General Question Orbital decay

[thor36]

Greetings fellows ! :tiphat:
I am new to Orbiter world, have just started running a few scenarios, taking things easy and getting acquainted with the interface. I must say it doesn't feel easy to me a.t.m., but I'm not giving up =)

Alright, now to the real thing. I ran a demo scenario with ISS orbiting the Earth with a space shuttle docked, and ran the simulation for a while. I noticed that the Apogee and Perigee values were not decreasing, but after some time even increased by the last digits. I don't understand this, as I know that the drag causes orbital decay and is a serious issue that the ISS operators are addressing by giving pushes to the ISS.

Thank you for reading, and I hope I picked the right subforum

 

[Poltergeist]

Greetings fellows ! :tiphat:
I am new to Orbiter world, have just started running a few scenarios, taking things easy and getting acquainted with the interface. I must say it doesn't feel easy to me a.t.m., but I'm not giving up =)

Alright, now to the real thing. I ran a demo scenario with ISS orbiting the Earth with a space shuttle docked, and ran the simulation for a while. I noticed that the Apogee and Perigee values were not decreasing, but after some time even increased by the last digits. I don't understand this, as I know that the drag causes orbital decay and is a serious issue that the ISS operators are addressing by giving pushes to the ISS.

Thank you for reading, and I hope I picked the right subforum

It is true that the ISS needs boosts now and again due to orbital decay, but think for a second about how it would be if this occurred in Orbiter. Lets say you go for an interplanetary trip, and plan to rendezvous with the ISS after you return and park the ship there for a while. If orbital decay was implemented, there would be no ISS to come back to. Unless you feel like stopping and saving your progress on the way to a destination, and sending a mission (or using a already docked craft) to boost the orbit of the ISS often. It's a cool, realistic thing, yet would make Orbiter less fun. Just imagine that when you aren't looking a Soyuz TMA comes up and does the orbital boost :thumbup:

 

[francisdrake]

Hi, welcome to the Orbiter forum.
We chose to fly Orbiter not because it is easy, but because it is hard !

You can adjust the atmospheric model in the Orbiter launchpad under
"Extra", + Celestical body configuration, + Atmosphere configuration

See Orbiter Doc\Technotes\earth_atm.pdf for details.

 

[thor36]

Poltergeist, thank you for reply =)
I agree, in a kind of situation that you have described orbital decay would be very unwelcome. However, in the example of ISS, the station does actually get boosts, and we could say "ok, let's forget about orbital decay for the ISS". But if one tried to park a spacecraft in an orbit, maybe deploy a satellite etc... Orbital decay would be an important factor to observe when dealing with long duration missions. So for example if I wanted to try this out ( had I enough programming skills ), how should I implement it, what formulas to use, what propagation algorithms ?

 

[romanasul]

Poltergeist, thank you for reply =)
I agree, in a kind of situation that you have described orbital decay would be very unwelcome. However, in the example of ISS, the station does actually get boosts, and we could say "ok, let's forget about orbital decay for the ISS". But if one tried to park a spacecraft in an orbit, maybe deploy a satellite etc... Orbital decay would be an important factor to observe when dealing with long duration missions. So for example if I wanted to try this out ( had I enough programming skills ), how should I implement it, what formulas to use, what propagation algorithms ?

Orbiter uses orbit stabilization which can be found in the launchpad under Extra/Time Propagation. This basically makes sure that your vessels and stations don't fall out of the sky under high time acceleration. If you want you can disable it and test it out.

 

[thor36]

Hi francisdrake, thank you for the welcome ! I missed your post the first time.
Haha yes , and by the end of this decade I will land D-Glider on the Moon

@romanasul - I checked the option you have mentioned and unticked it, but the results I get then with "Shuttle docked to ISS" get really crazy, I'll need to take a deeper look at what's happening there.

francisdrake, do you know how the atmospheric model from the doc file is used in the orbit propagation ? I did a little searching online and found out that there's a whole set of things to consider when doing orbit predictions ( planet oblateness, atmosphere, solar radiation ... ), and that only few propagators include them ( HPOP for example, and another one with a crazy enough name that I couldn't remember ). Do you know what is used in Orbiter ?
Thank you !

 

[francisdrake]

I'd say the physics engine of Orbiter is pretty impressive! It is much more sophisticated than a 'game' would require.

Supersonic wave drag - I assume - is calculated based on the atmospheric density given by the atmospheric model, the speed of the vessel, the cross section of the vessel and - probably - a simplified shape of the vessel. Don't know for sure.

In the launch pad "Parameters" tab nonspherical gravity sources (=planet oblateness) and radiation pressure can be turned on, along with gravity-gradient (=difference between mass center and gravity center of a vessel).

If you are new to Orbiter I suggest to leave all these extras off until you have a good grasp on flying and astrodynamics.

 

[boogabooga]

Alright, now to the real thing. I ran a demo scenario with ISS orbiting the Earth with a space shuttle docked, and ran the simulation for a while. I noticed that the Apogee and Perigee values were not decreasing, but after some time even increased by the last digits. I don't understand this, as I know that the drag causes orbital decay and is a serious issue that the ISS operators are addressing by giving pushes to the ISS.

