Home

Context Navigation

source: palm/trunk/SOURCE/timestep.f90 @ 3129

Last change on this file since 3129 was 3120, checked in by gronemeier, 6 years ago
implementation of dynamic sgs model
Property svn:keywords set to `Id`
File size: 16.3 KB

Line
1	!> @file timestep.f90
2	!------------------------------------------------------------------------------!
3	! This file is part of the PALM model system.
4	!
5	! PALM is free software: you can redistribute it and/or modify it under the
6	! terms of the GNU General Public License as published by the Free Software
7	! Foundation, either version 3 of the License, or (at your option) any later
8	! version.
9	!
10	! PALM is distributed in the hope that it will be useful, but WITHOUT ANY
11	! WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR
12	! A PARTICULAR PURPOSE. See the GNU General Public License for more details.
13	!
14	! You should have received a copy of the GNU General Public License along with
15	! PALM. If not, see <http://www.gnu.org/licenses/>.
16	!
17	! Copyright 1997-2018 Leibniz Universitaet Hannover
18	!------------------------------------------------------------------------------!
19	!
20	! Current revisions:
21	! ------------------
22	!
23	!
24	! Former revisions:
25	! -----------------
26	! $Id: timestep.f90 3120 2018-07-11 18:30:57Z gronemeier $
27	! Put ABS( km ) in computation of time step according to the diffusion criterion
28	!
29	! 3084 2018-06-19 15:30:55Z gronemeier
30	! limit increase of dt_3d only in case of RANS mode
31	!
32	! 3083 2018-06-19 14:03:12Z gronemeier
33	! limit dt_3d to be at maximum 2*old_dt; define old_dt at beginning of routine
34	! Add km/kh_max
35	!
36	! 3049 2018-05-29 13:52:36Z Giersch
37	! Error messages revised
38	!
39	! 3045 2018-05-28 07:55:41Z Giersch
40	! Error message revised
41	!
42	! 2718 2018-01-02 08:49:38Z maronga
43	! Corrected "Former revisions" section
44	!
45	! 2696 2017-12-14 17:12:51Z kanani
46	! Change in file header (GPL part)
47	!
48	! 2365 2017-08-21 14:59:59Z kanani
49	! Vertical grid nesting: Sync fine and coarse grid timestep (SadiqHuq)
50	!
51	! 2258 2017-06-08 07:55:13Z suehring
52	! Bugfix, add pre-preprocessor directives to enable non-parrallel mode
53	!
54	! 2168 2017-03-06 13:08:38Z suehring
55	!
56	! 2130 2017-01-24 16:25:39Z raasch
57	! bugfix: in case of nested runs the stop condition in case of too small
58	! timesteps is communicated to all parent/child processes,
59	! some formatting done
60	!
61	! 2118 2017-01-17 16:38:49Z raasch
62	! OpenACC directives and related part of code removed
63	!
64	! 2000 2016-08-20 18:09:15Z knoop
65	! Forced header and separation lines into 80 columns
66	!
67	! 1849 2016-04-08 11:33:18Z hoffmann
68	! Adapted for modularization of microphysics
69	!
70	! 1682 2015-10-07 23:56:08Z knoop
71	! Code annotations made doxygen readable
72	!
73	! 1484 2014-10-21 10:53:05Z kanani
74	! Changes due to new module structure of the plant canopy model:
75	! calculations and parameters related to the plant canopy model removed
76	! (the limitation of the canopy drag, i.e. that the canopy drag itself should
77	! not change the sign of the velocity components, is now assured for in the
78	! calculation of the canopy tendency terms in subroutine plant_canopy_model)
79	!
80	! 1342 2014-03-26 17:04:47Z kanani
81	! REAL constants defined as wp-kind
82	!
83	! 1322 2014-03-20 16:38:49Z raasch
84	! REAL functions provided with KIND-attribute
85	!
