New magnetic field configurations (work in progress)

[Cham]

I'm still waiting for a proper interpretation to my magnetic evolution mystery (Fridger, please !? I've also posted to some physics discussion groups and nobody appears to know). I'm now experimenting with another interesting field configuration : two magnetic dipoles in (static) interaction. It could be very interesting to explore this field in 3D. I intend to use it for an imaginary world of two magnetic planets in orbit around their barycenter (a planet and a big moon, if you prefer). Here's a preview as seen in Mathematica :



Of course, this one isn't finished at all yet. The final field will be much better looking.

EDIT : GEEZ ! I'm now having enough stuff for MONTHS ! (if not YEARS !). That new configuration is giving lots of EXTREMELLY interesting results (well, only for people who are interested in magnetic fields, I guess). There are so much parameters to vary (both dipoles orientations and relative strength, distance, etc), and for each set, the field has a very different pattern. Mathematica is doing a great job here, and I can now export all the lines directly to Celestia. This is trully fantastic !

The currently active CMOD projects I'm working on are :

1- finishing the rotating magnetic dipoles with radiation
2- light (and massive particles) geodesics in Kerr geometry
3- that new two dipoles configuration

I don't think I'll work on configurations with three dipoles or more, even if my Mathematica code is now trivial to modify. I don't think I'll add any quadrupole field for a very long time, since I already have so much to do.

I now feel like an over excited child ! Time will be my main problem now, especially since the job is calling back (new semester is beginning this monday).

[Cham]

Just for fun.

here are three gif pictures of the pair configuration :

[ANDREA]

Just for fun. here are three gif pictures of the pair configuration

Wow Cham, from a scientific point of view I absolutely know nothing on magnetic fields, but graphically I feel your work is extraordinary.
Please, as I already told some time ago, go on this way, even if you have no feedbacks, be sure that many people like me are following with attention your work.
Bye and thank you.

Andrea

[Cham]

In the case of the field shown in the previous B&W pictures, I made a small exponent mistake for the dipolar field.

The picture below shows the two dipoles interacting. Can you feel the magnet bars with iron filings?

[Cham]

Here's a small prototype of the magnetic planets. The match is perfect :



Because Celestia can't show evolving field lines, this kind of models can't be used for excentric orbits.

[Cham]

Here's a large model in Celestia. The frame rate is extremelly smooth !
(click to enlarge). The planets aren't there yet.

[jll]

Amazing work

It's funny how this video looks like your model http://www.youtube.com/watch?v=xVgPplOgB1g&mode=related&search= .

JLL

[Cham]

It's funny how this video looks like your model http://www.youtube.com/watch?v=xVgPplOgB1g&mode=related&search= .

This isn't the same beast at all. The "waves" you see in that video are a representation of "quadrupolar gravitational radiation". It isn't magnetic stuff.

[Cham]

Here's a magnetic model for testing purposes. Please, feedback would be appreciated. How many FPS do you get ? What are your impressions on this field ?

http://nho.ohn.free.fr/celestia/Cham/Magnetic_test.zip (3.8 MB zip file)

The magnetic "planets" (or stars ?) are temporarily located in the solar system.

[selden]

It looks great, although the version I downloaded isn't blue.

It also runs full speed (60fps) on my home system, but that's no longer close to being typical.

System:
1GB 3.4GHz P4-550, WinXP Pro SP2
256MB GF 7800GTX, ForceWare v93.71
Celestia from cvs

[t00fri]

It looks great, although the version I downloaded isn't blue.

It also runs full speed (60fps) on my home system, but that's no longer close to being typical.

System:
1GB 3.4GHz P4-550, WinXP Pro SP2
256MB GF 7800GTX, ForceWare v93.71
Celestia from cvs

Selden,

why so little RAM?? That's indeed not typical anymore
It's good for email, though...certainly not for doing any kind of image manipulation. RAM is more important than a fast CPU.

[selden]

Fridger,


That's what I got with the system two years ago and it works fine for me: memory occupancy is usually less than 512MB. I don't do large image manipulations. What few images I manipulate have to be 1K or less since they're model surface textures and need to be viewable with MS OpenGL software drivers.

