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 2162 by chrisfen, Mon Apr 11 20:19:22 2005 UTC vs.
Revision 2755 by chrisfen, Wed May 17 14:03:15 2006 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
47		use vector_class
48		use simulation
49		use status
50	+	use interpolation
51		#ifdef IS_MPI
52		use mpiSimulation
53		#endif
#	Line 54 | Line 55 | module electrostatic_module
55
56		PRIVATE
57
58	+
59	+	#define __FORTRAN90
60	+	#include "UseTheForce/DarkSide/fInteractionMap.h"
61	+	#include "UseTheForce/DarkSide/fElectrostaticSummationMethod.h"
62	+	#include "UseTheForce/DarkSide/fElectrostaticScreeningMethod.h"
63	+
64	+
65		!! these prefactors convert the multipole interactions into kcal / mol
66		!! all were computed assuming distances are measured in angstroms
67		!! Charge-Charge, assuming charges are measured in electrons
#	Line 68 | Line 76 | module electrostatic_module
76		!! This unit is also known affectionately as an esu centi-barn.
77		real(kind=dp), parameter :: pre14 = 69.13373_dp
78
79	+	real(kind=dp), parameter :: zero = 0.0_dp
80	+
81	+	!! number of points for electrostatic splines
82	+	integer, parameter :: np = 100
83	+
84	+	!! variables to handle different summation methods for long-range
85	+	!! electrostatics:
86	+	integer, save :: summationMethod = NONE
87	+	integer, save :: screeningMethod = UNDAMPED
88	+	logical, save :: summationMethodChecked = .false.
89	+	real(kind=DP), save :: defaultCutoff = 0.0_DP
90	+	real(kind=DP), save :: defaultCutoff2 = 0.0_DP
91	+	logical, save :: haveDefaultCutoff = .false.
92	+	real(kind=DP), save :: dampingAlpha = 0.0_DP
93	+	real(kind=DP), save :: alpha2 = 0.0_DP
94	+	logical, save :: haveDampingAlpha = .false.
95	+	real(kind=DP), save :: dielectric = 1.0_DP
96	+	logical, save :: haveDielectric = .false.
97	+	real(kind=DP), save :: constEXP = 0.0_DP
98	+	real(kind=dp), save :: rcuti = 0.0_DP
99	+	real(kind=dp), save :: rcuti2 = 0.0_DP
100	+	real(kind=dp), save :: rcuti3 = 0.0_DP
101	+	real(kind=dp), save :: rcuti4 = 0.0_DP
102	+	real(kind=dp), save :: alphaPi = 0.0_DP
103	+	real(kind=dp), save :: invRootPi = 0.0_DP
104	+	real(kind=dp), save :: rrf = 1.0_DP
105	+	real(kind=dp), save :: rt = 1.0_DP
106	+	real(kind=dp), save :: rrfsq = 1.0_DP
107	+	real(kind=dp), save :: preRF = 0.0_DP
108	+	real(kind=dp), save :: preRF2 = 0.0_DP
109	+	real(kind=dp), save :: f0 = 1.0_DP
110	+	real(kind=dp), save :: f1 = 1.0_DP
111	+	real(kind=dp), save :: f2 = 0.0_DP
112	+	real(kind=dp), save :: f3 = 0.0_DP
113	+	real(kind=dp), save :: f4 = 0.0_DP
114	+	real(kind=dp), save :: f0c = 1.0_DP
115	+	real(kind=dp), save :: f1c = 1.0_DP
116	+	real(kind=dp), save :: f2c = 0.0_DP
117	+	real(kind=dp), save :: f3c = 0.0_DP
118	+	real(kind=dp), save :: f4c = 0.0_DP
119	+	real(kind=dp), save :: df0 = 0.0_DP
120	+	type(cubicSpline), save :: f0spline
121	+	logical, save :: haveElectroSpline = .false.
122	+	real(kind=dp), save :: one_third = 1.0_DP / 3.0_DP
123	+
124	+	#if defined(__IFC) || defined(__PGI)
125	+	! error function for ifc version > 7.
126	+	double precision, external :: derfc
127	+	#endif
128	+
129	+	public :: setElectrostaticSummationMethod
130	+	public :: setScreeningMethod
131	+	public :: setElectrostaticCutoffRadius
132	+	public :: setDampingAlpha
133	+	public :: setReactionFieldDielectric
134	+	public :: buildElectroSpline
135		public :: newElectrostaticType
136		public :: setCharge
137		public :: setDipoleMoment
#	Line 76 | Line 140 | module electrostatic_module
140		public :: doElectrostaticPair
141		public :: getCharge
142		public :: getDipoleMoment
143	<	public :: pre22
143	>	public :: destroyElectrostaticTypes
144	>	public :: self_self
145	>	public :: rf_self_excludes
146
147	+
148		type :: Electrostatic
149		integer :: c_ident
150		logical :: is_Charge = .false.
151		logical :: is_Dipole = .false.
152		logical :: is_SplitDipole = .false.
153		logical :: is_Quadrupole = .false.
154	+	logical :: is_Tap = .false.
155		real(kind=DP) :: charge = 0.0_DP
156		real(kind=DP) :: dipole_moment = 0.0_DP
157		real(kind=DP) :: split_dipole_distance = 0.0_DP
#	Line 92 | Line 160 | contains
160
161		type(Electrostatic), dimension(:), allocatable :: ElectrostaticMap
162
163	+	logical, save :: hasElectrostaticMap
164	+
165		contains
166
167	<	subroutine newElectrostaticType(c_ident, is_Charge, is_Dipole, &
168	<	is_SplitDipole, is_Quadrupole, status)
167	>	subroutine setElectrostaticSummationMethod(the_ESM)
168	>	integer, intent(in) :: the_ESM
169	>
170	>	if ((the_ESM .le. 0) .or. (the_ESM .gt. REACTION_FIELD)) then
171	>	call handleError("setElectrostaticSummationMethod", "Unsupported Summation Method")
172	>	endif
173	>
174	>	summationMethod = the_ESM
175	>
176	>	end subroutine setElectrostaticSummationMethod
177	>
178	>	subroutine setScreeningMethod(the_SM)
179	>	integer, intent(in) :: the_SM
180	>	screeningMethod = the_SM
181	>	end subroutine setScreeningMethod
182	>
183	>	subroutine setElectrostaticCutoffRadius(thisRcut, thisRsw)
184	>	real(kind=dp), intent(in) :: thisRcut
185	>	real(kind=dp), intent(in) :: thisRsw
186	>	defaultCutoff = thisRcut
187	>	defaultCutoff2 = defaultCutoff*defaultCutoff
188	>	rrf = defaultCutoff
189	>	rt = thisRsw
190	>	haveDefaultCutoff = .true.
191	>	end subroutine setElectrostaticCutoffRadius
192	>
193	>	subroutine setDampingAlpha(thisAlpha)
194	>	real(kind=dp), intent(in) :: thisAlpha
195	>	dampingAlpha = thisAlpha
196	>	alpha2 = dampingAlpha*dampingAlpha
197	>	haveDampingAlpha = .true.
198	>	end subroutine setDampingAlpha
199	>
200	>	subroutine setReactionFieldDielectric(thisDielectric)
201	>	real(kind=dp), intent(in) :: thisDielectric
202	>	dielectric = thisDielectric
203	>	haveDielectric = .true.
204	>	end subroutine setReactionFieldDielectric
205	>
206	>	subroutine buildElectroSpline()
207	>	real( kind = dp ), dimension(np) :: xvals, yvals
208	>	real( kind = dp ) :: dx, rmin, rval
209	>	integer :: i
210	>
211	>	rmin = 0.0_dp
212	>
213	>	dx = (defaultCutoff-rmin) / dble(np-1)
214
215	+	do i = 1, np
216	+	rval = rmin + dble(i-1)*dx
217	+	xvals(i) = rval
218	+	yvals(i) = derfc(dampingAlpha*rval)
219	+	enddo
220	+
221	+	call newSpline(f0spline, xvals, yvals, .true.)
222	+
223	+	haveElectroSpline = .true.
224	+	end subroutine buildElectroSpline
225	+
226	+	subroutine newElectrostaticType(c_ident, is_Charge, is_Dipole, &
227	+	is_SplitDipole, is_Quadrupole, is_Tap, status)
228	+
229		integer, intent(in) :: c_ident
230		logical, intent(in) :: is_Charge
231		logical, intent(in) :: is_Dipole
232		logical, intent(in) :: is_SplitDipole
233		logical, intent(in) :: is_Quadrupole
234	+	logical, intent(in) :: is_Tap
235		integer, intent(out) :: status
236		integer :: nAtypes, myATID, i, j
237
238		status = 0
239		myATID = getFirstMatchingElement(atypes, "c_ident", c_ident)
240	<
240	>
241		!! Be simple-minded and assume that we need an ElectrostaticMap that
242		!! is the same size as the total number of atom types
243
244		if (.not.allocated(ElectrostaticMap)) then
245	<
245	>
246		nAtypes = getSize(atypes)
247	<
247	>
248		if (nAtypes == 0) then
249		status = -1
250		return
251		end if
252	<
253	<	if (.not. allocated(ElectrostaticMap)) then
254	<	allocate(ElectrostaticMap(nAtypes))
125	<	endif
126	<
252	>
253	>	allocate(ElectrostaticMap(nAtypes))
254	>
255		end if
256
257		if (myATID .gt. size(ElectrostaticMap)) then
258		status = -1
259		return
260		endif
261	<
261	>
262		! set the values for ElectrostaticMap for this atom type:
263
264		ElectrostaticMap(myATID)%c_ident = c_ident
#	Line 138 | Line 266 | contains
266		ElectrostaticMap(myATID)%is_Dipole = is_Dipole
267		ElectrostaticMap(myATID)%is_SplitDipole = is_SplitDipole
268		ElectrostaticMap(myATID)%is_Quadrupole = is_Quadrupole
269	<
269	>	ElectrostaticMap(myATID)%is_Tap = is_Tap
270	>
271	>	hasElectrostaticMap = .true.
