-
Notifications
You must be signed in to change notification settings - Fork 9
/
main.f90
1088 lines (1088 loc) · 46.5 KB
/
main.f90
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
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
! SSSSSSSSSSSSSSS NNNNNNNN NNNNNNNN CCCCCCCCCCCCC
! SS:::::::::::::::SN:::::::N N::::::N CCC::::::::::::C
! S:::::SSSSSS::::::SN::::::::N N::::::N CC:::::::::::::::C
! S:::::S SSSSSSSN:::::::::N N::::::N C:::::CCCCCCCC::::C
! S:::::S N::::::::::N N::::::N aaaaaaaaaaaaa C:::::C CCCCCC
! S:::::S N:::::::::::N N::::::N a::::::::::::a C:::::C
! S::::SSSS N:::::::N::::N N::::::N aaaaaaaaa:::::aC:::::C
! SS::::::SSSSS N::::::N N::::N N::::::N a::::aC:::::C
! SSS::::::::SS N::::::N N::::N:::::::N aaaaaaa:::::aC:::::C
! SSSSSS::::S N::::::N N:::::::::::N aa::::::::::::aC:::::C
! S:::::SN::::::N N::::::::::N a::::aaaa::::::aC:::::C
! S:::::SN::::::N N:::::::::Na::::a a:::::a C:::::C CCCCCC
! SSSSSSS S:::::SN::::::N N::::::::Na::::a a:::::a C:::::CCCCCCCC::::C
! S::::::SSSSSS:::::SN::::::N N:::::::Na:::::aaaa::::::a CC:::::::::::::::C
! S:::::::::::::::SS N::::::N N::::::N a::::::::::aa:::a CCC::::::::::::C
! SSSSSSSSSSSSSSS NNNNNNNN NNNNNNN aaaaaaaaaa aaaa CCCCCCCCCCCCC
!-----------------------------------------------------------------------------------
! **SNaC**
! Pedro Costa ([email protected])
!-----------------------------------------------------------------------------------
program snac
use iso_c_binding , only: C_PTR
use mpi_f08
use mod_bound , only: bounduvw,boundp,updt_rhs,inflow
use mod_chkdiv , only: chkdiv
use mod_chkdt , only: chkdt
use mod_common_mpi , only: myid,myid_block,comm_block
use mod_correc , only: correc
use mod_initflow , only: initflow,init_inflow
use mod_initgrid , only: initgrid,distribute_grid,bound_grid,save_grid
use mod_initmpi , only: initmpi
use mod_fillps , only: fillps
use mod_load , only: load
use mod_output , only: out0d,out1d,out2d,write_visu_3d
use mod_param , only: read_input, &
datadir, &
small, &
rkcoeff, &
cfl,dtmin,uref,lref,rey,visc, &
nstep,time_max,tw_max,stop_type, &
restart,is_overwrite_save, &
nthreadsmax, &
icheck,iout0d,iout1d,iout2d,iout3d,isave, &
dims,lo,hi,lmin,lmax, &
gt,gr, &
cbcvel,bcvel,cbcpre,bcpre, &
inflow_type, &
bforce,periods,inivel, &
vol_all,my_block,id_first,nblocks,nrank, &
is_periodic,l_periodic, &
lmax_max,lmin_min,lo_min,hi_max, &
hypre_tol,hypre_maxiter,hypre_solver_i
use mod_updt_pressure , only: updt_pressure
use mod_rk , only: rk_mom
use mod_sanity , only: test_sanity
use mod_solver , only: init_bc_rhs,init_matrix_3d,create_solver,setup_solver, &
add_constant_to_diagonal,solve_helmholtz,finalize_solver,finalize_matrix, &
hypre_solver
#if defined(_FFT_X) || defined(_FFT_Y) || defined(_FFT_Z)
use mod_solver , only: init_fft_reduction,init_n_2d_matrices,create_n_solvers,setup_n_solvers,solve_n_helmholtz_2d, &
init_n_3d_matrices,solve_n_helmholtz_3d, &
add_constant_to_n_diagonals,finalize_n_solvers,finalize_n_matrices
#ifdef _FFT_USE_SLABS
use mod_solver , only: alltoallw,init_comm_slab,init_transpose_slab_uneven,transpose_slab
#endif
use mod_fft , only: fft,fftend
use mod_sanity , only: test_sanity_fft
#endif
use mod_types
!$ use omp_lib
implicit none
integer , dimension(0:1,3) :: nb
logical , dimension(0:1,3) :: is_bound,is_bound_inflow
type(MPI_DATATYPE) , dimension( 3) :: halos
integer , dimension( 3) :: ng,lo_g,hi_g,lo_1,hi_1
real(rp), allocatable, dimension(:,:,:) :: u,v,w,p,up,vp,wp,pp,po
real(rp), allocatable, dimension(:,:,:) :: velin_x,velin_y,velin_z
#ifdef _IMPDIFF
real(rp), allocatable, dimension(:,:,:) :: uo,vo,wo
#endif
real(rp), allocatable, dimension(:,:,:) :: dudtrko,dvdtrko,dwdtrko
type rhs_bound
real(rp), allocatable, dimension(:,:,:) :: x
real(rp), allocatable, dimension(:,:,:) :: y
real(rp), allocatable, dimension(:,:,:) :: z
end type rhs_bound
type(rhs_bound) :: rhsp
type(rhs_bound) :: u_in,v_in,w_in
real(rp) :: alpha,alpha_bc(0:1,1:3)
#ifdef _IMPDIFF
type(rhs_bound) :: rhsu,rhsv,rhsw,bcu,bcv,bcw
#endif
real(rp), dimension(0:1,3) :: dl
#ifdef _IMPDIFF
real(rp), dimension(0:1,3) :: dlu,dlv,dlw
integer , dimension( 3) :: hiu,hiv,hiw
#endif
type(hypre_solver) :: psolver
#ifdef _IMPDIFF
type(hypre_solver) :: usolver,vsolver,wsolver
real(rp) :: alphai,alphaoi
#endif
!
real(rp) :: dt,dtmax,time,dtrk,divtot,divmax
integer :: irk,istep
real(rp), allocatable, dimension(:) :: dxc ,dxf ,xc ,xf , &
dyc ,dyf ,yc ,yf , &
dzc ,dzf ,zc ,zf , &
dxc_g,dxf_g,xc_g,xf_g, &
dyc_g,dyf_g,yc_g,yf_g, &
dzc_g,dzf_g,zc_g,zf_g
logical :: is_uniform_grid
logical, dimension(3) :: is_centered
!
#if defined(_FFT_X) || defined(_FFT_Y) || defined(_FFT_Z)
integer , dimension( 3) :: lo_a,hi_a
logical , dimension(0:1,3) :: is_bound_a
target :: dxc,dxf,dyc,dyf,dzc,dzf
real(rp), pointer, dimension(:) :: dl1_1,dl1_2,dl2_1,dl2_2
type(C_PTR) , dimension(2) :: arrplan_p
real(rp) :: normfft_p
real(rp), allocatable, dimension(:) :: lambda_p,lambda_p_a
real(rp) :: alpha_lambda_p
type(hypre_solver), allocatable, dimension(:) :: psolver_fft
type(MPI_COMM) , allocatable, dimension(:) :: comms_fft
integer :: npsolvers
#ifdef _IMPDIFF
integer , dimension( 3) :: hiu_a,hiv_a,hiw_a
real(rp), pointer, dimension(:) :: dlu1_1,dlu1_2,dlu2_1,dlu2_2, &
dlv1_1,dlv1_2,dlv2_1,dlv2_2, &
dlw1_1,dlw1_2,dlw2_1,dlw2_2
type(C_PTR) , dimension(2) :: arrplan_u,arrplan_v,arrplan_w
real(rp) :: normfft_u,normfft_v,normfft_w
real(rp), allocatable, dimension(:) :: lambda_u,lambda_v,lambda_w
real(rp), allocatable, dimension(:) :: lambda_u_a,lambda_v_a,lambda_w_a
real(rp) :: alpha_lambda_u, &
alpha_lambda_v, &
alpha_lambda_w
type(hypre_solver), allocatable, dimension(:) :: usolver_fft, &
vsolver_fft, &
wsolver_fft
#endif
#ifdef _FFT_USE_SLABS
target :: dxc_g,dxf_g,dyc_g,dyf_g,dzc_g,dzf_g
type(alltoallw), allocatable, dimension(:,:) :: t_params
integer , dimension(3) :: n_s,n_p,lo_s,hi_s
real(rp), pointer , dimension(:) :: dl1_1_g,dl1_2_g,dl2_1_g,dl2_2_g
real(rp), allocatable, dimension(:,:,:) :: pp_s
integer :: nrank_block
#ifdef _IMPDIFF
real(rp), allocatable, dimension(:,:,:) :: up_s,vp_s,wp_s
#endif
#endif
#ifdef _FFT_USE_SLICED_PENCILS
integer :: nslices
integer :: q
#ifndef _FFT_USE_SLABS
integer , dimension(3 ) :: n_s,lo_s,hi_s
#endif
integer, allocatable, dimension(:,:) :: lo_sp,hi_sp
#endif
#endif
!
