-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathfd_schemes.hh
451 lines (366 loc) · 11.3 KB
/
fd_schemes.hh
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
/*
fd_schemes.hh - This file is part of MUSIC -
a code to generate multi-scale initial conditions
for cosmological simulations
Copyright (C) 2010 Oliver Hahn
*/
#ifndef __FD_SCHEMES_HH
#define __FD_SCHEMES_HH
#include <vector>
#include <stdexcept>
//! abstract implementation of the Poisson/Force scheme
template< class L, class G, typename real_t=double >
class scheme
{
public:
typedef L laplacian;
typedef G gradient;
laplacian m_laplacian;
gradient m_gradient;
//! gradient along x-direction
template< class C >
inline real_t grad_x( const C&c, const int i, const int j, const int k )
{ return m_gradient.apply_x( c,i,j,k ); }
//! gradient along y-direction
template< class C >
inline real_t grad_y( const C&c, const int i, const int j, const int k )
{ return m_gradient.apply_y( c,i,j,k ); }
//! gradient along z-direction
template< class C >
inline real_t grad_z( const C&c, const int i, const int j, const int k )
{ return m_gradient.apply_z( c,i,j,k ); }
//! apply Laplace operator
template< class C >
inline real_t L_apply( const C&c, const int i, const int j, const int k )
{ return m_laplacian.apply( c,i,j,k ); }
//! compute an explicit solution for the central component of the discrete Poisson's equation
template< class C >
inline real_t L_rhs( const C&c, const int i, const int j, const int k )
{ return m_laplacian.rhs( c,i,j,k ); }
inline real_t ccoeff( void )
{ return m_laplacian.ccoeff(); }
};
//! base class for finite difference gradients
template< int nextent, typename T >
class gradient
{
typedef T real_t;
std::vector<real_t> m_stencil;
const unsigned nl;
public:
gradient()
: nl( 2*nextent+1 )
{
m_stencil.assign(nl*nl*nl,(real_t)0.0);
}
real_t& operator()(int i)
{ return m_stencil[i+nextent]; }
const real_t& operator()(int i) const
{ return m_stencil[i+nextent]; }
template< class C >
inline void apply( const C& c, C& f, int dir )
{
f = c;
int nx=c.size(0), ny=c.size(1), nz=c.size(2);
double hx = 1.0/(nx+1.0), hy = 1.0/(ny+1.0), hz = 1.0/(nz+1.0);
f.zero();
if( dir == 0 )
for( int i=0; i<nx; ++i )
for( int j=0; j<ny; ++j )
for( int k=0; k<nz; ++k )
for( int ii = -nextent; ii<=nextent; ++ii )
f(i,j,k) += (*this)(ii) * c(i+ii,j,k)/hx;
else if( dir == 1 )
for( int i=0; i<nx; ++i )
for( int j=0; j<ny; ++j )
for( int k=0; k<nz; ++k )
for( int jj = -nextent; jj<=nextent; ++jj )
f(i,j,k) += (*this)(jj) * c(i,j+jj,k)/hy;
else if( dir == 2 )
for( int i=0; i<nx; ++i )
for( int j=0; j<ny; ++j )
for( int k=0; k<nz; ++k )
for( int kk = -nextent; kk<=nextent; ++kk )
f(i,j,k) += (*this)(kk) * c(i,j,k+kk)/hz;
}
};
//! base class for finite difference stencils
template< int nextent, typename real_t >
class base_stencil
{
protected:
std::vector<real_t> m_stencil;
const unsigned nl;
public:
bool m_modsource;
public:
base_stencil( bool amodsource = false )
: nl( 2*nextent+1 ), m_modsource( amodsource )
{
m_stencil.assign(nl*nl*nl,(real_t)0.0);
}
real_t& operator()(int i, int j, int k)
{ return m_stencil[((i+nextent)*nl+(j+nextent))*nl+(k+nextent)]; }
const real_t& operator()(unsigned i, unsigned j, unsigned k) const
{ return m_stencil[((i+nextent)*nl+(j+nextent))*nl+(k+nextent)]; }
template< class C >
inline real_t rhs( const C& c, const int i, const int j, const int k )
{
real_t sum = this->apply( c, i, j, k );
sum -= (*this)(0,0,0) * c(i,j,k);
return sum;
}
inline real_t ccoeff( void )
{
return (*this)(0,0,0);
}
template< class C >
inline real_t apply( const C& c, const int i, const int j, const int k )
{
real_t sum = 0.0;
for( int ii=-nextent; ii<=nextent; ++ii )
for( int jj=-nextent; jj<=nextent; ++jj )
for( int kk=-nextent; kk<=nextent; ++kk )
sum += (*this)(ii,jj,kk) * c(i+ii,j+jj,k+kk);
return sum;
}
template< class C >
inline real_t modsource( const C& c, const int i, const int j, const int k )
{
return 0.0;
}
};
/***************************************************************************************/
/***************************************************************************************/
/***************************************************************************************/
//... Implementation of the Gradient schemes............................................
template< typename real_t >
class deriv_2P : public gradient<1,real_t>
{
public:
deriv_2P( void )
{
(*this)( 0 ) = 0.0;
(*this)(-1 ) = -0.5;
(*this)(+1 ) = +0.5;
}
};
//... Implementation of the Laplacian schemes..........................................
