Hi,
We had been using our own version of TwoPunctures, and I’ve recently switched to the ETK version since it doesn’t require solving for the masses separately. Junk radiation has always been present, but I’m finding that some of my recent simulations have particularly loud junk radiation, which also makes it harder avoid reflection of. So I’m guessing I’m not fully understanding the TwoPunctures parameters.
I’ve attached a parfile and a psi4 plot for a simulation I’ve started that’s having this large junk radiation, and was hoping someone could point out any mistakes I may have in the TwoPunctures parameters.
Thanks! Deborah
-------------------------------------- Deborah Ferguson Post Doc University of Texas at Austin
Deborah
One one hand, TwoPunctures has parameters to choose the resolution of the spectral grid it uses. I assume you are aware of these, and that you also know how to choose the 3D AMR grid setup parameters.
TwoPunctures only solves for the conformal factor. It does per se not directly choose a lapse and a shift, but their choice is important in practice. Choosing \alpha=1 and \beta^i=0 as initial conditions is often not good. Similarly, choosing the antisymmetry lapse (with \alpha=-1 at the puncture) is also a bad idea (it is ill-posed). The TwoPunctures toolkit in the Einstein Toolkit has several options for choosing the initial gauge conditions. Personally, I am partial to initial_lapse = "twopunctures-averaged", but I am sure that others have other preferences.
Of course, the BSSN gauge parameters used during evolution will also play a role.
You can look at the gallery example at http://einsteintoolkit.org/gallery/bbh/index.html for ideas.
-erik
On Tue, Aug 9, 2022 at 4:47 PM Ferguson, Deborah deborah.ferguson@austin.utexas.edu wrote:
Hi,
We had been using our own version of TwoPunctures, and I’ve recently switched to the ETK version since it doesn’t require solving for the masses separately. Junk radiation has always been present, but I’m finding that some of my recent simulations have particularly loud junk radiation, which also makes it harder avoid reflection of. So I’m guessing I’m not fully understanding the TwoPunctures parameters.
I’ve attached a parfile and a psi4 plot for a simulation I’ve started that’s having this large junk radiation, and was hoping someone could point out any mistakes I may have in the TwoPunctures parameters.
Thanks!
Deborah
Deborah Ferguson
Post Doc
University of Texas at Austin
Users mailing list Users@einsteintoolkit.org http://lists.einsteintoolkit.org/mailman/listinfo/users
Hi Deborah,
I'd be curious to see a comparison between the junk radiation you're seeing now versus before, for the same physical scenario, grids, etc. Also, the extra wiggles/reflections are typically caused not by junk radiation but by a sharp gauge feature (https://arxiv.org/abs/1404.6523). The two can be easily distinguished as the former propagates at c, and the latter at sqrt(2) c (see paper for details).
-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 Tue, Aug 9, 2022 at 4:41 PM Erik Schnetter schnetter@gmail.com wrote:
Deborah
One one hand, TwoPunctures has parameters to choose the resolution of the spectral grid it uses. I assume you are aware of these, and that you also know how to choose the 3D AMR grid setup parameters.
TwoPunctures only solves for the conformal factor. It does per se not directly choose a lapse and a shift, but their choice is important in practice. Choosing \alpha=1 and \beta^i=0 as initial conditions is often not good. Similarly, choosing the antisymmetry lapse (with \alpha=-1 at the puncture) is also a bad idea (it is ill-posed). The TwoPunctures toolkit in the Einstein Toolkit has several options for choosing the initial gauge conditions. Personally, I am partial to initial_lapse = "twopunctures-averaged", but I am sure that others have other preferences.
Of course, the BSSN gauge parameters used during evolution will also play a role.
You can look at the gallery example at http://einsteintoolkit.org/gallery/bbh/index.html for ideas.
-erik
On Tue, Aug 9, 2022 at 4:47 PM Ferguson, Deborah deborah.ferguson@austin.utexas.edu wrote:
Hi,
We had been using our own version of TwoPunctures, and I’ve recently
switched to the ETK version since it doesn’t require solving for the masses separately. Junk radiation has always been present, but I’m finding that some of my recent simulations have particularly loud junk radiation, which also makes it harder avoid reflection of. So I’m guessing I’m not fully understanding the TwoPunctures parameters.
