ViewVC Help
View File | Revision Log | Show Annotations | View Changeset | Root Listing
root/OpenMD/trunk/src/nonbonded/Electrostatic.cpp
Revision: 1900
Committed: Fri Jul 12 17:38:06 2013 UTC (11 years, 10 months ago) by gezelter
File size: 43053 byte(s)
Log Message:
Added self-interaction for shifted multipoles

File Contents

# Content
1 /*
2 * Copyright (c) 2005 The University of Notre Dame. All Rights Reserved.
3 *
4 * The University of Notre Dame grants you ("Licensee") a
5 * non-exclusive, royalty free, license to use, modify and
6 * redistribute this software in source and binary code form, provided
7 * that the following conditions are met:
8 *
9 * 1. Redistributions of source code must retain the above copyright
10 * notice, this list of conditions and the following disclaimer.
11 *
12 * 2. Redistributions in binary form must reproduce the above copyright
13 * notice, this list of conditions and the following disclaimer in the
14 * documentation and/or other materials provided with the
15 * distribution.
16 *
17 * This software is provided "AS IS," without a warranty of any
18 * kind. All express or implied conditions, representations and
19 * warranties, including any implied warranty of merchantability,
20 * fitness for a particular purpose or non-infringement, are hereby
21 * excluded. The University of Notre Dame and its licensors shall not
22 * be liable for any damages suffered by licensee as a result of
23 * using, modifying or distributing the software or its
24 * derivatives. In no event will the University of Notre Dame or its
25 * licensors be liable for any lost revenue, profit or data, or for
26 * direct, indirect, special, consequential, incidental or punitive
27 * damages, however caused and regardless of the theory of liability,
28 * arising out of the use of or inability to use software, even if the
29 * University of Notre Dame has been advised of the possibility of
30 * such damages.
31 *
32 * SUPPORT OPEN SCIENCE! If you use OpenMD or its source code in your
33 * research, please cite the appropriate papers when you publish your
34 * work. Good starting points are:
35 *
36 * [1] Meineke, et al., J. Comp. Chem. 26, 252-271 (2005).
37 * [2] Fennell & Gezelter, J. Chem. Phys. 124 234104 (2006).
38 * [3] Sun, Lin & Gezelter, J. Chem. Phys. 128, 234107 (2008).
39 * [4] Kuang & Gezelter, J. Chem. Phys. 133, 164101 (2010).
40 * [5] Vardeman, Stocker & Gezelter, J. Chem. Theory Comput. 7, 834 (2011).
41 */
42
43 #include <stdio.h>
44 #include <string.h>
45
46 #include <cmath>
47 #include "nonbonded/Electrostatic.hpp"
48 #include "utils/simError.h"
49 #include "types/NonBondedInteractionType.hpp"
50 #include "types/FixedChargeAdapter.hpp"
51 #include "types/FluctuatingChargeAdapter.hpp"
52 #include "types/MultipoleAdapter.hpp"
53 #include "io/Globals.hpp"
54 #include "nonbonded/SlaterIntegrals.hpp"
55 #include "utils/PhysicalConstants.hpp"
56 #include "math/erfc.hpp"
57 #include "math/SquareMatrix.hpp"
58
59 namespace OpenMD {
60
61 Electrostatic::Electrostatic(): name_("Electrostatic"), initialized_(false),
62 forceField_(NULL), info_(NULL),
63 haveCutoffRadius_(false),
64 haveDampingAlpha_(false),
65 haveDielectric_(false),
66 haveElectroSplines_(false)
67 {}
68
69 void Electrostatic::initialize() {
70
71 Globals* simParams_ = info_->getSimParams();
72
73 summationMap_["HARD"] = esm_HARD;
74 summationMap_["NONE"] = esm_HARD;
75 summationMap_["SWITCHING_FUNCTION"] = esm_SWITCHING_FUNCTION;
76 summationMap_["SHIFTED_POTENTIAL"] = esm_SHIFTED_POTENTIAL;
77 summationMap_["SHIFTED_FORCE"] = esm_SHIFTED_FORCE;
78 summationMap_["TAYLOR_SHIFTED"] = esm_TAYLOR_SHIFTED;
79 summationMap_["REACTION_FIELD"] = esm_REACTION_FIELD;
80 summationMap_["EWALD_FULL"] = esm_EWALD_FULL;
81 summationMap_["EWALD_PME"] = esm_EWALD_PME;
82 summationMap_["EWALD_SPME"] = esm_EWALD_SPME;
83 screeningMap_["DAMPED"] = DAMPED;
84 screeningMap_["UNDAMPED"] = UNDAMPED;
85
86 // these prefactors convert the multipole interactions into kcal / mol
87 // all were computed assuming distances are measured in angstroms
88 // Charge-Charge, assuming charges are measured in electrons
89 pre11_ = 332.0637778;
90 // Charge-Dipole, assuming charges are measured in electrons, and
91 // dipoles are measured in debyes
92 pre12_ = 69.13373;
93 // Dipole-Dipole, assuming dipoles are measured in Debye
94 pre22_ = 14.39325;
95 // Charge-Quadrupole, assuming charges are measured in electrons, and
96 // quadrupoles are measured in 10^-26 esu cm^2
97 // This unit is also known affectionately as an esu centibarn.