real(rp), dimension(100) :: var
character(len=3 ) :: cblock
character(len=7 ) :: fldnum
character(len=100) :: filename
integer :: iunit
!
real(rp) :: twi,tw
logical :: is_done,kill
#ifdef _TIMING
real(rp) :: dt12,dt12av,dt12min,dt12max
#endif
integer :: idir,ib,il,iu,iskip
!
call MPI_INIT()
call MPI_COMM_RANK(MPI_COMM_WORLD, myid)
twi = MPI_WTIME()
!
! read parameter file
!
call read_input()
lo_g(:) = lo(:)
hi_g(:) = hi(:)
!
! check sanity of input file
!
call test_sanity(lo,hi,dims,gr,stop_type,cbcvel,cbcpre)
#if defined(_FFT_X) || defined(_FFT_Y) || defined(_FFT_Z)
#if defined(_FFT_X) && !(defined(_FFT_Y) || defined(_FFT_Z))
idir = 1
#elif defined(_FFT_Y) && !(defined(_FFT_X) || defined(_FFT_Z))
idir = 2
#elif defined(_FFT_Z) && !(defined(_FFT_X) || defined(_FFT_Y))
idir = 3
#else
if(myid == 0) write(stderr,*) 'ERROR: there can be only one FFT direction; check the pre-processor flags.'
if(myid == 0) write(stderr,*) 'Aborting...'
call MPI_FINALIZE
error stop
#endif
call test_sanity_fft(dims(idir),lo(idir),hi(idir),lmin(idir),lmax(idir),gr(idir))
#endif
!
! initialize MPI/OpenMP
!
!!$call omp_set_num_threads(nthreadsmax) ! ! overwrites the input, disable for now
call initmpi(my_block,id_first,dims,cbcpre,bcpre,periods,lmin,lmax,gt,gr,lo,hi,ng,nb,is_bound,halos)
lo_1(:) = lo(:) - lo_g(:) + 1 ! lo(:) with 1 as first index in the beginning of each block
hi_1(:) = hi(:) - lo_g(:) + 1 ! hi(:) with 1 as first index in the beginning of each block
!
! allocate variables
!
allocate(u( lo(1)-1:hi(1)+1,lo(2)-1:hi(2)+1,lo(3)-1:hi(3)+1), &
v( lo(1)-1:hi(1)+1,lo(2)-1:hi(2)+1,lo(3)-1:hi(3)+1), &
w( lo(1)-1:hi(1)+1,lo(2)-1:hi(2)+1,lo(3)-1:hi(3)+1), &
p( lo(1)-1:hi(1)+1,lo(2)-1:hi(2)+1,lo(3)-1:hi(3)+1), &
up(lo(1)-1:hi(1)+1,lo(2)-1:hi(2)+1,lo(3)-1:hi(3)+1), &
vp(lo(1)-1:hi(1)+1,lo(2)-1:hi(2)+1,lo(3)-1:hi(3)+1), &
wp(lo(1)-1:hi(1)+1,lo(2)-1:hi(2)+1,lo(3)-1:hi(3)+1), &
pp(lo(1)-1:hi(1)+1,lo(2)-1:hi(2)+1,lo(3)-1:hi(3)+1), &
po(lo(1)-1:hi(1)+1,lo(2)-1:hi(2)+1,lo(3)-1:hi(3)+1))
allocate(u_in%x(lo(2)-1:hi(2)+1,lo(3)-1:hi(3)+1,0:1), &
u_in%y(lo(1)-1:hi(1)+1,lo(3)-1:hi(3)+1,0:1), &
u_in%z(lo(1)-1:hi(1)+1,lo(2)-1:hi(2)+1,0:1), &
v_in%x(lo(2)-1:hi(2)+1,lo(3)-1:hi(3)+1,0:1), &
v_in%y(lo(1)-1:hi(1)+1,lo(3)-1:hi(3)+1,0:1), &
v_in%z(lo(1)-1:hi(1)+1,lo(2)-1:hi(2)+1,0:1), &
w_in%x(lo(2)-1:hi(2)+1,lo(3)-1:hi(3)+1,0:1), &
w_in%y(lo(1)-1:hi(1)+1,lo(3)-1:hi(3)+1,0:1), &
w_in%z(lo(1)-1:hi(1)+1,lo(2)-1:hi(2)+1,0:1))
#ifdef _IMPDIFF
allocate(uo(lo(1)-1:hi(1)+1,lo(2)-1:hi(2)+1,lo(3)-1:hi(3)+1), &
vo(lo(1)-1:hi(1)+1,lo(2)-1:hi(2)+1,lo(3)-1:hi(3)+1), &
wo(lo(1)-1:hi(1)+1,lo(2)-1:hi(2)+1,lo(3)-1:hi(3)+1))
#endif
allocate(dudtrko(lo(1):hi(1),lo(2):hi(2),lo(3):hi(3)), &
dvdtrko(lo(1):hi(1),lo(2):hi(2),lo(3):hi(3)), &
dwdtrko(lo(1):hi(1),lo(2):hi(2),lo(3):hi(3)))
allocate(dxc(lo(1)-1:hi(1)+1), &
dxf(lo(1)-1:hi(1)+1), &
xc(lo(1)-1:hi(1)+1), &
xf(lo(1)-1:hi(1)+1), &
dyc(lo(2)-1:hi(2)+1), &
dyf(lo(2)-1:hi(2)+1), &
yc(lo(2)-1:hi(2)+1), &
yf(lo(2)-1:hi(2)+1), &
dzc(lo(3)-1:hi(3)+1), &
dzf(lo(3)-1:hi(3)+1), &
zc(lo(3)-1:hi(3)+1), &
zf(lo(3)-1:hi(3)+1))
allocate(dxc_g(lo_g(1)-1:hi_g(1)+1), &
dxf_g(lo_g(1)-1:hi_g(1)+1), &
xc_g(lo_g(1)-1:hi_g(1)+1), &
xf_g(lo_g(1)-1:hi_g(1)+1), &
dyc_g(lo_g(2)-1:hi_g(2)+1), &
dyf_g(lo_g(2)-1:hi_g(2)+1), &
yc_g(lo_g(2)-1:hi_g(2)+1), &
yf_g(lo_g(2)-1:hi_g(2)+1), &
dzc_g(lo_g(3)-1:hi_g(3)+1), &
dzf_g(lo_g(3)-1:hi_g(3)+1), &
zc_g(lo_g(3)-1:hi_g(3)+1), &
zf_g(lo_g(3)-1:hi_g(3)+1))
allocate(rhsp%x(lo(2):hi(2),lo(3):hi(3),0:1), &
rhsp%y(lo(1):hi(1),lo(3):hi(3),0:1), &
rhsp%z(lo(1):hi(1),lo(2):hi(2),0:1))
#ifdef _IMPDIFF
allocate(rhsu%x(lo(2):hi(2),lo(3):hi(3),0:1), &
rhsu%y(lo(1):hi(1),lo(3):hi(3),0:1), &
rhsu%z(lo(1):hi(1),lo(2):hi(2),0:1), &
rhsv%x(lo(2):hi(2),lo(3):hi(3),0:1), &
rhsv%y(lo(1):hi(1),lo(3):hi(3),0:1), &
rhsv%z(lo(1):hi(1),lo(2):hi(2),0:1), &
rhsw%x(lo(2):hi(2),lo(3):hi(3),0:1), &
rhsw%y(lo(1):hi(1),lo(3):hi(3),0:1), &
rhsw%z(lo(1):hi(1),lo(2):hi(2),0:1))
allocate( bcu%x(lo(2):hi(2),lo(3):hi(3),0:1), &
bcu%y(lo(1):hi(1),lo(3):hi(3),0:1), &
bcu%z(lo(1):hi(1),lo(2):hi(2),0:1), &
bcv%x(lo(2):hi(2),lo(3):hi(3),0:1), &
bcv%y(lo(1):hi(1),lo(3):hi(3),0:1), &
bcv%z(lo(1):hi(1),lo(2):hi(2),0:1), &
bcw%x(lo(2):hi(2),lo(3):hi(3),0:1), &
bcw%y(lo(1):hi(1),lo(3):hi(3),0:1), &
bcw%z(lo(1):hi(1),lo(2):hi(2),0:1))
#endif
#if defined(_FFT_X) || defined(_FFT_Y) || defined(_FFT_Z)
#ifdef _FFT_USE_SLICED_PENCILS
nslices = max(16,ng(idir))
if(nslices > ng(idir)) then
if(myid == 0) write(stderr,*) 'ERROR: number of pencil slices cannot exceed the number of grid points along the FFT direction.'
