Bound-free interactions #325

chvogl · 2015-06-11T16:42:53Z

Start of continuum implementation:
for testing purposes bound-free interactions are currently treated as true absorptions and are described
by a fixed grey opacity

TODOs

added bound-free absorption
add a flag to switch this on or off
unit tests
offline quantitative test

-added compute_distance2continuum as replacement to compute_distance2electron -introduced montecarlo_continuum_event_handler -made some changes to struct rpacket

wkerzendorf · 2015-06-11T19:06:23Z

Great that looks like a good start. I need to have a read through it.

wkerzendorf · 2015-06-15T19:10:05Z

tardis/montecarlo/src/cmontecarlo.c

+  	  // Compute the continuum oddities for a real packet
+//	  calculate_chi_bf(packet, storage);
+
+  chi_boundfree = 0.2e-15;	//calculate_chi_bf(packet, storage);


I think for these values it would be useful to write getters. these can for now produce preset results, but later be linked to atomic data.

-implemented function to calculate chi_bf: use rudimentary hardcoded list of continuum edges (similar to line_list) to test frequency dependent absorption

- introduced a dummy list of frequencies of continuum edges similar to the list of line transitions - added function to calculate bound-free opacities in a hydrogenic approximation using dummy data

wkerzendorf · 2015-06-24T10:52:21Z

hi @chvogl can you have a look at the bullet points in the PR. You probably have already added a flag to switch this on/off (then just tick this). With the unit tests @kaushik94 can help for this. Thanks

chvogl · 2015-06-24T11:57:32Z

hi @wkerzendorf. I have a flag to switch off the continuum processes. However, as it is implemented now it is only a temporary solution. Are the offline quantitative tests something I should concern myself with right now?

wkerzendorf · 2015-06-24T14:12:02Z

@chvogl great! A current temporary solution is great for this PR. The question about the quantitative tests is: It depends on what you are working on right now. You can also add more checkmarks on the PR list tell us what you think still needs to be done for this first part of the BF addition. have a look at #332 how @aoifeboyle does that. Finally, you can also always put a PNG of a plot in the description of the PR (you can always change it as you add more commits). let me know if there's anything I can help you with.

ssim · 2015-06-26T15:22:35Z

@chvogl I think the idea for a quantitative test is that you make some simple test problem up (a constant density spherical slab or whatever) where you can compute the expected depth of a continuum edge - and then (offline) check that the code produces it. Eventually this would become a proper test in the code, but for now having such a check (with appropriate yaml file to run it) would be good. @unoebauer will be able to help propose/guide on this too, I think.

-implemented a function which returns the frequency dependent bf cross-section (based on dummy data) -selection of i-packet or k-packet creation in a bf-absorption

chvogl · 2015-07-03T13:37:43Z

Quantitative test:
two Tardis simulations with and without bound-free interactions (see plot below in green and blue)
Parameters for simulations:

disable_electron_scattering: yes
velocity: start : 1 km/s ; stop : 20000 km/s
time_explosion: 1300 day
density type : uniform ; value : 1e-42 g/cm^3
all other parameters are as in the tardis_example.yml

Parameters for bound-free absorption:

frequency of continuum absorption edges : 9.0e14, 8.223e14, 6.0e14, 3.5e14, 3.0e14 Hz
chi_bf_partial = 0.0075* 0.25e-15 1/cm;
opacities = [chi_bf_partial, 0.0, 2.0 * chi_bf_partial, 0.3 * chi_bf_partial, 2.0 * chi_bf_partial]

Analytic solution:

one approximation: due to the large ratio between the radius of the inner and outer boundary
(1/20000) all rays are essentially moving radially
O(v/c) accuracy in transforming between observer and comoving frame
we can obtain the ratio between the luminosity with and without bound-free absorption from the
analytic solution of the radiative transfer equation (see plot below in red)

unoebauer · 2015-07-03T15:16:11Z

Hi @chvogl @wkerzendorf @ssim,

I think this test of the bf processes looks really good and encouraging. If you agree, I would call this milestone (offline testing) closed. Since only the unit tests are still missing, we could already merge now. What do you think?

