In orbit you have to slow down to speed up
55 points
7 days ago
| 13 comments
| wired.com
| HN
austin-cheney
8 minutes ago
[-]
I have always wondered about this with regard to Newton's Third law:

    For every action there is an equal and opposite reaction.
So if a craft departs from rendezvous with another craft it must do so by pushing away from that other craft. That means it is equally pushing on both crafts. If the rendezvous was in orbit does that mean departing from rendezvous pushes both crafts out of orbit? If so does the other craft have to correct for this to reestablish orbit or is orbit self-correcting as if in a third body scenario?

I ask because Earth is an third body scenario between the sun and Jupiter. Jupiter has enough gravity to occasionally pull the Earth slightly (not significantly) out of orbit from the Sun, but Earth's orbit to the Sun is self-correcting due to the difference in mass between the Sun and Jupiter. Quick web searching reveals Jupiter's pull on Earth is only approximately 0.005% of the Sun's after accounting for both mass and distance, but that number rises to 0.011% after accounting for syzygy with the moon.

reply
lproven
4 hours ago
[-]
As Larry Niven wrote in The Smoke Ring and The Integral Trees...

"West takes you In, In takes you East, East takes you Out, Out takes you West, North and South bring you back again."

https://forum.nasaspaceflight.com/index.php?topic=3341.0

reply
ChicagoBoy11
7 hours ago
[-]
For anyone who is remotely interested in this, a considerable chunk of the Gemini program was all about solving some of the practicalities involved with Rendezvous, and it is quite interesting even hearing some of the astronauts come to grips with some of the physics while orbiting in space trying the various types of rendezvous and docking maneuvers that were attempted.
reply
ubj
1 hour ago
[-]
Buzz Aldrin (who was the second person to walk on the moon) wrote his MIT thesis on methods for astronauts to handle the complexities of orbital dynamics when performing rendezvous maneuvers in orbit:

https://dspace.mit.edu/handle/1721.1/12652

reply
ralfd
6 hours ago
[-]
I have a hard time imagining physics. For example take a train moving 100 kmh to the north which wants to reverse direction to the south. It has to break and then accelerate again, a very costly operation. Except when the tracks make a turn? But how can a northward momentum change to a southward momentum?

The same confusion I have when trying to imagine satellites going around Earth or slingshot maneuvers. Would an X-Wing turn in space differently than in the atmosphere of Hoth? Would it in space just rotate, but keep its forward (now backwards) momentum instead of turning like a fighter jet?

reply
NitpickLawyer
5 hours ago
[-]
> The same confusion I have when trying to imagine satellites going around Earth or slingshot maneuvers.

I can't recommend KSP enough. It's a "silly" game with "on rails physics" (so not exactly 100% accurate wrt general relativity stuff) but it's got a very nice interface and it will make you "get" orbital mechanics by dragging stuff around. You'll get an intuition for it after a few hours of gameplay / yt video tutorials. Really cool game.

reply
montagg
5 hours ago
[-]
This is how I now “get” orbital mechanics better than I ever did trying to study it. Play is the best education.
reply
ErroneousBosh
4 hours ago
[-]
> For example take a train moving 100 kmh to the north which wants to reverse direction to the south. It has to break and then accelerate again, a very costly operation. Except when the tracks make a turn? But how can a northward momentum change to a southward momentum?

Your train is decelerating, and then accelerating southwards. It really is.

If you were on a train that was travelling in a straight line northwards and the driver applied the brakes, it would decelerate, which really is acceleration with a negative value (and I can hear that in my old high school physics teacher's voice, hope you're doing well, Mr Siwek). You would feel yourself being thrown forwards if the acceleration was strong enough because your momentum wants to keep you moving north.

If you were on a train that was travelling around a U-shaped bit of track looping from northbound to southbound, then you'd be thrown towards the outside of the curve. Guess what? The train is not moving north so fast, and your momentum is trying to keep you moving north.

