Closed rbooth200 closed 7 years ago
Looks fine to me - I think we could merge, although I'm not really familiar with the meshless and I think David should have the last word. After we have closed this and the tests branch I'll add the slope limiter tests to the tests run after every commit, at the moment we run only SPH tests.
It might be worth only running a subset of them since they take a few minutes for each of the gresho tests.
On Thu, 22 Sep 2016 23:23 giovanni-rosotti, notifications@github.com wrote:
Looks fine to me - I think we could merge, although I'm not really familiar with the meshless and I think David should have the last word. After we have closed this and the tests branch I'll add the slope limiter tests to the tests run after every commit, at the moment we run only SPH tests.
— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/gandalfcode/gandalf/pull/76#issuecomment-249045631, or mute the thread https://github.com/notifications/unsubscribe-auth/AOqItCIm4eNvwBX9aO9kT_ev9O8TxjG4ks5qsv_RgaJpZM4J-3U8 .
Just looking through some of the code and running through a batch of tests and all looks good so far. Will run a few more and if they're fine, I'll go ahead and merge.
All fine so I merged the branch. I guess we can now delete both the slope_limiter_testing and slope_limiters branch?
Please leave the slope_limiter_testing branch for now. I have some extra bits and pieces that I'm still playing around with. I'll either push more commits or delete it later...
On 23/09/16 10:58, dhubber wrote:
All fine so I merged the branch. I guess we can now delete both the slope_limiter_testing and slope_limiters branch?
— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/gandalfcode/gandalf/pull/76#issuecomment-249151706, or mute the thread https://github.com/notifications/unsubscribe-auth/AOqItO9op6HqjETFb5U6po30z-xWXDjpks5qs6LigaJpZM4J-3U8.
Sorry guys, deleted the branch as soon as I saw the e-mail - feel free to restore it if you need to
No worries, I'll do that later if need be.
On 23/09/16 11:03, giovanni-rosotti wrote:
Sorry guys, deleted the branch as soon as I saw the e-mail - feel free to restore it if you need to
— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/gandalfcode/gandalf/pull/76#issuecomment-249152542, or mute the thread https://github.com/notifications/unsubscribe-auth/AOqItD2cRsNx5kfG1FyUqv02Xc5OE6g2ks5qs6PigaJpZM4J-3U8.
I've modified the slope limiters and written some scripts to produce some comparisons between the various slope limiters. Since this branch is built upon the slope_limiter branch we had before I will delete that pull request and we should use this one instead. This pull request includes the following:
I've also attached a couple of results from the slope limiter tests for the discussion. These can be used to rank the limiters by how diffusive they are. I'd suggest the order from least to most diffusive is:
which agrees well with theory.