//! 7-point, 2nd order finite difference Laplacian
template< typename real_t >
class stencil_7P : public base_stencil<1,real_t>
{
public:
stencil_7P( void )
{
(*this)( 0, 0, 0) = -6.0;
(*this)(-1, 0, 0) = +1.0;
(*this)(+1, 0, 0) = +1.0;
(*this)( 0,-1, 0) = +1.0;
(*this)( 0,+1, 0) = +1.0;
(*this)( 0, 0,-1) = +1.0;
(*this)( 0, 0,+1) = +1.0;
}
template< class C >
inline real_t apply( const C& c, const int i, const int j, const int k ) const
{
//return c(i-1,j,k)+c(i+1,j,k)+c(i,j-1,k)+c(i,j+1,k)+c(i,j,k-1)+c(i,j,k+1)-6.0*c(i,j,k);
return (double)c(i-1,j,k)+(double)c(i+1,j,k)+(double)c(i,j-1,k)+(double)c(i,j+1,k)+(double)c(i,j,k-1)+(double)c(i,j,k+1)-6.0*(double)c(i,j,k);
}
template< class C >
inline real_t rhs( const C& c, const int i, const int j, const int k ) const
{
return c(i-1,j,k)+c(i+1,j,k)+c(i,j-1,k)+c(i,j+1,k)+c(i,j,k-1)+c(i,j,k+1);
}
inline real_t ccoeff( void )
{
return -6.0;
}
};
//! 13-point, 4th order finite difference Laplacian
template< typename real_t >
class stencil_13P : public base_stencil<2,real_t>
{
public:
stencil_13P( void )
{
(*this)( 0, 0, 0) = -90.0/12.;
(*this)(-1, 0, 0) =
(*this)(+1, 0, 0) =
(*this)( 0,-1, 0) =
(*this)( 0,+1, 0) =
(*this)( 0, 0,-1) =
(*this)( 0, 0,+1) = 16./12.;
(*this)(-2, 0, 0) =
(*this)(+2, 0, 0) =
(*this)( 0,-2, 0) =
(*this)( 0,+2, 0) =
(*this)( 0, 0,-2) =
(*this)( 0, 0,+2) = -1./12.;
}
template< class C >
inline real_t apply( const C& c, const int i, const int j, const int k )
{
return
(-1.0*(c(i-2,j,k)+c(i+2,j,k)+c(i,j-2,k)+c(i,j+2,k)+c(i,j,k-2)+c(i,j,k+2))
+16.0*(c(i-1,j,k)+c(i+1,j,k)+c(i,j-1,k)+c(i,j+1,k)+c(i,j,k-1)+c(i,j,k+1))
-90.0*c(i,j,k))/12.0;
}
template< class C >
inline real_t rhs( const C& c, const int i, const int j, const int k )
{
return
(-1.0*(c(i-2,j,k)+c(i+2,j,k)+c(i,j-2,k)+c(i,j+2,k)+c(i,j,k-2)+c(i,j,k+2))
+16.0*(c(i-1,j,k)+c(i+1,j,k)+c(i,j-1,k)+c(i,j+1,k)+c(i,j,k-1)+c(i,j,k+1)))/12.0;
}
inline real_t ccoeff( void )
{
return -90.0/12.0;
}
};
//! 19-point, 6th order finite difference Laplacian
template< typename real_t >
class stencil_19P : public base_stencil<3,real_t>
{
public:
stencil_19P( void )
{
(*this)( 0, 0, 0) = -1470./180.;
(*this)(-1, 0, 0) =
(*this)(+1, 0, 0) =
(*this)( 0,-1, 0) =
(*this)( 0,+1, 0) =
(*this)( 0, 0,-1) =
(*this)( 0, 0,+1) = 270./180.;
(*this)(-2, 0, 0) =
(*this)(+2, 0, 0) =
(*this)( 0,-2, 0) =
(*this)( 0,+2, 0) =
(*this)( 0, 0,-2) =
(*this)( 0, 0,+2) = -27./180.;
(*this)(-3, 0, 0) =
(*this)(+3, 0, 0) =
(*this)( 0,-3, 0) =
(*this)( 0,+3, 0) =
(*this)( 0, 0,-3) =
(*this)( 0, 0,+3) = 2./180.;
}
template< class C >
inline real_t apply( const C& c, const int i, const int j, const int k )
{
return
(2.0*(c(i-3,j,k)+c(i+3,j,k)+c(i,j-3,k)+c(i,j+3,k)+c(i,j,k-3)+c(i,j,k+3))
-27.0*(c(i-2,j,k)+c(i+2,j,k)+c(i,j-2,k)+c(i,j+2,k)+c(i,j,k-2)+c(i,j,k+2))
+270.0*(c(i-1,j,k)+c(i+1,j,k)+c(i,j-1,k)+c(i,j+1,k)+c(i,j,k-1)+c(i,j,k+1))
-1470.0*c(i,j,k))/180.0;
}
template< class C >
inline real_t rhs( const C& c, const int i, const int j, const int k )
{
return
(2.0*(c(i-3,j,k)+c(i+3,j,k)+c(i,j-3,k)+c(i,j+3,k)+c(i,j,k-3)+c(i,j,k+3))
-27.0*(c(i-2,j,k)+c(i+2,j,k)+c(i,j-2,k)+c(i,j+2,k)+c(i,j,k-2)+c(i,j,k+2))
+270.0*(c(i-1,j,k)+c(i+1,j,k)+c(i,j-1,k)+c(i,j+1,k)+c(i,j,k-1)+c(i,j,k+1)))/180.0;
}
inline real_t ccoeff( void )
{
