ViewVC Help

View File | Revision Log | Show Annotations | View Changeset | Root Listing

root/group/trunk/OOPSE-4/src/UseTheForce/DarkSide/electrostatic.F90

Comparing trunk/OOPSE-4/src/UseTheForce/DarkSide/electrostatic.F90 (file contents):
Revision 2118 by gezelter, Fri Mar 11 15:53:18 2005 UTC vs.
Revision 2411 by chrisfen, Wed Nov 2 21:01:21 2005 UTC

#	Line 40 | Line 40 | module electrostatic_module
40		!!
41
42		module electrostatic_module
43	<
43	>
44		use force_globals
45		use definitions
46		use atype_module
#	Line 54 | Line 54 | module electrostatic_module
54
55		PRIVATE
56
57	+
58	+	#define __FORTRAN90
59	+	#include "UseTheForce/DarkSide/fInteractionMap.h"
60	+	#include "UseTheForce/DarkSide/fElectrostaticSummationMethod.h"
61	+	#include "UseTheForce/DarkSide/fScreeningMethod.h"
62	+
63	+
64		!! these prefactors convert the multipole interactions into kcal / mol
65		!! all were computed assuming distances are measured in angstroms
66		!! Charge-Charge, assuming charges are measured in electrons
#	Line 68 | Line 75 | module electrostatic_module
75		!! This unit is also known affectionately as an esu centi-barn.
76		real(kind=dp), parameter :: pre14 = 69.13373_dp
77
78	+	!! variables to handle different summation methods for long-range
79	+	!! electrostatics:
80	+	integer, save :: summationMethod = NONE
81	+	integer, save :: screeningMethod = UNDAMPED
82	+	logical, save :: summationMethodChecked = .false.
83	+	real(kind=DP), save :: defaultCutoff = 0.0_DP
84	+	real(kind=DP), save :: defaultCutoff2 = 0.0_DP
85	+	logical, save :: haveDefaultCutoff = .false.
86	+	real(kind=DP), save :: dampingAlpha = 0.0_DP
87	+	logical, save :: haveDampingAlpha = .false.
88	+	real(kind=DP), save :: dielectric = 1.0_DP
89	+	logical, save :: haveDielectric = .false.
90	+	real(kind=DP), save :: constERFC = 0.0_DP
91	+	real(kind=DP), save :: constEXP = 0.0_DP
92	+	real(kind=dp), save :: rcuti = 0.0_DP
93	+	real(kind=dp), save :: rcuti2 = 0.0_DP
94	+	real(kind=dp), save :: rcuti3 = 0.0_DP
95	+	real(kind=dp), save :: rcuti4 = 0.0_DP
96	+	real(kind=dp), save :: alphaPi = 0.0_DP
97	+	real(kind=dp), save :: invRootPi = 0.0_DP
98	+	real(kind=dp), save :: rrf = 1.0_DP
99	+	real(kind=dp), save :: rt = 1.0_DP
100	+	real(kind=dp), save :: rrfsq = 1.0_DP
101	+	real(kind=dp), save :: preRF = 0.0_DP
102	+	real(kind=dp), save :: preRF2 = 0.0_DP
103	+
104	+	#ifdef __IFC
105	+	! error function for ifc version > 7.
106	+	double precision, external :: derfc
107	+	#endif
108	+
109	+	public :: setElectrostaticSummationMethod
110	+	public :: setScreeningMethod
111	+	public :: setElectrostaticCutoffRadius
112	+	public :: setDampingAlpha
113	+	public :: setReactionFieldDielectric
114		public :: newElectrostaticType
115		public :: setCharge
116		public :: setDipoleMoment
#	Line 76 | Line 119 | module electrostatic_module
119		public :: doElectrostaticPair
120		public :: getCharge
121		public :: getDipoleMoment
122	+	public :: destroyElectrostaticTypes
123	+	public :: self_self
124	+	public :: rf_self_excludes
125
126		type :: Electrostatic
127		integer :: c_ident
#	Line 83 | Line 129 | module electrostatic_module
129		logical :: is_Dipole = .false.
130		logical :: is_SplitDipole = .false.
131		logical :: is_Quadrupole = .false.
132	+	logical :: is_Tap = .false.
133		real(kind=DP) :: charge = 0.0_DP
134		real(kind=DP) :: dipole_moment = 0.0_DP
135		real(kind=DP) :: split_dipole_distance = 0.0_DP
#	Line 93 | Line 140 | contains
140
141		contains
142
143	+	subroutine setElectrostaticSummationMethod(the_ESM)
144	+	integer, intent(in) :: the_ESM
145	+
146	+	if ((the_ESM .le. 0) .or. (the_ESM .gt. REACTION_FIELD)) then
147	+	call handleError("setElectrostaticSummationMethod", "Unsupported Summation Method")
148	+	endif
149	+
150	+	summationMethod = the_ESM
151	+
152	+	end subroutine setElectrostaticSummationMethod
153	+
154	+	subroutine setScreeningMethod(the_SM)
155	+	integer, intent(in) :: the_SM
156	+	screeningMethod = the_SM
157	+	end subroutine setScreeningMethod
158	+
159	+	subroutine setElectrostaticCutoffRadius(thisRcut, thisRsw)
160	+	real(kind=dp), intent(in) :: thisRcut
161	+	real(kind=dp), intent(in) :: thisRsw
162	+	defaultCutoff = thisRcut
163	+	rrf = defaultCutoff
164	+	rt = thisRsw
165	+	haveDefaultCutoff = .true.
166	+	end subroutine setElectrostaticCutoffRadius
167	+
168	+	subroutine setDampingAlpha(thisAlpha)
169	+	real(kind=dp), intent(in) :: thisAlpha
170	+	dampingAlpha = thisAlpha
171	+	haveDampingAlpha = .true.
172	+	end subroutine setDampingAlpha
173	+
174	+	subroutine setReactionFieldDielectric(thisDielectric)
175	+	real(kind=dp), intent(in) :: thisDielectric
176	+	dielectric = thisDielectric
177	+	haveDielectric = .true.
178	+	end subroutine setReactionFieldDielectric
179	+
180		subroutine newElectrostaticType(c_ident, is_Charge, is_Dipole, &
181	<	is_SplitDipole, is_Quadrupole, status)
182	<
181	>	is_SplitDipole, is_Quadrupole, is_Tap, status)
182	>
183		integer, intent(in) :: c_ident
184		logical, intent(in) :: is_Charge
185		logical, intent(in) :: is_Dipole
186		logical, intent(in) :: is_SplitDipole
187		logical, intent(in) :: is_Quadrupole
188	+	logical, intent(in) :: is_Tap
189		integer, intent(out) :: status
190		integer :: nAtypes, myATID, i, j
191
192		status = 0
193		myATID = getFirstMatchingElement(atypes, "c_ident", c_ident)
194	<
194	>
195		!! Be simple-minded and assume that we need an ElectrostaticMap that
196		!! is the same size as the total number of atom types
197
198		if (.not.allocated(ElectrostaticMap)) then
199	<
199	>
200		nAtypes = getSize(atypes)
201	<
201	>
202		if (nAtypes == 0) then
203		status = -1
204		return
205		end if
206	<
206	>
207		if (.not. allocated(ElectrostaticMap)) then
208		allocate(ElectrostaticMap(nAtypes))
209		endif
210	<
210	>
211		end if
212
213		if (myATID .gt. size(ElectrostaticMap)) then
214		status = -1
215		return
216		endif
217	<
217	>
218		! set the values for ElectrostaticMap for this atom type:
219
220		ElectrostaticMap(myATID)%c_ident = c_ident
#	Line 137 | Line 222 | contains
222		ElectrostaticMap(myATID)%is_Dipole = is_Dipole
223		ElectrostaticMap(myATID)%is_SplitDipole = is_SplitDipole
224		ElectrostaticMap(myATID)%is_Quadrupole = is_Quadrupole
225	<
225	>	ElectrostaticMap(myATID)%is_Tap = is_Tap
226	>
227		end subroutine newElectrostaticType
228
229		subroutine setCharge(c_ident, charge, status)
#	Line 165 | Line 251 | contains
251		call handleError("electrostatic", "Attempt to setCharge of an atom type that is not a charge!")
252		status = -1
253		return
254	<	endif
254	>	endif
255
256		ElectrostaticMap(myATID)%charge = charge
257		end subroutine setCharge
#	Line 256 | Line 342 | contains
342		status = -1
343		return
344		endif
345	<
345	>
346		do i = 1, 3
347	<	ElectrostaticMap(myATID)%quadrupole_moments(i) = &
348	<	quadrupole_moments(i)
349	<	enddo
347	>	ElectrostaticMap(myATID)%quadrupole_moments(i) = &
348	>	quadrupole_moments(i)
349	>	enddo
350
351		end subroutine setQuadrupoleMoments
352
353	<
353	>
354		function getCharge(atid) result (c)
355		integer, intent(in) :: atid
356		integer :: localError
357		real(kind=dp) :: c
358	<
358	>
359		if (.not.allocated(ElectrostaticMap)) then
360		call handleError("electrostatic", "no ElectrostaticMap was present before first call of getCharge!")
361		return
362		end if
363	<
363	>
364		if (.not.ElectrostaticMap(atid)%is_Charge) then
365		call handleError("electrostatic", "getCharge was called for an atom type that isn't a charge!")
366		return
367		endif
368	<
368	>
369		c = ElectrostaticMap(atid)%charge
370		end function getCharge
371
#	Line 287 | Line 373 | contains
373		integer, intent(in) :: atid
374		integer :: localError
375		real(kind=dp) :: dm
376	<
376	>
377		if (.not.allocated(ElectrostaticMap)) then
378		call handleError("electrostatic", "no ElectrostaticMap was present before first call of getDipoleMoment!")
379		return
380		end if
381	<
381	>
382		if (.not.ElectrostaticMap(atid)%is_Dipole) then
383		call handleError("electrostatic", "getDipoleMoment was called for an atom type that isn't a dipole!")
384		return
385		endif
386	<
386	>
387		dm = ElectrostaticMap(atid)%dipole_moment
388		end function getDipoleMoment
389
390	+	subroutine checkSummationMethod()
391	+
392	+	if (.not.haveDefaultCutoff) then
393	+	call handleError("checkSummationMethod", "no Default Cutoff set!")
394	+	endif
395	+
396	+	rcuti = 1.0d0 / defaultCutoff
397	+	rcuti2 = rcuti*rcuti
398	+	rcuti3 = rcuti2*rcuti
399	+	rcuti4 = rcuti2*rcuti2
400	+
401	+	if (screeningMethod .eq. DAMPED) then
402	+	if (.not.haveDampingAlpha) then
403	+	call handleError("checkSummationMethod", "no Damping Alpha set!")
404	+	endif
405	+
406	+	if (.not.haveDefaultCutoff) then
407	+	call handleError("checkSummationMethod", "no Default Cutoff set!")
408	+	endif
409	+
410	+	constEXP = exp(-dampingAlpha*dampingAlpha*defaultCutoff*defaultCutoff)
411	+	constERFC = derfc(dampingAlpha*defaultCutoff)
412	+	invRootPi = 0.56418958354775628695d0
413	+	alphaPi = 2*dampingAlpha*invRootPi
414	+
415	+	endif
416	+
417	+	if (summationMethod .eq. REACTION_FIELD) then
418	+	if (haveDielectric) then
419	+	defaultCutoff2 = defaultCutoff*defaultCutoff
420	+	preRF = (dielectric-1.0d0) / &
421	+	((2.0d0*dielectric+1.0d0)*defaultCutoff2*defaultCutoff)
422	+	preRF2 = 2.0d0*preRF
423	+	else
424	+	call handleError("checkSummationMethod", "Dielectric not set")
425	+	endif
426	+
427	+	endif
428	+
429	+	summationMethodChecked = .true.