98 pre14_ = 69.13373;
99 // Dipole-Quadrupole, assuming dipoles are measured in debyes and
100 // quadrupoles in esu centibarns:
101 pre24_ = 14.39325;
102 // Quadrupole-Quadrupole, assuming esu centibarns:
103 pre44_ = 14.39325;
104
105 // conversions for the simulation box dipole moment
106 chargeToC_ = 1.60217733e-19;
107 angstromToM_ = 1.0e-10;
108 debyeToCm_ = 3.33564095198e-30;
109
110 // Default number of points for electrostatic splines
111 np_ = 100;
112
113 // variables to handle different summation methods for long-range
114 // electrostatics:
115 summationMethod_ = esm_HARD;
116 screeningMethod_ = UNDAMPED;
117 dielectric_ = 1.0;
118
119 // check the summation method:
120 if (simParams_->haveElectrostaticSummationMethod()) {
121 string myMethod = simParams_->getElectrostaticSummationMethod();
122 toUpper(myMethod);
123 map<string, ElectrostaticSummationMethod>::iterator i;
124 i = summationMap_.find(myMethod);
125 if ( i != summationMap_.end() ) {
126 summationMethod_ = (*i).second;
127 } else {
128 // throw error
129 sprintf( painCave.errMsg,
130 "Electrostatic::initialize: Unknown electrostaticSummationMethod.\n"
131 "\t(Input file specified %s .)\n"
132 "\telectrostaticSummationMethod must be one of: \"hard\",\n"
133 "\t\"shifted_potential\", \"shifted_force\",\n"
134 "\t\"taylor_shifted\", or \"reaction_field\".\n",
135 myMethod.c_str() );
136 painCave.isFatal = 1;
137 simError();
138 }
139 } else {
140 // set ElectrostaticSummationMethod to the cutoffMethod:
141 if (simParams_->haveCutoffMethod()){
142 string myMethod = simParams_->getCutoffMethod();
143 toUpper(myMethod);
144 map<string, ElectrostaticSummationMethod>::iterator i;
145 i = summationMap_.find(myMethod);
146 if ( i != summationMap_.end() ) {
147 summationMethod_ = (*i).second;
148 }
149 }
150 }
151
152 if (summationMethod_ == esm_REACTION_FIELD) {
153 if (!simParams_->haveDielectric()) {
154 // throw warning
155 sprintf( painCave.errMsg,
156 "SimInfo warning: dielectric was not specified in the input file\n\tfor the reaction field correction method.\n"
157 "\tA default value of %f will be used for the dielectric.\n", dielectric_);
158 painCave.isFatal = 0;
159 painCave.severity = OPENMD_INFO;
160 simError();
161 } else {
162 dielectric_ = simParams_->getDielectric();
163 }
164 haveDielectric_ = true;
165 }
166
167 if (simParams_->haveElectrostaticScreeningMethod()) {
168 string myScreen = simParams_->getElectrostaticScreeningMethod();
169 toUpper(myScreen);
170 map<string, ElectrostaticScreeningMethod>::iterator i;
171 i = screeningMap_.find(myScreen);
172 if ( i != screeningMap_.end()) {
173 screeningMethod_ = (*i).second;
174 } else {
175 sprintf( painCave.errMsg,
176 "SimInfo error: Unknown electrostaticScreeningMethod.\n"
177 "\t(Input file specified %s .)\n"
178 "\telectrostaticScreeningMethod must be one of: \"undamped\"\n"
179 "or \"damped\".\n", myScreen.c_str() );
180 painCave.isFatal = 1;
181 simError();
182 }
183 }
184
185 // check to make sure a cutoff value has been set:
186 if (!haveCutoffRadius_) {
187 sprintf( painCave.errMsg, "Electrostatic::initialize has no Default "
188 "Cutoff value!\n");
189 painCave.severity = OPENMD_ERROR;
190 painCave.isFatal = 1;
191 simError();
192 }
193
194 if (screeningMethod_ == DAMPED) {
195 if (!simParams_->haveDampingAlpha()) {
196 // first set a cutoff dependent alpha value
197 // we assume alpha depends linearly with rcut from 0 to 20.5 ang
198 dampingAlpha_ = 0.425 - cutoffRadius_* 0.02;
199 if (dampingAlpha_ < 0.0) dampingAlpha_ = 0.0;
200
201 // throw warning
202 sprintf( painCave.errMsg,
203 "Electrostatic::initialize: dampingAlpha was not specified in the\n"
204 "\tinput file. A default value of %f (1/ang) will be used for the\n"
205 "\tcutoff of %f (ang).\n",
206 dampingAlpha_, cutoffRadius_);
207 painCave.severity = OPENMD_INFO;
208 painCave.isFatal = 0;
209 simError();
210 } else {
211 dampingAlpha_ = simParams_->getDampingAlpha();
212 }
213 haveDampingAlpha_ = true;
214 }
215
216 Etypes.clear();
217 Etids.clear();
218 FQtypes.clear();
219 FQtids.clear();
220 ElectrostaticMap.clear();
221 Jij.clear();
222 nElectro_ = 0;
223 nFlucq_ = 0;
224
225 Etids.resize( forceField_->getNAtomType(), -1);
226 FQtids.resize( forceField_->getNAtomType(), -1);
227
228 set<AtomType*>::iterator at;
229 for (at = simTypes_.begin(); at != simTypes_.end(); ++at) {
230 if ((*at)->isElectrostatic()) nElectro_++;
231 if ((*at)->isFluctuatingCharge()) nFlucq_++;
232 }
233
234 Jij.resize(nFlucq_);
235
236 for (at = simTypes_.begin(); at != simTypes_.end(); ++at) {
237 if ((*at)->isElectrostatic()) addType(*at);
238 }
239
240 if (summationMethod_ == esm_REACTION_FIELD) {
241 preRF_ = (dielectric_ - 1.0) /
242 ((2.0 * dielectric_ + 1.0) * pow(cutoffRadius_,3) );
243 }
244
245 RealType b0c, b1c, b2c, b3c, b4c, b5c;
246 RealType db0c_1, db0c_2, db0c_3, db0c_4, db0c_5;
247 RealType a2, expTerm, invArootPi;
248
249 RealType r = cutoffRadius_;
250 RealType r2 = r * r;
251 RealType ric = 1.0 / r;
252 RealType ric2 = ric * ric;
253
254 if (screeningMethod_ == DAMPED) {
255 a2 = dampingAlpha_ * dampingAlpha_;
256 invArootPi = 1.0 / (dampingAlpha_ * sqrt(M_PI));
257 expTerm = exp(-a2 * r2);
258 // values of Smith's B_l functions at the cutoff radius:
259 b0c = erfc(dampingAlpha_ * r) / r;
260 b1c = ( b0c + 2.0*a2 * expTerm * invArootPi) / r2;
261 b2c = (3.0 * b1c + pow(2.0*a2, 2) * expTerm * invArootPi) / r2;
262 b3c = (5.0 * b2c + pow(2.0*a2, 3) * expTerm * invArootPi) / r2;
263 b4c = (7.0 * b3c + pow(2.0*a2, 4) * expTerm * invArootPi) / r2;
264 b5c = (9.0 * b4c + pow(2.0*a2, 5) * expTerm * invArootPi) / r2;
265 //selfMult1_ = - 2.0 * a2 * invArootPi;
266 //selfMult2_ = - 4.0 * a2 * a2 * invArootPi / 3.0;
267 //selfMult4_ = - 8.0 * a2 * a2 * a2 * invArootPi / 5.0;
268 // Half the Smith self piece:
269 selfMult1_ = - a2 * invArootPi;
270 selfMult2_ = - 2.0 * a2 * a2 * invArootPi / 3.0;
271 selfMult4_ = - 4.0 * a2 * a2 * a2 * invArootPi / 5.0;
272 } else {
273 a2 = 0.0;
274 b0c = 1.0 / r;
275 b1c = ( b0c) / r2;
276 b2c = (3.0 * b1c) / r2;
277 b3c = (5.0 * b2c) / r2;
278 b4c = (7.0 * b3c) / r2;
279 b5c = (9.0 * b4c) / r2;
280 selfMult1_ = 0.0;
281 selfMult2_ = 0.0;
282 selfMult4_ = 0.0;
283 }
284
285 // higher derivatives of B_0 at the cutoff radius:
286 db0c_1 = -r * b1c;
287 db0c_2 = -b1c + r2 * b2c;
288 db0c_3 = 3.0*r*b2c - r2*r*b3c;
289 db0c_4 = 3.0*b2c - 6.0*r2*b3c + r2*r2*b4c;
290 db0c_5 = -15.0*r*b3c + 10.0*r2*r*b4c - r2*r2*r*b5c;
291
292 selfMult1_ -= b0c;
293 selfMult2_ += (db0c_2 + 2.0*db0c_1*ric) / 3.0;
294 selfMult4_ -= (db0c_4 + 4.0*db0c_3*ric) / 15.0;
295
296 // working variables for the splines:
297 RealType ri, ri2;
298 RealType b0, b1, b2, b3, b4, b5;
299 RealType db0_1, db0_2, db0_3, db0_4, db0_5;
300 RealType f, fc, f0;
301 RealType g, gc, g0, g1, g2, g3, g4;
302 RealType h, hc, h1, h2, h3, h4;
303 RealType s, sc, s2, s3, s4;
304 RealType t, tc, t3, t4;
305 RealType u, uc, u4;
306
307 // working variables for Taylor expansion:
308 RealType rmRc, rmRc2, rmRc3, rmRc4;
309
310 // Approximate using splines using a maximum of 0.1 Angstroms
311 // between points.
