1 | SUBROUTINE inflow_turbulence |
---|
2 | |
---|
3 | !------------------------------------------------------------------------------! |
---|
4 | ! Current revisions: |
---|
5 | ! ----------------- |
---|
6 | ! |
---|
7 | ! Former revisions: |
---|
8 | ! ----------------- |
---|
9 | ! $Id: inflow_turbulence.f90 668 2010-12-23 13:22:58Z suehring $ |
---|
10 | ! |
---|
11 | ! 667 2010-12-23 12:06:00Z suehring/gryschka |
---|
12 | ! Using nbgp recycling planes for a better resolution of the turbulent flow |
---|
13 | ! near the inflow. |
---|
14 | ! |
---|
15 | ! 622 2010-12-10 08:08:13Z raasch |
---|
16 | ! optional barriers included in order to speed up collective operations |
---|
17 | ! |
---|
18 | ! 222 2009-01-12 16:04:16Z letzel |
---|
19 | ! Bugfix for nonparallel execution |
---|
20 | ! |
---|
21 | ! Initial version (2008/03/07) |
---|
22 | ! |
---|
23 | ! Description: |
---|
24 | ! ------------ |
---|
25 | ! Imposing turbulence at the respective inflow using the turbulence |
---|
26 | ! recycling method of Kataoka and Mizuno (2002). |
---|
27 | !------------------------------------------------------------------------------! |
---|
28 | |
---|
29 | USE arrays_3d |
---|
30 | USE control_parameters |
---|
31 | USE cpulog |
---|
32 | USE grid_variables |
---|
33 | USE indices |
---|
34 | USE interfaces |
---|
35 | USE pegrid |
---|
36 | |
---|
37 | |
---|
38 | IMPLICIT NONE |
---|
39 | |
---|
40 | INTEGER :: i, imax, j, k, l, ngp_ifd, ngp_pr |
---|
41 | |
---|
42 | REAL, DIMENSION(1:2) :: volume_flow_l, volume_flow_offset |
---|
43 | REAL, DIMENSION(nzb:nzt+1,5,nbgp) :: avpr, avpr_l |
---|
44 | REAL, DIMENSION(nzb:nzt+1,nysg:nyng,5,nbgp) :: inflow_dist |
---|
45 | |
---|
46 | CALL cpu_log( log_point(40), 'inflow_turbulence', 'start' ) |
---|
47 | |
---|
48 | ! |
---|
49 | !-- Carry out spanwise averaging in the recycling plane |
---|
50 | avpr_l = 0.0 |
---|
51 | ngp_pr = ( nzt - nzb + 2 ) * 5 * nbgp |
---|
52 | ngp_ifd = ngp_pr * ( nyn - nys + 1 + 2 * nbgp ) |
---|
53 | |
---|
54 | ! |
---|
55 | !-- First, local averaging within the recycling domain |
---|
56 | |
---|
57 | i = recycling_plane |
---|
58 | |
---|
59 | #if defined( __parallel ) |
---|
60 | IF ( myidx == id_recycling ) THEN |
---|
61 | |
---|
62 | DO l = 1, nbgp |
---|
63 | DO j = nys, nyn |
---|
64 | DO k = nzb, nzt + 1 |
---|
65 | |
---|
66 | avpr_l(k,1,l) = avpr_l(k,1,l) + u(k,j,i) |
---|
67 | avpr_l(k,2,l) = avpr_l(k,2,l) + v(k,j,i) |
---|
68 | avpr_l(k,3,l) = avpr_l(k,3,l) + w(k,j,i) |
---|
69 | avpr_l(k,4,l) = avpr_l(k,4,l) + pt(k,j,i) |
---|
70 | avpr_l(k,5,l) = avpr_l(k,5,l) + e(k,j,i) |
---|
71 | |
---|
72 | ENDDO |