430	+	end subroutine checkSummationMethod
431	+
432	+
433		subroutine doElectrostaticPair(atom1, atom2, d, rij, r2, sw, &
434		vpair, fpair, pot, eFrame, f, t, do_pot)
435	<
435	>
436		logical, intent(in) :: do_pot
437	<
437	>
438		integer, intent(in) :: atom1, atom2
439		integer :: localError
440
441		real(kind=dp), intent(in) :: rij, r2, sw
442		real(kind=dp), intent(in), dimension(3) :: d
443		real(kind=dp), intent(inout) :: vpair
444	<	real(kind=dp), intent(inout), dimension(3) :: fpair
444	>	real(kind=dp), intent(inout), dimension(3) :: fpair
445
446		real( kind = dp ) :: pot
447		real( kind = dp ), dimension(9,nLocal) :: eFrame
448		real( kind = dp ), dimension(3,nLocal) :: f
449	+	real( kind = dp ), dimension(3,nLocal) :: felec
450		real( kind = dp ), dimension(3,nLocal) :: t
321	–
322	–	real (kind = dp), dimension(3) :: ul_i
323	–	real (kind = dp), dimension(3) :: ul_j
451
452	+	real (kind = dp), dimension(3) :: ux_i, uy_i, uz_i
453	+	real (kind = dp), dimension(3) :: ux_j, uy_j, uz_j
454	+	real (kind = dp), dimension(3) :: dudux_i, duduy_i, duduz_i
455	+	real (kind = dp), dimension(3) :: dudux_j, duduy_j, duduz_j
456	+
457		logical :: i_is_Charge, i_is_Dipole, i_is_SplitDipole, i_is_Quadrupole
458		logical :: j_is_Charge, j_is_Dipole, j_is_SplitDipole, j_is_Quadrupole
459	+	logical :: i_is_Tap, j_is_Tap
460		integer :: me1, me2, id1, id2
461		real (kind=dp) :: q_i, q_j, mu_i, mu_j, d_i, d_j
462	+	real (kind=dp) :: qxx_i, qyy_i, qzz_i
463	+	real (kind=dp) :: qxx_j, qyy_j, qzz_j
464	+	real (kind=dp) :: cx_i, cy_i, cz_i
465	+	real (kind=dp) :: cx_j, cy_j, cz_j
466	+	real (kind=dp) :: cx2, cy2, cz2
467		real (kind=dp) :: ct_i, ct_j, ct_ij, a1
468		real (kind=dp) :: riji, ri, ri2, ri3, ri4
469	<	real (kind=dp) :: pref, vterm, epot, dudr
469	>	real (kind=dp) :: pref, vterm, epot, dudr, vterm1, vterm2
470		real (kind=dp) :: xhat, yhat, zhat
471		real (kind=dp) :: dudx, dudy, dudz
334	–	real (kind=dp) :: drdxj, drdyj, drdzj
335	–	real (kind=dp) :: duduix, duduiy, duduiz, dudujx, dudujy, dudujz
472		real (kind=dp) :: scale, sc2, bigR
473	+	real (kind=dp) :: varERFC, varEXP
474	+	real (kind=dp) :: limScale
475	+	real (kind=dp) :: preVal, rfVal
476
477		if (.not.allocated(ElectrostaticMap)) then
478		call handleError("electrostatic", "no ElectrostaticMap was present before first call of do_electrostatic_pair!")
479		return
480		end if
481
482	+	if (.not.summationMethodChecked) then
483	+	call checkSummationMethod()
484	+	endif
485	+
486		#ifdef IS_MPI
487		me1 = atid_Row(atom1)
488		me2 = atid_Col(atom2)
#	Line 348 | Line 491 | contains
491		me2 = atid(atom2)
492		#endif
493
494	+	!!$    if (rij .ge. defaultCutoff) then
495	+	!!$       write(*,*) 'warning: rij = ', rij, ' rcut = ', defaultCutoff, ' sw = ', sw
496	+	!!$    endif
497	+
498		!! some variables we'll need independent of electrostatic type:
499
500		riji = 1.0d0 / rij
501	<
501	>
502		xhat = d(1) * riji
503		yhat = d(2) * riji
504		zhat = d(3) * riji
505
359	–	drdxj = xhat
360	–	drdyj = yhat
361	–	drdzj = zhat
362	–
506		!! logicals
364	–
507		i_is_Charge = ElectrostaticMap(me1)%is_Charge
508		i_is_Dipole = ElectrostaticMap(me1)%is_Dipole
509		i_is_SplitDipole = ElectrostaticMap(me1)%is_SplitDipole
510		i_is_Quadrupole = ElectrostaticMap(me1)%is_Quadrupole
511	+	i_is_Tap = ElectrostaticMap(me1)%is_Tap
512
513		j_is_Charge = ElectrostaticMap(me2)%is_Charge
514		j_is_Dipole = ElectrostaticMap(me2)%is_Dipole
515		j_is_SplitDipole = ElectrostaticMap(me2)%is_SplitDipole
516		j_is_Quadrupole = ElectrostaticMap(me2)%is_Quadrupole
517	+	j_is_Tap = ElectrostaticMap(me2)%is_Tap
518
519		if (i_is_Charge) then
520		q_i = ElectrostaticMap(me1)%charge
521		endif
522	<
522	>
523		if (i_is_Dipole) then
524		mu_i = ElectrostaticMap(me1)%dipole_moment
525		#ifdef IS_MPI
526	<	ul_i(1) = eFrame_Row(3,atom1)
527	<	ul_i(2) = eFrame_Row(6,atom1)
528	<	ul_i(3) = eFrame_Row(9,atom1)
526	>	uz_i(1) = eFrame_Row(3,atom1)
527	>	uz_i(2) = eFrame_Row(6,atom1)
528	>	uz_i(3) = eFrame_Row(9,atom1)
529		#else
530	<	ul_i(1) = eFrame(3,atom1)
531	<	ul_i(2) = eFrame(6,atom1)
532	<	ul_i(3) = eFrame(9,atom1)
530	>	uz_i(1) = eFrame(3,atom1)
531	>	uz_i(2) = eFrame(6,atom1)
532	>	uz_i(3) = eFrame(9,atom1)
533		#endif
534	<	ct_i = ul_i(1)*drdxj + ul_i(2)*drdyj + ul_i(3)*drdzj
534	>	ct_i = uz_i(1)*xhat + uz_i(2)*yhat + uz_i(3)*zhat
535
536		if (i_is_SplitDipole) then
537		d_i = ElectrostaticMap(me1)%split_dipole_distance
538		endif
539	<
539	>
540		endif
541
542	+	if (i_is_Quadrupole) then
543	+	qxx_i = ElectrostaticMap(me1)%quadrupole_moments(1)
544	+	qyy_i = ElectrostaticMap(me1)%quadrupole_moments(2)
545	+	qzz_i = ElectrostaticMap(me1)%quadrupole_moments(3)
546	+	#ifdef IS_MPI
547	+	ux_i(1) = eFrame_Row(1,atom1)
548	+	ux_i(2) = eFrame_Row(4,atom1)
549	+	ux_i(3) = eFrame_Row(7,atom1)
550	+	uy_i(1) = eFrame_Row(2,atom1)
551	+	uy_i(2) = eFrame_Row(5,atom1)
552	+	uy_i(3) = eFrame_Row(8,atom1)
553	+	uz_i(1) = eFrame_Row(3,atom1)
554	+	uz_i(2) = eFrame_Row(6,atom1)
555	+	uz_i(3) = eFrame_Row(9,atom1)
556	+	#else
557	+	ux_i(1) = eFrame(1,atom1)
558	+	ux_i(2) = eFrame(4,atom1)
559	+	ux_i(3) = eFrame(7,atom1)
560	+	uy_i(1) = eFrame(2,atom1)
561	+	uy_i(2) = eFrame(5,atom1)
562	+	uy_i(3) = eFrame(8,atom1)
563	+	uz_i(1) = eFrame(3,atom1)
564	+	uz_i(2) = eFrame(6,atom1)
565	+	uz_i(3) = eFrame(9,atom1)
566	+	#endif
567	+	cx_i = ux_i(1)*xhat + ux_i(2)*yhat + ux_i(3)*zhat
568	+	cy_i = uy_i(1)*xhat + uy_i(2)*yhat + uy_i(3)*zhat
569	+	cz_i = uz_i(1)*xhat + uz_i(2)*yhat + uz_i(3)*zhat
570	+	endif
571	+
572		if (j_is_Charge) then
573		q_j = ElectrostaticMap(me2)%charge
574		endif
575	<
575	>
576		if (j_is_Dipole) then
577		mu_j = ElectrostaticMap(me2)%dipole_moment
578		#ifdef IS_MPI
579	<	ul_j(1) = eFrame_Col(3,atom2)
580	<	ul_j(2) = eFrame_Col(6,atom2)
581	<	ul_j(3) = eFrame_Col(9,atom2)
579	>	uz_j(1) = eFrame_Col(3,atom2)
580	>	uz_j(2) = eFrame_Col(6,atom2)
581	>	uz_j(3) = eFrame_Col(9,atom2)
582		#else
583	<	ul_j(1) = eFrame(3,atom2)
584	<	ul_j(2) = eFrame(6,atom2)
585	<	ul_j(3) = eFrame(9,atom2)
583	>	uz_j(1) = eFrame(3,atom2)
584	>	uz_j(2) = eFrame(6,atom2)
585	>	uz_j(3) = eFrame(9,atom2)
586		#endif
587	<	ct_j = ul_j(1)*drdxj + ul_j(2)*drdyj + ul_j(3)*drdzj
587	>	ct_j = uz_j(1)*xhat + uz_j(2)*yhat + uz_j(3)*zhat
588
589		if (j_is_SplitDipole) then
590		d_j = ElectrostaticMap(me2)%split_dipole_distance
591		endif
592		endif
593
594	+	if (j_is_Quadrupole) then
595	+	qxx_j = ElectrostaticMap(me2)%quadrupole_moments(1)
596	+	qyy_j = ElectrostaticMap(me2)%quadrupole_moments(2)
597	+	qzz_j = ElectrostaticMap(me2)%quadrupole_moments(3)
598	+	#ifdef IS_MPI
599	+	ux_j(1) = eFrame_Col(1,atom2)
600	+	ux_j(2) = eFrame_Col(4,atom2)
601	+	ux_j(3) = eFrame_Col(7,atom2)
602	+	uy_j(1) = eFrame_Col(2,atom2)
603	+	uy_j(2) = eFrame_Col(5,atom2)
604	+	uy_j(3) = eFrame_Col(8,atom2)
605	+	uz_j(1) = eFrame_Col(3,atom2)
606	+	uz_j(2) = eFrame_Col(6,atom2)
607	+	uz_j(3) = eFrame_Col(9,atom2)
608	+	#else
609	+	ux_j(1) = eFrame(1,atom2)
610	+	ux_j(2) = eFrame(4,atom2)
611	+	ux_j(3) = eFrame(7,atom2)
612	+	uy_j(1) = eFrame(2,atom2)
613	+	uy_j(2) = eFrame(5,atom2)
614	+	uy_j(3) = eFrame(8,atom2)
615	+	uz_j(1) = eFrame(3,atom2)
616	+	uz_j(2) = eFrame(6,atom2)
617	+	uz_j(3) = eFrame(9,atom2)
618	+	#endif
619	+	cx_j = ux_j(1)*xhat + ux_j(2)*yhat + ux_j(3)*zhat
620	+	cy_j = uy_j(1)*xhat + uy_j(2)*yhat + uy_j(3)*zhat
621	+	cz_j = uz_j(1)*xhat + uz_j(2)*yhat + uz_j(3)*zhat
622	+	endif
623	+
624		epot = 0.0_dp
625		dudx = 0.0_dp
626		dudy = 0.0_dp
627		dudz = 0.0_dp
628
629	<	duduix = 0.0_dp
630	<	duduiy = 0.0_dp
631	<	duduiz = 0.0_dp
629	>	dudux_i = 0.0_dp
630	>	duduy_i = 0.0_dp
631	>	duduz_i = 0.0_dp
632
633	<	dudujx = 0.0_dp
634	<	dudujy = 0.0_dp
635	<	dudujz = 0.0_dp
633	>	dudux_j = 0.0_dp
634	>	duduy_j = 0.0_dp
635	>	duduz_j = 0.0_dp
636
637		if (i_is_Charge) then
638
639		if (j_is_Charge) then
436	–
437	–	vterm = pre11 * q_i * q_j * riji
438	–	vpair = vpair + vterm
439	–	epot = epot + sw*vterm
640
641	<	dudr  = - sw * vterm * riji
641	>	if (summationMethod .eq. SHIFTED_POTENTIAL) then
642
643	<	dudx = dudx + dudr * drdxj
644	<	dudy = dudy + dudr * drdyj
645	<	dudz = dudz + dudr * drdzj
646	<
647	<	endif
643	>	vterm = pre11 * q_i * q_j * (riji - rcuti)
644	>	vpair = vpair + vterm
645	>	epot = epot + sw*vterm
646	>
647	>	dudr  = -sw*pre11*q_i*q_j * (riji*riji)
648	>
649	>	dudx = dudx + dudr * xhat
650	>	dudy = dudy + dudr * yhat
651	>	dudz = dudz + dudr * zhat
652
653	<	if (j_is_Dipole) then
653	>	!!$          elseif (summationMethod .eq. DAMPED_WOLF) then
654	>	!!$             varERFC = derfc(dampingAlpha*rij)
655	>	!!$             varEXP = exp(-dampingAlpha*dampingAlpha*rij*rij)
656	>	!!$             vterm = pre11 * q_i * q_j * (varERFC*riji - constERFC*rcuti)
657	>	!!$             vpair = vpair + vterm
658	>	!!$             epot = epot + sw*vterm
659	>	!!$
660	>	!!$!             dudr  = -sw*pre11*q_i*q_j * (((varERFC*riji*riji &
661	>	!!$!                  + alphaPi*varEXP*riji) - (constERFC*rcuti2 &
662	>	!!$!                  + alphaPi*constEXP*rcuti)) )
663	>	!!$             dudr  = -sw*pre11*q_i*q_j * (varERFC*riji*riji &
664	>	!!$                  + alphaPi*varEXP*riji)
665	>	!!$
666	>	!!$             dudx = dudx + dudr * xhat
667	>	!!$             dudy = dudy + dudr * yhat
668	>	!!$             dudz = dudz + dudr * zhat
669
670	<	if (j_is_SplitDipole) then
671	<	BigR = sqrt(r2 + 0.25_dp * d_j * d_j)
672	<	ri = 1.0_dp / BigR
673	<	scale = rij * ri
455	<	else
456	<	ri = riji
457	<	scale = 1.0_dp
458	<	endif
459	<
460	<	ri2 = ri * ri
461	<	ri3 = ri2 * ri
462	<	sc2 = scale * scale
670	>	elseif (summationMethod .eq. SHIFTED_FORCE) then
671	>	vterm = pre11 * q_i * q_j * (riji + rij*rcuti2 - 2.0d0*rcuti)
672	>	vpair = vpair + vterm
673	>	epot = epot + sw*vterm
674
675	<	pref = pre12 * q_i * mu_j
676	<	vterm = pref * ct_j * ri2 * scale
677	<	vpair = vpair + vterm
678	<	epot = epot + sw * vterm
675	>	dudr  = -sw*pre11*q_i*q_j * (riji*riji-rcuti2)
676	>
677	>	dudx = dudx + dudr * xhat
678	>	dudy = dudy + dudr * yhat
679	>	dudz = dudz + dudr * zhat
680
681	<	!! this has a + sign in the () because the rij vector is
682	<	!! r_j - r_i and the charge-dipole potential takes the origin
683	<	!! as the point dipole, which is atom j in this case.