312 int nptest = int((cutoffRadius_ + 2.0) / 0.1);
313 np_ = (np_ > nptest) ? np_ : nptest;
314
315 // Add a 2 angstrom safety window to deal with cutoffGroups that
316 // have charged atoms longer than the cutoffRadius away from each
317 // other. Splining is almost certainly the best choice here.
318 // Direct calls to erfc would be preferrable if it is a very fast
319 // implementation.
320
321 RealType dx = (cutoffRadius_ + 2.0) / RealType(np_);
322
323 // Storage vectors for the computed functions
324 vector<RealType> rv;
325 vector<RealType> v01v;
326 vector<RealType> v11v;
327 vector<RealType> v21v, v22v;
328 vector<RealType> v31v, v32v;
329 vector<RealType> v41v, v42v, v43v;
330
331 /*
332 vector<RealType> dv01v;
333 vector<RealType> dv11v;
334 vector<RealType> dv21v, dv22v;
335 vector<RealType> dv31v, dv32v;
336 vector<RealType> dv41v, dv42v, dv43v;
337 */
338
339 for (int i = 1; i < np_ + 1; i++) {
340 r = RealType(i) * dx;
341 rv.push_back(r);
342
343 ri = 1.0 / r;
344 ri2 = ri * ri;
345
346 r2 = r * r;
347 expTerm = exp(-a2 * r2);
348
349 // Taylor expansion factors (no need for factorials this way):
350 rmRc = r - cutoffRadius_;
351 rmRc2 = rmRc * rmRc / 2.0;
352 rmRc3 = rmRc2 * rmRc / 3.0;
353 rmRc4 = rmRc3 * rmRc / 4.0;
354
355 // values of Smith's B_l functions at r:
356 if (screeningMethod_ == DAMPED) {
357 b0 = erfc(dampingAlpha_ * r) * ri;
358 b1 = ( b0 + 2.0*a2 * expTerm * invArootPi) * ri2;
359 b2 = (3.0 * b1 + pow(2.0*a2, 2) * expTerm * invArootPi) * ri2;
360 b3 = (5.0 * b2 + pow(2.0*a2, 3) * expTerm * invArootPi) * ri2;
361 b4 = (7.0 * b3 + pow(2.0*a2, 4) * expTerm * invArootPi) * ri2;
362 b5 = (9.0 * b4 + pow(2.0*a2, 5) * expTerm * invArootPi) * ri2;
363 } else {
364 b0 = ri;
365 b1 = ( b0) * ri2;
366 b2 = (3.0 * b1) * ri2;
367 b3 = (5.0 * b2) * ri2;
368 b4 = (7.0 * b3) * ri2;
369 b5 = (9.0 * b4) * ri2;
370 }
371
372 // higher derivatives of B_0 at r:
373 db0_1 = -r * b1;
374 db0_2 = -b1 + r2 * b2;
375 db0_3 = 3.0*r*b2 - r2*r*b3;
376 db0_4 = 3.0*b2 - 6.0*r2*b3 + r2*r2*b4;
377 db0_5 = -15.0*r*b3 + 10.0*r2*r*b4 - r2*r2*r*b5;
378
379 f = b0;
380 fc = b0c;
381 f0 = f - fc - rmRc*db0c_1;
382
383 g = db0_1;
384 gc = db0c_1;
385 g0 = g - gc;
386 g1 = g0 - rmRc *db0c_2;
387 g2 = g1 - rmRc2*db0c_3;
388 g3 = g2 - rmRc3*db0c_4;
389 g4 = g3 - rmRc4*db0c_5;
390
391 h = db0_2;
392 hc = db0c_2;
393 h1 = h - hc;
394 h2 = h1 - rmRc *db0c_3;
395 h3 = h2 - rmRc2*db0c_4;
396 h4 = h3 - rmRc3*db0c_5;
397
398 s = db0_3;
399 sc = db0c_3;
400 s2 = s - sc;
401 s3 = s2 - rmRc *db0c_4;
402 s4 = s3 - rmRc2*db0c_5;
403
404 t = db0_4;
405 tc = db0c_4;
406 t3 = t - tc;
407 t4 = t3 - rmRc *db0c_5;
408
409 u = db0_5;
410 uc = db0c_5;
411 u4 = u - uc;
412
413 // in what follows below, the various v functions are used for
414 // potentials and torques, while the w functions show up in the
415 // forces.
416
417 switch (summationMethod_) {
418 case esm_SHIFTED_FORCE:
419
420 v01 = f - fc - rmRc*gc;
421 v11 = g - gc - rmRc*hc;
422 v21 = g*ri - gc*ric - rmRc*(hc - gc*ric)*ric;
423 v22 = h - g*ri - (hc - gc*ric) - rmRc*(sc - (hc - gc*ric)*ric);
424 v31 = (h-g*ri)*ri - (hc-gc*ric)*ric - rmRc*(sc-2.0*(hc-gc*ric)*ric)*ric;
425 v32 = (s - 3.0*(h-g*ri)*ri) - (sc - 3.0*(hc-gc*ric)*ric)
426 - rmRc*(tc - 3.0*(sc-2.0*(hc-gc*ric)*ric)*ric);
427 v41 = (h - g*ri)*ri2 - (hc - gc*ric)*ric2
428 - rmRc*(sc - 3.0*(hc-gc*ric)*ric)*ric2;
429 v42 = (s-3.0*(h-g*ri)*ri)*ri - (sc-3.0*(hc-gc*ric)*ric)*ric
430 - rmRc*(tc - (4.0*sc - 9.0*(hc - gc*ric)*ric)*ric)*ric;
431
432 v43 = (t - 3.0*(2.0*s - 5.0*(h - g*ri)*ri)*ri)
433 - (tc - 3.0*(2.0*sc - 5.0*(hc - gc*ric)*ric)*ric)
434 - rmRc*(uc-3.0*(2.0*tc - (7.0*sc - 15.0*(hc - gc*ric)*ric)*ric)*ric);
435
436 dv01 = g - gc;
437 dv11 = h - hc;
438 dv21 = (h - g*ri)*ri - (hc - gc*ric)*ric;
439 dv22 = (s - (h - g*ri)*ri) - (sc - (hc - gc*ric)*ric);
440 dv31 = (s - 2.0*(h-g*ri)*ri)*ri - (sc - 2.0*(hc-gc*ric)*ric)*ric;
441 dv32 = (t - 3.0*(s-2.0*(h-g*ri)*ri)*ri)
442 - (tc - 3.0*(sc-2.0*(hc-gc*ric)*ric)*ric);
443 dv41 = (s - 3.0*(h - g*ri)*ri)*ri2 - (sc - 3.0*(hc - gc*ric)*ric)*ric2;
444 dv42 = (t - (4.0*s - 9.0*(h-g*ri)*ri)*ri)*ri
445 - (tc - (4.0*sc - 9.0*(hc-gc*ric)*ric)*ric)*ric;
446 dv43 = (u - 3.0*(2.0*t - (7.0*s - 15.0*(h - g*ri)*ri)*ri)*ri)
447 - (uc - 3.0*(2.0*tc - (7.0*sc - 15.0*(hc - gc*ric)*ric)*ric)*ric);
448
449 break;
450
451 case esm_TAYLOR_SHIFTED:
452
453 v01 = f0;