86	! 1320 2014-03-20 08:40:49Z raasch
87	! ONLY-attribute added to USE-statements,
88	! kind-parameters added to all INTEGER and REAL declaration statements,
89	! kinds are defined in new module kinds,
90	! old module precision_kind is removed,
91	! revision history before 2012 removed,
92	! comment fields (!:) to be used for variable explanations added to
93	! all variable declaration statements
94	!
95	! 1257 2013-11-08 15:18:40Z raasch
96	! openacc porting
97	! bugfix for calculation of advective timestep in case of vertically stretched
98	! grids
99	!
100	! 1092 2013-02-02 11:24:22Z raasch
101	! unused variables removed
102	!
103	! 1053 2012-11-13 17:11:03Z hoffmann
104	! timestep is reduced in two-moment cloud scheme according to the maximum
105	! terminal velocity of rain drops
106	!
107	! 1036 2012-10-22 13:43:42Z raasch
108	! code put under GPL (PALM 3.9)
109	!
110	! 1001 2012-09-13 14:08:46Z raasch
111	! all actions concerning leapfrog scheme removed
112	!
113	! 978 2012-08-09 08:28:32Z fricke
114	! restriction of the outflow damping layer in the diffusion criterion removed
115	!
116	! 866 2012-03-28 06:44:41Z raasch
117	! bugfix for timestep calculation in case of Galilei transformation,
118	! special treatment in case of mirror velocity boundary condition removed
119	!
120	! Revision 1.1 1997/08/11 06:26:19 raasch
121	! Initial revision
122	!
123	!
124	! Description:
125	! ------------
126	!> Compute the time step under consideration of the FCL and diffusion criterion.
127	!------------------------------------------------------------------------------!
128	SUBROUTINE timestep
129
130
131	USE arrays_3d, &
132	ONLY: dzu, dzw, kh, km, u, v, w
133
134	USE control_parameters, &
135	ONLY: cfl_factor, coupling_mode, dt_3d, dt_fixed, dt_max, &
136	galilei_transformation, old_dt, message_string, rans_mode, &
137	stop_dt, terminate_coupled, terminate_coupled_remote, &
138	timestep_reason, u_gtrans, use_ug_for_galilei_tr, v_gtrans
139
140	USE cpulog, &
141	ONLY: cpu_log, log_point
142
143	USE grid_variables, &
144	ONLY: dx, dx2, dy, dy2
145
146	USE indices, &
147	ONLY: nxl, nxlg, nxr, nxrg, nyn, nyng, nys, nysg, nzb, nzt
148
149	USE interfaces
150
151	USE kinds
152
153	USE microphysics_mod, &
154	ONLY: dt_precipitation
155
156	USE pegrid
157
158	USE pmc_interface, &
159	ONLY: nested_run
160
161	USE statistics, &
162	ONLY: flow_statistics_called, hom, u_max, u_max_ijk, v_max, v_max_ijk,&
163	w_max, w_max_ijk
164
165	USE vertical_nesting_mod, &
166	ONLY: vnested, vnest_timestep_sync
167
168	IMPLICIT NONE
169
170	INTEGER(iwp) :: i !<
171	INTEGER(iwp) :: j !<
172	INTEGER(iwp) :: k !<
173	INTEGER(iwp) :: km_max_ijk(3) = -1 !< index values (i,j,k) of location where km_max occurs
174	INTEGER(iwp) :: kh_max_ijk(3) = -1 !< index values (i,j,k) of location where kh_max occurs
175
176	LOGICAL :: stop_dt_local !< local switch for controlling the time stepping
177
178	REAL(wp) :: div !<
179	REAL(wp) :: dt_diff !<
180	REAL(wp) :: dt_diff_l !<
181	REAL(wp) :: dt_u !<
182	REAL(wp) :: dt_u_l !<
183	REAL(wp) :: dt_v !<
184	REAL(wp) :: dt_v_l !<
185	REAL(wp) :: dt_w !<
186	REAL(wp) :: dt_w_l !<
187	REAL(wp) :: km_max !< maximum of Km in entire domain
188	REAL(wp) :: kh_max !< maximum of Kh in entire domain
189	REAL(wp) :: u_gtrans_l !<
190	REAL(wp) :: u_max_l !<
191	REAL(wp) :: u_min_l !<
192	REAL(wp) :: value !<
193	REAL(wp) :: v_gtrans_l !<
194	REAL(wp) :: v_max_l !<
195	REAL(wp) :: v_min_l !<
196	REAL(wp) :: w_max_l !<
197	REAL(wp) :: w_min_l !<
198
199	REAL(wp), DIMENSION(2) :: uv_gtrans !<
200	REAL(wp), DIMENSION(2) :: uv_gtrans_l !<
201	REAL(wp), DIMENSION(3) :: reduce !<
202	REAL(wp), DIMENSION(3) :: reduce_l !<
203	REAL(wp), DIMENSION(nzb+1:nzt) :: dxyz2_min !<
204
205
206	CALL cpu_log( log_point(12), 'calculate_timestep', 'start' )
207	!
