src/fix_qeq.rst
View File

@ -34,13 +34,14 @@ Syntax
* maxiter = maximum iterations to perform charge equilibration
* qfile = a filename with QEq parameters or *coul/streitz* or *reax/c*
* zero or more keyword/value pairs may be appended
* keyword = *alpha* or *qdamp* or *qstep*
* keyword = *alpha* or *qdamp* or *qstep* or *warn*
.. parsed-literal::
*alpha* value = Slater type orbital exponent (qeq/slater only)
*qdamp* value = damping factor for damped dynamics charge solver (qeq/dynamic and qeq/fire only)
*qstep* value = time step size for damped dynamics charge solver (qeq/dynamic and qeq/fire only)
*warn* value = do (=yes) or do not (=no) print a warning when the maximum number of iterations is reached
Examples
""""""""
@ -101,8 +102,8 @@ The QEq method minimizes the electrostatic energy of the system (or
equalizes the derivative of energy with respect to charge of all the
atoms) by adjusting the partial charge on individual atoms based on
interactions with their neighbors within *cutoff*\ . It requires a few
parameters, in *metal* units, for each atom type which provided in a
file specified by *qfile*\ . The file has the following format
parameters in the appropriate units for each atom type which are read
from a file specified by *qfile*\ . The file has the following format
.. parsed-literal::
@ -112,7 +113,7 @@ file specified by *qfile*\ . The file has the following format
Ntype chi eta gamma zeta qcore
There have to be parameters given for every atom type. Wildcard entries
are possible using the same syntax as elsewhere in LAMMPS
are possible using the same type range syntax as for "coeff" commands
(i.e., n\*m, n\*, \*m, \*). Later entries will overwrite previous ones.
Empty lines or any text following the pound sign (#) are ignored.
Each line starts with the atom type followed by five parameters.
@ -126,6 +127,14 @@ entries per line are required.
* *zeta* = Slater type orbital exponent defined by the :ref:`Streitz-Mintmire <Streitz1>` potential in reverse distance units
* *qcore* = charge of the nucleus defined by the :ref:`Streitz-Mintmire potential <Streitz1>` potential in charge units
The fix qeq styles will print a warning if the charges are not
equilibrated within *tolerance* by *maxiter* steps, unless the
*warn* keyword is used with "no" as argument. This latter option
may be useful for testing and benchmarking purposes, as it allows
to use a fixed number of QEq iterations when *tolerance* is set
to a small enough value to alway reach the *maxiter* limit. Turning
off warnings will avoid the excessive output in that case.
The *qeq/point* style describes partial charges on atoms as point
charges. Interaction between a pair of charged particles is 1/r,
which is the simplest description of the interaction between charges.
@ -229,7 +238,7 @@ Related commands
Default
"""""""
none
warn yes
----------

22
examples/qeq/buck.inc Normal file
View File

@ -0,0 +1,22 @@
kspace_style pppm 1e-6
neighbor 1.0 bin
neigh_modify delay 0 every 1 check yes
group type1 type 1
compute charge1 type1 property/atom q
compute q1 type1 reduce ave c_charge1
group type2 type 2
compute charge2 type2 property/atom q
compute q2 type2 reduce ave c_charge2
variable qtot equal count(type1)*c_q1+count(type2)*c_q2
variable nqeq equal f_2
thermo_style custom step pe c_q1 c_q2 v_qtot v_nqeq
thermo 10
thermo_modify format line "%4d %12.9g %12.8f %12.8f %16.12f %6.0f"
timestep 0.0001
fix 1 all nve

75
examples/qeq/in.qeq.buck
View File

@ -1,41 +1,68 @@
# This example demonstrates the use of various fix qeq variants with
# that defines and uses charges, in this case pair_style buck/coul/long
# a pair style using charges, in this case pair_style buck/coul/long
units metal
atom_style charge
read_data data.aC
replicate 2 2 2
#replicate 2 2 2
pair_style buck/coul/long 12.0
pair_coeff 2 2 1388.77 .3623188 175.0
pair_coeff 1 2 18003 .2052124 133.5381
pair_coeff 1 1 0 .1 0
kspace_style ewald 1e-6
neighbor 1.0 bin
neigh_modify delay 0 every 1 check yes
fix 2 all qeq/shielded 1 10 1.0e-20 10 param.qeq2
group type1 type 1
compute charge1 type1 property/atom q
compute q1 type1 reduce ave c_charge1
group type2 type 2
compute charge2 type2 property/atom q
compute q2 type2 reduce ave c_charge2
variable qtot equal count(type1)*c_q1+count(type2)*c_q2
thermo_style custom step pe c_q1 c_q2 v_qtot
thermo 10
timestep 0.0001
include buck.inc
velocity all create 300.0 1281937
fix 1 all nve
run 0 post no
#fix 2 all qeq/point 1 10 1.0e-6 100 param.qeq2
#fix 2 all qeq/shielded 1 10 1.0e-6 100 param.qeq2
#fix 2 all qeq/slater 1 10 1.0e-6 100 param.qeq2
#fix 2 all qeq/dynamic 1 10 1.0e-4 100 param.qeq2
fix 2 all qeq/fire 1 10 1.0e-4 100 param.qeq2
write_restart qeq.restart
run 100
clear
print "Using fix qeq/point"
read_restart qeq.restart
fix 2 all qeq/point 1 10 1.0e-6 100 param.qeq2
include buck.inc
run 100
clear
print "Using fix qeq/shielded"
read_restart qeq.restart
fix 2 all qeq/shielded 1 10 1.0e-6 100 param.qeq2
include buck.inc
run 100
clear
print "Using fix qeq/slater"
read_restart qeq.restart
fix 2 all qeq/shielded 1 10 1.0e-6 100 param.qeq2
include buck.inc
run 100
clear
print "Using fix qeq/dynamic"
read_restart qeq.restart
fix 2 all qeq/dynamic 1 10 1.0e-3 100 param.qeq2
include buck.inc
run 100
clear
print "Using fix qeq/fire"
read_restart qeq.restart
fix 2 all qeq/fire 1 10 1.0e-3 100 param.qeq2
include buck.inc
run 100