454 v11 = g1;
455 v21 = g2 * ri;
456 v22 = h2 - v21;
457 v31 = (h3 - g3 * ri) * ri;
458 v32 = s3 - 3.0*v31;
459 v41 = (h4 - g4 * ri) * ri2;
460 v42 = s4 * ri - 3.0*v41;
461 v43 = t4 - 6.0*v42 - 3.0*v41;
462
463 dv01 = g0;
464 dv11 = h1;
465 dv21 = (h2 - g2*ri)*ri;
466 dv22 = (s2 - (h2 - g2*ri)*ri);
467 dv31 = (s3 - 2.0*(h3-g3*ri)*ri)*ri;
468 dv32 = (t3 - 3.0*(s3-2.0*(h3-g3*ri)*ri)*ri);
469 dv41 = (s4 - 3.0*(h4 - g4*ri)*ri)*ri2;
470 dv42 = (t4 - (4.0*s4 - 9.0*(h4-g4*ri)*ri)*ri)*ri;
471 dv43 = (u4 - 3.0*(2.0*t4 - (7.0*s4 - 15.0*(h4 - g4*ri)*ri)*ri)*ri);
472
473 break;
474
475 case esm_SHIFTED_POTENTIAL:
476
477 v01 = f - fc;
478 v11 = g - gc;
479 v21 = g*ri - gc*ric;
480 v22 = h - g*ri - (hc - gc*ric);
481 v31 = (h-g*ri)*ri - (hc-gc*ric)*ric;
482 v32 = (s - 3.0*(h-g*ri)*ri) - (sc - 3.0*(hc-gc*ric)*ric);
483 v41 = (h - g*ri)*ri2 - (hc - gc*ric)*ric2;
484 v42 = (s-3.0*(h-g*ri)*ri)*ri - (sc-3.0*(hc-gc*ric)*ric)*ric;
485 v43 = (t - 3.0*(2.0*s - 5.0*(h - g*ri)*ri)*ri)
486 - (tc - 3.0*(2.0*sc - 5.0*(hc - gc*ric)*ric)*ric);
487
488 dv01 = g;
489 dv11 = h;
490 dv21 = (h - g*ri)*ri;
491 dv22 = (s - (h - g*ri)*ri);
492 dv31 = (s - 2.0*(h-g*ri)*ri)*ri;
493 dv32 = (t - 3.0*(s-2.0*(h-g*ri)*ri)*ri);
494 dv41 = (s - 3.0*(h - g*ri)*ri)*ri2;
495 dv42 = (t - (4.0*s - 9.0*(h-g*ri)*ri)*ri)*ri;
496 dv43 = (u - 3.0*(2.0*t - (7.0*s - 15.0*(h - g*ri)*ri)*ri)*ri);
497
498 break;
499
500 case esm_SWITCHING_FUNCTION:
501 case esm_HARD:
502
503 v01 = f;
504 v11 = g;
505 v21 = g*ri;
506 v22 = h - g*ri;
507 v31 = (h-g*ri)*ri;
508 v32 = (s - 3.0*(h-g*ri)*ri);
509 v41 = (h - g*ri)*ri2;
510 v42 = (s-3.0*(h-g*ri)*ri)*ri;
511 v43 = (t - 3.0*(2.0*s - 5.0*(h - g*ri)*ri)*ri);
512
513 dv01 = g;
514 dv11 = h;
515 dv21 = (h - g*ri)*ri;
516 dv22 = (s - (h - g*ri)*ri);
517 dv31 = (s - 2.0*(h-g*ri)*ri)*ri;
518 dv32 = (t - 3.0*(s-2.0*(h-g*ri)*ri)*ri);
519 dv41 = (s - 3.0*(h - g*ri)*ri)*ri2;
520 dv42 = (t - (4.0*s - 9.0*(h-g*ri)*ri)*ri)*ri;
521 dv43 = (u - 3.0*(2.0*t - (7.0*s - 15.0*(h - g*ri)*ri)*ri)*ri);
522
523 break;
524
525 case esm_REACTION_FIELD:
526
527 // following DL_POLY's lead for shifting the image charge potential:
528 f = b0 + preRF_ * r2;
529 fc = b0c + preRF_ * cutoffRadius_ * cutoffRadius_;
530
531 g = db0_1 + preRF_ * 2.0 * r;
532 gc = db0c_1 + preRF_ * 2.0 * cutoffRadius_;
533
534 h = db0_2 + preRF_ * 2.0;
535 hc = db0c_2 + preRF_ * 2.0;
536
537 v01 = f - fc;
538 v11 = g - gc;
539 v21 = g*ri - gc*ric;
540 v22 = h - g*ri - (hc - gc*ric);
541 v31 = (h-g*ri)*ri - (hc-gc*ric)*ric;
542 v32 = (s - 3.0*(h-g*ri)*ri) - (sc - 3.0*(hc-gc*ric)*ric);
543 v41 = (h - g*ri)*ri2 - (hc - gc*ric)*ric2;
544 v42 = (s-3.0*(h-g*ri)*ri)*ri - (sc-3.0*(hc-gc*ric)*ric)*ric;
545 v43 = (t - 3.0*(2.0*s - 5.0*(h - g*ri)*ri)*ri)
546 - (tc - 3.0*(2.0*sc - 5.0*(hc - gc*ric)*ric)*ric);
547
548 dv01 = g;
549 dv11 = h;
550 dv21 = (h - g*ri)*ri;
551 dv22 = (s - (h - g*ri)*ri);
552 dv31 = (s - 2.0*(h-g*ri)*ri)*ri;
553 dv32 = (t - 3.0*(s-2.0*(h-g*ri)*ri)*ri);
554 dv41 = (s - 3.0*(h - g*ri)*ri)*ri2;
555 dv42 = (t - (4.0*s - 9.0*(h-g*ri)*ri)*ri)*ri;
556 dv43 = (u - 3.0*(2.0*t - (7.0*s - 15.0*(h - g*ri)*ri)*ri)*ri);
557
558 break;
559
560 case esm_EWALD_FULL:
561 case esm_EWALD_PME:
562 case esm_EWALD_SPME:
563 default :
564 map<string, ElectrostaticSummationMethod>::iterator i;
565 std::string meth;
566 for (i = summationMap_.begin(); i != summationMap_.end(); ++i) {
567 if ((*i).second == summationMethod_) meth = (*i).first;
568 }
569 sprintf( painCave.errMsg,
570 "Electrostatic::initialize: electrostaticSummationMethod %s \n"
571 "\thas not been implemented yet. Please select one of:\n"
572 "\t\"hard\", \"shifted_potential\", or \"shifted_force\"\n",
573 meth.c_str() );
574 painCave.isFatal = 1;
575 simError();
576 break;
577 }
578
579 // Add these computed values to the storage vectors for spline creation:
580 v01v.push_back(v01);
581 v11v.push_back(v11);
582 v21v.push_back(v21);
583 v22v.push_back(v22);
584 v31v.push_back(v31);
585 v32v.push_back(v32);
586 v41v.push_back(v41);
587 v42v.push_back(v42);
588 v43v.push_back(v43);
589 /*
590 dv01v.push_back(dv01);
591 dv11v.push_back(dv11);
592 dv21v.push_back(dv21);
593 dv22v.push_back(dv22);
594 dv31v.push_back(dv31);
595 dv32v.push_back(dv32);
596 dv41v.push_back(dv41);
597 dv42v.push_back(dv42);
598 dv43v.push_back(dv43);
599 */
600 }
601
602 // construct the spline structures and fill them with the values we've