272	>
273		end subroutine newElectrostaticType
274
275		subroutine setCharge(c_ident, charge, status)
#	Line 150 | Line 281 | contains
281		status = 0
282		myATID = getFirstMatchingElement(atypes, "c_ident", c_ident)
283
284	<	if (.not.allocated(ElectrostaticMap)) then
284	>	if (.not.hasElectrostaticMap) then
285		call handleError("electrostatic", "no ElectrostaticMap was present before first call of setCharge!")
286		status = -1
287		return
#	Line 166 | Line 297 | contains
297		call handleError("electrostatic", "Attempt to setCharge of an atom type that is not a charge!")
298		status = -1
299		return
300	<	endif
300	>	endif
301
302		ElectrostaticMap(myATID)%charge = charge
303		end subroutine setCharge
#	Line 180 | Line 311 | contains
311		status = 0
312		myATID = getFirstMatchingElement(atypes, "c_ident", c_ident)
313
314	<	if (.not.allocated(ElectrostaticMap)) then
314	>	if (.not.hasElectrostaticMap) then
315		call handleError("electrostatic", "no ElectrostaticMap was present before first call of setDipoleMoment!")
316		status = -1
317		return
#	Line 210 | Line 341 | contains
341		status = 0
342		myATID = getFirstMatchingElement(atypes, "c_ident", c_ident)
343
344	<	if (.not.allocated(ElectrostaticMap)) then
344	>	if (.not.hasElectrostaticMap) then
345		call handleError("electrostatic", "no ElectrostaticMap was present before first call of setSplitDipoleDistance!")
346		status = -1
347		return
#	Line 240 | Line 371 | contains
371		status = 0
372		myATID = getFirstMatchingElement(atypes, "c_ident", c_ident)
373
374	<	if (.not.allocated(ElectrostaticMap)) then
374	>	if (.not.hasElectrostaticMap) then
375		call handleError("electrostatic", "no ElectrostaticMap was present before first call of setQuadrupoleMoments!")
376		status = -1
377		return
#	Line 257 | Line 388 | contains
388		status = -1
389		return
390		endif
391	<
391	>
392		do i = 1, 3
393	<	ElectrostaticMap(myATID)%quadrupole_moments(i) = &
394	<	quadrupole_moments(i)
395	<	enddo
393	>	ElectrostaticMap(myATID)%quadrupole_moments(i) = &
394	>	quadrupole_moments(i)
395	>	enddo
396
397		end subroutine setQuadrupoleMoments
398
399	<
399	>
400		function getCharge(atid) result (c)
401		integer, intent(in) :: atid
402		integer :: localError
403		real(kind=dp) :: c
404	<
405	<	if (.not.allocated(ElectrostaticMap)) then
404	>
405	>	if (.not.hasElectrostaticMap) then
406		call handleError("electrostatic", "no ElectrostaticMap was present before first call of getCharge!")
407		return
408		end if
409	<
409	>
410		if (.not.ElectrostaticMap(atid)%is_Charge) then
411		call handleError("electrostatic", "getCharge was called for an atom type that isn't a charge!")
412		return
413		endif
414	<
414	>
415		c = ElectrostaticMap(atid)%charge
416		end function getCharge
417
#	Line 288 | Line 419 | contains
419		integer, intent(in) :: atid
420		integer :: localError
421		real(kind=dp) :: dm
422	<
423	<	if (.not.allocated(ElectrostaticMap)) then
422	>
423	>	if (.not.hasElectrostaticMap) then
424		call handleError("electrostatic", "no ElectrostaticMap was present before first call of getDipoleMoment!")
425		return
426		end if
427	<
427	>
428		if (.not.ElectrostaticMap(atid)%is_Dipole) then
429		call handleError("electrostatic", "getDipoleMoment was called for an atom type that isn't a dipole!")
430		return
431		endif
432	<
432	>
433		dm = ElectrostaticMap(atid)%dipole_moment
434		end function getDipoleMoment
435
436	<	subroutine doElectrostaticPair(atom1, atom2, d, rij, r2, sw, &
437	<	vpair, fpair, pot, eFrame, f, t, do_pot)
438	<
436	>	subroutine checkSummationMethod()
437	>
438	>	if (.not.haveDefaultCutoff) then
439	>	call handleError("checkSummationMethod", "no Default Cutoff set!")
440	>	endif
441	>
442	>	rcuti = 1.0_dp / defaultCutoff
443	>	rcuti2 = rcuti*rcuti
444	>	rcuti3 = rcuti2*rcuti
445	>	rcuti4 = rcuti2*rcuti2
446	>
447	>	if (screeningMethod .eq. DAMPED) then
448	>	if (.not.haveDampingAlpha) then
449	>	call handleError("checkSummationMethod", "no Damping Alpha set!")
450	>	endif
451	>
452	>	if (.not.haveDefaultCutoff) then
453	>	call handleError("checkSummationMethod", "no Default Cutoff set!")
454	>	endif
455	>
456	>	constEXP = exp(-alpha2*defaultCutoff2)
457	>	invRootPi = 0.56418958354775628695_dp
458	>	alphaPi = 2.0_dp*dampingAlpha*invRootPi
459	>	f0c = derfc(dampingAlpha*defaultCutoff)
460	>	f1c = alphaPi*defaultCutoff*constEXP + f0c
461	>	f2c = alphaPi*2.0_dp*alpha2*constEXP
462	>	f3c = alphaPi*2.0_dp*alpha2*constEXP*defaultCutoff2*defaultCutoff
463	>	endif
464	>
465	>	if (summationMethod .eq. REACTION_FIELD) then
466	>	if (haveDielectric) then
467	>	defaultCutoff2 = defaultCutoff*defaultCutoff
468	>	preRF = (dielectric-1.0_dp) / &
469	>	((2.0_dp*dielectric+1.0_dp)*defaultCutoff2*defaultCutoff)
470	>	preRF2 = 2.0_dp*preRF
471	>	else
472	>	call handleError("checkSummationMethod", "Dielectric not set")
473	>	endif
474	>
475	>	endif
476	>
477	>	if (.not.haveElectroSpline) then
478	>	call buildElectroSpline()
479	>	end if
480	>
481	>	summationMethodChecked = .true.
482	>	end subroutine checkSummationMethod
483	>
484	>
485	>	subroutine doElectrostaticPair(atom1, atom2, d, rij, r2, rcut, sw, &
486	>	vpair, fpair, pot, eFrame, f, t, do_pot)
487	>
488		logical, intent(in) :: do_pot
489	<
489	>
490		integer, intent(in) :: atom1, atom2
491		integer :: localError
492
493	<	real(kind=dp), intent(in) :: rij, r2, sw
493	>	real(kind=dp), intent(in) :: rij, r2, sw, rcut
494		real(kind=dp), intent(in), dimension(3) :: d
495		real(kind=dp), intent(inout) :: vpair
496	<	real(kind=dp), intent(inout), dimension(3) :: fpair
496	>	real(kind=dp), intent(inout), dimension(3) :: fpair
497
498		real( kind = dp ) :: pot
499		real( kind = dp ), dimension(9,nLocal) :: eFrame
500		real( kind = dp ), dimension(3,nLocal) :: f
501	+	real( kind = dp ), dimension(3,nLocal) :: felec
502		real( kind = dp ), dimension(3,nLocal) :: t
503	<
503	>
504		real (kind = dp), dimension(3) :: ux_i, uy_i, uz_i
505		real (kind = dp), dimension(3) :: ux_j, uy_j, uz_j
506		real (kind = dp), dimension(3) :: dudux_i, duduy_i, duduz_i
#	Line 327 | Line 508 | contains
508
509		logical :: i_is_Charge, i_is_Dipole, i_is_SplitDipole, i_is_Quadrupole
510		logical :: j_is_Charge, j_is_Dipole, j_is_SplitDipole, j_is_Quadrupole
511	+	logical :: i_is_Tap, j_is_Tap
512		integer :: me1, me2, id1, id2
513		real (kind=dp) :: q_i, q_j, mu_i, mu_j, d_i, d_j
514		real (kind=dp) :: qxx_i, qyy_i, qzz_i
#	Line 334 | Line 516 | contains
516		real (kind=dp) :: cx_i, cy_i, cz_i
517		real (kind=dp) :: cx_j, cy_j, cz_j
518		real (kind=dp) :: cx2, cy2, cz2
519	<	real (kind=dp) :: ct_i, ct_j, ct_ij, a1
519	>	real (kind=dp) :: ct_i, ct_j, ct_ij, a0, a1
520		real (kind=dp) :: riji, ri, ri2, ri3, ri4
521	<	real (kind=dp) :: pref, vterm, epot, dudr
521	>	real (kind=dp) :: pref, vterm, epot, dudr, vterm1, vterm2
522		real (kind=dp) :: xhat, yhat, zhat
523		real (kind=dp) :: dudx, dudy, dudz
524		real (kind=dp) :: scale, sc2, bigR
525	+	real (kind=dp) :: varEXP
526	+	real (kind=dp) :: pot_term
527	+	real (kind=dp) :: preVal, rfVal
528	+	real (kind=dp) :: cti3, ctj3, ctidotj
529	+	real (kind=dp) :: ri7damp, ri5damp, prei3, prei4
530	+	real (kind=dp) :: xhatdot2, yhatdot2, zhatdot2
531	+	real (kind=dp) :: xhatdot5, yhatdot5, zhatdot5
532
533	<	if (.not.allocated(ElectrostaticMap)) then
534	<	call handleError("electrostatic", "no ElectrostaticMap was present before first call of do_electrostatic_pair!")