wkerzendorf · 2015-07-03T15:22:09Z

Sounds good to me. The last thing, that I would want to do is to link the script for the analytical model and plots, so that we can later possibly convert this into a test. I suggest using github gist (it will take you 30 seconds to put it there) and then put the link here.

chvogl · 2015-07-03T15:44:49Z

The program with the analytical solution can be found here: https://gist.github.com/chvogl/f193ce79d8ed19274ed8#file-quantitative_test-py

wkerzendorf · 2015-07-03T15:46:45Z

@chvogl thank you very much! @ssim I'll let you merge. Looks good from my side.

unoebauer · 2015-07-03T15:55:18Z

@wkerzendorf @ssim - please wait with the merging. @chvogl is currently refining the switch implementation so that it also affects the bf data preparation processes.

ssim · 2015-07-03T17:18:11Z

Ok - let me know when you are ready! I'm also very pleased with this and happy to merge whenever @chvogl has had time to deal with the switch (as above)

….c to montecarlo.pyx -disabled passing of continuum dummy data if continuum processes are switched off

chvogl · 2015-07-03T18:15:39Z

@ssim I still want to do some cleanup. I will get back to you as soon as I'm finished.

linked the plasma documentation again

wkerzendorf · 2015-07-06T14:32:57Z

@chvogl the merge of the parallel stuff screwed things up slightly. you will need to manually merge. Let me know if you have any questions about it.

Merge branch 'New_Continuum'

chvogl · 2015-07-07T17:41:00Z

@ssim @wkerzendorf I think we can merge now.

ssim · 2015-07-07T17:43:04Z

ok - I will merge today, unless @wkerzendorf has any objections?

ssim · 2015-07-08T04:24:54Z

I'm happy to merge this as it, but one thing to consider @chvogl @wkerzendorf @unoebauer ....is this all thread safe for the the openmp approach being used now?

E.g. I suspect that chi_bf_tmp_partial at least needs to be thread private, although there may be other things? Is this already taken care of @chvogl ?

ssim · 2015-07-08T04:27:07Z

By the way, even if it's not thread safe yet, I'd be inclined to merge anyway and just make an issue that we need to come back at some point and check this is all compatible with openmp (I don't think we necessarily need to hold up progress towards accepting this PR request) - thoughts?

chvogl · 2015-07-08T08:03:38Z

@ssim I'm aware that the chi_bf_tmp_partial is not thread safe yet. When I asked @unoebauer and @wkerzendorf about it, the general opinion was that this is something, which can be taken care of later on. Thus I would agree with you that we merge now and open an issue about the openmp compatibility.

unoebauer · 2015-07-08T08:10:13Z

Hi @ssim, @wkerzendorf:

As @chvogl said, we discussed yesterday potential incompatibilities of the bf scheme with the parallel version of Tardis. Since the current bf implementation is still at the early development stage and likely to change, my opinion was to not worry about this for the first merge. But we should definitely discuss this in more detail at the WW. Bottom line: I'd be in favour of merging the current version and opening and issue.

wkerzendorf · 2015-07-08T08:20:05Z

@chvogl @unoebauer @ssim looks good ... am merging. Please open an issue @chvogl to remind us that we need to make the mentioned array. Great job!

Bound-free interactions

chvogl added 5 commits June 10, 2015 14:07

Just a test commit. Start of continuum implementation.

f830fe1

Another test commit. Start of continuum implementation.

da842c6

Start of continuum implementation. Runs without error.

684d4d1

Test commit, please disregard. Start of continuum implementation.

