1. Please reply to the list rather than me personally.<div><br></div><div>2. Upgrade to petsc-3.3 for GAMG. You can also use ML or BoomerAMG with older versions of PETSc.</div><div><br></div><div>3. Send full convergence output.</div>
<div><br></div><div>4. Lots of general tips here: <a href="http://scicomp.stackexchange.com/questions/513/why-is-my-iterative-linear-solver-not-converging">http://scicomp.stackexchange.com/questions/513/why-is-my-iterative-linear-solver-not-converging</a><br>
</div><div class="gmail_extra"><br><br><div class="gmail_quote">On Mon, Nov 12, 2012 at 11:10 AM, Rob Kunz <span dir="ltr"><<a href="mailto:rfk102@arl.psu.edu" target="_blank">rfk102@arl.psu.edu</a>></span> wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">





<div lang="EN-US" link="blue" vlink="purple">
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1f497d">Hi Jed:<u></u><u></u></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1f497d">Thanks for your response.<u></u><u></u></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1f497d">I am currently doing all PETSC call from C.<u></u><u></u></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1f497d">Attached is the code snippet I am using based on ex54.c.html.<u></u><u></u></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1f497d">I get an error that PCGAMG is undefined.<u></u><u></u></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1f497d">Must I do something on the command line?<u></u><u></u></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1f497d">Thanks<u></u><u></u></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1f497d">Rob<u></u><u></u></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1f497d"><u></u> <u></u></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1f497d"><u></u> <u></u></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1f497d">Robert F. Kunz<u></u><u></u></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1f497d">Head, Computational Mechanics Division, Applied Research Laboratory<u></u><u></u></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1f497d">Professor, Aerospace Engineering<u></u><u></u></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1f497d">The Pennsylvania State University<u></u><u></u></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1f497d">PO Box 30<u></u><u></u></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1f497d">University Park, PA 16804<u></u><u></u></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1f497d"><a href="mailto:rfk102@arl.psu.edu" target="_blank">rfk102@arl.psu.edu</a><u></u><u></u></span></p>

<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1f497d"><a href="tel:814-360-1901" value="+18143601901" target="_blank">814-360-1901</a><u></u><u></u></span></p>

<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1f497d"><u></u> <u></u></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1f497d"><u></u> <u></u></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1f497d"><u></u> <u></u></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1f497d"><u></u> <u></u></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1f497d"><u></u> <u></u></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1f497d"><u></u> <u></u></span></p>
<div style="border:none;border-top:solid #b5c4df 1.0pt;padding:3.0pt 0in 0in 0in">
<p class="MsoNormal"><b><span style="font-size:10.0pt;font-family:"Tahoma","sans-serif"">From:</span></b><span style="font-size:10.0pt;font-family:"Tahoma","sans-serif""> <a href="mailto:petsc-users-bounces@mcs.anl.gov" target="_blank">petsc-users-bounces@mcs.anl.gov</a> [mailto:<a href="mailto:petsc-users-bounces@mcs.anl.gov" target="_blank">petsc-users-bounces@mcs.anl.gov</a>]
<b>On Behalf Of </b>Jed Brown<br>
<b>Sent:</b> Monday, November 12, 2012 11:31 AM<br>
<b>To:</b> PETSc users list<br>
<b>Subject:</b> Re: [petsc-users] Parallel vs. single processor<u></u><u></u></span></p>
</div><div><div class="h5">
<p class="MsoNormal"><u></u> <u></u></p>
<p class="MsoNormal">Try (e.g., -pc_type gamg) and additive Schwarz. Also try using ASM as a smoother.<u></u><u></u></p>
<div>
<p class="MsoNormal" style="margin-bottom:12.0pt"><u></u> <u></u></p>
<div>
<p class="MsoNormal">On Mon, Nov 12, 2012 at 10:21 AM, Rob Kunz <<a href="mailto:rfk102@arl.psu.edu" target="_blank">rfk102@arl.psu.edu</a>> wrote:<u></u><u></u></p>
<div>
<div>
<p class="MsoNormal">Hi:<u></u><u></u></p>
<p class="MsoNormal">I have a pressure based CFD code using PETSC for nearly SPD pressure equation.<u></u><u></u></p>
<p class="MsoNormal">PETSC default solver + preconditioner (GMRES+BJ for n-proc) is very reliable for me.<u></u><u></u></p>
<p class="MsoNormal">However I have a very tough bearing problem with extremely high aspect ratio cells and the 1-processor run works fine (GMRES+ILU0), but for nproc>1 the code does not converge.<u></u><u></u></p>
<p class="MsoNormal">Of course I can get the code to return identical 1-proc and 12-proc convergence histories if I choose a point preconditioner like Jacobi, but that kills the convergence of both.<u></u><u></u></p>
<p class="MsoNormal">Any recommendations?<u></u><u></u></p>
<p class="MsoNormal">Thanks<span style="color:#888888"><br>
<span>Rob Kunz</span></span><u></u><u></u></p>
</div>
</div>
</div>
<p class="MsoNormal"><u></u> <u></u></p>
</div>
</div></div></div>
</div>

</blockquote></div><br></div>