Calculate future positions of masses using orbital elements

[TheHappyKoala]

Happy almost Monday 😄 !

I remember you'd written some code for the Lambert solver that would stop it from calculating a trajectory that would pass through the planet the spacecraft is departing from, since, generally speaking, such trajectories do not make for healthy spacecraft 😅.

I'm asking because chances are I've consigned that code to the dust bin of history with some of the changes that I made to the solver last summer. Alternatively, that check is only done with respect to the planet whose sphere of influence the spacecraft is in, the issue being that when travelling from Earth to, say, Mars, the trajectory is actually calculated with respect to the Sun (yields much more accurate results, but when calculating a trajectory to the Moon, you do so with respect to Earth, still).

Not sure I'm making much sense here, but if there's a chance you remember anything related to the above, lemme know 😄 . Aint a big issue but I've probably lost 120 crews today alone 😭

[HugoGranstrom]

I'm afraid I never wrote any code that did that sadly 😅 I may have said that we would eventually need it (which we do obviously 😂) but no code was ever written. My reasoning was along the lines of that the nearest distance to the starting body (perihelion) of the orbit must be greater than the starting body's radius. But I never came any longer with that idea 😅

[HugoGranstrom]

So if we want to calculate the velocities with respect to the Sun but we start from the Earth, then we could get into this problem. My main concern is how we calculate the shortest distance to the starting body if the orbit is hyperbolic with respect to it

[HugoGranstrom]

I'm not sure how we would solve it if we does find a way either 😅 I think we have one variable to play with (how many revolutions around the sun before we arrive)

[TheHappyKoala]

Hahahaha, ah, got ya!

The risk of this happening is only high enough to be bothered about when you are in a really tight orbit around the planet, so what we could implement is a simple brute force check, and if that check determines that there will be collision, you simply display a validation message to the user telling them that they have to try a trajectory with a different tof or depart from a different point in the orbit.

[HugoGranstrom]

Found a formula we perhaps could use here: https://en.m.wikipedia.org/wiki/Hyperbolic_trajectory#Impact_parameter_and_the_distance_of_closest_approach

[HugoGranstrom]

Yeah that seems like a sane way to handle it 😄👍😊 The user should be the one to handle it for us 😂

[TheHappyKoala]

The formula is probably the way to go, but I'll deal with that once the orbital insertion controls are done and dusted... Don't want to get distracted - happens very easily to me, I have the attention span of a 🐶!

On the subject of getting distracted... Look, our first procedurally generated planet! Nothing fancy, but soon enough their appearance will be determined by their temperature, mass and so on... Cool times ahead 😄

Have a mighty good start to February... Almost spring now 🌞 !

[TheHappyKoala]

The white are ocean basis / lava plains... Clearly I'm yet to figure out how to work with colors 🤣

[TheHappyKoala]

And we need procedural gas giants, neptunes and what not!

[TheHappyKoala]

Man, this is so cool I'm shitting ma pants!

[HugoGranstrom]

😮😮😮😮😮😮😮😮🤩🤩🤩🤩🤩🤩🤩🤩🤩🤩🤩🤩 THAT'S BLOODY BRILLIANT!!! 🌎🌍🌏🚀🚀🚀🚀🚀🚀 It looks so good! Is there algorithms for doing this or what's going on?

[TheHappyKoala]

It's gonna take the exoplanet scenarios to a whole other level :D But yes, there are! I'm using something called perlin noise subject to the constraints I've already mentioned; temperature, mass and so on.

Gonna make a big commit this weekend... If you like algorithms I think you'll enjoy reading through the code 😄

[HugoGranstrom]

Yes it really will take it to a new level 🚀😉👍

Gonna check it out, and YES you can bet I'm going to enjoy reading it 😁

Haven't really started with the Orbital Elements yet 😅 I've been prioritizing working through the exercises in my programming course, but now the next one is due at the end of February so I'll do my best not to procastinate i anymore 😅😂

[TheHappyKoala]

No rush! I work at a very slow pace, too. And I just realised I'm shit posting in your issue... Sorry 😇

[HugoGranstrom]

Nah, this is how we roll 😎 we post our shit wherever we like 😉

[HugoGranstrom]

Finally the puzzle pieces in my head has fallen into place! 🤣 I have found formulas for converting to and from orbital elements here and here. And on Wikipedia I found how to propagate them forward in time. So now it's just a matter of implementing them.

