#Orbital mechanics

1 messages · Page 1 of 1 (latest)

plain osprey
#

Patched-conics orbital mechanics like in KSP with each planetary body moving independently of each other in the sky on rails.

For the scale of the game, I feel this should be tweaked such that gravity is way higher then in real life (specifically, it should drop off much slower such that higher orbits still have you moving at a good speed), such that an orbit should take no longer then 6-7m around an average sized planet and moving between planets should ideally only take around 15m when a transfer window is available (obviously this would take much longer otherwise)

fast falcon
#

Do instead of being inversely proportional to r^2, have the force of gravity be on a lower exponent, closer to 1? Or even lower than 1? This might be interesting.

ivory rampart
#

outer wilds is epic so I'm all for it lol

midnight sun
#

It would be so fun to slingshot around a planet and see your builds from space while doing so!

storm aurora
#

I'm so sorry to rain on your parade, but gravity needs to have an r^2 dropoff if you want to be able to orbit planets. Stable orbits are only possible in r^2 gravity. If the gravity exponent is x<2, you gain more energy falling closer than it takes you to stay in orbit and you eventually spiral outwards to infinity. If it's x>2, you gain less energy falling closer than it takes you to say in orbit and you eventually spiral inwards to hit the planet. Only orbits that are perfect circles stay stable in gravity other than r^2. Off by even a hair in any direction, and you pretty rapidly spiral to doom. In r^2 gravity, the energy you gain from falling in towards the planet is exactly equal to the amount of energy it takes you to maintain orbit, so any conic section is stable (as long as it doesn't intersect with the ground)

#

(Source: I'm an aeronautical engineer with formal training in orbital mechanics.)

#

Now, if you put the planets on rails, you don't have to worry about them doing anything weird. But any craft you make won't be able to orbit them. You could still fly between planets and land on them, but you couldn't make space stations.

#

Your gravity doesn't need to extend infinitely far. You can put artifical cutoffs for SOI's to keep the math easier to calculate. But it needs to be precisely r^2.

storm aurora
#

Luckily, R of these planets is pretty low. If your world has a radius of only 6 kilometers, say, and a surface gravity equal to the earth, then orbits are 1/1000 as large and orbital velocity is 31.6 times lower, meaning you'd need only about 250 m/s to achieve orbit and a low orbit would only take about 4 minutes. It might not drop off faster than IRL, but if everything is scaled down by a factor of ~1000.... then it'll still feel root 1000 = 31.6 times faster.

#

Even a 30 km radius planet, which is quite a bit bigger than the current colossal setting, would still be about 200x smaller than IRL and feel root 200 = 14.1 times faster.

sly furnaceBOT
#

GG @storm aurora, you just advanced to level 1!

austere atlas
#

thats a lot of numbers I dont understand

pure raft
#

The thing is, the game doesn't need to have realistic physics, and also probably shouldn't. Given the scale of these planets, having a gravity inversely proportional to r^2 would result in the force dropping off really quickly when gaining just a little bit of altitude. It would be either too strong on the surface or too light in the sky (I'm thinking close to world height). The planet radius would have to be massive for gravity to appear somewhat consistent.

#

It'd make more sense for orbiting bodies to be "on rails", as it was put before

sly furnaceBOT
#

GG @pure raft, you just advanced to level 3!

austere atlas
#

like whats the problem

pure raft
#

Basically, realistic physics aren't fun

austere atlas
#

ok

#

so whats the solution?

#

like about in the middle?

#

or fully unrealistic

pure raft
#

To not make the game into a physics simulation

austere atlas
#

ok

#

it still prob has to have some logic to it tho

pure raft
#

It's been mentioned at the top of the thread, but Outer Wilds did this really well

#

It uses linear gravity for its planets instead of a squared one, which makes if feel a lot better

#

As for Planetsmith, I think you could still be able to build space stations and the like, and when you lauch them from a planet it could calculate what distance+speed would make it seem realistic, and then having the building be set on those rails

austere atlas
#

well

#

thats cool ig

#

cus yeah at the end of the day this is a game

storm aurora
#

World height is 512 meters. Even with a radius of 6 km, about the size of the current large setting, gravity would only be about 5% stronger at the 'core' and 5% weaker at the top. If I can be casually dismissed for the size of the blocks changing by that much, I feel like gravity can deal. And I'd personally like to be able to orbit planets for real, rather than fake it. The idea that this was going to have some kind of space travel and multiple planets was the main reason I got interested in this game in the first place.

pure raft
#

That's fair

#

If the "buildable" zone is within 6km and 6.5km of the core (assuming 1 hex = 1 meter) and gravity is calculated using r^2, then the relation of forces would be something like 6^2/6.5^2 which is close to 0.85, so around a 15% decrease of force (if my math is mathing correctly)

austere atlas
pure raft
#

I agree that the difference isn't as drastic as I thought when I wrote the first message and could very well still be playable, but it should probably be also tweaked for smaller planet sizes as it would be a bigger difference since you would be closer to the core (I don't know each planet size so I won't be calculating those)

pure raft
# austere atlas what grade do u learn this in?