call MPI_FINALIZE()
error stop
end if
#if defined(_IMPDIFF) || defined(_FFT_USE_SLABS)
#if defined(_IMPDIFF)
if(myid == 0) write(stderr,*) 'ERROR: implicit diffusion not yet supported with "_FFT_USE_SLICED_PENCILS".'
#elif defined(_FFT_USE_SLABS)
if(myid == 0) write(stderr,*) 'ERROR: cannot select both "_FFT_USE_SLICED_PENCILS" and "_FFT_USE_SLABS".'
#endif
if(myid == 0) write(stderr,*) 'Aborting...'
call MPI_FINALIZE()
error stop
#endif
#endif
lo_a(:) = lo(:)
hi_a(:) = hi(:)
#ifdef _FFT_USE_SLABS
nrank_block = product(dims(:))
if(ng(idir) >= nrank_block) then
n_p = hi(:)-lo(:)+1
#ifdef _FFT_X
n_s = [ng(1)/nrank_block,ng(2),ng(3)]
#elif _FFT_Y
n_s = [ng(1),ng(2)/nrank_block,ng(3)]
#elif _FFT_Z
n_s = [ng(1),ng(2),ng(3)/nrank_block]
#endif
else
write(stderr,*) 'ERROR: number of tasks in block', my_block, 'exceeds the number of grid points along direction', idir,':'
write(stderr,*) nrank_block, ' > ', ng(idir)
write(stderr,*) 'Aborting...'
call MPI_FINALIZE()
stop
end if
!
! distribute slab subdomains as evenly as possible
!
irk = mod(ng(idir),nrank_block)
if(myid_block+1 <= irk) n_s(idir) = n_s(idir) + 1
lo_s(:) = lo_g(:)
hi_s(:) = hi_g(:)
lo_s(idir) = (myid_block*n_s(idir) + 1) - (lo_g(idir) - 1)
hi_s(idir) = lo_s(idir) + n_s(idir) - 1
if(myid_block+1 > irk) then
lo_s(idir) = lo_s(idir) + irk
hi_s(idir) = hi_s(idir) + irk
end if
!
lo_a(:) = lo_s(:)
hi_a(:) = hi_s(:)
allocate(t_params(nrank_block,2))
deallocate(po)
allocate(pp_s(lo_s(1)-0:hi_s(1)+0,lo_s(2)-0:hi_s(2)+0,lo_s(3)-0:hi_s(3)+0), &
po( lo_s(1)-0:hi_s(1)+0,lo_s(2)-0:hi_s(2)+0,lo_s(3)-0:hi_s(3)+0))
pp_s(:,:,:) = 0._rp
#ifdef _IMPDIFF
deallocate(uo,vo,wo)
allocate(up_s(lo_s(1)-0:hi_s(1)+0,lo_s(2)-0:hi_s(2)+0,lo_s(3)-0:hi_s(3)+0), &
vp_s(lo_s(1)-0:hi_s(1)+0,lo_s(2)-0:hi_s(2)+0,lo_s(3)-0:hi_s(3)+0), &
wp_s(lo_s(1)-0:hi_s(1)+0,lo_s(2)-0:hi_s(2)+0,lo_s(3)-0:hi_s(3)+0), &
uo( lo_s(1)-0:hi_s(1)+0,lo_s(2)-0:hi_s(2)+0,lo_s(3)-0:hi_s(3)+0), &
vo( lo_s(1)-0:hi_s(1)+0,lo_s(2)-0:hi_s(2)+0,lo_s(3)-0:hi_s(3)+0), &
wo( lo_s(1)-0:hi_s(1)+0,lo_s(2)-0:hi_s(2)+0,lo_s(3)-0:hi_s(3)+0))
up_s(:,:,:) = 0._rp
vp_s(:,:,:) = 0._rp
wp_s(:,:,:) = 0._rp
#endif
#endif
npsolvers = hi_a(idir)-lo_a(idir)+1
#ifdef _FFT_USE_SLICED_PENCILS
!
! distribute sliced pencils as evenly as possible
!
allocate(lo_sp(3,nslices),hi_sp(3,nslices))
do q=1,nslices
n_s(:) = ng(:)
n_s(idir) = ng(idir)/nslices
irk = mod(ng(idir),nslices)
if(q <= irk) n_s(idir) = n_s(idir) + 1
lo_s(:) = lo(:)
hi_s(:) = hi(:)
lo_s(idir) = lo(idir) + (q-1)*(n_s(idir))
hi_s(idir) = lo_s(idir) + n_s(idir) - 1
if(q > irk) then
lo_s(idir) = lo_s(idir) + irk
hi_s(idir) = hi_s(idir) + irk
end if
lo_sp(:,q) = lo_s(:)
hi_sp(:,q) = hi_s(:)
end do
npsolvers = nslices
#endif
#endif
!
if(myid == 0) then
write(stdout,*) '*******************************'
write(stdout,*) '*** Beginning of simulation ***'
write(stdout,*) '*******************************'
write(stdout,*) ''
end if
!
! generate grid
!
