Martin's meeting outline
I will talk about my progress in the following projets (2011).
29 nov
- Magnetic tower paper. Writing the conclusions. Should have it ready for Adam's and Eric's final revision by the end of the week.
- CRL 618. The new 4 runs, with AMR (i.e. no particle grids), are in bluegene's queue. I've experienced some code running problems there, but Jonathan and I are working to solve them. In addition, Bin and I are running two of the sims in bluehive. Finish time estimation of 4 day/sim.
- Binary.
-Setup comparison:
http://www.pas.rochester.edu/~martinhe/2011/binary/diffSETUPS.gif
I've been working on two simulations: one with an initial Keplerian disk and the other one with the orbiting binaries.
-Initial Keplerian Disks, gamma=1.001.
+5amr, no ambient wind, the disk is unstable and tilts. I'mm still running this sim further:
http://www.pas.rochester.edu/~martinhe/2011/binary/29nov-densCONTOURS-pole.gif
http://www.pas.rochester.edu/~martinhe/2011/binary/29nov-densCONTOURS-edge.gif
http://www.pas.rochester.edu/~martinhe/2011/binary/29nov-densCONTOURS-angle.gif
Now trying 6amr but keeping the same gravity soft radius than in the 5amr sim.
+The disk is embedded in an abient medium with dens=densdisk/1000 and vx= 20kms-1:
http://www.pas.rochester.edu/~martinhe/2011/binary/29nova.gif
Currently running one without any wind and 5 amr levels.
-Binaries. I've ran a test simulation with the primary injecting the wind, while the secondary pulls gas form t=0 with 4amr levels. I see that a disk forms and tilts (apologies for the ugly image, the orbital plane is the XY one):
http://www.pas.rochester.edu/~martinhe/2011/binary/29nova.png
Currently running one with 5 amr levels and a secondary's gravity field which will be tunned on once the primary's wind goes beyond the grid boundaries. Snapshot:
http://www.pas.rochester.edu/~martinhe/2011/binary/29novf.png
21 nov
- CRL 618. New runs with diffusion and artificial viscosity seem to get rid of carbuncles. Also, seems that, as expected, 5amr, 64 cells/rclump are necessary to resolve the clumps properly. The two movies below show runs with a density contrast of 100 and a particle grid which follows the clump or the jet.
- Clump, rhoamb=const, 4amr
http://www.pas.rochester.edu/~martinhe/2011/crl/21nova.gif
- Clump, rhoamb=r-1, 4amr
http://www.pas.rochester.edu/~martinhe/2011/crl/21novd.gif
- Jet, rhoamb=const, 5amr
http://www.pas.rochester.edu/~martinhe/2011/crl/21novb.gif
(in this test the particle is a bit offset relatie to the jet's head, this will be corrected for future runs).
- Jet, rhoamb=r-1, 4amr. The bowshock seems too thick. I'll reduce the diffusion for the next run.
http://www.pas.rochester.edu/~martinhe/2011/crl/21novc.gif
- Disks. I found that the latest datasets yield a stable disk at the centra of the grid with a Keplerian vel profile fir both gamma=1.001 and 1.667. I've defer a simulation which will include an inflowing wind because I want to get the binary simulations running soon for my aas talk. The binary implementation is almost ready. My tests are not long enough to see any potential disk tilt.
- Magnetic towers. New paragraph and plot (which substitutes the one I showed during the last meeting) where I quantitatively show that our towers are PFD,
see pages 21 and 23 of the latest draft:
http://www.pas.rochester.edu/~martinhe/2011/magTOWER/paper.pdf
I plan to have the paper ready by the beginning of next week?
- PN paper. I've not yet let the other co-authors know about the latest referee's response.
What now?
Are we submitting the paper as it is now to MNRAS or New Astronomy?
- Students. Ruka has given me the 1st draft of his report.
16nov
- Students. Ruka's Bondi module is now working! (see his post). Matt's final part of the testing scripts seem to be working too. We're ready to start testing (Matt ?). We've included new tests to catch bugs related to Hypre and threading compilation flags.
- Accretion disks. My implementation keeps showing disk tilt, but velocity plots suggest it is caused by material that flows along the disk poles and collides at the centre. Even with a disk to ambient density contrast of 100 and gravity potential softening. I also tried gamma=5/3, but the disk ended up launching some sort of jet, very early in its evo.