603 // computed:
604
605 v01s = new CubicSpline();
606 v01s->addPoints(rv, v01v);
607 v11s = new CubicSpline();
608 v11s->addPoints(rv, v11v);
609 v21s = new CubicSpline();
610 v21s->addPoints(rv, v21v);
611 v22s = new CubicSpline();
612 v22s->addPoints(rv, v22v);
613 v31s = new CubicSpline();
614 v31s->addPoints(rv, v31v);
615 v32s = new CubicSpline();
616 v32s->addPoints(rv, v32v);
617 v41s = new CubicSpline();
618 v41s->addPoints(rv, v41v);
619 v42s = new CubicSpline();
620 v42s->addPoints(rv, v42v);
621 v43s = new CubicSpline();
622 v43s->addPoints(rv, v43v);
623
624 /*
625 dv01s = new CubicSpline();
626 dv01s->addPoints(rv, dv01v);
627 dv11s = new CubicSpline();
628 dv11s->addPoints(rv, dv11v);
629 dv21s = new CubicSpline();
630 dv21s->addPoints(rv, dv21v);
631 dv22s = new CubicSpline();
632 dv22s->addPoints(rv, dv22v);
633 dv31s = new CubicSpline();
634 dv31s->addPoints(rv, dv31v);
635 dv32s = new CubicSpline();
636 dv32s->addPoints(rv, dv32v);
637 dv41s = new CubicSpline();
638 dv41s->addPoints(rv, dv41v);
639 dv42s = new CubicSpline();
640 dv42s->addPoints(rv, dv42v);
641 dv43s = new CubicSpline();
642 dv43s->addPoints(rv, dv43v);
643 */
644
645 haveElectroSplines_ = true;
646
647 initialized_ = true;
648 }
649
650 void Electrostatic::addType(AtomType* atomType){
651
652 ElectrostaticAtomData electrostaticAtomData;
653 electrostaticAtomData.is_Charge = false;
654 electrostaticAtomData.is_Dipole = false;
655 electrostaticAtomData.is_Quadrupole = false;
656 electrostaticAtomData.is_Fluctuating = false;
657
658 FixedChargeAdapter fca = FixedChargeAdapter(atomType);
659
660 if (fca.isFixedCharge()) {
661 electrostaticAtomData.is_Charge = true;
662 electrostaticAtomData.fixedCharge = fca.getCharge();
663 }
664
665 MultipoleAdapter ma = MultipoleAdapter(atomType);
666 if (ma.isMultipole()) {
667 if (ma.isDipole()) {
668 electrostaticAtomData.is_Dipole = true;
669 electrostaticAtomData.dipole = ma.getDipole();
670 }
671 if (ma.isQuadrupole()) {
672 electrostaticAtomData.is_Quadrupole = true;
673 electrostaticAtomData.quadrupole = ma.getQuadrupole();
674 }
675 }
676
677 FluctuatingChargeAdapter fqa = FluctuatingChargeAdapter(atomType);
678
679 if (fqa.isFluctuatingCharge()) {
680 electrostaticAtomData.is_Fluctuating = true;
681 electrostaticAtomData.electronegativity = fqa.getElectronegativity();
682 electrostaticAtomData.hardness = fqa.getHardness();
683 electrostaticAtomData.slaterN = fqa.getSlaterN();
684 electrostaticAtomData.slaterZeta = fqa.getSlaterZeta();
685 }
686
687 int atid = atomType->getIdent();
688 int etid = Etypes.size();
689 int fqtid = FQtypes.size();
690
691 pair<set<int>::iterator,bool> ret;
692 ret = Etypes.insert( atid );
693 if (ret.second == false) {
694 sprintf( painCave.errMsg,
695 "Electrostatic already had a previous entry with ident %d\n",
696 atid);
697 painCave.severity = OPENMD_INFO;
698 painCave.isFatal = 0;
699 simError();
700 }
701
702 Etids[ atid ] = etid;
703 ElectrostaticMap.push_back(electrostaticAtomData);
704
705 if (electrostaticAtomData.is_Fluctuating) {
706 ret = FQtypes.insert( atid );
707 if (ret.second == false) {
708 sprintf( painCave.errMsg,
709 "Electrostatic already had a previous fluctuating charge entry with ident %d\n",
710 atid );
711 painCave.severity = OPENMD_INFO;
712 painCave.isFatal = 0;
713 simError();
714 }
715 FQtids[atid] = fqtid;
716 Jij[fqtid].resize(nFlucq_);
717
718 // Now, iterate over all known fluctuating and add to the coulomb integral map:
719
720 std::set<int>::iterator it;
721 for( it = FQtypes.begin(); it != FQtypes.end(); ++it) {
722 int etid2 = Etids[ (*it) ];
723 int fqtid2 = FQtids[ (*it) ];
724 ElectrostaticAtomData eaData2 = ElectrostaticMap[ etid2 ];
725 RealType a = electrostaticAtomData.slaterZeta;
726 RealType b = eaData2.slaterZeta;
727 int m = electrostaticAtomData.slaterN;
728 int n = eaData2.slaterN;
729
730 // Create the spline of the coulombic integral for s-type
731 // Slater orbitals. Add a 2 angstrom safety window to deal
732 // with cutoffGroups that have charged atoms longer than the
733 // cutoffRadius away from each other.
734
735 RealType rval;
736 RealType dr = (cutoffRadius_ + 2.0) / RealType(np_ - 1);
737 vector<RealType> rvals;
738 vector<RealType> Jvals;
739 // don't start at i = 0, as rval = 0 is undefined for the
740 // slater overlap integrals.