The difference here is that if you brake the train to a stop and throw it in reverse then you're dissipating energy as heat to stop it, and then applying more energy from the drivetrain to get it moving again, but if you go round a U-shaped track the energy going north is now energy going east. You have not added or removed energy, just pointed it a different direction.

reply
GeneralMayhem
4 hours ago
[-]
Turning around a track definitely dissipates some heat energy through increased friction with the rails. Imagine taking a semicircle turn and making it tighter and tighter. At the limit, the train is basically hitting a solid wall and rebounding in the other direction, which would certainly transfer some energy.

The energy question is this: going from a 100kmh-due-north momentum to a 100kmh-due-south momentum via slowing, stopping, and accelerating again clearly takes energy. You can also switch the momentum vector by driving in a semicircle. Turning around a semicircle takes some energy, but how much - and where does it come from? Does it depend on how tight the circle is - or does that just spread it out over a wider time/distance? If you had an electric train with zero loss from battery to wheels, and you needed to get it from going north to going south, what would be the most efficient way to do it?

reply
floxy
1 hour ago
[-]
There is no "required" energy to change direction, even for a zero-radius change, think of a bouncing ball:

https://www.youtube.com/watch?v=QpuCtzdvix4

reply
10000truths
34 minutes ago
[-]
This only applies in perfectly elastic systems, where the bodies can convert kinetic energy to potential energy and back with perfect restitution. Which, thanks to the second law of thermodynamics, doesn't exist in reality. It's only a question of how much energy is lost. (Unless, of course, you include the medium into which the energy dissipates as heat into the system itself. But such a model is not useful in almost all practical scenarios.)
reply
altcognito
41 minutes ago
[-]
A bouncing ball is elastic. There is some loss in the process of storing the energy from the movement into the ball and then releasing it into the opposite direction. Good example though!
reply
ErroneousBosh
40 minutes ago
[-]
> Turning around a track definitely dissipates some heat energy through increased friction with the rails.

No it doesn't, but we're talking about identical spherical frictionless trains in a vacuum.

reply
btilly
5 hours ago
[-]
What's going on here is that your momentum changes whenever you experience a force. Your energy changes whenever you experience a force towards or from the direction that you are traveling.

The force from the rails at all points is at right angles to the direction of motion. So your energy doesn't change. Your momentum is constantly changing. And you're doing it by shoving the Earth the other way. But the Earth is big enough that nobody notices.

Now to the orbital example. In the Newtonian approximation, an orbit works similarly. In a circular orbit, you're exchanging momentum with the planet, but your energy remains the same. The closer the orbit, the more speed you need to maintain this against a stronger gravity, and the faster you have to move.

In an elliptical orbit, you're constantly exchanging momentum with the planet, but now you're also exchanging between gravitational potential energy, and kinetic energy. You speed up as you fall in, and slow down as you move out. Which means that you are moving below orbital speed at the far end of your orbit, and above when you are close.

Now to this paradox. Slowing down causes you to shift which elliptical orbit you are in, to one which is overall faster. Therefore slowing down puts you ahead in half an orbit, and then you'll never stop being ahead.

reply
ambicapter
5 hours ago
[-]
In your train example, the rails exert a force on the train as it turns. In orbit, the planets are constantly exerting a force on the satellite.
reply
boothby
3 hours ago
[-]
What's happening is that you exchange forward momentum for angular momentum. When the track straightens out again, you trade the angular momentum for forward momentum again. The train pays for this in friction losses; the orbital maneuver costs some fuel for steering.
reply
Xerox9213
5 hours ago
[-]
When you brake you generate a ton of heat.

Doing a U-turn generates less heat, but still quite a bit. The train will have to slow down depending on the radius of the curve, and even then the turn will slow it down some more.

But yeah, less heat generation means kinetic energy is conserved.

Cars have to slow down when they turn because it’s too much to ask of the tires to accelerate (throttle) and turn, since turning is in itself acceleration.

reply
NetMageSCW
5 hours ago
[-]
Caveat: when the tires are already at the limit of adhesion (e.g. on an F1 car). In a road car, you are not normally turning at 1g and probably can’t accelerate at 1g so you can turn and accelerate when you have enough margin.

