[cig-commits] r19149 - seismo/3D/SPECFEM3D/trunk/src/shared

[liuqy at geodynamics.org]

Author: liuqy
Date: 2011-11-04 07:08:43 -0700 (Fri, 04 Nov 2011)
New Revision: 19149

Modified:
   seismo/3D/SPECFEM3D/trunk/src/shared/check_mesh_resolution.f90
Log:
Fix bug related to pmax_glob for elastic/acoustic medium.



Modified: seismo/3D/SPECFEM3D/trunk/src/shared/check_mesh_resolution.f90
===================================================================
--- seismo/3D/SPECFEM3D/trunk/src/shared/check_mesh_resolution.f90	2011-11-04 01:40:19 UTC (rev 19148)
+++ seismo/3D/SPECFEM3D/trunk/src/shared/check_mesh_resolution.f90	2011-11-04 14:08:43 UTC (rev 19149)
@@ -456,10 +456,13 @@
     !          also, keep in mind that the minimum period is just an estimation and
     !          there is no such sharp cut-off period for valid synthetics.
     !          seismograms become just more and more inaccurate for periods shorter than this estimate.
-    pmax = avg_distance / min( vpmin,vp2min,vsmin ) * NPTS_PER_WAVELENGTH
+    if (has_vp2_zero) then
+        pmax = avg_distance / min( vpmin,vsmin ) * NPTS_PER_WAVELENGTH
+    else
+        pmax = avg_distance / min( vpmin,vp2max,vsmin ) * NPTS_PER_WAVELENGTH
+    endif
     pmax_glob = max(pmax_glob,pmax)
 
-
     ! old: based on GLL distance, i.e. on maximum ratio ( gridspacing / velocity )
     !pmax = distance_max / min( vpmin,vsmin ) * NELEM_PER_WAVELENGTH
     !pmax_glob = max(pmax_glob,pmax)