I'm now working with Jonathan's implementation. Differences:
Mine | His | ||
Old revision | latest one | ||
gamma=1.001 | gamma=5/3 | ||
Particle AMR | normal (gradients) AMR | —> has no effect on the disk, but speeds up the sims. | |
Disk initialized explicitly | Disk initialized with as an object | ||
Open BC | periodic BC |
I see a stable disk. I'm now running two sims: one with the disk alone and one with an incoming wind too. I'll next couple Jonathan's implementation with my binary one. I expect to have this in a week time.
- Magnetic towers. New plot: Poyting to kinetic fluxes:
http://www.pas.rochester.edu/~martinhe/2011/magTOWER/pfd.pdf
Andrea: "The pictures look fine indeed and pretty convincing. However be aware that the 1/ (beta M2) relation is only (very) approximative and may not apply everywhere. The only way to really show it is PDF is to calculate the Poynitng flux and kinetic flux along the jet at different height/radii and compare. I really think it is worth the effort, even for future work.".
I'll so some plots of the ratio of Poyting to kinetic flux: [ Bz2/(8pi) ] vz / ( .5 rho vz3 )
- AGN jet truncation due to stellar winds. Tests keep going well. The following show an AGN radio jet with an opening angle of 20o, and a star with a strong wind, but not so to truncate the jet.
http://www.pas.rochester.edu/~martinhe/2011/agn/2d-dens-10nov11.gif ← small grid
http://www.pas.rochester.edu/~martinhe/2011/agn/2d-dens-16nov11.gif ← large grid
We expect larger opening angles close to the host galaxy (http://www.cita.utoronto.ca/~everett/agnOverview/31xopmon.jpg) and we also want the star to have some velocity and a stronger wind. Coming up soon.
- CRL 618. New implementation to follow the clump faster and see the effect of artificial viscosity, difusion and the code's limiters on the spurious features that we saw (along the head of the clump) before (see Bin's blog).
http://www.pas.rochester.edu/~martinhe/2011/crl/16nov.png
Note that the time figures are larger by a factor of ~10 because these runs have been compiled with check flags on. Future simulations will run ~10 times faster.
Running with a massless particle allows us to make full sims in a coupe of days. Once we're happy with the diffusion setup and the morphology of the clumps, we can either switch to the gradients AMR or combine it with the particle's, in order to capture the clump/jets' cavity with higher resolution.
8nov
- Magnetic tower paper, sent to all co-authors. Got responses from Andrea and Pat. Waiting for comments from Adam, Eric and Sergey.
- Binary. 6amr run has almost completed 20 orbits, like the 5 amr one. 5+6amr run is still running.
*Disk tests. These are set up as flat (radius/height=4) cylinders at the center of the grid, with a constant disk to ambient density ratio of 10 and a Keplerian azimuthal velocity about the central, static, point gravity particle. The rest of the setup (gamma=1.001, massparticle, grid size, scales, etc.) are as in the binary simulations (see previous posts). No wind is injected, and the duration of the movie is equivalent to 2 orbital periods of the binary simulations.
-The disk does not reach a stable state, at least for the simulations' times explored, but I'm still running these tests.
-The link below shows logarithmic density maps, upper view, of a 5amr (left part) and a 6amr (right part) disk.
http://www.pas.rochester.edu/~martinhe/2011/binary/8nov11.5amrVS6amr.gif
-The link below shows logarithmic 3d density iso-coutours. 5amr (top) and a 6amr (bottom). Edge-on (left) and pole-on (right). The viewing position is the same in all panels; the bottom ones actually show a smaller disk. The beginning of the simulations is particularly surprising (left column) for I see oscillations of the gas that is in the central most orbits.
http://www.pas.rochester.edu/~martinhe/2011/binary/8nov11.5amrVS6amr.3d.gif
-The initial oscillations are likely due to ambient medium which is nor rotating about the central gravity particle and is thus pulled by gravity along the poles. Also, the ambient medium and the disk have temperatures of 100 and 10 K, respectively. The idea was to have pressure equilibrium as initial conditions. Yet, may be the low temperature of the disk is such that the scale defined by the contrast of Keplerian to sound speeds is rather small and so I do not resolve it, even with 6 amr levels. This is motivated by the fact that I see the inner part of the disk very rapidly collapses in to a 2 cell thick structure. I'm now working to improve the implementation.
- CRL618. Bin and I have the new setup which includes diffusion and carbuncles prevention. Our runs (jet/clump to ambient density ratio of 100 with a constant, and stratified, density ambient) are waiting in bluehive's queue.
- AGN jet truncation due to stellar winds. Tests are going well:
http://www.pas.rochester.edu/~martinhe/2011/agn/agn.8nov.gif
http://www.pas.rochester.edu/~martinhe/2011/agn/agn.8nov.2.gif
This is a logarithmic false colour map of the grid gas density. The jet enters the grid on the left vertical face. The star has a spherical wind (Mdot~5e-6, v~500km/s) and is at the intersection of the three planes, at the middle of the grid. The flow structure after the collision of the stellar wind and the jet is very interesting, potentially observationally distinctive.
1 nov
* Binary.
- Tests with a simple disk about one star (only) are running (see link below, 2d slice, left: the disk just before interacting with the wind; right: the disk just after interacting with the wind). Results soon.
http://www.pas.rochester.edu/~martinhe/2011/binary/1nov11a.png
- Soothed plots: Angm. mom. projection, mass and accretion rate (which seem to be rather high) vs time, for 5 and 6 amr levels:
http://www.pas.rochester.edu/~martinhe/2011/binary/1nov11.pdf
25 oct
- Magnetic tower project.
*Hydro comparison: http://www.pas.rochester.edu/~martinhe/2011/magTOWER/25oct11.pdf.
*Working on Andrea's comments.
- Binary project.
*Run with 6amr levels is still running and it has gotten up to time=14.8 orbits. The disk tilts; I see and angle ~ 10o-50o between the orbital ang. mom. vector (vertical direction) and that of the disk.
*Run with 5amr unitl time=3 orbits, and then restart with 6amr levels is running too. It is rather slow though. I'm working to fix this.
*Run with 7amr unpractically slow; quite a bit of fractional steps between every level 0 step. I'm now:
*updating my implementation to use threading *running with a slightly large gravity softening radius (which will make the conditions at the inner disk less dramatic)
to see if it helps.
- CRL618 project.
*bear2fix and shape are now working. Bin's in charge of producing the synthetic emission and PV diagrams.
*Also running another clump-no rings case with more resolution at the flow behind the clump, as requested by Bruce via email.
- AGN jet truncation project. Implementing it.
18 oct
- I worked on the 3 proceeding from the magnetic fields in the universe 3 conference. I submitted them yesterday. Will post them in astro-ph this week.
- Magnetic tower. I've been writing the intro and Andrea's comments. The hydro-rotating run is in bluegene's queue. I expect to have the data in a few days time.
- Binary. We have a 5 amr and a 6amr runs. Both have completed about 10 orbits.
- rdisk(6amr) < rdisk(5amr).
- 6amr disk shows a rather moderate tilt, whereas the 5amr disk, as we saw before, tilts all the way to the orbital plane.
- Projected angular momentum plots (something's wrong): http://www.pas.rochester.edu/~martinhe/2011/binary/18oct11.pdf
- Density movies: http://www.pas.rochester.edu/~martinhe/2011/binary/18oct11.gif
Meeting resolutions: (i) continue running the 6 amr run (ii) run a 7amr version (iii) run a 5 amr until orbit 3, and then restart with 6amr (iiii) revise the angular momentum projection angles plots routine.
- CRL 618. I'm working on the synthetic maps. First test of the jet with ambient rings: http://www.pas.rochester.edu/~martinhe/2011/18ocr11-crl.jpg
- AGN jet truncation. I met with Eric yesterday to discuss simulations to explore http://adsabs.harvard.edu/abs/2006MNRAS.371.1717H. We agreed in a couple of test I'll run soon, in which a light 3D hydro jet will run into a sink particle with a spherical wind of M.~10-6 Msuny-1 and vwind~100 km s-1.
- I'll be out of town this friday.
11 oct
- Binary wind capture problem. I've finished the Tenerife proceedings and registereed to the aas meeting in January, at TX. Since it's close to the holidays I was planning to change my plane tickets so I can aslo give some seminars at Mexico's Institutes of astronomy and astroph.
High resolution run is still running. I've found that the disk that forms has a radius about 4 times smaller than disks we've seen so far. Here's a 3D dens iso-contour snapshot after 3.5 orbits:
http://www.pas.rochester.edu/~martinhe/2011/binary/11oct11a.png
There's a slight tilt of the disk.
Previous run: angular momentum vector projection angle plots:
http://www.pas.rochester.edu/~martinhe/2011/binary/11oct11b.png
movie:
http://www.pas.rochester.edu/~martinhe/2011/binary/11oct11b.gif
- Magnetic tower. I've been working on the intro of the paper. I'll then work on the rotation section to include Andrea's comments, and we should then be ready to submit.
We have enough data to make a second paper that analyses the hydro vs. PFD regimes further, as Eric has been suggesting. We should talk about that with him. i.e. how would that affect the current text. Also, I want to use the data to do some statistical analysis on synthetic polarimetry for a third paper.
I've also been writing the Poland's proceedings about the mag tower+ experiments talk. The one about the tower poster is ready.
- AGN: I've been writing the proceedings about the AGN poster that I took to Poland (http://www.pas.rochester.edu/~martinhe/AGNposter.pdf). Ready for submission too.
- CRL 618: I've been trying to make shape synthetic emission, and PV, images. They are not ready yet, but I estimate another week for it.
- Testing: We've been having regular meetings. The RT module, data files and testing page (https://clover.pas.rochester.edu/trac/astrobear/wiki/TestSuite/RayleighTaylor2D) are ready. Edd has also been working in documenting the FieldLoopAdvention one. Bin is working on her radiative shock bear2fix data files and in updating their documentation. Ruka's now fluently playing with the code and has started the implementation of the 2.5D hydro Bondi accretion test. Matt is working on the testing scripts. We should have the 1st version of the suite ready very soon now.
4 oct
- I've been writing the NSF proposal.
- All crl616 run complete. We should have a telecom with Bruce.
- Binary-formed accretion disk. I ran the latest simulation from longer. Another one, with an extra refinement level, is running now.
- I'm writing the Tenerife proceedings. Want to send them by Friday.
- I'm writing the AAS, TX, abstract about the binary problem.
- How's the magnetic tower paper going?
27 sep
- CRL 618. Problem: jet run is much slower than the clump one. The chombos show maxDensclump~97, but maxDensjet(head)~900, at tcomp~0.0084. Problem in the module. Fixed and running both jet with and without rings. Quick ugly density plots:
http://www.pas.rochester.edu/~martinhe/27sep-crl618jets.png
- Advances in the binary problem. See Accretion-disc post.
- Working on the NSF proposal and different proceedings we have.
- Updates to the testing wiki page. Matt and I will get our hands on the testing scripts and related bear2fix routines this week.
- The hydro cooling case of the magnetic tower has completed:
http://www.pas.rochester.edu/~martinhe/2011/magTOWER/27sep.png. The next one is the hydro rotating which I'll run soon.
20 sep
- Binary simulations. The new run, which has a grid twice as large -in every direction- than the previous run, in on going. Should have a movie for the research meeting.
- CRL 618. The clump run, with and without ambient rings, have completed (see Bin's blog). The jet runs, with and without ambient rings, are slowly on going. We should have data to show to Bruce Balick in a week time.
- Magnetic tower. Waiting for Sergey's and Andrea's comments. Writing the talk proceedings. Running (gbene) the hydro+cooling case, as requested by Eric.
- Writing the Tenerife and Poland (AGN) proceedings.
13 Sep.
- Tenerife PN proceedings?
- I have been working on:
- Binary problem, bear2fix plots of mean angular momentum vector direction vs time
- Updating testing documentation and scripts
- Matt?
6 Sep.
-Present a summary of the MfuThree meeting, and discuss about some potential collaborations with people I saw there.
-Possibly related to the magnetic tower sims:
-Discus some ideas about the binary, wind capture, accretion disk formation simulations.
-Discus about Bin's and Ruka's (the new undergrad) projects.
15 aug New initial conditions for the binary problem which seem to affect the inclination of the disk found before.
I'll present a preview of my jet Poland talk at the journal club. I'll leave this Friday and will be back on Tuesday the 30th.
Still working on the magnetic tower paper. Plan to give the 1st draft to Adam and Eric this Thursday.
27 June
Accretion disk formation https://clover.pas.rochester.edu/trac/astrobear/blog/category/Accretion-disk. Do the resolution and outflow radius affect the average orientation of the disk?
Magnetic tower. Paper results analysis.
Bin and I have been working on the new CRL-618 runs with velocities of 200 and 500km/s.