681	>	!!$          elseif (summationMethod .eq. DAMPED_SHIFTED_FORCE) then
682	>	!!$             varERFC = derfc(dampingAlpha*rij)
683	>	!!$             varEXP = exp(-dampingAlpha*dampingAlpha*rij*rij)
684	>	!!$             vterm = pre11 * q_i * q_j * (varERFC*riji - constERFC*rcuti)
685	>	!!$             vpair = vpair + vterm
686	>	!!$             epot = epot + sw*vterm
687	>	!!$
688	>	!!$!             dudr  = -sw*pre11*q_i*q_j * (((varERFC*riji*riji &
689	>	!!$!                  + alphaPi*varEXP*riji) - (constERFC*rcuti2 &
690	>	!!$!                  + alphaPi*constEXP*rcuti)) )
691	>	!!$             dudr  = -sw*pre11*q_i*q_j * (varERFC*riji*riji &
692	>	!!$                  + alphaPi*varEXP*riji)
693	>	!!$
694	>	!!$             dudx = dudx + dudr * xhat
695	>	!!$             dudy = dudy + dudr * yhat
696	>	!!$             dudz = dudz + dudr * zhat
697
698	<	dudx = dudx + pref * sw * ri3 * ( ul_j(1) + 3.0d0*ct_j*xhat*sc2)
699	<	dudy = dudy + pref * sw * ri3 * ( ul_j(2) + 3.0d0*ct_j*yhat*sc2)
700	<	dudz = dudz + pref * sw * ri3 * ( ul_j(3) + 3.0d0*ct_j*zhat*sc2)
698	>	elseif (summationMethod .eq. REACTION_FIELD) then
699	>	preVal = pre11 * q_i * q_j
700	>	rfVal = preRF*rij*rij
701	>	vterm = preVal * ( riji + rfVal )
702	>
703	>	vpair = vpair + vterm
704	>	epot = epot + sw*vterm
705	>
706	>	dudr  = sw * preVal * ( 2.0d0*rfVal - riji )*riji
707	>
708	>	dudx = dudx + dudr * xhat
709	>	dudy = dudy + dudr * yhat
710	>	dudz = dudz + dudr * zhat
711
477	–	dudujx = dudujx - pref * sw * ri2 * xhat * scale
478	–	dudujy = dudujy - pref * sw * ri2 * yhat * scale
479	–	dudujz = dudujz - pref * sw * ri2 * zhat * scale
480	–
481	–	endif
482	–
483	–	endif
484	–
485	–	if (i_is_Dipole) then
486	–
487	–	if (j_is_Charge) then
488	–
489	–	if (i_is_SplitDipole) then
490	–	BigR = sqrt(r2 + 0.25_dp * d_i * d_i)
491	–	ri = 1.0_dp / BigR
492	–	scale = rij * ri
712		else
713	<	ri = riji
714	<	scale = 1.0_dp
715	<	endif
497	<
498	<	ri2 = ri * ri
499	<	ri3 = ri2 * ri
500	<	sc2 = scale * scale
713	>	vterm = pre11 * q_i * q_j * riji
714	>	vpair = vpair + vterm
715	>	epot = epot + sw*vterm
716
717	<	pref = pre12 * q_j * mu_i
718	<	vterm = pref * ct_i * ri2 * scale
719	<	vpair = vpair + vterm
720	<	epot = epot + sw * vterm
717	>	dudr  = - sw * vterm * riji
718	>
719	>	dudx = dudx + dudr * xhat
720	>	dudy = dudy + dudr * yhat
721	>	dudz = dudz + dudr * zhat
722
723	<	dudx = dudx + pref * sw * ri3 * ( ul_i(1) - 3.0d0 * ct_i * xhat*sc2)
508	<	dudy = dudy + pref * sw * ri3 * ( ul_i(2) - 3.0d0 * ct_i * yhat*sc2)
509	<	dudz = dudz + pref * sw * ri3 * ( ul_i(3) - 3.0d0 * ct_i * zhat*sc2)
723	>	endif
724
511	–	duduix = duduix + pref * sw * ri2 * xhat * scale
512	–	duduiy = duduiy + pref * sw * ri2 * yhat * scale
513	–	duduiz = duduiz + pref * sw * ri2 * zhat * scale
725		endif
726
727		if (j_is_Dipole) then
728
729	<	if (i_is_SplitDipole) then
730	<	if (j_is_SplitDipole) then
731	<	BigR = sqrt(r2 + 0.25_dp * d_i * d_i + 0.25_dp * d_j * d_j)
732	<	else
733	<	BigR = sqrt(r2 + 0.25_dp * d_i * d_i)
734	<	endif
735	<	ri = 1.0_dp / BigR
736	<	scale = rij * ri
729	>	pref = pre12 * q_i * mu_j
730	>
731	>	!!$          if (summationMethod .eq. UNDAMPED_WOLF) then
732	>	!!$             ri2 = riji * riji
733	>	!!$             ri3 = ri2 * riji
734	>	!!$
735	>	!!$             pref = pre12 * q_i * mu_j
736	>	!!$             vterm = - pref * ct_j * (ri2 - rcuti2)
737	>	!!$             vpair = vpair + vterm
738	>	!!$             epot = epot + sw*vterm
739	>	!!$
740	>	!!$             !! this has a + sign in the () because the rij vector is
741	>	!!$             !! r_j - r_i and the charge-dipole potential takes the origin
742	>	!!$             !! as the point dipole, which is atom j in this case.
743	>	!!$
744	>	!!$             dudx = dudx - sw*pref * ( ri3*( uz_j(1) - 3.0d0*ct_j*xhat) &
745	>	!!$                  - rcuti3*( uz_j(1) - 3.0d0*ct_j*d(1)*rcuti ) )
746	>	!!$             dudy = dudy - sw*pref * ( ri3*( uz_j(2) - 3.0d0*ct_j*yhat) &
747	>	!!$                  - rcuti3*( uz_j(2) - 3.0d0*ct_j*d(2)*rcuti ) )
748	>	!!$             dudz = dudz - sw*pref * ( ri3*( uz_j(3) - 3.0d0*ct_j*zhat) &
749	>	!!$                  - rcuti3*( uz_j(3) - 3.0d0*ct_j*d(3)*rcuti ) )
750	>	!!$
751	>	!!$             duduz_j(1) = duduz_j(1) - sw*pref*( ri2*xhat - d(1)*rcuti3 )
752	>	!!$             duduz_j(2) = duduz_j(2) - sw*pref*( ri2*yhat - d(2)*rcuti3 )
753	>	!!$             duduz_j(3) = duduz_j(3) - sw*pref*( ri2*zhat - d(3)*rcuti3 )
754	>	!!$
755	>	!!$          elseif (summationMethod .eq. REACTION_FIELD) then
756	>
757	>	if (summationMethod .eq. REACTION_FIELD) then
758	>	ri2 = riji * riji
759	>	ri3 = ri2 * riji
760	>
761	>	pref = pre12 * q_i * mu_j
762	>	vterm = - pref * ct_j * ( ri2 - preRF2*rij )
763	>	vpair = vpair + vterm
764	>	epot = epot + sw*vterm
765	>
766	>	!! this has a + sign in the () because the rij vector is
767	>	!! r_j - r_i and the charge-dipole potential takes the origin
768	>	!! as the point dipole, which is atom j in this case.
769	>
770	>	dudx = dudx - sw*pref*( ri3*(uz_j(1) - 3.0d0*ct_j*xhat) - &
771	>	preRF2*uz_j(1) )
772	>	dudy = dudy - sw*pref*( ri3*(uz_j(2) - 3.0d0*ct_j*yhat) - &
773	>	preRF2*uz_j(2) )
774	>	dudz = dudz - sw*pref*( ri3*(uz_j(3) - 3.0d0*ct_j*zhat) - &
775	>	preRF2*uz_j(3) )
776	>	duduz_j(1) = duduz_j(1) - sw*pref * xhat * ( ri2 - preRF2*rij )
777	>	duduz_j(2) = duduz_j(2) - sw*pref * yhat * ( ri2 - preRF2*rij )
778	>	duduz_j(3) = duduz_j(3) - sw*pref * zhat * ( ri2 - preRF2*rij )
779	>
780		else
781		if (j_is_SplitDipole) then
782		BigR = sqrt(r2 + 0.25_dp * d_j * d_j)
783		ri = 1.0_dp / BigR
784	<	scale = rij * ri
785	<	else
784	>	scale = rij * ri
785	>	else
786		ri = riji
787		scale = 1.0_dp
788		endif
789	<	endif
789	>
790	>	ri2 = ri * ri
791	>	ri3 = ri2 * ri
792	>	sc2 = scale * scale
793
794	<	ct_ij = ul_i(1)*ul_j(1) + ul_i(2)*ul_j(2) + ul_i(3)*ul_j(3)
794	>	pref = pre12 * q_i * mu_j
795	>	vterm = - pref * ct_j * ri2 * scale
796	>	vpair = vpair + vterm
797	>	epot = epot + sw*vterm
798	>
799	>	!! this has a + sign in the () because the rij vector is
800	>	!! r_j - r_i and the charge-dipole potential takes the origin
801	>	!! as the point dipole, which is atom j in this case.