208	!-- Save former time step as reference
209	old_dt = dt_3d
210
211	!
212	!-- In case of Galilei-transform not using the geostrophic wind as translation
213	!-- velocity, compute the volume-averaged horizontal velocity components, which
214	!-- will then be subtracted from the horizontal wind for the time step and
215	!-- horizontal advection routines.
216	IF ( galilei_transformation .AND. .NOT. use_ug_for_galilei_tr ) THEN
217	IF ( flow_statistics_called ) THEN
218	!
219	!-- Horizontal averages already existent, just need to average them
220	!-- vertically.
221	u_gtrans = 0.0_wp
222	v_gtrans = 0.0_wp
223	DO k = nzb+1, nzt
224	u_gtrans = u_gtrans + hom(k,1,1,0)
225	v_gtrans = v_gtrans + hom(k,1,2,0)
226	ENDDO
227	u_gtrans = u_gtrans / REAL( nzt - nzb, KIND=wp )
228	v_gtrans = v_gtrans / REAL( nzt - nzb, KIND=wp )
229	ELSE
230	!
231	!-- Averaging over the entire model domain.
232	u_gtrans_l = 0.0_wp
233	v_gtrans_l = 0.0_wp
234	DO i = nxl, nxr
235	DO j = nys, nyn
236	DO k = nzb+1, nzt
237	u_gtrans_l = u_gtrans_l + u(k,j,i)
238	v_gtrans_l = v_gtrans_l + v(k,j,i)
239	ENDDO
240	ENDDO
241	ENDDO
242	uv_gtrans_l(1) = u_gtrans_l / &
243	REAL( (nxr-nxl+1)(nyn-nys+1)(nzt-nzb), KIND=wp )
244	uv_gtrans_l(2) = v_gtrans_l / &
245	REAL( (nxr-nxl+1)(nyn-nys+1)(nzt-nzb), KIND=wp )
246	#if defined( __parallel )
247	IF ( collective_wait ) CALL MPI_BARRIER( comm2d, ierr )
248	CALL MPI_ALLREDUCE( uv_gtrans_l, uv_gtrans, 2, MPI_REAL, MPI_SUM, &
249	comm2d, ierr )
250	u_gtrans = uv_gtrans(1) / REAL( numprocs, KIND=wp )
251	v_gtrans = uv_gtrans(2) / REAL( numprocs, KIND=wp )
252	#else
253	u_gtrans = uv_gtrans_l(1)
254	v_gtrans = uv_gtrans_l(2)
255	#endif
256	ENDIF
257	ENDIF
258
259	!
260	!-- Determine the maxima of the velocity components, including their
261	!-- grid index positions.
262	CALL global_min_max( nzb, nzt+1, nysg, nyng, nxlg, nxrg, u, 'abs', 0.0_wp, &
263	u_max, u_max_ijk )
264	CALL global_min_max( nzb, nzt+1, nysg, nyng, nxlg, nxrg, v, 'abs', 0.0_wp, &
265	v_max, v_max_ijk )
266	CALL global_min_max( nzb, nzt+1, nysg, nyng, nxlg, nxrg, w, 'abs', 0.0_wp, &
267	w_max, w_max_ijk )
268
269	IF ( .NOT. dt_fixed ) THEN
270	!