741 for (int i = 1; i < np_+1; i++) {
742 rval = RealType(i) * dr;
743 rvals.push_back(rval);
744 Jvals.push_back(sSTOCoulInt( a, b, m, n, rval *
745 PhysicalConstants::angstromToBohr ) *
746 PhysicalConstants::hartreeToKcal );
747 }
748
749 CubicSpline* J = new CubicSpline();
750 J->addPoints(rvals, Jvals);
751 Jij[fqtid][fqtid2] = J;
752 Jij[fqtid2].resize( nFlucq_ );
753 Jij[fqtid2][fqtid] = J;
754 }
755 }
756 return;
757 }
758
759 void Electrostatic::setCutoffRadius( RealType rCut ) {
760 cutoffRadius_ = rCut;
761 haveCutoffRadius_ = true;
762 }
763
764 void Electrostatic::setElectrostaticSummationMethod( ElectrostaticSummationMethod esm ) {
765 summationMethod_ = esm;
766 }
767 void Electrostatic::setElectrostaticScreeningMethod( ElectrostaticScreeningMethod sm ) {
768 screeningMethod_ = sm;
769 }
770 void Electrostatic::setDampingAlpha( RealType alpha ) {
771 dampingAlpha_ = alpha;
772 haveDampingAlpha_ = true;
773 }
774 void Electrostatic::setReactionFieldDielectric( RealType dielectric ){
775 dielectric_ = dielectric;
776 haveDielectric_ = true;
777 }
778
779 void Electrostatic::calcForce(InteractionData &idat) {
780
781 if (!initialized_) initialize();
782
783 data1 = ElectrostaticMap[Etids[idat.atid1]];
784 data2 = ElectrostaticMap[Etids[idat.atid2]];
785
786 U = 0.0; // Potential
787 F.zero(); // Force
788 Ta.zero(); // Torque on site a
789 Tb.zero(); // Torque on site b
790 Ea.zero(); // Electric field at site a
791 Eb.zero(); // Electric field at site b
792 dUdCa = 0.0; // fluctuating charge force at site a
793 dUdCb = 0.0; // fluctuating charge force at site a
794
795 // Indirect interactions mediated by the reaction field.
796 indirect_Pot = 0.0; // Potential
797 indirect_F.zero(); // Force
798 indirect_Ta.zero(); // Torque on site a
799 indirect_Tb.zero(); // Torque on site b
800
801 // Excluded potential that is still computed for fluctuating charges
802 excluded_Pot= 0.0;
803
804
805 // some variables we'll need independent of electrostatic type:
806
807 ri = 1.0 / *(idat.rij);
808 rhat = *(idat.d) * ri;
809
810 // logicals
811
812 a_is_Charge = data1.is_Charge;
813 a_is_Dipole = data1.is_Dipole;
814 a_is_Quadrupole = data1.is_Quadrupole;
815 a_is_Fluctuating = data1.is_Fluctuating;
816
817 b_is_Charge = data2.is_Charge;
818 b_is_Dipole = data2.is_Dipole;
819 b_is_Quadrupole = data2.is_Quadrupole;
820 b_is_Fluctuating = data2.is_Fluctuating;
821
822 // Obtain all of the required radial function values from the
823 // spline structures:
824
825 // needed for fields (and forces):
826 if (a_is_Charge || b_is_Charge) {
827 v01s->getValueAndDerivativeAt( *(idat.rij), v01, dv01);
828 }
829 if (a_is_Dipole || b_is_Dipole) {
830 v11s->getValueAndDerivativeAt( *(idat.rij), v11, dv11);
831 v11or = ri * v11;
832 }
833 if (a_is_Quadrupole || b_is_Quadrupole || (a_is_Dipole && b_is_Dipole)) {
834 v21s->getValueAndDerivativeAt( *(idat.rij), v21, dv21);
835 v22s->getValueAndDerivativeAt( *(idat.rij), v22, dv22);
836 v22or = ri * v22;
837 }
838
839 // needed for potentials (and forces and torques):
840 if ((a_is_Dipole && b_is_Quadrupole) ||
841 (b_is_Dipole && a_is_Quadrupole)) {
842 v31s->getValueAndDerivativeAt( *(idat.rij), v31, dv31);
843 v32s->getValueAndDerivativeAt( *(idat.rij), v32, dv32);
844 v31or = v31 * ri;
845 v32or = v32 * ri;
846 }
847 if (a_is_Quadrupole && b_is_Quadrupole) {
848 v41s->getValueAndDerivativeAt( *(idat.rij), v41, dv41);
849 v42s->getValueAndDerivativeAt( *(idat.rij), v42, dv42);
850 v43s->getValueAndDerivativeAt( *(idat.rij), v43, dv43);
851 v42or = v42 * ri;
852 v43or = v43 * ri;
853 }
854
855 // calculate the single-site contributions (fields, etc).
856
857 if (a_is_Charge) {
858 C_a = data1.fixedCharge;
859
860 if (a_is_Fluctuating) {
861 C_a += *(idat.flucQ1);
862 }
863
864 if (idat.excluded) {
865 *(idat.skippedCharge2) += C_a;
866 } else {
867 // only do the field if we're not excluded:
868 Eb -= C_a * pre11_ * dv01 * rhat;
869 }
870 }
871
872 if (a_is_Dipole) {
873 D_a = *(idat.dipole1);
874 rdDa = dot(rhat, D_a);
875 rxDa = cross(rhat, D_a);
876 if (!idat.excluded)
877 Eb -= pre12_ * ((dv11-v11or) * rdDa * rhat + v11or * D_a);
878 }
879
880 if (a_is_Quadrupole) {
881 Q_a = *(idat.quadrupole1);
882 trQa = Q_a.trace();
883 Qar = Q_a * rhat;
884 rQa = rhat * Q_a;
885 rdQar = dot(rhat, Qar);
886 rxQar = cross(rhat, Qar);
887 if (!idat.excluded)
888 Eb -= pre14_ * (trQa * rhat * dv21 + 2.0 * Qar * v22or
889 + rdQar * rhat * (dv22 - 2.0*v22or));
890 }
891
892 if (b_is_Charge) {
893 C_b = data2.fixedCharge;
894
895 if (b_is_Fluctuating)
896 C_b += *(idat.flucQ2);
897
898 if (idat.excluded) {
899 *(idat.skippedCharge1) += C_b;
900 } else {
901 // only do the field if we're not excluded:
902 Ea += C_b * pre11_ * dv01 * rhat;
903 }
904 }
905
906 if (b_is_Dipole) {
907 D_b = *(idat.dipole2);
908 rdDb = dot(rhat, D_b);
909 rxDb = cross(rhat, D_b);
910 if (!idat.excluded)
911 Ea += pre12_ * ((dv11-v11or) * rdDb * rhat + v11or * D_b);
912 }
913
914 if (b_is_Quadrupole) {
915 Q_b = *(idat.quadrupole2);
916 trQb = Q_b.trace();
917 Qbr = Q_b * rhat;
918 rQb = rhat * Q_b;
919 rdQbr = dot(rhat, Qbr);
920 rxQbr = cross(rhat, Qbr);
921 if (!idat.excluded)
922 Ea += pre14_ * (trQb * rhat * dv21 + 2.0 * Qbr * v22or
923 + rdQbr * rhat * (dv22 - 2.0*v22or));
924 }
925
926 if ((a_is_Fluctuating || b_is_Fluctuating) && idat.excluded) {
927 J = Jij[FQtids[idat.atid1]][FQtids[idat.atid2]];
928 }
929
930 if (a_is_Charge) {
931
932 if (b_is_Charge) {
933 pref = pre11_ * *(idat.electroMult);
934 U += C_a * C_b * pref * v01;
935 F += C_a * C_b * pref * dv01 * rhat;
936
937 // If this is an excluded pair, there are still indirect
938 // interactions via the reaction field we must worry about:
939
940 if (summationMethod_ == esm_REACTION_FIELD && idat.excluded) {
941 rfContrib = preRF_ * pref * C_a * C_b * *(idat.r2);
942 indirect_Pot += rfContrib;
943 indirect_F += rfContrib * 2.0 * ri * rhat;
944 }
945
946 // Fluctuating charge forces are handled via Coulomb integrals
947 // for excluded pairs (i.e. those connected via bonds) and
948 // with the standard charge-charge interaction otherwise.