535	<	return
347	<	end if
533	>	if (.not.summationMethodChecked) then
534	>	call checkSummationMethod()
535	>	endif
536
537		#ifdef IS_MPI
538		me1 = atid_Row(atom1)
#	Line 356 | Line 544 | contains
544
545		!! some variables we'll need independent of electrostatic type:
546
547	<	riji = 1.0d0 / rij
548	<
547	>	riji = 1.0_dp / rij
548	>
549		xhat = d(1) * riji
550		yhat = d(2) * riji
551		zhat = d(3) * riji
552
553		!! logicals
366	–
554		i_is_Charge = ElectrostaticMap(me1)%is_Charge
555		i_is_Dipole = ElectrostaticMap(me1)%is_Dipole
556		i_is_SplitDipole = ElectrostaticMap(me1)%is_SplitDipole
557		i_is_Quadrupole = ElectrostaticMap(me1)%is_Quadrupole
558	+	i_is_Tap = ElectrostaticMap(me1)%is_Tap
559
560		j_is_Charge = ElectrostaticMap(me2)%is_Charge
561		j_is_Dipole = ElectrostaticMap(me2)%is_Dipole
562		j_is_SplitDipole = ElectrostaticMap(me2)%is_SplitDipole
563		j_is_Quadrupole = ElectrostaticMap(me2)%is_Quadrupole
564	+	j_is_Tap = ElectrostaticMap(me2)%is_Tap
565
566		if (i_is_Charge) then
567		q_i = ElectrostaticMap(me1)%charge
568		endif
569	<
569	>
570		if (i_is_Dipole) then
571		mu_i = ElectrostaticMap(me1)%dipole_moment
572		#ifdef IS_MPI
#	Line 394 | Line 583 | contains
583		if (i_is_SplitDipole) then
584		d_i = ElectrostaticMap(me1)%split_dipole_distance
585		endif
586	<
586	>	duduz_i = zero
587		endif
588
589		if (i_is_Quadrupole) then
#	Line 425 | Line 614 | contains
614		cx_i = ux_i(1)*xhat + ux_i(2)*yhat + ux_i(3)*zhat
615		cy_i = uy_i(1)*xhat + uy_i(2)*yhat + uy_i(3)*zhat
616		cz_i = uz_i(1)*xhat + uz_i(2)*yhat + uz_i(3)*zhat
617	+	dudux_i = zero
618	+	duduy_i = zero
619	+	duduz_i = zero
620		endif
621
430	–
622		if (j_is_Charge) then
623		q_j = ElectrostaticMap(me2)%charge
624		endif
625	<
625	>
626		if (j_is_Dipole) then
627		mu_j = ElectrostaticMap(me2)%dipole_moment
628		#ifdef IS_MPI
#	Line 448 | Line 639 | contains
639		if (j_is_SplitDipole) then
640		d_j = ElectrostaticMap(me2)%split_dipole_distance
641		endif
642	+	duduz_j = zero
643		endif
644
645		if (j_is_Quadrupole) then
#	Line 478 | Line 670 | contains
670		cx_j = ux_j(1)*xhat + ux_j(2)*yhat + ux_j(3)*zhat
671		cy_j = uy_j(1)*xhat + uy_j(2)*yhat + uy_j(3)*zhat
672		cz_j = uz_j(1)*xhat + uz_j(2)*yhat + uz_j(3)*zhat
673	+	dudux_j = zero
674	+	duduy_j = zero
675	+	duduz_j = zero
676		endif
677	+
678	+	epot = zero
679	+	dudx = zero
680	+	dudy = zero
681	+	dudz = zero
682
683	<	epot = 0.0_dp
484	<	dudx = 0.0_dp
485	<	dudy = 0.0_dp
486	<	dudz = 0.0_dp
683	>	if (i_is_Charge) then
684
685	<	dudux_i = 0.0_dp
686	<	duduy_i = 0.0_dp
687	<	duduz_i = 0.0_dp
685	>	if (j_is_Charge) then
686	>	if (screeningMethod .eq. DAMPED) then
687	>	! assemble the damping variables
688	>	call lookupUniformSpline1d(f0spline, rij, f0, df0)
689	>	f1 = -rij * df0 + f0
690	>	endif
691
692	<	dudux_j = 0.0_dp
493	<	duduy_j = 0.0_dp
494	<	duduz_j = 0.0_dp
692	>	preVal = pre11 * q_i * q_j
693
694	<	if (i_is_Charge) then
694	>	if (summationMethod .eq. SHIFTED_POTENTIAL) then
695	>	vterm = preVal * (riji*f0 - rcuti*f0c)
696	>
697	>	dudr  = -sw * preVal * riji * riji * f1
698	>
699	>	elseif (summationMethod .eq. SHIFTED_FORCE) then
700	>	vterm = preVal * ( riji*f0 - rcuti*f0c + &
701	>	f1c*rcuti2*(rij-defaultCutoff) )
702	>
703	>	dudr  = -sw*preVal * (riji*riji*f1 - rcuti2*f1c)
704	>
705	>	elseif (summationMethod .eq. REACTION_FIELD) then
706	>	rfVal = preRF*rij*rij
707	>	vterm = preVal * ( riji + rfVal )
708	>
709	>	dudr  = sw * preVal * ( 2.0_dp*rfVal - riji )*riji
710	>
711	>	else
712	>	vterm = preVal * riji*f0
713	>
714	>	dudr  = - sw * preVal * riji*riji*f1
715	>
716	>	endif
717
498	–	if (j_is_Charge) then
499	–
500	–	vterm = pre11 * q_i * q_j * riji
718		vpair = vpair + vterm
719		epot = epot + sw*vterm
720
504	–	dudr  = - sw * vterm * riji
505	–
721		dudx = dudx + dudr * xhat
722		dudy = dudy + dudr * yhat
723		dudz = dudz + dudr * zhat
724	<
724	>
725		endif
726
727		if (j_is_Dipole) then
728	<
729	<	if (j_is_SplitDipole) then
730	<	BigR = sqrt(r2 + 0.25_dp * d_j * d_j)
731	<	ri = 1.0_dp / BigR
732	<	scale = rij * ri
518	<	else
519	<	ri = riji
520	<	scale = 1.0_dp
728	>	if (screeningMethod .eq. DAMPED) then
729	>	! assemble the damping variables
730	>	call lookupUniformSpline1d(f0spline, rij, f0, df0)
731	>	f1 = -rij * df0 + f0
732	>	f3 = -2.0_dp*alpha2*df0*rij*rij*rij
733		endif
734
523	–	ri2 = ri * ri
524	–	ri3 = ri2 * ri
525	–	sc2 = scale * scale
526	–
735		pref = pre12 * q_i * mu_j
528	–	vterm = - pref * ct_j * ri2 * scale
529	–	vpair = vpair + vterm
530	–	epot = epot + sw * vterm
736
737	<	!! this has a + sign in the () because the rij vector is
738	<	!! r_j - r_i and the charge-dipole potential takes the origin
739	<	!! as the point dipole, which is atom j in this case.