call initgrid(lo_g(1),hi_g(1),gt(1),gr(1),lmin(1),lmax(1),dxc_g,dxf_g,xc_g,xf_g)
call initgrid(lo_g(2),hi_g(2),gt(2),gr(2),lmin(2),lmax(2),dyc_g,dyf_g,yc_g,yf_g)
call initgrid(lo_g(3),hi_g(3),gt(3),gr(3),lmin(3),lmax(3),dzc_g,dzf_g,zc_g,zf_g)
write(cblock,'(i3.3)') my_block
if(myid_block == 0) then
call save_grid(trim(datadir)//'grid_x_b_'//cblock,lo_g(1),hi_g(1),xf_g,xc_g,dxf_g,dxc_g)
call save_grid(trim(datadir)//'grid_y_b_'//cblock,lo_g(2),hi_g(2),yf_g,yc_g,dyf_g,dyc_g)
call save_grid(trim(datadir)//'grid_z_b_'//cblock,lo_g(3),hi_g(3),zf_g,zc_g,dzf_g,dzc_g)
open(newunit=iunit,status='replace',file=trim(datadir)//'geometry_b_'//cblock//'.out')
write(iunit,*) lo_g(1),lo_g(2),lo_g(3)
write(iunit,*) hi_g(1),hi_g(2),hi_g(3)
write(iunit,*) lmin(1),lmin(2),lmin(3)
write(iunit,*) lmax(1),lmax(2),lmax(3)
close(iunit)
end if
call distribute_grid(lo_g(1),lo(1),hi(1),dxc_g,dxc)
call distribute_grid(lo_g(1),lo(1),hi(1),dxf_g,dxf)
call distribute_grid(lo_g(1),lo(1),hi(1), xc_g, xc)
call distribute_grid(lo_g(1),lo(1),hi(1), xf_g, xf)
call distribute_grid(lo_g(2),lo(2),hi(2),dyc_g,dyc)
call distribute_grid(lo_g(2),lo(2),hi(2),dyf_g,dyf)
call distribute_grid(lo_g(2),lo(2),hi(2), yc_g, yc)
call distribute_grid(lo_g(2),lo(2),hi(2), yf_g, yf)
call distribute_grid(lo_g(3),lo(3),hi(3),dzc_g,dzc)
call distribute_grid(lo_g(3),lo(3),hi(3),dzf_g,dzf)
call distribute_grid(lo_g(3),lo(3),hi(3), zc_g, zc)
call distribute_grid(lo_g(3),lo(3),hi(3), zf_g, zf)
call bound_grid(lo_g(1),hi_g(1),lo(1),hi(1),nb(0:1,1),is_periodic(1),lo_min(1),hi_max(1),dxf,dxc)
call bound_grid(lo_g(2),hi_g(2),lo(2),hi(2),nb(0:1,2),is_periodic(2),lo_min(2),hi_max(2),dyf,dyc)
call bound_grid(lo_g(3),hi_g(3),lo(3),hi(3),nb(0:1,3),is_periodic(3),lo_min(3),hi_max(3),dzf,dzc)
#if defined(_FFT_X) || defined(_FFT_Y) || defined(_FFT_Z)
#ifdef _FFT_USE_SLABS
dxc_g(lo_g(1)-1) = 0._rp
dxf_g(lo_g(1)-1) = 0._rp
if(lo(1) == lo_g(1)) then
dxc_g(lo_g(1)-1) = dxc(lo(1)-1)
dxf_g(lo_g(1)-1) = dxf(lo(1)-1)
end if
call MPI_ALLREDUCE(MPI_IN_PLACE,dxc_g(lo_g(1)-1),1,MPI_REAL_RP,MPI_MAX,comm_block)
call MPI_ALLREDUCE(MPI_IN_PLACE,dxf_g(lo_g(1)-1),1,MPI_REAL_RP,MPI_MAX,comm_block)
dxc_g(hi_g(1) ) = 0._rp
dxf_g(hi_g(1) ) = 0._rp
dxc_g(hi_g(1)+1) = 0._rp
dxf_g(hi_g(1)+1) = 0._rp
if(hi(1) == hi_g(1)) then
dxc_g(hi_g(1) ) = dxc(hi(1) )
dxf_g(hi_g(1) ) = dxf(hi(1) )
dxc_g(hi_g(1)+1) = dxc(hi(1)+1)
dxf_g(hi_g(1)+1) = dxf(hi(1)+1)
end if
call MPI_ALLREDUCE(MPI_IN_PLACE,dxc_g(hi_g(1) ),1,MPI_REAL_RP,MPI_MAX,comm_block)
call MPI_ALLREDUCE(MPI_IN_PLACE,dxf_g(hi_g(1) ),1,MPI_REAL_RP,MPI_MAX,comm_block)
call MPI_ALLREDUCE(MPI_IN_PLACE,dxc_g(hi_g(1)+1),1,MPI_REAL_RP,MPI_MAX,comm_block)
call MPI_ALLREDUCE(MPI_IN_PLACE,dxf_g(hi_g(1)+1),1,MPI_REAL_RP,MPI_MAX,comm_block)
dyc_g(lo_g(2)-1) = 0._rp
dyf_g(lo_g(2)-1) = 0._rp
if(lo(2) == lo_g(2)) then
dyc_g(lo_g(2)-1) = dyc(lo(2)-1)
dyf_g(lo_g(2)-1) = dyf(lo(2)-1)
end if
call MPI_ALLREDUCE(MPI_IN_PLACE,dyc_g(lo_g(2)-1),1,MPI_REAL_RP,MPI_MAX,comm_block)
call MPI_ALLREDUCE(MPI_IN_PLACE,dyf_g(lo_g(2)-1),1,MPI_REAL_RP,MPI_MAX,comm_block)
dyc_g(hi_g(2) ) = 0._rp
dyf_g(hi_g(2) ) = 0._rp
dyc_g(hi_g(2)+1) = 0._rp
dyf_g(hi_g(2)+1) = 0._rp
if(hi(2) == hi_g(2)) then
dyc_g(hi_g(2) ) = dyc(hi(2) )
dyf_g(hi_g(2) ) = dyf(hi(2) )
dyc_g(hi_g(2)+1) = dyc(hi(2)+1)
dyf_g(hi_g(2)+1) = dyf(hi(2)+1)
end if
call MPI_ALLREDUCE(MPI_IN_PLACE,dyc_g(hi_g(2) ),1,MPI_REAL_RP,MPI_MAX,comm_block)
call MPI_ALLREDUCE(MPI_IN_PLACE,dyf_g(hi_g(2) ),1,MPI_REAL_RP,MPI_MAX,comm_block)
call MPI_ALLREDUCE(MPI_IN_PLACE,dyc_g(hi_g(2)+1),1,MPI_REAL_RP,MPI_MAX,comm_block)
call MPI_ALLREDUCE(MPI_IN_PLACE,dyf_g(hi_g(2)+1),1,MPI_REAL_RP,MPI_MAX,comm_block)
dzc_g(lo_g(3)-1) = 0._rp
dzf_g(lo_g(3)-1) = 0._rp
if(lo(3) == lo_g(3)) then
dzc_g(lo_g(3)-1) = dzc(lo(3)-1)
dzf_g(lo_g(3)-1) = dzf(lo(3)-1)
end if
call MPI_ALLREDUCE(MPI_IN_PLACE,dzc_g(lo_g(3)-1),1,MPI_REAL_RP,MPI_MAX,comm_block)
call MPI_ALLREDUCE(MPI_IN_PLACE,dzf_g(lo_g(3)-1),1,MPI_REAL_RP,MPI_MAX,comm_block)
dzc_g(hi_g(3) ) = 0._rp
dzf_g(hi_g(3) ) = 0._rp
dzc_g(hi_g(3)+1) = 0._rp
dzf_g(hi_g(3)+1) = 0._rp
if(hi(3) == hi_g(3)) then
dzc_g(hi_g(3) ) = dzc(hi(3) )
dzf_g(hi_g(3) ) = dzf(hi(3) )
dzc_g(hi_g(3)+1) = dzc(hi(3)+1)
dzf_g(hi_g(3)+1) = dzf(hi(3)+1)
end if
call MPI_ALLREDUCE(MPI_IN_PLACE,dzc_g(hi_g(3) ),1,MPI_REAL_RP,MPI_MAX,comm_block)
call MPI_ALLREDUCE(MPI_IN_PLACE,dzf_g(hi_g(3) ),1,MPI_REAL_RP,MPI_MAX,comm_block)
call MPI_ALLREDUCE(MPI_IN_PLACE,dzc_g(hi_g(3)+1),1,MPI_REAL_RP,MPI_MAX,comm_block)
call MPI_ALLREDUCE(MPI_IN_PLACE,dzf_g(hi_g(3)+1),1,MPI_REAL_RP,MPI_MAX,comm_block)
#endif
#endif
is_uniform_grid = all(dzf(:) == dzf(lo(3))) .and. &
all(dyf(:) == dyf(lo(2))) .and. &
all(dxf(:) == dxf(lo(1))) .and. &
#ifdef _FFT_X
dyf(lo(2)) == dzf(lo(3))
#elif _FFT_Y
dxf(lo(1)) == dzf(lo(3))
#elif _FFT_Z
dxf(lo(1)) == dyf(lo(2))
#else
dxf(lo(1)) == dyf(lo(2)) .and. &
dxf(lo(1)) == dzf(lo(3)) .and. &
dyf(lo(2)) == dzf(lo(3))
#endif
call mpi_allreduce(MPI_IN_PLACE,is_uniform_grid,1,MPI_LOGICAL,MPI_LAND,MPI_COMM_WORLD)
!
