Celestia Forums

Release:
The Flight of Buran

---------------------------------------------

After many weeks of work, we can finally present the Flight of Buran for Celestia.

This is an Add-On for Celestia 1.5 that shows the entire flight of november 15, 1988 - from lift-off,
through two full orbits around the Earth to landing at Baikonur.

The Add-On also comes with a CELX script that lets you experience the flight movie-style.
This is short - only 2.5 minutes - since i don??t like those loooong meditating scripts. It
is just an introduction, so you can discover the various maneuvers the ship goes through
on its journey. Everything is there, every burn or turn. And, it is exact.

Everything happens exactly like it did in reality, at the correct time and altitude. The only
approximation is the inclination of its orbit, which changed several times and i did not
have very exact information for.

Let me point out, for example: when the four strap-on boosters separate, it happens insanely
fast. You must slow down time to 1/100 to really appreciate what is happening. We did not
believe it when we wrote it in, and had to check and verify it - but three separate sources
agree, so that is how it happened... See for yourself the insane speed they drop back to
Earth!

A collaboration
The Flight of Buran is a collaboration between ElChristou, Andrea and myself. To be specific,
- we got permission to use the models on http://www.buran.ru. However, these are very complex, and
much to detailed to be used for Celestia. So, we had to use them as a basis only, and re-work
everything.

- Andrea did all the research and documented each detail we used
- I collected and cleaned up all the meshes, and put the models together
- ElChristou optimized all the models for real-time rendering: simplified them, in
many cases recreated them completely.
- I then texture-mapped each model, and wrote the SSC and XYZ files
- I also created the "special effects".

A note on the trajectory
The trajectory is reasonably exact - within the limits of Celestia, of course. Using Andrea??s
information, i plotted it by hand, with 5-minute intervals for the plot points. I achieved
this by putting out geostationery satellites all over its real trajectory and aligning the
XYZ file to them, hitting each satellite at the right moment. Then i removed the
satellites...

The flight
You can witness the launch, the entire flight and all maneuvers during - and you can
choose to follow Buran, the Energia rocket, or each of the A4 strap-on boosters all the way.

Buran
Buran was the Soviet Union??s version of the Space Shuttle. They look pretty similar,
but in reality, they are very different machines - the US shuttle is a proper launcher,
while Buran is only a payload for the Energia rocket. In spite of this, Buran could lift
more, fly almost twice as high, and stay in space twice as long as the shuttle.

The A4 boosters could perform soft landings via parachutes and retrofire, and the
Energia rocket was planned to be able to land on a conventional airfield - not just
a tank, it was a spaceship in its own right.

However, none of this would ever happen: Buran flew only once. The entire program
was abandoned when the Soviet Union fell apart.


Some screenshots:


(click for full size)

The Add-On can be downloaded from Celestial Matters.
Here you will also find Andrea??s article about the Buran program.

Thanks to:
- Buggs Moran for calculating the range of the boosters
- Ivvi for translating to russian for us
- Cartrite for beta-testing

... And, of course, thanks to Andrea and ElChristou for the many, many hours they have put into this project!

Enjoy!
- rthorvald

Thanks to ALL Involve it will be a KOOL add on

Amazing screenshots ! If you 1 reason to get Celestia this must be it !

Stunning work, Runar, Andrea, Chris, Buggs, and the rest of the contributors! I'm downloading this right now.

--Chris

Guys, I have to add that I haven't done 100% of my homework (because of not enough time), so later I'll send to Runar some more fixes in the models and this will be eventually an update of the addon.

ElChristou wrote:Guys, I have to add that I haven't done 100% of my homework (because of not enough time), so later I'll send to Runar some more fixes in the models and this will be eventually and update of the addon.

That is correct; a few meshes are not optimized yet. The remaining detail have no effect on the appearance or workings on the add-on, but may help get a better framerate on lower-end macines.


- rthorvald

rthorvald wrote:Buran flew only once. The entire program was abandoned when the Soviet Union fell apart.

This I never knew.


This work is just... wow!

Very !

Absolutely stunning and beautiful work. Everything turned out great. Glad I could play a small part in it.

Here's a quick shot of what I see on the launch pad...
This add-on is very very nice, and thanks to Runar and friends for
the spectacular work here.

One question though... I keep seeing the 'paused' flag on in the
upper right-hand corner. Is this supposed to be set? If I hit the space
bar to unlock, things seem to work as they should.

Not a criticism, just an observation. Many thanks once again for this
marvelous add-on.

