<html>
  <head>
    <meta content="text/html; charset=ISO-8859-1"
      http-equiv="Content-Type">
  </head>
  <body bgcolor="#FFFFFF" text="#000000">
    Thank you very much!<br>
    I will make some tests and let you know how it goes.<br>
    <br>
    Michele<br>
    <br>
    <div class="moz-cite-prefix">On 09/23/2013 01:24 PM, Mark F. Adams
      wrote:<br>
    </div>
    <blockquote cite="mid:B92918FB-21D0-49D5-983C-6C9D80F9D0AC@lbl.gov"
      type="cite">
      <meta http-equiv="Content-Type" content="text/html;
        charset=ISO-8859-1">
      Ah yes, I agree, geometric MG with Galerkin is worth trying. 
      <div><br>
      </div>
      <div><br>
        <div>
          <div>On Sep 23, 2013, at 3:36 PM, Dave May <<a
              moz-do-not-send="true"
              href="mailto:dave.mayhem23@gmail.com">dave.mayhem23@gmail.com</a>>
            wrote:</div>
          <br class="Apple-interchange-newline">
          <blockquote type="cite">
            <div dir="ltr">
              <div>Hey Michele,<br>
                <br>
                I think Galerkin MG, even with a jump in the coefficient
                of 10^3, can be made more robust and faster for your
                problem.<br>
              </div>
              I'd start by trying different smoothers. Try
              cheby/(bjacobi or asm), or gmres/(bjacobi or asm). When
              using cheby, make sure you ask for the spectrum estimate.<br>
              Definitely do more than one application of KSP on each
              level.<br>
              <div><br>
                <div>I don't think you should switch to AMG without
                  first trying more robust smoothers and profiling how
                  much time is spent in the coarse grid solve.<br>
                </div>
                <div>I solve similar problems (geometry and size of the
                  jump) in connection with variable visocsity stokes
                  problems on structured grids, and I'm yet to see a
                  case with AMG is more robust and or faster than
                  Galerkin MG.<br>
                </div>
                <br>
                <br>
              </div>
              <div>Cheers,<br>
              </div>
              <div>  Dave<br>
              </div>
              <div> <br>
              </div>
            </div>
            <div class="gmail_extra"><br>
              <br>
              <div class="gmail_quote">On 23 September 2013 20:52, Mark
                F. Adams <span dir="ltr"><<a moz-do-not-send="true"
                    href="mailto:mfadams@lbl.gov" target="_blank">mfadams@lbl.gov</a>></span>
                wrote:<br>
                <blockquote class="gmail_quote" style="margin:0 0 0
                  .8ex;border-left:1px #ccc solid;padding-left:1ex">
                  <div class="im"><br>
                    >><br>
                    > The simulation does start with both phases and
                    the geometry is supposed to become more complex as
                    the simulation progresses.<br>
                    > But so far the run is stopped before there are
                    significant changes in the shape of the droplet.<br>
                    <br>
                  </div>
                  Humm,  not sure why you are seeing degradation then.
                   I imagine the initial geometry is not grid aligned so
                  I'm not sure why you are seeing degradation in
                  convergence rate.<br>
                  <div class="im"><br>
                    > I can give a shot to AMG: which options would
                    you suggest to use.<br>
                    <br>
                  </div>
                  -pc_type gamg -pc_gamg_type agg -pc_gamg_agg_nsmooths
                  1<br>
                  <div class="im"><br>
                    > Also, how can I project out the constant from
                    the rhs? Thanks a lot!<br>
                    <br>
                  </div>
                  <a moz-do-not-send="true"
href="http://www.mcs.anl.gov/petsc/petsc-current/docs/manualpages/KSP/KSPSetNullSpace.html#KSPSetNullSpace"
                    target="_blank">http://www.mcs.anl.gov/petsc/petsc-current/docs/manualpages/KSP/KSPSetNullSpace.html#KSPSetNullSpace</a><br>
                  <br>
                </blockquote>
              </div>
              <br>
            </div>
          </blockquote>
        </div>
        <br>
      </div>
    </blockquote>
    <br>
  </body>
</html>