<div dir="ltr">I'm not sure I can use a periodic boundary condition. Could I use another solver like tree gravity instead? <div><br></div><div>I have read that the only restriction is that NBX = NBY = NBZ and it works with AMR structure.</div><div><br></div><div><br></div><div><br></div><div>Fernanda Clever</div></div><div class="gmail_extra"><br><div class="gmail_quote">2018-07-05 16:49 GMT-05:00 Klaus Weide <span dir="ltr"><<a href="mailto:klaus@flash.uchicago.edu" target="_blank">klaus@flash.uchicago.edu</a>></span>:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><span class="">On Thu, 5 Jul 2018, María Fernanda Clever Uribe wrote:<br>
<br>
> I am trying to setup a problem that requires using RHD and self-gravity,<br>
> and I have a problem while running the simulation.<br>
> <br>
> I am using 2D cartesian coordinates, and the multigrid solver which is<br>
> specified in Config file:<br>
> <br>
> physics/Gravity/GravityMain/<wbr>Poisson/Multigrid<br>
> <br>
> I compiled using the following line:<br>
> <br>
> ./setup JetGravity -auto -2d +noDefaultMpole +gravMgrid -maxblocks=2000<br>
> <br>
> The error message is:<br>
> <br>
> abort_message: gr_isoMpoleInit: FATAL: unsupported geometry!<br>
> <br>
> I do not understand why there is an error. Is it due to geometry, since the<br>
> multigrid solver should solve gravity with this kind of geometry?<br>
<br>
</span>I think you will find that the problem comes from the combination of 2D <br>
multigrid with an "isolated" gravity boundary condition. Multigrid with <br>
this boundary condition uses the auxiliary multipole solver in <br>
Grid/GridSolvers/<wbr>IsoBndMultipole to "compute the boundary values of the <br>
image mass potential"; this is currently only supported in 3D Cartesian <br>
geometry.<br>
<br>
If you can use a different gravity boundary condition ("periodic", most <br>
likely), that should work.<br>
<span class="HOEnZb"><font color="#888888"><br>
Klaus</font></span></blockquote></div><br></div>