737	>	if (summationMethod .eq. REACTION_FIELD) then
738	>	ri2 = riji * riji
739	>	ri3 = ri2 * riji
740	>
741	>	vterm = - pref * ct_j * ( ri2 - preRF2*rij )
742	>	vpair = vpair + vterm
743	>	epot = epot + sw*vterm
744	>
745	>	dudx = dudx - sw*pref*( ri3*(uz_j(1) - 3.0_dp*ct_j*xhat) - &
746	>	preRF2*uz_j(1) )
747	>	dudy = dudy - sw*pref*( ri3*(uz_j(2) - 3.0_dp*ct_j*yhat) - &
748	>	preRF2*uz_j(2) )
749	>	dudz = dudz - sw*pref*( ri3*(uz_j(3) - 3.0_dp*ct_j*zhat) - &
750	>	preRF2*uz_j(3) )
751	>	duduz_j(1) = duduz_j(1) - sw*pref * xhat * ( ri2 - preRF2*rij )
752	>	duduz_j(2) = duduz_j(2) - sw*pref * yhat * ( ri2 - preRF2*rij )
753	>	duduz_j(3) = duduz_j(3) - sw*pref * zhat * ( ri2 - preRF2*rij )
754
755	<	dudx = dudx - pref * sw * ri3 * ( uz_j(1) - 3.0d0*ct_j*xhat*sc2)
756	<	dudy = dudy - pref * sw * ri3 * ( uz_j(2) - 3.0d0*ct_j*yhat*sc2)
757	<	dudz = dudz - pref * sw * ri3 * ( uz_j(3) - 3.0d0*ct_j*zhat*sc2)
755	>	else
756	>	if (j_is_SplitDipole) then
757	>	BigR = sqrt(r2 + 0.25_dp * d_j * d_j)
758	>	ri = 1.0_dp / BigR
759	>	scale = rij * ri
760	>	else
761	>	ri = riji
762	>	scale = 1.0_dp
763	>	endif
764	>
765	>	ri2 = ri * ri
766	>	ri3 = ri2 * ri
767	>	sc2 = scale * scale
768
769	<	duduz_j(1) = duduz_j(1) - pref * sw * ri2 * xhat * scale
770	<	duduz_j(2) = duduz_j(2) - pref * sw * ri2 * yhat * scale
771	<	duduz_j(3) = duduz_j(3) - pref * sw * ri2 * zhat * scale
772	<
769	>	pot_term =  ri2 * scale * f1
770	>	vterm = -pref * ct_j * pot_term
771	>	vpair = vpair + vterm
772	>	epot = epot + sw*vterm
773	>
774	>	prei3 = sw*pref*ri3
775	>	ri5damp = 3.0_dp*f1 + f3
776	>
777	>	dudx = dudx - prei3 * ( uz_j(1)*f1 - ct_j*xhat*sc2*ri5damp )
778	>	dudy = dudy - prei3 * ( uz_j(2)*f1 - ct_j*yhat*sc2*ri5damp )
779	>	dudz = dudz - prei3 * ( uz_j(3)*f1 - ct_j*zhat*sc2*ri5damp )
780	>
781	>	duduz_j(1) = duduz_j(1) - sw*pref * pot_term * xhat
782	>	duduz_j(2) = duduz_j(2) - sw*pref * pot_term * yhat
783	>	duduz_j(3) = duduz_j(3) - sw*pref * pot_term * zhat
784	>
785	>	endif
786		endif
787
788		if (j_is_Quadrupole) then
789	+	if (screeningMethod .eq. DAMPED) then
790	+	! assemble the damping variables
791	+	call lookupUniformSpline1d(f0spline, rij, f0, df0)
792	+	f1 = -rij * df0 + f0
793	+	f2 = -2.0_dp*alpha2*df0
794	+	f3 = f2*r2*rij
795	+	f4 = 0.4_dp*alpha2*f3*r2
796	+	endif
797	+	ri5damp = f1 + f3*one_third
798	+	ri7damp = ri5damp + f4
799	+
800		ri2 = riji * riji
801		ri3 = ri2 * riji
549	–	ri4 = ri2 * ri2
802		cx2 = cx_j * cx_j
803		cy2 = cy_j * cy_j
804		cz2 = cz_j * cz_j
805
806	+	pref =  pre14 * q_i * one_third
807
808	<	pref =  pre14 * q_i / 3.0_dp
809	<	vterm = pref * ri3 * (qxx_j * (3.0_dp*cx2 - 1.0_dp) + &
810	<	qyy_j * (3.0_dp*cy2 - 1.0_dp) + &
811	<	qzz_j * (3.0_dp*cz2 - 1.0_dp))
808	>	pot_term = ri3*( qxx_j*(3.0_dp*cx2*ri5damp - f1) + &
809	>	qyy_j*(3.0_dp*cy2*ri5damp - f1) + &
810	>	qzz_j*(3.0_dp*cz2*ri5damp - f1) )
811	>	vterm = pref * pot_term
812		vpair = vpair + vterm
813	<	epot = epot + sw * vterm
813	>	epot = epot + sw*vterm
814
815	<	dudx = dudx - 5.0_dp*sw*vterm*riji*xhat + pref * sw * ri4 * ( &
816	<	qxx_j*(6.0_dp*cx_j*ux_j(1) - 2.0_dp*xhat) + &
817	<	qyy_j*(6.0_dp*cy_j*uy_j(1) - 2.0_dp*xhat) + &
818	<	qzz_j*(6.0_dp*cz_j*uz_j(1) - 2.0_dp*xhat) )
819	<	dudy = dudy - 5.0_dp*sw*vterm*riji*yhat + pref * sw * ri4 * ( &
820	<	qxx_j*(6.0_dp*cx_j*ux_j(2) - 2.0_dp*yhat) + &
821	<	qyy_j*(6.0_dp*cy_j*uy_j(2) - 2.0_dp*yhat) + &
822	<	qzz_j*(6.0_dp*cz_j*uz_j(2) - 2.0_dp*yhat) )
823	<	dudz = dudz - 5.0_dp*sw*vterm*riji*zhat + pref * sw * ri4 * ( &
824	<	qxx_j*(6.0_dp*cx_j*ux_j(3) - 2.0_dp*zhat) + &
825	<	qyy_j*(6.0_dp*cy_j*uy_j(3) - 2.0_dp*zhat) + &
826	<	qzz_j*(6.0_dp*cz_j*uz_j(3) - 2.0_dp*zhat) )
815	>	! precompute variables for convenience (and obfuscation unfortunatly)
816	>	prei3 = 3.0_dp*sw*pref*ri3
817	>	prei4 = prei3*riji
818	>	xhatdot2 = xhat*2.0_dp * ri5damp
819	>	yhatdot2 = yhat*2.0_dp * ri5damp
820	>	zhatdot2 = zhat*2.0_dp * ri5damp
821	>	xhatdot5 = xhat*5.0_dp * ri7damp
822	>	yhatdot5 = yhat*5.0_dp * ri7damp
823	>	zhatdot5 = zhat*5.0_dp * ri7damp
824	>
825	>	dudx = dudx - prei4 * ( &
826	>	qxx_j*(cx2*xhatdot5 - (2.0_dp*cx_j*ux_j(1) + xhat)*ri5damp) + &
827	>	qyy_j*(cy2*xhatdot5 - (2.0_dp*cy_j*uy_j(1) + xhat)*ri5damp) + &
828	>	qzz_j*(cz2*xhatdot5 - (2.0_dp*cz_j*uz_j(1) + xhat)*ri5damp) )
829	>	dudy = dudy - prei4 * ( &
830	>	qxx_j*(cx2*yhatdot5 - (2.0_dp*cx_j*ux_j(2) + yhat)*ri5damp) + &
831	>	qyy_j*(cy2*yhatdot5 - (2.0_dp*cy_j*uy_j(2) + yhat)*ri5damp) + &
832	>	qzz_j*(cz2*yhatdot5 - (2.0_dp*cz_j*uz_j(2) + yhat)*ri5damp) )
833	>	dudz = dudz - prei4 * ( &
834	>	qxx_j*(cx2*zhatdot5 - (2.0_dp*cx_j*ux_j(3) + zhat)*ri5damp) + &
835	>	qyy_j*(cy2*zhatdot5 - (2.0_dp*cy_j*uy_j(3) + zhat)*ri5damp) + &
836	>	qzz_j*(cz2*zhatdot5 - (2.0_dp*cz_j*uz_j(3) + zhat)*ri5damp) )
837
838	<	dudux_j(1) = dudux_j(1) + pref * sw * ri3 * (qxx_j*6.0_dp*cx_j*xhat)
839	<	dudux_j(2) = dudux_j(2) + pref * sw * ri3 * (qxx_j*6.0_dp*cx_j*yhat)
840	<	dudux_j(3) = dudux_j(3) + pref * sw * ri3 * (qxx_j*6.0_dp*cx_j*zhat)
838	>	dudux_j(1) = dudux_j(1) + prei3*(qxx_j*cx_j*xhatdot2)
839	>	dudux_j(2) = dudux_j(2) + prei3*(qxx_j*cx_j*yhatdot2)
840	>	dudux_j(3) = dudux_j(3) + prei3*(qxx_j*cx_j*zhatdot2)
841	>
842	>	duduy_j(1) = duduy_j(1) + prei3*(qyy_j*cy_j*xhatdot2)
843	>	duduy_j(2) = duduy_j(2) + prei3*(qyy_j*cy_j*yhatdot2)
844	>	duduy_j(3) = duduy_j(3) + prei3*(qyy_j*cy_j*zhatdot2)
845	>
846	>	duduz_j(1) = duduz_j(1) + prei3*(qzz_j*cz_j*xhatdot2)
847	>	duduz_j(2) = duduz_j(2) + prei3*(qzz_j*cz_j*yhatdot2)
848	>	duduz_j(3) = duduz_j(3) + prei3*(qzz_j*cz_j*zhatdot2)
849
850	<	duduy_j(1) = duduy_j(1) + pref * sw * ri3 * (qyy_j*6.0_dp*cy_j*xhat)
580	<	duduy_j(2) = duduy_j(2) + pref * sw * ri3 * (qyy_j*6.0_dp*cy_j*yhat)