[selden]

The model runs at 15-20fps on my system at work. It's also not so typical, though:

2GB 1.86GHz Core2Duo, WinXP Pro SP2
128MB Quadro FX 550 ForceWare v91.36
Celestia from cvs

Benchmarks seem to indicate that each of the CPUs in this Core2Duo has about the same performance as the CPU in my home system. Apparently the speed of OpenGL lines depends primarily on the performance of the graphics card.

[Cham]

I had a discussion with a friend about the magnetic field lines color. What color should be the lines ? Is there any convention about that ? After few experimentations, I've found that the current translucent yellow is very good, while red is generally odd. Blue is very good too but I feel it's less appropriate for magnetic field lines, I don't know why (maybe because blue reminds electric effects and is more appropriate for electric fields ?). Green is nice too.

I suspect that yellow is nice because it reminds in some way the hot plasma flowing on the Sun's magnetic field lines, as seen on popular pictures.

EDIT : Here's a nice variation of the previous model. I think it could be a nice one for a white dwarfs binary (WD also have strong magnetic fields with a dominant dipolar component) :

[buggs_moran]

I personally like the yellow (gold) or blue the most. That white dwarf interaction is fantastic.

[Boux]

Beautiful!
I like the gold color.
Blue/purple is not bad either.
Over 150 fps here with 4xAA, 8xAF, quality max.
Well, it is not a typical system

[Cham]

Here's a field configuration with a charged particle moving in it. The particle is a Carbon 12 nucleus pitched at 80% light speed. The magnetic field strength at a star's surface is about 100 gauss (about 100 times our Sun's mean field at its surface). I know the field strength is very modest for a white dwarf, but setting a stronger field gives some tiny loops (1 mm !) on the particle's helicoidal paths. Since there's a mathematical scale invariance, you can also interpret that field as a much stronger one (with, say, a 1000000 gauss field strength) and a much more massive particle with an higher velocity, so the trajectory isn't really affected.

The blue curve shown here takes special relativity into account and the particle takes 49.8 sec to move on it !

I've also added a very nice fade-in and fade-out effect on the particle's path, so there isn't any apparent beginning and ending cuts on the curve.

[Cham]

I just found an extraordinary path. The charged particle is coming from infinity, get trapped in the field for a while, doing all sorts of crazy things, then return to infinity :



The purple curve shown on this picture is really the same relativistic particle moving all the way down until it escape, all this in 2 min 19 sec !

Unfortunately, I wont release that path since the numerical resolution doesn't take into acount the fact that the white dwarf binary is rotating. That curve is valid only for a static field. The curves I'm building are valid for the rotating field only if the time spent in the field is small compared to the rotation period of the binary (13 min).

[buggs_moran]

Oh that is fun! Maybe someday we could follow the path in Celestia? Can you export the cmod as an xyz? That might be neat too...

[Cham]

Oh that is fun! Maybe someday we could follow the path in Celestia? Can you export the cmod as an xyz? That might be neat too...

I already tried an xyz file, back in september, for my Magnetic-Earth addon. It didn't worked since the particle was taking few seconds to follow its entire trajectory shown in Celestia.

In the case shown above, it takes about 2 minutes to follow a pretty long trajectory : the particle is moving at 80% light speed in a space roughly equivalent to our Earth-Moon system. So I don't think it would be visible in real time, in Celestia (too fast).

And there's a second, more important problem here : the field is actually rotating in 13 minutes since its source are two white dwarfs moving around their barycenter. The trajectory shown above is valid for a static field only and was a testing prototype only.

I'm thinking about integrating a real trajectory, in Mathematica, for a rotating field. Once done, the complete trajectory (defined for the time of a complete rotation period) would be shown as a static curve in Celestia, while the field is steadily rotating. I'll have to add two reference points rotating with the field , associated to the starting and ending points on the curve . That way, the user could see **when** the particle begins its journey, and **when** it's ending it.

I'm not sure if this willl works, since I'll have to find some initial conditions which brings the particle inside the field and brings it back outside after a complete rotation period.