! initialization of the flow fields
!
u(:,:,:) = 0._rp
v(:,:,:) = 0._rp
w(:,:,:) = 0._rp
p(:,:,:) = 0._rp
if(.not.restart) then
istep = 0
time = 0._rp
call initflow(inivel,.false.,lo,hi,lo_g,hi_g,lmax-lmin,uref,lref,visc,bforce(1), &
xc,xf,yc,yf,zc,zf,dxc,dxf,dyc,dyf,dzc,dzf,u,v,w,p)
if(myid == 0) write(stdout,*) '*** Initial condition succesfully set ***'
else
call load('r',trim(datadir)//'fld_b_'//cblock//'.bin',comm_block,ng,[1,1,1],lo_1,hi_1,u,v,w,p,po,time,istep)
if(myid == 0) write(stdout,*) '*** Checkpoint loaded at time = ', time, 'time step = ', istep, '. ***'
end if
call bounduvw(cbcvel,lo,hi,bcvel,.false.,halos,is_bound,nb, &
dxc,dxf,dyc,dyf,dzc,dzf,u,v,w)
call boundp( cbcpre,lo,hi,bcpre,halos,is_bound,nb,dxc,dyc,dzc,p)
u_in%x(:,:,:) = 0._rp
v_in%x(:,:,:) = 0._rp
w_in%x(:,:,:) = 0._rp
u_in%y(:,:,:) = 0._rp
v_in%y(:,:,:) = 0._rp
w_in%y(:,:,:) = 0._rp
u_in%z(:,:,:) = 0._rp
v_in%z(:,:,:) = 0._rp
w_in%z(:,:,:) = 0._rp
#ifdef _IMPDIFF
do ib=0,1
bcu%x(:,:,ib) = bcvel(ib,1,1)
bcu%y(:,:,ib) = bcvel(ib,2,1)
bcu%z(:,:,ib) = bcvel(ib,3,1)
bcv%x(:,:,ib) = bcvel(ib,1,2)
bcv%y(:,:,ib) = bcvel(ib,2,2)
bcv%z(:,:,ib) = bcvel(ib,3,2)
bcw%x(:,:,ib) = bcvel(ib,1,3)
bcw%y(:,:,ib) = bcvel(ib,2,3)
bcw%z(:,:,ib) = bcvel(ib,3,3)
end do
#endif
do idir=1,3
do ib=0,1
is_bound_inflow(ib,idir) = is_bound(ib,idir).and.inflow_type(ib,idir)>0.and.cbcvel(ib,idir,idir)=='D'
if(is_bound_inflow(ib,idir)) then
select case(idir)
case(1)
il = 2;iu = 3;iskip = 1
call init_inflow(inflow_type(ib,idir),periods(il:iu:iskip), &
lo(il:iu:iskip),hi(il:iu:iskip),lmin(il:iu:iskip),lmax(il:iu:iskip), &
yc,yf,zc,zf,bcvel(ib,idir,idir),lref,visc,u_in%x(:,:,ib),v_in%x(:,:,ib),w_in%x(:,:,ib))
#ifdef _IMPDIFF
bcu%x(:,:,ib) = u_in%x(lo(2):hi(2),lo(3):hi(3),ib)
bcv%x(:,:,ib) = v_in%x(lo(2):hi(2),lo(3):hi(3),ib)
bcw%x(:,:,ib) = w_in%x(lo(2):hi(2),lo(3):hi(3),ib)
#endif
case(2)
il = 1;iu = 3;iskip = 2
call init_inflow(inflow_type(ib,idir),periods(il:iu:iskip), &
lo(il:iu:iskip),hi(il:iu:iskip),lmin(il:iu:iskip),lmax(il:iu:iskip), &
xc,xf,zc,zf,bcvel(ib,idir,idir),lref,visc,v_in%y(:,:,ib),u_in%y(:,:,ib),w_in%y(:,:,ib))
#ifdef _IMPDIFF
bcu%y(:,:,ib) = u_in%y(lo(1):hi(1),lo(3):hi(3),ib)
bcv%y(:,:,ib) = v_in%y(lo(1):hi(1),lo(3):hi(3),ib)
bcw%y(:,:,ib) = w_in%y(lo(1):hi(1),lo(3):hi(3),ib)
#endif
case(3)
il = 1;iu = 2;iskip = 1
call init_inflow(inflow_type(ib,idir),periods(il:iu:iskip), &
lo(il:iu:iskip),hi(il:iu:iskip),lmin(il:iu:iskip),lmax(il:iu:iskip), &
xc,xf,yc,yf,bcvel(ib,idir,idir),lref,visc,w_in%z(:,:,ib),u_in%z(:,:,ib),v_in%z(:,:,ib))
#ifdef _IMPDIFF
bcu%z(:,:,ib) = u_in%z(lo(1):hi(1),lo(2):hi(2),ib)
bcv%z(:,:,ib) = v_in%z(lo(1):hi(1),lo(2):hi(2),ib)
bcw%z(:,:,ib) = w_in%z(lo(1):hi(1),lo(2):hi(2),ib)
#endif
end select
end if
end do
end do
call inflow(is_bound_inflow,.false.,lo,hi,u_in%x,v_in%x,w_in%x, &
u_in%y,v_in%y,w_in%y, &
u_in%z,v_in%z,w_in%z,u,v,w)
up(:,:,:) = 0._rp
vp(:,:,:) = 0._rp
wp(:,:,:) = 0._rp
pp(:,:,:) = 0._rp
po(:,:,:) = 0._rp
dudtrko(:,:,:) = 0._rp
dvdtrko(:,:,:) = 0._rp
dwdtrko(:,:,:) = 0._rp
#ifdef _IMPDIFF
uo(:,:,:) = 0._rp
vo(:,:,:) = 0._rp
wo(:,:,:) = 0._rp
alphaoi = 0._rp
#endif
!
! post-process and write initial condition
!
write(fldnum,'(i7.7)') istep
include 'out1d.h90'
include 'out3d.h90'
!
! determine time step
!
call chkdt(lo,hi,dxc,dxf,dyc,dyf,dzc,dzf,visc,u,v,w,dtmax)
dt = min(cfl*dtmax,dtmin)
if(myid == 0) write(stdout,*) 'dtmax = ', dtmax, 'dt = ',dt
!
! initialize Poisson solver
!
#if defined(_FFT_X) || defined(_FFT_Y) || defined(_FFT_Z)
#ifdef _FFT_X
idir = 1
il = 2;iu = 3;iskip = 1
#ifndef _FFT_USE_SLABS
dl1_1 => dyc;dl1_2 => dyf;dl2_1 => dzc;dl2_2 => dzf
#else
dl1_1 => dyc_g;dl1_2 => dyf_g;dl2_1 => dzc_g;dl2_2 => dzf_g
#endif
#ifdef _IMPDIFF
#ifndef _FFT_USE_SLABS
dlu1_1 => dyc;dlu1_2 => dyf;dlu2_1 => dzc;dlu2_2 => dzf
dlv1_1 => dyf;dlv1_2 => dyc;dlv2_1 => dzc;dlv2_2 => dzf
dlw1_1 => dyc;dlw1_2 => dyf;dlw2_1 => dzf;dlw2_2 => dzc
#else
dlu1_1 => dyc_g;dlu1_2 => dyf_g;dlu2_1 => dzc_g;dlu2_2 => dzf_g
dlv1_1 => dyf_g;dlv1_2 => dyc_g;dlv2_1 => dzc_g;dlv2_2 => dzf_g
dlw1_1 => dyc_g;dlw1_2 => dyf_g;dlw2_1 => dzf_g;dlw2_2 => dzc_g
#endif
#endif
#elif _FFT_Y
idir = 2
il = 1;iu = 3;iskip = 2
#ifndef _FFT_USE_SLABS
dl1_1 => dxc;dl1_2 => dxf;dl2_1 => dzc;dl2_2 => dzf
#else
dl1_1 => dxc_g;dl1_2 => dxf_g;dl2_1 => dzc_g;dl2_2 => dzf_g
#endif
#ifdef _IMPDIFF
#ifndef _FFT_USE_SLABS
dlu1_1 => dxf;dlu1_2 => dxc;dlu2_1 => dzc;dlu2_2 => dzf
dlv1_1 => dxc;dlv1_2 => dxf;dlv2_1 => dzc;dlv2_2 => dzf
dlw1_1 => dxc;dlw1_2 => dxf;dlw2_1 => dzf;dlw2_2 => dzc
#else
dlu1_1 => dxf_g;dlu1_2 => dxc_g;dlu2_1 => dzc_g;dlu2_2 => dzf_g
dlv1_1 => dxc_g;dlv1_2 => dxf_g;dlv2_1 => dzc_g;dlv2_2 => dzf_g
dlw1_1 => dxc_g;dlw1_2 => dxf_g;dlw2_1 => dzf_g;dlw2_2 => dzc_g
#endif
#endif
#elif _FFT_Z
idir = 3
il = 1;iu = 2;iskip = 1
#ifndef _FFT_USE_SLABS
dl1_1 => dxc;dl1_2 => dxf;dl2_1 => dyc;dl2_2 => dyf
#else
dl1_1 => dxc_g;dl1_2 => dxf_g;dl2_1 => dyc_g;dl2_2 => dyf_g
#endif
#ifdef _IMPDIFF
#ifndef _FFT_USE_SLABS
dlu1_1 => dxf;dlu1_2 => dxc;dlu2_1 => dyc;dlu2_2 => dyf
dlv1_1 => dxc;dlv1_2 => dxf;dlv2_1 => dyf;dlv2_2 => dyc
dlw1_1 => dxc;dlw1_2 => dxf;dlw2_1 => dyc;dlw2_2 => dyf
#else
dlu1_1 => dxf_g;dlu1_2 => dxc_g;dlu2_1 => dyc_g;dlu2_2 => dyf_g
dlv1_1 => dxc_g;dlv1_2 => dxf_g;dlv2_1 => dyf_g;dlv2_2 => dyc_g
dlw1_1 => dxc_g;dlw1_2 => dxf_g;dlw2_1 => dyc_g;dlw2_2 => dyf_g
#endif
#endif
#endif
#endif
dl = reshape([dxc_g(lo_g(1)-1),dxc_g(hi_g(1)), &
dyc_g(lo_g(2)-1),dyc_g(hi_g(2)), &
dzc_g(lo_g(3)-1),dzc_g(hi_g(3))],shape(dl))
is_centered(:) = [.true.,.true.,.true.]