650
examples/qeq/log.20Apr21.qeq.buck.g++.1 Normal file
View File

@ -0,0 +1,650 @@
LAMMPS (8 Apr 2021)
using 1 OpenMP thread(s) per MPI task
# This example demonstrates the use of various fix qeq variants with
# a pair style using charges, in this case pair_style buck/coul/long
units metal
atom_style charge
read_data data.aC
Reading data file ...
orthogonal box = (0.0000000 0.0000000 0.0000000) to (25.158320 25.158320 28.020256)
1 by 1 by 1 MPI processor grid
reading atoms ...
1200 atoms
read_data CPU = 0.009 seconds
#replicate 2 2 2
pair_style buck/coul/long 12.0
pair_coeff 2 2 1388.77 .3623188 175.0
pair_coeff 1 2 18003 .2052124 133.5381
pair_coeff 1 1 0 .1 0
fix 2 all qeq/shielded 1 10 1.0e-20 10 param.qeq2
include buck.inc
kspace_style pppm 1e-6
neighbor 1.0 bin
neigh_modify delay 0 every 1 check yes
group type1 type 1
400 atoms in group type1
compute charge1 type1 property/atom q
compute q1 type1 reduce ave c_charge1
group type2 type 2
800 atoms in group type2
compute charge2 type2 property/atom q
compute q2 type2 reduce ave c_charge2
variable qtot equal count(type1)*c_q1+count(type2)*c_q2
variable nqeq equal f_2
thermo_style custom step pe c_q1 c_q2 v_qtot v_nqeq
thermo 10
thermo_modify format line "%4d %12.9g %12.8f %12.8f %16.12f %6.0f"
timestep 0.0001
fix 1 all nve
velocity all create 300.0 1281937
run 0 post no
PPPM initialization ...
using 12-bit tables for long-range coulomb (src/kspace.cpp:339)
G vector (1/distance) = 0.30705229
grid = 48 48 54
stencil order = 5
estimated absolute RMS force accuracy = 1.8909403e-05
estimated relative force accuracy = 1.3131854e-06
using double precision FFTW3
3d grid and FFT values/proc = 184525 124416
Neighbor list info ...
update every 1 steps, delay 0 steps, check yes
max neighbors/atom: 2000, page size: 100000
master list distance cutoff = 13
ghost atom cutoff = 13
binsize = 6.5, bins = 4 4 5
2 neighbor lists, perpetual/occasional/extra = 2 0 0
(1) pair buck/coul/long, perpetual, half/full from (2)
attributes: half, newton on
pair build: halffull/newton
stencil: none
bin: none
(2) fix qeq/shielded, perpetual
attributes: full, newton on
pair build: full/bin/atomonly
stencil: full/bin/3d
bin: standard
WARNING: Fix qeq CG convergence failed (4.357840257025601e-19) after 10 iterations at step 0 (src/QEQ/fix_qeq.cpp:410)
WARNING: Fix qeq CG convergence failed (5.274094378414531e-18) after 10 iterations at step 0 (src/QEQ/fix_qeq.cpp:410)
Per MPI rank memory allocation (min/avg/max) = 38.75 | 38.75 | 38.75 Mbytes
Step PotEng c_q1 c_q2 v_qtot v_nqeq
0 -2879.00327 0.76536977 -0.38268489 0.000000000000 10
Loop time of 1.66893e-06 on 1 procs for 0 steps with 1200 atoms
write_restart qeq.restart
System init for write_restart ...
PPPM initialization ...
using 12-bit tables for long-range coulomb (src/kspace.cpp:339)
G vector (1/distance) = 0.27644401
grid = 27 27 30
stencil order = 5
estimated absolute RMS force accuracy = 1.4502702e-05
estimated relative force accuracy = 1.0071569e-06
using double precision FFTW3
3d grid and FFT values/proc = 42772 21870
clear
using 1 OpenMP thread(s) per MPI task
print "Using fix qeq/point"
Using fix qeq/point
read_restart qeq.restart
Reading restart file ...
restart file = 8 Apr 2021, LAMMPS = 8 Apr 2021
restoring atom style charge from restart
orthogonal box = (0.0000000 0.0000000 0.0000000) to (25.158320 25.158320 28.020256)
1 by 1 by 1 MPI processor grid
restoring pair style buck/coul/long from restart
1200 atoms
read_restart CPU = 0.001 seconds
fix 2 all qeq/point 1 10 1.0e-6 100 param.qeq2
include buck.inc
kspace_style pppm 1e-6
neighbor 1.0 bin
neigh_modify delay 0 every 1 check yes
group type1 type 1
400 atoms in group type1
compute charge1 type1 property/atom q
compute q1 type1 reduce ave c_charge1
group type2 type 2
800 atoms in group type2
compute charge2 type2 property/atom q
compute q2 type2 reduce ave c_charge2
variable qtot equal count(type1)*c_q1+count(type2)*c_q2
variable nqeq equal f_2
thermo_style custom step pe c_q1 c_q2 v_qtot v_nqeq
thermo 10
thermo_modify format line "%4d %12.9g %12.8f %12.8f %16.12f %6.0f"
timestep 0.0001
fix 1 all nve
run 100
PPPM initialization ...
using 12-bit tables for long-range coulomb (src/kspace.cpp:339)
G vector (1/distance) = 0.27644401
grid = 27 27 30
stencil order = 5
estimated absolute RMS force accuracy = 1.4502702e-05
estimated relative force accuracy = 1.0071569e-06
using double precision FFTW3
3d grid and FFT values/proc = 42772 21870
Neighbor list info ...
update every 1 steps, delay 0 steps, check yes
max neighbors/atom: 2000, page size: 100000
master list distance cutoff = 13
ghost atom cutoff = 13
binsize = 6.5, bins = 4 4 5
2 neighbor lists, perpetual/occasional/extra = 2 0 0
(1) pair buck/coul/long, perpetual, half/full from (2)
attributes: half, newton on
pair build: halffull/newton
stencil: none
bin: none
(2) fix qeq/point, perpetual
attributes: full, newton on
pair build: full/bin/atomonly
stencil: full/bin/3d
bin: standard
Per MPI rank memory allocation (min/avg/max) = 24.69 | 24.69 | 24.69 Mbytes
Step PotEng c_q1 c_q2 v_qtot v_nqeq
0 -3432.17988 0.85228288 -0.42614144 -0.000000000000 3
10 -3452.03328 0.85475605 -0.42737803 -0.000000000000 8
20 -3497.57515 0.85994936 -0.42997468 0.000000000000 8
30 -3568.22095 0.86767937 -0.43383969 0.000000000001 8
40 -3633.24956 0.87335551 -0.43667775 0.000000000000 8
50 -3700.10219 0.87805056 -0.43902528 0.000000000000 8
60 -3784.36769 0.88402303 -0.44201151 -0.000000000000 8
70 -3877.51378 0.89008950 -0.44504475 0.000000000000 8
80 -3965.29722 0.89431515 -0.44715757 -0.000000000001 8
90 -4048.36764 0.89698588 -0.44849294 0.000000000000 8
100 -4118.65809 0.89719102 -0.44859551 0.000000000000 8
Loop time of 11.5935 on 1 procs for 100 steps with 1200 atoms
Performance: 0.075 ns/day, 322.041 hours/ns, 8.626 timesteps/s
99.9% CPU use with 1 MPI tasks x 1 OpenMP threads
MPI task timing breakdown:
Section | min time | avg time | max time |%varavg| %total
---------------------------------------------------------------
Pair | 2.8257 | 2.8257 | 2.8257 | 0.0 | 24.37
Kspace | 1.2136 | 1.2136 | 1.2136 | 0.0 | 10.47
Neigh | 0 | 0 | 0 | 0.0 | 0.00
Comm | 0.015541 | 0.015541 | 0.015541 | 0.0 | 0.13
Output | 0.0014489 | 0.0014489 | 0.0014489 | 0.0 | 0.01
Modify | 7.5351 | 7.5351 | 7.5351 | 0.0 | 64.99
Other | | 0.00206 | | | 0.02
Nlocal: 1200.00 ave 1200 max 1200 min
Histogram: 1 0 0 0 0 0 0 0 0 0
Nghost: 8100.00 ave 8100 max 8100 min
Histogram: 1 0 0 0 0 0 0 0 0 0
Neighs: 367600.0 ave 367600 max 367600 min
Histogram: 1 0 0 0 0 0 0 0 0 0
FullNghs: 735200.0 ave 735200 max 735200 min
Histogram: 1 0 0 0 0 0 0 0 0 0
Total # of neighbors = 735200
Ave neighs/atom = 612.66667
Neighbor list builds = 0
Dangerous builds = 0
clear
using 1 OpenMP thread(s) per MPI task
print "Using fix qeq/shielded"
Using fix qeq/shielded
read_restart qeq.restart
Reading restart file ...
restart file = 8 Apr 2021, LAMMPS = 8 Apr 2021
restoring atom style charge from restart
orthogonal box = (0.0000000 0.0000000 0.0000000) to (25.158320 25.158320 28.020256)
1 by 1 by 1 MPI processor grid