This is an Add-On for Celestia 1.5 that shows the entire flight of november 15, 1988 - from lift-off,
through two full orbits around the Earth to landing at Baikonur.

Great works Buran looks terrible and this addon works fine on my computer, and the smoke...beautifull.
IBut my N1 pad make big flashes on celestia 1.5 i will look at this.

Did you use a calculation program to tune this flight ?


Linuxman

IBut my N1 pad make big flashes on celestia 1.5 i will look at this.

I think it is a little too close to the ground. It is really a problem with Celestia, not the pad - but it might help to float the pad a little higher.

Did you use a calculation program to tune this flight ?

No. We had good info on where it was at any time, how fast it was going, etc. So i marked 5-minute points along this trajectory (i hung up satellites as markers) and then wrote the trajectory so that it intersected each satellite at the right time. Of course, that time was an average exept for landing and takeoff.

- rthorvald

Very nice work guys (especially if it was all done in just a few weeks). Don't you guys have day-jobs!?

Just one minor comment (not a biggy)... on my machine if I double-click on the script, it starts up with the Sun at the background, which makes the white text of the menu a little hard to read.
Perhaps the script should first go to Baikonaur (or somewhere else in space) before displaying the Menu in order to start with a darker background.

How accurate is the Orbiter's orientation? Is that also based on actual data, or is there some artistic licence involved. It seems to do it's orbital manouvers, roll-overs, etc.. in just a few seconds ...(a lot faster than I expected).

Chuft-Captain wrote:How accurate is the Orbiter's orientation? Is that also based on actual data, or is there some artistic licence involved. It seems to do it's orbital manouvers, roll-overs, etc.. in just a few seconds ...(a lot faster than I expected).

Its orientation is entirely accurate, but the roll-overs happens quicker than it probably did in reality.

This is because it is impossible to animate it in real-time in an SSC document without writing, say, 12 definition per second for the entire sequence = thousands of SSC declarations. So, instead of writing, for example, 1 degree of rotation per second (which would look very choppy), i chose to make the three times it turns smooth movements in realtime.

But they happen exactly when and where they did in reality.

- rthorvald

I understand the issue Runar. I still like the addon, especially when combined with the N1-Lune addon.

rthorvald wrote:

Chuft-Captain wrote:How accurate is the Orbiter's orientation? Is that also based on actual data, or is there some artistic licence involved. It seems to do it's orbital manouvers, roll-overs, etc.. in just a few seconds ...(a lot faster than I expected).

Its orientation is entirely accurate, but the roll-overs happens quicker than it probably did in reality.

This is because it is impossible to animate it in real-time in an SSC document without writing, say, 12 definition per second for the entire sequence = thousands of SSC declarations. So, instead of writing, for example, 1 degree of rotation per second (which would look very choppy), i chose to make the three times it turns smooth movements in realtime.

But they happen exactly when and where they did in reality.

You can greatly simplify this add-on by using a SampledOrientation instead of multiple ssc instances. I can help you do this, and write some tools (like an Euler angle to quaternion converter) to make it easier to produce add-ons in the future.

Ideally, the Buran spacecraft should be a single ssc object. I know that there are some things that make this inconvenient right now, and I want to fix those. In addition to making it easier to construct sampled orientation (.q) files, I also want to add support for multiple trajectory segments, each of which can have a different orbit type and reference frame. For example, the first segment of the Buran flight is on the launch pad. You'd use an Earth body-fixed reference frame and a FixedPosition. After launch, it might be more convenient to switch to an Earth-centered equatorial frame. The boosters would be in the body fixed frame of the spacecraft initially, then in the Earth-centered equatorial frame after detaching.

Finally, I have some ideas for better rendering of ground based objects. They suffer from some depth sorting problems now, and lighting is much to harsh--scattered light from the atmosphere isn't taken into account at the moment.

Since SampledOrientation is supported already, that's the best place to start optimizing Buran.

--Chris

chris wrote:You can greatly simplify this add-on by using a SampledOrientation instead of multiple ssc instances. I can help you do this, and write some tools (like an Euler angle to quaternion converter) to make it easier to produce add-ons in the future.

That would be great because we do need more examples. Same for the Apollo mission, I'm sure reference frame ARE the way to go but actually people don't have enough (simple) examples to study...

chris wrote:You can greatly simplify this add-on by using a SampledOrientation instead of multiple ssc instances. I can help you do this, and write some tools (like an Euler angle to quaternion converter) to make it easier to produce add-ons in the future.