802	>
803	>	dudx = dudx - sw*pref * ri3 * ( uz_j(1) - 3.0d0*ct_j*xhat*sc2)
804	>	dudy = dudy - sw*pref * ri3 * ( uz_j(2) - 3.0d0*ct_j*yhat*sc2)
805	>	dudz = dudz - sw*pref * ri3 * ( uz_j(3) - 3.0d0*ct_j*zhat*sc2)
806	>
807	>	duduz_j(1) = duduz_j(1) - sw*pref * ri2 * xhat * scale
808	>	duduz_j(2) = duduz_j(2) - sw*pref * ri2 * yhat * scale
809	>	duduz_j(3) = duduz_j(3) - sw*pref * ri2 * zhat * scale
810
811	<	ri2 = ri * ri
812	<	ri3 = ri2 * ri
811	>	endif
812	>	endif
813	>
814	>	if (j_is_Quadrupole) then
815	>	ri2 = riji * riji
816	>	ri3 = ri2 * riji
817		ri4 = ri2 * ri2
818	<	sc2 = scale * scale
818	>	cx2 = cx_j * cx_j
819	>	cy2 = cy_j * cy_j
820	>	cz2 = cz_j * cz_j
821
822	<	pref = pre22 * mu_i * mu_j
823	<	vterm = pref * ri3 * (ct_ij - 3.0d0 * ct_i * ct_j * sc2)
824	<	vpair = vpair + vterm
825	<	epot = epot + sw * vterm
822	>	!!$          if (summationMethod .eq. UNDAMPED_WOLF) then
823	>	!!$             pref =  pre14 * q_i / 3.0_dp
824	>	!!$             vterm1 = pref * ri3*( qxx_j * (3.0_dp*cx2 - 1.0_dp) + &
825	>	!!$                  qyy_j * (3.0_dp*cy2 - 1.0_dp) + &
826	>	!!$                  qzz_j * (3.0_dp*cz2 - 1.0_dp) )
827	>	!!$             vterm2 = pref * rcuti3*( qxx_j * (3.0_dp*cx2 - 1.0_dp) + &
828	>	!!$                  qyy_j * (3.0_dp*cy2 - 1.0_dp) + &
829	>	!!$                  qzz_j * (3.0_dp*cz2 - 1.0_dp) )
830	>	!!$             vpair = vpair + ( vterm1 - vterm2 )
831	>	!!$             epot = epot + sw*( vterm1 - vterm2 )
832	>	!!$
833	>	!!$             dudx = dudx - (5.0_dp * &
834	>	!!$                  (vterm1*riji*xhat - vterm2*rcuti2*d(1))) + sw*pref * ( &
835	>	!!$                  (ri4 - rcuti4)*(qxx_j*(6.0_dp*cx_j*ux_j(1)) - &
836	>	!!$                  qxx_j*2.0_dp*(xhat - rcuti*d(1))) + &
837	>	!!$                  (ri4 - rcuti4)*(qyy_j*(6.0_dp*cy_j*uy_j(1)) - &
838	>	!!$                  qyy_j*2.0_dp*(xhat - rcuti*d(1))) + &
839	>	!!$                  (ri4 - rcuti4)*(qzz_j*(6.0_dp*cz_j*uz_j(1)) - &
840	>	!!$                  qzz_j*2.0_dp*(xhat - rcuti*d(1))) )
841	>	!!$             dudy = dudy - (5.0_dp * &
842	>	!!$                  (vterm1*riji*yhat - vterm2*rcuti2*d(2))) + sw*pref * ( &
843	>	!!$                  (ri4 - rcuti4)*(qxx_j*(6.0_dp*cx_j*ux_j(2)) - &
844	>	!!$                  qxx_j*2.0_dp*(yhat - rcuti*d(2))) + &
845	>	!!$                  (ri4 - rcuti4)*(qyy_j*(6.0_dp*cy_j*uy_j(2)) - &
846	>	!!$                  qyy_j*2.0_dp*(yhat - rcuti*d(2))) + &
847	>	!!$                  (ri4 - rcuti4)*(qzz_j*(6.0_dp*cz_j*uz_j(2)) - &
848	>	!!$                  qzz_j*2.0_dp*(yhat - rcuti*d(2))) )
849	>	!!$             dudz = dudz - (5.0_dp * &
850	>	!!$                  (vterm1*riji*zhat - vterm2*rcuti2*d(3))) + sw*pref * ( &
851	>	!!$                  (ri4 - rcuti4)*(qxx_j*(6.0_dp*cx_j*ux_j(3)) - &
852	>	!!$                  qxx_j*2.0_dp*(zhat - rcuti*d(3))) + &
853	>	!!$                  (ri4 - rcuti4)*(qyy_j*(6.0_dp*cy_j*uy_j(3)) - &
854	>	!!$                  qyy_j*2.0_dp*(zhat - rcuti*d(3))) + &
855	>	!!$                  (ri4 - rcuti4)*(qzz_j*(6.0_dp*cz_j*uz_j(3)) - &
856	>	!!$                  qzz_j*2.0_dp*(zhat - rcuti*d(3))) )
857	>	!!$
858	>	!!$             dudux_j(1) = dudux_j(1) + sw*pref*(ri3*(qxx_j*6.0_dp*cx_j*xhat) -&
859	>	!!$                  rcuti4*(qxx_j*6.0_dp*cx_j*d(1)))
860	>	!!$             dudux_j(2) = dudux_j(2) + sw*pref*(ri3*(qxx_j*6.0_dp*cx_j*yhat) -&
861	>	!!$                  rcuti4*(qxx_j*6.0_dp*cx_j*d(2)))
862	>	!!$             dudux_j(3) = dudux_j(3) + sw*pref*(ri3*(qxx_j*6.0_dp*cx_j*zhat) -&
863	>	!!$                  rcuti4*(qxx_j*6.0_dp*cx_j*d(3)))
864	>	!!$
865	>	!!$             duduy_j(1) = duduy_j(1) + sw*pref*(ri3*(qyy_j*6.0_dp*cy_j*xhat) -&
866	>	!!$                  rcuti4*(qyy_j*6.0_dp*cx_j*d(1)))
867	>	!!$             duduy_j(2) = duduy_j(2) + sw*pref*(ri3*(qyy_j*6.0_dp*cy_j*yhat) -&
868	>	!!$                  rcuti4*(qyy_j*6.0_dp*cx_j*d(2)))
869	>	!!$             duduy_j(3) = duduy_j(3) + sw*pref*(ri3*(qyy_j*6.0_dp*cy_j*zhat) -&
870	>	!!$                  rcuti4*(qyy_j*6.0_dp*cx_j*d(3)))
871	>	!!$
872	>	!!$             duduz_j(1) = duduz_j(1) + sw*pref*(ri3*(qzz_j*6.0_dp*cz_j*xhat) -&
873	>	!!$                  rcuti4*(qzz_j*6.0_dp*cx_j*d(1)))
874	>	!!$             duduz_j(2) = duduz_j(2) + sw*pref*(ri3*(qzz_j*6.0_dp*cz_j*yhat) -&
875	>	!!$                  rcuti4*(qzz_j*6.0_dp*cx_j*d(2)))
876	>	!!$             duduz_j(3) = duduz_j(3) + sw*pref*(ri3*(qzz_j*6.0_dp*cz_j*zhat) -&
877	>	!!$                  rcuti4*(qzz_j*6.0_dp*cx_j*d(3)))
878	>	!!$
879	>	!!$          else
880	>	pref =  pre14 * q_i / 3.0_dp
881	>	vterm = pref * ri3 * (qxx_j * (3.0_dp*cx2 - 1.0_dp) + &
882	>	qyy_j * (3.0_dp*cy2 - 1.0_dp) + &
883	>	qzz_j * (3.0_dp*cz2 - 1.0_dp))
884	>	vpair = vpair + vterm
885	>	epot = epot + sw*vterm
886	>
887	>	dudx = dudx - 5.0_dp*sw*vterm*riji*xhat + sw*pref * ri4 * ( &
888	>	qxx_j*(6.0_dp*cx_j*ux_j(1) - 2.0_dp*xhat) + &
889	>	qyy_j*(6.0_dp*cy_j*uy_j(1) - 2.0_dp*xhat) + &
890	>	qzz_j*(6.0_dp*cz_j*uz_j(1) - 2.0_dp*xhat) )
891	>	dudy = dudy - 5.0_dp*sw*vterm*riji*yhat + sw*pref * ri4 * ( &
892	>	qxx_j*(6.0_dp*cx_j*ux_j(2) - 2.0_dp*yhat) + &
893	>	qyy_j*(6.0_dp*cy_j*uy_j(2) - 2.0_dp*yhat) + &
894	>	qzz_j*(6.0_dp*cz_j*uz_j(2) - 2.0_dp*yhat) )
895	>	dudz = dudz - 5.0_dp*sw*vterm*riji*zhat + sw*pref * ri4 * ( &
896	>	qxx_j*(6.0_dp*cx_j*ux_j(3) - 2.0_dp*zhat) + &
897	>	qyy_j*(6.0_dp*cy_j*uy_j(3) - 2.0_dp*zhat) + &
898	>	qzz_j*(6.0_dp*cz_j*uz_j(3) - 2.0_dp*zhat) )
899	>
900	>	dudux_j(1) = dudux_j(1) + sw*pref * ri3*(qxx_j*6.0_dp*cx_j*xhat)
901	>	dudux_j(2) = dudux_j(2) + sw*pref * ri3*(qxx_j*6.0_dp*cx_j*yhat)
902	>	dudux_j(3) = dudux_j(3) + sw*pref * ri3*(qxx_j*6.0_dp*cx_j*zhat)
903	>
904	>	duduy_j(1) = duduy_j(1) + sw*pref * ri3*(qyy_j*6.0_dp*cy_j*xhat)
905	>	duduy_j(2) = duduy_j(2) + sw*pref * ri3*(qyy_j*6.0_dp*cy_j*yhat)
906	>	duduy_j(3) = duduy_j(3) + sw*pref * ri3*(qyy_j*6.0_dp*cy_j*zhat)
907	>
908	>	duduz_j(1) = duduz_j(1) + sw*pref * ri3*(qzz_j*6.0_dp*cz_j*xhat)
909	>	duduz_j(2) = duduz_j(2) + sw*pref * ri3*(qzz_j*6.0_dp*cz_j*yhat)
910	>	duduz_j(3) = duduz_j(3) + sw*pref * ri3*(qzz_j*6.0_dp*cz_j*zhat)
911
912	<	a1 = 5.0d0 * ct_i * ct_j * sc2 - ct_ij
912	>	!!$          endif
913	>	endif
914	>	endif
915
916	<	dudx=dudx+pref*sw*3.0d0*ri4*scale*(a1*xhat-ct_i*ul_j(1)-ct_j*ul_i(1))
552	<	dudy=dudy+pref*sw*3.0d0*ri4*scale*(a1*yhat-ct_i*ul_j(2)-ct_j*ul_i(2))
553	<	dudz=dudz+pref*sw*3.0d0*ri4*scale*(a1*zhat-ct_i*ul_j(3)-ct_j*ul_i(3))
916	>	if (i_is_Dipole) then
917
918	<	duduix = duduix + pref*sw*ri3*(ul_j(1) - 3.0d0*ct_j*xhat*sc2)
919	<	duduiy = duduiy + pref*sw*ri3*(ul_j(2) - 3.0d0*ct_j*yhat*sc2)
920	<	duduiz = duduiz + pref*sw*ri3*(ul_j(3) - 3.0d0*ct_j*zhat*sc2)
918	>	if (j_is_Charge) then
919	>
920	>	pref = pre12 * q_j * mu_i
921	>
922	>	!!$          if (summationMethod .eq. UNDAMPED_WOLF) then
923	>	!!$             ri2 = riji * riji
924	>	!!$             ri3 = ri2 * riji
925	>	!!$
926	>	!!$             pref = pre12 * q_j * mu_i
927	>	!!$             vterm = pref * ct_i * (ri2 - rcuti2)