call init_bc_rhs(cbcpre,bcpre,dl,is_bound,is_centered,lo,hi,periods, &
dxc,dxf,dyc,dyf,dzc,dzf,rhsp%x,rhsp%y,rhsp%z)
alpha = 0._rp
alpha_bc(:,:) = 0._rp
#if defined(_FFT_X) || defined(_FFT_Y) || defined(_FFT_Z)
allocate(lambda_p(hi(idir)-lo(idir)+1))
call init_fft_reduction(idir,hi(:)-lo(:)+1,cbcpre(:,idir),.true.,dl(0,idir),arrplan_p,normfft_p,lambda_p)
alpha_lambda_p = 0._rp
allocate(psolver_fft(npsolvers))
allocate(comms_fft(hi_a(idir)-lo_a(idir)+1))
allocate(lambda_p_a(hi_a(idir)-lo_a(idir)+1))
is_bound_a(:,:) = is_bound(:,:)
#ifndef _FFT_USE_SLABS
comms_fft(:) = MPI_COMM_WORLD
lambda_p_a(:) = lambda_p
#else
do ib=1,3 ! determine is_bound pertaining to the slabs
is_bound_a(:,ib) = .false.
where(cbcpre(0:1,ib) /= 'F') is_bound_a(0:1,ib) = .true.
end do
call init_comm_slab(lo(idir),hi(idir),lo_s(idir),hi_s(idir),myid,comms_fft)
lambda_p_a(:) = lambda_p(lo_s(idir)-lo(idir)+1:hi_s(idir)-lo(idir)+1)
call init_transpose_slab_uneven(idir,1,0,dims,lo_1-1,lo_s-1,n_p,n_s,comm_block,t_params)
#endif
#ifndef _FFT_USE_SLICED_PENCILS
call init_n_2d_matrices(cbcpre(:,il:iu:iskip),bcpre(:,il:iu:iskip),dl(:,il:iu:iskip), &
is_uniform_grid,is_bound_a(:,il:iu:iskip),is_centered(il:iu:iskip), &
lo_a(idir),hi_a(idir),lo_a(il:iu:iskip),hi_a(il:iu:iskip),periods(il:iu:iskip), &
dl1_1,dl1_2,dl2_1,dl2_2,alpha,alpha_bc,lambda_p_a,comms_fft,psolver_fft)
#else
call init_n_3d_matrices(idir,nslices,cbcpre,bcpre,dl,is_uniform_grid,is_bound,is_centered,lo,periods, &
lo_sp,hi_sp,dxc,dxf,dyc,dyf,dzc,dzf,alpha,alpha_bc,lambda_p_a,psolver_fft)
#endif
call create_n_solvers(npsolvers,hypre_maxiter,hypre_tol,hypre_solver_i,psolver_fft)
call setup_n_solvers(npsolvers,psolver_fft)
#else
call init_matrix_3d(cbcpre,bcpre,dl,is_uniform_grid,is_bound,is_centered,lo,hi,periods, &
dxc,dxf,dyc,dyf,dzc,dzf,alpha,alpha_bc,psolver)
call create_solver(hypre_maxiter,hypre_tol,hypre_solver_i,psolver)
call setup_solver(psolver)
#endif
#ifdef _IMPDIFF
dlu = reshape([dxf_g(lo_g(1)-0),dxf_g(hi_g(1)), &
dyc_g(lo_g(2)-1),dyc_g(hi_g(2)), &
dzc_g(lo_g(3)-1),dzc_g(hi_g(3))],shape(dlu))
hiu(:) = hi(:)
if(is_bound(1,1)) hiu(:) = hiu(:)-[1,0,0]
is_centered(:) = [.false.,.true.,.true.]
call init_bc_rhs(cbcvel(:,:,1),bcvel(:,:,1),dlu,is_bound,is_centered,lo,hiu,periods, &
dxf,dxc,dyc,dyf,dzc,dzf,rhsu%x,rhsu%y,rhsu%z,bcu%x,bcu%y,bcu%z)
#if defined(_FFT_X) || defined(_FFT_Y) || defined(_FFT_Z)
allocate(lambda_u(hi(idir)-lo(idir)+1))
call init_fft_reduction(idir,hi(:)-lo(:)+1,cbcvel(:,idir,1),is_centered(idir),dlu(0,idir),arrplan_u,normfft_u,lambda_u)
alpha_lambda_u = 0._rp
allocate(lambda_u_a(hi_a(idir)-lo_a(idir)+1))
allocate(usolver_fft(hi_a(idir)-lo_a(idir)+1))
#ifndef _FFT_USE_SLABS
lambda_u_a(:) = lambda_u(:)
hiu_a(:) = hiu(:)
#else
lambda_u_a(:) = lambda_u(lo_s(idir)-lo(idir)+1:hi_s(idir)-lo(idir)+1)
hiu_a(:) = hi_s(:)
if(periods(1) == 0) hiu_a(:) = hiu_a(:)-[1,0,0]
#endif
call init_n_2d_matrices(cbcvel(:,il:iu:iskip,1),bcvel(:,il:iu:iskip,1),dlu(:,il:iu:iskip), &
is_uniform_grid,is_bound_a(:,il:iu:iskip),is_centered(il:iu:iskip), &
lo_a(idir),hiu_a(idir),lo_a(il:iu:iskip),hiu_a(il:iu:iskip),periods(il:iu:iskip), &
dlu1_1,dlu1_2,dlu2_1,dlu2_2,alpha,alpha_bc,lambda_u_a,comms_fft,usolver_fft)
#else
call init_matrix_3d(cbcvel(:,:,1),bcvel(:,:,1),dlu,is_uniform_grid,is_bound,is_centered,lo,hiu,periods, &
dxf,dxc,dyc,dyf,dzc,dzf,alpha,alpha_bc,usolver)
#endif
dlv = reshape([dxc_g(lo_g(1)-1),dxc_g(hi_g(1)), &
dyf_g(lo_g(2)-0),dyf_g(hi_g(2)), &
dzc_g(lo_g(3)-1),dzc_g(hi_g(3))],shape(dlv))
hiv(:) = hi(:)
if(is_bound(1,2)) hiv(:) = hiv(:)-[0,1,0]
is_centered(:) = [.true.,.false.,.true.]