restoring pair style buck/coul/long from restart
1200 atoms
read_restart CPU = 0.001 seconds
fix 2 all qeq/shielded 1 10 1.0e-6 100 param.qeq2
include buck.inc
kspace_style pppm 1e-6
neighbor 1.0 bin
neigh_modify delay 0 every 1 check yes
group type1 type 1
400 atoms in group type1
compute charge1 type1 property/atom q
compute q1 type1 reduce ave c_charge1
group type2 type 2
800 atoms in group type2
compute charge2 type2 property/atom q
compute q2 type2 reduce ave c_charge2
variable qtot equal count(type1)*c_q1+count(type2)*c_q2
variable nqeq equal f_2
thermo_style custom step pe c_q1 c_q2 v_qtot v_nqeq
thermo 10
thermo_modify format line "%4d %12.9g %12.8f %12.8f %16.12f %6.0f"
timestep 0.0001
fix 1 all nve
run 100
PPPM initialization ...
using 12-bit tables for long-range coulomb (src/kspace.cpp:339)
G vector (1/distance) = 0.27644401
grid = 27 27 30
stencil order = 5
estimated absolute RMS force accuracy = 1.4502702e-05
estimated relative force accuracy = 1.0071569e-06
using double precision FFTW3
3d grid and FFT values/proc = 42772 21870
Neighbor list info ...
update every 1 steps, delay 0 steps, check yes
max neighbors/atom: 2000, page size: 100000
master list distance cutoff = 13
ghost atom cutoff = 13
binsize = 6.5, bins = 4 4 5
2 neighbor lists, perpetual/occasional/extra = 2 0 0
(1) pair buck/coul/long, perpetual, half/full from (2)
attributes: half, newton on
pair build: halffull/newton
stencil: none
bin: none
(2) fix qeq/shielded, perpetual
attributes: full, newton on
pair build: full/bin/atomonly
stencil: full/bin/3d
bin: standard
Per MPI rank memory allocation (min/avg/max) = 24.69 | 24.69 | 24.69 Mbytes
Step PotEng c_q1 c_q2 v_qtot v_nqeq
0 -2879.00309 0.76536977 -0.38268489 -0.000000000000 3
10 -2882.50998 0.76536972 -0.38268486 0.000000000000 2
20 -2893.89472 0.76536950 -0.38268475 0.000000000000 2
30 -2913.6181 0.76536875 -0.38268438 0.000000000001 1
40 -2942.24129 0.76536939 -0.38268470 -0.000000000001 1
50 -2980.18817 0.76536780 -0.38268390 0.000000000000 2
60 -3027.60957 0.76536804 -0.38268402 0.000000000000 2
70 -3084.12552 0.76536573 -0.38268287 0.000000000000 2
80 -3148.8697 0.76536550 -0.38268275 0.000000000001 1
90 -3220.43086 0.76536380 -0.38268190 0.000000000000 2
100 -3297.0618 0.76536251 -0.38268126 0.000000000000 2
Loop time of 7.93936 on 1 procs for 100 steps with 1200 atoms
Performance: 0.109 ns/day, 220.538 hours/ns, 12.595 timesteps/s
99.9% CPU use with 1 MPI tasks x 1 OpenMP threads
MPI task timing breakdown:
Section | min time | avg time | max time |%varavg| %total
---------------------------------------------------------------
Pair | 2.8061 | 2.8061 | 2.8061 | 0.0 | 35.34
Kspace | 1.2176 | 1.2176 | 1.2176 | 0.0 | 15.34
Neigh | 0 | 0 | 0 | 0.0 | 0.00
Comm | 0.015528 | 0.015528 | 0.015528 | 0.0 | 0.20
Output | 0.0014365 | 0.0014365 | 0.0014365 | 0.0 | 0.02
Modify | 3.8966 | 3.8966 | 3.8966 | 0.0 | 49.08
Other | | 0.002076 | | | 0.03
Nlocal: 1200.00 ave 1200 max 1200 min
Histogram: 1 0 0 0 0 0 0 0 0 0
Nghost: 8100.00 ave 8100 max 8100 min
Histogram: 1 0 0 0 0 0 0 0 0 0
Neighs: 367600.0 ave 367600 max 367600 min
Histogram: 1 0 0 0 0 0 0 0 0 0
FullNghs: 735200.0 ave 735200 max 735200 min
Histogram: 1 0 0 0 0 0 0 0 0 0
Total # of neighbors = 735200
Ave neighs/atom = 612.66667
Neighbor list builds = 0
Dangerous builds = 0
clear
using 1 OpenMP thread(s) per MPI task
print "Using fix qeq/slater"
Using fix qeq/slater
read_restart qeq.restart
Reading restart file ...
restart file = 8 Apr 2021, LAMMPS = 8 Apr 2021
restoring atom style charge from restart
orthogonal box = (0.0000000 0.0000000 0.0000000) to (25.158320 25.158320 28.020256)
1 by 1 by 1 MPI processor grid
restoring pair style buck/coul/long from restart
1200 atoms
read_restart CPU = 0.001 seconds
fix 2 all qeq/shielded 1 10 1.0e-6 100 param.qeq2
include buck.inc
kspace_style pppm 1e-6
neighbor 1.0 bin
neigh_modify delay 0 every 1 check yes
group type1 type 1
400 atoms in group type1
compute charge1 type1 property/atom q
compute q1 type1 reduce ave c_charge1
group type2 type 2
800 atoms in group type2
compute charge2 type2 property/atom q
compute q2 type2 reduce ave c_charge2
variable qtot equal count(type1)*c_q1+count(type2)*c_q2
variable nqeq equal f_2
thermo_style custom step pe c_q1 c_q2 v_qtot v_nqeq
thermo 10
thermo_modify format line "%4d %12.9g %12.8f %12.8f %16.12f %6.0f"
timestep 0.0001
fix 1 all nve
run 100
PPPM initialization ...
using 12-bit tables for long-range coulomb (src/kspace.cpp:339)
G vector (1/distance) = 0.27644401
grid = 27 27 30
stencil order = 5
estimated absolute RMS force accuracy = 1.4502702e-05
estimated relative force accuracy = 1.0071569e-06
using double precision FFTW3
3d grid and FFT values/proc = 42772 21870
Neighbor list info ...
update every 1 steps, delay 0 steps, check yes
max neighbors/atom: 2000, page size: 100000
master list distance cutoff = 13
ghost atom cutoff = 13
binsize = 6.5, bins = 4 4 5
2 neighbor lists, perpetual/occasional/extra = 2 0 0
(1) pair buck/coul/long, perpetual, half/full from (2)
attributes: half, newton on
pair build: halffull/newton
stencil: none
bin: none
(2) fix qeq/shielded, perpetual
attributes: full, newton on
pair build: full/bin/atomonly
stencil: full/bin/3d
bin: standard
Per MPI rank memory allocation (min/avg/max) = 24.69 | 24.69 | 24.69 Mbytes
Step PotEng c_q1 c_q2 v_qtot v_nqeq
0 -2879.00309 0.76536977 -0.38268489 -0.000000000000 3
10 -2882.50998 0.76536972 -0.38268486 0.000000000000 2
20 -2893.89472 0.76536950 -0.38268475 0.000000000000 2
30 -2913.6181 0.76536875 -0.38268438 0.000000000001 1
40 -2942.24129 0.76536939 -0.38268470 -0.000000000001 1
50 -2980.18817 0.76536780 -0.38268390 0.000000000000 2
60 -3027.60957 0.76536804 -0.38268402 0.000000000000 2
70 -3084.12552 0.76536573 -0.38268287 0.000000000000 2
80 -3148.8697 0.76536550 -0.38268275 0.000000000001 1
90 -3220.43086 0.76536380 -0.38268190 0.000000000000 2
100 -3297.0618 0.76536251 -0.38268126 0.000000000000 2
Loop time of 7.9652 on 1 procs for 100 steps with 1200 atoms
Performance: 0.108 ns/day, 221.256 hours/ns, 12.555 timesteps/s
99.9% CPU use with 1 MPI tasks x 1 OpenMP threads
MPI task timing breakdown:
Section | min time | avg time | max time |%varavg| %total
---------------------------------------------------------------
Pair | 2.809 | 2.809 | 2.809 | 0.0 | 35.27
Kspace | 1.2214 | 1.2214 | 1.2214 | 0.0 | 15.33
Neigh | 0 | 0 | 0 | 0.0 | 0.00
Comm | 0.015635 | 0.015635 | 0.015635 | 0.0 | 0.20
Output | 0.0014393 | 0.0014393 | 0.0014393 | 0.0 | 0.02
Modify | 3.9157 | 3.9157 | 3.9157 | 0.0 | 49.16
Other | | 0.002091 | | | 0.03
Nlocal: 1200.00 ave 1200 max 1200 min
Histogram: 1 0 0 0 0 0 0 0 0 0
Nghost: 8100.00 ave 8100 max 8100 min
Histogram: 1 0 0 0 0 0 0 0 0 0
Neighs: 367600.0 ave 367600 max 367600 min
Histogram: 1 0 0 0 0 0 0 0 0 0
FullNghs: 735200.0 ave 735200 max 735200 min
Histogram: 1 0 0 0 0 0 0 0 0 0
Total # of neighbors = 735200
Ave neighs/atom = 612.66667
Neighbor list builds = 0
Dangerous builds = 0
clear
using 1 OpenMP thread(s) per MPI task
print "Using fix qeq/dynamic"
Using fix qeq/dynamic
read_restart qeq.restart
