[cig-commits] commit: Merge
Mercurial
hg at geodynamics.org
Wed Aug 29 19:41:43 PDT 2012
changeset: 148:b3e79c713b83
tag: tip
parent: 147:4ee9a2c82c91
parent: 146:9694227b0dc1
user: Matthew G. Knepley <knepley at gmail.com>
date: Wed Aug 29 21:41:37 2012 -0500
files: faultRup.tex
description:
Merge
diff -r 4ee9a2c82c91 -r b3e79c713b83 faultRup.tex
--- a/faultRup.tex Wed Aug 29 21:41:09 2012 -0500
+++ b/faultRup.tex Wed Aug 29 21:41:37 2012 -0500
@@ -47,6 +47,16 @@
% ==================================================================
\begin{document}
+
+\begin{quote}{\it
+ This information is distributed solely for the purpose of
+ predissemination peer review and must not be disclosed, released, or
+ published until after approval by the U.S. Geological Survey
+ (USGS). It is deliberative and predecisional information and the
+ findings and conclusions in the document have not been formally
+ approved for release by the USGS. It does not represent and should
+ not be construed to represent any USGS determination or policy.
+}\end{quote}
\title{A Domain Decomposition Approach to Implementing Fault Slip in
Finite-Element Models of Quasi-static and Dynamic Crustal
@@ -152,8 +162,8 @@ found several effects that required reso
found several effects that required resolving both the interseismic
deformation and the rapid slip during dynamic rupture; the
low-rigidity layers reduced the nucleation size, amplified slip rates
-during dynamic rupture, increased the recurrent interval, and reduced
-the amount of aseismic slip
+during dynamic rupture, increased the recurrence interval, and reduced
+the amount of aseismic slip.
Collectively, these studies suggest a set of desirable features for
models of the earthquake cycle in order to capture both the slow
@@ -475,8 +485,9 @@ the action on the increment in the solut
the action on the increment in the solution, just as we did for the
quasi-static simulations. In this case we associate the increment in
the solution with the temporal discretization. We march forward in
-time using explicit time stepping via Newmark's method with a central
-difference scheme wherein the acceleration and velocity are given by
+time using explicit time stepping via Newmark's method
+\citep{Newmark:1959} with a central difference scheme wherein the
+acceleration and velocity are given by
\begin{gather}
\frac{\partial^2 \bm{u}}{\partial t^2}(t) =
\frac{1}{\Delta t^2} \left(
@@ -508,7 +519,7 @@ finite-element formulations.
Earthquake ruptures in which the slip has a short rise time tend to
introduce deformation at short length scales (high frequencies) that
-the numerical models cannot resolve accurately. This is especially true
+numerical models cannot resolve accurately. This is especially true
in spontaneous rupture simulations, because the rise time is sensitive
to the evolution of the fault rupture. In order to reduce the
introduction of deformation at such short length scales we add
@@ -572,7 +583,7 @@ a constitutive model controls the tracti
a constitutive model controls the tractions on the fault surface. The
fault slip evolves based on the fault tractions as driven by the
initial conditions, boundary conditions and deformation. In our
-formulation of fault slip, fault slip is assumed to be known and the
+formulation of fault slip, slip is assumed to be known and the
fault tractions (Lagrange multipliers) are part of the solution
(unknowns). The fault constitutive model places bounds on the Lagrange
multipliers and the system of equations is nonlinear-- when a location
@@ -868,7 +879,7 @@ level operators and interpolation matric
level operators and interpolation matrices using only the system
matrix, treated as a weighted graph, rather than a separate
description of the problem geometry, such as a mesh. We split the
-elastic block from the fault block, and also to manage Schur
+elastic block from the fault block and also manage the Schur
complements. In this way, all block preconditioners, including those
nested with multigrid, can be controlled from the options file without
recompilation or special code.