It’s just the average driver doesn’t realize how much margin is available.

reply
wongarsu
5 hours ago
[-]
A train has momentum in the direction of the track. If the track makes a 180° turn the train will lose some momentum to increased friction with the track during the turn, but essentially the momentum still follows the track.

A fighter jet (or X-Wing in orbit) kind of generates its own "track" with the guiding forces of the wings. You can still do a 180° turn and keep a significant part of your momentum. Though the guiding effects are a lot softer, so your losses are a lot worse

A satellite (or an X-Wing in orbit) has no rails that can go in arbitrary directions. Any momentum is in "orbit direction", but orbits work in weirder ways. If you make your orbit highly elliptical then at the highest point you will have traded nearly all your kinetic energy for potential energy and can make a 180° turn pretty cheaply (because it's only a small change in speed)

reply
kbelder
3 hours ago
[-]
A very related physics issue that boggles my mind is when you roll a disk, like a wheel. You can roll the disk north, and it'll lean, curve, and end up going south. What force changed the direction of the wheel?

I understand it, intellectually. It's pushing sideways against the surface as it leans and spins, but it just doesn't feel right. I have no intuition for it.

reply
max51
8 minutes ago
[-]
If you are talking about the gyroscopic precession effect that happens when you push on a spinning disc, this is the best video I've seen so far that explains it in an intuitive way: youtube.com/watch?v=n5bKzBZ7XuM
reply
lucianbr
4 hours ago
[-]
You too can change direction easier if there is an object (like a pole or something) you can push/pull against. Try it, maybe it will help your intuition.

Run towards a pole and then try to come back around it, once without touching it and once using it to swing around. That's the role the curved tracks play. You exchange momentum with the object, and in the end with the Earth.

reply
aaronblohowiak
6 hours ago
[-]
play KSP, it will click after a few days.
reply
exe34
5 hours ago
[-]
I feel like none of the answers have addressed you train example correctly. The momentum is exchanged with the Earth. So the Earth+train still have the same total momentum. The energy is mostly conserved (ignoring the friction that's needed to stay on the track). You can do the same by running past a lamp post and extending a hand to grab it - you'll change direction.
reply
stoneforger
7 hours ago
[-]
They should really teach physics using KSP.
reply
vannucci
6 hours ago
[-]
I tried to teach a group of HS students about orbital mechanics as a high school physics teacher using KSP. It was... difficult. Not impossible. But I agree it's an excellent learning tool.
reply
hobs
6 hours ago
[-]
Right, the UI/UX is a lot to just get to the rocket part. KSP is probably the best game that forces that into your head with a classic simulation that's fun, but I gotta say something like Rocket League was better at building my intuition for rocket behaviors.
reply
matheusmoreira
6 hours ago
[-]
Yeah, it's amazing. With enough docking and maneuvering practice I developed some kind of intuition for moving in space. I could maneuver without meticulously planning the burns.

Still can't leave Eve though...

reply
dabluecaboose
6 hours ago
[-]
I'm a professional astrodynamicist and I owe my base level understanding of orbital mechanics to KSP. It's a fantastic resource for learning the basics of Keplerian motion.