581	<	duduy_j(3) = duduy_j(3) + pref * sw * ri3 * (qyy_j*6.0_dp*cy_j*zhat)
582	<
583	<	duduz_j(1) = duduz_j(1) + pref * sw * ri3 * (qzz_j*6.0_dp*cz_j*xhat)
584	<	duduz_j(2) = duduz_j(2) + pref * sw * ri3 * (qzz_j*6.0_dp*cz_j*yhat)
585	<	duduz_j(3) = duduz_j(3) + pref * sw * ri3 * (qzz_j*6.0_dp*cz_j*zhat)
850	>
851		endif
587	–
852		endif
853	<
853	>
854		if (i_is_Dipole) then
591	–
592	–	if (j_is_Charge) then
855
856	<	if (i_is_SplitDipole) then
857	<	BigR = sqrt(r2 + 0.25_dp * d_i * d_i)
858	<	ri = 1.0_dp / BigR
859	<	scale = rij * ri
860	<	else
861	<	ri = riji
600	<	scale = 1.0_dp
856	>	if (j_is_Charge) then
857	>	if (screeningMethod .eq. DAMPED) then
858	>	! assemble the damping variables
859	>	call lookupUniformSpline1d(f0spline, rij, f0, df0)
860	>	f1 = -rij * df0 + f0
861	>	f3 = -2.0_dp*alpha2*df0*r2*rij
862		endif
863	+
864	+	pref = pre12 * q_j * mu_i
865	+
866	+	if (summationMethod .eq. REACTION_FIELD) then
867
868	<	ri2 = ri * ri
869	<	ri3 = ri2 * ri
870	<	sc2 = scale * scale
868	>	ri2 = riji * riji
869	>	ri3 = ri2 * riji
870	>
871	>	vterm = pref * ct_i * ( ri2 - preRF2*rij )
872	>	vpair = vpair + vterm
873	>	epot = epot + sw*vterm
874
875	<	pref = pre12 * q_j * mu_i
876	<	vterm = pref * ct_i * ri2 * scale
877	<	vpair = vpair + vterm
878	<	epot = epot + sw * vterm
875	>	dudx = dudx + sw*pref * ( ri3*(uz_i(1) - 3.0_dp*ct_i*xhat) - &
876	>	preRF2*uz_i(1) )
877	>	dudy = dudy + sw*pref * ( ri3*(uz_i(2) - 3.0_dp*ct_i*yhat) - &
878	>	preRF2*uz_i(2) )
879	>	dudz = dudz + sw*pref * ( ri3*(uz_i(3) - 3.0_dp*ct_i*zhat) - &
880	>	preRF2*uz_i(3) )
881	>
882	>	duduz_i(1) = duduz_i(1) + sw*pref * xhat * ( ri2 - preRF2*rij )
883	>	duduz_i(2) = duduz_i(2) + sw*pref * yhat * ( ri2 - preRF2*rij )
884	>	duduz_i(3) = duduz_i(3) + sw*pref * zhat * ( ri2 - preRF2*rij )
885
612	–	dudx = dudx + pref * sw * ri3 * ( uz_i(1) - 3.0d0 * ct_i * xhat*sc2)
613	–	dudy = dudy + pref * sw * ri3 * ( uz_i(2) - 3.0d0 * ct_i * yhat*sc2)
614	–	dudz = dudz + pref * sw * ri3 * ( uz_i(3) - 3.0d0 * ct_i * zhat*sc2)
615	–
616	–	duduz_i(1) = duduz_i(1) + pref * sw * ri2 * xhat * scale
617	–	duduz_i(2) = duduz_i(2) + pref * sw * ri2 * yhat * scale
618	–	duduz_i(3) = duduz_i(3) + pref * sw * ri2 * zhat * scale
619	–	endif
620	–
621	–	if (j_is_Dipole) then
622	–
623	–	if (i_is_SplitDipole) then
624	–	if (j_is_SplitDipole) then
625	–	BigR = sqrt(r2 + 0.25_dp * d_i * d_i + 0.25_dp * d_j * d_j)
626	–	else
627	–	BigR = sqrt(r2 + 0.25_dp * d_i * d_i)
628	–	endif
629	–	ri = 1.0_dp / BigR
630	–	scale = rij * ri
886		else
887	<	if (j_is_SplitDipole) then
888	<	BigR = sqrt(r2 + 0.25_dp * d_j * d_j)
887	>	if (i_is_SplitDipole) then
888	>	BigR = sqrt(r2 + 0.25_dp * d_i * d_i)
889		ri = 1.0_dp / BigR
890	<	scale = rij * ri
891	<	else
890	>	scale = rij * ri
891	>	else
892		ri = riji
893		scale = 1.0_dp
894		endif
895	+
896	+	ri2 = ri * ri
897	+	ri3 = ri2 * ri
898	+	sc2 = scale * scale
899	+
900	+	pot_term = ri2 * f1 * scale
901	+	vterm = pref * ct_i * pot_term
902	+	vpair = vpair + vterm
903	+	epot = epot + sw*vterm
904	+
905	+	prei3 = sw*pref*ri3
906	+	ri5damp = 3.0_dp*f1 + f3
907	+
908	+	dudx = dudx + prei3 * ( uz_i(1)*f1 - ct_i*xhat*sc2*ri5damp )
909	+	dudy = dudy + prei3 * ( uz_i(2)*f1 - ct_i*yhat*sc2*ri5damp )
910	+	dudz = dudz + prei3 * ( uz_i(3)*f1 - ct_i*zhat*sc2*ri5damp )
911	+
912	+	duduz_i(1) = duduz_i(1) + sw*pref * pot_term * xhat
913	+	duduz_i(2) = duduz_i(2) + sw*pref * pot_term * yhat
914	+	duduz_i(3) = duduz_i(3) + sw*pref * pot_term * zhat
915	+
916		endif
917	+	endif
918	+
919	+	if (j_is_Dipole) then
920	+	!!$          if (screeningMethod .eq. DAMPED) then
921	+	!!$             ! assemble the damping variables
922	+	!!$             call lookupUniformSpline1d(f0spline, rij, f0, df0)
923	+	!!$             f1 = -rij * df0 + f0
924	+	!!$             f2 = -2.0_dp*alpha2*df0
925	+	!!$             f3 = f2*r2*rij
926	+	!!$             f4 = 0.4_dp*alpha2*f3*r2
927	+	!!$          endif
928
929		ct_ij = uz_i(1)*uz_j(1) + uz_i(2)*uz_j(2) + uz_i(3)*uz_j(3)
930	<
931	<	ri2 = ri * ri
932	<	ri3 = ri2 * ri
930	>
931	>	ri2 = riji * riji
932	>	ri3 = ri2 * riji
933		ri4 = ri2 * ri2
647	–	sc2 = scale * scale
648	–
649	–	pref = pre22 * mu_i * mu_j
650	–	vterm = pref * ri3 * (ct_ij - 3.0d0 * ct_i * ct_j * sc2)
651	–	vpair = vpair + vterm
652	–	epot = epot + sw * vterm
934
935	<	a1 = 5.0d0 * ct_i * ct_j * sc2 - ct_ij
935	>	pref = pre22 * mu_i * mu_j
936
937	<	dudx=dudx+pref*sw*3.0d0*ri4*scale*(a1*xhat-ct_i*uz_j(1)-ct_j*uz_i(1))
938	<	dudy=dudy+pref*sw*3.0d0*ri4*scale*(a1*yhat-ct_i*uz_j(2)-ct_j*uz_i(2))
939	<	dudz=dudz+pref*sw*3.0d0*ri4*scale*(a1*zhat-ct_i*uz_j(3)-ct_j*uz_i(3))
937	>	if (summationMethod .eq. REACTION_FIELD) then
938	>	vterm = pref*( ri3*(ct_ij - 3.0_dp * ct_i * ct_j) - &
939	>	preRF2*ct_ij )
940	>	vpair = vpair + vterm
941	>	epot = epot + sw*vterm
942	>
943	>	a1 = 5.0_dp * ct_i * ct_j - ct_ij
944	>
945	>	dudx = dudx + sw*pref*3.0_dp*ri4 &
946	>	* (a1*xhat-ct_i*uz_j(1)-ct_j*uz_i(1))
947	>	dudy = dudy + sw*pref*3.0_dp*ri4 &
948	>	* (a1*yhat-ct_i*uz_j(2)-ct_j*uz_i(2))
949	>	dudz = dudz + sw*pref*3.0_dp*ri4 &
950	>	* (a1*zhat-ct_i*uz_j(3)-ct_j*uz_i(3))
951	>
952	>	duduz_i(1) = duduz_i(1) + sw*pref*(ri3*(uz_j(1)-3.0_dp*ct_j*xhat) &
953	>	- preRF2*uz_j(1))
954	>	duduz_i(2) = duduz_i(2) + sw*pref*(ri3*(uz_j(2)-3.0_dp*ct_j*yhat) &
955	>	- preRF2*uz_j(2))
956	>	duduz_i(3) = duduz_i(3) + sw*pref*(ri3*(uz_j(3)-3.0_dp*ct_j*zhat) &
957	>	- preRF2*uz_j(3))
958	>	duduz_j(1) = duduz_j(1) + sw*pref*(ri3*(uz_i(1)-3.0_dp*ct_i*xhat) &
959	>	- preRF2*uz_i(1))
960	>	duduz_j(2) = duduz_j(2) + sw*pref*(ri3*(uz_i(2)-3.0_dp*ct_i*yhat) &
961	>	- preRF2*uz_i(2))