Reading restart file ...
restart file = 8 Apr 2021, LAMMPS = 8 Apr 2021
restoring atom style charge from restart
orthogonal box = (0.0000000 0.0000000 0.0000000) to (25.158320 25.158320 28.020256)
1 by 1 by 1 MPI processor grid
restoring pair style buck/coul/long from restart
1200 atoms
read_restart CPU = 0.001 seconds
fix 2 all qeq/dynamic 1 10 1.0e-3 100 param.qeq2
include buck.inc
kspace_style pppm 1e-6
neighbor 1.0 bin
neigh_modify delay 0 every 1 check yes
group type1 type 1
400 atoms in group type1
compute charge1 type1 property/atom q
compute q1 type1 reduce ave c_charge1
group type2 type 2
800 atoms in group type2
compute charge2 type2 property/atom q
compute q2 type2 reduce ave c_charge2
variable qtot equal count(type1)*c_q1+count(type2)*c_q2
variable nqeq equal f_2
thermo_style custom step pe c_q1 c_q2 v_qtot v_nqeq
thermo 10
thermo_modify format line "%4d %12.9g %12.8f %12.8f %16.12f %6.0f"
timestep 0.0001
fix 1 all nve
run 100
PPPM initialization ...
using 12-bit tables for long-range coulomb (src/kspace.cpp:339)
G vector (1/distance) = 0.27644401
grid = 27 27 30
stencil order = 5
estimated absolute RMS force accuracy = 1.4502702e-05
estimated relative force accuracy = 1.0071569e-06
using double precision FFTW3
3d grid and FFT values/proc = 42772 21870
Neighbor list info ...
update every 1 steps, delay 0 steps, check yes
max neighbors/atom: 2000, page size: 100000
master list distance cutoff = 13
ghost atom cutoff = 13
binsize = 6.5, bins = 4 4 5
2 neighbor lists, perpetual/occasional/extra = 2 0 0
(1) pair buck/coul/long, perpetual
attributes: half, newton on
pair build: half/bin/atomonly/newton
stencil: half/bin/3d/newton
bin: standard
(2) fix qeq/dynamic, perpetual, copy from (1)
attributes: half, newton on
pair build: copy
stencil: none
bin: none
Per MPI rank memory allocation (min/avg/max) = 17.87 | 17.87 | 17.87 Mbytes
Step PotEng c_q1 c_q2 v_qtot v_nqeq
0 -3432.38094 0.85231286 -0.42615643 0.000000000001 43
10 -3452.05217 0.85475894 -0.42737947 -0.000000000001 17
20 -3497.8643 0.85999180 -0.42999590 -0.000000000007 22
30 -3568.53169 0.86772479 -0.43386239 -0.000000000006 22
40 -3633.43753 0.87338291 -0.43669146 -0.000000000006 22
50 -3700.27953 0.87807632 -0.43903816 -0.000000000005 22
60 -3784.4004 0.88402822 -0.44201411 0.000000000002 17
70 -3877.73706 0.89012201 -0.44506100 0.000000000002 22
80 -3965.36111 0.89432486 -0.44716243 0.000000000008 17
90 -4048.57901 0.89701688 -0.44850844 -0.000000000004 22
100 -4118.62736 0.89718691 -0.44859346 -0.000000000026 17
Loop time of 18.5333 on 1 procs for 100 steps with 1200 atoms
Performance: 0.047 ns/day, 514.815 hours/ns, 5.396 timesteps/s
99.9% CPU use with 1 MPI tasks x 1 OpenMP threads
MPI task timing breakdown:
Section | min time | avg time | max time |%varavg| %total
---------------------------------------------------------------
Pair | 2.8268 | 2.8268 | 2.8268 | 0.0 | 15.25
Kspace | 1.2138 | 1.2138 | 1.2138 | 0.0 | 6.55
Neigh | 0 | 0 | 0 | 0.0 | 0.00
Comm | 0.015407 | 0.015407 | 0.015407 | 0.0 | 0.08
Output | 0.0014303 | 0.0014303 | 0.0014303 | 0.0 | 0.01
Modify | 14.474 | 14.474 | 14.474 | 0.0 | 78.10
Other | | 0.001973 | | | 0.01
Nlocal: 1200.00 ave 1200 max 1200 min
Histogram: 1 0 0 0 0 0 0 0 0 0
Nghost: 8100.00 ave 8100 max 8100 min
Histogram: 1 0 0 0 0 0 0 0 0 0
Neighs: 367600.0 ave 367600 max 367600 min
Histogram: 1 0 0 0 0 0 0 0 0 0
Total # of neighbors = 367600
Ave neighs/atom = 306.33333
Neighbor list builds = 0
Dangerous builds = 0
clear
using 1 OpenMP thread(s) per MPI task
print "Using fix qeq/fire"
Using fix qeq/fire
read_restart qeq.restart
Reading restart file ...
restart file = 8 Apr 2021, LAMMPS = 8 Apr 2021
restoring atom style charge from restart
orthogonal box = (0.0000000 0.0000000 0.0000000) to (25.158320 25.158320 28.020256)
1 by 1 by 1 MPI processor grid
restoring pair style buck/coul/long from restart
1200 atoms
read_restart CPU = 0.001 seconds
fix 2 all qeq/fire 1 10 1.0e-3 100 param.qeq2
include buck.inc
kspace_style pppm 1e-6
neighbor 1.0 bin
neigh_modify delay 0 every 1 check yes
group type1 type 1
400 atoms in group type1
compute charge1 type1 property/atom q
compute q1 type1 reduce ave c_charge1
group type2 type 2
800 atoms in group type2
compute charge2 type2 property/atom q
compute q2 type2 reduce ave c_charge2
variable qtot equal count(type1)*c_q1+count(type2)*c_q2
variable nqeq equal f_2
thermo_style custom step pe c_q1 c_q2 v_qtot v_nqeq
thermo 10
thermo_modify format line "%4d %12.9g %12.8f %12.8f %16.12f %6.0f"
timestep 0.0001
fix 1 all nve
run 100
PPPM initialization ...
using 12-bit tables for long-range coulomb (src/kspace.cpp:339)
G vector (1/distance) = 0.27644401
grid = 27 27 30
stencil order = 5
estimated absolute RMS force accuracy = 1.4502702e-05
estimated relative force accuracy = 1.0071569e-06
using double precision FFTW3
3d grid and FFT values/proc = 42772 21870
Neighbor list info ...
update every 1 steps, delay 0 steps, check yes
max neighbors/atom: 2000, page size: 100000
master list distance cutoff = 13
ghost atom cutoff = 13
binsize = 6.5, bins = 4 4 5
2 neighbor lists, perpetual/occasional/extra = 2 0 0
(1) pair buck/coul/long, perpetual
attributes: half, newton on
pair build: half/bin/atomonly/newton
stencil: half/bin/3d/newton
bin: standard
(2) fix qeq/fire, perpetual, copy from (1)
attributes: half, newton on
pair build: copy
stencil: none
bin: none
Per MPI rank memory allocation (min/avg/max) = 17.87 | 17.87 | 17.87 Mbytes
Step PotEng c_q1 c_q2 v_qtot v_nqeq
0 -3432.06113 0.85226679 -0.42613339 0.000000000004 37
10 -3452.0494 0.85475813 -0.42737906 0.000000000001 10
20 -3497.83503 0.85998739 -0.42999370 0.000000000003 13
30 -3568.47507 0.86771599 -0.43385799 0.000000000004 13
40 -3633.35368 0.87337029 -0.43668514 0.000000000004 13
50 -3700.15601 0.87805847 -0.43902924 0.000000000005 13
60 -3784.32042 0.88401635 -0.44200818 0.000000000000 11
70 -3877.59818 0.89010162 -0.44505081 0.000000000000 13
80 -3965.28426 0.89431356 -0.44715678 0.000000000000 11
90 -4048.3338 0.89698069 -0.44849034 0.000000000001 13
100 -4118.63638 0.89718818 -0.44859409 0.000000000003 12
Loop time of 13.0492 on 1 procs for 100 steps with 1200 atoms
Performance: 0.066 ns/day, 362.479 hours/ns, 7.663 timesteps/s
99.9% CPU use with 1 MPI tasks x 1 OpenMP threads
MPI task timing breakdown:
Section | min time | avg time | max time |%varavg| %total
---------------------------------------------------------------
Pair | 2.7996 | 2.7996 | 2.7996 | 0.0 | 21.45
Kspace | 1.2141 | 1.2141 | 1.2141 | 0.0 | 9.30
Neigh | 0 | 0 | 0 | 0.0 | 0.00
Comm | 0.015527 | 0.015527 | 0.015527 | 0.0 | 0.12
Output | 0.0014405 | 0.0014405 | 0.0014405 | 0.0 | 0.01
Modify | 9.0166 | 9.0166 | 9.0166 | 0.0 | 69.10
Other | | 0.001981 | | | 0.02
Nlocal: 1200.00 ave 1200 max 1200 min
Histogram: 1 0 0 0 0 0 0 0 0 0
Nghost: 8100.00 ave 8100 max 8100 min
Histogram: 1 0 0 0 0 0 0 0 0 0
Neighs: 367600.0 ave 367600 max 367600 min
Histogram: 1 0 0 0 0 0 0 0 0 0
Total # of neighbors = 367600
Ave neighs/atom = 306.33333
Neighbor list builds = 0
Dangerous builds = 0
Total wall time: 0:01:00