Also, obligatory XKCD: https://xkcd.com/1356/

reply
taneq
5 hours ago
[-]
Arguably aerodynamics is confusing on a whole other level to mere orbital dynamics. :D
reply
dabluecaboose
4 hours ago
[-]
I washed my hands of aerodynamics after I got my first job in satellite navigation. Messy stuff, that Navier-Stokes business
reply
M95D
5 hours ago
[-]
No, it's simple. Just make sure the airplane falls nose-first if it ever stops (speed<stall).
reply
aaronblohowiak
5 hours ago
[-]
I wish ksp 2 hadn't been a boondoggle
reply
mikkupikku
5 hours ago
[-]
I haven't kept up with it, but hopefully Kitten Space Agency will be able to take up the torch.
reply
delichon
7 hours ago
[-]
I doubt SpaceX could put a satellite in orbit with KSP physics. Just the absence of realistic thermal conduction would prevent it. The outer skin temperature typically peaks around 300–600 °C during the densest part of the atmosphere. If you calculate those forces wrong the rocket has a bad day. Best case it is over engineered and has a reduced payload. They might as well do their calculations with pi equal to 3.
reply
0_____0
6 hours ago
[-]
What does thermal conduction affect? Is it mostly practical spacecraft construction, or actually related to orbital mechanics?
reply
bogzz
6 hours ago
[-]
The FAR mod is touted as being realistic; I haven't played it though.
reply
iso1631
5 hours ago
[-]
https://xkcd.com/2205/ comes to mind with your pi approximation.

Nobody is saying KSP physics is perfect.

Until I played KSP, I had no idea how hard orbit was compared with just going up into space (and generally the greater population thinks the same -- they think that sending New Shephard upto 100km is about the same as sending a Dragon into orbit). I had no idea how you move in orbit, how getting from low earth equitorial orbit to Jupiter takes less energy than getting from the same ship to a polar orbit (and even then that the only real way to change your orbit like that is to go out beyond the moon and back), etc.

reply
yabones
6 hours ago
[-]
Forwards is up, up is back, back is down, down is forwards.
reply
nomel
12 minutes ago
[-]
It's much easier to reason about when you frame it closer to reality: you're not on a circular path, you're continuously falling, and because you're moving forward, you're continuously missing the earth, with its pull decreasing with distance.
reply
jayknight
1 hour ago
[-]
How related to this is the helicopter 90-degree phase lag thing?
reply
geon
5 hours ago
[-]
Or in the words of Larry Niven (The integral trees)

East takes you Out

Out takes you West

West takes you In

In takes you East

reply
NitpickLawyer
5 hours ago
[-]
Down is where the enemy gate is.
reply
taneq
5 hours ago
[-]
So precise, he piss on a plate and never splash.
reply
Scubabear68
4 hours ago
[-]
This brings back fond memories of Heinlein's juvenile sci fi series.
reply
brudgers
5 days ago
[-]
A similar thing is true when cornering a race car when measuring time through the corner.
reply
taneq
5 hours ago
[-]
Slow is smooth, smooth is fast.
reply
dotancohen
7 hours ago
[-]
How so?
reply
anonymars
7 hours ago
[-]
"Faster" (higher speed) = wider cornering radius = more distance = slower
reply
wrigby
6 hours ago
[-]
But you exit going faster, which means you make up time on the straight after the corner.
reply
rtkwe
6 hours ago
[-]
It's a balance, many of these cars can accelerate and decelerate very hard so the time to get back to the full speed for the next section is fairly short reducing the effect of slowing down. The effect of taking a too wide racing line though means a large multiple in the distance travelled.
reply
HPsquared
5 hours ago
[-]
Cars can usually brake and turn harder than they can accelerate.

You also tend to spend more time on the straight after the corner, than in the corner itself

So you mostly optimise for corner exit speed, especially if the car has particularly slow acceleration and a long straight comes after the corner.

reply
drivebyhooting
35 minutes ago
[-]
For F1 I was under the impression exit speed wasn’t as important as minimizing arc length of the turn.
reply
oarsinsync
6 hours ago
[-]
Assuming there is a long enough straight before the next corner
reply
taneq
5 hours ago
[-]
Yeah depends on the corner but the general thumb-suck approximation is sound.
reply
rascul
6 hours ago
[-]
If there's banking, it can change things.
reply
everyone
6 hours ago
[-]
The distance has no effect.. Its all about speed, you want to take the line that lets you get through the corner while maintaining the highest speed. If you are going faster and spend as little time as possible breaking and accelerating you will gain time. Also a higher exit speed means you will be going faster for the entire straight after the corner making a very big time difference.

Your car, depending on how much grip it has + other variables, will have a theoretical minimum diameter circle it can drive around at various speeds. The higher the speed the bigger the circle. Finding your racing line is just a matter of fitting the biggest circular arc inside the space available in the corner.

