Algol and other binaries

[Colin_hutcheson]

I may being very stupid - but does celestia have binary systems programmed into it, in particular Algol? I don't see the second star when I go to Algol, even when I'm more than the separtation of the two stars. What am I doing wrong?!

Colin

[t00fri]

I may being very stupid - but does celestia have binary systems programmed into it, in particular Algol? I don't see the second star when I go to Algol, even when I'm more than the separtation of the two stars. What am I doing wrong?!

Colin

We do have severlal hundred of visual and spectroscopic binary orbits from the best scientific sources in Celestia. You may check which we have by inspecting my visualbins.stc and spectbins.stc files in the distribution.

The file nearstars.stc moreover has some pretty multiple star orbits that are quite nearby.

Many double stars are of course lacking in Celestia, since we can only display the orbits in 3d, if ALL required data are available. Notably we need distances and masses of the components! The binaries that I entered into Celestia represent the subset where all required information is known with reasonably small uncertainties.

Algol (beta Persei) is certainly a famous eclipsing binary (actually a triple system). On the other hand, Celestia contains always entire scientific catalogs of binaries of a certain type. I have included what was contained in these catalogs (visual and spectroscopic types), provided the required 3d information was adequate.

Bye Fridger

[jogad]

Hello,


Sorry to dig out an old 3 years topic, but what about now?

I really miss Algol.

I know the accuracy of Celestia and I am grateful to its authors for this.
But we know that Algol is actually an eclipsing star (and a triple system).
To find here a single star is more than inaccurate; it is a big mistake.
Even if data are not absolutely precise, it is better than no data at all.

As an emblematic representative of eclipsing stars, I think Algol deserves special attention and greater efforts than for other anonymous stars.

A little search gave me the following 3 pages paper:

http://articles.adsabs.harvard.edu/cgi- ... lassic=YES


I am not a scientist and I'm sure that more competent people than me can find other useful data.

Would this allow to consider the entrance of Algol's system in Celestia?

[t00fri]

Jogad,

Algol (beta Perseus) is a triple star system, which immediately makes the situation considerably more complex than the rendering of binary stars.

Celestia is devoted to scientific standing, hence "incorrect" orbits will not be implemented officially. We will not make an exception for one star (Algol) of many others that have not been implemented because of incomplete orbit measurements.

Add-on creators are free to do whatever they please, of course.

But there is hope. Some time ago Andrew and I discussed already that we might collaborate in implementing this most interesting

Multiple star catalogue (MSC) (Tokovinin 1997-1999)
http://aas.aanda.org/index.php?option=c ... Itemid=129
that specifically concentrates on systems with MORE than two components (3 -7 (!)) .

It turned out that both of us had independently discovered this great multiple star catalog with growing plans to implement it...

Depending on the various orbit-radii and the observer's position, multiple systems can often be reduced in good approximation to trees of effective binary systems with some components being in reality again (narrow) binary stars.

Like so:


However, it has to be critically examined in each case, whether such assumptions would be justifiable or not. This takes plenty of extra time --which right now-- looks tight (for me at least).

This catalog does include the Algol system. However, for determining the missing parameters like the individual star magnitudes and colors, my large extensions from my forthcoming Celestia.Sci release are definitely required. No idea, how to go about this problematics... Correspondingly, Celestia.Sci will also include a complete orbit rendering of most 6th catalog visual double stars (~1300) instead of only 154 in the regular Celestia distribution. Analogously for the SB9 catalog of spectroscopic doubles.

Fridger

[chris]

This catalog does include the Algol system. However, for determining the missing parameters like the individual star magnitudes and colors, my large extensions from my forthcoming Celestia.Sci release are definitely required. No idea, how to go about this problematics... Correspondingly, Celestia.Sci will also include a complete orbit rendering of most 6th catalog visual double stars (~1300) instead of only 154 in the regular Celestia distribution. Analogously for the SB9 catalog of spectroscopic doubles.

Looking forward to using this catalog in the base Celestia package as well.

--Chris

[ajtribick]

One of the reasons that implementing the MSC would be tricky is that it overlaps with the other binaries catalogues we are using. Notably, the Pourbaix et al. (2000) catalogue used in spectbins.stc uses combined visual and double-lined spectroscopic binary orbits (VB+SB2), which give fully 3D versions of the system. If an MSC system happens to include one of these, we must then find a reliable way of mapping which subsystem is being dealt with, and which data to use. (Plus we also need to carefully determine what coordinate and orbital parameter conventions are being used: it is fairly easy to make mistakes given that there are various different conventions out there.)

[t00fri]

One of the reasons that implementing the MSC would be tricky is that it overlaps with the other binaries catalogues we are using. Notably, the Pourbaix et al. (2000) catalogue used in spectbins.stc uses combined visual and double-lined spectroscopic binary orbits (VB+SB2), which give fully 3D versions of the system. If an MSC system happens to include one of these, we must then find a reliable way of mapping which subsystem is being dealt with, and which data to use. (Plus we also need to carefully determine what coordinate and orbital parameter conventions are being used: it is fairly easy to make mistakes given that there are various different conventions out there.)

as our recent respective discussions have shown "convincingly"...
viewtopic.php?f=3&t=15994

[ajtribick]

Which reminds me, I said I was going to try to figure out getting the Upsilon Andromedae system in 3D glory into Celestia at some point...

[jogad]

Hello,

I see that things which seem simple are sometimes more complicated than expected.

In any case, I sincerely thank all of you who have considered this problem.

If you are not appalled by some approximations and since we are in the Celestia forum users I propose this small STC file until we have more reliable data. (And better used than by me.)

