From: Roland Haas rhaas@illinois.edu
Sent: Tuesday, March 3, 2020 6:29 PM To: Hamilton, Maria Cc: Peter Diener Subject: Re: Questions about the nsnstohmns.par
Hello Maria,
- The parameter has refinement levels, and regrid radius, but
does not have time_refinement_factors. Is this on purpose? Should it be there?
Without setting time_refinement_factors it uses the default which is a factor of 2 between each level. Many of the BBH parfiles use a factor of 1 between the coarsest level, however for NS this is often not required (the differing factor is only there to cure an instability in the shift condition).
- CarpetRegrid2 is given 3 num_centers, if first and second are
the NSs, what is the role of the third, why does it have only 1 num_level, and why is not centered at 0?
The third one is created without any boxes. I will contain the merged star / black hole once it forms. The Triger thorn (see your question 4) is used to enable refinement levels for this center once a horizon is about to form (lapse drops below a given value). Since it is only a placeholder initially it is set to 1 level and since it will capture the black hole that forms at the origin it is centered on the origin.
- CarpetRegrid2 is given position_x_1 and position_x_2, if those
are the centers for the NSs, why is 15.1875 for a 45 radius? If I want the radius to be 50, how do I change this?
45 is in km. So 45km is ~30Msun distance which gives a radius of ~15Msun (Msun being the code unit as well).
- Why do I need Trigger?
Trigger is needed to enable grid changes once a horizon is about to form.
- If I want to search for a BH as final result, how do I add
AHFinderDirect?
Isn't it already there? If not then you can add it as usual but set eg the parameter "find_individual_after[n]" of the horizon to some huge value (1e6 or so) and have the Trigger thorn steer that value to "0." once it enables the central grid patch so that the horizonfinder starts looking for horizons (there may also be a more direct "enable" parameter listed in AHFinderDirect's param.ccl that you could use instead of hacking find_invidual_after, or whatever its name actually is.
- SphericalSurface is given 5 surface, but surface [2] is
skipped. Why?
That one was likely intended / used for a black hole horizon. There is no reason for the set of used surfaces being contiguous really.
Thank you for your time, and I hope you won't mind my questions.
No worries. As usual I would like to encourage these question to be asked on the users mailing list so that others can contribute (and benefit from the answers eventually).
Yours, Roland
-- My email is as private as my paper mail. I therefore support encrypting and signing email messages. Get my PGP key from http://pgp.mit.edu .