return -1470.0/180.0;
}
};
//! flux operator for the 4th order FD Laplacian
template< typename real_t >
class Laplace_flux_O4
{
public:
/*! computes flux across a surface normal to x-direction
* @param idir idir is -1 for left boundary, +1 for right boundary
* @param c array on which to apply the operator
* @param i grid x-index
* @param j grid y-index
* @param k grid z-index
* @returns flux value
*/
template< class C >
inline double apply_x( int idir, const C& c, const int i, const int j, const int k )
{
double fac = -((double)idir)/12.0;
return fac*(-c(i-2,j,k)+15.0*c(i-1,j,k)-15.0*c(i,j,k)+c(i+1,j,k));
}
/*! computes flux across a surface normal to y-direction
* @param idir idir is -1 for left boundary, +1 for right boundary
* @param c array on which to apply the operator
* @param i grid x-index
* @param j grid y-index
* @param k grid z-index
* @returns flux value
*/
template< class C >
inline double apply_y( int idir, const C& c, const int i, const int j, const int k )
{
double fac = -((double)idir)/12.0;
return fac*(-c(i,j-2,k)+15.0*c(i,j-1,k)-15.0*c(i,j,k)+c(i,j+1,k));
}
/*! computes flux across a surface normal to z-direction
* @param idir idir is -1 for left boundary, +1 for right boundary
* @param c array on which to apply the operator
* @param i grid x-index
* @param j grid y-index
* @param k grid z-index
* @returns flux value
*/
template< class C >
inline double apply_z( int idir, const C& c, const int i, const int j, const int k )
{
double fac = -((double)idir)/12.0;
return fac*(-c(i,j,k-2)+15.0*c(i,j,k-1)-15.0*c(i,j,k)+c(i,j,k+1));
}
};
//! flux operator for the 6th order FD Laplacian
template< typename real_t >
class Laplace_flux_O6
{
public:
/*! computes flux across a surface normal to x-direction
* @param idir idir is -1 for left boundary, +1 for right boundary
* @param c array on which to apply the operator
* @param i grid x-index
* @param j grid y-index
* @param k grid z-index
* @returns flux value
*/
template< class C >
inline double apply_x( int idir, const C& c, const int i, const int j, const int k )
{
double fac = -((double)idir)/180.0;
return fac*(2.*c(i-3,j,k)-25.*c(i-2,j,k)+245.*c(i-1,j,k)-245.0*c(i,j,k)+25.*c(i+1,j,k)-2.*c(i+2,j,k));
}
/*! computes flux across a surface normal to y-direction
* @param idir idir is -1 for left boundary, +1 for right boundary
* @param c array on which to apply the operator
* @param i grid x-index
* @param j grid y-index
* @param k grid z-index
* @returns flux value
*/
template< class C >
inline double apply_y( int idir, const C& c, const int i, const int j, const int k )
{
double fac = -((double)idir)/180.0;
return fac*(2.*c(i,j-3,k)-25.*c(i,j-2,k)+245.*c(i,j-1,k)-245.0*c(i,j,k)+25.*c(i,j+1,k)-2.*c(i,j+2,k));
}
/*! computes flux across a surface normal to z-direction
* @param idir idir is -1 for left boundary, +1 for right boundary
* @param c array on which to apply the operator
* @param i grid x-index
* @param j grid y-index
* @param k grid z-index
* @returns flux value
*/
template< class C >
inline double apply_z( int idir, const C& c, const int i, const int j, const int k )
{
double fac = -((double)idir)/180.0;
return fac*(2.*c(i,j,k-3)-25.*c(i,j,k-2)+245.*c(i,j,k-1)-245.0*c(i,j,k)+25.*c(i,j,k+1)-2.*c(i,j,k+2));
}
};
#endif