962	>	duduz_j(3) = duduz_j(3) + sw*pref*(ri3*(uz_i(3)-3.0_dp*ct_i*zhat) &
963	>	- preRF2*uz_i(3))
964
965	<	duduz_i(1) = duduz_i(1) + pref*sw*ri3*(uz_j(1) - 3.0d0*ct_j*xhat*sc2)
966	<	duduz_i(2) = duduz_i(2) + pref*sw*ri3*(uz_j(2) - 3.0d0*ct_j*yhat*sc2)
967	<	duduz_i(3) = duduz_i(3) + pref*sw*ri3*(uz_j(3) - 3.0d0*ct_j*zhat*sc2)
965	>	else
966	>	if (i_is_SplitDipole) then
967	>	if (j_is_SplitDipole) then
968	>	BigR = sqrt(r2 + 0.25_dp * d_i * d_i + 0.25_dp * d_j * d_j)
969	>	else
970	>	BigR = sqrt(r2 + 0.25_dp * d_i * d_i)
971	>	endif
972	>	ri = 1.0_dp / BigR
973	>	scale = rij * ri
974	>	else
975	>	if (j_is_SplitDipole) then
976	>	BigR = sqrt(r2 + 0.25_dp * d_j * d_j)
977	>	ri = 1.0_dp / BigR
978	>	scale = rij * ri
979	>	else
980	>	ri = riji
981	>	scale = 1.0_dp
982	>	endif
983	>	endif
984	>
985	>	sc2 = scale * scale
986
987	<	duduz_j(1) = duduz_j(1) + pref*sw*ri3*(uz_i(1) - 3.0d0*ct_i*xhat*sc2)
988	<	duduz_j(2) = duduz_j(2) + pref*sw*ri3*(uz_i(2) - 3.0d0*ct_i*yhat*sc2)
989	<	duduz_j(3) = duduz_j(3) + pref*sw*ri3*(uz_i(3) - 3.0d0*ct_i*zhat*sc2)
990	<	endif
987	>	pot_term = (ct_ij - 3.0_dp * ct_i * ct_j * sc2)
988	>	!!$             vterm = pref * ( ri3*pot_term*f1 + (ct_i * ct_j * sc2)*f2 )
989	>	vterm = pref * ri3 * pot_term
990	>	vpair = vpair + vterm
991	>	epot = epot + sw*vterm
992	>
993	>	! precompute variables for convenience (and obfuscation
994	>	! unfortunatly)
995	>	!!$             ri5damp = f1 + f3*one_third
996	>	!!$             ri7damp = 5.0_dp*(ri5damp + f4)
997	>	prei3 = sw*pref*ri3
998	>	prei4 = 3.0_dp*sw*pref*ri4*scale
999	>	!!$             cti3 = 3.0_dp*ct_i*sc2*ri5damp
1000	>	!!$             ctj3 = 3.0_dp*ct_j*sc2*ri5damp
1001	>	cti3 = 3.0_dp*ct_i*sc2
1002	>	ctj3 = 3.0_dp*ct_j*sc2
1003	>	ctidotj = ct_i * ct_j * sc2
1004
1005	+	dudx = dudx + prei4 * ( 5.0_dp*ctidotj*xhat - &
1006	+	(ct_i*uz_j(1) + ct_j*uz_i(1) + ct_ij*xhat) )
1007	+	dudy = dudy + prei4 * ( 5.0_dp*ctidotj*yhat - &
1008	+	(ct_i*uz_j(2) + ct_j*uz_i(2) + ct_ij*yhat) )
1009	+	dudz = dudz + prei4 * ( 5.0_dp*ctidotj*zhat - &
1010	+	(ct_i*uz_j(3) + ct_j*uz_i(3) + ct_ij*zhat) )
1011	+
1012	+	duduz_i(1) = duduz_i(1) + prei3 * ( uz_j(1) - ctj3*xhat )
1013	+	duduz_i(2) = duduz_i(2) + prei3 * ( uz_j(2) - ctj3*yhat )
1014	+	duduz_i(3) = duduz_i(3) + prei3 * ( uz_j(3) - ctj3*zhat )
1015	+
1016	+	duduz_j(1) = duduz_j(1) + prei3 * ( uz_i(1) - cti3*xhat )
1017	+	duduz_j(2) = duduz_j(2) + prei3 * ( uz_i(2) - cti3*yhat )
1018	+	duduz_j(3) = duduz_j(3) + prei3 * ( uz_i(3) - cti3*zhat )
1019	+
1020	+	!!$             dudx = dudx + prei4 * ( ctidotj*xhat*ri7damp - &
1021	+	!!$                  (ct_i*uz_j(1) + ct_j*uz_i(1) + ct_ij*xhat)*ri5damp )
1022	+	!!$             dudy = dudy + prei4 * ( ctidotj*yhat*ri7damp - &
1023	+	!!$                  (ct_i*uz_j(2) + ct_j*uz_i(2) + ct_ij*yhat)*ri5damp )
1024	+	!!$             dudz = dudz + prei4 * ( ctidotj*zhat*ri7damp - &
1025	+	!!$                  (ct_i*uz_j(3) + ct_j*uz_i(3) + ct_ij*zhat)*ri5damp )
1026	+	!!$
1027	+	!!$             duduz_i(1) = duduz_i(1) + prei3 * ( uz_j(1)*f1 - ctj3*xhat )
1028	+	!!$             duduz_i(2) = duduz_i(2) + prei3 * ( uz_j(2)*f1 - ctj3*yhat )
1029	+	!!$             duduz_i(3) = duduz_i(3) + prei3 * ( uz_j(3)*f1 - ctj3*zhat )
1030	+	!!$
1031	+	!!$             duduz_j(1) = duduz_j(1) + prei3 * ( uz_i(1)*f1 - cti3*xhat )
1032	+	!!$             duduz_j(2) = duduz_j(2) + prei3 * ( uz_i(2)*f1 - cti3*yhat )
1033	+	!!$             duduz_j(3) = duduz_j(3) + prei3 * ( uz_i(3)*f1 - cti3*zhat )
1034	+
1035	+	endif
1036	+	endif
1037		endif
1038
1039		if (i_is_Quadrupole) then
1040		if (j_is_Charge) then
1041	<
1041	>	if (screeningMethod .eq. DAMPED) then
1042	>	! assemble the damping variables
1043	>	call lookupUniformSpline1d(f0spline, rij, f0, df0)
1044	>	f1 = -rij * df0 + f0
1045	>	f2 = -2.0_dp*alpha2*df0
1046	>	f3 = f2*r2*rij
1047	>	f4 = 0.4_dp*alpha2*f3*r2
1048	>	endif
1049	>	ri5damp = f1 + f3*one_third
1050	>	ri7damp = ri5damp + f4
1051	>
1052		ri2 = riji * riji
1053		ri3 = ri2 * riji
1054		ri4 = ri2 * ri2
1055		cx2 = cx_i * cx_i
1056		cy2 = cy_i * cy_i
1057		cz2 = cz_i * cz_i
1058	<
1059	<	pref = pre14 * q_j / 3.0_dp
1060	<	vterm = pref * ri3 * (qxx_i * (3.0_dp*cx2 - 1.0_dp) + &
1061	<	qyy_i * (3.0_dp*cy2 - 1.0_dp) + &
1062	<	qzz_i * (3.0_dp*cz2 - 1.0_dp))
1058	>
1059	>	pref = pre14 * q_j * one_third
1060	>
1061	>	pot_term = ri3 * ( qxx_i * (3.0_dp*cx2*ri5damp - f1) + &
1062	>	qyy_i * (3.0_dp*cy2*ri5damp - f1) + &
1063	>	qzz_i * (3.0_dp*cz2*ri5damp - f1) )
1064	>
1065	>	vterm = pref * pot_term
1066		vpair = vpair + vterm
1067	<	epot = epot + sw * vterm
1068	<
1069	<	dudx = dudx - 5.0_dp*sw*vterm*riji*xhat + pref * sw * ri4 * ( &
1070	<	qxx_i*(6.0_dp*cx_i*ux_i(1) - 2.0_dp*xhat) + &
1071	<	qyy_i*(6.0_dp*cy_i*uy_i(1) - 2.0_dp*xhat) + &
1072	<	qzz_i*(6.0_dp*cz_i*uz_i(1) - 2.0_dp*xhat) )
1073	<	dudy = dudy - 5.0_dp*sw*vterm*riji*yhat + pref * sw * ri4 * ( &
1074	<	qxx_i*(6.0_dp*cx_i*ux_i(2) - 2.0_dp*yhat) + &
1075	<	qyy_i*(6.0_dp*cy_i*uy_i(2) - 2.0_dp*yhat) + &
1076	<	qzz_i*(6.0_dp*cz_i*uz_i(2) - 2.0_dp*yhat) )
1077	<	dudz = dudz - 5.0_dp*sw*vterm*riji*zhat + pref * sw * ri4 * ( &
1078	<	qxx_i*(6.0_dp*cx_i*ux_i(3) - 2.0_dp*zhat) + &
1079	<	qyy_i*(6.0_dp*cy_i*uy_i(3) - 2.0_dp*zhat) + &
1080	<	qzz_i*(6.0_dp*cz_i*uz_i(3) - 2.0_dp*zhat) )
1081	<
1082	<	dudux_i(1) = dudux_i(1) + pref * sw * ri3 * (qxx_i*6.0_dp*cx_i*xhat)
1083	<	dudux_i(2) = dudux_i(2) + pref * sw * ri3 * (qxx_i*6.0_dp*cx_i*yhat)
1084	<	dudux_i(3) = dudux_i(3) + pref * sw * ri3 * (qxx_i*6.0_dp*cx_i*zhat)
1085	<
1086	<	duduy_i(1) = duduy_i(1) + pref * sw * ri3 * (qyy_i*6.0_dp*cy_i*xhat)
1087	<	duduy_i(2) = duduy_i(2) + pref * sw * ri3 * (qyy_i*6.0_dp*cy_i*yhat)
