Author Topic: VSFL (Virtual SpaceFlight Network)  (Read 1308 times)

Offline Black Phoenix

  • Member
  • Posts: 1
  • Kyiv, Ukraine
    • Black Phoenix's Brain Blog
  • Liked: 0
  • Likes Given: 0
VSFL (Virtual SpaceFlight Network)
« on: 08/13/2012 08:20 PM »
This is a project I work on as a hobby. It's still mostly closed to the public, though by now it reached a stage when I'm interested in finding more interested people.



VSFL Network is an online spaceflight network oriented towards the aerospace engineering enthusiasts. In simple words it's a spaceflight simulator which runs 24/7 and propagates internal and external state of all vessels in its orbit in realtime.

This simulation is interfaced with other simulations through various ways (networked protocol to communicate with a simulator I wrote extending X-Plane, and in future towards Orbiter and my own separate simulator).

It doesn't have any renderer yet, but it's actually a simulator I originally wrote to extend X-Plane flight simulator:


It serves two goals actually - I want to learn more about aerospace (I don't have money to build any rockets, gotta use mathematical models instead), and it would be fun to work with people in learning together.

I've only got the nerve to post this because I got a working virtual satellite in orbit now, it transmits time and magnetic field data continiously.


Maybe the whole idea will be more clear if I list features (these are features of the simulation itself and the website):
* Synchronizes state between all clients (aka multiplayer, unified protocol for multiplayer spaceflight).
* Online flight tracking, keeps statistics about vessels and their flight
* Radio transmission simulation: spacecraft are able to communicate using radiowaves on a predefined set of channels. Right now the basic wave propagation is simulated (line of sight check, a model which accounts for propagation time and atmospheric effects is written out but not programmed yet). The model is digital transmissions only. There can't be analog signals, I don't have CPU power for that.
* Persistency: the realistic server allows launching payloads into orbit permanently. They will be calculated on server, and will stay there 24/7. It is possible to have payloads with some additional logic (for example flight computers, electric systems, etc).
* Online docking: work-in-progress. The VSFL Network will allow docking several vessels together (temporary or permanently - aka build a space station). This is mostly for playing out rezendvous scenarios.
* Roleplaying for aerospace engineering enthusiasts: any VSA/VA registered on the website can publish their materials, register their vessels (realistic server keeps strict track of allocated and used vessels ), publish mission schedules, make use of the realistic server. It's all about launching and designing spacecraft.
* Ground tracking system simulation - the "virtual company" that I do stuff under, XSAG, provides ground tracking network with six telemetry downlink/command uplink channels. They can be used, but alternatively a "good enough" VSA can create own ground tracking network.
* Full physics for objects in orbit - atmospheric drag, non-linear gravity, magnetic model included for navigation, perturbations from other bodies. Atmospheric drag is simulated based on satellite geometry and shape. Re-entry heating is simulated, but right now the simulation cuts off at 100 km (formal reentry altitude).
* The scope of VSFL Network covers Earth & Moon at the moment, though that's going to be solved in distant future. If you want to fly beyond Earth & Moon you have to use your own simulator.
* Internal Vessels Systems Simulator. A side project I'm working on, you define systems of a spacecraft (electric wires, sensors, computers, hydraulics, solar panels, etc) and it models them in realtime.


Okay, so there's still a lot of things missing from here: docking together (only parent-child stacking is supported, not arbitrary docking), lacks perturbations from Sun/Jupiter, right now Moon is in wrong position (just a simple fix to that though), no solar pressure. Lots of small details missing (no inertia coupling, lots of hacks for X-Plane, no support for arbitrary coordinate systems, just the predefined ones).

I will fix that sometime. I want to write a new simulator, from scratch (the server will move seamlessly).

So now it's time for some images! I'm sorry that most flights so far did not carry any cameras or anything, so it's mostly telemetry.

This is the tracking system available so far:


The whole thing is welcoming towards working on aerospace designs - unbuilt spacecraft, or making new systems for old ones. Pretty much backseat engineering, it's encouraged to play with concepts (if you need to go beyond simply running it in your local simulator).

