[Users] Eccentricity in BNS evolution with Lorene

[Bing-Jyun Tsao]

Hi Zach and Samuel,

Thank you both for the great tips. To comment on some of the suggestions:

1. Resolution: I tested with a higher amount of resolution when I first
encountered the problem, but for that particular case I still found the
eccentricity. However, I do agree that resolution could not be ruled out as
it might still not be high enough.

2. NS tracking grids: I used VisIt to visualize the grids following the two
stars and they seem to cover most of the stars until they are tidally
disrupted. The boxes are only slightly (1.2x) bigger than the star so I
will increase the radius to what Zach recommended.

3. Initial data being too close: I was also wondering if 45 km was too
close as it is about 3 * R_star. It is reassuring to hear from Samuel's
test with FUKA that ADM angular momentum is consistent with PN results for
a nearly same case.

In summary, I will try a few simulations with much higher resolution,
bigger tracking grids to see if the problem persists. I will also make sure
that I am measuring the ADM angular momentum correctly when comparing the
PN and Lorene ID.

Again, thanks for the help. I will keep this thread updated
whether it works out or not.

Best,
Johnny Tsao


On Thu, Oct 6, 2022 at 12:21 PM Samuel Tootle <tootle at itp.uni-frankfurt.de>
wrote:

> Hi Johnny,
>
> This is quite interesting given my understanding is that the inspiral for
> equal mass Lorene ID is usually quite reasonable. Specifically your note
> regarding the 20% difference between ADM Angular momentum for
> quasi-equilibrium and that resulting from your PN estimate ID seems quite
> wrong. Maybe the Lorene ID is at a poor resolution?
>
> I did a test with the FUKA BNS code for a 1.4-1.4 BNS using a polytope at
> 45km and the difference in the ADM Angular momentum between a coarse
> solution using quasi-equilibrium and 3.5PN estimates was < 1%.
>
> Cheers,
> Samuel Tootle
> Goethe Universität Frankfurt
>
> *From: *Zach Etienne <zachetie at gmail.com>
> *To: *Bing-Jyun Tsao <johnny.tsao.880724 at gmail.com>
> *CC: *users at einsteintoolkit.org
> *Date: *Oct 5, 2022 21:02:01
> *Subject: *Re: [Users] Eccentricity in BNS evolution with Lorene
>
> Hi Johnny,
>
> I noticed you set
> ADMBase::initial_shift           = "zero"
> so you eliminated one possible contributor to coordinate eccentricity
> (Lorene's initial shift condition results in significant eccentricity).
> Good choice.
>
> I don't use GRHydro/ML_BSSN, so I cannot comment on your choices there.
> Generic advice:
> * Resolution may be too low -- try higher resolution & see if your
> eccentricity reduces.
> * Initial data is too close (Lorene assumes a helical Killing vector,
> which is violated more and more as the initial separation of the stars
> decreases)
> * Check that you are properly tracking the NSs such that they are centered
> within high-resolution AMR boxes, and that the radius of the AMR boxes is
> at least 1.5x the radius of each NS. A movie visualizing the density of the
> stars in the orbital plane would be very very useful.
>
> Hope this helps!
>
> -Zach
>
> *     *     *
> Zachariah Etienne
> Assoc. Prof. of Physics, U. of Idaho
> Adjunct Assoc. Prof. of Physics & Astronomy, West Virginia U.
> https://etienneresearch.com
> https://blackholesathome.net
>
>
> On Wed, Oct 5, 2022 at 11:41 AM Bing-Jyun Tsao <
> johnny.tsao.880724 at gmail.com> wrote:
>
>> Dear Einstein Toolkit community,
>>
>> I am Bing-Jyun (Johnny) Tsao, a graduate student at University of Texas
>> at Austin. I am currently working on a project comparing our local version
>> of BNS initial data (M. Clark, P. Laguna, 2016 Physical Review D 94 064058)
>> with Lorene.
>>
>> When using Lorene, I found that the trajectory of the stars is showing
>> eccentricity (as shown in the plot below), and this occurred when I ran
>> Lorene with equal-mass BNS with ADM mass = 1.4 Msun, and with the
>> publicly available Lorene data on their website "G2_I14vs14_D4R33_45km",
>> both of which have an initial separation of 45 km. I use
>> VolumeIntegral_GRMHD to track the stars, and GRHydro + ML_BSSN to evolve.
>> Additionally, I also found that the ADM angular momentum from Lorene is
>> about 20% smaller than that from our code which uses post-Newtonian
>> calculations.
>>
>> My question is:
>> From my understanding, Lorene gives quasi-circular initial data. Thus,
>> is it abnormal to see eccentricity, or is it a physical artifact, perhaps
>> from tidal deformation, that always shows up when the stars are very close
>> to each other?
>>
>> Here I attached my parfile and the trajectory for a run using the Lorene
>> publicly available data "G2_I14vs14_D4R33_45km".
>>
>> Thanks in advance.
>>
>> Best,
>> Johnny Tsao
>>
>> _______________________________________________
>> Users mailing list
>> Users at einsteintoolkit.org
>> http://lists.einsteintoolkit.org/mailman/listinfo/users
>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.einsteintoolkit.org/pipermail/users/attachments/20221006/661b7880/attachment-0001.html