Comparison with O+16a, attempt 2

Recap of Last Post

  • In the last post, I performed a run with the same parameters as the Ohlmann+16a run.
  • However the simulation went from 2-body to 3-body when the secondary was mistakenly reintroduced after about 7 days.
  • This was not a bug in the code, just me forgetting to flip a 'switch'.

New Work

  • I flipped the switch and reran the code up to 20 days.

Summary of New Results

  • The results match qualitatively those of O+16a. Differences are likely caused by remaining differences in the numerical setups, including softening radius, ambient density, resolution, and degree of corotation.

Detailed Results

I) Circular orbit as in Ohlmann16a
Damp088) Extrapolated hydro BCs, Multipole expansion Poisson BCs, ambient dyne/cm
(comet compute 1728 cores, 2 cores per task to increase memory per task)
( cm, , 5 levels AMR)
Restarted from run Damp062, at s, just after Damping stopped, to s
​2d density
​2d density (zoomed)
​2d density and velocity (zoomed)
​2d density edge-on slice (zoomed)
​2d particle velocity (zoomed)
​2d particle position (zoomed)
​2d Particle mass (zoomed)

Here are PROJECTIONS, instead of slices, zoomed in with a better color scheme, and extended to go up to 20 days:
​2d density projection
​2d density projection edge-on

Here is a version of the SLICE in the orbital plane, going up to 20 days, rotated 180 degrees to match the orbit of O+16a, color range shifted to match O+16a, and with a better color scheme:
​2d density
…or with blue labels for better visibility: ​2d density
​2d density (zoomed)
Here is temperature:
​2d temperature
…and Mach number:
​Mach number

Below I've plotted Fig 1 of ​Ohlmann+16a and the equivalent figure with this simulation for comparison. In my plot of separation, I've shown the softening radius as a dotted horizontal line. In the inset showing particle trajectories, I've inverted the axes and shifted the origin to allow direct comparison (our simulation is rotated by 180 degrees with respect to theirs). Circles represent the softening radius, while the green dot shows the initial center of mass of the two stars. The very small green square in the upper right shows the smallest resolution element.
http://www.pas.rochester.edu/~lchamandy/Graphics/Ohlmann/fig1.pdf http://www.pas.rochester.edu/~lchamandy/Graphics/RGB/Post-sink_particle/Post-modified_Lane_Emden/Damp088/p20d.gif

Below are snapshots of density in the orbital plane from t = 10 days and t = 20 days. Note that to compare directly, I have rotated each figure by 180 deg. In the O+16a figure, the '+' denotes the primary core particle, while the 'x' denotes the secondary point particle. The core and secondary are denoted as '0' and '1' respectively in my plot.
http://www.pas.rochester.edu/~lchamandy/Graphics/RGB/Post-sink_particle/Post-modified_Lane_Emden/Damp088/rho2d_zoom_blue_0043.png http://www.pas.rochester.edu/~lchamandy/Graphics/RGB/Post-sink_particle/Post-modified_Lane_Emden/Damp088/rho2d_zoom_blue_0086.png
http://www.pas.rochester.edu/~lchamandy/Graphics/Ohlmann/fig3.pdf

Discussion

  • The results match qualitatively those of Ohlmann+16a, but differ in detail.
  • The orbit of the secondary is larger, while that of the primary is slightly smaller, compared to O+16a. The separation does not become as small at the first minimum or as large at the first maximum, compared to O+16a. This may be due to the ~5 times larger gravitational softening length used.
  • At t = 10 days, the density in the orbital plane appears similar to that of O+16a, but the spiral arm is less pronounced and features are less detailed. This could be due to the resolution being lower by a factor of somewhere between 2 and 12. It may be partly caused by the weaker gravitational interaction due to larger softening radius mentioned above.
  • Also, the common envelope has not expanded as much into the ambient medium. This could be due to the >107 times larger ambient density used compared to that of O+16a.
  • Recall the differences between our setup and O+16a, outlined in the Table 2 of the following ​pdf file. The main differences between the two setups are:
    • 5x larger softening radius
    • 107 times larger ambient density
    • up to 12 times smaller resolution
    • 9x smaller box size
    • no initial rotation vs 95% solid-body corotation with orbit
    • some differences in the relaxation run to prepare the red giant star (detailed in the table and in previous posts)
    • extrapolated instead of periodic BCs

Next Steps

  • The easiest and most obvious thing to do is to reduce the softening radius (= cutoff radius defening RG core. Recall that for r smaller than this radius, we replace the core with a modified Lane-Emden solution). However, reducing this softening radius to 1 Rsun and keeping the same max resolution would mean that we resolve the softening radius by only 3.5 resolution elements, whereas O+16a claim that 10 resolution elements is needed.
  • That suggests that we should increase the max resolution by a factor of 2 or 4, but this would slow down the code considerably. It would be best to first lower the softening length without changing the resolution.
  • It would also be worth testing lower ambient densities. In past tests the code has crashed, but it should be possible to lower the ambient density by at least a factor of 10 or 30.
  • So this suggests the following 3 runs (starting from the beginning of the relaxation run):
    • As 088 but reduce the ambient density by 10 or 30
    • As 088 but reduce the softening radius by 4.8 to equal that of O+16a
    • As 088 but with both of the above changes

Comments

No comments.