1088	<	duduy_i(3) = duduy_i(3) + pref * sw * ri3 * (qyy_i*6.0_dp*cy_i*zhat)
1067	>	epot = epot + sw*vterm
1068	>
1069	>	! precompute variables for convenience (and obfuscation unfortunatly)
1070	>	prei3 = 3.0_dp*sw*pref*ri3
1071	>	prei4 = prei3*riji
1072	>	xhatdot2 = xhat*2.0_dp * ri5damp
1073	>	yhatdot2 = yhat*2.0_dp * ri5damp
1074	>	zhatdot2 = zhat*2.0_dp * ri5damp
1075	>	xhatdot5 = xhat*5.0_dp * ri7damp
1076	>	yhatdot5 = yhat*5.0_dp * ri7damp
1077	>	zhatdot5 = zhat*5.0_dp * ri7damp
1078	>
1079	>	dudx = dudx - prei4 * ( &
1080	>	qxx_i*(cx2*xhatdot5 - (2.0_dp*cx_i*ux_i(1) + xhat)*ri5damp) + &
1081	>	qyy_i*(cy2*xhatdot5 - (2.0_dp*cy_i*uy_i(1) + xhat)*ri5damp) + &
1082	>	qzz_i*(cz2*xhatdot5 - (2.0_dp*cz_i*uz_i(1) + xhat)*ri5damp) )
1083	>	dudy = dudy - prei4 * ( &
1084	>	qxx_i*(cx2*yhatdot5 - (2.0_dp*cx_i*ux_i(2) + yhat)*ri5damp) + &
1085	>	qyy_i*(cy2*yhatdot5 - (2.0_dp*cy_i*uy_i(2) + yhat)*ri5damp) + &
1086	>	qzz_i*(cz2*yhatdot5 - (2.0_dp*cz_i*uz_i(2) + yhat)*ri5damp) )
1087	>	dudz = dudz - prei4 * ( &
1088	>	qxx_i*(cx2*zhatdot5 - (2.0_dp*cx_i*ux_i(3) + zhat)*ri5damp) + &
1089	>	qyy_i*(cy2*zhatdot5 - (2.0_dp*cy_i*uy_i(3) + zhat)*ri5damp) + &
1090	>	qzz_i*(cz2*zhatdot5 - (2.0_dp*cz_i*uz_i(3) + zhat)*ri5damp) )
1091
1092	<	duduz_i(1) = duduz_i(1) + pref * sw * ri3 * (qzz_i*6.0_dp*cz_i*xhat)
1093	<	duduz_i(2) = duduz_i(2) + pref * sw * ri3 * (qzz_i*6.0_dp*cz_i*yhat)
1094	<	duduz_i(3) = duduz_i(3) + pref * sw * ri3 * (qzz_i*6.0_dp*cz_i*zhat)
1092	>	dudux_i(1) = dudux_i(1) + prei3*(qxx_i*cx_i*xhatdot2)
1093	>	dudux_i(2) = dudux_i(2) + prei3*(qxx_i*cx_i*yhatdot2)
1094	>	dudux_i(3) = dudux_i(3) + prei3*(qxx_i*cx_i*zhatdot2)
1095	>
1096	>	duduy_i(1) = duduy_i(1) + prei3*(qyy_i*cy_i*xhatdot2)
1097	>	duduy_i(2) = duduy_i(2) + prei3*(qyy_i*cy_i*yhatdot2)
1098	>	duduy_i(3) = duduy_i(3) + prei3*(qyy_i*cy_i*zhatdot2)
1099	>
1100	>	duduz_i(1) = duduz_i(1) + prei3*(qzz_i*cz_i*xhatdot2)
1101	>	duduz_i(2) = duduz_i(2) + prei3*(qzz_i*cz_i*yhatdot2)
1102	>	duduz_i(3) = duduz_i(3) + prei3*(qzz_i*cz_i*zhatdot2)
1103		endif
1104		endif
1105	<
1106	<
1105	>
1106	>
1107		if (do_pot) then
1108		#ifdef IS_MPI
1109	<	pot_row(atom1) = pot_row(atom1) + 0.5d0*epot
1110	<	pot_col(atom2) = pot_col(atom2) + 0.5d0*epot
1109	>	pot_row(ELECTROSTATIC_POT,atom1) = pot_row(ELECTROSTATIC_POT,atom1) + 0.5_dp*epot
1110	>	pot_col(ELECTROSTATIC_POT,atom2) = pot_col(ELECTROSTATIC_POT,atom2) + 0.5_dp*epot
1111		#else
1112		pot = pot + epot
1113		#endif
1114		endif
1115	<
1115	>
1116		#ifdef IS_MPI
1117		f_Row(1,atom1) = f_Row(1,atom1) + dudx
1118		f_Row(2,atom1) = f_Row(2,atom1) + dudy
1119		f_Row(3,atom1) = f_Row(3,atom1) + dudz
1120	<
1120	>
1121		f_Col(1,atom2) = f_Col(1,atom2) - dudx
1122		f_Col(2,atom2) = f_Col(2,atom2) - dudy
1123		f_Col(3,atom2) = f_Col(3,atom2) - dudz
1124	<
1124	>
1125		if (i_is_Dipole .or. i_is_Quadrupole) then
1126		t_Row(1,atom1)=t_Row(1,atom1) - uz_i(2)*duduz_i(3) + uz_i(3)*duduz_i(2)
1127		t_Row(2,atom1)=t_Row(2,atom1) - uz_i(3)*duduz_i(1) + uz_i(1)*duduz_i(3)
#	Line 765 | Line 1156 | contains
1156		f(1,atom1) = f(1,atom1) + dudx
1157		f(2,atom1) = f(2,atom1) + dudy
1158		f(3,atom1) = f(3,atom1) + dudz
1159	<
1159	>
1160		f(1,atom2) = f(1,atom2) - dudx
1161		f(2,atom2) = f(2,atom2) - dudy
1162		f(3,atom2) = f(3,atom2) - dudz
1163	<
1163	>
1164		if (i_is_Dipole .or. i_is_Quadrupole) then
1165		t(1,atom1)=t(1,atom1) - uz_i(2)*duduz_i(3) + uz_i(3)*duduz_i(2)
1166		t(2,atom1)=t(2,atom1) - uz_i(3)*duduz_i(1) + uz_i(1)*duduz_i(3)
#	Line 801 | Line 1192 | contains
1192		endif
1193
1194		#endif
1195	<
1195	>
1196		#ifdef IS_MPI
1197		id1 = AtomRowToGlobal(atom1)
1198		id2 = AtomColToGlobal(atom2)
#	Line 811 | Line 1202 | contains
1202		#endif
1203
1204		if (molMembershipList(id1) .ne. molMembershipList(id2)) then
1205	<
1205	>
1206		fpair(1) = fpair(1) + dudx
1207		fpair(2) = fpair(2) + dudy
1208		fpair(3) = fpair(3) + dudz
#	Line 820 | Line 1211 | contains
1211
1212		return
1213		end subroutine doElectrostaticPair
1214	<
1214	>
1215	>	subroutine destroyElectrostaticTypes()
1216	>
1217	>	if(allocated(ElectrostaticMap)) deallocate(ElectrostaticMap)
1218	>
1219	>	end subroutine destroyElectrostaticTypes
1220	>
1221	>	subroutine self_self(atom1, eFrame, mypot, t, do_pot)
1222	>	logical, intent(in) :: do_pot
1223	>	integer, intent(in) :: atom1
1224	>	integer :: atid1
1225	>	real(kind=dp), dimension(9,nLocal) :: eFrame
1226	>	real(kind=dp), dimension(3,nLocal) :: t
1227	>	real(kind=dp) :: mu1, c1
1228	>	real(kind=dp) :: preVal, epot, mypot
1229	>	real(kind=dp) :: eix, eiy, eiz
1230	>
1231	>	! this is a local only array, so we use the local atom type id's:
1232	>	atid1 = atid(atom1)
1233	>
1234	>	if (.not.summationMethodChecked) then
1235	>	call checkSummationMethod()
1236	>	endif
1237	>
1238	>	if (summationMethod .eq. REACTION_FIELD) then
1239	>	if (ElectrostaticMap(atid1)%is_Dipole) then
1240	>	mu1 = getDipoleMoment(atid1)
1241	>
1242	>	preVal = pre22 * preRF2 * mu1*mu1
1243	>	mypot = mypot - 0.5_dp*preVal
1244	>
1245	>	! The self-correction term adds into the reaction field vector
1246	>
1247	>	eix = preVal * eFrame(3,atom1)
1248	>	eiy = preVal * eFrame(6,atom1)
1249	>	eiz = preVal * eFrame(9,atom1)
1250	>
1251	>	! once again, this is self-self, so only the local arrays are needed
1252	>	! even for MPI jobs:
1253	>
1254	>	t(1,atom1)=t(1,atom1) - eFrame(6,atom1)*eiz + &
1255	>	eFrame(9,atom1)*eiy
1256	>	t(2,atom1)=t(2,atom1) - eFrame(9,atom1)*eix + &
1257	>	eFrame(3,atom1)*eiz
1258	>	t(3,atom1)=t(3,atom1) - eFrame(3,atom1)*eiy + &
1259	>	eFrame(6,atom1)*eix
1260	>
1261	>	endif
1262	>
1263	>	elseif ( (summationMethod .eq. SHIFTED_FORCE) .or. &