---|
73 | ENDDO |
---|
74 | i = i + 1 |
---|
75 | ENDDO |
---|
76 | |
---|
77 | ENDIF |
---|
78 | ! |
---|
79 | !-- Now, averaging over all PEs |
---|
80 | IF ( collective_wait ) CALL MPI_BARRIER( comm2d, ierr ) |
---|
81 | CALL MPI_ALLREDUCE( avpr_l(nzb,1,1), avpr(nzb,1,1), ngp_pr, & |
---|
82 | MPI_REAL, MPI_SUM, comm2d, ierr ) |
---|
83 | |
---|
84 | #else |
---|
85 | DO l = 1, nbgp |
---|
86 | DO j = nys, nyn |
---|
87 | DO k = nzb, nzt + 1 |
---|
88 | |
---|
89 | avpr_l(k,1,l) = avpr_l(k,1,l) + u(k,j,i) |
---|
90 | avpr_l(k,2,l) = avpr_l(k,2,l) + v(k,j,i) |
---|
91 | avpr_l(k,3,l) = avpr_l(k,3,l) + w(k,j,i) |
---|
92 | avpr_l(k,4,l) = avpr_l(k,4,l) + pt(k,j,i) |
---|
93 | avpr_l(k,5,l) = avpr_l(k,5,l) + e(k,j,i) |
---|
94 | |
---|
95 | ENDDO |
---|
96 | ENDDO |
---|
97 | i = i + 1 |
---|
98 | ENDDO |
---|
99 | |
---|
100 | avpr = avpr_l |
---|
101 | #endif |
---|
102 | |
---|
103 | avpr = avpr / ( ny + 1 ) |
---|
104 | ! |
---|
105 | !-- Calculate the disturbances at the recycling plane |
---|
106 | i = recycling_plane |
---|
107 | |
---|
108 | #if defined( __parallel ) |
---|
109 | IF ( myidx == id_recycling ) THEN |
---|
110 | DO l = 1, nbgp |
---|
111 | DO j = nysg, nyng |
---|
112 | DO k = nzb, nzt + 1 |
---|
113 | |
---|
114 | inflow_dist(k,j,1,l) = u(k,j,i+1) - avpr(k,1,l) |
---|
115 | inflow_dist(k,j,2,l) = v(k,j,i) - avpr(k,2,l) |
---|
116 | inflow_dist(k,j,3,l) = w(k,j,i) - avpr(k,3,l) |
---|
117 | inflow_dist(k,j,4,l) = pt(k,j,i) - avpr(k,4,l) |
---|
118 | inflow_dist(k,j,5,l) = e(k,j,i) - avpr(k,5,l) |
---|
119 | |
---|
120 | ENDDO |
---|
121 | ENDDO |
---|
122 | i = i + 1 |
---|
123 | ENDDO |
---|
124 | |
---|
125 | ENDIF |
---|
126 | #else |
---|
127 | DO l = 1, nbgp |
---|
128 | DO j = nysg, nyng |
---|
129 | DO k = nzb, nzt+1 |
---|
130 | |
---|
131 | inflow_dist(k,j,1,l) = u(k,j,i+1) - avpr(k,1,l) |
---|
132 | inflow_dist(k,j,2,l) = v(k,j,i) - avpr(k,2,l) |
---|
133 | inflow_dist(k,j,3,l) = w(k,j,i) - avpr(k,3,l) |
---|
134 | inflow_dist(k,j,4,l) = pt(k,j,i) - avpr(k,4,l) |
---|
135 | inflow_dist(k,j,5,l) = e(k,j,i) - avpr(k,5,l) |
---|
136 | |
---|
137 | ENDDO |
---|
138 | ENDDO |
---|
139 | i = i + 1 |
---|
140 | ENDDO |
---|
141 | #endif |
---|
142 | |
---|
143 | ! |
---|
144 | !-- For parallel runs, send the disturbances to the respective inflow PE |
---|
145 | #if defined( __parallel ) |
---|
146 | IF ( myidx == id_recycling .AND. myidx /= id_inflow ) THEN |
---|
147 | |
---|
148 | CALL MPI_SEND( inflow_dist(nzb,nysg,1,1), ngp_ifd, MPI_REAL, & |
---|
149 | id_inflow, 1, comm1dx, ierr ) |
---|
150 | |
---|
151 | ELSEIF ( myidx /= id_recycling .AND. myidx == id_inflow ) THEN |
