<html><body style="word-wrap: break-word; -webkit-nbsp-mode: space; -webkit-line-break: after-white-space; ">
Dear Magma-philes,<div> </div><div>I am pleased to announce the updated release of the Magma Dynamics Demonstration Suite (MADDs) version FP_0.1.2</div><div><br></div><div>This is a suite of demonstration and benchmark codes for magma dynamics problems. </div><div>The FP implementation is written using a combination of FEniCS (<a href="http://www.fenics.org">www.fenics.org</a>) and PETSc (<a href="http://www.mcs-anl.gov/petsc-as/petsc">www.mcs-anl.gov/petsc-as/petsc</a>) libraries for rapid and Flexible Programming of multi-physics problems.</div><div><br></div><div>Specific Geoscience Benchmarks include:<span class="Apple-tab-span" style="white-space:pre"> </span></div><div><span class="Apple-tab-span" style="white-space:pre"> </span>* MADDs-1: High resolution Stokes flow solution for solid mantle flow in a mid-ocean ridge geometry with accurate continuous pressure.</div><div><span class="Apple-tab-span" style="white-space:pre"> </span>* MADDs-2: Post-processing for the melt flow velocity field, assuming constant porosity and the solid flow and dynamic pressure from MADDs-1</div><div><span class="Apple-tab-span" style="white-space:pre"> </span>* MADDs-4: Benchmark codes for time-dependent 2-D and 3-D non-linear solitary waves including code for generating high accuracy Solitary waves initial conditions for 1,2 and 3-D waves (from G Simpson)</div><div><span class="Apple-tab-span" style="white-space:pre"> </span>* MADDs-5a: Time dependent 2-D code for calculating porosity and melt flow evolution at mid-ocean ridges with forced adiabatic melting and the Solid flow field from MADDs-1</div><div><span class="Apple-tab-span" style="white-space:pre"> </span>* Ridgemesh: An auxiliary program for generating corner refined unstructured meshes for mid-ocean ridge corner flow (used in MADDs-1,2 and 5a)</div><div><br></div><div><div>Computational Features of MADDs-FP include:</div><div><span class="Apple-tab-span" style="white-space: pre; "> </span>* Automatic code generation from a high-level description of the PDE's (FEniCS Form Compiler) allowing ease of modifying the governing equations</div><div><span class="Apple-tab-span" style="white-space: pre; "> </span>* Automatic differentiation for calculating exact Jacobians for non-linear problems</div><div><span class="Apple-tab-span" style="white-space: pre; "> </span>* Access to the full range of linear and non-linear solvers in PETSc with command line options (in particular FieldSplit block preconditioners for multi-physics problems) </div><div><span class="Apple-tab-span" style="white-space: pre; "> </span>* Semi-Lagrangian Advection schemes on unstructured triangular/tetrahedral meshes</div><div><br></div><div>Source code can be downloaded from the CIG mercurial repositories with</div><div><br></div><div><span class="Apple-tab-span" style="white-space:pre"> </span>hg clone <a href="http://geodynamics.org/hg/magma/3D/MADDs-FP">http://geodynamics.org/hg/magma/3D/MADDs-FP</a></div><div><br></div><div><span class="Apple-tab-span" style="white-space:pre"> </span>(see <a href="http://geodynamics.org/cig/software/Repository/mercurial/">http://geodynamics.org/cig/software/Repository/mercurial/</a> or <a href="http://www.selenic.com/mercurial/wiki/">http://www.selenic.com/mercurial/wiki/</a> for information on mercurial)</div><div><span class="Apple-tab-span" style="white-space:pre"> </span>The codes will require building the current releases of FEniCS (e.g Dolfin 0.9.5 etc) and PETSc-3.0.0 and are written in C++ and Python.</div><div><span class="Apple-tab-span" style="white-space:pre"> </span>Build and run instructions are given in the documentation. </div><div><br></div><div>I will be discussing these codes briefly at the CIG business meeting at AGU (Tuesday, <a href="http://geodynamics.org/cig/upcoming/business-mtg-09">http://geodynamics.org/cig/upcoming/business-mtg-09</a>).</div><div> If you have any questions or other magma related business you would like me to discuss, please let me know.</div><div><br></div><div>Cheers</div><div>Marc</div><div><br></div><div><br></div><div> <span class="Apple-style-span" style="font-size: 12px; "><div>----------------------------------------------------</div><div>Marc Spiegelman</div><div>Lamont-Doherty Earth Observatory</div><div>Dept. of Applied Physics/Applied Math</div><div>Columbia University</div><div><a href="http://www.ldeo.columbia.edu/~mspieg">http://www.ldeo.columbia.edu/~mspieg</a></div><div>tel: 845 704 2323 (SkypeIn)</div><div>----------------------------------------------------</div><br class="Apple-interchange-newline"></span> </div><br></div></body></html>