928	>	!!$             vpair = vpair + vterm
929	>	!!$             epot = epot + sw*vterm
930	>	!!$
931	>	!!$             dudx = dudx + sw*pref * ( ri3*( uz_i(1) - 3.0d0*ct_i*xhat) &
932	>	!!$                  - rcuti3*( uz_i(1) - 3.0d0*ct_i*d(1)*rcuti ) )
933	>	!!$             dudy = dudy + sw*pref * ( ri3*( uz_i(2) - 3.0d0*ct_i*yhat) &
934	>	!!$                  - rcuti3*( uz_i(2) - 3.0d0*ct_i*d(2)*rcuti ) )
935	>	!!$             dudz = dudz + sw*pref * ( ri3*( uz_i(3) - 3.0d0*ct_i*zhat) &
936	>	!!$                  - rcuti3*( uz_i(3) - 3.0d0*ct_i*d(3)*rcuti ) )
937	>	!!$
938	>	!!$             duduz_i(1) = duduz_i(1) + sw*pref*( ri2*xhat - d(1)*rcuti3 )
939	>	!!$             duduz_i(2) = duduz_i(2) + sw*pref*( ri2*yhat - d(2)*rcuti3 )
940	>	!!$             duduz_i(3) = duduz_i(3) + sw*pref*( ri2*zhat - d(3)*rcuti3 )
941	>	!!$
942	>	!!$          elseif (summationMethod .eq. REACTION_FIELD) then
943	>	if (summationMethod .eq. REACTION_FIELD) then
944	>	ri2 = riji * riji
945	>	ri3 = ri2 * riji
946
947	<	dudujx = dudujx + pref*sw*ri3*(ul_i(1) - 3.0d0*ct_i*xhat*sc2)
948	<	dudujy = dudujy + pref*sw*ri3*(ul_i(2) - 3.0d0*ct_i*yhat*sc2)
949	<	dudujz = dudujz + pref*sw*ri3*(ul_i(3) - 3.0d0*ct_i*zhat*sc2)
947	>	pref = pre12 * q_j * mu_i
948	>	vterm = pref * ct_i * ( ri2 - preRF2*rij )
949	>	vpair = vpair + vterm
950	>	epot = epot + sw*vterm
951	>
952	>	dudx = dudx + sw*pref * ( ri3*(uz_i(1) - 3.0d0*ct_i*xhat) - &
953	>	preRF2*uz_i(1) )
954	>	dudy = dudy + sw*pref * ( ri3*(uz_i(2) - 3.0d0*ct_i*yhat) - &
955	>	preRF2*uz_i(2) )
956	>	dudz = dudz + sw*pref * ( ri3*(uz_i(3) - 3.0d0*ct_i*zhat) - &
957	>	preRF2*uz_i(3) )
958	>
959	>	duduz_i(1) = duduz_i(1) + sw*pref * xhat * ( ri2 - preRF2*rij )
960	>	duduz_i(2) = duduz_i(2) + sw*pref * yhat * ( ri2 - preRF2*rij )
961	>	duduz_i(3) = duduz_i(3) + sw*pref * zhat * ( ri2 - preRF2*rij )
962	>
963	>	else
964	>	if (i_is_SplitDipole) then
965	>	BigR = sqrt(r2 + 0.25_dp * d_i * d_i)
966	>	ri = 1.0_dp / BigR
967	>	scale = rij * ri
968	>	else
969	>	ri = riji
970	>	scale = 1.0_dp
971	>	endif
972	>
973	>	ri2 = ri * ri
974	>	ri3 = ri2 * ri
975	>	sc2 = scale * scale
976	>
977	>	pref = pre12 * q_j * mu_i
978	>	vterm = pref * ct_i * ri2 * scale
979	>	vpair = vpair + vterm
980	>	epot = epot + sw*vterm
981	>
982	>	dudx = dudx + sw*pref * ri3 * ( uz_i(1) - 3.0d0 * ct_i * xhat*sc2)
983	>	dudy = dudy + sw*pref * ri3 * ( uz_i(2) - 3.0d0 * ct_i * yhat*sc2)
984	>	dudz = dudz + sw*pref * ri3 * ( uz_i(3) - 3.0d0 * ct_i * zhat*sc2)
985	>
986	>	duduz_i(1) = duduz_i(1) + sw*pref * ri2 * xhat * scale
987	>	duduz_i(2) = duduz_i(2) + sw*pref * ri2 * yhat * scale
988	>	duduz_i(3) = duduz_i(3) + sw*pref * ri2 * zhat * scale
989	>	endif
990		endif
991	+
992	+	if (j_is_Dipole) then
993
994	+	!!$          if (summationMethod .eq. UNDAMPED_WOLF) then
995	+	!!$             ct_ij = uz_i(1)*uz_j(1) + uz_i(2)*uz_j(2) + uz_i(3)*uz_j(3)
996	+	!!$
997	+	!!$             ri2 = riji * riji
998	+	!!$             ri3 = ri2 * riji
999	+	!!$             ri4 = ri2 * ri2
1000	+	!!$
1001	+	!!$             pref = pre22 * mu_i * mu_j
1002	+	!!$             vterm = pref * (ri3 - rcuti3) * (ct_ij - 3.0d0 * ct_i * ct_j)
1003	+	!!$             vpair = vpair + vterm
1004	+	!!$             epot = epot + sw*vterm
1005	+	!!$
1006	+	!!$             a1 = 5.0d0 * ct_i * ct_j - ct_ij
1007	+	!!$
1008	+	!!$             dudx = dudx + sw*pref*3.0d0*( &
1009	+	!!$                  ri4*(a1*xhat-ct_i*uz_j(1)-ct_j*uz_i(1)) &
1010	+	!!$                  - rcuti4*(a1*xhat-ct_i*uz_j(1)-ct_j*uz_i(1)) )
1011	+	!!$             dudy = dudy + sw*pref*3.0d0*( &
1012	+	!!$                  ri4*(a1*yhat-ct_i*uz_j(2)-ct_j*uz_i(2)) &
1013	+	!!$                  - rcuti4*(a1*yhat-ct_i*uz_j(2)-ct_j*uz_i(2)) )
1014	+	!!$             dudz = dudz + sw*pref*3.0d0*( &
1015	+	!!$                  ri4*(a1*zhat-ct_i*uz_j(3)-ct_j*uz_i(3)) &
1016	+	!!$                  - rcuti4*(a1*zhat-ct_i*uz_j(3)-ct_j*uz_i(3)) )
1017	+	!!$
1018	+	!!$             duduz_i(1) = duduz_i(1) + sw*pref*(ri3*(uz_j(1)-3.0d0*ct_j*xhat) &
1019	+	!!$                  - rcuti3*(uz_j(1) - 3.0d0*ct_j*xhat))
1020	+	!!$             duduz_i(2) = duduz_i(2) + sw*pref*(ri3*(uz_j(2)-3.0d0*ct_j*yhat) &
1021	+	!!$                  - rcuti3*(uz_j(2) - 3.0d0*ct_j*yhat))
1022	+	!!$             duduz_i(3) = duduz_i(3) + sw*pref*(ri3*(uz_j(3)-3.0d0*ct_j*zhat) &
1023	+	!!$                  - rcuti3*(uz_j(3) - 3.0d0*ct_j*zhat))
1024	+	!!$             duduz_j(1) = duduz_j(1) + sw*pref*(ri3*(uz_i(1)-3.0d0*ct_i*xhat) &
1025	+	!!$                  - rcuti3*(uz_i(1) - 3.0d0*ct_i*xhat))
1026	+	!!$             duduz_j(2) = duduz_j(2) + sw*pref*(ri3*(uz_i(2)-3.0d0*ct_i*yhat) &
1027	+	!!$                  - rcuti3*(uz_i(2) - 3.0d0*ct_i*yhat))
1028	+	!!$             duduz_j(3) = duduz_j(3) + sw*pref*(ri3*(uz_i(3)-3.0d0*ct_i*zhat) &
1029	+	!!$                  - rcuti3*(uz_i(3) - 3.0d0*ct_i*zhat))
1030	+	!!$
1031	+	!!$          elseif (summationMethod .eq. DAMPED_WOLF) then
1032	+	!!$             ct_ij = uz_i(1)*uz_j(1) + uz_i(2)*uz_j(2) + uz_i(3)*uz_j(3)
1033	+	!!$
1034	+	!!$             ri2 = riji * riji
1035	+	!!$             ri3 = ri2 * riji
1036	+	!!$             ri4 = ri2 * ri2
1037	+	!!$             sc2 = scale * scale
1038	+	!!$
1039	+	!!$             pref = pre22 * mu_i * mu_j
1040	+	!!$             vterm = pref * ri3 * (ct_ij - 3.0d0 * ct_i * ct_j)
1041	+	!!$             vpair = vpair + vterm
1042	+	!!$             epot = epot + sw*vterm
1043	+	!!$
1044	+	!!$             a1 = 5.0d0 * ct_i * ct_j * sc2 - ct_ij
1045	+	!!$
1046	+	!!$             dudx = dudx + sw*pref*3.0d0*ri4*(a1*xhat-ct_i*uz_j(1)-ct_j*uz_i(1))
1047	+	!!$             dudy = dudy + sw*pref*3.0d0*ri4*(a1*yhat-ct_i*uz_j(2)-ct_j*uz_i(2))
1048	+	!!$             dudz = dudz + sw*pref*3.0d0*ri4*(a1*zhat-ct_i*uz_j(3)-ct_j*uz_i(3))
1049	+	!!$
1050	+	!!$             duduz_i(1) = duduz_i(1) + sw*pref*ri3 *(uz_j(1) - 3.0d0*ct_j*xhat)
1051	+	!!$             duduz_i(2) = duduz_i(2) + sw*pref*ri3 *(uz_j(2) - 3.0d0*ct_j*yhat)
1052	+	!!$             duduz_i(3) = duduz_i(3) + sw*pref*ri3 *(uz_j(3) - 3.0d0*ct_j*zhat)
1053	+	!!$
1054	+	!!$             duduz_j(1) = duduz_j(1) + sw*pref*ri3 *(uz_i(1) - 3.0d0*ct_i*xhat)
1055	+	!!$             duduz_j(2) = duduz_j(2) + sw*pref*ri3 *(uz_i(2) - 3.0d0*ct_i*yhat)
1056	+	!!$             duduz_j(3) = duduz_j(3) + sw*pref*ri3 *(uz_i(3) - 3.0d0*ct_i*zhat)
1057	+	!!$
1058	+	!!$          elseif (summationMethod .eq. REACTION_FIELD) then
1059	+	if (summationMethod .eq. REACTION_FIELD) then
1060	+	ct_ij = uz_i(1)*uz_j(1) + uz_i(2)*uz_j(2) + uz_i(3)*uz_j(3)
1061	+
1062	+	ri2 = riji * riji
1063	+	ri3 = ri2 * riji
1064	+	ri4 = ri2 * ri2
1065	+
1066	+	pref = pre22 * mu_i * mu_j
1067	+
1068	+	vterm = pref*( ri3*(ct_ij - 3.0d0 * ct_i * ct_j) - &
1069	+	preRF2*ct_ij )
1070	+	vpair = vpair + vterm
1071	+	epot = epot + sw*vterm
1072	+
1073	+	a1 = 5.0d0 * ct_i * ct_j - ct_ij
1074	+
1075	+	dudx = dudx + sw*pref*3.0d0*ri4 &
1076	+	* (a1*xhat-ct_i*uz_j(1)-ct_j*uz_i(1))