1264	>	(summationMethod .eq. SHIFTED_POTENTIAL) ) then
1265	>	if (ElectrostaticMap(atid1)%is_Charge) then
1266	>	c1 = getCharge(atid1)
1267	>
1268	>	if (screeningMethod .eq. DAMPED) then
1269	>	mypot = mypot - (f0c * rcuti * 0.5_dp + &
1270	>	dampingAlpha*invRootPi) * c1 * c1
1271	>
1272	>	else
1273	>	mypot = mypot - (rcuti * 0.5_dp * c1 * c1)
1274	>
1275	>	endif
1276	>	endif
1277	>	endif
1278	>
1279	>	return
1280	>	end subroutine self_self
1281	>
1282	>	subroutine rf_self_excludes(atom1, atom2, sw, eFrame, d, rij, vpair, myPot, &
1283	>	f, t, do_pot)
1284	>	logical, intent(in) :: do_pot
1285	>	integer, intent(in) :: atom1
1286	>	integer, intent(in) :: atom2
1287	>	logical :: i_is_Charge, j_is_Charge
1288	>	logical :: i_is_Dipole, j_is_Dipole
1289	>	integer :: atid1
1290	>	integer :: atid2
1291	>	real(kind=dp), intent(in) :: rij
1292	>	real(kind=dp), intent(in) :: sw
1293	>	real(kind=dp), intent(in), dimension(3) :: d
1294	>	real(kind=dp), intent(inout) :: vpair
1295	>	real(kind=dp), dimension(9,nLocal) :: eFrame
1296	>	real(kind=dp), dimension(3,nLocal) :: f
1297	>	real(kind=dp), dimension(3,nLocal) :: t
1298	>	real (kind = dp), dimension(3) :: duduz_i
1299	>	real (kind = dp), dimension(3) :: duduz_j
1300	>	real (kind = dp), dimension(3) :: uz_i
1301	>	real (kind = dp), dimension(3) :: uz_j
1302	>	real(kind=dp) :: q_i, q_j, mu_i, mu_j
1303	>	real(kind=dp) :: xhat, yhat, zhat
1304	>	real(kind=dp) :: ct_i, ct_j
1305	>	real(kind=dp) :: ri2, ri3, riji, vterm
1306	>	real(kind=dp) :: pref, preVal, rfVal, myPot
1307	>	real(kind=dp) :: dudx, dudy, dudz, dudr
1308	>
1309	>	if (.not.summationMethodChecked) then
1310	>	call checkSummationMethod()
1311	>	endif
1312	>
1313	>	dudx = zero
1314	>	dudy = zero
1315	>	dudz = zero
1316	>
1317	>	riji = 1.0_dp/rij
1318	>
1319	>	xhat = d(1) * riji
1320	>	yhat = d(2) * riji
1321	>	zhat = d(3) * riji
1322	>
1323	>	! this is a local only array, so we use the local atom type id's:
1324	>	atid1 = atid(atom1)
1325	>	atid2 = atid(atom2)
1326	>	i_is_Charge = ElectrostaticMap(atid1)%is_Charge
1327	>	j_is_Charge = ElectrostaticMap(atid2)%is_Charge
1328	>	i_is_Dipole = ElectrostaticMap(atid1)%is_Dipole
1329	>	j_is_Dipole = ElectrostaticMap(atid2)%is_Dipole
1330	>
1331	>	if (i_is_Charge.and.j_is_Charge) then
1332	>	q_i = ElectrostaticMap(atid1)%charge
1333	>	q_j = ElectrostaticMap(atid2)%charge
1334	>
1335	>	preVal = pre11 * q_i * q_j
1336	>	rfVal = preRF*rij*rij
1337	>	vterm = preVal * rfVal
1338	>
1339	>	myPot = myPot + sw*vterm
1340	>
1341	>	dudr  = sw*preVal * 2.0_dp*rfVal*riji
1342	>
1343	>	dudx = dudx + dudr * xhat
1344	>	dudy = dudy + dudr * yhat
1345	>	dudz = dudz + dudr * zhat
1346	>
1347	>	elseif (i_is_Charge.and.j_is_Dipole) then
1348	>	q_i = ElectrostaticMap(atid1)%charge
1349	>	mu_j = ElectrostaticMap(atid2)%dipole_moment
1350	>	uz_j(1) = eFrame(3,atom2)
1351	>	uz_j(2) = eFrame(6,atom2)
1352	>	uz_j(3) = eFrame(9,atom2)
1353	>	ct_j = uz_j(1)*xhat + uz_j(2)*yhat + uz_j(3)*zhat
1354	>
1355	>	ri2 = riji * riji
1356	>	ri3 = ri2 * riji
1357	>
1358	>	pref = pre12 * q_i * mu_j
1359	>	vterm = - pref * ct_j * ( ri2 - preRF2*rij )
1360	>	myPot = myPot + sw*vterm
1361	>
1362	>	dudx = dudx - sw*pref*( ri3*(uz_j(1)-3.0_dp*ct_j*xhat) &
1363	>	- preRF2*uz_j(1) )
1364	>	dudy = dudy - sw*pref*( ri3*(uz_j(2)-3.0_dp*ct_j*yhat) &
1365	>	- preRF2*uz_j(2) )
1366	>	dudz = dudz - sw*pref*( ri3*(uz_j(3)-3.0_dp*ct_j*zhat) &
1367	>	- preRF2*uz_j(3) )
1368	>
1369	>	duduz_j(1) = duduz_j(1) - sw * pref * xhat * ( ri2 - preRF2*rij )
1370	>	duduz_j(2) = duduz_j(2) - sw * pref * yhat * ( ri2 - preRF2*rij )
1371	>	duduz_j(3) = duduz_j(3) - sw * pref * zhat * ( ri2 - preRF2*rij )
1372	>
1373	>	elseif (i_is_Dipole.and.j_is_Charge) then
1374	>	mu_i = ElectrostaticMap(atid1)%dipole_moment
1375	>	q_j = ElectrostaticMap(atid2)%charge
1376	>	uz_i(1) = eFrame(3,atom1)
1377	>	uz_i(2) = eFrame(6,atom1)
1378	>	uz_i(3) = eFrame(9,atom1)
1379	>	ct_i = uz_i(1)*xhat + uz_i(2)*yhat + uz_i(3)*zhat
1380	>
1381	>	ri2 = riji * riji
1382	>	ri3 = ri2 * riji
1383	>
1384	>	pref = pre12 * q_j * mu_i
1385	>	vterm = pref * ct_i * ( ri2 - preRF2*rij )
1386	>	myPot = myPot + sw*vterm
1387	>
1388	>	dudx = dudx + sw*pref*( ri3*(uz_i(1)-3.0_dp*ct_i*xhat) &
1389	>	- preRF2*uz_i(1) )
1390	>	dudy = dudy + sw*pref*( ri3*(uz_i(2)-3.0_dp*ct_i*yhat) &
1391	>	- preRF2*uz_i(2) )
1392	>	dudz = dudz + sw*pref*( ri3*(uz_i(3)-3.0_dp*ct_i*zhat) &
1393	>	- preRF2*uz_i(3) )
1394	>
1395	>	duduz_i(1) = duduz_i(1) + sw * pref * xhat * ( ri2 - preRF2*rij )
1396	>	duduz_i(2) = duduz_i(2) + sw * pref * yhat * ( ri2 - preRF2*rij )
1397	>	duduz_i(3) = duduz_i(3) + sw * pref * zhat * ( ri2 - preRF2*rij )
1398	>
1399	>	endif
1400	>
1401	>
1402	>	! accumulate the forces and torques resulting from the self term
1403	>	f(1,atom1) = f(1,atom1) + dudx
1404	>	f(2,atom1) = f(2,atom1) + dudy
1405	>	f(3,atom1) = f(3,atom1) + dudz
1406	>
1407	>	f(1,atom2) = f(1,atom2) - dudx
1408	>	f(2,atom2) = f(2,atom2) - dudy
1409	>	f(3,atom2) = f(3,atom2) - dudz
1410	>
1411	>	if (i_is_Dipole) then
1412	>	t(1,atom1)=t(1,atom1) - uz_i(2)*duduz_i(3) + uz_i(3)*duduz_i(2)
1413	>	t(2,atom1)=t(2,atom1) - uz_i(3)*duduz_i(1) + uz_i(1)*duduz_i(3)
1414	>	t(3,atom1)=t(3,atom1) - uz_i(1)*duduz_i(2) + uz_i(2)*duduz_i(1)
1415	>	elseif (j_is_Dipole) then
1416	>	t(1,atom2)=t(1,atom2) - uz_j(2)*duduz_j(3) + uz_j(3)*duduz_j(2)
1417	>	t(2,atom2)=t(2,atom2) - uz_j(3)*duduz_j(1) + uz_j(1)*duduz_j(3)
1418	>	t(3,atom2)=t(3,atom2) - uz_j(1)*duduz_j(2) + uz_j(2)*duduz_j(1)
1419	>	endif
1420	>
1421	>	return
1422	>	end subroutine rf_self_excludes
1423	>
1424		end module electrostatic_module

Diff Legend

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