<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
  <head>
    <meta content="text/html; charset=ISO-8859-1"
      http-equiv="Content-Type">
  </head>
  <body text="#000000" bgcolor="#ffffff">
    <tt>Jesus -<br>
      <br>
      What you most likely want to resolve is the characteristic
      pressure scale height in your problem.<br>
      <br>
      Hope this helps.<br>
      <br>
      Tomek<br>
    </tt>--<br>
    On 3/17/2011 6:20 PM, Jesús Zavala Franco wrote:
    <blockquote
      cite="mid:AANLkTimsZOOxhs4KRHt08s207BOX=JLfmVqsz_MngcQw@mail.gmail.com"
      type="cite">Hi John,
      <div><br>
      </div>
      <div>Thank you for your reply, I have tried different
        configurations and I always get the same situation: the gas
        spreading outwards with velocities of ~20 km/s after 300 Myrs of
        evolution. This happens whether I spread the dark matter
        particles or not. By the way, I tried using the point mass
        potential that is implemented as an option in FLASH as you
        suggested. So I removed the particles completely. I get exactly
        the same situation, so it is not related to the particles.<br>
        <br>
      </div>
      <div>I'm thinking that the problem is related to what Colin
        mentioned in his answer (check following e-mail)</div>
      <div><br>
      </div>
      <div>Cheers,</div>
      <div>Jesus</div>
      <div><br>
        <div class="gmail_quote">2011/3/17 John ZuHone <span dir="ltr"><<a
              moz-do-not-send="true"
              href="mailto:jzuhone@cfa.harvard.edu" target="_blank">jzuhone@cfa.harvard.edu</a>></span><br>
          <blockquote class="gmail_quote" style="margin: 0pt 0pt 0pt
            0.8ex; border-left: 1px solid rgb(204, 204, 204);
            padding-left: 1ex;">
            <div style="word-wrap: break-word;">Jesus,
              <div><br>
              </div>
              <div>I would actually not expect this setup to work all
                that well. Most certainly the central cell with the
                particles and the gas is not going to be in hydrostatic
                equilibrium. This would be true of the central cell even
                if you had the dark matter spread throughout, but in
                that case it would not be much of an issue since what
                you would get is some minor flattening of the gas
                density profile near the center.</div>
              <div><br>
              </div>
              <div>But if I read you correctly you have the particles
                all dropped into this central zone, and the mass of this
                central zone alone is much larger than the mass of gas
                in the entire system, which means each of your particles
                is very massive. What velocities do you have them set
                to? Even if you set them initially to zero and they are
                all at the center you're going to get some spurious
                velocities. Since the particles are so massive the
                potential is probably changing a lot and this is
                throwing things out of equilibrium in the center.</div>
              <div><br>
              </div>
              <div>Since the HSE is definitely broken in this cell and
                probably in a few cells surrounding it, I don't think
                it's surprising that you're seeing what you are. You are
                trying to simulate a point mass but severely
                underresolving it both in terms of spatial and mass
                resolution. </div>
              <div><br>
              </div>
              <div>Is there a particular reason you're trying to
                represent a point mass in this way? There is a point
                mass gravitational acceleration option in FLASH that
                would probably be better suited for this purpose. </div>
              <div><br>
              </div>
              <div>Best,</div>
              <div><br>
              </div>
              <font color="#888888">
                <div>John ZuHone</div>
              </font>
              <div>
                <div>
                  <div><br>
                  </div>
                  <div>
                    <div>
                      <div>On Mar 17, 2011, at 12:55 PM, Jesús Zavala
                        Franco wrote:</div>
                      <br>
                      <blockquote type="cite">Dear all,
                        <div><br>
                        </div>
                        <div>I'm having problems setting up a sphere of
                          gas in hydrostatic equilibrium under its own
                          gravity + the gravity of a particle
                          distribution.</div>
                        <div><br>
                        </div>
                        <div>
                          In the general case, I would like to have this
                          particle distribution with the same density
                          profile as the gas distribution (so I'm aiming
                          at setting up a dark matter halo with a
                          particle distribution, with gas inside), but
                          to describe the problem I'm having, I will
                          avoid distributing particles in the sphere and
                          simply:</div>
                        <div><br>
                        </div>
                        <div>1) put 1000 particles in the centre of the
                          sphere</div>
                        <div>2) put gas distributed spherical around
                          this centre with a radial density profile (a
                          NFW profile), and pressure given by the
                          condition of hydrostatic equilibrium:</div>
                        <div><br>
                        </div>
                        <div>dP(r)/dr = -(GM(<r)/r^2) * rho_gas(r)</div>
                        <div><br>
                        </div>