650
examples/qeq/log.20Apr21.qeq.buck.g++.4 Normal file
View File

@ -0,0 +1,650 @@
LAMMPS (8 Apr 2021)
using 1 OpenMP thread(s) per MPI task
# This example demonstrates the use of various fix qeq variants with
# a pair style using charges, in this case pair_style buck/coul/long
units metal
atom_style charge
read_data data.aC
Reading data file ...
orthogonal box = (0.0000000 0.0000000 0.0000000) to (25.158320 25.158320 28.020256)
1 by 2 by 2 MPI processor grid
reading atoms ...
1200 atoms
read_data CPU = 0.009 seconds
#replicate 2 2 2
pair_style buck/coul/long 12.0
pair_coeff 2 2 1388.77 .3623188 175.0
pair_coeff 1 2 18003 .2052124 133.5381
pair_coeff 1 1 0 .1 0
fix 2 all qeq/shielded 1 10 1.0e-20 10 param.qeq2
include buck.inc
kspace_style pppm 1e-6
neighbor 1.0 bin
neigh_modify delay 0 every 1 check yes
group type1 type 1
400 atoms in group type1
compute charge1 type1 property/atom q
compute q1 type1 reduce ave c_charge1
group type2 type 2
800 atoms in group type2
compute charge2 type2 property/atom q
compute q2 type2 reduce ave c_charge2
variable qtot equal count(type1)*c_q1+count(type2)*c_q2
variable nqeq equal f_2
thermo_style custom step pe c_q1 c_q2 v_qtot v_nqeq
thermo 10
thermo_modify format line "%4d %12.9g %12.8f %12.8f %16.12f %6.0f"
timestep 0.0001
fix 1 all nve
velocity all create 300.0 1281937
run 0 post no
PPPM initialization ...
using 12-bit tables for long-range coulomb (src/kspace.cpp:339)
G vector (1/distance) = 0.30705229
grid = 48 48 54
stencil order = 5
estimated absolute RMS force accuracy = 1.8909403e-05
estimated relative force accuracy = 1.3131854e-06
using double precision FFTW3
3d grid and FFT values/proc = 57970 32256
Neighbor list info ...
update every 1 steps, delay 0 steps, check yes
max neighbors/atom: 2000, page size: 100000
master list distance cutoff = 13
ghost atom cutoff = 13
binsize = 6.5, bins = 4 4 5
2 neighbor lists, perpetual/occasional/extra = 2 0 0
(1) pair buck/coul/long, perpetual, half/full from (2)
attributes: half, newton on
pair build: halffull/newton
stencil: none
bin: none
(2) fix qeq/shielded, perpetual
attributes: full, newton on
pair build: full/bin/atomonly
stencil: full/bin/3d
bin: standard
WARNING: Fix qeq CG convergence failed (4.299911728887494e-19) after 10 iterations at step 0 (src/QEQ/fix_qeq.cpp:410)
WARNING: Fix qeq CG convergence failed (5.273380778822746e-18) after 10 iterations at step 0 (src/QEQ/fix_qeq.cpp:410)
Per MPI rank memory allocation (min/avg/max) = 14.97 | 15.02 | 15.08 Mbytes
Step PotEng c_q1 c_q2 v_qtot v_nqeq
0 -2879.00327 0.76536977 -0.38268489 0.000000000000 10
Loop time of 3.33786e-06 on 4 procs for 0 steps with 1200 atoms
write_restart qeq.restart
System init for write_restart ...
PPPM initialization ...
using 12-bit tables for long-range coulomb (src/kspace.cpp:339)
G vector (1/distance) = 0.27644401
grid = 27 27 30
stencil order = 5
estimated absolute RMS force accuracy = 1.4502702e-05
estimated relative force accuracy = 1.0071569e-06
using double precision FFTW3
3d grid and FFT values/proc = 14960 5832
clear
using 1 OpenMP thread(s) per MPI task
print "Using fix qeq/point"
Using fix qeq/point
read_restart qeq.restart
Reading restart file ...
restart file = 8 Apr 2021, LAMMPS = 8 Apr 2021
restoring atom style charge from restart
orthogonal box = (0.0000000 0.0000000 0.0000000) to (25.158320 25.158320 28.020256)
1 by 2 by 2 MPI processor grid
restoring pair style buck/coul/long from restart
1200 atoms
read_restart CPU = 0.001 seconds
fix 2 all qeq/point 1 10 1.0e-6 100 param.qeq2
include buck.inc
kspace_style pppm 1e-6
neighbor 1.0 bin
neigh_modify delay 0 every 1 check yes
group type1 type 1
400 atoms in group type1
compute charge1 type1 property/atom q
compute q1 type1 reduce ave c_charge1
group type2 type 2
800 atoms in group type2
compute charge2 type2 property/atom q
compute q2 type2 reduce ave c_charge2
variable qtot equal count(type1)*c_q1+count(type2)*c_q2
variable nqeq equal f_2
thermo_style custom step pe c_q1 c_q2 v_qtot v_nqeq
thermo 10
thermo_modify format line "%4d %12.9g %12.8f %12.8f %16.12f %6.0f"
timestep 0.0001
fix 1 all nve
run 100
PPPM initialization ...
using 12-bit tables for long-range coulomb (src/kspace.cpp:339)
G vector (1/distance) = 0.27644401
grid = 27 27 30
stencil order = 5
estimated absolute RMS force accuracy = 1.4502702e-05
estimated relative force accuracy = 1.0071569e-06
using double precision FFTW3
3d grid and FFT values/proc = 14960 5832
Neighbor list info ...
update every 1 steps, delay 0 steps, check yes
max neighbors/atom: 2000, page size: 100000
master list distance cutoff = 13
ghost atom cutoff = 13
binsize = 6.5, bins = 4 4 5
2 neighbor lists, perpetual/occasional/extra = 2 0 0
(1) pair buck/coul/long, perpetual, half/full from (2)
attributes: half, newton on
pair build: halffull/newton
stencil: none
bin: none
(2) fix qeq/point, perpetual
attributes: full, newton on
pair build: full/bin/atomonly
stencil: full/bin/3d
bin: standard
Per MPI rank memory allocation (min/avg/max) = 11.10 | 11.14 | 11.16 Mbytes
Step PotEng c_q1 c_q2 v_qtot v_nqeq
0 -3432.17988 0.85228288 -0.42614144 -0.000000000000 3
10 -3452.03328 0.85475605 -0.42737803 -0.000000000000 8
20 -3497.57515 0.85994936 -0.42997468 0.000000000000 8
30 -3568.22095 0.86767937 -0.43383969 0.000000000000 8
40 -3633.24956 0.87335551 -0.43667775 -0.000000000000 8
50 -3700.10219 0.87805056 -0.43902528 0.000000000000 8
60 -3784.36769 0.88402303 -0.44201151 0.000000000000 8
70 -3877.51378 0.89008950 -0.44504475 0.000000000000 8
80 -3965.29722 0.89431515 -0.44715757 0.000000000000 8
90 -4048.36764 0.89698588 -0.44849294 -0.000000000000 8
100 -4118.65809 0.89719102 -0.44859551 0.000000000000 8
Loop time of 3.30911 on 4 procs for 100 steps with 1200 atoms
Performance: 0.261 ns/day, 91.920 hours/ns, 30.220 timesteps/s
99.0% CPU use with 4 MPI tasks x 1 OpenMP threads
MPI task timing breakdown:
Section | min time | avg time | max time |%varavg| %total
---------------------------------------------------------------
Pair | 0.67613 | 0.68904 | 0.71562 | 1.9 | 20.82
Kspace | 0.36056 | 0.3881 | 0.39892 | 2.6 | 11.73
Neigh | 0 | 0 | 0 | 0.0 | 0.00
Comm | 0.013339 | 0.017982 | 0.019974 | 2.0 | 0.54
Output | 0.0006721 | 0.00099713 | 0.0019572 | 0.0 | 0.03
Modify | 2.2109 | 2.211 | 2.211 | 0.0 | 66.81
Other | | 0.002041 | | | 0.06
Nlocal: 300.000 ave 300 max 300 min
Histogram: 4 0 0 0 0 0 0 0 0 0
Nghost: 4875.00 ave 4880 max 4870 min
Histogram: 2 0 0 0 0 0 0 0 0 2
Neighs: 91900.0 ave 91900 max 91900 min
Histogram: 4 0 0 0 0 0 0 0 0 0
FullNghs: 183800.0 ave 183800 max 183800 min
Histogram: 4 0 0 0 0 0 0 0 0 0
Total # of neighbors = 735200
Ave neighs/atom = 612.66667
Neighbor list builds = 0
Dangerous builds = 0
clear
using 1 OpenMP thread(s) per MPI task
print "Using fix qeq/shielded"
Using fix qeq/shielded
read_restart qeq.restart
Reading restart file ...
restart file = 8 Apr 2021, LAMMPS = 8 Apr 2021
restoring atom style charge from restart
orthogonal box = (0.0000000 0.0000000 0.0000000) to (25.158320 25.158320 28.020256)
1 by 2 by 2 MPI processor grid
restoring pair style buck/coul/long from restart
1200 atoms
read_restart CPU = 0.006 seconds
fix 2 all qeq/shielded 1 10 1.0e-6 100 param.qeq2
include buck.inc
kspace_style pppm 1e-6
neighbor 1.0 bin
neigh_modify delay 0 every 1 check yes
group type1 type 1
400 atoms in group type1
compute charge1 type1 property/atom q
compute q1 type1 reduce ave c_charge1
group type2 type 2
800 atoms in group type2
compute charge2 type2 property/atom q
compute q2 type2 reduce ave c_charge2
variable qtot equal count(type1)*c_q1+count(type2)*c_q2
variable nqeq equal f_2
thermo_style custom step pe c_q1 c_q2 v_qtot v_nqeq
thermo 10
thermo_modify format line "%4d %12.9g %12.8f %12.8f %16.12f %6.0f"
timestep 0.0001
fix 1 all nve
run 100
PPPM initialization ...
using 12-bit tables for long-range coulomb (src/kspace.cpp:339)
G vector (1/distance) = 0.27644401
grid = 27 27 30
stencil order = 5
estimated absolute RMS force accuracy = 1.4502702e-05
estimated relative force accuracy = 1.0071569e-06
using double precision FFTW3
3d grid and FFT values/proc = 14960 5832
Neighbor list info ...
update every 1 steps, delay 0 steps, check yes
max neighbors/atom: 2000, page size: 100000
master list distance cutoff = 13
ghost atom cutoff = 13
binsize = 6.5, bins = 4 4 5
2 neighbor lists, perpetual/occasional/extra = 2 0 0
(1) pair buck/coul/long, perpetual, half/full from (2)
attributes: half, newton on
pair build: halffull/newton
stencil: none
bin: none
(2) fix qeq/shielded, perpetual
attributes: full, newton on
pair build: full/bin/atomonly
stencil: full/bin/3d
bin: standard
Per MPI rank memory allocation (min/avg/max) = 11.10 | 11.14 | 11.16 Mbytes
Step PotEng c_q1 c_q2 v_qtot v_nqeq
0 -2879.00309 0.76536977 -0.38268489 0.000000000000 3
10 -2882.50998 0.76536972 -0.38268486 0.000000000000 2
20 -2893.89472 0.76536950 -0.38268475 -0.000000000000 2
30 -2913.6181 0.76536875 -0.38268438 -0.000000000000 1
40 -2942.24129 0.76536939 -0.38268470 0.000000000000 1
50 -2980.18817 0.76536780 -0.38268390 0.000000000000 2
60 -3027.60957 0.76536804 -0.38268402 -0.000000000000 2
70 -3084.12552 0.76536573 -0.38268287 0.000000000000 2
80 -3148.8697 0.76536550 -0.38268275 0.000000000000 1
90 -3220.43086 0.76536380 -0.38268190 0.000000000000 2
100 -3297.0618 0.76536251 -0.38268126 0.000000000000 2
Loop time of 2.25559 on 4 procs for 100 steps with 1200 atoms
Performance: 0.383 ns/day, 62.655 hours/ns, 44.334 timesteps/s
97.9% CPU use with 4 MPI tasks x 1 OpenMP threads
MPI task timing breakdown:
Section | min time | avg time | max time |%varavg| %total
---------------------------------------------------------------
Pair | 0.67442 | 0.69181 | 0.70907 | 2.0 | 30.67
Kspace | 0.39381 | 0.41151 | 0.43023 | 2.6 | 18.24
Neigh | 0 | 0 | 0 | 0.0 | 0.00
Comm | 0.012851 | 0.01426 | 0.015146 | 0.7 | 0.63
Output | 0.00066686 | 0.00098681 | 0.0019395 | 0.0 | 0.04
Modify | 1.1349 | 1.135 | 1.135 | 0.0 | 50.32
Other | | 0.002035 | | | 0.09
Nlocal: 300.000 ave 300 max 300 min
Histogram: 4 0 0 0 0 0 0 0 0 0
Nghost: 4875.00 ave 4880 max 4870 min
Histogram: 2 0 0 0 0 0 0 0 0 2
Neighs: 91900.0 ave 91900 max 91900 min
Histogram: 4 0 0 0 0 0 0 0 0 0
FullNghs: 183800.0 ave 183800 max 183800 min
Histogram: 4 0 0 0 0 0 0 0 0 0
Total # of neighbors = 735200
Ave neighs/atom = 612.66667
Neighbor list builds = 0
Dangerous builds = 0
clear
using 1 OpenMP thread(s) per MPI task
print "Using fix qeq/slater"
Using fix qeq/slater
read_restart qeq.restart
Reading restart file ...
restart file = 8 Apr 2021, LAMMPS = 8 Apr 2021
restoring atom style charge from restart
