Schmidt law for the star formation rate

Adding the option to use a Schmidt law for star formation within the EAGLE star formation module.

The code defaults to the pressure law with no extra parameter necessary in the yaml file. In order to use the Schmidt law, add the first or both lines to the yaml file:

  SF_model:                     SchmidtLaw     # (Optional) If set to "SchmidtLaw", the SFR is calculated using a Schmidt law, defaults to the pressure law
  star_formation_efficiency:         0.01      # (Optional) If a Schmidt law is used, this regulates the star formation efficiency per free-fall time (defaults to 0.01)

Mostly copied over from @folkert's work for COLIBRE with smaller adaptations. I also made sure that the IsolatedGalaxy_feedback example runs out of the box again (download the new cooling tables and minor changes to the python script that tests the SFH). The condition for star formation remains untouched, the only part that is affected is the calculation of the star formation rate.

The code is compiling and running for both the EAGLE pressure law and the Schmidt law (tested with IsolatedGalaxy_feedback).

added 1 commit

• 620a1fdf - Minor change to how the plotting script is called in run.sh

Compare with previous version

assigned to @matthieu

added EAGLE SPH feature request labels

Is that ready to be looked at or is it being tested first?

Ready to be looked at. I have started some runs yesterday and they all look as expected.

I like the changes and the way things are separated.

This now inspires me to go one level further and copy what you did over into the pressure-law. That is grouping the relevant quantities also in a sub-structure. I would also not make the master switch an optional parameter. I would demand that the user explicitly states what they want to run in the parameter file. This way we always know what is about to happen. (It's a bit of a pain to update all the param files but overall worth it for clarity I think.)

What do you think? I am happy to make the changes.

Yes, in the COLIBRE code, @folkert also had the pressure law properties in a separate substructure and the SF_model parameter is mandatory. I went for the minimally changed version but if you think it's worth updating the parameter files, then of course: go ahead!

added 1 commit

• 4e875a8b - Also move the pressure law parameters in their own sub-structure. Force the...

Compare with previous version

added 1 commit

• db56982e - Rename some of the variables related to the direct conversion into stars....

Compare with previous version

added 2 commits

• 3f5ec732 - Update all the relevant YAML files
• 07f51426 - Document the fact that in the XL model we are not using the threshold for automatic SF any more

Compare with previous version

@folkert what do you think of the updates? Since you originally contributed all this code, I want to make sure you are happy with this new version.

It looks quite similar to my change in the COLIBRE branch, is there a specific reason why we do not use an union in the struct anymore? Besides this the code looks good to me.

I guess we could but that is just saving a couple of bytes overall. Maybe that's not worth it if it reduces clarity.

merged

mentioned in commit cff4960c