rss
Update on common envelope simulations
Recap of Last Post
- In the last post I presented a simulation of a RG translating across the grid.
- I also presented a common envelope simulation but later realized that the initial velocities of the primary and secondary were wrong (about a factor of 2 too small). I corrected the script to calculate the initial velocities and tested it against the Sun-Earth system.
New Work
I corrected the initial velocities and simulated the following cases:
- RG star of mass 1.956 Msun including 0.369 Msun core (point mass), with secondary point mass of 0.978 Msun (half the RG mass), at initial separation 49 Rsun, just outside of the RG outer radius of about 48 Rsun. Very close to parameters of Ohlmann+16a.
- As above but larger separation of 60 Rsun instead of 49 Rsun.
- As above but even larger separation of 98 Rsun instead of 49 Rsun.
- As the first case but now make secondary mass equal to only 0.001 Msun so that problem reduces to "1-body".
- Replace RG with a point particle of the same mass and run in low resolution with uniform background to test orbital dynamics.
- Same but with original point masses of the first case, i.e. the RG core and the secondary, without RG envelope.
Summary of New Results
- We do not achieve a circular orbit in the CE sims. In each case the orbit is looser than it should be. The binary separation grows with time and the speed of the secondary reduces with time.
- When the RG is replaced by a point mass then the correct circular orbit is obtained (in low res with a uniform ambient medium).
- If the RG envelope is excluded and the point masses representing the core and secondary are made to orbit in a uniform ambient medium, we obtain almost the identical (non-circular) orbit as for the full CE simulation.
- The natural explanation for these results is that the point masses can feel each other's gravity, the gas can feel the gravity of the point masses, but the point masses do not feel gravity from the gas! This feature must obviously be included in the next set of runs.
Results
I) Circular orbit as in Ohlmann16a
Damp084) 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 and velocity
2d density and particle velocity
2d density edge-on
2d Temperature
2d Particle position
2d Particle mass
II) As 
above but wider orbit (binary separation of 60Rsun instead of 49Rsun)
Damp086) Extrapolated hydro BCs, Multipole expansion Poisson BCs, ambient dyne/cm
(bluehive standard 120 cores)
( cm, , 5 levels AMR)
Restarted from run Damp062, at s, just after Damping stopped
2d density
2d density and velocity
III) As above but even wider orbit (binary separation of 98Rsun instead of 49Rsun)
Damp087) 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
2d density and velocity
IV) As above but very low mass secondary (0.001Msun instead of 0.978Msun)
Damp085) Extrapolated hydro BCs, Multipole expansion Poisson BCs, ambient dyne/cm
(stampede normal 1024 cores)
( cm, , 5 levels AMR)
Restarted from run Damp062, at s, just after Damping stopped
2d density andparticle velocity
V) As above but RG represented by a point mass, so two point masses orbiting on uniform background
Kepler083) Extrapolated hydro BCs, Multipole expansion Poisson BCs, ambient density and pressure set to 1d-10
( cm, , 0 levels AMR)
(Started from t=0)
2D projected particle positions
particle velocities
2D projected particle positions, edge-on
VI) As (V) above but only mass of RG core is included in primary point mass, so same point masses as in
Kepler083core) Extrapolated hydro BCs, Multipole expansion Poisson BCs, ambient density and pressure set to 1d-10
(stampede normal 1024 cores)
( cm, , 0 levels AMR)
(Started from t=0)
2D projected particle positions
particle velocities
Comparison between and (VI)
Particle velocities
COMMENTS: Clearly, the secondary does not feel the gravity of the RG envelope. Thus, point masses do not feel gravity of the gas. This obviously needs to be corrected for the next set of runs.
Next Steps
Once this problems is resolved, we need to redo run in particular. If everything is in order, we can then plot the orbital separation as a function of time, and compare the result with Ohlmann+16a
Figure 1.
Here are the other figures from that paper for reference:
Figure 2
Figure 3
Figure 4
Aside from resolution and other details, the biggest difference between our setup and that of Ohlmann+ is that their RG is initiated with 95% corotation with the orbit, whereas ours is not rotating initially.
Other points worth mentioning
- My quota on bluehive has been reduced from 24 TB to 16 TB. I will look to reduce it more in the next two weeks. My quota on blue streak has been reduced from 5 TB to 2 TB.
- There is a nice paper by MacLeod et al. about accretion onto the secondary during the common envelope that I'm reading at the moment. Aside from being interesting, it is quite pedgogically written. I'm doing a literature review to prepare for the upcoming conference.
Comments
No comments.