call init_bc_rhs(cbcvel(:,:,2),bcvel(:,:,2),dlv,is_bound,is_centered,lo,hiv,periods, &
dxc,dxf,dyf,dyc,dzc,dzf,rhsv%x,rhsv%y,rhsv%z,bcv%x,bcv%y,bcv%z)
#if defined(_FFT_X) || defined(_FFT_Y) || defined(_FFT_Z)
allocate(lambda_v(hi(idir)-lo(idir)+1))
call init_fft_reduction(idir,hi(:)-lo(:)+1,cbcvel(:,idir,2),is_centered(idir),dlv(0,idir),arrplan_v,normfft_v,lambda_v)
alpha_lambda_v = 0._rp
allocate(lambda_v_a(hi_a(idir)-lo_a(idir)+1))
allocate(vsolver_fft(hi_a(idir)-lo_a(idir)+1))
#ifndef _FFT_USE_SLABS
lambda_v_a(:) = lambda_v(:)
hiv_a(:) = hiv(:)
#else
lambda_v_a(:) = lambda_v(lo_s(idir)-lo(idir)+1:hi_s(idir)-lo(idir)+1)
hiv_a(:) = hi_s(:)
if(periods(2) == 0) hiv_a(:) = hiv_a(:)-[0,1,0]
#endif
call init_n_2d_matrices(cbcvel(:,il:iu:iskip,2),bcvel(:,il:iu:iskip,2),dlv(:,il:iu:iskip), &
is_uniform_grid,is_bound_a(:,il:iu:iskip),is_centered(il:iu:iskip), &
lo_a(idir),hiv_a(idir),lo_a(il:iu:iskip),hiv_a(il:iu:iskip),periods(il:iu:iskip), &
dlv1_1,dlv1_2,dlv2_1,dlv2_2,alpha,alpha_bc,lambda_v_a,comms_fft,vsolver_fft)
#else
call init_matrix_3d(cbcvel(:,:,2),bcvel(:,:,2),dlv,is_uniform_grid,is_bound,is_centered,lo,hiv,periods, &
dxc,dxf,dyf,dyc,dzc,dzf,alpha,alpha_bc,vsolver)
#endif
dlw = reshape([dxc_g(lo_g(1)-1),dxc_g(hi_g(1)), &
dyc_g(lo_g(2)-1),dyc_g(hi_g(2)), &
dzf_g(lo_g(3)-0),dzf_g(hi_g(3))],shape(dlw))
hiw(:) = hi(:)
if(is_bound(1,3)) hiw(:) = hiw(:)-[0,0,1]
is_centered(:) = [.true.,.true.,.false.]
call init_bc_rhs(cbcvel(:,:,3),bcvel(:,:,3),dlw,is_bound,is_centered,lo,hiw,periods, &
dxc,dxf,dyc,dyf,dzf,dzc,rhsw%x,rhsw%y,rhsw%z,bcw%x,bcw%y,bcw%z)
#if defined(_FFT_X) || defined(_FFT_Y) || defined(_FFT_Z)
allocate(lambda_w(hi(idir)-lo(idir)+1))
call init_fft_reduction(idir,hi(:)-lo(:)+1,cbcvel(:,idir,3),is_centered(idir),dlw(0,idir),arrplan_w,normfft_w,lambda_w)
alpha_lambda_w = 0._rp
allocate(lambda_w_a(hi_a(idir)-lo_a(idir)+1))
allocate(wsolver_fft(hi_a(idir)-lo_a(idir)+1))
#ifndef _FFT_USE_SLABS
lambda_w_a(:) = lambda_w(:)
hiw_a(:) = hiw(:)
#else
lambda_w_a(:) = lambda_w(lo_s(idir)-lo(idir)+1:hi_s(idir)-lo(idir)+1)
hiw_a(:) = hi_s(:)
if(periods(3) == 0) hiw_a(:) = hiw_a(:)-[0,0,1]
#endif
call init_n_2d_matrices(cbcvel(:,il:iu:iskip,3),bcvel(:,il:iu:iskip,3),dlw(:,il:iu:iskip), &
is_uniform_grid,is_bound_a(:,il:iu:iskip),is_centered(il:iu:iskip), &
lo_a(idir),hiw_a(idir),lo_a(il:iu:iskip),hiw_a(il:iu:iskip),periods(il:iu:iskip), &
dlw1_1,dlw1_2,dlw2_1,dlw2_2,alpha,alpha_bc,lambda_w_a,comms_fft,wsolver_fft)
#else
call init_matrix_3d(cbcvel(:,:,3),bcvel(:,:,3),dlw,is_uniform_grid,is_bound,is_centered,lo,hiw,periods, &
dxc,dxf,dyc,dyf,dzf,dzc,alpha,alpha_bc,wsolver)
#endif
#endif
!
! main loop
!
if(myid == 0) write(stdout,*) '*** Calculation loop starts now ***'
kill = .false.
is_done = .false.
do while(.not.is_done)
#ifdef _TIMING
dt12 = MPI_WTIME()
#endif
istep = istep + 1
time = time + dt
if(myid == 0) write(stdout,*) 'Timestep #', istep, 'Time = ', time
do irk=1,3
dtrk = sum(rkcoeff(:,irk))*dt
alpha = -visc*dtrk/2._rp
call rk_mom(rkcoeff(:,irk),lo,hi,dxc,dxf,dyc,dyf,dzc,dzf,dt,bforce, &
visc,u,v,w,p,dudtrko,dvdtrko,dwdtrko,up,vp,wp)
#ifdef _IMPDIFF
alphai = alpha**(-1)
!
!$OMP WORKSHARE
up(lo(1):hi(1),lo(2):hi(2),lo(3):hi(3)) = up(lo(1):hi(1),lo(2):hi(2),lo(3):hi(3))*alphai
!$OMP END WORKSHARE
call updt_rhs(lo,hiu,is_bound,rhsu%x,rhsu%y,rhsu%z,up)
#if defined(_FFT_X) || defined(_FFT_Y) || defined(_FFT_Z)
call add_constant_to_n_diagonals(hiu_a(idir)-lo_a(idir)+1,lo_a(il:iu:iskip),hiu_a(il:iu:iskip), &
alphai-alphaoi,usolver_fft(:)%mat) ! correct diagonal term
call create_n_solvers(hiu_a(idir)-lo_a(idir)+1,hypre_maxiter,hypre_tol,hypre_solver_i,usolver_fft)
call setup_n_solvers(hiu_a(idir)-lo_a(idir)+1,usolver_fft)
call fft(arrplan_u(1),up(lo(1):hi(1),lo(2):hi(2),lo(3):hi(3)))
#ifndef _FFT_USE_SLABS
call solve_n_helmholtz_2d(usolver_fft,lo(idir),hiu(idir),1,lo(il:iu:iskip),hiu(il:iu:iskip),up,uo)
#else
call transpose_slab(1,0,t_params(:,1:2:1 ),comm_block,up,up_s)
call solve_n_helmholtz_2d(usolver_fft,lo_a(idir),hiu_a(idir),0,lo_a(il:iu:iskip),hiu_a(il:iu:iskip),up_s,uo)
call transpose_slab(0,1,t_params(:,2:1:-1),comm_block,up_s,up)
#endif
call fft(arrplan_u(2),up(lo(1):hi(1),lo(2):hi(2),lo(3):hi(3)))
!$OMP WORKSHARE
up(lo(1):hi(1),lo(2):hi(2),lo(3):hi(3)) = up(lo(1):hi(1),lo(2):hi(2),lo(3):hi(3))*normfft_u
!$OMP END WORKSHARE
call finalize_n_solvers(hiu_a(idir)-lo_a(idir)+1,usolver_fft)
#else
call add_constant_to_diagonal(lo,hiu,alphai-alphaoi,usolver%mat) ! correct diagonal term
call create_solver(hypre_maxiter,hypre_tol,hypre_solver_i,usolver)
call setup_solver(usolver)
call solve_helmholtz(usolver,lo,hiu,up,uo)
call finalize_solver(usolver)
#endif
!