@@ -1220,7 +1231,7 @@ case) is almost completely controlled by
case) is almost completely controlled by the available memory
bandwidth. Good illustrations of the memory system performance are
given by the \texttt{VecAXPY}, \texttt{VecMAXPY} and \texttt{VecMDot}
-operations reported in the log summary \cite{PETSc:manual}. These
+operations reported in the log summary \citep{PETSc:manual}. These
operations are limited by available memory bandwidth rather than the
rate at which a processor can perform floating points operations. From
Table~\ref{tab:solvertest:memory:events}, we see that we saturate the
@@ -1606,7 +1617,7 @@ rupture propagation.
\begin{figure*}[h]
\noindent\includegraphics{figs/sieve}
\caption{Direct acyclic graph representations of the covering
- relation for 2-D meshes with tri3 and quad4 cells. The graphs for
+ relation for 2-D meshes with triangular and quadrilateral cells. The graphs for
interpolated meshes (a) and (b) include all levels of the topology
whereas the graphs for optimized meshes (c) and (d) only include
the top and bottom levels. The graphs for interpolated meshes in
@@ -2072,7 +2083,7 @@ VecMAXPY & 1 & 1.0 & 1733 \\
\hline
\end{tabular}
\tablenotetext{a}{Basic simulation parameters for the SCEC
- dynamic spontaneous rupture benchmark TPV13. Full specification of
+ dynamic spontaneous rupture benchmark TPV13. A complete list of
the parameters can be found in \citet{Harris:etal:SRL:2011}.}
\end{table}
diff -r 4ee9a2c82c91 -r b3e79c713b83 figs/cohesivecell.pdf
Binary file figs/cohesivecell.pdf has changed
diff -r 4ee9a2c82c91 -r b3e79c713b83 figs/sieve.tex
--- a/figs/sieve.tex Wed Aug 29 21:41:09 2012 -0500
+++ b/figs/sieve.tex Wed Aug 29 21:41:37 2012 -0500
@@ -26,7 +26,7 @@
\colorlet{coledge}{blue}
\colorlet{colcell}{red}
-\tikzstyle{title}=[text width=55mm,text centered,font=\bfseries\large]
+\tikzstyle{title}=[text width=62mm,text centered,font=\bfseries\large]
\tikzstyle{annotate}=[font=\bfseries]
\tikzstyle{vertex}=[circle,color=black,fill=colvertex,inner sep=1.5pt]
@@ -52,7 +52,7 @@
% ----------------------------------------------------------------------
% Interpolated tri mesh
-\node[title] at ($ (o1)+(0,15mm) $) {(a) Interpolated tri3 mesh};
+\node[title] at ($ (o1)+(0,15mm) $) {(a) Interpolated triangular mesh};
% Vertices
\node[vertex] (a_p7) at ($ (o1)+(-15mm,+0mm) $) {}; \node[vlabel,left] at (a_p7) {7};
@@ -113,9 +113,9 @@
\draw[->] (a_g6.south) -- (a_g1.north);
% ----------------------------------------------------------------------
-% Interpolated quad4 mesh
+% Interpolated quadrilateral mesh
-\node[title] at ($ (o2)+(0,15mm) $) {(b) Interpolated quad4 mesh};
+\node[title] at ($ (o2)+(0,15mm) $) {(b) Interpolated quadrilateral mesh};
% Vertices
\node[vertex] (b_p9) at ($ (o2)+(-15mm,+7.5mm) $) {}; \node[vlabel,left] at (b_p9) {9};
@@ -190,7 +190,7 @@
% ----------------------------------------------------------------------
% Optimized tri mesh
-\node[title] at ($ (o3)+(0,15mm) $) {(c) Optimized tri3 mesh};
+\node[title] at ($ (o3)+(0,15mm) $) {(c) Optimized triangular mesh};
% Vertices
\node[vertex] (c_p2) at ($ (o3)+(-15mm,+0mm) $) {}; \node[vlabel,left] at (c_p2) {2};
@@ -232,9 +232,9 @@
\draw[->] (c_g5.south) -- (c_g1.north);
% ----------------------------------------------------------------------
-% Optimized quad4 mesh
+% Optimized quadrilateral mesh
-\node[title] at ($ (o4)+(0,15mm) $) {(d) Optimized quad4 mesh};
+\node[title] at ($ (o4)+(0,15mm) $) {(d) Optimized quadrilateral mesh};
% Vertices
\node[vertex] (d_p2) at ($ (o4)+(-15mm,+7.5mm) $) {}; \node[vlabel,left] at (d_p2) {2};
diff -r 4ee9a2c82c91 -r b3e79c713b83 figs/solvertest_geometry.pdf
Binary file figs/solvertest_geometry.pdf has changed
diff -r 4ee9a2c82c91 -r b3e79c713b83 figs/tpv13_geometry.pdf
Binary file figs/tpv13_geometry.pdf has changed
More information about the CIG-COMMITS
mailing list