It is currently Tue, 12-12-17, 14:06 GMT

All times are UTC




Post new topic Reply to topic  [ 13 posts ] 
Author Message
PostPosted: Sun, 11-01-15, 12:33 GMT 
Offline
User avatar

Joined: Tue, 04-09-07, 2:32 GMT
Posts: 430
Location: South Korea
I've been working with an engineer at KARI (Korean Aerospace Research Institute - the Korean space agency) on an addon simulating the Korean Lunar Orbiter mission concept in celestia.Sci.
(the addon also works just fine using Celestia trunk code)

KARI has just begun developing a lunar orbiter and has set a goal of 2018 for launching it on a US launch vehicle (originally the plan was to launch on Korea's next gen launch vehicle KSLV-2, but budget cuts this year make that unlikely). The orbiter will be followed up by a surface rover.

This is a YouTube of the mission concept (produced before the switch to a US launch): http://youtu.be/4xyLTlhO7TU

The KARI engineer provided me with a medium-res 3ds model and .xyz trajectory. Using Blender, I split the 3ds model into different movable components such as the body, solar panels, and high-gain antenna and converted the lot to cmod using 3dstocmod and cmodfix:
Attachment:
LO-model.jpg
LO-model.jpg [ 128.33 KiB | Viewed 4062 times ]

Attachment:
LO-modelsplit.jpg
LO-modelsplit.jpg [ 95.75 KiB | Viewed 4062 times ]


Blender has a high learning curve and the 3ds output requires hand-editing to remove extraneous opacity attributes (after conversion to ascii cmod format). I wish there was something better that is also free and works on the Mac...

In the .ssc, I then linked all the components together using BodyFixed frame directives.

Here is the spacecraft in LEO (missing the upper stage), along with the trajectory overview:
Attachment:
LO-earthsep.png
LO-earthsep.png [ 273.73 KiB | Viewed 4062 times ]


I had to use the Spice orbits for the Earth and Moon, by following the instructions given in the CM Spice Kernel Files forum. There are some differences between the default orbits and Spice ones that prevent the .xyz trajectory from lining up with the default orbits. After testing again, the default orbit does provide a good match for the .xyz trajectory, so there is no need to use the Spice orbits! I must have mixed up something when testing previously.

The solar panels automatically align themselves to the Sun. This is achieved by setting the target of one the axes of the BodyFrame of the solar panels to "Sol":
Attachment:
LO-solarpanels.jpg
LO-solarpanels.jpg [ 126.67 KiB | Viewed 4062 times ]


This is the orbiter in the dark side of the Moon. Eclipse shadows on the spacecraft are handled correctly.
Attachment:
LO-lunareclipse.png
LO-lunareclipse.png [ 238.43 KiB | Viewed 4062 times ]


The spacecraft follows a roughly polar lunar orbit, coming as close as about 100 km from the lunar surface. I used JVV's LRO WAC map and the LOLA normal map from here. Planetographic grid has been turned on to show lunar longitude and latitude.
Note that the orbiter trajectory appears to intersect the Moon, but that's ok because the Moon will move away along its own orbit to prevent a premature crash ;-)
Attachment:
LO-observing.png
LO-observing.png [ 444.72 KiB | Viewed 4062 times ]


Top
 Profile  
 
PostPosted: Mon, 19-01-15, 0:18 GMT 
Offline
User avatar

Joined: Tue, 04-09-07, 2:32 GMT
Posts: 430
Location: South Korea
I noticed that with the provided trajectory .xyz file, the spacecraft would periodically speed up, slow down, then speed up again, etc.
To illustrate, here is a Matlab plot of the spacecraft velocity:
Attachment:
velocitygaps.jpg
velocitygaps.jpg [ 35.39 KiB | Viewed 4036 times ]


To see why, let's take a look at the .xyz file:
Code:
 2458330.207 -2956.863152  4765.437104  3027.647441
 2458330.207 -2957.134383  4765.280585  3027.653452
 2458330.207  -3042.98079  4756.968872  3041.183781
 2458330.208 -3244.743734  4682.744325  3039.987276
 2458330.209 -3525.597221   4541.81125  3018.195342
 2458330.209 -3856.933379  4338.323259   2973.66722
  2458330.21 -4210.611914  4078.911213  2905.378869


It seems that the time (first column) was not written with sufficient number of decimal places, causing several rows to have the same time.
So I tried to simply delete any rows with duplicated times. But this didn't work, because it caused the distance between the same timestamps to vary.
In the above example, deleting duplicate .207 and .209 entries would cause the distance to jump suddenly between .207 and .208, and .209 to .21.
Since velocity=distance/time, this causes the velocity to jump suddenly also.