---|
152 | |
---|
153 | inflow_dist = 0.0 |
---|
154 | CALL MPI_RECV( inflow_dist(nzb,nysg,1,1), ngp_ifd, MPI_REAL, & |
---|
155 | id_recycling, 1, comm1dx, status, ierr ) |
---|
156 | |
---|
157 | ENDIF |
---|
158 | #endif |
---|
159 | |
---|
160 | ! |
---|
161 | !-- Add the disturbance at the inflow |
---|
162 | IF ( nxl == 0 ) THEN |
---|
163 | |
---|
164 | DO j = nysg, nyng |
---|
165 | DO k = nzb, nzt + 1 |
---|
166 | |
---|
167 | u(k,j,-nbgp+1:0) = mean_inflow_profiles(k,1) + & |
---|
168 | inflow_dist(k,j,1,1:nbgp) * inflow_damping_factor(k) |
---|
169 | v(k,j,-nbgp:-1) = mean_inflow_profiles(k,2) + & |
---|
170 | inflow_dist(k,j,2,1:nbgp) * inflow_damping_factor(k) |
---|
171 | w(k,j,-nbgp:-1) = inflow_dist(k,j,3,1:nbgp) * inflow_damping_factor(k) |
---|
172 | pt(k,j,-nbgp:-1) = mean_inflow_profiles(k,4) + & |
---|
173 | inflow_dist(k,j,4,1:nbgp) * inflow_damping_factor(k) |
---|
174 | e(k,j,-nbgp:-1) = mean_inflow_profiles(k,5) + & |
---|
175 | inflow_dist(k,j,5,1:nbgp) * inflow_damping_factor(k) |
---|
176 | e(k,j,-nbgp:-1) = MAX( e(k,j,-nbgp:-1), 0.0 ) |
---|
177 | |
---|
178 | ENDDO |
---|
179 | ENDDO |
---|
180 | |
---|
181 | ENDIF |
---|
182 | |
---|
183 | ! |
---|
184 | !-- Conserve the volume flow at the inflow in order to avoid generation of |
---|
185 | !-- waves in the stable layer |
---|
186 | ! IF ( conserve_volume_flow .AND. inflow_l ) THEN |
---|
187 | |
---|
188 | ! volume_flow(1) = 0.0 |
---|
189 | ! volume_flow_l(1) = 0.0 |
---|
190 | |
---|
191 | ! i = 0 |
---|
192 | |
---|
193 | ! DO j = nys, nyn |
---|
194 | ! |
---|
195 | !-- Sum up the volume flow through the south/north boundary |
---|
196 | ! DO k = nzb_2d(j,i) + 1, nzt |
---|
197 | ! volume_flow_l(1) = volume_flow_l(1) + u(k,j,i) * dzu(k) |
---|
198 | ! ENDDO |
---|
199 | ! ENDDO |
---|
200 | |
---|
201 | #if defined( __parallel ) |
---|
202 | ! IF ( collective_wait ) CALL MPI_BARRIER( comm2d, ierr ) |
---|
203 | ! CALL MPI_ALLREDUCE( volume_flow_l(1), volume_flow(1), 1, MPI_REAL, & |
---|
204 | ! MPI_SUM, comm1dy, ierr ) |
---|
205 | #else |
---|
206 | ! volume_flow = volume_flow_l |
---|
207 | #endif |
---|
208 | ! volume_flow_offset(1) = ( volume_flow_initial(1) - volume_flow(1) ) & |
---|
209 | ! / volume_flow_area(1) |
---|
210 | |
---|
211 | ! DO j = nys-1, nyn+1 |
---|
212 | ! DO k = nzb_v_inner(j,i) + 1, nzt |
---|
213 | ! u(k,j,i) = u(k,j,i) + volume_flow_offset(1) |
---|
214 | ! ENDDO |
---|
215 | ! ENDDO |
---|
216 | |
---|
217 | ! ENDIF |
---|
218 | |
---|
219 | CALL cpu_log( log_point(40), 'inflow_turbulence', 'stop' ) |
---|
220 | |
---|
221 | |
---|
222 | END SUBROUTINE inflow_turbulence |
---|