On what timescale are you talking about? Over the course of one orbit, it is usual to see apogee and perigee "increasing" and "decreasing"- this is an effect of non-spherical gravity. Actually, it is the reading on Orbit MFD that increases and decreases because it calculates your ideal orbit at every instant- your actual orbit will be fairly stable, but not quite what Orbit MFD predicts.

Atmospheric drag is calculated to an altitude of about 2000 km. However, as mentioned, it is disabled at high time acceleration-for good reason. Never forget that Orbiter is a numerical simulation and susceptible to "non-physical" results if the time-step is too high. In other words, you don't want to mess with it.

If you want to see atmospheric drag in action, put the simulation on about 1000x acceleration and wait about 2 real-world hours.

 

[thor36]

Well, I kind of agree about getting a good grasp on flying and astrodynamics, that is probably the primary educational goal of Orbiter. But I don't think it is bad if I am a little bit more curious about what is happening in the program, what things are being calculated and what not, and how that is being done. I find it fun and motivating =) I can leave the detailed questions for later though.

@boogabooga - I am talking about much bigger timescales than few orbits, I fired up the time warp to maximum value and watched what would happen. I was expecting the "average" altitude to be gradually decreasing, as well as perigee and apogee due to, not least, the effects we've been discussing in this thread. But values were increasing and decreasing to extremely low to extremely high in an alternating fashion. I thought about numerical computational issues as well, there is probably a warning regarding using big time warps somewhere in the manual, that I hadn't reached yet.

---------- Post added at 05:08 PM ---------- Previous post was at 12:44 AM ----------

Another funny thing this time... I ran the DG in circular Earth orbit scenario, pushed the DG higher in altitude to get into a higher, elliptical orbit ( in normal time speed ) , and then sped the time a little to observe the motion of the spacecraft "square" around the ellipse on the monitor. What happened then is the elliptical orbit turned into a hyperbolical one, with DG sailing away for good. What happened in the simulation ?

 

[N_Molson]

If you want to keep high precision, avoid time warp > x10 in LEO. x100 clearly reduces the accuracy of the simulation, and anything higher should be reserved to interplanetary trips. For an accurate Earth to Moon flight I recommend a maximum of x100, with x10 until you're out of the atmospheric model (altitude > 2500 km). Of course, you might have to let run the computer at night. To experience the simulator at its best, time acceleration should be kept at x1. But yeah, one week for a Lunar Pass is a bit long.

 

[thor36]

Haha, sure, and doing Voyager mission reconstruction with x1 would be a blast too. Thanks for the explanation, it all makes sense now.

Is all the propagation in Orbiter done with Runge-Kutta methods ? With the equation being the sum of all forces acting on the spacecraft ?

 

[N_Molson]

Is all the propagation in Orbiter done with Runge-Kutta methods ? With the equation being the sum of all forces acting on the spacecraft ?

Such advanced physics are chinese to me, but I think that the exact equations are somewhere in the documentation. That the kind of things that martins is probably the only one to know (I dare to say that even super-Urwumpe will fail at this point).

 

[orb]

Is all the propagation in Orbiter done with Runge-Kutta methods ?

By default, yes, but you can change the propagators on the Extra tab in the Orbiter Launchpad (there's additionally Orbit stabilization, which you can change, too, or disable).

 

[boogabooga]

If you want to keep high precision, avoid time warp > x10 in LEO. x100 clearly reduces the accuracy of the simulation, and anything higher should be reserved to interplanetary trips. For an accurate Earth to Moon flight I recommend a maximum of x100, with x10 until you're out of the atmospheric model (altitude > 2500 km). Of course, you might have to let run the computer at night. To experience the simulator at its best, time acceleration should be kept at x1. But yeah, one week for a Lunar Pass is a bit long.

I use a rule of thumb to not let more than one orbital period pass in every real world second. So 1 000x is okay (but pushing it) for LEO, 10 000x for HEO, lunar transfers, etc. But the 100 000x should be reserved for heliocentric operations- and even at that reserved for long-period trips to outer planets.

 

[thor36]

I use a rule of thumb to not let more than one orbital period pass in every real world second. So 1 000x is okay (but pushing it) for LEO, 10 000x for HEO, lunar transfers, etc. But the 100 000x should be reserved for heliocentric operations- and even at that reserved for long-period trips to outer planets.

Wow, your rule of thumb seems quite dare at first sight, are your simulations actually performing well with such time warps ?

@N_Molson - I suppose you meant technotes/dynamics.pdf , that's the text I have found while looking for equations. I believe I have found my answer there. Though exactly what I was looking for is not stated directly ( probably because you could almost write a book about it ! ) I deduced the answer, hopefully correctly. I admire how you folks know astrodynamics so well and understand all that's going on while flying spacecraft, without being bothered about the details of what kind of routines run in the background !:hail: Now that's Chinese to me !

 

[VincentMcConnell]

You can also get your orbit to perturb realistically if you turn:
Nonspherical gravity sources,
Radiation Pressure,
Gravity Gradient torque

ON in the parameters tab in the Orbiter Launchpad startup. It makes things a bit harder, as your active vessel or target orbit keeps changing -- but it's a bit more realistic.

 

[thor36]

Thank you for the tips VincentMcConnell. I am trying to make some simple software of my own, and would use Orbiter to compare results. I'm still not entirely sure about the actual implementation of orbit propagation, I have an idea but it seems just too crude and simple to me to seriously consider it