I’ve attached a parfile and a psi4 plot for a simulation I’ve started
that’s having this large junk radiation, and was hoping someone could point out any mistakes I may have in the TwoPunctures parameters.
Thanks!
Deborah
Deborah Ferguson
Post Doc
University of Texas at Austin
Users mailing list Users@einsteintoolkit.org http://lists.einsteintoolkit.org/mailman/listinfo/users
-- Erik Schnetter schnetter@gmail.com http://www.perimeterinstitute.ca/personal/eschnetter/ _______________________________________________ Users mailing list Users@einsteintoolkit.org http://lists.einsteintoolkit.org/mailman/listinfo/users
Hello Deborah, all,
Having used both TwoPunctures codes in the past: have you eg dumped the TP coefficients of both codes to disk after a solve to compare those? The two may have quite different ways of specifying the mass (e.g. one can hunt for a given mass on its own, the other uses Baum).
The quantities to output for comparison would be the v.d0 array in src/TwoPunctures.c .
Yours, Roland
Hi Deborah,
I'd be curious to see a comparison between the junk radiation you're seeing now versus before, for the same physical scenario, grids, etc. Also, the extra wiggles/reflections are typically caused not by junk radiation but by a sharp gauge feature (https://urldefense.com/v3/__https://arxiv.org/abs/1404.6523__;!!DZ3fjg!8pjNi... ). The two can be easily distinguished as the former propagates at c, and the latter at sqrt(2) c (see paper for details).
-Zach
Zachariah Etienne Assoc. Prof. of Physics, U. of Idaho Adjunct Assoc. Prof. of Physics & Astronomy, West Virginia U. https://urldefense.com/v3/__https://etienneresearch.com__;!!DZ3fjg!8pjNitnVp... https://urldefense.com/v3/__https://blackholesathome.net__;!!DZ3fjg!8pjNitnV...
On Tue, Aug 9, 2022 at 4:41 PM Erik Schnetter schnetter@gmail.com wrote:
Deborah
One one hand, TwoPunctures has parameters to choose the resolution of the spectral grid it uses. I assume you are aware of these, and that you also know how to choose the 3D AMR grid setup parameters.
TwoPunctures only solves for the conformal factor. It does per se not directly choose a lapse and a shift, but their choice is important in practice. Choosing \alpha=1 and \beta^i=0 as initial conditions is often not good. Similarly, choosing the antisymmetry lapse (with \alpha=-1 at the puncture) is also a bad idea (it is ill-posed). The TwoPunctures toolkit in the Einstein Toolkit has several options for choosing the initial gauge conditions. Personally, I am partial to initial_lapse = "twopunctures-averaged", but I am sure that others have other preferences.
Of course, the BSSN gauge parameters used during evolution will also play a role.
You can look at the gallery example at <https://urldefense.com/v3/__http://einsteintoolkit.org/gallery/bbh/index.htm... > for ideas.
-erik
On Tue, Aug 9, 2022 at 4:47 PM Ferguson, Deborah deborah.ferguson@austin.utexas.edu wrote:
Hi,
We had been using our own version of TwoPunctures, and I’ve recently
switched to the ETK version since it doesn’t require solving for the masses separately. Junk radiation has always been present, but I’m finding that some of my recent simulations have particularly loud junk radiation, which also makes it harder avoid reflection of. So I’m guessing I’m not fully understanding the TwoPunctures parameters.
I’ve attached a parfile and a psi4 plot for a simulation I’ve started
that’s having this large junk radiation, and was hoping someone could point out any mistakes I may have in the TwoPunctures parameters.
Thanks!
Deborah
Deborah Ferguson
Post Doc
University of Texas at Austin
Users mailing list Users@einsteintoolkit.org https://urldefense.com/v3/__http://lists.einsteintoolkit.org/mailman/listinf...
-- Erik Schnetter schnetter@gmail.com https://urldefense.com/v3/__http://www.perimeterinstitute.ca/personal/eschne... _______________________________________________ Users mailing list Users@einsteintoolkit.org https://urldefense.com/v3/__http://lists.einsteintoolkit.org/mailman/listinf...
Yours, Roland
users@lists.einsteintoolkit.org
-
Erik Schnetter -
Ferguson, Deborah -
Roland Haas -
Zach Etienne