1077	+	dudy = dudy + sw*pref*3.0d0*ri4 &
1078	+	* (a1*yhat-ct_i*uz_j(2)-ct_j*uz_i(2))
1079	+	dudz = dudz + sw*pref*3.0d0*ri4 &
1080	+	* (a1*zhat-ct_i*uz_j(3)-ct_j*uz_i(3))
1081	+
1082	+	duduz_i(1) = duduz_i(1) + sw*pref*(ri3*(uz_j(1)-3.0d0*ct_j*xhat) &
1083	+	- preRF2*uz_j(1))
1084	+	duduz_i(2) = duduz_i(2) + sw*pref*(ri3*(uz_j(2)-3.0d0*ct_j*yhat) &
1085	+	- preRF2*uz_j(2))
1086	+	duduz_i(3) = duduz_i(3) + sw*pref*(ri3*(uz_j(3)-3.0d0*ct_j*zhat) &
1087	+	- preRF2*uz_j(3))
1088	+	duduz_j(1) = duduz_j(1) + sw*pref*(ri3*(uz_i(1)-3.0d0*ct_i*xhat) &
1089	+	- preRF2*uz_i(1))
1090	+	duduz_j(2) = duduz_j(2) + sw*pref*(ri3*(uz_i(2)-3.0d0*ct_i*yhat) &
1091	+	- preRF2*uz_i(2))
1092	+	duduz_j(3) = duduz_j(3) + sw*pref*(ri3*(uz_i(3)-3.0d0*ct_i*zhat) &
1093	+	- preRF2*uz_i(3))
1094	+
1095	+	else
1096	+	if (i_is_SplitDipole) then
1097	+	if (j_is_SplitDipole) then
1098	+	BigR = sqrt(r2 + 0.25_dp * d_i * d_i + 0.25_dp * d_j * d_j)
1099	+	else
1100	+	BigR = sqrt(r2 + 0.25_dp * d_i * d_i)
1101	+	endif
1102	+	ri = 1.0_dp / BigR
1103	+	scale = rij * ri
1104	+	else
1105	+	if (j_is_SplitDipole) then
1106	+	BigR = sqrt(r2 + 0.25_dp * d_j * d_j)
1107	+	ri = 1.0_dp / BigR
1108	+	scale = rij * ri
1109	+	else
1110	+	ri = riji
1111	+	scale = 1.0_dp
1112	+	endif
1113	+	endif
1114	+
1115	+	ct_ij = uz_i(1)*uz_j(1) + uz_i(2)*uz_j(2) + uz_i(3)*uz_j(3)
1116	+
1117	+	ri2 = ri * ri
1118	+	ri3 = ri2 * ri
1119	+	ri4 = ri2 * ri2
1120	+	sc2 = scale * scale
1121	+
1122	+	pref = pre22 * mu_i * mu_j
1123	+	vterm = pref * ri3 * (ct_ij - 3.0d0 * ct_i * ct_j * sc2)
1124	+	vpair = vpair + vterm
1125	+	epot = epot + sw*vterm
1126	+
1127	+	a1 = 5.0d0 * ct_i * ct_j * sc2 - ct_ij
1128	+
1129	+	dudx = dudx + sw*pref*3.0d0*ri4*scale &
1130	+	*(a1*xhat-ct_i*uz_j(1)-ct_j*uz_i(1))
1131	+	dudy = dudy + sw*pref*3.0d0*ri4*scale &
1132	+	*(a1*yhat-ct_i*uz_j(2)-ct_j*uz_i(2))
1133	+	dudz = dudz + sw*pref*3.0d0*ri4*scale &
1134	+	*(a1*zhat-ct_i*uz_j(3)-ct_j*uz_i(3))
1135	+
1136	+	duduz_i(1) = duduz_i(1) + sw*pref*ri3 &
1137	+	*(uz_j(1) - 3.0d0*ct_j*xhat*sc2)
1138	+	duduz_i(2) = duduz_i(2) + sw*pref*ri3 &
1139	+	*(uz_j(2) - 3.0d0*ct_j*yhat*sc2)
1140	+	duduz_i(3) = duduz_i(3) + sw*pref*ri3 &
1141	+	*(uz_j(3) - 3.0d0*ct_j*zhat*sc2)
1142	+
1143	+	duduz_j(1) = duduz_j(1) + sw*pref*ri3 &
1144	+	*(uz_i(1) - 3.0d0*ct_i*xhat*sc2)
1145	+	duduz_j(2) = duduz_j(2) + sw*pref*ri3 &
1146	+	*(uz_i(2) - 3.0d0*ct_i*yhat*sc2)
1147	+	duduz_j(3) = duduz_j(3) + sw*pref*ri3 &
1148	+	*(uz_i(3) - 3.0d0*ct_i*zhat*sc2)
1149	+	endif
1150	+	endif
1151		endif
1152	<
1152	>
1153	>	if (i_is_Quadrupole) then
1154	>	if (j_is_Charge) then
1155	>
1156	>	ri2 = riji * riji
1157	>	ri3 = ri2 * riji
1158	>	ri4 = ri2 * ri2
1159	>	cx2 = cx_i * cx_i
1160	>	cy2 = cy_i * cy_i
1161	>	cz2 = cz_i * cz_i
1162	>
1163	>	!!$          if (summationMethod .eq. UNDAMPED_WOLF) then
1164	>	!!$             pref = pre14 * q_j / 3.0_dp
1165	>	!!$             vterm1 = pref * ri3*( qxx_i * (3.0_dp*cx2 - 1.0_dp) + &
1166	>	!!$                  qyy_i * (3.0_dp*cy2 - 1.0_dp) + &
1167	>	!!$                  qzz_i * (3.0_dp*cz2 - 1.0_dp) )
1168	>	!!$             vterm2 = pref * rcuti3*( qxx_i * (3.0_dp*cx2 - 1.0_dp) + &
1169	>	!!$                  qyy_i * (3.0_dp*cy2 - 1.0_dp) + &
1170	>	!!$                  qzz_i * (3.0_dp*cz2 - 1.0_dp) )
1171	>	!!$             vpair = vpair + ( vterm1 - vterm2 )
1172	>	!!$             epot = epot + sw*( vterm1 - vterm2 )
1173	>	!!$
1174	>	!!$             dudx = dudx - sw*(5.0_dp*(vterm1*riji*xhat-vterm2*rcuti2*d(1))) +&
1175	>	!!$                  sw*pref * ( (ri4 - rcuti4)*(qxx_i*(6.0_dp*cx_i*ux_i(1)) - &
1176	>	!!$                  qxx_i*2.0_dp*(xhat - rcuti*d(1))) + &
1177	>	!!$                  (ri4 - rcuti4)*(qyy_i*(6.0_dp*cy_i*uy_i(1)) - &
1178	>	!!$                  qyy_i*2.0_dp*(xhat - rcuti*d(1))) + &
1179	>	!!$                  (ri4 - rcuti4)*(qzz_i*(6.0_dp*cz_i*uz_i(1)) - &
1180	>	!!$                  qzz_i*2.0_dp*(xhat - rcuti*d(1))) )
1181	>	!!$             dudy = dudy - sw*(5.0_dp*(vterm1*riji*yhat-vterm2*rcuti2*d(2))) +&
1182	>	!!$                  sw*pref * ( (ri4 - rcuti4)*(qxx_i*(6.0_dp*cx_i*ux_i(2)) - &
1183	>	!!$                  qxx_i*2.0_dp*(yhat - rcuti*d(2))) + &
1184	>	!!$                  (ri4 - rcuti4)*(qyy_i*(6.0_dp*cy_i*uy_i(2)) - &
1185	>	!!$                  qyy_i*2.0_dp*(yhat - rcuti*d(2))) + &
1186	>	!!$                  (ri4 - rcuti4)*(qzz_i*(6.0_dp*cz_i*uz_i(2)) - &
1187	>	!!$                  qzz_i*2.0_dp*(yhat - rcuti*d(2))) )
1188	>	!!$             dudz = dudz - sw*(5.0_dp*(vterm1*riji*zhat-vterm2*rcuti2*d(3))) +&
1189	>	!!$                  sw*pref * ( (ri4 - rcuti4)*(qxx_i*(6.0_dp*cx_i*ux_i(3)) - &
1190	>	!!$                  qxx_i*2.0_dp*(zhat - rcuti*d(3))) + &
1191	>	!!$                  (ri4 - rcuti4)*(qyy_i*(6.0_dp*cy_i*uy_i(3)) - &
1192	>	!!$                  qyy_i*2.0_dp*(zhat - rcuti*d(3))) + &
1193	>	!!$                  (ri4 - rcuti4)*(qzz_i*(6.0_dp*cz_i*uz_i(3)) - &
1194	>	!!$                  qzz_i*2.0_dp*(zhat - rcuti*d(3))) )
1195	>	!!$
1196	>	!!$             dudux_i(1) = dudux_i(1) + sw*pref*(ri3*(qxx_i*6.0_dp*cx_i*xhat) -&
1197	>	!!$                  rcuti4*(qxx_i*6.0_dp*cx_i*d(1)))
1198	>	!!$             dudux_i(2) = dudux_i(2) + sw*pref*(ri3*(qxx_i*6.0_dp*cx_i*yhat) -&
1199	>	!!$                  rcuti4*(qxx_i*6.0_dp*cx_i*d(2)))
1200	>	!!$             dudux_i(3) = dudux_i(3) + sw*pref*(ri3*(qxx_i*6.0_dp*cx_i*zhat) -&
1201	>	!!$                  rcuti4*(qxx_i*6.0_dp*cx_i*d(3)))
1202	>	!!$
1203	>	!!$             duduy_i(1) = duduy_i(1) + sw*pref*(ri3*(qyy_i*6.0_dp*cy_i*xhat) -&
1204	>	!!$                  rcuti4*(qyy_i*6.0_dp*cx_i*d(1)))
1205	>	!!$             duduy_i(2) = duduy_i(2) + sw*pref*(ri3*(qyy_i*6.0_dp*cy_i*yhat) -&
1206	>	!!$                  rcuti4*(qyy_i*6.0_dp*cx_i*d(2)))
1207	>	!!$             duduy_i(3) = duduy_i(3) + sw*pref*(ri3*(qyy_i*6.0_dp*cy_i*zhat) -&
1208	>	!!$                  rcuti4*(qyy_i*6.0_dp*cx_i*d(3)))
1209	>	!!$
1210	>	!!$             duduz_i(1) = duduz_i(1) + sw*pref*(ri3*(qzz_i*6.0_dp*cz_i*xhat) -&
1211	>	!!$                  rcuti4*(qzz_i*6.0_dp*cx_i*d(1)))
1212	>	!!$             duduz_i(2) = duduz_i(2) + sw*pref*(ri3*(qzz_i*6.0_dp*cz_i*yhat) -&
1213	>	!!$                  rcuti4*(qzz_i*6.0_dp*cx_i*d(2)))
1214	>	!!$             duduz_i(3) = duduz_i(3) + sw*pref*(ri3*(qzz_i*6.0_dp*cz_i*zhat) -&
1215	>	!!$                  rcuti4*(qzz_i*6.0_dp*cx_i*d(3)))
1216	>	!!$
1217	>	!!$          else
1218	>	pref = pre14 * q_j / 3.0_dp
1219	>	vterm = pref * ri3 * (qxx_i * (3.0_dp*cx2 - 1.0_dp) + &
1220	>	qyy_i * (3.0_dp*cy2 - 1.0_dp) + &
1221	>	qzz_i * (3.0_dp*cz2 - 1.0_dp))
1222	>	vpair = vpair + vterm
1223	>	epot = epot + sw*vterm
1224	>
1225	>	dudx = dudx - 5.0_dp*sw*vterm*riji*xhat + sw*pref*ri4 * ( &
1226	>	qxx_i*(6.0_dp*cx_i*ux_i(1) - 2.0_dp*xhat) + &
1227	>	qyy_i*(6.0_dp*cy_i*uy_i(1) - 2.0_dp*xhat) + &
1228	>	qzz_i*(6.0_dp*cz_i*uz_i(1) - 2.0_dp*xhat) )
1229	>	dudy = dudy - 5.0_dp*sw*vterm*riji*yhat + sw*pref*ri4 * ( &
1230	>	qxx_i*(6.0_dp*cx_i*ux_i(2) - 2.0_dp*yhat) + &
