Solving on a disk

[rsferna...@gmail.com]

Hi,

I'm trying to solve a simple Poisson's equation on a full disk structure.

I'm wondering whether you would recommend that I use the double-cover method 

or whether there is some spectral basis that would be better.

Thanks in advance.

[Louis-Alexandre Couston]

Hi,

Did you figure out the best way to solve some equations on a full 2D disk?

[Jeffrey S. Oishi]

Hi,

This is a bit belated, but our 2D disk bases are ready for testing in Dedalus 3. It's very pre-release at this point, but if you'd like to try, we can provide an example.

Jeff

--
You received this message because you are subscribed to the Google Groups "Dedalus Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dedalus-user...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/dedalus-users/5b410b07-9790-4d97-81ee-70c320a43217n%40googlegroups.com.

[Louis-Alexandre Couston]

Hi Jeff,

I would love to try it!!!!!!!! :DDD

Should I install dedalus from the github repo to test it?

Best,

Louis

[Nicolas Grisouard]

I would be interested to know what is under the hood. I have a low-burning interest on topic since the summer of 2009 (https://sites.physics.utoronto.ca/nicolasgrisouard/images/otherpublications/2009GrisouardUnknown)

Cheers, Nico.

[Geoffrey Vasil]

Hi Nico, 

As Jeff mentioned, we are still finalising the APIs for the disk. So the code implementation is still a bit in flux. But it will be getting sorted out reasonably soon. 

But if you want to read about what really under the hood you might kind the following useful:

https://arxiv.org/abs/1509.07624

We followed that up with similar results in the sphere:

https://arxiv.org/abs/1804.10320

Even if the disk is what you are really after, the sphere does help put things in context a bit. 

Hope that helps!  

Cheers, 

Geoff 

To view this discussion on the web visit https://groups.google.com/d/msgid/dedalus-users/07f405ae-4e53-44bf-830a-57f9e1f4b45an%40googlegroups.com.

[Nicolas Grisouard]

Exactly what I was looking for, thanks! -- Nico.

[Louis-Alexandre Couston]

Hi all,

I've had a first run through those papers.

Would you say that the best way foward to setting up a problem for Rayleigh-Benard convection in a disk geometry is to start from the convection benchmark problem at https://github.com/DedalusProject/dedalus_sphere/blob/master/paper/marti_conv_benchmark.py ?

I understand it may change the definition of the matrices but is it doable to remove the latitudinal dependence from this script?

Or would you recommend working on a new script?

Happy to provide more details on the problem set up if helpful!

Best,

Louis

[Keaton Burns]

Hi Louis,

I think the easiest thing would actually be to wait a couple more weeks, when we’ll have the real “beta release” of v3 of Dedalus.  There’s a few things we’re trying to clean up before then, which are taking a little longer than expected, but the payoff is that you’ll be able to use disks and spheres in a much more “dedalus-like” symbolic syntax without having to deal with the matrices directly, at all.  And a simple setup with Rayleigh-Benard in a sideways disk is one of the included examples, so that would definitely be the easiest place to start from.

Best,

-Keaton

To view this discussion on the web visit https://groups.google.com/d/msgid/dedalus-users/78ae55d8-de14-4137-b2ea-ef2131ace438n%40googlegroups.com.

[Louis-Alexandre Couston]

Hi Keaton,

Ok. This is rather convincing :)

Looking forward to it (and thank you so much for working on all these new features!!)!

Best,

Louis

[Louis-Alexandre Couston]

Hello,

Has the beta version of dedalus v3 been released?

Cheers,