orthogonal box = (0.0000000 0.0000000 0.0000000) to (25.158320 25.158320 28.020256)
1 by 2 by 2 MPI processor grid
restoring pair style buck/coul/long from restart
1200 atoms
read_restart CPU = 0.012 seconds
fix 2 all qeq/shielded 1 10 1.0e-6 100 param.qeq2
include buck.inc
kspace_style pppm 1e-6
neighbor 1.0 bin
neigh_modify delay 0 every 1 check yes
group type1 type 1
400 atoms in group type1
compute charge1 type1 property/atom q
compute q1 type1 reduce ave c_charge1
group type2 type 2
800 atoms in group type2
compute charge2 type2 property/atom q
compute q2 type2 reduce ave c_charge2
variable qtot equal count(type1)*c_q1+count(type2)*c_q2
variable nqeq equal f_2
thermo_style custom step pe c_q1 c_q2 v_qtot v_nqeq
thermo 10
thermo_modify format line "%4d %12.9g %12.8f %12.8f %16.12f %6.0f"
timestep 0.0001
fix 1 all nve
run 100
PPPM initialization ...
using 12-bit tables for long-range coulomb (src/kspace.cpp:339)
G vector (1/distance) = 0.27644401
grid = 27 27 30
stencil order = 5
estimated absolute RMS force accuracy = 1.4502702e-05
estimated relative force accuracy = 1.0071569e-06
using double precision FFTW3
3d grid and FFT values/proc = 14960 5832
Neighbor list info ...
update every 1 steps, delay 0 steps, check yes
max neighbors/atom: 2000, page size: 100000
master list distance cutoff = 13
ghost atom cutoff = 13
binsize = 6.5, bins = 4 4 5
2 neighbor lists, perpetual/occasional/extra = 2 0 0
(1) pair buck/coul/long, perpetual, half/full from (2)
attributes: half, newton on
pair build: halffull/newton
stencil: none
bin: none
(2) fix qeq/shielded, perpetual
attributes: full, newton on
pair build: full/bin/atomonly
stencil: full/bin/3d
bin: standard
Per MPI rank memory allocation (min/avg/max) = 11.10 | 11.14 | 11.16 Mbytes
Step PotEng c_q1 c_q2 v_qtot v_nqeq
0 -2879.00309 0.76536977 -0.38268489 0.000000000000 3
10 -2882.50998 0.76536972 -0.38268486 0.000000000000 2
20 -2893.89472 0.76536950 -0.38268475 -0.000000000000 2
30 -2913.6181 0.76536875 -0.38268438 -0.000000000000 1
40 -2942.24129 0.76536939 -0.38268470 0.000000000000 1
50 -2980.18817 0.76536780 -0.38268390 0.000000000000 2
60 -3027.60957 0.76536804 -0.38268402 -0.000000000000 2
70 -3084.12552 0.76536573 -0.38268287 0.000000000000 2
80 -3148.8697 0.76536550 -0.38268275 0.000000000000 1
90 -3220.43086 0.76536380 -0.38268190 0.000000000000 2
100 -3297.0618 0.76536251 -0.38268126 0.000000000000 2
Loop time of 2.39249 on 4 procs for 100 steps with 1200 atoms
Performance: 0.361 ns/day, 66.458 hours/ns, 41.797 timesteps/s
96.0% CPU use with 4 MPI tasks x 1 OpenMP threads
MPI task timing breakdown:
Section | min time | avg time | max time |%varavg| %total
---------------------------------------------------------------
Pair | 0.6751 | 0.70301 | 0.71919 | 2.1 | 29.38
Kspace | 0.45569 | 0.47315 | 0.49885 | 2.6 | 19.78
Neigh | 0 | 0 | 0 | 0.0 | 0.00
Comm | 0.012967 | 0.018681 | 0.020909 | 2.4 | 0.78
Output | 0.00066733 | 0.00099397 | 0.0019579 | 0.0 | 0.04
Modify | 1.1945 | 1.1946 | 1.1947 | 0.0 | 49.93
Other | | 0.002046 | | | 0.09
Nlocal: 300.000 ave 300 max 300 min
Histogram: 4 0 0 0 0 0 0 0 0 0
Nghost: 4875.00 ave 4880 max 4870 min
Histogram: 2 0 0 0 0 0 0 0 0 2
Neighs: 91900.0 ave 91900 max 91900 min
Histogram: 4 0 0 0 0 0 0 0 0 0
FullNghs: 183800.0 ave 183800 max 183800 min
Histogram: 4 0 0 0 0 0 0 0 0 0
Total # of neighbors = 735200
Ave neighs/atom = 612.66667
Neighbor list builds = 0
Dangerous builds = 0
clear
using 1 OpenMP thread(s) per MPI task
print "Using fix qeq/dynamic"
Using fix qeq/dynamic
read_restart qeq.restart
Reading restart file ...
restart file = 8 Apr 2021, LAMMPS = 8 Apr 2021
restoring atom style charge from restart
orthogonal box = (0.0000000 0.0000000 0.0000000) to (25.158320 25.158320 28.020256)
1 by 2 by 2 MPI processor grid
restoring pair style buck/coul/long from restart
1200 atoms
read_restart CPU = 0.002 seconds
fix 2 all qeq/dynamic 1 10 1.0e-3 100 param.qeq2
include buck.inc
kspace_style pppm 1e-6
neighbor 1.0 bin
neigh_modify delay 0 every 1 check yes
group type1 type 1
400 atoms in group type1
compute charge1 type1 property/atom q
compute q1 type1 reduce ave c_charge1
group type2 type 2
800 atoms in group type2
compute charge2 type2 property/atom q
compute q2 type2 reduce ave c_charge2
variable qtot equal count(type1)*c_q1+count(type2)*c_q2
variable nqeq equal f_2
thermo_style custom step pe c_q1 c_q2 v_qtot v_nqeq
thermo 10
thermo_modify format line "%4d %12.9g %12.8f %12.8f %16.12f %6.0f"
timestep 0.0001
fix 1 all nve
run 100
PPPM initialization ...
using 12-bit tables for long-range coulomb (src/kspace.cpp:339)
G vector (1/distance) = 0.27644401
grid = 27 27 30
stencil order = 5
estimated absolute RMS force accuracy = 1.4502702e-05
estimated relative force accuracy = 1.0071569e-06
using double precision FFTW3
3d grid and FFT values/proc = 14960 5832
Neighbor list info ...
update every 1 steps, delay 0 steps, check yes
max neighbors/atom: 2000, page size: 100000
master list distance cutoff = 13
ghost atom cutoff = 13
binsize = 6.5, bins = 4 4 5
2 neighbor lists, perpetual/occasional/extra = 2 0 0
(1) pair buck/coul/long, perpetual
attributes: half, newton on
pair build: half/bin/atomonly/newton
stencil: half/bin/3d/newton
bin: standard
(2) fix qeq/dynamic, perpetual, copy from (1)
attributes: half, newton on
pair build: copy
stencil: none
bin: none
Per MPI rank memory allocation (min/avg/max) = 9.195 | 9.246 | 9.278 Mbytes
Step PotEng c_q1 c_q2 v_qtot v_nqeq
0 -3432.38094 0.85231286 -0.42615643 -0.000000000001 43
10 -3452.05217 0.85475894 -0.42737947 -0.000000000003 17
20 -3497.8643 0.85999180 -0.42999590 0.000000000000 22
30 -3568.53169 0.86772479 -0.43386239 -0.000000000000 22
40 -3633.43753 0.87338291 -0.43669146 0.000000000006 22
50 -3700.27953 0.87807632 -0.43903816 0.000000000003 22
60 -3784.4004 0.88402822 -0.44201411 0.000000000009 17
70 -3877.73706 0.89012201 -0.44506100 0.000000000010 22
80 -3965.36111 0.89432486 -0.44716243 0.000000000011 17
90 -4048.57901 0.89701688 -0.44850844 0.000000000012 22
100 -4118.62736 0.89718691 -0.44859346 0.000000000013 17
Loop time of 5.27704 on 4 procs for 100 steps with 1200 atoms
Performance: 0.164 ns/day, 146.584 hours/ns, 18.950 timesteps/s
98.5% CPU use with 4 MPI tasks x 1 OpenMP threads
MPI task timing breakdown:
Section | min time | avg time | max time |%varavg| %total
---------------------------------------------------------------
Pair | 0.68437 | 0.69096 | 0.69826 | 0.7 | 13.09
Kspace | 0.38484 | 0.38941 | 0.39524 | 0.7 | 7.38
Neigh | 0 | 0 | 0 | 0.0 | 0.00
Comm | 0.012609 | 0.01529 | 0.016842 | 1.3 | 0.29
Output | 0.00067735 | 0.0010006 | 0.0019588 | 1.7 | 0.02
Modify | 4.1783 | 4.1783 | 4.1784 | 0.0 | 79.18
Other | | 0.002027 | | | 0.04
Nlocal: 300.000 ave 300 max 300 min
Histogram: 4 0 0 0 0 0 0 0 0 0
Nghost: 4875.00 ave 4880 max 4870 min
Histogram: 2 0 0 0 0 0 0 0 0 2
Neighs: 91900.0 ave 93081 max 90719 min
Histogram: 2 0 0 0 0 0 0 0 0 2
Total # of neighbors = 367600
Ave neighs/atom = 306.33333
Neighbor list builds = 0
Dangerous builds = 0
clear
using 1 OpenMP thread(s) per MPI task
print "Using fix qeq/fire"
Using fix qeq/fire
read_restart qeq.restart
Reading restart file ...
restart file = 8 Apr 2021, LAMMPS = 8 Apr 2021
restoring atom style charge from restart
orthogonal box = (0.0000000 0.0000000 0.0000000) to (25.158320 25.158320 28.020256)
1 by 2 by 2 MPI processor grid
restoring pair style buck/coul/long from restart
1200 atoms
read_restart CPU = 0.001 seconds
fix 2 all qeq/fire 1 10 1.0e-3 100 param.qeq2
include buck.inc
kspace_style pppm 1e-6
neighbor 1.0 bin
neigh_modify delay 0 every 1 check yes
group type1 type 1
400 atoms in group type1
compute charge1 type1 property/atom q
compute q1 type1 reduce ave c_charge1
group type2 type 2
800 atoms in group type2
compute charge2 type2 property/atom q
compute q2 type2 reduce ave c_charge2
variable qtot equal count(type1)*c_q1+count(type2)*c_q2
variable nqeq equal f_2
thermo_style custom step pe c_q1 c_q2 v_qtot v_nqeq
thermo 10
thermo_modify format line "%4d %12.9g %12.8f %12.8f %16.12f %6.0f"
timestep 0.0001
fix 1 all nve
run 100
PPPM initialization ...
using 12-bit tables for long-range coulomb (src/kspace.cpp:339)
G vector (1/distance) = 0.27644401
grid = 27 27 30
stencil order = 5
estimated absolute RMS force accuracy = 1.4502702e-05
estimated relative force accuracy = 1.0071569e-06
using double precision FFTW3
3d grid and FFT values/proc = 14960 5832
Neighbor list info ...
update every 1 steps, delay 0 steps, check yes
max neighbors/atom: 2000, page size: 100000
master list distance cutoff = 13
ghost atom cutoff = 13
binsize = 6.5, bins = 4 4 5
2 neighbor lists, perpetual/occasional/extra = 2 0 0
(1) pair buck/coul/long, perpetual
attributes: half, newton on
pair build: half/bin/atomonly/newton
stencil: half/bin/3d/newton
bin: standard
(2) fix qeq/fire, perpetual, copy from (1)
attributes: half, newton on
pair build: copy
stencil: none
bin: none
Per MPI rank memory allocation (min/avg/max) = 9.195 | 9.246 | 9.278 Mbytes
Step PotEng c_q1 c_q2 v_qtot v_nqeq
0 -3432.05316 0.85226679 -0.42613339 0.000000000001 37
10 -3452.04937 0.85475813 -0.42737906 0.000000000001 10
20 -3497.83659 0.85998739 -0.42999370 0.000000000002 13
30 -3568.47793 0.86771599 -0.43385799 0.000000000002 13
40 -3633.35326 0.87337029 -0.43668514 0.000000000002 13
50 -3700.16079 0.87805847 -0.43902924 0.000000000000 13
60 -3784.31906 0.88401635 -0.44200818 -0.000000000001 11
70 -3877.60163 0.89010162 -0.44505081 -0.000000000000 13
80 -3965.28179 0.89431356 -0.44715678 0.000000000001 11
90 -4048.33861 0.89698069 -0.44849034 0.000000000001 13
100 -4118.63861 0.89718818 -0.44859409 0.000000000002 12
Loop time of 3.88026 on 4 procs for 100 steps with 1200 atoms
Performance: 0.223 ns/day, 107.785 hours/ns, 25.771 timesteps/s
98.0% CPU use with 4 MPI tasks x 1 OpenMP threads
MPI task timing breakdown:
Section | min time | avg time | max time |%varavg| %total
---------------------------------------------------------------
Pair | 0.68424 | 0.69912 | 0.73572 | 2.5 | 18.02
Kspace | 0.38093 | 0.41715 | 0.43168 | 3.2 | 10.75
Neigh | 0 | 0 | 0 | 0.0 | 0.00
Comm | 0.012711 | 0.013318 | 0.014003 | 0.4 | 0.34
Output | 0.00066566 | 0.00098735 | 0.0019317 | 0.0 | 0.03
Modify | 2.7477 | 2.7477 | 2.7477 | 0.0 | 70.81
Other | | 0.002004 | | | 0.05
Nlocal: 300.000 ave 300 max 300 min
Histogram: 4 0 0 0 0 0 0 0 0 0
Nghost: 4875.00 ave 4880 max 4870 min
Histogram: 2 0 0 0 0 0 0 0 0 2
Neighs: 91900.0 ave 93081 max 90719 min
Histogram: 2 0 0 0 0 0 0 0 0 2
Total # of neighbors = 367600
Ave neighs/atom = 306.33333
Neighbor list builds = 0
Dangerous builds = 0
Total wall time: 0:00:17