I'm thinking of creating a TS type called OrbitalElements so we don't have to specify all of them as inputs to our functions. What do you think about that? 😄

[HugoGranstrom]

Is there anything from your branches that could cause any headache-inducing merge conflicts or am I free to do whatever I want in lambert.ts? 😅

[TheHappyKoala]

Go for it, and yay 😄 !!! That will make the space flight stuff so much more interesting since you won't have to wait for the trajectory cruncher to generate the future position of a mass. Maybe we could add a date control, too, so that people could view the solar system at different points in time???

[TheHappyKoala]

And I really need to merge the gatsby branch soon... It's turning into a monster, haha

[HugoGranstrom]

Haha yeah it takes quite some time right now 😅🤣 The date control should work, but I don't know how accurate it would be for longer timescales 😅 My Earth is already off by 0.02 AU after one year. Don't know how the numbers look for our integrators but they might be off by the same 🤣 It would be a neat feature nonetheless 😉 It wouldn't be to hard to do once we can do it for one mass

[HugoGranstrom]

We have our monster of our own we could use to tame Gatsby: RKN12 🤣 🤣

[TheHappyKoala]

I should have left it at implementing Gatsby but instead I doubled down on the gitigans by baking other features like new textures, procedural stars etc into it... Will I ever learn 👽 ?

Yea, the farther into the future, or past for that matter, you try to ascertain the position of a planet, the less accurate it will be, especially for comets that are often perturbed by Jupiter. Would be kind of cool to create one scenario where you have a set of clones running on orbital elements and compare their positions to a set of planets whose positions are determined by an integrator and get a sense of how important perturbations are to our solar system.

Yes, we should wire reality up to RKN12 and make things happen 🤣 .

[HugoGranstrom]

Gitigans have a certain charm to them 😂 it's hard to resist sometimes. You have gotten loads of work done nonetheless 😉

Oh yeah, that would be super cool 🚀🚀😄😄😄😎 I think we would pretty quickly see the effects for certain bodies (mercury) and some would probably not change much (Neptune)

[HugoGranstrom]

Do you prefer to have the Orbital Elements in degrees or radians? For doing all the future telling it's more convenient to have them in radians all along. Do you have any opinions on the matter? 😄

[TheHappyKoala]

Nah, doesn't really matter. For whatever reason I have a natural preference for degrees (guess they feel more familliar!), but at the end of the day converting between the two is a piece of cake. Fact of the matter is that switching to radians might make more sense seeing as THREE.js uses radians and not degrees 🚀 !

[TheHappyKoala]

First commit for the procedural planets is on its way... Just the terrestial ones, but they are trickier to generate than the gas giants, so there's still a yay factor to this commit, I'd say 🤣

[HugoGranstrom]

Radians it is then! 😉 Felt better getting your opinion on this so I didn't suddenly change it to radians and all your code assuming degrees stopped working 🤣

Awesome! 🚀 Looking forward to populating a whole galaxy with procedural planets 🤣

[HugoGranstrom]

I have btw managed to get the stateToKepler, keplerToState and propagateOrbitalElements to work so now I just have to make functions for traversing a SOITree and doing all the propagation for all bodies.

I'm thinking of making two different functions:

• One we will use in the trajectory cruncher that only cares about the future position of one body. It will propagate the body and all it's parents forward to get the final position. Ex Titan: first spin Titan around Saturn and then Saturn around the Sun.
• The other function calculates the future positions of ALL bodies in the system so it does the job the integrators would have otherwise done.