The solution was to recalculate all the timestamps. I simply took the total time interval between the first row of the .xyz file and the last row, and divided by the total number of .xyz entries minus 1. Then I made all the timestamps increment by this constant time interval.

This is the resulting Matlab plot:
Attachment:
velocitynogaps.jpg
velocitynogaps.jpg [ 37.68 KiB | Viewed 4036 times ]


As you can see, there are no more gaps and moreover the velocity curve is smooth. The motion of the spacecraft in celestia.Sci is also much smoother. However, the velocities are also higher than the original; I'm still investigating why.


Top
 Profile  
 
PostPosted: Mon, 19-01-15, 1:44 GMT 
Offline
User avatar

Joined: Tue, 04-09-07, 2:32 GMT
Posts: 430
Location: South Korea
I think that the smoothed velocities are correct (i.e., the unsmoothed velocities with gaps are wrong).

Here is a plot comparing the smoothed and unsmoothed velocities of the spacecraft at launch time:
Attachment:
velocity-smoothcomp.png
velocity-smoothcomp.png [ 49.15 KiB | Viewed 4035 times ]


For both, the velocity profiles can be broken down into 3 main parts:
1. Launch: Steep increase in velocity
2. LEO orbit: Constant velocity (circular orbit)
3. Trans-lunar Injection (TLI): Delta-v increase to push spacecraft towards the Moon

For the unsmoothed profile, the velocity evens out at only ~5.5 km/s, and the delta-v of TLI is only about +2 km/s.
These numbers are too small, especially considering that a typical orbital velocity of a circular LEO should be ~8 km/s.

By comparison, the smoothed profile flattens out at ~7.8 km/s, which is a reasonable velocity for LEO. Also, TLI delta-v is roughly +3 km/s, which is also in the ballpark for typical TLI delta-v's. :D


Top
 Profile  
 
PostPosted: Fri, 23-01-15, 10:06 GMT 
Offline
User avatar

Joined: Tue, 04-09-07, 2:32 GMT
Posts: 430
Location: South Korea
While preparing this mission concept, I also took the opportunity to generate a fresh 32k normal map of the Moon using the latest 2014 LOLA data.
A previous version from 2011 is available from http://imbrium.mit.edu/EXTRAS/CELESTIA/ and also the Motherlode.

Like the 2011 version, I used LDEM_128.IMG from here: http://imbrium.mit.edu/DATA/LOLA_GDR/CYLINDRICAL/IMG/
Then I used Fridger's nmtools to generate a DXT5nm virtual texture. I took the opportunity to compile nmtools again for the Mac in order to try out nmtilesDXT (a nmtools program that generates DXT5nm tiles). This worked, although I had to change the Makefile and also compile NVIDIA texture tools from scratch.

Anyway here is a side-by-side comparison of the original 2011 version vs the 2014 normal map generated by me:

Attachment:
PDS2011vs2014.jpg
PDS2011vs2014.jpg [ 382.34 KiB | Viewed 3998 times ]


As you can see, there is a clear improvement in quality in the 2014 data.
Details are sharper, and there are less artifacts as compared to the 2011 data. It goes without saying that the DXT5nm compressed format also saves disk space: 589 MB vs 946 MB for the 2011 version (which uses PNG).


Top
 Profile  
 
PostPosted: Fri, 23-01-15, 15:00 GMT 
Offline
Site Admin
User avatar

Joined: Fri, 31-08-07, 7:01 GMT
Posts: 4514
Location: Hamburg, Germany
Hi Dawoon,

looking pretty good from what is visible above!

Good to read that my NmTools / F-TexTools packages worked for you (Ever tried my matching F-TexTools??, take versions 2.0.pre2 by all means!). After you had provided binary MacOS (universal) binaries for the early 1.0.x version, you had suddenly vanished from the "scene" for about 5 years(!), which even triggered concerned threads like this one ;-):

Where on Earth is Dirkpitt?
viewtopic.php?f=2&t=369

So no surprise that you needed to recompile the latest 2.0pre2 versions on your MAC in 2015 ;-) . For the current versions, I only was able to provide binaries for Win 7 and Linux.

The 2.0x versions run a factor of 2 faster than the earlier 1.0x versions, which is very convenient for tiling monster textures...and the like.

Actually, some years ago I did a lot of work about a 16k colored version of the Moon using the Chang'e 2 surface texture and hand-fitted LRO-Lola normalmaps.

Perhaps you are interested about my reports of this very tedious venture that used scrientific color profiles (UVVIS, 5 band colors, Clementine mission) for the coloration. The most tedious part was a by-hand matching process of the Lola NM-map to the Chang'e 2 surface texture! Unfortunately, after examining several current data sets, LRO-WAC (Nov 2011), LRO-LOLA (2011), Kaguya,..., I found regions of misalignment in most cases so far. For instance around the big Schroedinger crater etc userpix/4_moon_16k_snapshot8_1.jpg

Here are the links with MANY neat views| screenshots I think. Have a look:
In there, I have also discussed pros and cons about the surface and normalmap textures available then.


Fridger

PS: After the recent CM upgrade only Browser Fullscreen (F11 toggle for FireFox) seem to work for "kingsize" displays. Please look at the images with F11 fullscreen whenever possible. Unfortunately, the text width has turned much too large after the phpBB upgrade (due to the large image sizes). The standard add-on managing all this does unfortunately not work anymore.

If I find a little time, I might redisplay the images in form of size-adaptable "attachments"! But that is rather time consuming.

_________________
Image


Top
 Profile  
 
PostPosted: Fri, 23-01-15, 22:28 GMT 
Offline
User avatar

Joined: Tue, 04-09-07, 21:55 GMT
Posts: 775
Location: N 42.38846 W 83.45456
Quote:
After testing again, the default orbit does provide a good match for the .xyz trajectory, so there is no need to use the Spice orbits! I must have mixed up something when testing previously.

For the planets ,and the Earths Moon, the default is rather good
the biggest difference is with all the other moons and minor planets

_________________
"I don't pitch Linux to my friends, I let Microsoft do that for me."
Using OpenSUSE 42.1 & Scientific Linux 6.7


Top
 Profile  
 
PostPosted: Sat, 24-01-15, 0:24 GMT 
Offline
User avatar

Joined: Tue, 04-09-07, 2:32 GMT
Posts: 430
Location: South Korea
Yes, I'm indeed out of "hiding" and it's both a great feeling as well as a cause of anxiety... I'm in that situation where I'm looking for a job while also fresh out of grad school, so I have a lot of time (at the moment) to be able to do all these projects while also the lack of job is a constant source of pressure ^^'

t00fri wrote:
I found regions of misalignment in most cases so far. For instance around the big Schroedinger crater etc userpix/4_moon_16k_snapshot8_1.jpg


You have more locations displaying in that screenshot than me.
Which addon are you using to display all those extra locations?

I'm also aware that the LRO WAC color texture (I'm using JohnVV's version) has some shading issues. Yet even in the Chang'e 2 map, there is some (perhaps inevitable?) shadowing towards the polar regions (I'm referring to viewtopic.php?f=6&t=428). But the Chang'e coloring is definitely an improvement vs LRO WAC. Would you be able to offer this as a download?

Also what Specular color and power settings are you using?

You should definitely try the 2014 version of the LRO LOLA data. The quality is visibly improved.
Just be careful that LDEM_128.IMG starts from longitude 180 deg, meaning that after conversion with nmtools I had to shift around tile filenames to match celestia.Sci's expectation that the data start from 0 deg.