Chris,
This is exactly why i posted some questions about it a while back - i thought it was possible, but did not know how. I did some test, read the replies and available documentation and tried, but i just don??t understand it yet. After wrestling with the stuff for a while, i simply decided to do it the "old fashioned" way, and hope to get my head around the subject later on. Else, the production would stop for a long time...

Any help you can give on this would be greatly appreciated! The three turns the spacecraft makes:

- from vertical to upside down-and-forward at launch
- from upside-down and forward to right-side-up and backward right before retrofire
- from right-side-up-and-backward to right-side-up-and-forward before aerobraking

... Are the only segments of the Add-On that looks artificial. Getting that right would be fantastic...

Some of these movements are duplicated for the various models, since Buran is composed of several distinct objects (for example, the landing wheels is a separate model, so the holes where they come out is hidden with yet a separate model) - but this has no relevance for producing a sampledorientation file; all the models can share the same sampledorientation.

I have the SSC file as separate files for each model, to make it easier to work on, and also a julian timetable that describes the events. Would you like to take a look?

We are working on a 1.1 version already, as there are some other details we are not happy with. Getting this in place too, would really be great...

- rthorvald

rthorvald wrote:

chris wrote:You can greatly simplify this add-on by using a SampledOrientation instead of multiple ssc instances. I can help you do this, and write some tools (like an Euler angle to quaternion converter) to make it easier to produce add-ons in the future.

Chris,
This is exactly why i posted some questions about it a while back - i thought it was possible, but did not know how. I did some test, read the replies and available documentation and tried, but i just don??t understand it yet. After wrestling with the stuff for a while, i simply decided to do it the "old fashioned" way, and hope to get my head around the subject later on. Else, the production would stop for a long time...

Any help you can give on this would be greatly appreciated! The three turns the spacecraft makes:

- from vertical to upside down-and-forward at launch
- from upside-down and forward to right-side-up and backward right before retrofire
- from right-side-up-and-backward to right-side-up-and-forward before aerobraking

... Are the only segments of the Add-On that looks artificial. Getting that right would be fantastic...

Some of these movements are duplicated for the various models, since Buran is composed of several distinct objects (for example, the landing wheels is a separate model, so the holes where they come out is hidden with yet a separate model) - but this has no relevance for producing a sampledorientation file; all the models can share the same sampledorientation.

The components probably don't need to use a SampledOrientation at all--since they're fixed relative to the main model, they can just use FixedRotations.

I have the SSC file as separate files for each model, to make it easier to work on, and also a julian timetable that describes the events. Would you like to take a look?

We are working on a 1.1 version already, as there are some other details we are not happy with. Getting this in place too, would really be great...

Yes, both of these would be useful. Please mail them to me--use my gmail .com address, not shatters.net.

--Chris

The components probably don't need to use a SampledOrientation at all--since they're fixed relative to the main model, they can just use FixedRotations.

No, Each model use either Burans XYZ or have their own. (Each of the rockets have their own, while the fire / exhaust effects, landing wheels, parachute and wheel doors use Buran??s. They fly alongside the ship, not as part of it. But of course their movements are the same.

Please mail them to me--use my gmail .com address, not shatters.net.

I don??t remember your gmail, so i just upped it to my website. It is a 16 Kb zip, here:
http://runar.thorvaldsen.net/burandocs.zip

Note:
The SSC files will replace the one in the Buran Add-On. If you are on linux, you must rename them so they load in the correct order (they are numbered), but on Win/Mac it is just plug-and-play.

The timeline included is just the highlights you need. Note here that we do not know *how long* it took to turn for retrofire, but it cannot have been more than a minute or so, to fit the timetable. And for the launch, it is self-evident: it should turn with its trajectory, from lift-off until it is at about 60 km altitude, instead of doing one swift turn just before reaching orbit like it does now.

If there is anything else i can give you or help out with, please send me a note!

Thanks,
- rthorvald

rthorvald, chris,
I had the Buran Addon in the extras folder that I install the latest build from svn. I'm not sure when this first started because I when start Celestia I go to Mars but when I right click on the earth with the Buran addon active, my system locks up. The pop up that appears when right clicking on a planet never quite appears. The only choice is to shut down the program. Svn KDE version 4123 . When I remove the Buran addon everything goes back to normal.

I noticed that this system lock up does not happen with older cvs builds. When I right click on earth and chose orbiting bodies-spacecraft, there are 5 Buran entries and 13 blank entries extra. I guess this may be affecting the svn 4123 version in a strange way.
cartrite