It's just a division. Maybe the reasoning behind would be taught in secondary/ high school? And I'm thinking about physics I learned in my first year of college, but you could just look up the equations online, although going deeper into astrophysics would probably require a degree in itself

austere atlas
#

I mean I did start learning physics this year

#

rn we have only learnt about measuring or smth for some reason

pure raft
#

Still, I think it would be possible to make a structure orbit around the planet. Maybe it would be an oval instead of a circle, but still an orbit nonetheless. In simple terms, if you fall into the planet when going slow and drift away when going fast, then there has to be a point in the middle where you stay in orbit

#

Maybe to simplify things and not have to worry about being stable accounting for all celestial bodies in the system, gravity could have a range around each planet/star after which it is rounded down to 0 (maybe when below 0.05m/s^2). That way, orbiting bodies would have to stay within that space but they wouldn't be affected by other close planets (assuming those are separated by at least 2 times that same distance)

storm aurora
#

There is a point in the middle where you stay in orbit - it's a perfect circle. But only perfect circles work with anything other than r^2 gravity. If you're going precisely the right speed, you will stay at exactly the same distance from the core. But be off by even a hair's width, and you'll gain the wrong amount of energy and spiral inwards or outwards. It is a metastable point, not a stable one. With r^2 gravity dropoff, you gain exactly the right amount of energy to stay in not just a circular orbit, but any non-hyperbolic conic section orbit. It's a stable system, not a metastable one. The difference is between trying to get a pendulum to be vertical by flipping it upside down and trying to keep it balanced vs just letting it hang down on its own. You can technically balance it upside-down, but even the tiniest disturbance will have it falling over to one side or another. That's how all gravity factors other than r^2 behave.

#

And, yes, you're basically suggesting a system known as spheres of influence, which is how we normally simplify orbital mechanics for simulations. I'm not going to pretend that I wouldn't prefer to have n-body interactions (for our vessels, the planets can still be on rails), but that's how KSP does it and I'm happy to accept an SOI system if it make the game more approachable. I care more about being able to orbit planets properly than it working exactly like it should IRL. Making space stations is cool.

pure raft
#

If a stable orbit is not possible, then maybe when you approach that metastable point the game automatically puts that body "on rails", giving the illusion of a stable one. Then, whenever that body experiences any external force it is out of the rails and back on the fully simulated state. Of course, I'd make sense if the player ship was unable to enter that stable orbit unless aided by an autopilot.

#

Since we are talking about ships and stations, this could be justified as rockets giving small impulses to course-correct themselves.

storm aurora
#

That'd be better than just not being able to do it at all, but I think the easiest solution is to just use r^2 gravity in the first place and not have the problem at all.

sly furnaceBOT
#

GG @storm aurora, you just advanced to level 2!

austere atlas
#

a

#

planets

#

and smithing

storm aurora
#

I also beleive that the goal is to not make this game about advanced technology, and rather a more fantasy feel. So having computer controlled rockets seems like it might be out of scope, but just having physics that works like normal and like a steampunk jetpack for throwing yourself into orbit seems like it might be closer to what we're going for.

#

Or some kind of wind-sail powered contraption. After all, in this scenario orbital velocity is less than the speed of sound, so there are 'realistic' options for pushing structures up to those speeds with 'normal' magic and fantasy ideas.

austere atlas
#

imagine dif types

#

like a water one

#

and stuff

storm aurora
#

Yeah. I figure, a fantasy setting with planets you can travel between fits very well with steampunk as it's technological cap.

pure raft
#

Jetpacks are always cool

pure raft
storm aurora
#

That's fair. The game does have to come first; it's supposed to be a game. I'm very confident that there's no issue, but if there somehow turns out to be, it would be better to have some option rather than none.

#

Another solution would be to just not have gravity drop off until you're out of the build limit. Treat the outside of the build limit as r and have no decrease inside. That would make it so you couldn't have any low orbits, but otherwise everything would work fine.

pure raft
sly furnaceBOT
#

GG @pure raft, you just advanced to level 4!

storm aurora
#

In general, you could have any gravity function you pleased as long as it's confined to below where you want orbits to happen.

pure raft
#

Maybe the player and ship are subjected to different types of forces

storm aurora
#

As long as it's r^2 in the region you want to have orbits, you can make it whatever you want outside of those regions. The math doesn't care what the distribution looks like outside of the region you're in.

#

I still think just having it be r^2 everywhere is the easiest solution and I don't anticipate any issues from it, but if you have to change it for some reason, there are ways to do it.

plain osprey
#

as for the building system, I'd imagine space stations and the such should probs be on their own grid independent of the planet and then they could be stored independently as vehicles.
Parts for a spaceship would have to be initially built on some kind of platform that makes them sit on a flat surface instead of the slightly curved planetary one (so you can dock things together), and once they are orbiting they would be vehicles independent of any ship

#

having them be linked to the planetary build limit in any way would probs just cause issues

#

when a ship lands on a planet it would either land on some kind of landing legs that enable it to sit on the surface while still being largely independent from it or it would dock to a specialized flat block that is linked to the flat platform needed to consturct a space vessel or some sort of flexible tube that extends from the space ship or a dock on the surface or both (so that nothing gets distorted)

fading solar
#

I'm all for it. A 3d ksp-like map, with orbital lines visible, heck, even space stations would be possible. I love this idea, combined with my spaceship system this would be epic

storm aurora
#

Yeah, the thing about a realistic system is that we already know how it works. That tends to give us more options and opportunities, not fewer.

#

The game isn't a rocket simulator, and it shouldn't become one. Building a spacecraft should be simple, thrust should be assumed to be through a central axis and not generate moments, ect. But the physics of their movements should be realistic, because realistic orbital mechanics is already interesting and highly capable.

old dew
#

So, I was thinking about this, and I realized that there is another way to make the original idea work. Just change the time values, so that in-real life you still have the desired quantity of transfer windows (say, 1 every 15 minutes). This means that any orbit still works, as you still have the r^2. It could be that space time and planet time are at different speeds too, or we could presume slower planetary rotations to match.

That being said, I believe Fifteen1413 already said that the orbit time is ~4 minutes, so I think I'm finding a solution for a non-existent problem.