Science Forums 2

[CraigD]

Originally Posted by TheBigDog

How far beyond a full orbit does it go? It should be short by 278.349 meters in 86164 seconds considering the sidereal rotation time of the earth.

A lot beyond – about 119001 m, or 39 s.

I suspect my high-precision calculations of GEOSTAT orbit are off vs. the constant of gravity and Earth mass in the simulation – I used slightly different sources calculating them. I can tweak the orbital radius up ~ 6315 m to get nearly exactly a 86164 s orbital period, like this:

XGRAVSIM4>Earth.M=5.9736e24
XGRAVSIM4>Ship.M=10 .L=42170455,0,0    .V=0,3074.661175977908351,0
XGRAVSIM4>$.t=43082 ship.l .v
#SHIP.L=-42164136.41712641552,-11.14168741885003,0 .V=.0008125255939590325,-3075.121934885605702,0
#note:                 42164136.41712788759 3075.121934885713047
XGRAVSIM4>$.t=86164 ship.l .v
#SHIP.L=42170454.9999941175,22.279930461983273,0 .V=-.0016245603985481512,3074.661175977479037,0
#note:                 42170455.00000000309 3074.661175977908222

though the resulting orbit is still slightly (~ 6319 m) non-circular.

All in all, though, I pretty happy with the accuracy of my GRAVSIM* simulators – I’ve used them to practice various maneuvers, mostly in Earth orbit, and they always agree pretty closely with published data. I much prefer the new, object-y syntax, though still need to add some essential useful $ (simulator/system) object properties, such as a “dump all” function – small, easy enhancements.

As I mentioned in my previous post, the neatest thing about GRAVSIM4 is that you can add “programs” to each object, the ship being the most interesting object. Since that’s all done with MUMPS xecute code, it’s easy to make these programs “modular” – a rocket motor program can run with throttle and thrust vector data from a guidance program, etc. – and “big bump” events like stage separations can be separate programs that wait for a defined condition to adjust the ship’s net mass and current engine characteristics. Weird stuff like atmospheric drag (for launch or aerobraking), magnetic or solar sail propulsion (not in scope for an orbit-to-moon mission, but handy in general) can be written and dropped in. I hope to build a small library of useful .X programs ASAP, with parameters for real-world commercial motors, making it possible to quickly try different configurations of ship hardware.

Guidance programming remains the hard part. It’s real rocket science, and not something I’ve ever done professionally, so rather over my head (unless I can smooze in with some local NASA folk ). I’ve a tendency to come up with rough calculations, then “fly by the seat of my pants” to compensate for their roughness – not, I’m pretty sure, how the pros do it.

Graphics are another problem area – I don’t have anything handy that will give me a real-time graphic diagram, and find myself spending a lot of time struggling to visualize the numeric data. Though a browser interface would be slickest, I’ll probably write a small VB app to run separate from my terminal window – but either approach will take worktime, which I’ve not got a lot of extra free right now. I’d also like to learn more latex, so that I could put graphics in posts using [math] tags, though this isn’t very helpful for realtime visualizing.

For the next day or so, I fear I’ll be consumed by the much-less interesting (but better paying ) world of SOX. Gah!