271	!-- Variable time step:
272	!-- Calculate the maximum time step according to the CFL-criterion,
273	!-- individually for each velocity component
274	dt_u_l = 999999.9_wp
275	dt_v_l = 999999.9_wp
276	dt_w_l = 999999.9_wp
277	DO i = nxl, nxr
278	DO j = nys, nyn
279	DO k = nzb+1, nzt
280	dt_u_l = MIN( dt_u_l, ( dx / &
281	( ABS( u(k,j,i) - u_gtrans ) + 1.0E-10_wp ) ) )
282	dt_v_l = MIN( dt_v_l, ( dy / &
283	( ABS( v(k,j,i) - v_gtrans ) + 1.0E-10_wp ) ) )
284	dt_w_l = MIN( dt_w_l, ( dzu(k) / &
285	( ABS( w(k,j,i) ) + 1.0E-10_wp ) ) )
286	ENDDO
287	ENDDO
288	ENDDO
289
290	#if defined( __parallel )
291	reduce_l(1) = dt_u_l
292	reduce_l(2) = dt_v_l
293	reduce_l(3) = dt_w_l
294	IF ( collective_wait ) CALL MPI_BARRIER( comm2d, ierr )
295	CALL MPI_ALLREDUCE( reduce_l, reduce, 3, MPI_REAL, MPI_MIN, comm2d, ierr )
296	dt_u = reduce(1)
297	dt_v = reduce(2)
298	dt_w = reduce(3)
299	#else
300	dt_u = dt_u_l
301	dt_v = dt_v_l
302	dt_w = dt_w_l
303	#endif
304
305	!
306	!-- Compute time step according to the diffusion criterion.
307	!-- First calculate minimum grid spacing which only depends on index k.
308	!-- When using the dynamic subgrid model, negative km are possible.
309	dt_diff_l = 999999.0_wp
310
311	DO k = nzb+1, nzt
312	dxyz2_min(k) = MIN( dx2, dy2, dzw(k)dzw(k) ) 0.125_wp
313	ENDDO
314
315	!$OMP PARALLEL private(i,j,k,value) reduction(MIN: dt_diff_l)
316	!$OMP DO
317	DO i = nxl, nxr
318	DO j = nys, nyn
319	DO k = nzb+1, nzt
320	dt_diff_l = MIN( dt_diff_l, dxyz2_min(k) / &
321	( MAX( kh(k,j,i), ABS( km(k,j,i) ) ) &
322	+ 1E-20_wp ) )
323	ENDDO
324	ENDDO
325	ENDDO
326	!$OMP END PARALLEL
327	#if defined( __parallel )
328	IF ( collective_wait ) CALL MPI_BARRIER( comm2d, ierr )
329	CALL MPI_ALLREDUCE( dt_diff_l, dt_diff, 1, MPI_REAL, MPI_MIN, comm2d, &
330	ierr )
331	#else
332	dt_diff = dt_diff_l
333	#endif
334
335	!
336	!-- The time step is the minimum of the 3-4 components and the diffusion time
337	!-- step minus a reduction (cfl_factor) to be on the safe side.
338	!-- The time step must not exceed the maximum allowed value.
339	dt_3d = cfl_factor * MIN( dt_diff, dt_u, dt_v, dt_w, dt_precipitation )
340	dt_3d = MIN( dt_3d, dt_max )
341	!
342	!-- In RANS mode, the time step must not increase by more than a factor of 2
343	IF ( rans_mode ) dt_3d = MIN( dt_3d, dt_max, 2.0_wp * old_dt )
344
345	!
346	!-- Remember the restricting time step criterion for later output.
347	IF ( MIN( dt_u, dt_v, dt_w ) < dt_diff ) THEN
348	timestep_reason = 'A'
349	ELSE
350	timestep_reason = 'D'
351	ENDIF
352
353	!
354	!-- Set flag if the time step becomes too small.
355	IF ( dt_3d < ( 0.00001_wp * dt_max ) ) THEN
356	stop_dt = .TRUE.
