Changes between Version 364 and Version 365 of doc/app/initialization_parameters

Timestamp:

Sep 5, 2017 4:17:34 PM (8 years ago)

Author:

gronemeier

Comment:

--

doc/app/initialization_parameters

@@ -897 +897 @@
 }}}
 {{{#!td style="vertical-align:top"
--1
+4
 }}}
 {{{#!td
 Number of cycles to be used with the multigrid scheme.\\\\
 This parameter determines the number of cycles to be carried out in the multigrid method used for solving the Poisson equation for perturbation pressure (see [#psolver psolver]). The type of the cycles can be set with [#cycle_mg cycle_mg].\\\\
-By default ('''mg_cyles''' = -''1''), the number of cycles depends on the requested accuracy of the scheme (see [#residual_limit residual_limit]) and may vary from time step to time step. In this case, the CPU time for a run will be difficult to estimate since it heavily depends on the total number of the cycles to be carried out.\\\\
-By assigning '''mg_cycles''' a value (>= ''1''), the number of cycles can be fixed so that the CPU time can be clearly estimated.\\\\
-'''Note:'''\\ 
-When using a fixed number of cycles, the user must examine the local file [../iofiles#RUN_CONTROL RUN_CONTROL] regularly to check whether the divergence of the velocity field is sufficiently reduced by the pressure solver. It should be reduced at least by two orders of magnitude. For cyclic boundary conditions along both horizontal directions (see [#bc_lr bc_lr] and [#bc_ns bc_ns]) '''mg_cycles''' = ''2'' is typically a good choice, for non-cyclic lateral boundary conditions '''mg_cycles''' = ''4'' may be sufficient.
+By default ('''mg_cyles''' = ''4''), a fixed number of cycles (4) is carried out. This number of cycles might not be sufficient for every simulation! The user must examine the local file [../iofiles#RUN_CONTROL RUN_CONTROL] regularly to check whether the divergence of the velocity field is sufficiently reduced by the pressure solver. It should be reduced at least by two orders of magnitude. For cyclic boundary conditions along both horizontal directions (see [#bc_lr bc_lr] and [#bc_ns bc_ns]) '''mg_cycles''' = ''2'' is typically a good choice, for non-cyclic lateral boundary conditions '''mg_cycles''' = ''4'' may be sufficient.\\\\
+If set '''mg_cyles''' = ''-1'', the number of cycles depends on the requested accuracy of the scheme (see [#residual_limit residual_limit]) and may vary from time step to time step. In this case, the CPU time for a run will be difficult to estimate since it heavily depends on the total number of the cycles to be carried out.
 }}}
 |----------------
@@ -1020 +1018 @@
       This solver is specially optimized for 1d domain decompositions. Vectorization is optimized for domain decompositions along x only.\\\\
 '' 'multigrid' ''\\\\
-      Multigrid scheme (see [[Diplomarbeit J. Uhlenbrock]], in German only). v- and w-cycles (see [#cycle_mg cycle_mg]) are implemented. The convergence of the iterative scheme can be steered by the number of v-/w-cycles to be carried out for each call of the scheme ([#mg_cycles mg_cycles]) and by the number of Gauss-Seidel iterations (see [#ngsrb ngsrb]) to be carried out on each grid level. Instead, the requested accuracy can be given via [#residual_limit residual_limit]. This is the default! The smaller this limit is, the more cycles have to be carried out in this case and the number of cycles may vary from time step to time step.\\\\
+      Multigrid scheme (see [[Diplomarbeit J. Uhlenbrock]], in German only). v- and w-cycles (see [#cycle_mg cycle_mg]) are implemented. The convergence of the iterative scheme can be steered by the number of v-/w-cycles to be carried out for each call of the scheme ([#mg_cycles mg_cycles]) and by the number of Gauss-Seidel iterations (see [#ngsrb ngsrb]) to be carried out on each grid level. Instead, the requested accuracy can be given via [#residual_limit residual_limit]. The smaller this limit is, the more cycles have to be carried out in this case and the number of cycles may vary from time step to time step.\\\\
       If [#mg_cycles mg_cycles] is set to its optimal value, the computing time of the multigrid scheme amounts approximately to that of the direct solver '' 'poisfft','' as long as the number of grid points in the three directions of space corresponds to a power-of-two (2^n^) where ''n'' >= 5 must hold. With large ''n'', the multigrid scheme can even be faster than the direct solver (although its accuracy is several orders of magnitude worse, but this does not affect the accuracy of the simulation). Nevertheless, the user should always carry out some test runs in order to find out the optimum value for [#mg_cycles mg_cycles], because the CPU time of a run very critically depends on this parameter.\\\\
       This scheme requires that the number of grid points of the subdomains (or of the total domain, if only one PE is used) along each of the directions can, at least, be divided once by 2 without rest.\\\\