| 1 |
!! |
| 2 |
!! Copyright (c) 2005 The University of Notre Dame. All Rights Reserved. |
| 3 |
!! |
| 4 |
!! The University of Notre Dame grants you ("Licensee") a |
| 5 |
!! non-exclusive, royalty free, license to use, modify and |
| 6 |
!! redistribute this software in source and binary code form, provided |
| 7 |
!! that the following conditions are met: |
| 8 |
!! |
| 9 |
!! 1. Acknowledgement of the program authors must be made in any |
| 10 |
!! publication of scientific results based in part on use of the |
| 11 |
!! program. An acceptable form of acknowledgement is citation of |
| 12 |
!! the article in which the program was described (Matthew |
| 13 |
!! A. Meineke, Charles F. Vardeman II, Teng Lin, Christopher |
| 14 |
!! J. Fennell and J. Daniel Gezelter, "OOPSE: An Object-Oriented |
| 15 |
!! Parallel Simulation Engine for Molecular Dynamics," |
| 16 |
!! J. Comput. Chem. 26, pp. 252-271 (2005)) |
| 17 |
!! |
| 18 |
!! 2. Redistributions of source code must retain the above copyright |
| 19 |
!! notice, this list of conditions and the following disclaimer. |
| 20 |
!! |
| 21 |
!! 3. Redistributions in binary form must reproduce the above copyright |
| 22 |
!! notice, this list of conditions and the following disclaimer in the |
| 23 |
!! documentation and/or other materials provided with the |
| 24 |
!! distribution. |
| 25 |
!! |
| 26 |
!! This software is provided "AS IS," without a warranty of any |
| 27 |
!! kind. All express or implied conditions, representations and |
| 28 |
!! warranties, including any implied warranty of merchantability, |
| 29 |
!! fitness for a particular purpose or non-infringement, are hereby |
| 30 |
!! excluded. The University of Notre Dame and its licensors shall not |
| 31 |
!! be liable for any damages suffered by licensee as a result of |
| 32 |
!! using, modifying or distributing the software or its |
| 33 |
!! derivatives. In no event will the University of Notre Dame or its |
| 34 |
!! licensors be liable for any lost revenue, profit or data, or for |
| 35 |
!! direct, indirect, special, consequential, incidental or punitive |
| 36 |
!! damages, however caused and regardless of the theory of liability, |
| 37 |
!! arising out of the use of or inability to use software, even if the |
| 38 |
!! University of Notre Dame has been advised of the possibility of |
| 39 |
!! such damages. |
| 40 |
!! |
| 41 |
|
| 42 |
!! This Module Calculates forces due to SSD potential and VDW interactions |
| 43 |
!! [Chandra and Ichiye, J. Chem. Phys. 111, 2701 (1999)]. |
| 44 |
|
| 45 |
!! This module contains the Public procedures: |
| 46 |
|
| 47 |
|
| 48 |
!! Corresponds to the force field defined in ssd_FF.cpp |
| 49 |
!! @author Charles F. Vardeman II |
| 50 |
!! @author Matthew Meineke |
| 51 |
!! @author Christopher Fennell |
| 52 |
!! @author J. Daniel Gezelter |
| 53 |
!! @version $Id: sticky.F90,v 1.14 2005-08-26 21:30:41 chrisfen Exp $, $Date: 2005-08-26 21:30:41 $, $Name: not supported by cvs2svn $, $Revision: 1.14 $ |
| 54 |
|
| 55 |
module sticky |
| 56 |
|
| 57 |
use force_globals |
| 58 |
use definitions |
| 59 |
use atype_module |
| 60 |
use vector_class |
| 61 |
use simulation |
| 62 |
use status |
| 63 |
#ifdef IS_MPI |
| 64 |
use mpiSimulation |
| 65 |
#endif |
| 66 |
implicit none |
| 67 |
|
| 68 |
PRIVATE |
| 69 |
|
| 70 |
public :: newStickyType |
| 71 |
public :: do_sticky_pair |
| 72 |
public :: destroyStickyTypes |
| 73 |
public :: do_sticky_power_pair |
| 74 |
public :: getStickyCut |
| 75 |
public :: getStickyPowerCut |
| 76 |
|
| 77 |
type :: StickyList |
| 78 |
integer :: c_ident |
| 79 |
real( kind = dp ) :: w0 = 0.0_dp |
| 80 |
real( kind = dp ) :: v0 = 0.0_dp |
| 81 |
real( kind = dp ) :: v0p = 0.0_dp |
| 82 |
real( kind = dp ) :: rl = 0.0_dp |
| 83 |
real( kind = dp ) :: ru = 0.0_dp |
| 84 |
real( kind = dp ) :: rlp = 0.0_dp |
| 85 |
real( kind = dp ) :: rup = 0.0_dp |
| 86 |
real( kind = dp ) :: rbig = 0.0_dp |
| 87 |
end type StickyList |
| 88 |
|
| 89 |
type(StickyList), dimension(:),allocatable :: StickyMap |
| 90 |
|
| 91 |
contains |
| 92 |
|
| 93 |
subroutine newStickyType(c_ident, w0, v0, v0p, rl, ru, rlp, rup, isError) |
| 94 |
|
| 95 |
integer, intent(in) :: c_ident |
| 96 |
integer, intent(inout) :: isError |
| 97 |
real( kind = dp ), intent(in) :: w0, v0, v0p |
| 98 |
real( kind = dp ), intent(in) :: rl, ru |
| 99 |
real( kind = dp ), intent(in) :: rlp, rup |
| 100 |
integer :: nATypes, myATID |
| 101 |
|
| 102 |
|
| 103 |
isError = 0 |
| 104 |
myATID = getFirstMatchingElement(atypes, "c_ident", c_ident) |
| 105 |
|
| 106 |
!! Be simple-minded and assume that we need a StickyMap that |
| 107 |
!! is the same size as the total number of atom types |
| 108 |
|
| 109 |
if (.not.allocated(StickyMap)) then |
| 110 |
|
| 111 |
nAtypes = getSize(atypes) |
| 112 |
|
| 113 |
if (nAtypes == 0) then |
| 114 |
isError = -1 |
| 115 |
return |
| 116 |
end if |
| 117 |
|
| 118 |
if (.not. allocated(StickyMap)) then |
| 119 |
allocate(StickyMap(nAtypes)) |
| 120 |
endif |
| 121 |
|
| 122 |
end if |
| 123 |
|
| 124 |
if (myATID .gt. size(StickyMap)) then |
| 125 |
isError = -1 |
| 126 |
return |
| 127 |
endif |
| 128 |
|
| 129 |
! set the values for StickyMap for this atom type: |
| 130 |
|
| 131 |
StickyMap(myATID)%c_ident = c_ident |
| 132 |
|
| 133 |
! we could pass all 5 parameters if we felt like it... |
| 134 |
|
| 135 |
StickyMap(myATID)%w0 = w0 |
| 136 |
StickyMap(myATID)%v0 = v0 |
| 137 |
StickyMap(myATID)%v0p = v0p |
| 138 |
StickyMap(myATID)%rl = rl |
| 139 |
StickyMap(myATID)%ru = ru |
| 140 |
StickyMap(myATID)%rlp = rlp |
| 141 |
StickyMap(myATID)%rup = rup |
| 142 |
|
| 143 |
if (StickyMap(myATID)%ru .gt. StickyMap(myATID)%rup) then |
| 144 |
StickyMap(myATID)%rbig = StickyMap(myATID)%ru |
| 145 |
else |
| 146 |
StickyMap(myATID)%rbig = StickyMap(myATID)%rup |
| 147 |
endif |
| 148 |
|
| 149 |
return |
| 150 |
end subroutine newStickyType |
| 151 |
|
| 152 |
function getStickyCut(atomID) result(cutValue) |
| 153 |
integer, intent(in) :: atomID |
| 154 |
real(kind=dp) :: cutValue |
| 155 |
|
| 156 |
cutValue = StickyMap(atomID)%rbig |
| 157 |
end function getStickyCut |
| 158 |
|
| 159 |
function getStickyPowerCut(atomID) result(cutValue) |
| 160 |
integer, intent(in) :: atomID |
| 161 |
real(kind=dp) :: cutValue |
| 162 |
|
| 163 |
cutValue = StickyMap(atomID)%rbig |
| 164 |
end function getStickyPowerCut |
| 165 |
|
| 166 |
subroutine do_sticky_pair(atom1, atom2, d, rij, r2, sw, vpair, fpair, & |
| 167 |
pot, A, f, t, do_pot) |
| 168 |
|
| 169 |
!! This routine does only the sticky portion of the SSD potential |
| 170 |
!! [Chandra and Ichiye, J. Chem. Phys. 111, 2701 (1999)]. |
| 171 |
!! The Lennard-Jones and dipolar interaction must be handled separately. |
| 172 |
|
| 173 |
!! We assume that the rotation matrices have already been calculated |
| 174 |
!! and placed in the A array. |
| 175 |
|
| 176 |
!! i and j are pointers to the two SSD atoms |
| 177 |
|
| 178 |
integer, intent(in) :: atom1, atom2 |
| 179 |
real (kind=dp), intent(inout) :: rij, r2 |
| 180 |
real (kind=dp), dimension(3), intent(in) :: d |
| 181 |
real (kind=dp), dimension(3), intent(inout) :: fpair |
| 182 |
real (kind=dp) :: pot, vpair, sw |
| 183 |
real (kind=dp), dimension(9,nLocal) :: A |
| 184 |
real (kind=dp), dimension(3,nLocal) :: f |
| 185 |
real (kind=dp), dimension(3,nLocal) :: t |
| 186 |
logical, intent(in) :: do_pot |
| 187 |
|
| 188 |
real (kind=dp) :: xi, yi, zi, xj, yj, zj, xi2, yi2, zi2, xj2, yj2, zj2 |
| 189 |
real (kind=dp) :: r3, r5, r6, s, sp, dsdr, dspdr |
| 190 |
real (kind=dp) :: wi, wj, w, wip, wjp, wp |
| 191 |
real (kind=dp) :: dwidx, dwidy, dwidz, dwjdx, dwjdy, dwjdz |
| 192 |
real (kind=dp) :: dwipdx, dwipdy, dwipdz, dwjpdx, dwjpdy, dwjpdz |
| 193 |
real (kind=dp) :: dwidux, dwiduy, dwiduz, dwjdux, dwjduy, dwjduz |
| 194 |
real (kind=dp) :: dwipdux, dwipduy, dwipduz, dwjpdux, dwjpduy, dwjpduz |
| 195 |
real (kind=dp) :: zif, zis, zjf, zjs, uglyi, uglyj |
| 196 |
real (kind=dp) :: drdx, drdy, drdz |
| 197 |
real (kind=dp) :: txi, tyi, tzi, txj, tyj, tzj |
| 198 |
real (kind=dp) :: fxii, fyii, fzii, fxjj, fyjj, fzjj |
| 199 |
real (kind=dp) :: fxij, fyij, fzij, fxji, fyji, fzji |
| 200 |
real (kind=dp) :: fxradial, fyradial, fzradial |
| 201 |
real (kind=dp) :: rijtest, rjitest |
| 202 |
real (kind=dp) :: radcomxi, radcomyi, radcomzi |
| 203 |
real (kind=dp) :: radcomxj, radcomyj, radcomzj |
| 204 |
integer :: id1, id2 |
| 205 |
integer :: me1, me2 |
| 206 |
real (kind=dp) :: w0, v0, v0p, rl, ru, rlp, rup, rbig |
| 207 |
|
| 208 |
if (.not.allocated(StickyMap)) then |
| 209 |
call handleError("sticky", "no StickyMap was present before first call of do_sticky_pair!") |
| 210 |
return |
| 211 |
end if |
| 212 |
|
| 213 |
#ifdef IS_MPI |
| 214 |
me1 = atid_Row(atom1) |
| 215 |
me2 = atid_Col(atom2) |
| 216 |
#else |
| 217 |
me1 = atid(atom1) |
| 218 |
me2 = atid(atom2) |
| 219 |
#endif |
| 220 |
|
| 221 |
if (me1.eq.me2) then |
| 222 |
w0 = StickyMap(me1)%w0 |
| 223 |
v0 = StickyMap(me1)%v0 |
| 224 |
v0p = StickyMap(me1)%v0p |
| 225 |
rl = StickyMap(me1)%rl |
| 226 |
ru = StickyMap(me1)%ru |
| 227 |
rlp = StickyMap(me1)%rlp |
| 228 |
rup = StickyMap(me1)%rup |
| 229 |
rbig = StickyMap(me1)%rbig |
| 230 |
else |
| 231 |
! This is silly, but if you want 2 sticky types in your |
| 232 |
! simulation, we'll let you do it with the Lorentz- |
| 233 |
! Berthelot mixing rules. |
| 234 |
! (Warning: you'll be SLLLLLLLLLLLLLLLOOOOOOOOOOWWWWWWWWWWW) |
| 235 |
rl = 0.5_dp * ( StickyMap(me1)%rl + StickyMap(me2)%rl ) |
| 236 |
ru = 0.5_dp * ( StickyMap(me1)%ru + StickyMap(me2)%ru ) |
| 237 |
rlp = 0.5_dp * ( StickyMap(me1)%rlp + StickyMap(me2)%rlp ) |
| 238 |
rup = 0.5_dp * ( StickyMap(me1)%rup + StickyMap(me2)%rup ) |
| 239 |
rbig = max(ru, rup) |
| 240 |
w0 = sqrt( StickyMap(me1)%w0 * StickyMap(me2)%w0 ) |
| 241 |
v0 = sqrt( StickyMap(me1)%v0 * StickyMap(me2)%v0 ) |
| 242 |
v0p = sqrt( StickyMap(me1)%v0p * StickyMap(me2)%v0p ) |
| 243 |
endif |
| 244 |
|
| 245 |
if ( rij .LE. rbig ) then |
| 246 |
|
| 247 |
r3 = r2*rij |
| 248 |
r5 = r3*r2 |
| 249 |
|
| 250 |
drdx = d(1) / rij |
| 251 |
drdy = d(2) / rij |
| 252 |
drdz = d(3) / rij |
| 253 |
|
| 254 |
#ifdef IS_MPI |
| 255 |
! rotate the inter-particle separation into the two different |
| 256 |
! body-fixed coordinate systems: |
| 257 |
|
| 258 |
xi = A_row(1,atom1)*d(1) + A_row(2,atom1)*d(2) + A_row(3,atom1)*d(3) |
| 259 |
yi = A_row(4,atom1)*d(1) + A_row(5,atom1)*d(2) + A_row(6,atom1)*d(3) |
| 260 |
zi = A_row(7,atom1)*d(1) + A_row(8,atom1)*d(2) + A_row(9,atom1)*d(3) |
| 261 |
|
| 262 |
! negative sign because this is the vector from j to i: |
| 263 |
|
| 264 |
xj = -(A_Col(1,atom2)*d(1) + A_Col(2,atom2)*d(2) + A_Col(3,atom2)*d(3)) |
| 265 |
yj = -(A_Col(4,atom2)*d(1) + A_Col(5,atom2)*d(2) + A_Col(6,atom2)*d(3)) |
| 266 |
zj = -(A_Col(7,atom2)*d(1) + A_Col(8,atom2)*d(2) + A_Col(9,atom2)*d(3)) |
| 267 |
#else |
| 268 |
! rotate the inter-particle separation into the two different |
| 269 |
! body-fixed coordinate systems: |
| 270 |
|
| 271 |
xi = a(1,atom1)*d(1) + a(2,atom1)*d(2) + a(3,atom1)*d(3) |
| 272 |
yi = a(4,atom1)*d(1) + a(5,atom1)*d(2) + a(6,atom1)*d(3) |
| 273 |
zi = a(7,atom1)*d(1) + a(8,atom1)*d(2) + a(9,atom1)*d(3) |
| 274 |
|
| 275 |
! negative sign because this is the vector from j to i: |
| 276 |
|
| 277 |
xj = -(a(1,atom2)*d(1) + a(2,atom2)*d(2) + a(3,atom2)*d(3)) |
| 278 |
yj = -(a(4,atom2)*d(1) + a(5,atom2)*d(2) + a(6,atom2)*d(3)) |
| 279 |
zj = -(a(7,atom2)*d(1) + a(8,atom2)*d(2) + a(9,atom2)*d(3)) |
| 280 |
#endif |
| 281 |
|
| 282 |
xi2 = xi*xi |
| 283 |
yi2 = yi*yi |
| 284 |
zi2 = zi*zi |
| 285 |
|
| 286 |
xj2 = xj*xj |
| 287 |
yj2 = yj*yj |
| 288 |
zj2 = zj*zj |
| 289 |
|
| 290 |
call calc_sw_fnc(rij, rl, ru, rlp, rup, s, sp, dsdr, dspdr) |
| 291 |
|
| 292 |
wi = 2.0d0*(xi2-yi2)*zi / r3 |
| 293 |
wj = 2.0d0*(xj2-yj2)*zj / r3 |
| 294 |
w = wi+wj |
| 295 |
|
| 296 |
zif = zi/rij - 0.6d0 |
| 297 |
zis = zi/rij + 0.8d0 |
| 298 |
|
| 299 |
zjf = zj/rij - 0.6d0 |
| 300 |
zjs = zj/rij + 0.8d0 |
| 301 |
|
| 302 |
wip = zif*zif*zis*zis - w0 |
| 303 |
wjp = zjf*zjf*zjs*zjs - w0 |
| 304 |
wp = wip + wjp |
| 305 |
|
| 306 |
vpair = vpair + 0.5d0*(v0*s*w + v0p*sp*wp) |
| 307 |
if (do_pot) then |
| 308 |
#ifdef IS_MPI |
| 309 |
pot_row(atom1) = pot_row(atom1) + 0.25d0*(v0*s*w + v0p*sp*wp)*sw |
| 310 |
pot_col(atom2) = pot_col(atom2) + 0.25d0*(v0*s*w + v0p*sp*wp)*sw |
| 311 |
#else |
| 312 |
pot = pot + 0.5d0*(v0*s*w + v0p*sp*wp)*sw |
| 313 |
#endif |
| 314 |
endif |
| 315 |
|
| 316 |
dwidx = 4.0d0*xi*zi/r3 - 6.0d0*xi*zi*(xi2-yi2)/r5 |
| 317 |
dwidy = - 4.0d0*yi*zi/r3 - 6.0d0*yi*zi*(xi2-yi2)/r5 |
| 318 |
dwidz = 2.0d0*(xi2-yi2)/r3 - 6.0d0*zi2*(xi2-yi2)/r5 |
| 319 |
|
| 320 |
dwjdx = 4.0d0*xj*zj/r3 - 6.0d0*xj*zj*(xj2-yj2)/r5 |
| 321 |
dwjdy = - 4.0d0*yj*zj/r3 - 6.0d0*yj*zj*(xj2-yj2)/r5 |
| 322 |
dwjdz = 2.0d0*(xj2-yj2)/r3 - 6.0d0*zj2*(xj2-yj2)/r5 |
| 323 |
|
| 324 |
uglyi = zif*zif*zis + zif*zis*zis |
| 325 |
uglyj = zjf*zjf*zjs + zjf*zjs*zjs |
| 326 |
|
| 327 |
dwipdx = -2.0d0*xi*zi*uglyi/r3 |
| 328 |
dwipdy = -2.0d0*yi*zi*uglyi/r3 |
| 329 |
dwipdz = 2.0d0*(1.0d0/rij - zi2/r3)*uglyi |
| 330 |
|
| 331 |
dwjpdx = -2.0d0*xj*zj*uglyj/r3 |
| 332 |
dwjpdy = -2.0d0*yj*zj*uglyj/r3 |
| 333 |
dwjpdz = 2.0d0*(1.0d0/rij - zj2/r3)*uglyj |
| 334 |
|
| 335 |
dwidux = 4.0d0*(yi*zi2 + 0.5d0*yi*(xi2-yi2))/r3 |
| 336 |
dwiduy = 4.0d0*(xi*zi2 - 0.5d0*xi*(xi2-yi2))/r3 |
| 337 |
dwiduz = - 8.0d0*xi*yi*zi/r3 |
| 338 |
|
| 339 |
dwjdux = 4.0d0*(yj*zj2 + 0.5d0*yj*(xj2-yj2))/r3 |
| 340 |
dwjduy = 4.0d0*(xj*zj2 - 0.5d0*xj*(xj2-yj2))/r3 |
| 341 |
dwjduz = - 8.0d0*xj*yj*zj/r3 |
| 342 |
|
| 343 |
dwipdux = 2.0d0*yi*uglyi/rij |
| 344 |
dwipduy = -2.0d0*xi*uglyi/rij |
| 345 |
dwipduz = 0.0d0 |
| 346 |
|
| 347 |
dwjpdux = 2.0d0*yj*uglyj/rij |
| 348 |
dwjpduy = -2.0d0*xj*uglyj/rij |
| 349 |
dwjpduz = 0.0d0 |
| 350 |
|
| 351 |
! do the torques first since they are easy: |
| 352 |
! remember that these are still in the body fixed axes |
| 353 |
|
| 354 |
txi = 0.5d0*(v0*s*dwidux + v0p*sp*dwipdux)*sw |
| 355 |
tyi = 0.5d0*(v0*s*dwiduy + v0p*sp*dwipduy)*sw |
| 356 |
tzi = 0.5d0*(v0*s*dwiduz + v0p*sp*dwipduz)*sw |
| 357 |
|
| 358 |
txj = 0.5d0*(v0*s*dwjdux + v0p*sp*dwjpdux)*sw |
| 359 |
tyj = 0.5d0*(v0*s*dwjduy + v0p*sp*dwjpduy)*sw |
| 360 |
tzj = 0.5d0*(v0*s*dwjduz + v0p*sp*dwjpduz)*sw |
| 361 |
|
| 362 |
! go back to lab frame using transpose of rotation matrix: |
| 363 |
|
| 364 |
#ifdef IS_MPI |
| 365 |
t_Row(1,atom1) = t_Row(1,atom1) + a_Row(1,atom1)*txi + & |
| 366 |
a_Row(4,atom1)*tyi + a_Row(7,atom1)*tzi |
| 367 |
t_Row(2,atom1) = t_Row(2,atom1) + a_Row(2,atom1)*txi + & |
| 368 |
a_Row(5,atom1)*tyi + a_Row(8,atom1)*tzi |
| 369 |
t_Row(3,atom1) = t_Row(3,atom1) + a_Row(3,atom1)*txi + & |
| 370 |
a_Row(6,atom1)*tyi + a_Row(9,atom1)*tzi |
| 371 |
|
| 372 |
t_Col(1,atom2) = t_Col(1,atom2) + a_Col(1,atom2)*txj + & |
| 373 |
a_Col(4,atom2)*tyj + a_Col(7,atom2)*tzj |
| 374 |
t_Col(2,atom2) = t_Col(2,atom2) + a_Col(2,atom2)*txj + & |
| 375 |
a_Col(5,atom2)*tyj + a_Col(8,atom2)*tzj |
| 376 |
t_Col(3,atom2) = t_Col(3,atom2) + a_Col(3,atom2)*txj + & |
| 377 |
a_Col(6,atom2)*tyj + a_Col(9,atom2)*tzj |
| 378 |
#else |
| 379 |
t(1,atom1) = t(1,atom1) + a(1,atom1)*txi + a(4,atom1)*tyi + a(7,atom1)*tzi |
| 380 |
t(2,atom1) = t(2,atom1) + a(2,atom1)*txi + a(5,atom1)*tyi + a(8,atom1)*tzi |
| 381 |
t(3,atom1) = t(3,atom1) + a(3,atom1)*txi + a(6,atom1)*tyi + a(9,atom1)*tzi |
| 382 |
|
| 383 |
t(1,atom2) = t(1,atom2) + a(1,atom2)*txj + a(4,atom2)*tyj + a(7,atom2)*tzj |
| 384 |
t(2,atom2) = t(2,atom2) + a(2,atom2)*txj + a(5,atom2)*tyj + a(8,atom2)*tzj |
| 385 |
t(3,atom2) = t(3,atom2) + a(3,atom2)*txj + a(6,atom2)*tyj + a(9,atom2)*tzj |
| 386 |
#endif |
| 387 |
! Now, on to the forces: |
| 388 |
|
| 389 |
! first rotate the i terms back into the lab frame: |
| 390 |
|
| 391 |
radcomxi = (v0*s*dwidx+v0p*sp*dwipdx)*sw |
| 392 |
radcomyi = (v0*s*dwidy+v0p*sp*dwipdy)*sw |
| 393 |
radcomzi = (v0*s*dwidz+v0p*sp*dwipdz)*sw |
| 394 |
|
| 395 |
radcomxj = (v0*s*dwjdx+v0p*sp*dwjpdx)*sw |
| 396 |
radcomyj = (v0*s*dwjdy+v0p*sp*dwjpdy)*sw |
| 397 |
radcomzj = (v0*s*dwjdz+v0p*sp*dwjpdz)*sw |
| 398 |
|
| 399 |
#ifdef IS_MPI |
| 400 |
fxii = a_Row(1,atom1)*(radcomxi) + & |
| 401 |
a_Row(4,atom1)*(radcomyi) + & |
| 402 |
a_Row(7,atom1)*(radcomzi) |
| 403 |
fyii = a_Row(2,atom1)*(radcomxi) + & |
| 404 |
a_Row(5,atom1)*(radcomyi) + & |
| 405 |
a_Row(8,atom1)*(radcomzi) |
| 406 |
fzii = a_Row(3,atom1)*(radcomxi) + & |
| 407 |
a_Row(6,atom1)*(radcomyi) + & |
| 408 |
a_Row(9,atom1)*(radcomzi) |
| 409 |
|
| 410 |
fxjj = a_Col(1,atom2)*(radcomxj) + & |
| 411 |
a_Col(4,atom2)*(radcomyj) + & |
| 412 |
a_Col(7,atom2)*(radcomzj) |
| 413 |
fyjj = a_Col(2,atom2)*(radcomxj) + & |
| 414 |
a_Col(5,atom2)*(radcomyj) + & |
| 415 |
a_Col(8,atom2)*(radcomzj) |
| 416 |
fzjj = a_Col(3,atom2)*(radcomxj)+ & |
| 417 |
a_Col(6,atom2)*(radcomyj) + & |
| 418 |
a_Col(9,atom2)*(radcomzj) |
| 419 |
#else |
| 420 |
fxii = a(1,atom1)*(radcomxi) + & |
| 421 |
a(4,atom1)*(radcomyi) + & |
| 422 |
a(7,atom1)*(radcomzi) |
| 423 |
fyii = a(2,atom1)*(radcomxi) + & |
| 424 |
a(5,atom1)*(radcomyi) + & |
| 425 |
a(8,atom1)*(radcomzi) |
| 426 |
fzii = a(3,atom1)*(radcomxi) + & |
| 427 |
a(6,atom1)*(radcomyi) + & |
| 428 |
a(9,atom1)*(radcomzi) |
| 429 |
|
| 430 |
fxjj = a(1,atom2)*(radcomxj) + & |
| 431 |
a(4,atom2)*(radcomyj) + & |
| 432 |
a(7,atom2)*(radcomzj) |
| 433 |
fyjj = a(2,atom2)*(radcomxj) + & |
| 434 |
a(5,atom2)*(radcomyj) + & |
| 435 |
a(8,atom2)*(radcomzj) |
| 436 |
fzjj = a(3,atom2)*(radcomxj)+ & |
| 437 |
a(6,atom2)*(radcomyj) + & |
| 438 |
a(9,atom2)*(radcomzj) |
| 439 |
#endif |
| 440 |
|
| 441 |
fxij = -fxii |
| 442 |
fyij = -fyii |
| 443 |
fzij = -fzii |
| 444 |
|
| 445 |
fxji = -fxjj |
| 446 |
fyji = -fyjj |
| 447 |
fzji = -fzjj |
| 448 |
|
| 449 |
! now assemble these with the radial-only terms: |
| 450 |
|
| 451 |
fxradial = 0.5d0*(v0*dsdr*drdx*w + v0p*dspdr*drdx*wp + fxii + fxji) |
| 452 |
fyradial = 0.5d0*(v0*dsdr*drdy*w + v0p*dspdr*drdy*wp + fyii + fyji) |
| 453 |
fzradial = 0.5d0*(v0*dsdr*drdz*w + v0p*dspdr*drdz*wp + fzii + fzji) |
| 454 |
|
| 455 |
#ifdef IS_MPI |
| 456 |
f_Row(1,atom1) = f_Row(1,atom1) + fxradial |
| 457 |
f_Row(2,atom1) = f_Row(2,atom1) + fyradial |
| 458 |
f_Row(3,atom1) = f_Row(3,atom1) + fzradial |
| 459 |
|
| 460 |
f_Col(1,atom2) = f_Col(1,atom2) - fxradial |
| 461 |
f_Col(2,atom2) = f_Col(2,atom2) - fyradial |
| 462 |
f_Col(3,atom2) = f_Col(3,atom2) - fzradial |
| 463 |
#else |
| 464 |
f(1,atom1) = f(1,atom1) + fxradial |
| 465 |
f(2,atom1) = f(2,atom1) + fyradial |
| 466 |
f(3,atom1) = f(3,atom1) + fzradial |
| 467 |
|
| 468 |
f(1,atom2) = f(1,atom2) - fxradial |
| 469 |
f(2,atom2) = f(2,atom2) - fyradial |
| 470 |
f(3,atom2) = f(3,atom2) - fzradial |
| 471 |
#endif |
| 472 |
|
| 473 |
#ifdef IS_MPI |
| 474 |
id1 = AtomRowToGlobal(atom1) |
| 475 |
id2 = AtomColToGlobal(atom2) |
| 476 |
#else |
| 477 |
id1 = atom1 |
| 478 |
id2 = atom2 |
| 479 |
#endif |
| 480 |
|
| 481 |
if (molMembershipList(id1) .ne. molMembershipList(id2)) then |
| 482 |
|
| 483 |
fpair(1) = fpair(1) + fxradial |
| 484 |
fpair(2) = fpair(2) + fyradial |
| 485 |
fpair(3) = fpair(3) + fzradial |
| 486 |
|
| 487 |
endif |
| 488 |
endif |
| 489 |
end subroutine do_sticky_pair |
| 490 |
|
| 491 |
!! calculates the switching functions and their derivatives for a given |
| 492 |
subroutine calc_sw_fnc(r, rl, ru, rlp, rup, s, sp, dsdr, dspdr) |
| 493 |
|
| 494 |
real (kind=dp), intent(in) :: r, rl, ru, rlp, rup |
| 495 |
real (kind=dp), intent(inout) :: s, sp, dsdr, dspdr |
| 496 |
|
| 497 |
! distances must be in angstroms |
| 498 |
|
| 499 |
if (r.lt.rl) then |
| 500 |
s = 1.0d0 |
| 501 |
dsdr = 0.0d0 |
| 502 |
elseif (r.gt.ru) then |
| 503 |
s = 0.0d0 |
| 504 |
dsdr = 0.0d0 |
| 505 |
else |
| 506 |
s = ((ru + 2.0d0*r - 3.0d0*rl) * (ru-r)**2) / & |
| 507 |
((ru - rl)**3) |
| 508 |
dsdr = 6.0d0*(r-ru)*(r-rl)/((ru - rl)**3) |
| 509 |
endif |
| 510 |
|
| 511 |
if (r.lt.rlp) then |
| 512 |
sp = 1.0d0 |
| 513 |
dspdr = 0.0d0 |
| 514 |
elseif (r.gt.rup) then |
| 515 |
sp = 0.0d0 |
| 516 |
dspdr = 0.0d0 |
| 517 |
else |
| 518 |
sp = ((rup + 2.0d0*r - 3.0d0*rlp) * (rup-r)**2) / & |
| 519 |
((rup - rlp)**3) |
| 520 |
dspdr = 6.0d0*(r-rup)*(r-rlp)/((rup - rlp)**3) |
| 521 |
endif |
| 522 |
|
| 523 |
return |
| 524 |
end subroutine calc_sw_fnc |
| 525 |
|
| 526 |
subroutine destroyStickyTypes() |
| 527 |
if(allocated(StickyMap)) deallocate(StickyMap) |
| 528 |
end subroutine destroyStickyTypes |
| 529 |
|
| 530 |
subroutine do_sticky_power_pair(atom1, atom2, d, rij, r2, sw, vpair, fpair, & |
| 531 |
pot, A, f, t, do_pot) |
| 532 |
!! We assume that the rotation matrices have already been calculated |
| 533 |
!! and placed in the A array. |
| 534 |
|
| 535 |
!! i and j are pointers to the two SSD atoms |
| 536 |
|
| 537 |
integer, intent(in) :: atom1, atom2 |
| 538 |
real (kind=dp), intent(inout) :: rij, r2 |
| 539 |
real (kind=dp), dimension(3), intent(in) :: d |
| 540 |
real (kind=dp), dimension(3), intent(inout) :: fpair |
| 541 |
real (kind=dp) :: pot, vpair, sw |
| 542 |
real (kind=dp), dimension(9,nLocal) :: A |
| 543 |
real (kind=dp), dimension(3,nLocal) :: f |
| 544 |
real (kind=dp), dimension(3,nLocal) :: t |
| 545 |
logical, intent(in) :: do_pot |
| 546 |
|
| 547 |
real (kind=dp) :: xi, yi, zi, xj, yj, zj, xi2, yi2, zi2, xj2, yj2, zj2 |
| 548 |
real (kind=dp) :: xihat, yihat, zihat, xjhat, yjhat, zjhat |
| 549 |
real (kind=dp) :: rI, rI2, rI3, rI4, rI5, rI6, rI7, s, sp, dsdr, dspdr |
| 550 |
real (kind=dp) :: wi, wj, w, wi2, wj2, eScale, v0scale |
| 551 |
real (kind=dp) :: dwidx, dwidy, dwidz, dwjdx, dwjdy, dwjdz |
| 552 |
real (kind=dp) :: dwidux, dwiduy, dwiduz, dwjdux, dwjduy, dwjduz |
| 553 |
real (kind=dp) :: drdx, drdy, drdz |
| 554 |
real (kind=dp) :: txi, tyi, tzi, txj, tyj, tzj |
| 555 |
real (kind=dp) :: fxii, fyii, fzii, fxjj, fyjj, fzjj |
| 556 |
real (kind=dp) :: fxij, fyij, fzij, fxji, fyji, fzji |
| 557 |
real (kind=dp) :: fxradial, fyradial, fzradial |
| 558 |
real (kind=dp) :: rijtest, rjitest |
| 559 |
real (kind=dp) :: radcomxi, radcomyi, radcomzi |
| 560 |
real (kind=dp) :: radcomxj, radcomyj, radcomzj |
| 561 |
integer :: id1, id2 |
| 562 |
integer :: me1, me2 |
| 563 |
real (kind=dp) :: w0, v0, v0p, rl, ru, rlp, rup, rbig |
| 564 |
real (kind=dp) :: zi3, zi4, zi5, zj3, zj4, zj5 |
| 565 |
real (kind=dp) :: frac1, frac2 |
| 566 |
|
| 567 |
if (.not.allocated(StickyMap)) then |
| 568 |
call handleError("sticky", "no StickyMap was present before first call of do_sticky_power_pair!") |
| 569 |
return |
| 570 |
end if |
| 571 |
|
| 572 |
#ifdef IS_MPI |
| 573 |
me1 = atid_Row(atom1) |
| 574 |
me2 = atid_Col(atom2) |
| 575 |
#else |
| 576 |
me1 = atid(atom1) |
| 577 |
me2 = atid(atom2) |
| 578 |
#endif |
| 579 |
|
| 580 |
if (me1.eq.me2) then |
| 581 |
w0 = StickyMap(me1)%w0 |
| 582 |
v0 = StickyMap(me1)%v0 |
| 583 |
v0p = StickyMap(me1)%v0p |
| 584 |
rl = StickyMap(me1)%rl |
| 585 |
ru = StickyMap(me1)%ru |
| 586 |
rlp = StickyMap(me1)%rlp |
| 587 |
rup = StickyMap(me1)%rup |
| 588 |
rbig = StickyMap(me1)%rbig |
| 589 |
else |
| 590 |
! This is silly, but if you want 2 sticky types in your |
| 591 |
! simulation, we'll let you do it with the Lorentz- |
| 592 |
! Berthelot mixing rules. |
| 593 |
! (Warning: you'll be SLLLLLLLLLLLLLLLOOOOOOOOOOWWWWWWWWWWW) |
| 594 |
rl = 0.5_dp * ( StickyMap(me1)%rl + StickyMap(me2)%rl ) |
| 595 |
ru = 0.5_dp * ( StickyMap(me1)%ru + StickyMap(me2)%ru ) |
| 596 |
rlp = 0.5_dp * ( StickyMap(me1)%rlp + StickyMap(me2)%rlp ) |
| 597 |
rup = 0.5_dp * ( StickyMap(me1)%rup + StickyMap(me2)%rup ) |
| 598 |
rbig = max(ru, rup) |
| 599 |
w0 = sqrt( StickyMap(me1)%w0 * StickyMap(me2)%w0 ) |
| 600 |
v0 = sqrt( StickyMap(me1)%v0 * StickyMap(me2)%v0 ) |
| 601 |
v0p = sqrt( StickyMap(me1)%v0p * StickyMap(me2)%v0p ) |
| 602 |
endif |
| 603 |
|
| 604 |
if ( rij .LE. rbig ) then |
| 605 |
|
| 606 |
rI = 1.0d0/rij |
| 607 |
rI2 = rI*rI |
| 608 |
rI3 = rI2*rI |
| 609 |
rI4 = rI2*rI2 |
| 610 |
rI5 = rI3*rI2 |
| 611 |
rI6 = rI3*rI3 |
| 612 |
rI7 = rI4*rI3 |
| 613 |
|
| 614 |
drdx = d(1) * rI |
| 615 |
drdy = d(2) * rI |
| 616 |
drdz = d(3) * rI |
| 617 |
|
| 618 |
#ifdef IS_MPI |
| 619 |
! rotate the inter-particle separation into the two different |
| 620 |
! body-fixed coordinate systems: |
| 621 |
|
| 622 |
xi = A_row(1,atom1)*d(1) + A_row(2,atom1)*d(2) + A_row(3,atom1)*d(3) |
| 623 |
yi = A_row(4,atom1)*d(1) + A_row(5,atom1)*d(2) + A_row(6,atom1)*d(3) |
| 624 |
zi = A_row(7,atom1)*d(1) + A_row(8,atom1)*d(2) + A_row(9,atom1)*d(3) |
| 625 |
|
| 626 |
! negative sign because this is the vector from j to i: |
| 627 |
|
| 628 |
xj = -(A_Col(1,atom2)*d(1) + A_Col(2,atom2)*d(2) + A_Col(3,atom2)*d(3)) |
| 629 |
yj = -(A_Col(4,atom2)*d(1) + A_Col(5,atom2)*d(2) + A_Col(6,atom2)*d(3)) |
| 630 |
zj = -(A_Col(7,atom2)*d(1) + A_Col(8,atom2)*d(2) + A_Col(9,atom2)*d(3)) |
| 631 |
#else |
| 632 |
! rotate the inter-particle separation into the two different |
| 633 |
! body-fixed coordinate systems: |
| 634 |