My own VSA XSAG (eXperimental Space & Aeronautics Group, just a name that I used since it was just messing around) is mostly dedicated to developing new highly-detailed models of spacecraft and aircraft (guidance computers and software, internal systems etc). This is a separate topic to discuss :)


To give you a general idea of the aerospace focus of this network, here's a short summary of first fivesix space missions done so far (so far they all have been done with a derivative of Ares/Orion stack, both SRB and Zenit as first stage). All the missions have been autonomous, meaning that I only rely on data from telemetry (where applies):
* XS1A, XS1B: qualification suborbital flights testing the rocket. About 12 flight control system issues were found and fixed during these flights
* XS1: first orbital flight, mostly served as testing of a slightly updated guidance system and autonomous flight capability of the capsule
* XS2: second orbital flight, testing long-term autonomous flight. This flight lasted 1 day (and 16 orbits). There has been a scanning camera on-board, and it downlinked a fairly large amount of pictures (see below). Also has tested command uplinking from ground stations.
* XS3A: qualification flight of a modified RV-553 rocket (Ares upper stage, Zenit first stage, dummy satellite as a payload). This has launched two objects into permanent orbit (well, decaying).
* XS3B: qualification flight of the third stage that should get my payloads into higher orbits (10,000 kg into LEO, 4,000 kg into MEO, 2,000 kg into GEO)

All the missions were flown by the computer, I only worked with it via the radio transmissions (simulated: spacecraft encodes binary data, ground stations decode binary data and feed it to the website).


Selected data from various missions:
An example of a downlinked image (you can see slight noise from the errors in transmission):


Downlinked XS3B ascent telemetry (runs on a guidance system that's roughly based on PEG, noise can be seen during switchover from via-pad telemetry to Sary Shagan tracking):




Realtime heating:
Something I'm working on, only used in X-Space. Estimates heating by roughly calculating heating transfer between TPS, hull, environment, based on empiric models and just generally wrong. But gives at least some data...

Might be interesting too... in scope of a potential aerospace simulator maybe.


Involvement:
Appreciated: suggestions for improvement, pointing out errors (physics errors, errors in guidance, various bad properties of control systems preferably by the last operational flight).

It takes plenty of work to get one design running (though I've been mostly doing stuff from scratch), so enthusiasts are appreciated too. I generally like concept designs, but no need to post them here (PM or e-mail them to me if you want).

Satellites! The network can fly satellites, and I have a 8086 VM for them right now. It would be cool to make some more satellites (maybe test real world designs!), any virtual machine I can control via CLI or link up to my server will work. 8086 works, but I also have (untested) VM for ARM, PowerPC, z80, 6502, 68000 (good support claimed), some vague small development boards.

I am severely limited in CPU time, but a few operational satellites using low-end CPU's shouldn't hurt too much.


More information (website):
http://vsfl.wireos.com/ (the main website, flight archives)
http://vsfl.wireos.com/realtime-tracking (realtime tracking)
http://vsfl.wireos.com/launch-schedule (launch schedule)
http://vsfl.wireos.com/vessels (list of all registered/launched vessels)
http://vsfl.wireos.com/archives/364 (all of the downlinked images for XS2)

http://vsfl.wireos.com/screens/xsag_es602.php (telemetry from the satellite, available only while in coverage of Evpatoriya).
http://vsfl.wireos.com/screens/xsag_rv550.php (RV-55x, which is Ares/Zenit, tracking system, offline since there's no rocket).
vsfl.wireos.com:19500 (TCP port that streams data from time satellite. The format is http://www.everfall.com/paste/id.php?c70s0sr0hnat if you want to parse it yourself. Telnet there if you want to have random characters running over your screen?).


I will not apologize for materials on the website being too rough, because I will loose my sanity if I try to fix everything before presenting it to public. The whole thing is incomplete and I'm not actually an aerospace engineer, it's just my hobby.

I've typed everything up and probably made mistakes, I'll try to fix those.

I would love to work on something real and tangible, but my location cuts me off from all fun places, and I can't just sit and do nothing. I'm not satisfied with small model rockets. Hosting this on my home computer.

Tags: