[cig-commits] r16992 - in short/3D/PyLith/trunk: . doc/releasenotes

brad at geodynamics.org brad at geodynamics.org
Fri Jun 11 10:16:59 PDT 2010


Author: brad
Date: 2010-06-11 10:16:58 -0700 (Fri, 11 Jun 2010)
New Revision: 16992

Modified:
   short/3D/PyLith/trunk/README
   short/3D/PyLith/trunk/doc/releasenotes/announce_v1.5.0.txt
Log:
Small edits to README and release announcement.

Modified: short/3D/PyLith/trunk/README
===================================================================
--- short/3D/PyLith/trunk/README	2010-06-11 17:02:19 UTC (rev 16991)
+++ short/3D/PyLith/trunk/README	2010-06-11 17:16:58 UTC (rev 16992)
@@ -27,9 +27,8 @@
 
 PyLith is under active development and we expect a number of additions
 and improvements in the near future. Likely enhancements will include
-additional constitutive models for both volume elements and cohesive
-(fault) elements and the generation of Green's functions to be used in
-inversions.
+additional bulk and fault constitutive models and the generation of
+Green's functions to be used in inversions.
 
 ======================================================================
 TIPS
@@ -42,24 +41,24 @@
       - Command line arguments
 
         --petsc.pc_type=asm
-        --petsc.ksp_max_it=100
-        --petsc.ksp_gmres_restart=50
+        --petsc.ksp_max_it=400
+        --petsc.ksp_gmres_restart=100
         --petsc.ksp_rtol=1.0e-08
 
       - pylithapp.cfg (or your other favorite .cfg file)
 
         [pylithapp.petsc]
         pc_type = asm
-        ksp_max_it = 200
-        ksp_gmres_restart = 50
+        ksp_max_it = 400
+        ksp_gmres_restart = 100
         ksp_rtol = 1.0e-08
 
-  * If the solve takes more than a few hundred iterations for a
-    large problem (use the --petsc.ksp_monitor=1 and
-    --petsc.ksp_view=1 to see diagnostic information for the
-    solver), this is an indication that something is wrong. Either
-    the preconditioner is inappropriate for the type of problem you
-    are solving or there is an error in the problem setup.
+  * If the solve takes more than a few hundred iterations for a large
+    problem (use the --petsc.ksp_monitor=1 and --petsc.ksp_view=1 to
+    see diagnostic information for the solver), this is usually an
+    indication that something is wrong. Either the preconditioner is
+    inappropriate for the type of problem you are solving or there is
+    an error in the problem setup.
 
 
 ======================================================================
@@ -71,10 +70,10 @@
 
 (1) Recent releases of CUBIT include nodeset names in the Exodus file
 and PyLith now uses them to associate vertices with boundary
-conditions and faults. Use the netcdf utility ncdump to examine the
+conditions and faults. Use the NetCDF utility ncdump to examine the
 contents of the Exodus (.exo) file to see it it includes the variable
-ns_names. If it does use nodeset names rather than nodeset ids for
-boundary condition label properties. If your Exodus file does not
+ns_names. If it does, then use nodeset names rather than nodeset ids
+for boundary condition label properties. If your Exodus file does not
 contain nodeset names, then set the MeshIOCubit property
 use_nodeset_names to False to continue to use nodeset id values for
 boundary condition labels.
@@ -89,13 +88,13 @@
 described in the 'Material Models' section of the manual.
 
 (3) The fault property 'normal_dir' is obsolete. Only the property
-'up_dir' is required to enforce positive slip is left-lateral,
+'up_dir' is required to enforce that positive slip is left-lateral,
 reverse, and fault-opening for dipping faults in 2-D and horizontal
 fault surfaces in 3-D. Previously, in 2-D positive slip was always
 left-lateral, but now the up-direction is used to enforce positive
-slip is reverse motion for dipping faults. For horizontal fault
-surfaces in 3-D a normal of (0,0,1) is assumed in determining the
-up-dip direction.
+slip corresponds to reverse motion for dipping faults. For horizontal
+fault surfaces in 3-D a normal of (0,0,1) is assumed in determining
+the up-dip direction.
 
 ----------------------------------------------------------------------
 Version 1.5.0
@@ -103,17 +102,18 @@
 
   * Fault constitutive models
 
-    Added fault friction interface conditions with static
-    friction, linear slip-weakening friction, and rate- and
-    state-friction with the ageing law. The implementation can be used
-    in static, quasi-static, or dynamic problems.
+    Added fault friction interface conditions with static friction,
+    linear slip-weakening friction, and rate- and state-friction with
+    the ageing law. The implementation can be used in static,
+    quasi-static, and dynamic problems.
 