118
examples/qeq/log.27Nov18.qeq.buck.g++.1
View File

@ -1,118 +0,0 @@
LAMMPS (27 Nov 2018)
using 1 OpenMP thread(s) per MPI task
# This example demonstrates the use of various fix qeq variants with
# that defines and uses charges, in this case pair_style buck/coul/long
units metal
atom_style charge
read_data data.aC
orthogonal box = (0 0 0) to (25.1583 25.1583 28.0203)
1 by 1 by 1 MPI processor grid
reading atoms ...
1200 atoms
replicate 2 2 2
orthogonal box = (0 0 0) to (50.3166 50.3166 56.0405)
1 by 1 by 1 MPI processor grid
9600 atoms
Time spent = 0.00114894 secs
pair_style buck/coul/long 12.0
pair_coeff 2 2 1388.77 .3623188 175.0
pair_coeff 1 2 18003 .2052124 133.5381
pair_coeff 1 1 0 .1 0
kspace_style ewald 1e-6
neighbor 1.0 bin
neigh_modify delay 0 every 1 check yes
group type1 type 1
3200 atoms in group type1
compute charge1 type1 property/atom q
compute q1 type1 reduce ave c_charge1
group type2 type 2
6400 atoms in group type2
compute charge2 type2 property/atom q
compute q2 type2 reduce ave c_charge2
variable qtot equal count(type1)*c_q1+count(type2)*c_q2
thermo_style custom step pe c_q1 c_q2 v_qtot spcpu
thermo 10
timestep 0.0001
velocity all create 300.0 1281937
fix 1 all nve
#fix 2 all qeq/point 1 10 1.0e-6 100 param.qeq2
#fix 2 all qeq/shielded 1 10 1.0e-6 100 param.qeq2
#fix 2 all qeq/slater 1 10 1.0e-6 100 param.qeq2
#fix 2 all qeq/dynamic 1 10 1.0e-4 100 param.qeq2
fix 2 all qeq/fire 1 10 1.0e-4 100 param.qeq2
run 100
Ewald initialization ...
using 12-bit tables for long-range coulomb (src/kspace.cpp:321)
G vector (1/distance) = 0.305064
estimated absolute RMS force accuracy = 2.07629e-05
estimated relative force accuracy = 1.44191e-06
KSpace vectors: actual max1d max3d = 13556 20 34460
kxmax kymax kzmax = 18 18 20
Neighbor list info ...
update every 1 steps, delay 0 steps, check yes
max neighbors/atom: 2000, page size: 100000
master list distance cutoff = 13
ghost atom cutoff = 13
binsize = 6.5, bins = 8 8 9
2 neighbor lists, perpetual/occasional/extra = 2 0 0
(1) pair buck/coul/long, perpetual
attributes: half, newton on
pair build: half/bin/atomonly/newton
stencil: half/bin/3d/newton
bin: standard
(2) fix qeq/fire, perpetual, copy from (1)
attributes: half, newton on
pair build: copy
stencil: none
bin: none
Per MPI rank memory allocation (min/avg/max) = 134 | 134 | 134 Mbytes
Step PotEng c_q1 c_q2 v_qtot S/CPU
0 -27457.219 0.85227886 -0.42613943 -2.1827873e-10 0
10 -27626.057 0.85486228 -0.42743114 -2.0372681e-10 0.64313877
20 -27975.085 0.85968531 -0.42984266 -1.036824e-10 0.55119179
30 -28552.628 0.86755661 -0.4337783 1.3051249e-10 0.53160643
40 -29133.643 0.87426387 -0.43713193 1.1368684e-10 0.53075341
50 -29697.011 0.8794039 -0.43970195 1.200533e-10 0.52358127
60 -30342.001 0.88478594 -0.44239297 6.002665e-11 0.5366762
70 -31081.138 0.8906973 -0.44534865 -4.7293724e-11 0.55904546
80 -31792.732 0.89506635 -0.44753317 -4.3200998e-11 0.59606079
90 -32424.749 0.89714841 -0.44857421 -1.1596057e-10 0.58047419
100 -32998.353 0.89755721 -0.44877861 -1.0231815e-10 0.59444001
Loop time of 177.79 on 1 procs for 100 steps with 9600 atoms
Performance: 0.005 ns/day, 4938.612 hours/ns, 0.562 timesteps/s
99.8% CPU use with 1 MPI tasks x 1 OpenMP threads
MPI task timing breakdown:
Section | min time | avg time | max time |%varavg| %total
---------------------------------------------------------------
Pair | 11.518 | 11.518 | 11.518 | 0.0 | 6.48
Kspace | 107.37 | 107.37 | 107.37 | 0.0 | 60.39
Neigh | 0 | 0 | 0 | 0.0 | 0.00
Comm | 0.019721 | 0.019721 | 0.019721 | 0.0 | 0.01
Output | 0.002218 | 0.002218 | 0.002218 | 0.0 | 0.00
Modify | 58.869 | 58.869 | 58.869 | 0.0 | 33.11
Other | | 0.007197 | | | 0.00
Nlocal: 9600 ave 9600 max 9600 min
Histogram: 1 0 0 0 0 0 0 0 0 0
Nghost: 22125 ave 22125 max 22125 min
Histogram: 1 0 0 0 0 0 0 0 0 0
Neighs: 2.9408e+06 ave 2.9408e+06 max 2.9408e+06 min
Histogram: 1 0 0 0 0 0 0 0 0 0
Total # of neighbors = 2940800
Ave neighs/atom = 306.333
Neighbor list builds = 0
Dangerous builds = 0
Total wall time: 0:03:01

118
examples/qeq/log.27Nov18.qeq.buck.g++.4
View File

@ -1,118 +0,0 @@
LAMMPS (27 Nov 2018)
using 1 OpenMP thread(s) per MPI task
# This example demonstrates the use of various fix qeq variants with
# that defines and uses charges, in this case pair_style buck/coul/long
units metal
atom_style charge
read_data data.aC
orthogonal box = (0 0 0) to (25.1583 25.1583 28.0203)
1 by 2 by 2 MPI processor grid
reading atoms ...
1200 atoms
replicate 2 2 2
orthogonal box = (0 0 0) to (50.3166 50.3166 56.0405)
1 by 2 by 2 MPI processor grid
9600 atoms
Time spent = 0.000675201 secs
pair_style buck/coul/long 12.0
pair_coeff 2 2 1388.77 .3623188 175.0
pair_coeff 1 2 18003 .2052124 133.5381
pair_coeff 1 1 0 .1 0
kspace_style ewald 1e-6
neighbor 1.0 bin
neigh_modify delay 0 every 1 check yes
group type1 type 1
3200 atoms in group type1
compute charge1 type1 property/atom q
compute q1 type1 reduce ave c_charge1
group type2 type 2
6400 atoms in group type2
compute charge2 type2 property/atom q
compute q2 type2 reduce ave c_charge2
variable qtot equal count(type1)*c_q1+count(type2)*c_q2
thermo_style custom step pe c_q1 c_q2 v_qtot spcpu
thermo 10
timestep 0.0001
velocity all create 300.0 1281937
fix 1 all nve
#fix 2 all qeq/point 1 10 1.0e-6 100 param.qeq2
#fix 2 all qeq/shielded 1 10 1.0e-6 100 param.qeq2
#fix 2 all qeq/slater 1 10 1.0e-6 100 param.qeq2
#fix 2 all qeq/dynamic 1 10 1.0e-4 100 param.qeq2
fix 2 all qeq/fire 1 10 1.0e-4 100 param.qeq2
run 100
Ewald initialization ...
using 12-bit tables for long-range coulomb (src/kspace.cpp:321)
G vector (1/distance) = 0.305064
estimated absolute RMS force accuracy = 2.07629e-05
estimated relative force accuracy = 1.44191e-06
KSpace vectors: actual max1d max3d = 13556 20 34460
kxmax kymax kzmax = 18 18 20
Neighbor list info ...
update every 1 steps, delay 0 steps, check yes
max neighbors/atom: 2000, page size: 100000
master list distance cutoff = 13
ghost atom cutoff = 13
binsize = 6.5, bins = 8 8 9
2 neighbor lists, perpetual/occasional/extra = 2 0 0
(1) pair buck/coul/long, perpetual
attributes: half, newton on
pair build: half/bin/atomonly/newton
stencil: half/bin/3d/newton
bin: standard
(2) fix qeq/fire, perpetual, copy from (1)
attributes: half, newton on
pair build: copy
stencil: none
bin: none
Per MPI rank memory allocation (min/avg/max) = 53.06 | 53.13 | 53.21 Mbytes
Step PotEng c_q1 c_q2 v_qtot S/CPU
0 -27457.215 0.85227886 -0.42613943 2.1373125e-11 0
10 -27626.057 0.85486228 -0.42743114 3.0468073e-11 2.4245312
20 -27975.085 0.85968531 -0.42984266 1.0095391e-10 2.0185316
30 -28552.627 0.86755661 -0.4337783 1.3096724e-10 1.9605335
40 -29133.643 0.87426387 -0.43713193 1.5279511e-10 1.9624139
50 -29697.01 0.8794039 -0.43970195 1.6461854e-10 1.8113263
60 -30342 0.88478594 -0.44239297 1.7826096e-10 1.9537722
70 -31081.139 0.89069733 -0.44534866 1.4733814e-10 2.058406
80 -31792.732 0.89506635 -0.44753317 1.3824319e-10 2.2160813
90 -32424.752 0.89714841 -0.44857421 1.2914825e-10 2.0952145
100 -32998.353 0.89755721 -0.44877861 1.4824764e-10 2.1292486
Loop time of 48.7541 on 4 procs for 100 steps with 9600 atoms
Performance: 0.018 ns/day, 1354.281 hours/ns, 2.051 timesteps/s
97.4% CPU use with 4 MPI tasks x 1 OpenMP threads
MPI task timing breakdown:
Section | min time | avg time | max time |%varavg| %total
---------------------------------------------------------------
Pair | 2.9747 | 3.0315 | 3.0758 | 2.1 | 6.22
Kspace | 27.873 | 28.264 | 28.63 | 5.3 | 57.97
Neigh | 0 | 0 | 0 | 0.0 | 0.00
Comm | 0.53835 | 0.8523 | 1.2286 | 28.2 | 1.75
Output | 0.0012984 | 0.001591 | 0.0024178 | 1.2 | 0.00
Modify | 16.58 | 16.59 | 16.601 | 0.3 | 34.03
Other | | 0.01409 | | | 0.03
Nlocal: 2400 ave 2400 max 2400 min
Histogram: 4 0 0 0 0 0 0 0 0 0
Nghost: 11550 ave 11550 max 11550 min
Histogram: 4 0 0 0 0 0 0 0 0 0
Neighs: 735200 ave 740758 max 729642 min
Histogram: 2 0 0 0 0 0 0 0 0 2
Total # of neighbors = 2940800
Ave neighs/atom = 306.333
Neighbor list builds = 0
Dangerous builds = 0
Total wall time: 0:00:49

15
src/QEQ/fix_qeq.cpp
View File

@ -54,10 +54,15 @@ FixQEq::FixQEq(LAMMPS *lmp, int narg, char **arg) :
{
if (narg < 8) error->all(FLERR,"Illegal fix qeq command");
scalar_flag = 1;
extscalar = 0;
nevery = utils::inumeric(FLERR,arg[3],false,lmp);
cutoff = utils::numeric(FLERR,arg[4],false,lmp);
tolerance = utils::numeric(FLERR,arg[5],false,lmp);
maxiter = utils::inumeric(FLERR,arg[6],false,lmp);
maxwarn = 1;
matvecs = 0;
// check for sane arguments
if ((nevery <= 0) || (cutoff <= 0.0) || (tolerance <= 0.0) || (maxiter <= 0))
@ -123,7 +128,6 @@ FixQEq::FixQEq(LAMMPS *lmp, int narg, char **arg) :
} else {
read_file(arg[7]);
}
}
/* ---------------------------------------------------------------------- */
@ -277,6 +281,13 @@ void FixQEq::reallocate_matrix()
/* ---------------------------------------------------------------------- */
double FixQEq::compute_scalar()
{
return matvecs;
}
/* ---------------------------------------------------------------------- */
void FixQEq::init_list(int /*id*/, NeighList *ptr)
{
list = ptr;
@ -395,7 +406,7 @@ int FixQEq::CG( double *b, double *x )
vector_sum( d, 1., p, beta, d, inum );
}
if ((comm->me == 0) && (loop >= maxiter))
if ((comm->me == 0) && maxwarn && (loop >= maxiter))
error->warning(FLERR,fmt::format("Fix qeq CG convergence failed ({}) "
"after {} iterations at step {}",
sqrt(sig_new)/b_norm,loop,

5
src/QEQ/fix_qeq.h
View File

@ -35,6 +35,8 @@ class FixQEq : public Fix {
void pre_force_respa(int, int, int);
void min_pre_force(int);
virtual double compute_scalar();
// derived child classes must provide these functions
virtual void init() = 0;
@ -64,7 +66,8 @@ class FixQEq : public Fix {
double swa, swb; // lower/upper Taper cutoff radius
double Tap[8]; // Taper function
double tolerance; // tolerance for the norm of the rel residual in CG
int maxiter; // maximum number of QEq iterations
int maxiter; // maximum number of QEq iterations
int maxwarn; // print warning when max iterations was reached
double cutoff, cutoff_sq; // neighbor cutoff
double *chi,*eta,*gamma,*zeta,*zcore; // qeq parameters

11
src/QEQ/fix_qeq_dynamic.cpp
View File

@ -53,6 +53,12 @@ FixQEqDynamic::FixQEqDynamic(LAMMPS *lmp, int narg, char **arg) :
if (iarg+2 > narg) error->all(FLERR,"Illegal fix qeq/dynamic command");
qstep = atof(arg[iarg+1]);
iarg += 2;
} else if (strcmp(arg[iarg],"warn") == 0) {
if (iarg+2 > narg) error->all(FLERR,"Illegal fix qeq/dynamic command");
if (strcmp(arg[iarg+1],"no") == 0) maxwarn = 0;
else if (strcmp(arg[iarg+1],"yes") == 0) maxwarn = 1;
else error->all(FLERR,"Illegal fix qeq/dynamic command");
iarg += 2;
} else error->all(FLERR,"Illegal fix qeq/dynamic command");
}
}
@ -145,7 +151,7 @@ void FixQEqDynamic::pre_force(int /*vflag*/)
MPI_Allreduce(&enegmax,&enegmaxall,1,MPI_DOUBLE,MPI_MAX,world);
enegmax = enegmaxall;
if (enegchk <= tolerance && enegmax <= 100.0*tolerance) break;
if ((enegchk <= tolerance) && (enegmax <= 100.0*tolerance)) break;
for (ii = 0; ii < inum; ii++) {
i = ilist[ii];
@ -153,8 +159,9 @@ void FixQEqDynamic::pre_force(int /*vflag*/)
q1[i] += qf[i]*dtq2 - qdamp*q1[i];
}
}
matvecs = iloop;
if ((comm->me == 0) && (iloop >= maxiter))
if ((comm->me == 0) && maxwarn && (iloop >= maxiter))
error->warning(FLERR,fmt::format("Charges did not converge at step "
"{}: {}",update->ntimestep,enegchk));

9
src/QEQ/fix_qeq_fire.cpp
View File

@ -63,6 +63,12 @@ FixQEqFire::FixQEqFire(LAMMPS *lmp, int narg, char **arg) :
if (iarg+2 > narg) error->all(FLERR,"Illegal fix qeq/fire command");
qstep = atof(arg[iarg+1]);
iarg += 2;
} else if (strcmp(arg[iarg],"warn") == 0) {
if (iarg+2 > narg) error->all(FLERR,"Illegal fix qeq/fire command");
if (strcmp(arg[iarg+1],"no") == 0) maxwarn = 0;
else if (strcmp(arg[iarg+1],"yes") == 0) maxwarn = 1;
else error->all(FLERR,"Illegal fix qeq/fire command");
iarg += 2;
} else error->all(FLERR,"Illegal fix qeq/fire command");
}
}
@ -213,8 +219,9 @@ void FixQEqFire::pre_force(int /*vflag*/)
if (enegchk < tolerance) break;
}
matvecs = iloop;
if ((comm->me == 0) && (iloop >= maxiter))
if ((comm->me == 0) && maxwarn && (iloop >= maxiter))
error->warning(FLERR,fmt::format("Charges did not converge at step "
"{}: {}",update->ntimestep,enegchk));

11
src/QEQ/fix_qeq_point.cpp
View File

@ -38,7 +38,15 @@ using namespace LAMMPS_NS;
/* ---------------------------------------------------------------------- */
FixQEqPoint::FixQEqPoint(LAMMPS *lmp, int narg, char **arg) :
FixQEq(lmp, narg, arg) {}
FixQEq(lmp, narg, arg) {
if (narg == 10) {
if (strcmp(arg[8],"warn") == 0) {
if (strcmp(arg[9],"no") == 0) maxwarn = 0;
else if (strcmp(arg[9],"yes") == 0) maxwarn = 1;
else error->all(FLERR,"Illegal fix qeq/point command");
} else error->all(FLERR,"Illegal fix qeq/point command");
} else if (narg > 8) error->all(FLERR,"Illegal fix qeq/point command");
}
/* ---------------------------------------------------------------------- */
@ -80,6 +88,7 @@ void FixQEqPoint::pre_force(int /*vflag*/)
init_matvec();
matvecs = CG(b_s, s); // CG on s - parallel
matvecs += CG(b_t, t); // CG on t - parallel
matvecs /= 2;
calculate_Q();
if (force->kspace) force->kspace->qsum_qsq();

8
src/QEQ/fix_qeq_shielded.cpp
View File

@ -40,6 +40,13 @@ using namespace LAMMPS_NS;
FixQEqShielded::FixQEqShielded(LAMMPS *lmp, int narg, char **arg) :
FixQEq(lmp, narg, arg) {
if (narg == 10) {
if (strcmp(arg[8],"warn") == 0) {
if (strcmp(arg[9],"no") == 0) maxwarn = 0;
else if (strcmp(arg[9],"yes") == 0) maxwarn = 1;
else error->all(FLERR,"Illegal fix qeq/shielded command");
} else error->all(FLERR,"Illegal fix qeq/shielded command");
} else if (narg > 8) error->all(FLERR,"Illegal fix qeq/shielded command");
if (reax_flag) extract_reax();
}
@ -143,6 +150,7 @@ void FixQEqShielded::pre_force(int /*vflag*/)
init_matvec();
matvecs = CG(b_s, s); // CG on s - parallel
matvecs += CG(b_t, t); // CG on t - parallel
matvecs /= 2;
calculate_Q();
if (force->kspace) force->kspace->qsum_qsq();

7
src/QEQ/fix_qeq_slater.cpp
View File

@ -52,6 +52,12 @@ FixQEqSlater::FixQEqSlater(LAMMPS *lmp, int narg, char **arg) :
if (iarg+2 > narg) error->all(FLERR,"Illegal fix qeq/slater command");
alpha = atof(arg[iarg+1]);
iarg += 2;
} else if (strcmp(arg[iarg],"warn") == 0) {
if (iarg+2 > narg) error->all(FLERR,"Illegal fix qeq/slater command");
if (strcmp(arg[iarg+1],"no") == 0) maxwarn = 0;
else if (strcmp(arg[iarg+1],"yes") == 0) maxwarn = 1;
else error->all(FLERR,"Illegal fix qeq/slater command");
iarg += 2;
} else error->all(FLERR,"Illegal fix qeq/slater command");
}
@ -120,6 +126,7 @@ void FixQEqSlater::pre_force(int /*vflag*/)
init_matvec();
matvecs = CG(b_s, s); // CG on s - parallel
matvecs += CG(b_t, t); // CG on t - parallel
matvecs /= 2;
calculate_Q();
if (force->kspace) force->kspace->qsum_qsq();