Ideally you want to break in a straight line before the corner and reach the speed your car can drive the circle at at just the moment you enter it.

Theres more nuance when it comes to compound corners, FR vs FF cars, oversteer understeer, hills bumps etc. But the basic theory is simply fitting circles.

https://ibb.co/VY11TpTM

reply
bluGill
6 hours ago
[-]
You are generally not taking a perfect circle corner. You can/should be slowing down as you enter the corner and then speeding up even before you exit. In this way you can shorten the distance traveled while getting a higher exit speed - sometimes higher than the largest possible circle corner. Optimizing this for the car/track/conditions is what makes for a great driver.
reply
everyone
5 hours ago
[-]
The distance is irrelevant.. It is true that depending on the car you may gain time breaking and accelerating while turning.

But that is a more subtle and advanced concept though (like dealing with elevation changes).. People should understand the big circle first.

reply
bluGill
5 hours ago
[-]
In the context of winning a race you need to get the subtle and advanced concepts right or you will be in last place. If you are just driving on the street it doesn't matter.
reply
everyone
5 hours ago
[-]
Most times Ive seen anyone playing a racing game they seem to be totally clueless.. They dont even comprehend the big circle. They always go into corners way too fast, break super hard and then crawl out.

Its so common it surprises me racing games have always been so popular.

What I have also noticed is that over time racing games have changed their physics to be totally wacky in order to meet the general public's wacky expectations.. (eg. mario kart or GTA5) I cant play those games cus the physics are so strange.

reply
bluGill
4 hours ago
[-]
I was referring to real world races where we cannot ignore physics.

Racing games are very different. They tend to have adaptive AI - you are more likely to win with the naive approach you describe than the physically perfect route. The physically perfect result will get your through the race several minutes faster, but the AI opponents become impossible to beat. Thus the ideal path is the worst thing you can learn. (I haven't played games in years, but IIRC the games you mention don't pretend to be about racing, I wonder how ones that pretend to be a real race compare)

reply
everyone
3 hours ago
[-]
?? I guess you havent played modern racing games. No-one races against AI, its all against other people. Games like Assetto Corsa and iRacing have very good physics models. Real race drivers use them to train and are often seen online.

The circle thing is aimed at most people here. If your average person implemented that they would dramatically improve their times.. All the other stuff (of which of course there is a lot) would result in relatively marginal improvements.

reply
HPsquared
5 hours ago
[-]
Life and business are often the same.
reply
everyone
6 hours ago
[-]
I thought I understood all this until I played Kerbal Space Program.
reply
elephanlemon
6 hours ago
[-]
reply
QuiCasseRien
5 hours ago
[-]
could it be possible to flag a thread when you need to pay or register to read an article ??

very annoyoing, the subject looks good, open tab and rohhhhhhhh... paid or register.

reply
ceejayoz
5 hours ago
[-]
reply
f4uCL9dNSnQm
4 hours ago
[-]
There is the small, tiny issue of people commenting just based on title, without even reading the article.

In this case I expected it just links to https://www.youtube.com/watch?v=bcvnfQlz1x4 and didn't even notice in links to Wired.

reply
embedding-shape
5 hours ago
[-]
Allowing paywalls vs not been discussed for a long time. Latest comment from dang about it seems to be this:

> The answer is that paywalls are allowed when there are workarounds (such as archive links) which allow ordinary readers to read the article without paying or subscribing, while hardwalled domains (i.e., without such workarounds) are banned. - https://news.ycombinator.com/item?id=43876575

https://hn.algolia.com/?dateRange=all&page=0&prefix=true&que...

reply
pfdietz
7 hours ago
[-]
This is a consequence of the virial theorem of mechanics.

https://en.wikipedia.org/wiki/Virial_theorem

This theorem also lets you conclude that as a nondegenerate star becomes more tightly bound (smaller, for a given mass) it must also become hotter.

(Why did someone downvote this?)

reply