8a3b416

Start of continuum implementation:

5f612a4

-added compute_distance2continuum as replacement to compute_distance2electron -introduced montecarlo_continuum_event_handler -made some changes to struct rpacket

wkerzendorf reviewed Jun 15, 2015
View reviewed changes

chvogl added 3 commits June 16, 2015 10:35

Start of continuum implementation:

eb86222

Start of continuum implementation:

e7fcc50

-implemented function to calculate chi_bf: use rudimentary hardcoded list of continuum edges (similar to line_list) to test frequency dependent absorption

- added a manual option to switch off continuum interactions

c8fc9af

- introduced a dummy list of frequencies of continuum edges similar to the list of line transitions - added function to calculate bound-free opacities in a hydrogenic approximation using dummy data

chvogl added 2 commits June 29, 2015 16:41

-added selection of continuum for bf-absorption

204b04a

-implemented a function which returns the frequency dependent bf cross-section (based on dummy data) -selection of i-packet or k-packet creation in a bf-absorption

-selection of continuum for bf-absorption slightly changed

928b342

-migrated switch for turning off continuum processes from cmontecarlo…

892931b

….c to montecarlo.pyx -disabled passing of continuum dummy data if continuum processes are switched off

Merge pull request tardis-sn#346 from wkerzendorf/docs/nlte_physics

2994cdc

linked the plasma documentation again

chvogl added 3 commits July 6, 2015 18:12

- small changes and cleanup

2ed8adc

some conflicts resolved

bd4fa8e

Merge branch 'New_Continuum'

finished merging of New_Continuum with current master branch

7c536c6

minor changes

594e1c6

wkerzendorf added a commit that referenced this pull request Jul 8, 2015

Merge pull request #325 from chvogl/New_Continuum

f0bbeb7

Bound-free interactions

wkerzendorf merged commit f0bbeb7 into tardis-sn:master Jul 8, 2015

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Bound-free interactions #325

Bound-free interactions #325

chvogl commented Jun 11, 2015

wkerzendorf commented Jun 11, 2015

wkerzendorf Jun 15, 2015

wkerzendorf commented Jun 24, 2015

chvogl commented Jun 24, 2015

wkerzendorf commented Jun 24, 2015

ssim commented Jun 26, 2015

chvogl commented Jul 3, 2015

unoebauer commented Jul 3, 2015

wkerzendorf commented Jul 3, 2015

chvogl commented Jul 3, 2015

wkerzendorf commented Jul 3, 2015

unoebauer commented Jul 3, 2015

ssim commented Jul 3, 2015

chvogl commented Jul 3, 2015

wkerzendorf commented Jul 6, 2015

chvogl commented Jul 7, 2015

ssim commented Jul 7, 2015

ssim commented Jul 8, 2015

ssim commented Jul 8, 2015

chvogl commented Jul 8, 2015

unoebauer commented Jul 8, 2015

wkerzendorf commented Jul 8, 2015

Bound-free interactions #325

Bound-free interactions #325

Conversation

chvogl commented Jun 11, 2015

wkerzendorf commented Jun 11, 2015

wkerzendorf Jun 15, 2015

Choose a reason for hiding this comment

wkerzendorf commented Jun 24, 2015

chvogl commented Jun 24, 2015

wkerzendorf commented Jun 24, 2015

ssim commented Jun 26, 2015

chvogl commented Jul 3, 2015

unoebauer commented Jul 3, 2015

wkerzendorf commented Jul 3, 2015

chvogl commented Jul 3, 2015

wkerzendorf commented Jul 3, 2015

unoebauer commented Jul 3, 2015

ssim commented Jul 3, 2015

chvogl commented Jul 3, 2015

wkerzendorf commented Jul 6, 2015

chvogl commented Jul 7, 2015

ssim commented Jul 7, 2015

ssim commented Jul 8, 2015

ssim commented Jul 8, 2015

chvogl commented Jul 8, 2015

unoebauer commented Jul 8, 2015

wkerzendorf commented Jul 8, 2015