|
| 635 |
xi = a(1,atom1)*d(1) + a(2,atom1)*d(2) + a(3,atom1)*d(3) |
| 636 |
yi = a(4,atom1)*d(1) + a(5,atom1)*d(2) + a(6,atom1)*d(3) |
| 637 |
zi = a(7,atom1)*d(1) + a(8,atom1)*d(2) + a(9,atom1)*d(3) |
| 638 |
|
| 639 |
! negative sign because this is the vector from j to i: |
| 640 |
|
| 641 |
xj = -(a(1,atom2)*d(1) + a(2,atom2)*d(2) + a(3,atom2)*d(3)) |
| 642 |
yj = -(a(4,atom2)*d(1) + a(5,atom2)*d(2) + a(6,atom2)*d(3)) |
| 643 |
zj = -(a(7,atom2)*d(1) + a(8,atom2)*d(2) + a(9,atom2)*d(3)) |
| 644 |
#endif |
| 645 |
|
| 646 |
xi2 = xi*xi |
| 647 |
yi2 = yi*yi |
| 648 |
zi2 = zi*zi |
| 649 |
zi3 = zi2*zi |
| 650 |
zi4 = zi2*zi2 |
| 651 |
zi5 = zi3*zi2 |
| 652 |
xihat = xi*rI |
| 653 |
yihat = yi*rI |
| 654 |
zihat = zi*rI |
| 655 |
|
| 656 |
xj2 = xj*xj |
| 657 |
yj2 = yj*yj |
| 658 |
zj2 = zj*zj |
| 659 |
zj3 = zj2*zj |
| 660 |
zj4 = zj2*zj2 |
| 661 |
zj5 = zj3*zj2 |
| 662 |
xjhat = xj*rI |
| 663 |
yjhat = yj*rI |
| 664 |
zjhat = zj*rI |
| 665 |
|
| 666 |
call calc_sw_fnc(rij, rl, ru, rlp, rup, s, sp, dsdr, dspdr) |
| 667 |
|
| 668 |
frac1 = 0.25d0 |
| 669 |
frac2 = 0.75d0 |
| 670 |
|
| 671 |
wi = 2.0d0*(xi2-yi2)*zi*rI3 |
| 672 |
wj = 2.0d0*(xj2-yj2)*zj*rI3 |
| 673 |
|
| 674 |
wi2 = wi*wi |
| 675 |
wj2 = wj*wj |
| 676 |
|
| 677 |
w = frac1*wi*wi2 + frac2*wi + frac1*wj*wj2 + frac2*wj + v0p |
| 678 |
|
| 679 |
vpair = vpair + 0.5d0*(v0*s*w) |
| 680 |
|
| 681 |
if (do_pot) then |
| 682 |
#ifdef IS_MPI |
| 683 |
pot_row(atom1) = pot_row(atom1) + 0.25d0*(v0*s*w)*sw |
| 684 |
pot_col(atom2) = pot_col(atom2) + 0.25d0*(v0*s*w)*sw |
| 685 |
#else |
| 686 |
pot = pot + 0.5d0*(v0*s*w)*sw |
| 687 |
#endif |
| 688 |
endif |
| 689 |
|
| 690 |
dwidx = ( 4.0d0*xi*zi*rI3 - 6.0d0*xi*zi*(xi2-yi2)*rI5 ) |
| 691 |
dwidy = ( -4.0d0*yi*zi*rI3 - 6.0d0*yi*zi*(xi2-yi2)*rI5 ) |
| 692 |
dwidz = ( 2.0d0*(xi2-yi2)*rI3 - 6.0d0*zi2*(xi2-yi2)*rI5 ) |
| 693 |
|
| 694 |
dwidx = frac1*3.0d0*wi2*dwidx + frac2*dwidx |
| 695 |
dwidy = frac1*3.0d0*wi2*dwidy + frac2*dwidy |
| 696 |
dwidz = frac1*3.0d0*wi2*dwidz + frac2*dwidz |
| 697 |
|
| 698 |
dwjdx = ( 4.0d0*xj*zj*rI3 - 6.0d0*xj*zj*(xj2-yj2)*rI5 ) |
| 699 |
dwjdy = ( -4.0d0*yj*zj*rI3 - 6.0d0*yj*zj*(xj2-yj2)*rI5 ) |
| 700 |
dwjdz = ( 2.0d0*(xj2-yj2)*rI3 - 6.0d0*zj2*(xj2-yj2)*rI5 ) |
| 701 |
|
| 702 |
dwjdx = frac1*3.0d0*wj2*dwjdx + frac2*dwjdx |
| 703 |
dwjdy = frac1*3.0d0*wj2*dwjdy + frac2*dwjdy |
| 704 |
dwjdz = frac1*3.0d0*wj2*dwjdz + frac2*dwjdz |
| 705 |
|
| 706 |
dwidux = ( 4.0d0*(yi*zi2 + 0.5d0*yi*(xi2-yi2))*rI3 ) |
| 707 |
dwiduy = ( 4.0d0*(xi*zi2 - 0.5d0*xi*(xi2-yi2))*rI3 ) |
| 708 |
dwiduz = ( -8.0d0*xi*yi*zi*rI3 ) |
| 709 |
|
| 710 |
dwidux = frac1*3.0d0*wi2*dwidux + frac2*dwidux |
| 711 |
dwiduy = frac1*3.0d0*wi2*dwiduy + frac2*dwiduy |
| 712 |
dwiduz = frac1*3.0d0*wi2*dwiduz + frac2*dwiduz |
| 713 |
|
| 714 |
dwjdux = ( 4.0d0*(yj*zj2 + 0.5d0*yj*(xj2-yj2))*rI3 ) |
| 715 |
dwjduy = ( 4.0d0*(xj*zj2 - 0.5d0*xj*(xj2-yj2))*rI3 ) |
| 716 |
dwjduz = ( -8.0d0*xj*yj*zj*rI3 ) |
| 717 |
|
| 718 |
dwjdux = frac1*3.0d0*wj2*dwjdux + frac2*dwjdux |
| 719 |
dwjduy = frac1*3.0d0*wj2*dwjduy + frac2*dwjduy |
| 720 |
dwjduz = frac1*3.0d0*wj2*dwjduz + frac2*dwjduz |
| 721 |
|
| 722 |
! do the torques first since they are easy: |
| 723 |
! remember that these are still in the body fixed axes |
| 724 |
|
| 725 |
txi = 0.5d0*(v0*s*dwidux)*sw |
| 726 |
tyi = 0.5d0*(v0*s*dwiduy)*sw |
| 727 |
tzi = 0.5d0*(v0*s*dwiduz)*sw |
| 728 |
|
| 729 |
txj = 0.5d0*(v0*s*dwjdux)*sw |
| 730 |
tyj = 0.5d0*(v0*s*dwjduy)*sw |
| 731 |
tzj = 0.5d0*(v0*s*dwjduz)*sw |
| 732 |
|
| 733 |
! go back to lab frame using transpose of rotation matrix: |
| 734 |
|
| 735 |
#ifdef IS_MPI |
| 736 |
t_Row(1,atom1) = t_Row(1,atom1) + a_Row(1,atom1)*txi + & |
| 737 |
a_Row(4,atom1)*tyi + a_Row(7,atom1)*tzi |
| 738 |
t_Row(2,atom1) = t_Row(2,atom1) + a_Row(2,atom1)*txi + & |
| 739 |
a_Row(5,atom1)*tyi + a_Row(8,atom1)*tzi |
| 740 |
t_Row(3,atom1) = t_Row(3,atom1) + a_Row(3,atom1)*txi + & |
| 741 |
a_Row(6,atom1)*tyi + a_Row(9,atom1)*tzi |
| 742 |
|
| 743 |
t_Col(1,atom2) = t_Col(1,atom2) + a_Col(1,atom2)*txj + & |
| 744 |
a_Col(4,atom2)*tyj + a_Col(7,atom2)*tzj |
| 745 |
t_Col(2,atom2) = t_Col(2,atom2) + a_Col(2,atom2)*txj + & |