Top
 Profile  
 
PostPosted: Sat, 24-01-15, 1:16 GMT 
Offline
User avatar

Joined: Tue, 04-09-07, 21:55 GMT
Posts: 775
Location: N 42.38846 W 83.45456
i was going to start a new post
( this tangent discussion can be moved if more than just a few posts )

i just looked at the 2014 lola 512 ppd dem and compared it to the 2013 100M / 303 ppd ( remapped and bumped to 512 ) loc_wac STEREO dem
using ISIS3's "map2map"

http://lroc.sese.asu.edu/data/LRO-L-LRO ... AC_GLD100/

i like using a "Emboss" filter on 16 and 32 bit gray images
a 12 bit color lcd ( 24 bit if a "retina" apple display) i hard to read a 16/32 bit gray image
the Lola image is on the left and the loc_wac stereo DEM is on the right

i will let the images speak for them selves ( 1920x1080 )
Image Image Image
Image Image

mind you the "WAC_GLD100" is bumped UP in size to 512 ppd from 303 ppd



there is a narrow angle cam stereo dems of selected areas
http://lroc.sese.asu.edu/data/LRO-L-LRO ... P/NAC_DTM/

_________________
"I don't pitch Linux to my friends, I let Microsoft do that for me."
Using OpenSUSE 42.1 & Scientific Linux 6.7


Top
 Profile  
 
PostPosted: Mon, 26-01-15, 13:57 GMT 
Offline
Site Admin
User avatar

Joined: Fri, 31-08-07, 7:01 GMT
Posts: 4514
Location: Hamburg, Germany
John,

in the bottom-left image for example, finer surface details are clearly visible, yet there are these vertical stripes. On the other hand it is no problem whatsoever to get rid of these using the appropriate ISIS3 routine.

I think a good stategy could be to try and find the BEST matching pair of normalmap and surface map textures. Any mismatches in the raw data are much more tedious to correct for lateron...

Fridger

_________________
Image


Top
 Profile  
 
PostPosted: Sat, 25-04-15, 5:39 GMT 
Offline
User avatar

Joined: Tue, 04-09-07, 2:32 GMT
Posts: 430
Location: South Korea
Here is an animated thruster exhaust experiment.
This is done entirely using a sprite cmod and transparent particle texture, and Lua ScriptedOrbit for particle trajectories. The plume consists of 500 particles. Frame rate is just above 20 fps. Thanks to Martin (Cham) for his Tintin rocket exhaust sample, on which this is based!

Attachment:
LO1-thrusting.gif
LO1-thrusting.gif [ 475 KiB | Viewed 3698 times ]


Top
 Profile  
 
PostPosted: Sun, 26-04-15, 1:56 GMT 
Offline
User avatar

Joined: Tue, 04-09-07, 2:32 GMT
Posts: 430
Location: South Korea
I made a YouTube video of the thruster maneuver: http://youtu.be/JkB06GzPpSo

The (unrealistic!) sound effect was added at the request of the KARI engineer.
In fact the sound is played in realtime from a ScriptedRotation script using os.execute() to run a sound playing program (the builtin afplay command on OS X).
I also made a VBScript that will do the same on Windows, by loading WMPlayer.ocx.


Top
 Profile  
 
PostPosted: Mon, 06-07-15, 16:15 GMT 
Offline

Joined: Sun, 02-12-07, 19:36 GMT
Posts: 31
Hi there,

after a long, quiet time I am saying hello again.

This KARI designation just jolted me a bit, because I am currently working (real life) on another project of them the GK2.
Nothing really exciting like a designer or structural engineer mind you, I am just a paper checker (quality assurance).

Keep up the exciting work.

Best regards,

Guckytos


Top
 Profile  
 
PostPosted: Tue, 07-07-15, 0:39 GMT 
Offline
User avatar

Joined: Tue, 04-09-07, 2:32 GMT
Posts: 430
Location: South Korea
That's great to hear! I'm not very familiar with the Geo-Kompsat 2 program, but it looks like their launches could happen at around the same time (2018? 2019?) as the Korean lunar pathfinder mission.


Top
 Profile  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 13 posts ] 

All times are UTC


Who is online

Users browsing this forum: No registered users and 1 guest


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to:  
cron
Powered by phpBB® Forum Software © phpBB Group