Not sure how to handle the particles though... 🤔

Does these sound like reasonable cases? Or do you have other cases you would want implemented as well? 😄

[TheHappyKoala]

Ohhhh... Let me get back to you tomorrow with some feedback regarding the points you brought up... Right now I'm obsessively trying to wrap up the procedural planets... So close but there's this little issue pertaining to the interpolation of colors that is driving me nuuuuuuts 🤣

[HugoGranstrom]

No hurry 😉👍

Oh that is so annoying when there's just one small thing left and it's not cooperating 😔😦🦕

[TheHappyKoala]

LOL. Search terms that brought people to our website from google... One of those guys must have been VERY disappointed when he saw what our site was actually about 🤣 . I mean, wtf! Should I laugh or loose faith in humanity?

[HugoGranstrom]

Hahahahaha, what the #$_* 🤣🤣🤣🤣🤣🤣🤣🤣🤣🤣 An idea for the future perhaps?

[TheHappyKoala]

X-Rated HOS... 🤣 No, somebody will have to fork HOS and create it themselves

[TheHappyKoala]

I would love it if HOS was the basis of the world's first, ehrm, 3D space sex simulator

[HugoGranstrom]

🤣🤣 I would be hugely impressed if someone managed that feat

[TheHappyKoala]

Totally. How you doing? Any extravagant plans for the weekend?

[HugoGranstrom]

I'm doing fine, school is really chill right now and life is rolling on 😄 the weather does seem too nice in the weekend so hopefully I will get some coding done. How are you and what's your plan? 😊

[TheHappyKoala]

I'm doing pretty good. Work's still very challening and a lot of fun, and on my way home today I bought two beers at bolaget, so tonight I'm gonna drink beer and work on the procedural planets 🤣 I'm such a bad boy.

[HugoGranstrom]

Sounds great!😄👍 Haha maybe a more "open" mind can be helpful there 😅🤣

[TheHappyKoala]

[TheHappyKoala]

Creating world's is so much fun!!! This is a cool pic because it is the first time I have generated a planet based on its temperature, mass and radius

[HugoGranstrom]

Wow! 😮🤩🤩🤩 That looks so realistic but still a bit random 😉 what's thr difference between green and brown?

[TheHappyKoala]

Brown is supposed to represent mountains... I need a more realistic model for how vegetation changes with elevation. Habitable planets are tricky... World's like Mars and Mercury are more gratifying 🤣. Also, I need to take the longitude of a pixel into account, so that poles are covered by ice caps, whereas somewhat above and below the equator you get desserts.

[HugoGranstrom]

Ahhh I get it 😄👍 Sounds like a lot of work 😅

[TheHappyKoala]

Hms, no idea why I hadn't thought of this before, but by passing a uniform to the shader and calculating the temperature (black body, taking greenhouse gases into account is beyond ma paygrade) of Earth based on its distance from the Sun, I can change the look of Earth... So take it out into the asteroid belt and it freezes over, and inwards towards Mercury and it becomes a dessert world. That'll have to wait for later, though... Need to make small and manageable commits 😅

[TheHappyKoala]

OMG, OMG, OMG!!!

You know how to make sure that some property on an object is not undefined or null, you have to do a check like this:

data !== null && data.person && data.person.name && doSomething(data.person.name);

Well... With the newest version of JavaScript, you can do the same check like so:

data?.person?.name? doSomething(data.person.name);

🎈 💯 🏅 🎉

[HugoGranstrom]

Wow, that's such a brilliant idea in its simplicity 😉😊😎 Seeing how Earth changes depending on the distance to the Sun

Didn't actually know you had to go through all that trouble tbh 🤣😂🤣😂 but the new way is much neater and shorter, guess it saves you a lot of time?😊

[TheHappyKoala]

Doesn't save you that much time, it's more the experience of reading code that improves... It's the difference between caligraphy and writing like this "hOw u dUiN?!!" with bright red marker 🤣

[TheHappyKoala]

Maaannn.... So close now, even figured out how to make procedural gas giants!!! Gonna tweak the model that determines the type of planet based on its characteristics, and then I'm gonna put these guys to the side and merge this branch, and make HOS great again 🤣