I took the data from the article by Pan, Shao and Colavita to the orbit of the C component.
Regarding the pair AB, I did my best by taking the values available that I could find.
The only arbitrary choice on my part is the radius of the stars. There are indeed different values from an article to another. I took "average" values (Toofri would say "for an 'average' result" ) that gave an acceptable result for me.

The result is of course far from perfect but good enough for demonstration or educational purpose.
It may in particular be used to find the periods of the minima of algol this year.


System Overview. The upper right star is the famous eclipsing pair Algol A - Algol B


Zoom to the pair A-B


And the next minimum July 10 at 1h 13mn UTC as indicated by the ephemeris.
http://www.tmsc.org/astronomy/minima.html

Sorry to have zipped the file. It is because the forum doesn't accept the "stc" extension.

[ajtribick]

Of course in multiple systems the issues of accounting for light time effects becomes more complex too. For example in this case there are going to be eclipse timing variations due to the outer orbit, which an ephemeris would need to take into account.

And for a decent rendering of Algol we probably need to have Roche lobe distortion going on.

[Hungry4info]

Your script for contact binaries.. how hard would it be to get it to produce models for non-contact binaries like this?

[VikingTechJPL]

ajtribick wrote:

Of course in multiple systems the issues of accounting for light time effects becomes more complex too. For example in this case there are going to be eclipse timing variations due to the outer orbit, which an ephemeris would need to take into account.

Yes, and there are light-time effects when viewing a binary from any place other than Earth. As discussed in this thread ( http://www.shatters.net/forum/viewtopic.php?f=2&t=15817 ) an accurate "global-light-travel-time" paradigm is an issue in Celestia within our own Solar System too—though a considerably smaller one. But the farther a binary (or anything else) is from Earth, and the farther apart a binary's components are, the greater the problem becomes. I haven't heard whether Chris is considering implementing this anytime soon (or at all); his posts suggest that it will require a major rewrite. And as Fridger cautions, there are non-Newtonian issues that should be addressed since Celestia allows observes to travel close to and faster than light.

Whatever the case, it's great to see all the work being done to bring more binaries and other multiples to Celestia.

[t00fri]

Of course in multiple systems the issues of accounting for light time effects becomes more complex too. For example in this case there are going to be eclipse timing variations due to the outer orbit, which an ephemeris would need to take into account.
...

High-precision radial-velocity techniques, are now sensitive to O((v/c)^2) relativistic effects in the data of spectroscopic binary stars (SBs), given that the typical component velocities of close binaries can be as high as v =150 km/sec, whence ? = v/c = O(10^-4). These effects can be used to derive the absolute masses of the components of eclipsing single-lined SBs and double-lined SBs from Doppler measurements alone. High-precision stellar spectroscopy can thus substantially increase the number of measured stellar masses...

Here is the original paper
http://arxiv.org/abs/astro-ph/0609753

Quite an interesting application of special relativity!

Fridger

[ajtribick]

Your script for contact binaries.. how hard would it be to get it to produce models for non-contact binaries like this?

While there's definitely stuff in there that can be reused, there are enough differences to go about doing it a different way. For starters, the UV mapping I use for the contact binaries maps the top of the map to one "end" of the binary, while for normal stars the UV mapping puts the top of the map at the rotation pole. There may be other issues to do with the geometry.

In any case, my gut feeling is that such routines properly belong inside Celestia, so you can just specify the relevant parameters in the .stc and Celestia will calculate the shape for you.

[Hungry4info]

In any case, my gut feeling is that such routines properly belong inside Celestia, so you can just specify the relevant parameters in the .stc and Celestia will calculate the shape for you.

While I agree, it may not be within Celestia's design philosophy to calculate such things. I recall asking about Celestia calculating tidal heating and being told that Celestia is less a physics simulator, and more a visualisation tool. I suspect that such answer is applicable to simulating tidal distortion of stellar shape, which in any case would require more parameters than are typically declared in a .stc file.

[ajtribick]

In any case, my gut feeling is that such routines properly belong inside Celestia, so you can just specify the relevant parameters in the .stc and Celestia will calculate the shape for you.

While I agree, it may not be within Celestia's design philosophy to calculate such things. I recall asking about Celestia calculating tidal heating and being told that Celestia is less a physics simulator, and more a visualisation tool. I suspect that such answer is applicable to simulating tidal distortion of stellar shape, which in any case would require more parameters than are typically declared in a .stc file.

I think Roche lobe rendering does fall into the design philosophy: it is after all a visual rendering, unlike tidal heating which is calculation of some temperature quantity that does not actually affect the rendering. In any case, the first approximation to the Roche lobe shape (point masses surrounded by massless fluid) can be specified using very few parameters: in fact IIRC you can get all the information directly out of everything that is already specified in the .stc, provided you are willing to assume the system is Keplerian. (Failing that you could add into the .stc the masses of the stars).

Of course in the real case the non-spherical star would distort this somewhat, but I think such approximations are well accepted within Celestia (e.g. representing rapidly rotating stars as oblate spheroids, which applies in the case of uniform density rather than for actual stars)

[eburacum45]

A year or two ago I made a rough-n-ready model of Algol for Orion's Arm; it includes a transparent, flat billboard between the stars with a hand-drawn texture to simulate matter transfer.
This page has an image of the model, although I must admit I smeared the image of orange star in Gimp afterwards to give it a non-spherical look.
http://www.orionsarm.com/eg-article/45ecdf18385f5

I based the model on images of the system found on the Internet, like this one
http://www.glyphweb.com/esky/stars/algol.html

I'd really like to see an accurate version of this system one day, with the shape and mass transfer included somehow.