!$OMP WORKSHARE
vp(lo(1):hi(1),lo(2):hi(2),lo(3):hi(3)) = vp(lo(1):hi(1),lo(2):hi(2),lo(3):hi(3))*alphai
!$OMP END WORKSHARE
call updt_rhs(lo,hiv,is_bound,rhsv%x,rhsv%y,rhsv%z,vp)
#if defined(_FFT_X) || defined(_FFT_Y) || defined(_FFT_Z)
call add_constant_to_n_diagonals(hiv_a(idir)-lo_a(idir)+1,lo_a(il:iu:iskip),hiv_a(il:iu:iskip), &
alphai-alphaoi,vsolver_fft(:)%mat) ! correct diagonal term
call create_n_solvers(hiv_a(idir)-lo_a(idir)+1,hypre_maxiter,hypre_tol,hypre_solver_i,vsolver_fft)
call setup_n_solvers(hiv_a(idir)-lo_a(idir)+1,vsolver_fft)
call fft(arrplan_v(1),vp(lo(1):hi(1),lo(2):hi(2),lo(3):hi(3)))
#ifndef _FFT_USE_SLABS
call solve_n_helmholtz_2d(vsolver_fft,lo(idir),hiv(idir),1,lo(il:iu:iskip),hiv(il:iu:iskip),vp,vo)
#else
call transpose_slab(1,0,t_params(:,1:2:1 ),comm_block,vp,vp_s)
call solve_n_helmholtz_2d(vsolver_fft,lo_a(idir),hiv_a(idir),0,lo_a(il:iu:iskip),hiv_a(il:iu:iskip),vp_s,vo)
call transpose_slab(0,1,t_params(:,2:1:-1),comm_block,vp_s,vp)
#endif
call fft(arrplan_v(2),vp(lo(1):hi(1),lo(2):hi(2),lo(3):hi(3)))
!$OMP WORKSHARE
vp(lo(1):hi(1),lo(2):hi(2),lo(3):hi(3)) = vp(lo(1):hi(1),lo(2):hi(2),lo(3):hi(3))*normfft_v
!$OMP END WORKSHARE
call finalize_n_solvers(hiv_a(idir)-lo_a(idir)+1,vsolver_fft)
#else
call add_constant_to_diagonal(lo,hiv,alphai-alphaoi,vsolver%mat) ! correct diagonal term
call create_solver(hypre_maxiter,hypre_tol,hypre_solver_i,vsolver)
call setup_solver(vsolver)
call solve_helmholtz(vsolver,lo,hiv,vp,vo)
call finalize_solver(vsolver)
#endif
!
!$OMP WORKSHARE
wp(lo(1):hi(1),lo(2):hi(2),lo(3):hi(3)) = wp(lo(1):hi(1),lo(2):hi(2),lo(3):hi(3))*alphai
!$OMP END WORKSHARE
call updt_rhs(lo,hiw,is_bound,rhsw%x,rhsw%y,rhsw%z,wp)
#if defined(_FFT_X) || defined(_FFT_Y) || defined(_FFT_Z)
call add_constant_to_n_diagonals(hiw_a(idir)-lo_a(idir)+1,lo_a(il:iu:iskip),hiw_a(il:iu:iskip), &
alphai-alphaoi,wsolver_fft(:)%mat) ! correct diagonal term
call create_n_solvers(hiw_a(idir)-lo_a(idir)+1,hypre_maxiter,hypre_tol,hypre_solver_i,wsolver_fft)
call setup_n_solvers(hiw_a(idir)-lo_a(idir)+1,wsolver_fft)
call fft(arrplan_w(1),wp(lo(1):hi(1),lo(2):hi(2),lo(3):hi(3)))
#ifndef _FFT_USE_SLABS
call solve_n_helmholtz_2d(wsolver_fft,lo(idir),hiw(idir),1,lo(il:iu:iskip),hiw(il:iu:iskip),wp,wo)
#else
call transpose_slab(1,0,t_params(:,1:2:1 ),comm_block,wp,wp_s)
call solve_n_helmholtz_2d(wsolver_fft,lo_a(idir),hiw_a(idir),0,lo_a(il:iu:iskip),hiw_a(il:iu:iskip),wp_s,wo)
call transpose_slab(0,1,t_params(:,2:1:-1),comm_block,wp_s,wp)
#endif
call fft(arrplan_w(2),wp(lo(1):hi(1),lo(2):hi(2),lo(3):hi(3)))
!$OMP WORKSHARE
wp(lo(1):hi(1),lo(2):hi(2),lo(3):hi(3)) = wp(lo(1):hi(1),lo(2):hi(2),lo(3):hi(3))*normfft_w
!$OMP END WORKSHARE
call finalize_n_solvers(hiw_a(idir)-lo_a(idir)+1,wsolver_fft)
#else
call add_constant_to_diagonal(lo,hiw,alphai-alphaoi,wsolver%mat) ! correct diagonal term
call create_solver(hypre_maxiter,hypre_tol,hypre_solver_i,wsolver)
call setup_solver(wsolver)
call solve_helmholtz(wsolver,lo,hiw,wp,wo)
call finalize_solver(wsolver)
#endif
!
alphaoi = alphai
#endif
call bounduvw(cbcvel,lo,hi,bcvel,.false.,halos,is_bound,nb, &
dxc,dxf,dyc,dyf,dzc,dzf,up,vp,wp)
call inflow(is_bound_inflow,.false.,lo,hi,u_in%x,v_in%x,w_in%x, &
u_in%y,v_in%y,w_in%y, &
u_in%z,v_in%z,w_in%z,up,vp,wp)
#if !defined(_IMPDIFF) && defined(_ONE_PRESS_CORR)
dtrk = dt
if(irk < 3) then ! pressure correction only at the last RK step
!$OMP WORKSHARE
u(:,:,:) = up(:,:,:)
v(:,:,:) = vp(:,:,:)
w(:,:,:) = wp(:,:,:)
!$OMP END WORKSHARE
cycle
end if
#endif
call fillps(lo,hi,dxf,dyf,dzf,dtrk,up,vp,wp,pp)
call updt_rhs(lo,hi,is_bound,rhsp%x,rhsp%y,rhsp%z,pp)
#if defined(_FFT_X) || defined(_FFT_Y) || defined(_FFT_Z)
call fft(arrplan_p(1),pp(lo(1):hi(1),lo(2):hi(2),lo(3):hi(3)))
#ifndef _FFT_USE_SLABS
#ifdef _FFT_USE_SLICED_PENCILS
call solve_n_helmholtz_3d(psolver_fft,nslices,lo_sp,hi_sp,1,lo,hi,pp,po)
#else
call solve_n_helmholtz_2d(psolver_fft,lo(idir),hi(idir),1,lo(il:iu:iskip),hi(il:iu:iskip),pp,po)
#endif
#else
call transpose_slab(1,0,t_params(:,1:2:1 ),comm_block,pp,pp_s)
call solve_n_helmholtz_2d(psolver_fft,lo_s(idir),hi_s(idir),0,lo_s(il:iu:iskip),hi_s(il:iu:iskip),pp_s,po)
call transpose_slab(0,1,t_params(:,2:1:-1),comm_block,pp_s,pp)
#endif
call fft(arrplan_p(2),pp(lo(1):hi(1),lo(2):hi(2),lo(3):hi(3)))
!$OMP WORKSHARE
pp(lo(1):hi(1),lo(2):hi(2),lo(3):hi(3)) = pp(lo(1):hi(1),lo(2):hi(2),lo(3):hi(3))*normfft_p
!$OMP END WORKSHARE
#else
call solve_helmholtz(psolver,lo,hi,pp,po)
#endif
call boundp( cbcpre,lo,hi,bcpre,halos,is_bound,nb,dxc,dyc,dzc,pp)
call correc(lo,hi,dxc,dyc,dzc,dtrk,pp,up,vp,wp,u,v,w)
call bounduvw(cbcvel,lo,hi,bcvel,.true.,halos,is_bound,nb, &
dxc,dxf,dyc,dyf,dzc,dzf,u,v,w)
call inflow(is_bound_inflow,.true.,lo,hi,u_in%x,v_in%x,w_in%x, &
u_in%y,v_in%y,w_in%y, &
u_in%z,v_in%z,w_in%z,u,v,w)
call updt_pressure(lo,hi,dxc,dxf,dyc,dyf,dzc,dzf,alpha,pp,p)
call boundp( cbcpre,lo,hi,bcpre,halos,is_bound,nb,dxc,dyc,dzc,p)
end do
!
! check simulation stopping criteria
!
if(stop_type(1)) then ! maximum number of time steps reached
if(istep >= nstep ) is_done = is_done.or..true.
end if
if(stop_type(2)) then ! maximum simulation time reached
if(time >= time_max) is_done = is_done.or..true.
end if
if(stop_type(3)) then ! maximum wall-clock time reached
tw = (MPI_WTIME()-twi)/3600._rp
call MPI_ALLREDUCE(MPI_IN_PLACE,tw,1,MPI_REAL_RP,MPI_MAX,MPI_COMM_WORLD)
if(tw >= tw_max ) is_done = is_done.or..true.
end if
if(mod(istep,icheck) == 0) then
if(myid == 0) write(stdout,*) 'Checking stability and divergence...'
!
call chkdt(lo,hi,dxc,dxf,dyc,dyf,dzc,dzf,visc,u,v,w,dtmax)
dt = min(cfl*dtmax,dtmin)