1231	>	qyy_i*(6.0_dp*cy_i*uy_i(2) - 2.0_dp*yhat) + &
1232	>	qzz_i*(6.0_dp*cz_i*uz_i(2) - 2.0_dp*yhat) )
1233	>	dudz = dudz - 5.0_dp*sw*vterm*riji*zhat + sw*pref*ri4 * ( &
1234	>	qxx_i*(6.0_dp*cx_i*ux_i(3) - 2.0_dp*zhat) + &
1235	>	qyy_i*(6.0_dp*cy_i*uy_i(3) - 2.0_dp*zhat) + &
1236	>	qzz_i*(6.0_dp*cz_i*uz_i(3) - 2.0_dp*zhat) )
1237	>
1238	>	dudux_i(1) = dudux_i(1) + sw*pref*ri3*(qxx_i*6.0_dp*cx_i*xhat)
1239	>	dudux_i(2) = dudux_i(2) + sw*pref*ri3*(qxx_i*6.0_dp*cx_i*yhat)
1240	>	dudux_i(3) = dudux_i(3) + sw*pref*ri3*(qxx_i*6.0_dp*cx_i*zhat)
1241	>
1242	>	duduy_i(1) = duduy_i(1) + sw*pref*ri3*(qyy_i*6.0_dp*cy_i*xhat)
1243	>	duduy_i(2) = duduy_i(2) + sw*pref*ri3*(qyy_i*6.0_dp*cy_i*yhat)
1244	>	duduy_i(3) = duduy_i(3) + sw*pref*ri3*(qyy_i*6.0_dp*cy_i*zhat)
1245	>
1246	>	duduz_i(1) = duduz_i(1) + sw*pref*ri3*(qzz_i*6.0_dp*cz_i*xhat)
1247	>	duduz_i(2) = duduz_i(2) + sw*pref*ri3*(qzz_i*6.0_dp*cz_i*yhat)
1248	>	duduz_i(3) = duduz_i(3) + sw*pref*ri3*(qzz_i*6.0_dp*cz_i*zhat)
1249	>	!!$          endif
1250	>	endif
1251	>	endif
1252	>
1253	>
1254		if (do_pot) then
1255		#ifdef IS_MPI
1256	<	pot_row(atom1) = pot_row(atom1) + 0.5d0*epot
1257	<	pot_col(atom2) = pot_col(atom2) + 0.5d0*epot
1256	>	pot_row(ELECTROSTATIC_POT,atom1) = pot_row(ELECTROSTATIC_POT,atom1) + 0.5d0*epot
1257	>	pot_col(ELECTROSTATIC_POT,atom2) = pot_col(ELECTROSTATIC_POT,atom2) + 0.5d0*epot
1258		#else
1259		pot = pot + epot
1260		#endif
1261		endif
1262	<
1262	>
1263		#ifdef IS_MPI
1264		f_Row(1,atom1) = f_Row(1,atom1) + dudx
1265		f_Row(2,atom1) = f_Row(2,atom1) + dudy
1266		f_Row(3,atom1) = f_Row(3,atom1) + dudz
1267	<
1267	>
1268		f_Col(1,atom2) = f_Col(1,atom2) - dudx
1269		f_Col(2,atom2) = f_Col(2,atom2) - dudy
1270		f_Col(3,atom2) = f_Col(3,atom2) - dudz
1271	<
1271	>
1272		if (i_is_Dipole .or. i_is_Quadrupole) then
1273	<	t_Row(1,atom1) = t_Row(1,atom1) - ul_i(2)*duduiz + ul_i(3)*duduiy
1274	<	t_Row(2,atom1) = t_Row(2,atom1) - ul_i(3)*duduix + ul_i(1)*duduiz
1275	<	t_Row(3,atom1) = t_Row(3,atom1) - ul_i(1)*duduiy + ul_i(2)*duduix
1273	>	t_Row(1,atom1)=t_Row(1,atom1) - uz_i(2)*duduz_i(3) + uz_i(3)*duduz_i(2)
1274	>	t_Row(2,atom1)=t_Row(2,atom1) - uz_i(3)*duduz_i(1) + uz_i(1)*duduz_i(3)
1275	>	t_Row(3,atom1)=t_Row(3,atom1) - uz_i(1)*duduz_i(2) + uz_i(2)*duduz_i(1)
1276		endif
1277	+	if (i_is_Quadrupole) then
1278	+	t_Row(1,atom1)=t_Row(1,atom1) - ux_i(2)*dudux_i(3) + ux_i(3)*dudux_i(2)
1279	+	t_Row(2,atom1)=t_Row(2,atom1) - ux_i(3)*dudux_i(1) + ux_i(1)*dudux_i(3)
1280	+	t_Row(3,atom1)=t_Row(3,atom1) - ux_i(1)*dudux_i(2) + ux_i(2)*dudux_i(1)
1281
1282	+	t_Row(1,atom1)=t_Row(1,atom1) - uy_i(2)*duduy_i(3) + uy_i(3)*duduy_i(2)
1283	+	t_Row(2,atom1)=t_Row(2,atom1) - uy_i(3)*duduy_i(1) + uy_i(1)*duduy_i(3)
1284	+	t_Row(3,atom1)=t_Row(3,atom1) - uy_i(1)*duduy_i(2) + uy_i(2)*duduy_i(1)
1285	+	endif
1286	+
1287		if (j_is_Dipole .or. j_is_Quadrupole) then
1288	<	t_Col(1,atom2) = t_Col(1,atom2) - ul_j(2)*dudujz + ul_j(3)*dudujy
1289	<	t_Col(2,atom2) = t_Col(2,atom2) - ul_j(3)*dudujx + ul_j(1)*dudujz
1290	<	t_Col(3,atom2) = t_Col(3,atom2) - ul_j(1)*dudujy + ul_j(2)*dudujx
1288	>	t_Col(1,atom2)=t_Col(1,atom2) - uz_j(2)*duduz_j(3) + uz_j(3)*duduz_j(2)
1289	>	t_Col(2,atom2)=t_Col(2,atom2) - uz_j(3)*duduz_j(1) + uz_j(1)*duduz_j(3)
1290	>	t_Col(3,atom2)=t_Col(3,atom2) - uz_j(1)*duduz_j(2) + uz_j(2)*duduz_j(1)
1291		endif
1292	+	if (j_is_Quadrupole) then
1293	+	t_Col(1,atom2)=t_Col(1,atom2) - ux_j(2)*dudux_j(3) + ux_j(3)*dudux_j(2)
1294	+	t_Col(2,atom2)=t_Col(2,atom2) - ux_j(3)*dudux_j(1) + ux_j(1)*dudux_j(3)
1295	+	t_Col(3,atom2)=t_Col(3,atom2) - ux_j(1)*dudux_j(2) + ux_j(2)*dudux_j(1)
1296
1297	+	t_Col(1,atom2)=t_Col(1,atom2) - uy_j(2)*duduy_j(3) + uy_j(3)*duduy_j(2)
1298	+	t_Col(2,atom2)=t_Col(2,atom2) - uy_j(3)*duduy_j(1) + uy_j(1)*duduy_j(3)
1299	+	t_Col(3,atom2)=t_Col(3,atom2) - uy_j(1)*duduy_j(2) + uy_j(2)*duduy_j(1)
1300	+	endif
1301	+
1302		#else
1303		f(1,atom1) = f(1,atom1) + dudx
1304		f(2,atom1) = f(2,atom1) + dudy
1305		f(3,atom1) = f(3,atom1) + dudz
1306	<
1306	>
1307		f(1,atom2) = f(1,atom2) - dudx
1308		f(2,atom2) = f(2,atom2) - dudy
1309		f(3,atom2) = f(3,atom2) - dudz
1310	<
1310	>
1311		if (i_is_Dipole .or. i_is_Quadrupole) then
1312	<	t(1,atom1) = t(1,atom1) - ul_i(2)*duduiz + ul_i(3)*duduiy
1313	<	t(2,atom1) = t(2,atom1) - ul_i(3)*duduix + ul_i(1)*duduiz
1314	<	t(3,atom1) = t(3,atom1) - ul_i(1)*duduiy + ul_i(2)*duduix
1312	>	t(1,atom1)=t(1,atom1) - uz_i(2)*duduz_i(3) + uz_i(3)*duduz_i(2)
1313	>	t(2,atom1)=t(2,atom1) - uz_i(3)*duduz_i(1) + uz_i(1)*duduz_i(3)
1314	>	t(3,atom1)=t(3,atom1) - uz_i(1)*duduz_i(2) + uz_i(2)*duduz_i(1)
1315		endif
1316	<
1316	>	if (i_is_Quadrupole) then
1317	>	t(1,atom1)=t(1,atom1) - ux_i(2)*dudux_i(3) + ux_i(3)*dudux_i(2)
1318	>	t(2,atom1)=t(2,atom1) - ux_i(3)*dudux_i(1) + ux_i(1)*dudux_i(3)
1319	>	t(3,atom1)=t(3,atom1) - ux_i(1)*dudux_i(2) + ux_i(2)*dudux_i(1)
1320	>
1321	>	t(1,atom1)=t(1,atom1) - uy_i(2)*duduy_i(3) + uy_i(3)*duduy_i(2)
1322	>	t(2,atom1)=t(2,atom1) - uy_i(3)*duduy_i(1) + uy_i(1)*duduy_i(3)
1323	>	t(3,atom1)=t(3,atom1) - uy_i(1)*duduy_i(2) + uy_i(2)*duduy_i(1)
1324	>	endif
1325	>
1326		if (j_is_Dipole .or. j_is_Quadrupole) then
1327	<	t(1,atom2) = t(1,atom2) - ul_j(2)*dudujz + ul_j(3)*dudujy
1328	<	t(2,atom2) = t(2,atom2) - ul_j(3)*dudujx + ul_j(1)*dudujz
1329	<	t(3,atom2) = t(3,atom2) - ul_j(1)*dudujy + ul_j(2)*dudujx
1327	>	t(1,atom2)=t(1,atom2) - uz_j(2)*duduz_j(3) + uz_j(3)*duduz_j(2)
1328	>	t(2,atom2)=t(2,atom2) - uz_j(3)*duduz_j(1) + uz_j(1)*duduz_j(3)
1329	>	t(3,atom2)=t(3,atom2) - uz_j(1)*duduz_j(2) + uz_j(2)*duduz_j(1)
1330		endif
1331	+	if (j_is_Quadrupole) then
1332	+	t(1,atom2)=t(1,atom2) - ux_j(2)*dudux_j(3) + ux_j(3)*dudux_j(2)
1333	+	t(2,atom2)=t(2,atom2) - ux_j(3)*dudux_j(1) + ux_j(1)*dudux_j(3)
1334	+	t(3,atom2)=t(3,atom2) - ux_j(1)*dudux_j(2) + ux_j(2)*dudux_j(1)
1335	+
1336	+	t(1,atom2)=t(1,atom2) - uy_j(2)*duduy_j(3) + uy_j(3)*duduy_j(2)
1337	+	t(2,atom2)=t(2,atom2) - uy_j(3)*duduy_j(1) + uy_j(1)*duduy_j(3)
1338	+	t(3,atom2)=t(3,atom2) - uy_j(1)*duduy_j(2) + uy_j(2)*duduy_j(1)
1339	+	endif
1340	+
1341		#endif
1342	<
1342	>
1343		#ifdef IS_MPI
1344		id1 = AtomRowToGlobal(atom1)
1345		id2 = AtomColToGlobal(atom2)
#	Line 624 | Line 1349 | contains
1349		#endif
1350
1351		if (molMembershipList(id1) .ne. molMembershipList(id2)) then
1352	<
1352	>
1353		fpair(1) = fpair(1) + dudx
1354		fpair(2) = fpair(2) + dudy
1355		fpair(3) = fpair(3) + dudz
#	Line 633 | Line 1358 | contains
1358
1359		return
1360		end subroutine doElectrostaticPair
1361	<
1361	>
1362	>	subroutine destroyElectrostaticTypes()
1363	>
1364	>	if(allocated(ElectrostaticMap)) deallocate(ElectrostaticMap)
1365	>
1366	>	end subroutine destroyElectrostaticTypes
1367	>
1368	>	subroutine self_self(atom1, eFrame, mypot, t, do_pot)
1369	>	logical, intent(in) :: do_pot
1370	>	integer, intent(in) :: atom1
1371	>	integer :: atid1
1372	>	real(kind=dp), dimension(9,nLocal) :: eFrame
1373	>	real(kind=dp), dimension(3,nLocal) :: t
1374	>	real(kind=dp) :: mu1, c1
