| 1 |
subroutine accumulate_rf(atom1, atom2, rij) |
| 2 |
|
| 3 |
include 'sizes.inc' |
| 4 |
include 'simulation.inc' |
| 5 |
|
| 6 |
integer atom1, atom2 |
| 7 |
double precision taper, rij |
| 8 |
|
| 9 |
if (rij.le.rrf) then |
| 10 |
|
| 11 |
if (rij.lt.rt) then |
| 12 |
taper = 1.0d0 |
| 13 |
else |
| 14 |
taper = (rrf + 2.0d0*rij - 3.0d0*rt)*(rrf-rij)**2/ ((rrf-rt)**3) |
| 15 |
endif |
| 16 |
|
| 17 |
rflx(atom1) = rflx(atom1) + ulx(atom2)*mu(atom2)*taper |
| 18 |
rfly(atom1) = rfly(atom1) + uly(atom2)*mu(atom2)*taper |
| 19 |
rflz(atom1) = rflz(atom1) + ulz(atom2)*mu(atom2)*taper |
| 20 |
|
| 21 |
rflx(atom2) = rflx(atom2) + ulx(atom1)*mu(atom1)*taper |
| 22 |
rfly(atom2) = rfly(atom2) + uly(atom1)*mu(atom1)*taper |
| 23 |
rflz(atom2) = rflz(atom2) + ulz(atom1)*mu(atom1)*taper |
| 24 |
|
| 25 |
endif |
| 26 |
return |
| 27 |
|
| 28 |
end subroutine accumulate_rf |
| 29 |
|
| 30 |
subroutine accumulate_self_rf() |
| 31 |
|
| 32 |
include 'sizes.inc' |
| 33 |
include 'simulation.inc' |
| 34 |
|
| 35 |
integer i, ia, a1, atype1 |
| 36 |
logical is_dipole_atype |
| 37 |
external is_dipole_atype |
| 38 |
|
| 39 |
do i = 1, nmol |
| 40 |
do ia = 1, na(i) |
| 41 |
a1 = atom_index(i,ia) |
| 42 |
|
| 43 |
atype1 = atype(a1) |
| 44 |
|
| 45 |
if (is_dipole_atype(atype1)) then |
| 46 |
|
| 47 |
rflx(a1) = rflx(a1) + ulx(a1)*mu(a1) |
| 48 |
rfly(a1) = rfly(a1) + uly(a1)*mu(a1) |
| 49 |
rflz(a1) = rflz(a1) + ulz(a1)*mu(a1) |
| 50 |
|
| 51 |
endif |
| 52 |
enddo |
| 53 |
enddo |
| 54 |
|
| 55 |
return |
| 56 |
end subroutine accumulate_self_rf |
| 57 |
|
| 58 |
subroutine reaction_field(pot) |
| 59 |
|
| 60 |
include 'sizes.inc' |
| 61 |
include 'simulation.inc' |
| 62 |
|
| 63 |
double precision rrfsq, pre |
| 64 |
integer i, ia, a1, atype1 |
| 65 |
logical is_dipole_atype |
| 66 |
external is_dipole_atype |
| 67 |
|
| 68 |
! do single loop to compute torques on dipoles: |
| 69 |
! pre converts from mu in units of debye to kcal/mol |
| 70 |
|
| 71 |
rrfsq = rrf * rrf |
| 72 |
pre = 14.38362d0*2.0d0*(dielect-1.0d0)/((2.0d0*dielect+1.0d0)*rrfsq*rrf) |
| 73 |
|
| 74 |
do i = 1, nmol |
| 75 |
do ia = 1, na(i) |
| 76 |
a1 = atom_index(i,ia) |
| 77 |
|
| 78 |
atype1 = atype(a1) |
| 79 |
|
| 80 |
if (is_dipole_atype(atype1)) then |
| 81 |
|
| 82 |
! The torque contribution is dipole cross reaction_field |
| 83 |
|
| 84 |
tlx(a1) = tlx(a1) + pre*mu(a1)*(uly(a1)*rflz(a1) - ulz(a1)*rfly(a1)) |
| 85 |
tly(a1) = tly(a1) + pre*mu(a1)*(ulz(a1)*rflx(a1) - ulx(a1)*rflz(a1)) |
| 86 |
tlz(a1) = tlz(a1) + pre*mu(a1)*(ulx(a1)*rfly(a1) - uly(a1)*rflx(a1)) |
| 87 |
|
| 88 |
! the potential contribution is -1/2 dipole dot reaction_field |
| 89 |
|
| 90 |
pot = pot - 0.5d0 * pre * mu(a1) * & |
| 91 |
(rflx(a1)*ulx(a1) + rfly(a1)*uly(a1) + rflz(a1)*ulz(a1)) |
| 92 |
|
| 93 |
endif |
| 94 |
|
| 95 |
enddo |
| 96 |
enddo |
| 97 |
|
| 98 |
end subroutine reaction_field |
| 99 |
|
| 100 |
subroutine rf_correct_forces(atom1, atom2, dx, dy, dz, rij) |
| 101 |
include 'sizes.inc' |
| 102 |
include 'simulation.inc' |
| 103 |
|
| 104 |
integer atom1, atom2 |
| 105 |
double precision dtdr, rrfsq, prerf, rij |
| 106 |
double precision dudx, dudy, dudz, u1dotu2, dx, dy, dz |
| 107 |
|
| 108 |
rrfsq = rrf * rrf |
| 109 |
prerf = 14.38362d0*2.0d0*(dielect-1.0d0)/((2.0d0*dielect+1.0d0)*rrfsq*rrf) |
| 110 |
|
| 111 |
if (rij.le.rrf) then |
| 112 |
|
| 113 |
! cubic taper |
| 114 |
if (rij.lt.rt) then |
| 115 |
dtdr = 0.0d0 |
| 116 |
else |
| 117 |
dtdr = 6.0d0*(rij*rij - rij*rt - rij*rrf +rrf*rt)/((rrf-rt)**3) |
| 118 |
endif |
| 119 |
|
| 120 |
u1dotu2 = ulx(atom1)*ulx(atom2) + uly(atom1)*uly(atom2) + & |
| 121 |
ulz(atom1)*ulz(atom2) |
| 122 |
|
| 123 |
dudx = - prerf*mu(atom1)*mu(atom2)*u1dotu2*dtdr*dx/rij |
| 124 |
dudy = - prerf*mu(atom1)*mu(atom2)*u1dotu2*dtdr*dy/rij |
| 125 |
dudz = - prerf*mu(atom1)*mu(atom2)*u1dotu2*dtdr*dz/rij |
| 126 |
|
| 127 |
flx(atom1) = flx(atom1) + dudx |
| 128 |
fly(atom1) = fly(atom1) + dudy |
| 129 |
flz(atom1) = flz(atom1) + dudz |
| 130 |
|
| 131 |
flx(atom2) = flx(atom2) - dudx |
| 132 |
fly(atom2) = fly(atom2) - dudy |
| 133 |
flz(atom2) = flz(atom2) - dudz |
| 134 |
|
| 135 |
! add contribution to the virial as well |
| 136 |
virial = virial + ( dx*dudx + dy*dudy + dz*dudz ) |
| 137 |
|
| 138 |
endif |
| 139 |
|
| 140 |
return |
| 141 |
end subroutine rf_correct_forces |