949
950 if (idat.excluded) {
951 if (a_is_Fluctuating || b_is_Fluctuating) {
952 coulInt = J->getValueAt( *(idat.rij) );
953 if (a_is_Fluctuating) dUdCa += coulInt * C_b;
954 if (b_is_Fluctuating) dUdCb += coulInt * C_a;
955 excluded_Pot += C_a * C_b * coulInt;
956 }
957 } else {
958 if (a_is_Fluctuating) dUdCa += C_b * pref * v01;
959 if (a_is_Fluctuating) dUdCb += C_a * pref * v01;
960 }
961 }
962
963 if (b_is_Dipole) {
964 pref = pre12_ * *(idat.electroMult);
965 U += C_a * pref * v11 * rdDb;
966 F += C_a * pref * ((dv11 - v11or) * rdDb * rhat + v11or * D_b);
967 Tb += C_a * pref * v11 * rxDb;
968
969 if (a_is_Fluctuating) dUdCa += pref * v11 * rdDb;
970
971 // Even if we excluded this pair from direct interactions, we
972 // still have the reaction-field-mediated charge-dipole
973 // interaction:
974
975 if (summationMethod_ == esm_REACTION_FIELD && idat.excluded) {
976 rfContrib = C_a * pref * preRF_ * 2.0 * *(idat.rij);
977 indirect_Pot += rfContrib * rdDb;
978 indirect_F += rfContrib * D_b / (*idat.rij);
979 indirect_Tb += C_a * pref * preRF_ * rxDb;
980 }
981 }
982
983 if (b_is_Quadrupole) {
984 pref = pre14_ * *(idat.electroMult);
985 U += C_a * pref * (v21 * trQb + v22 * rdQbr);
986 F += C_a * pref * (trQb * dv21 * rhat + 2.0 * Qbr * v22or);
987 F += C_a * pref * rdQbr * rhat * (dv22 - 2.0*v22or);
988 Tb += C_a * pref * 2.0 * rxQbr * v22;
989
990 if (a_is_Fluctuating) dUdCa += pref * (v21 * trQb + v22 * rdQbr);
991 }
992 }
993
994 if (a_is_Dipole) {
995
996 if (b_is_Charge) {
997 pref = pre12_ * *(idat.electroMult);
998
999 U -= C_b * pref * v11 * rdDa;
1000 F -= C_b * pref * ((dv11-v11or) * rdDa * rhat + v11or * D_a);
1001 Ta -= C_b * pref * v11 * rxDa;
1002
1003 if (b_is_Fluctuating) dUdCb -= pref * v11 * rdDa;
1004
1005 // Even if we excluded this pair from direct interactions,
1006 // we still have the reaction-field-mediated charge-dipole
1007 // interaction:
1008 if (summationMethod_ == esm_REACTION_FIELD && idat.excluded) {
1009 rfContrib = C_b * pref * preRF_ * 2.0 * *(idat.rij);
1010 indirect_Pot -= rfContrib * rdDa;
1011 indirect_F -= rfContrib * D_a / (*idat.rij);
1012 indirect_Ta -= C_b * pref * preRF_ * rxDa;
1013 }
1014 }
1015
1016 if (b_is_Dipole) {
1017 pref = pre22_ * *(idat.electroMult);
1018 DadDb = dot(D_a, D_b);
1019 DaxDb = cross(D_a, D_b);
1020
1021 U -= pref * (DadDb * v21 + rdDa * rdDb * v22);
1022 F -= pref * (dv21 * DadDb * rhat + v22or * (rdDb * D_a + rdDa * D_b));
1023 F -= pref * (rdDa * rdDb) * (dv22 - 2.0*v22or) * rhat;
1024 Ta += pref * ( v21 * DaxDb - v22 * rdDb * rxDa);
1025 Tb += pref * (-v21 * DaxDb - v22 * rdDa * rxDb);
1026
1027 // Even if we excluded this pair from direct interactions, we
1028 // still have the reaction-field-mediated dipole-dipole
1029 // interaction:
1030 if (summationMethod_ == esm_REACTION_FIELD && idat.excluded) {
1031 rfContrib = -pref * preRF_ * 2.0;
1032 indirect_Pot += rfContrib * DadDb;
1033 indirect_Ta += rfContrib * DaxDb;
1034 indirect_Tb -= rfContrib * DaxDb;
1035 }
1036 }
1037
1038 if (b_is_Quadrupole) {
1039 pref = pre24_ * *(idat.electroMult);
1040 DadQb = D_a * Q_b;
1041 DadQbr = dot(D_a, Qbr);
1042 DaxQbr = cross(D_a, Qbr);
1043
1044 U -= pref * ((trQb*rdDa + 2.0*DadQbr)*v31 + rdDa*rdQbr*v32);
1045 F -= pref * (trQb*D_a + 2.0*DadQb) * v31or;
1046 F -= pref * (trQb*rdDa + 2.0*DadQbr) * (dv31-v31or) * rhat;
1047 F -= pref * (D_a*rdQbr + 2.0*rdDa*rQb) * v32or;
1048 F -= pref * (rdDa * rdQbr * rhat * (dv32-3.0*v32or));
1049 Ta += pref * ((-trQb*rxDa + 2.0 * DaxQbr)*v31 - rxDa*rdQbr*v32);
1050 Tb += pref * ((2.0*cross(DadQb, rhat) - 2.0*DaxQbr)*v31
1051 - 2.0*rdDa*rxQbr*v32);
1052 }
1053 }
1054
1055 if (a_is_Quadrupole) {
1056 if (b_is_Charge) {
1057 pref = pre14_ * *(idat.electroMult);
1058 U += C_b * pref * (v21 * trQa + v22 * rdQar);
1059 F += C_b * pref * (trQa * rhat * dv21 + 2.0 * Qar * v22or);
1060 F += C_b * pref * rdQar * rhat * (dv22 - 2.0*v22or);
1061 Ta += C_b * pref * 2.0 * rxQar * v22;
1062
1063 if (b_is_Fluctuating) dUdCb += pref * (v21 * trQa + v22 * rdQar);
1064 }
1065 if (b_is_Dipole) {
1066 pref = pre24_ * *(idat.electroMult);
1067 DbdQa = D_b * Q_a;
1068 DbdQar = dot(D_b, Qar);
1069 DbxQar = cross(D_b, Qar);
1070
1071 U += pref * ((trQa*rdDb + 2.0*DbdQar)*v31 + rdDb*rdQar*v32);
1072 F += pref * (trQa*D_b + 2.0*DbdQa) * v31or;
1073 F += pref * (trQa*rdDb + 2.0*DbdQar) * (dv31-v31or) * rhat;
1074 F += pref * (D_b*rdQar + 2.0*rdDb*rQa) * v32or;
1075 F += pref * (rdDb * rdQar * rhat * (dv32-3.0*v32or));
1076 Ta += pref * ((-2.0*cross(DbdQa, rhat) + 2.0*DbxQar)*v31
1077 + 2.0*rdDb*rxQar*v32);
1078 Tb += pref * ((trQa*rxDb - 2.0 * DbxQar)*v31 + rxDb*rdQar*v32);
1079 }
1080 if (b_is_Quadrupole) {
1081 pref = pre44_ * *(idat.electroMult); // yes
1082 QaQb = Q_a * Q_b;
1083 trQaQb = QaQb.trace();
