Pop III and Star Cluster Formation/Feedback Updates

#293 Merged at d99a67e
  1. John Wise

This PR has many changes to the Population III star formation and radiating star cluster formation routines, methods 3 and 5, respectively. It does not affect other star formation routines.

  • New feature: Option to write an interpolated output just after Pop III star formation or a supernova. Parameter PopIIIOutputOnFeedback.

  • New feature: Adding a timestep limiter based on cooling time. New parameters: UseCoolingTimestep, CoolingTimestepSafetyFactor (default = 0.1). Note: Accidentally duplicated in the chem_energy_changes PR when categorizing changes. I had transplated it from Britton Smith's fork twice.

  • New feature: Added new parameters to halt the simulation when a metal-enriched gas reaches some density. Controlled by parameters, StopFirstTimeAtMetalEnrichedDensity, specified in units of g/cm^3, and EnrichedMetalFraction, specified in absolute metal fraction (default = 1e-8).

  • New feature: Added core-collapse supernova models for Pop III stars, using the ejecta mass from Nomoto et al (2006) and a mass-independent 1e51 erg explosion energy. Triggered by PopIIIUseHypernovae == FALSE.

  • New feature: Added new parameter, PopIIISupernovaExplosions, which can completely turn off Pop III supernovae. Default == TRUE.

  • Improvement: Changes to the mass removal after star formation: The surrounding region is still assumed to be ionized and heated as the initial Stroemgren sphere forms. To avoid artifical cooling before the central source radiates, a 1/r^2 photo-heating and photo-ionizing rates are injected into the sphere.

  • Improvement: Changes to the supernovae energy injection: Species fractions are now located initialized as ionized. The chem/energy solver will take care of this at high temperatures. Metals are now added evenly throughout the sphere; before, the gas was forced to have a uniform metallicity.

  • Improvement: Optimizations to the routines that determine the overlapping grids with a feedback sphere.

  • Model change: Changed time for a Pop III stellar radiation to increase to its ZAMS luminosity from 50 kyr to 10 kyr (Susa 2014).

  • Solver change: Increasing maximum chem/energy iteration to 100k.

  • Code change: Put the H2 dissociation radiation calculation from point sources into its own routine.

  • BUGFIX: Corrected a bug that removed gas from the wrong cell after BH accretion.

  • BUGFIX: Corrected a bug that incorrectly calculated the SN feedback sphere's radius incorrectly when r < dx.

Comments (10)

  1. John Wise author

    I've fixed all of the conflicts that arose after the acceptance of the other PRs.

    1. Brian OShea

      This PR now compiles and passes the test suite. Once we get a third person (Britton?) to provide feedback and sign off, it's ready to be accepted.

      1. Brian OShea

        Following up on this PR: we still desperately need a third person -- preferably somebody who has some familiarity with this code -- to take a quick look over it!

  2. Greg Bryan

    I just looked it over and it seems good to me! I had two minor comments, but they are just related to presentation and John can either fix or ignore.

    1. John Wise author

      Thanks, Greg! I've addressed both of your comments in the code now. It's ready to merge.