1375	>	real(kind=dp) :: preVal, epot, mypot
1376	>	real(kind=dp) :: eix, eiy, eiz
1377	>
1378	>	! this is a local only array, so we use the local atom type id's:
1379	>	atid1 = atid(atom1)
1380	>
1381	>	if (.not.summationMethodChecked) then
1382	>	call checkSummationMethod()
1383	>	endif
1384	>
1385	>	if (summationMethod .eq. REACTION_FIELD) then
1386	>	if (ElectrostaticMap(atid1)%is_Dipole) then
1387	>	mu1 = getDipoleMoment(atid1)
1388	>
1389	>	preVal = pre22 * preRF2 * mu1*mu1
1390	>	mypot = mypot - 0.5d0*preVal
1391	>
1392	>	! The self-correction term adds into the reaction field vector
1393	>
1394	>	eix = preVal * eFrame(3,atom1)
1395	>	eiy = preVal * eFrame(6,atom1)
1396	>	eiz = preVal * eFrame(9,atom1)
1397	>
1398	>	! once again, this is self-self, so only the local arrays are needed
1399	>	! even for MPI jobs:
1400	>
1401	>	t(1,atom1)=t(1,atom1) - eFrame(6,atom1)*eiz + &
1402	>	eFrame(9,atom1)*eiy
1403	>	t(2,atom1)=t(2,atom1) - eFrame(9,atom1)*eix + &
1404	>	eFrame(3,atom1)*eiz
1405	>	t(3,atom1)=t(3,atom1) - eFrame(3,atom1)*eiy + &
1406	>	eFrame(6,atom1)*eix
1407	>
1408	>	endif
1409	>
1410	>	!!$    elseif (summationMethod .eq. UNDAMPED_WOLF) then
1411	>	!!$       if (ElectrostaticMap(atid1)%is_Charge) then
1412	>	!!$          c1 = getCharge(atid1)
1413	>	!!$
1414	>	!!$          mypot = mypot - (rcuti * 0.5_dp * c1 * c1)
1415	>	!!$       endif
1416	>	!!$
1417	>	!!$    elseif (summationMethod .eq. DAMPED_WOLF) then
1418	>	!!$       if (ElectrostaticMap(atid1)%is_Charge) then
1419	>	!!$          c1 = getCharge(atid1)
1420	>	!!$
1421	>	!!$          mypot = mypot - (constERFC * rcuti * 0.5_dp + &
1422	>	!!$               dampingAlpha*invRootPi) * c1 * c1
1423	>	!!$       endif
1424	>	endif
1425	>
1426	>	return
1427	>	end subroutine self_self
1428	>
1429	>	subroutine rf_self_excludes(atom1, atom2, sw, eFrame, d, rij, vpair, myPot, &
1430	>	f, t, do_pot)
1431	>	logical, intent(in) :: do_pot
1432	>	integer, intent(in) :: atom1
1433	>	integer, intent(in) :: atom2
1434	>	logical :: i_is_Charge, j_is_Charge
1435	>	logical :: i_is_Dipole, j_is_Dipole
1436	>	integer :: atid1
1437	>	integer :: atid2
1438	>	real(kind=dp), intent(in) :: rij
1439	>	real(kind=dp), intent(in) :: sw
1440	>	real(kind=dp), intent(in), dimension(3) :: d
1441	>	real(kind=dp), intent(inout) :: vpair
1442	>	real(kind=dp), dimension(9,nLocal) :: eFrame
1443	>	real(kind=dp), dimension(3,nLocal) :: f
1444	>	real(kind=dp), dimension(3,nLocal) :: t
1445	>	real (kind = dp), dimension(3) :: duduz_i
1446	>	real (kind = dp), dimension(3) :: duduz_j
1447	>	real (kind = dp), dimension(3) :: uz_i
1448	>	real (kind = dp), dimension(3) :: uz_j
1449	>	real(kind=dp) :: q_i, q_j, mu_i, mu_j
1450	>	real(kind=dp) :: xhat, yhat, zhat
1451	>	real(kind=dp) :: ct_i, ct_j
1452	>	real(kind=dp) :: ri2, ri3, riji, vterm
1453	>	real(kind=dp) :: pref, preVal, rfVal, myPot
1454	>	real(kind=dp) :: dudx, dudy, dudz, dudr
1455	>
1456	>	if (.not.summationMethodChecked) then
1457	>	call checkSummationMethod()
1458	>	endif
1459	>
1460	>	dudx = 0.0d0
1461	>	dudy = 0.0d0
1462	>	dudz = 0.0d0
1463	>
1464	>	riji = 1.0d0/rij
1465	>
1466	>	xhat = d(1) * riji
1467	>	yhat = d(2) * riji
1468	>	zhat = d(3) * riji
1469	>
1470	>	! this is a local only array, so we use the local atom type id's:
1471	>	atid1 = atid(atom1)
1472	>	atid2 = atid(atom2)
1473	>	i_is_Charge = ElectrostaticMap(atid1)%is_Charge
1474	>	j_is_Charge = ElectrostaticMap(atid2)%is_Charge
1475	>	i_is_Dipole = ElectrostaticMap(atid1)%is_Dipole
1476	>	j_is_Dipole = ElectrostaticMap(atid2)%is_Dipole
1477	>
1478	>	if (i_is_Charge.and.j_is_Charge) then
1479	>	q_i = ElectrostaticMap(atid1)%charge
1480	>	q_j = ElectrostaticMap(atid2)%charge
1481	>
1482	>	preVal = pre11 * q_i * q_j
1483	>	rfVal = preRF*rij*rij
1484	>	vterm = preVal * rfVal
1485	>
1486	>	myPot = myPot + sw*vterm
1487	>
1488	>	dudr  = sw*preVal * 2.0d0*rfVal*riji
1489	>
1490	>	dudx = dudx + dudr * xhat
1491	>	dudy = dudy + dudr * yhat
1492	>	dudz = dudz + dudr * zhat
1493	>
1494	>	elseif (i_is_Charge.and.j_is_Dipole) then
1495	>	q_i = ElectrostaticMap(atid1)%charge
1496	>	mu_j = ElectrostaticMap(atid2)%dipole_moment
1497	>	uz_j(1) = eFrame(3,atom2)
1498	>	uz_j(2) = eFrame(6,atom2)
1499	>	uz_j(3) = eFrame(9,atom2)
1500	>	ct_j = uz_j(1)*xhat + uz_j(2)*yhat + uz_j(3)*zhat
1501	>
1502	>	ri2 = riji * riji
1503	>	ri3 = ri2 * riji
1504	>
1505	>	pref = pre12 * q_i * mu_j
1506	>	vterm = - pref * ct_j * ( ri2 - preRF2*rij )
1507	>	myPot = myPot + sw*vterm
1508	>
1509	>	dudx = dudx - sw*pref*( ri3*(uz_j(1)-3.0d0*ct_j*xhat) &
1510	>	- preRF2*uz_j(1) )
1511	>	dudy = dudy - sw*pref*( ri3*(uz_j(2)-3.0d0*ct_j*yhat) &
1512	>	- preRF2*uz_j(2) )
1513	>	dudz = dudz - sw*pref*( ri3*(uz_j(3)-3.0d0*ct_j*zhat) &
1514	>	- preRF2*uz_j(3) )
1515	>
1516	>	duduz_j(1) = duduz_j(1) - sw * pref * xhat * ( ri2 - preRF2*rij )
1517	>	duduz_j(2) = duduz_j(2) - sw * pref * yhat * ( ri2 - preRF2*rij )
1518	>	duduz_j(3) = duduz_j(3) - sw * pref * zhat * ( ri2 - preRF2*rij )
1519	>
1520	>	elseif (i_is_Dipole.and.j_is_Charge) then
1521	>	mu_i = ElectrostaticMap(atid1)%dipole_moment
1522	>	q_j = ElectrostaticMap(atid2)%charge
1523	>	uz_i(1) = eFrame(3,atom1)
1524	>	uz_i(2) = eFrame(6,atom1)
1525	>	uz_i(3) = eFrame(9,atom1)
1526	>	ct_i = uz_i(1)*xhat + uz_i(2)*yhat + uz_i(3)*zhat
1527	>
1528	>	ri2 = riji * riji
1529	>	ri3 = ri2 * riji
1530	>
1531	>	pref = pre12 * q_j * mu_i
1532	>	vterm = pref * ct_i * ( ri2 - preRF2*rij )
1533	>	myPot = myPot + sw*vterm
1534	>
1535	>	dudx = dudx + sw*pref*( ri3*(uz_i(1)-3.0d0*ct_i*xhat) &
1536	>	- preRF2*uz_i(1) )
1537	>	dudy = dudy + sw*pref*( ri3*(uz_i(2)-3.0d0*ct_i*yhat) &
1538	>	- preRF2*uz_i(2) )
1539	>	dudz = dudz + sw*pref*( ri3*(uz_i(3)-3.0d0*ct_i*zhat) &
1540	>	- preRF2*uz_i(3) )
1541	>
1542	>	duduz_i(1) = duduz_i(1) + sw * pref * xhat * ( ri2 - preRF2*rij )
1543	>	duduz_i(2) = duduz_i(2) + sw * pref * yhat * ( ri2 - preRF2*rij )
1544	>	duduz_i(3) = duduz_i(3) + sw * pref * zhat * ( ri2 - preRF2*rij )
1545	>
1546	>	endif
1547	>
1548	>
1549	>	! accumulate the forces and torques resulting from the self term
1550	>	f(1,atom1) = f(1,atom1) + dudx
1551	>	f(2,atom1) = f(2,atom1) + dudy
1552	>	f(3,atom1) = f(3,atom1) + dudz
1553	>
1554	>	f(1,atom2) = f(1,atom2) - dudx
1555	>	f(2,atom2) = f(2,atom2) - dudy
1556	>	f(3,atom2) = f(3,atom2) - dudz
1557	>
1558	>	if (i_is_Dipole) then
1559	>	t(1,atom1)=t(1,atom1) - uz_i(2)*duduz_i(3) + uz_i(3)*duduz_i(2)
1560	>	t(2,atom1)=t(2,atom1) - uz_i(3)*duduz_i(1) + uz_i(1)*duduz_i(3)
1561	>	t(3,atom1)=t(3,atom1) - uz_i(1)*duduz_i(2) + uz_i(2)*duduz_i(1)
1562	>	elseif (j_is_Dipole) then
1563	>	t(1,atom2)=t(1,atom2) - uz_j(2)*duduz_j(3) + uz_j(3)*duduz_j(2)
1564	>	t(2,atom2)=t(2,atom2) - uz_j(3)*duduz_j(1) + uz_j(1)*duduz_j(3)
1565	>	t(3,atom2)=t(3,atom2) - uz_j(1)*duduz_j(2) + uz_j(2)*duduz_j(1)
1566	>	endif
1567	>
1568	>	return
1569	>	end subroutine rf_self_excludes
1570	>
1571		end module electrostatic_module

Diff Legend

–	Removed lines
+	Added lines
<	Changed lines
>	Changed lines