1084 rQaQb = rhat * QaQb;
1085 QaQbr = QaQb * rhat;
1086 QaxQb = cross(Q_a, Q_b);
1087 rQaQbr = dot(rQa, Qbr);
1088 rQaxQbr = cross(rQa, Qbr);
1089
1090 U += pref * (trQa * trQb + 2.0 * trQaQb) * v41;
1091 U += pref * (trQa * rdQbr + trQb * rdQar + 4.0 * rQaQbr) * v42;
1092 U += pref * (rdQar * rdQbr) * v43;
1093
1094 F += pref * rhat * (trQa * trQb + 2.0 * trQaQb)*dv41;
1095 F += pref*rhat*(trQa*rdQbr + trQb*rdQar + 4.0*rQaQbr)*(dv42-2.0*v42or);
1096 F += pref * rhat * (rdQar * rdQbr)*(dv43 - 4.0*v43or);
1097
1098 F += pref * 2.0 * (trQb*rQa + trQa*rQb) * v42or;
1099 F += pref * 4.0 * (rQaQb + QaQbr) * v42or;
1100 F += pref * 2.0 * (rQa*rdQbr + rdQar*rQb) * v43or;
1101
1102 Ta += pref * (- 4.0 * QaxQb * v41);
1103 Ta += pref * (- 2.0 * trQb * cross(rQa, rhat)
1104 + 4.0 * cross(rhat, QaQbr)
1105 - 4.0 * rQaxQbr) * v42;
1106 Ta += pref * 2.0 * cross(rhat,Qar) * rdQbr * v43;
1107
1108
1109 Tb += pref * (+ 4.0 * QaxQb * v41);
1110 Tb += pref * (- 2.0 * trQa * cross(rQb, rhat)
1111 - 4.0 * cross(rQaQb, rhat)
1112 + 4.0 * rQaxQbr) * v42;
1113 // Possible replacement for line 2 above:
1114 // + 4.0 * cross(rhat, QbQar)
1115
1116 Tb += pref * 2.0 * cross(rhat,Qbr) * rdQar * v43;
1117
1118 }
1119 }
1120
1121 if (idat.doElectricField) {
1122 *(idat.eField1) += Ea * *(idat.electroMult);
1123 *(idat.eField2) += Eb * *(idat.electroMult);
1124 }
1125
1126 if (a_is_Fluctuating) *(idat.dVdFQ1) += dUdCa * *(idat.sw);
1127 if (b_is_Fluctuating) *(idat.dVdFQ2) += dUdCb * *(idat.sw);
1128
1129 if (!idat.excluded) {
1130
1131 *(idat.vpair) += U;
1132 (*(idat.pot))[ELECTROSTATIC_FAMILY] += U * *(idat.sw);
1133 *(idat.f1) += F * *(idat.sw);
1134
1135 if (a_is_Dipole || a_is_Quadrupole)
1136 *(idat.t1) += Ta * *(idat.sw);
1137
1138 if (b_is_Dipole || b_is_Quadrupole)
1139 *(idat.t2) += Tb * *(idat.sw);
1140
1141 } else {
1142
1143 // only accumulate the forces and torques resulting from the
1144 // indirect reaction field terms.
1145
1146 *(idat.vpair) += indirect_Pot;
1147 (*(idat.excludedPot))[ELECTROSTATIC_FAMILY] += excluded_Pot;
1148 (*(idat.pot))[ELECTROSTATIC_FAMILY] += *(idat.sw) * indirect_Pot;
1149 *(idat.f1) += *(idat.sw) * indirect_F;
1150
1151 if (a_is_Dipole || a_is_Quadrupole)
1152 *(idat.t1) += *(idat.sw) * indirect_Ta;
1153
1154 if (b_is_Dipole || b_is_Quadrupole)
1155 *(idat.t2) += *(idat.sw) * indirect_Tb;
1156 }
1157 return;
1158 }
1159
1160 void Electrostatic::calcSelfCorrection(SelfData &sdat) {
1161
1162 if (!initialized_) initialize();
1163
1164 ElectrostaticAtomData data = ElectrostaticMap[Etids[sdat.atid]];
1165
1166 // logicals
1167 bool i_is_Charge = data.is_Charge;
1168 bool i_is_Dipole = data.is_Dipole;
1169 bool i_is_Quadrupole = data.is_Quadrupole;
1170 bool i_is_Fluctuating = data.is_Fluctuating;
1171 RealType C_a = data.fixedCharge;
1172 RealType self(0.0), preVal, DdD, trQ, trQQ;
1173
1174 if (i_is_Dipole) {
1175 DdD = data.dipole.lengthSquare();
1176 }
1177
1178 if (i_is_Fluctuating) {
1179 C_a += *(sdat.flucQ);
1180 // dVdFQ is really a force, so this is negative the derivative
1181 *(sdat.dVdFQ) -= *(sdat.flucQ) * data.hardness + data.electronegativity;
1182 (*(sdat.excludedPot))[ELECTROSTATIC_FAMILY] += (*sdat.flucQ) *
1183 (*(sdat.flucQ) * data.hardness * 0.5 + data.electronegativity);
1184 }
1185
1186 switch (summationMethod_) {
1187 case esm_REACTION_FIELD:
1188
1189 if (i_is_Charge) {
1190 // Self potential [see Wang and Hermans, "Reaction Field
1191 // Molecular Dynamics Simulation with Friedman’s Image Charge
1192 // Method," J. Phys. Chem. 99, 12001-12007 (1995).]
1193 preVal = pre11_ * preRF_ * C_a * C_a;
1194 (*(sdat.pot))[ELECTROSTATIC_FAMILY] -= 0.5 * preVal / cutoffRadius_;
1195 }
1196
1197 if (i_is_Dipole) {
1198 (*(sdat.pot))[ELECTROSTATIC_FAMILY] -= pre22_ * preRF_ * DdD;
1199 }
1200
1201 break;
1202
1203 case esm_SHIFTED_FORCE:
1204 case esm_SHIFTED_POTENTIAL:
1205 if (i_is_Charge)
1206 self += selfMult1_ * pre11_ * C_a * (C_a + *(sdat.skippedCharge));
1207 if (i_is_Dipole)
1208 self += selfMult2_ * pre22_ * DdD;
1209 if (i_is_Quadrupole) {
1210 trQ = data.quadrupole.trace();
1211 trQQ = (data.quadrupole * data.quadrupole).trace();
1212 self += selfMult4_ * pre44_ * (2.0*trQQ + trQ*trQ);
1213 if (i_is_Charge)
1214 self -= selfMult2_ * pre14_ * 2.0 * C_a * trQ;
1215 }
1216 (*(sdat.pot))[ELECTROSTATIC_FAMILY] += self;
1217 break;
1218 default:
1219 break;
1220 }
1221 }
1222
1223 RealType Electrostatic::getSuggestedCutoffRadius(pair<AtomType*, AtomType*> atypes) {
1224 // This seems to work moderately well as a default. There's no
1225 // inherent scale for 1/r interactions that we can standardize.
1226 // 12 angstroms seems to be a reasonably good guess for most
1227 // cases.
1228 return 12.0;
1229 }
1230 }

Properties

Name Value
svn:eol-style native