357
358	!
359	!-- Determine the maxima of the diffusion coefficients, including their
360	!-- grid index positions.
361	CALL global_min_max( nzb, nzt+1, nysg, nyng, nxlg, nxrg, km, 'abs', &
362	0.0_wp, km_max, km_max_ijk )
363	CALL global_min_max( nzb, nzt+1, nysg, nyng, nxlg, nxrg, kh, 'abs', &
364	0.0_wp, kh_max, kh_max_ijk )
365
366	WRITE( message_string, * ) 'Time step has reached minimum limit.', &
367	'&dt = ', dt_3d, ' s Simulation is terminated.', &
368	'&old_dt = ', old_dt, ' s', &
369	'&dt_u = ', dt_u, ' s', &
370	'&dt_v = ', dt_v, ' s', &
371	'&dt_w = ', dt_w, ' s', &
372	'&dt_diff = ', dt_diff, ' s', &
373	'&u_max = ', u_max, ' m/s k=', u_max_ijk(1), &
374	' j=', u_max_ijk(2), ' i=', u_max_ijk(3), &
375	'&v_max = ', v_max, ' m/s k=', v_max_ijk(1), &
376	' j=', v_max_ijk(2), ' i=', v_max_ijk(3), &
377	'&w_max = ', w_max, ' m/s k=', w_max_ijk(1), &
378	' j=', w_max_ijk(2), ' i=', w_max_ijk(3), &
379	'&km_max = ', km_max, ' m2/s2 k=', km_max_ijk(1), &
380	' j=', km_max_ijk(2), ' i=', km_max_ijk(3), &
381	'&kh_max = ', kh_max, ' m2/s2 k=', kh_max_ijk(1), &
382	' j=', kh_max_ijk(2), ' i=', kh_max_ijk(3)
383	CALL message( 'timestep', 'PA0312', 0, 1, 0, 6, 0 )
384	!
385	!-- In case of coupled runs inform the remote model of the termination
386	!-- and its reason, provided the remote model has not already been
387	!-- informed of another termination reason (terminate_coupled > 0) before.
388	#if defined( __parallel )
389	IF ( coupling_mode /= 'uncoupled' .AND. terminate_coupled == 0 ) THEN
390	terminate_coupled = 2
391	IF ( myid == 0 ) THEN
392	CALL MPI_SENDRECV( &
393	terminate_coupled, 1, MPI_INTEGER, target_id, 0, &
394	terminate_coupled_remote, 1, MPI_INTEGER, target_id, 0, &
395	comm_inter, status, ierr )
396	ENDIF
397	CALL MPI_BCAST( terminate_coupled_remote, 1, MPI_INTEGER, 0, &
398	comm2d, ierr)
399	ENDIF
400	#endif
401	ENDIF
402
403	!
404	!-- In case of nested runs all parent/child processes have to terminate if
405	!-- one process has set the stop flag, i.e. they need to set the stop flag
406	!-- too.
407	IF ( nested_run ) THEN
408	stop_dt_local = stop_dt
409	#if defined( __parallel )
410	CALL MPI_ALLREDUCE( stop_dt_local, stop_dt, 1, MPI_LOGICAL, MPI_LOR, &
411	MPI_COMM_WORLD, ierr )
412	#endif
413	ENDIF
414
415	!
416	!-- Ensure a smooth value (two significant digits) of the timestep.
417	div = 1000.0_wp
418	DO WHILE ( dt_3d < div )
419	div = div / 10.0_wp
420	ENDDO
421	dt_3d = NINT( dt_3d * 100.0_wp / div ) * div / 100.0_wp
422
423	ENDIF
424
425	!
426	!-- Vertical nesting: coarse and fine grid timestep has to be identical
427	IF ( vnested ) CALL vnest_timestep_sync
428
429	CALL cpu_log( log_point(12), 'calculate_timestep', 'stop' )
430
431	END SUBROUTINE timestep

Note: See TracBrowser for help on using the repository browser.

Download in other formats:

| Impressum | ©Leibniz Universität Hannover |