COMMON ENVELOPE SIMULATIONS

Plan for XSEDE proposal 2018

Summary of current allocation

  • A bit less than 130,000 node hours remain.
  • It is estimated (still very rough though) that we can complete about 15 runs like the non-accretion run from paper I (run 143).
  • We have thus far used about 23% of the allocation so about 77% remains.
  • To get this down to <50% by Oct 15, we need to use up about 27% of the full allocation or a little over 1/3 of what remains.
  • This translates to about 5-6 runs by Oct 15.

Plan for now until Oct 15

The following runs do not involve major changes to the code so can hopefully be done in this time frame:

  1. Test dependence on secondary mass (parameter regime 1).
    1. Run 149: As 143 but with secondary mass smaller by a factor of 2, i.e. m2=0.5 Msun instead of 1.0 Msun (running).
  2. Test dependence on secondary mass (parameter regime 2).
    1. Run 151: As 143 but with secondary mass smaller by a factor of 4, i.e. m2=0.25 Msun instead of 1.0 Msun (submitted).
  3. Test convergence with respect to softening length and resolution.
    1. Run 152: As 143 restarted from frame 72, which is the frame where the softening radius and smallest resolution element were halved in 143, but now do not halve the softening length nor the resolution element (submitted).
    2. As 152 but halve softening length only.
    3. As 152 but halve smallest resolution element only.
  4. Roche lobe overflow.
    1. Put secondary at Roche limit separation.
    2. Must think about refinement.
    3. Increase q to 1 both to reduce Roche limit separation and also increase .
  5. Test dependence on initial spin of the RG.
    1. Repeat run 143 but now initialize RG in solid body rotation at the initial orbital angular velocity.
  6. Test convergence with respect to size of refinement region (parameter regime 1).
    1. Repeat run 143 using a more liberal choice for the refinement radius for each interval (c.f. Fig 3 in Paper I)
  7. Test convergence with respect to size of refinement region (parameter regime 1).
    1. Based on the results of the above, repeat run once more using either an even more liberal refinement radius or else a refinement radius that is more conservative than run 143.
  • Another possibility is to restart run 143 from the end of the simulation at t = 40 days. This is easy to do but the drawback is that material will flow out of the box soon after.

Plan for Oct 15 to Dec 31

We can expect about 10 more runs equivalent to run 143 but I provide 15 possibilities below:

Questions to answer for jet project:

  • How does the jet affect the morphology of the envelope?
  • How does the jet affect the ejection of the envelope?
  • How does the jet evolve, does it get quenched?
  • What is the dependence of these questions on when the jet gets turned on?
  • What is the dependence on accretion rate?
  • Opening angle?
  1. Simulate CEE with a jet from the secondary (parameter regime 1) (Restart from run 143)
  2. Simulate CEE with a jet from the secondary (parameter regime 2) (Restart from run 143)
  3. Simulate CEE with a jet from the secondary (parameter regime 3) (Restart from run 143)
  4. Simulate CEE with a jet from the secondary (parameter regime 4) (Restart from run 143)
  5. Simulate CEE with a jet from the secondary (parameter regime 5) (Restart from run 143)
  6. Simulate RGB/AGB star in improved simulation including (part I of a long run)
    1. larger box (but smaller ambient resolution): the larger the box the longer we can run it before material starts leaving the box,
    2. ambient consisting of hydrostatic atmosphere surrounded by very low density and low pressure gas,
    3. start from Roche limit separation (optional)
  7. Simulate RGB/AGB star in improved simulation including (part II of a long run)
  8. Simulate RGB/AGB star in improved simulation including (part III of a long run)
  9. Simulate RGB/AGB star in improved simulation including (part IV of a long run)
  10. Simulate RGB/AGB star in improved simulation including (part V of a long run)
  11. Triple system involving a planet or tertiary star orbiting the secondary… (parameter regime 1)
  12. Triple system involving a planet or tertiary star orbiting the secondary… (parameter regime 2)
  13. Triple system involving a planet or tertiary star orbiting the secondary… (parameter regime 3)
  14. Triple system involving a planet or tertiary star orbiting the secondary… (parameter regime 4)
  15. Triple system involving a planet or tertiary star orbiting the secondary… (parameter regime 5)

Storage

  • The size of run 143 data is about 10 TB.
  • To do 15 more runs like 143 would require 150 TB.
  • This would cost an additional $15,000 per year on bluehive which is unaffordable.
  • We need to look for other options like free space somewhare on blue hive.
  • We probably need to keep only ½ of the data, i.e. every second chombo file.

Comments

No comments.