                        <div>where M(<r) is the total enclosed mass,
                          and rho_gas(r) is the gas density. M(<r)=
                          M_DM+Mgas(<r), with M_DM the total mass of
                          the 1000 particles, which by the way exceeds
                          the total mass of the gas by almost an order
                          of magnitude. To solve the equation I impose a
                          boundary condition outside the sphere setting
                          a pressure and a density which are reasonable
                          according to the problem and the density
                          profile I'm putting.</div>
                        <div><br>
                        </div>
                        <div>So in short, what I have is a sphere of gas
                          within a gravitational potential dominated by
                          essentially a point source in the centre and
                          with a pressure that should give support
                          agains the gravitational collapse.</div>
                        <div><br>
                        </div>
                        <div>I'm using a gird of 8^3 with 4 levels of
                          refinement, an ideal gas equation of state, a
                          Pfft Multigrid gravity solver, I'm using the
                          default operator splitting technique to
                          advance the solution.</div>
                        <div><br>
                        </div>
                        <div>What I notice is that since the first time
                          step, regardless of my choice of dtinit, the
                          cells just next to the central cell (the one
                          containing the 1000 particles), acquire a
                          radial outwards velocity, whit a size
                          depending of the time step, and that once the
                          time evolution reaches the typical times of
                          the problem (~100 Myrs), this results in the
                          gas in the inner parts propagating outwards,
                          reducing the density in the core, after 1Gyr
                          or so, this propagating gas reaches the
                          boundaries of the sphere. In other words, the
                          equilibrium set at first is broken.</div>
                        <div><br>
                        </div>
                        <div>I have tried using the option
                          ppm_modifystates=.true. since as described in
                          the user guide:</div>
                        <div><br>
                        </div>
                        <div>
                          <div style="margin: 0px;">"The version of PPM
                            in the FLASH code has an option to more
                            closely couple the hydrodynamic solver</div>
                          <div style="margin: 0px;">with a gravitational
                            source term. This can noticeably reduce
                            spurious velocities caused by the operator</div>
                          <div style="margin: 0px;">
                            splitting of the gravitational acceleration
                            from the hydrodynamics"</div>
                        </div>
                        <div><br>
                        </div>
                        <div>I thought this could help, but it didn't.
                          I'm aware that if I wouldn't put the particles
                          in the centre, this behaviour is just a
                          resolution problem since the core is not being
                          properly resolved and the density and enclosed
                          mass is underestimated there, the pressure is
                          therefore too high (since is initially set by
                          solving the hydrostatic equilibrium equation
                          with the assumed analytical profile) and it
                          pushed the gas outwards. However, since I'm
                          setting a significant gravity source in the
                          center I wouldn't expect this to be the case.
                          I have tried actually increasing the mass of
                          the particles by a factor of 10, and not
                          considering this mass increase in the solution
                          to the pressure equation, this sets the
                          overall pressure too low and you would expect
                          a collapse. Even though when at first the
                          velocities of the contiguous cells to the
                          centre, point inwards, after a while, they are
                          overwhelmed by an outward flow that develops
                          in the cells next to these, and to my
                          surprise, the sphere expands as well.</div>
                        <div><br>
                        </div>
                        <div>I'm running out of ideas on the initial
                          conditions setup, so I'm thinking this could
                          either be a resolution issue, or a bad choice
                          of hydro, gravity solver.</div>
                        <div><br>
                        </div>
                        <div>Any hel will be much appreciated. </div>
                        <div><br>
                        </div>
                        <div>Cheers,</div>
                        <div>Jesus Zavala</div>
                        <div>Department of Physics and Astronomy</div>
                        <div>University of Waterloo</div>
                      </blockquote>
                    </div>
                    <br>
                  </div>
                </div>
              </div>
            </div>
          </blockquote>
        </div>
        <br>
      </div>
    </blockquote>
  </body>
</html>