   * Drucker-Prager elastoplastic bulk rheology
 
-    Added a Drucker-Prager elastoplastic bulk rheology. This is a perfect
-    plasticity implementation (no hardening). This is a nonlinear
-    constitutive model, so the nonlinear solver is required when this
-    rheology is used. Refer to the 'Material Models' section of the manual.
+    Added a Drucker-Prager elastoplastic bulk rheology. This is a
+    perfect plasticity implementation (no hardening). This is a
+    nonlinear constitutive model, so the nonlinear solver is required
+    when this rheology is used. Refer to the 'Material Models' section
+    of the manual.
 
   * Plane strain Maxwell viscoelastic bulk rheology
 
@@ -138,9 +138,8 @@
     Added optimized elasticity objects for the most popular cell types
     and basis functions (linear polynomials). For tri3 and tet4 cells
     with one quadrature point, the optimized implementations do not
-    use mapped cells to reduce the number of operations. For quad4 and
-    hex8 cells the optimized objects provide hardwired loop bounds
-    which can lead to faster performance for some compilers.
+    use reference (mapped) cells in order to reduce the number of
+    operations. 
 
   * Scientific notation for ASCII VTK files
 

Modified: short/3D/PyLith/trunk/doc/releasenotes/announce_v1.5.0.txt
===================================================================
--- short/3D/PyLith/trunk/doc/releasenotes/announce_v1.5.0.txt	2010-06-11 17:02:19 UTC (rev 16991)
+++ short/3D/PyLith/trunk/doc/releasenotes/announce_v1.5.0.txt	2010-06-11 17:16:58 UTC (rev 16992)
@@ -5,13 +5,13 @@
 viscoelastic problems in tectonic deformation.
 
 This release adds several new features to PyLith, including (1) fault
-friction with several widely-used fault constitutive models, (2) an
+friction with a few widely-used fault constitutive models, (2) an
 optimized solver for explicit time-stepping with a lumped Jacobian
-sparse matrix, (3) a total Langrangian formulation for rigid-body
-motion and small strains, (4) use of scientific notation in VTK output
-files with user-specified precision, and (5) use of nodeset names in
-CUBIT Exodus files. We encourage all users of previous PyLith releases
-to switch to this latest release.
+matrix, (3) a total Langrangian formulation for rigid-body motion and
+small strains, (4) use of scientific notation in VTK output files with
+user-specified precision, and (5) use of nodeset names in CUBIT Exodus
+files. We encourage all users of previous PyLith releases to switch to
+this latest release.
 
 We strongly recommend all users of previous PyLith releases to switch
 to this latest release. In addition to adding features this release
@@ -35,14 +35,15 @@
     Added fault friction interface conditions with static
     friction, linear slip-weakening friction, and rate- and
     state-friction with the ageing law. The implementation can be used
-    in static, quasi-static, or dynamic problems.
+    in static, quasi-static, and dynamic problems.
 
   * Drucker-Prager elastoplastic bulk rheology
 
-    Added a Drucker-Prager elastoplastic bulk rheology. This is a perfect
-    plasticity implementation (no hardening). This is a nonlinear
-    constitutive model, so the nonlinear solver is required when this
-    rheology is used. Refer to the 'Material Models' section of the manual.
+    Added a Drucker-Prager elastoplastic bulk rheology. This is a
+    perfect plasticity implementation (no hardening). This is a
+    nonlinear constitutive model, so the nonlinear solver is required
+    when this rheology is used. Refer to the 'Material Models' section
+    of the manual.
 
  * Plane strain Maxwell viscoelastic bulk rheology
 
@@ -52,7 +53,7 @@
 
     Added a finite-deformation (rigid body motion and small strains)
     implementation of elasticity with stress calculated using the
-    Second Piola Kirchhoff stress tensor and strains calculated using
+    Second Piola-Kirchhoff stress tensor and strains calculated using
     the Green-Lagrange strain tensor.
 
   * Lumped Jacobian for explicit-time stepping
@@ -67,9 +68,7 @@
     Added optimized elasticity objects for the most popular cell types
     and basis functions (linear polynomials). For tri3 and tet4 cells
     with one quadrature point, the optimized implementations do not
-    use mapped cells to reduce the number of operations. For quad4 and
-    hex8 cells the optimized objects provide hardwired loop bounds
-    which can lead to faster performance for some compilers.
+    use mapped cells to reduce the number of operations. 
 
   * Scientific notation for ASCII VTK files
 
@@ -104,5 +103,3 @@
 
     - Fixed memory bug for a fault in a 1-D mesh when constructing the
       cohesive cells.
-
-



More information about the CIG-COMMITS mailing list