Hello EinsteinToolkit-Community,
first of all i want thank you very much for the fast and helpful answer to my last Email.
I mentioned i'm at the beginning of my phd. So before i'm wasting too much time on working at something an other already knows, i want to ask some perhaps a little bit naively sounding questions.
I have studied smoothed particle hydrodynamics on a fixed spacetime, so i get the idea devolping a sph based thorn. I found no publications, so is perhaps anybody already working on this problem, or has one shown that this isn't possible? Because this is an community based development, it is quite interesting whether it could be an useful thorn.
I'm not sure whether my questions are answerable, but perhaps you can help me a second time.
Thank you very much, Stefan Ruehe
Hi Stefan
On Wed, Jan 15, 2014 at 03:05:05PM +0100, sruehe@astrophysik.uni-kiel.de wrote:
I have studied smoothed particle hydrodynamics on a fixed spacetime, so i get the idea devolping a sph based thorn. I found no publications, so is perhaps anybody already working on this problem, or has one shown that this isn't possible?
I am not aware of an SPH implementation in Cactus, at least not in recent years. It would certainly be interesting, but it would also pose some challenges. One of the nice things Cactus (through Carpet) can offer is a pretty decent parallelization of (some form of) grids. A similar mechanism for SPH particles is currently not available in Cactus, but I could see interest for some projects. The questions whether it is worthwhile implementing it depends much on what it should be used for, and how alternative approaches look like.
I'm not sure whether my questions are answerable, but perhaps you can help me a second time.
This is what this email list is for. We actually _like_ users to ask questions here. :)
Frank
On Jan 15, 2014, at 9:05 , sruehe@astrophysik.uni-kiel.de wrote:
Hello EinsteinToolkit-Community,
first of all i want thank you very much for the fast and helpful answer to my last Email.
I mentioned i'm at the beginning of my phd. So before i'm wasting too much time on working at something an other already knows, i want to ask some perhaps a little bit naively sounding questions.
I have studied smoothed particle hydrodynamics on a fixed spacetime, so i get the idea devolping a sph based thorn. I found no publications, so is perhaps anybody already working on this problem, or has one shown that this isn't possible? Because this is an community based development, it is quite interesting whether it could be an useful thorn.
I'm not sure whether my questions are answerable, but perhaps you can help me a second time.
Stefan
Some years ago I developed an SPH thorn for Newtonian hydrodynamics, and coupled this to a grid-based elliptic solver for self-gravity. This code was able to evolve e.g. single stars. This was a proof of concept; I did not push this into the direction new physics.
A GR SPH implementation would be easier since it does not need a Newtonian solver. Instead, one would need to define how to set a grid-base T_ab (for spacetime evolution, if desired), and to either evaluate or interpolate the metric at the particle locations. For both, Cactus offers supporting infrastructure.
The missing parts in Cactus have to do with handling particles. Cactus offers parallelized 1D arrays where particles can be stored. It does not yet offer methods to re-distribute particles between different processes. Of course, these can be added, and I would be very interesting in helping out.
Would you have time for a Google Hangout in the coming days to chat? I'm located in the EST time zone, six hours behind you, so your late afternoon would be a good time for me.
-erik
I'd be interested in helping also and I might be of some assistance on this one. I'm currently working on a thorn that handles particles and uses boost's serialization library to keep them with their local grids. It currently 'works' but is in need of more testing and bug fixes before it'll be ready for science runs. It's also worth nothing that I abandoned using grid array's some time ago (for the reasons Erik mentioned) in favor of my own vector's of particles.
If there's a discussion on google hangouts, let me know as I'd be very interested in talking about this.
-Matt
On Wed, Jan 15, 2014 at 10:45 AM, Erik Schnetter schnetter@cct.lsu.eduwrote:
On Jan 15, 2014, at 9:05 , sruehe@astrophysik.uni-kiel.de wrote:
Hello EinsteinToolkit-Community,
first of all i want thank you very much for the fast and helpful answer
to
my last Email.
I mentioned i'm at the beginning of my phd. So before i'm wasting too
much
time on working at something an other already knows, i want to ask some perhaps a little bit naively sounding questions.
I have studied smoothed particle hydrodynamics on a fixed spacetime, so i get the idea devolping a sph based thorn. I found no publications, so is perhaps anybody already working on this problem, or has one shown that this isn't possible? Because this is an community based development, it is quite interesting whether it could be an useful thorn.
I'm not sure whether my questions are answerable, but perhaps you can
help
me a second time.
Stefan
Some years ago I developed an SPH thorn for Newtonian hydrodynamics, and coupled this to a grid-based elliptic solver for self-gravity. This code was able to evolve e.g. single stars. This was a proof of concept; I did not push this into the direction new physics.
A GR SPH implementation would be easier since it does not need a Newtonian solver. Instead, one would need to define how to set a grid-base T_ab (for spacetime evolution, if desired), and to either evaluate or interpolate the metric at the particle locations. For both, Cactus offers supporting infrastructure.
The missing parts in Cactus have to do with handling particles. Cactus offers parallelized 1D arrays where particles can be stored. It does not yet offer methods to re-distribute particles between different processes. Of course, these can be added, and I would be very interesting in helping out.
Would you have time for a Google Hangout in the coming days to chat? I'm located in the EST time zone, six hours behind you, so your late afternoon would be a good time for me.
-erik
-- Erik Schnetter schnetter@cct.lsu.edu http://www.perimeterinstitute.ca/personal/eschnetter/
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/.
Users mailing list Users@einsteintoolkit.org http://lists.einsteintoolkit.org/mailman/listinfo/users
On Wed, Jan 15, 2014 at 11:25:03AM -0500, Matt Kinsey wrote:
I'd be interested in helping also
If there is that much interst almost instantly we should maybe talk about this next Monday at our usual meeting to reach maybe even more people.
Stefan, Matt: Could you call in next Monday and, within 5 minutes or less, tell everyone what you are up to?
Frank
Sounds good to me, I'll be on the call.
-Matt
On Wed, Jan 15, 2014 at 11:38 AM, Frank Loeffler knarf@cct.lsu.edu wrote:
On Wed, Jan 15, 2014 at 11:25:03AM -0500, Matt Kinsey wrote:
I'd be interested in helping also
If there is that much interst almost instantly we should maybe talk about this next Monday at our usual meeting to reach maybe even more people.
Stefan, Matt: Could you call in next Monday and, within 5 minutes or less, tell everyone what you are up to?
Frank
users@lists.einsteintoolkit.org
-
Erik Schnetter -
Frank Loeffler -
Matt Kinsey -
sruehe＠astrophysik.uni-kiel.de