| 746 |
a_Col(5,atom2)*tyj + a_Col(8,atom2)*tzj |
| 747 |
t_Col(3,atom2) = t_Col(3,atom2) + a_Col(3,atom2)*txj + & |
| 748 |
a_Col(6,atom2)*tyj + a_Col(9,atom2)*tzj |
| 749 |
#else |
| 750 |
t(1,atom1) = t(1,atom1) + a(1,atom1)*txi + a(4,atom1)*tyi + a(7,atom1)*tzi |
| 751 |
t(2,atom1) = t(2,atom1) + a(2,atom1)*txi + a(5,atom1)*tyi + a(8,atom1)*tzi |
| 752 |
t(3,atom1) = t(3,atom1) + a(3,atom1)*txi + a(6,atom1)*tyi + a(9,atom1)*tzi |
| 753 |
|
| 754 |
t(1,atom2) = t(1,atom2) + a(1,atom2)*txj + a(4,atom2)*tyj + a(7,atom2)*tzj |
| 755 |
t(2,atom2) = t(2,atom2) + a(2,atom2)*txj + a(5,atom2)*tyj + a(8,atom2)*tzj |
| 756 |
t(3,atom2) = t(3,atom2) + a(3,atom2)*txj + a(6,atom2)*tyj + a(9,atom2)*tzj |
| 757 |
#endif |
| 758 |
! Now, on to the forces: |
| 759 |
|
| 760 |
! first rotate the i terms back into the lab frame: |
| 761 |
|
| 762 |
radcomxi = (v0*s*dwidx)*sw |
| 763 |
radcomyi = (v0*s*dwidy)*sw |
| 764 |
radcomzi = (v0*s*dwidz)*sw |
| 765 |
|
| 766 |
radcomxj = (v0*s*dwjdx)*sw |
| 767 |
radcomyj = (v0*s*dwjdy)*sw |
| 768 |
radcomzj = (v0*s*dwjdz)*sw |
| 769 |
|
| 770 |
#ifdef IS_MPI |
| 771 |
fxii = a_Row(1,atom1)*(radcomxi) + & |
| 772 |
a_Row(4,atom1)*(radcomyi) + & |
| 773 |
a_Row(7,atom1)*(radcomzi) |
| 774 |
fyii = a_Row(2,atom1)*(radcomxi) + & |
| 775 |
a_Row(5,atom1)*(radcomyi) + & |
| 776 |
a_Row(8,atom1)*(radcomzi) |
| 777 |
fzii = a_Row(3,atom1)*(radcomxi) + & |
| 778 |
a_Row(6,atom1)*(radcomyi) + & |
| 779 |
a_Row(9,atom1)*(radcomzi) |
| 780 |
|
| 781 |
fxjj = a_Col(1,atom2)*(radcomxj) + & |
| 782 |
a_Col(4,atom2)*(radcomyj) + & |
| 783 |
a_Col(7,atom2)*(radcomzj) |
| 784 |
fyjj = a_Col(2,atom2)*(radcomxj) + & |
| 785 |
a_Col(5,atom2)*(radcomyj) + & |
| 786 |
a_Col(8,atom2)*(radcomzj) |
| 787 |
fzjj = a_Col(3,atom2)*(radcomxj)+ & |
| 788 |
a_Col(6,atom2)*(radcomyj) + & |
| 789 |
a_Col(9,atom2)*(radcomzj) |
| 790 |
#else |
| 791 |
fxii = a(1,atom1)*(radcomxi) + & |
| 792 |
a(4,atom1)*(radcomyi) + & |
| 793 |
a(7,atom1)*(radcomzi) |
| 794 |
fyii = a(2,atom1)*(radcomxi) + & |
| 795 |
a(5,atom1)*(radcomyi) + & |
| 796 |
a(8,atom1)*(radcomzi) |
| 797 |
fzii = a(3,atom1)*(radcomxi) + & |
| 798 |
a(6,atom1)*(radcomyi) + & |
| 799 |
a(9,atom1)*(radcomzi) |
| 800 |
|
| 801 |
fxjj = a(1,atom2)*(radcomxj) + & |
| 802 |
a(4,atom2)*(radcomyj) + & |
| 803 |
a(7,atom2)*(radcomzj) |
| 804 |
fyjj = a(2,atom2)*(radcomxj) + & |
| 805 |
a(5,atom2)*(radcomyj) + & |
| 806 |
a(8,atom2)*(radcomzj) |
| 807 |
fzjj = a(3,atom2)*(radcomxj)+ & |
| 808 |
a(6,atom2)*(radcomyj) + & |
| 809 |
a(9,atom2)*(radcomzj) |
| 810 |
#endif |
| 811 |
|
| 812 |
fxij = -fxii |
| 813 |
fyij = -fyii |
| 814 |
fzij = -fzii |
| 815 |
|
| 816 |
fxji = -fxjj |
| 817 |
fyji = -fyjj |
| 818 |
fzji = -fzjj |
| 819 |
|
| 820 |
! now assemble these with the radial-only terms: |
| 821 |
|
| 822 |
fxradial = 0.5d0*(v0*dsdr*w*drdx + fxii + fxji) |
| 823 |
fyradial = 0.5d0*(v0*dsdr*w*drdy + fyii + fyji) |
| 824 |
fzradial = 0.5d0*(v0*dsdr*w*drdz + fzii + fzji) |
| 825 |
|
| 826 |
#ifdef IS_MPI |
| 827 |
f_Row(1,atom1) = f_Row(1,atom1) + fxradial |
| 828 |
f_Row(2,atom1) = f_Row(2,atom1) + fyradial |
| 829 |
f_Row(3,atom1) = f_Row(3,atom1) + fzradial |
| 830 |
|
| 831 |
f_Col(1,atom2) = f_Col(1,atom2) - fxradial |
| 832 |
f_Col(2,atom2) = f_Col(2,atom2) - fyradial |
| 833 |
f_Col(3,atom2) = f_Col(3,atom2) - fzradial |
| 834 |
#else |
| 835 |
f(1,atom1) = f(1,atom1) + fxradial |
| 836 |
f(2,atom1) = f(2,atom1) + fyradial |
| 837 |
f(3,atom1) = f(3,atom1) + fzradial |
| 838 |
|
| 839 |
f(1,atom2) = f(1,atom2) - fxradial |
| 840 |
f(2,atom2) = f(2,atom2) - fyradial |
| 841 |
f(3,atom2) = f(3,atom2) - fzradial |
| 842 |
#endif |
| 843 |
|
| 844 |
#ifdef IS_MPI |
| 845 |
id1 = AtomRowToGlobal(atom1) |
| 846 |
id2 = AtomColToGlobal(atom2) |
| 847 |
#else |
| 848 |
id1 = atom1 |
| 849 |
id2 = atom2 |
| 850 |
#endif |
| 851 |
|
| 852 |
if (molMembershipList(id1) .ne. molMembershipList(id2)) then |
| 853 |
|
| 854 |
fpair(1) = fpair(1) + fxradial |
| 855 |
fpair(2) = fpair(2) + fyradial |
| 856 |
fpair(3) = fpair(3) + fzradial |
| 857 |
|
| 858 |
endif |
| 859 |
endif |
| 860 |
end subroutine do_sticky_power_pair |
| 861 |
|
| 862 |
end module sticky |
