how effective are gravity assists anyways?

[jevry]

im doing a project about spaceflight and orbital mechanics and i've figured some stuff about the math behind it now
now i've tried using my cranky old pc to (barely) run RSS and assuming going to jupiter takes about 7600 dV how much would it cost if i tried to slingshot as much as i could and go via earth-venus-earth-earth-jupiter
i though il just use transfer planner to plan a rough route and see what happens 
sadly i burned up in the earths atmosphere during a gravity assist because i was unsure about the height of the atmosphere and foudn out that indeed yes, it was above 70 km
but i noticed that the dV expended i wrote down quickly added up to above 7600 m/s. you already need 3000 m/s to get to venus so that means for the entire journey after that has to be under ~4500 or its a waste of time and dV
i guess if i planned more strictly i could have avoided an unnessesary second 3000 m/s burn but even then, 4500 m/s max saved seems so harsh. i mean its great and all since thats a potential 25% fuel saved
but around 1500 m/s is surely going to interplanetary course corrections. and unless really well planned that other 3000 m/s is gone sooner than you think.

then i checked the internet and according to a deltaV map of the solar system the difference in dV between venus and jupiter is even narrower, less than 3000
so an already tight planning has to be even tighter

so how usefull is it even planning for them? is it usefull to do a single assist with venus? or do you need to go full on galileo probe style and hit 3 assists to get meaningfull milage out of it?

 

and on another note why is there no transfer window calculator that can calculate a route from planet A to C via slingshot planet B? it should be possible right?

 

Edited February 1, 2020 by jevry

[UmbralRaptor]

In KSP, rockets are easy while compute is limited. In real life, compute is plentiful and rockets are hard. So going through a bunch of simulations to save a few hundred to a few thousand m/s actually gets you significantly more payload for little effort. There's also an upper limit to how much velocity you can gain from a single slingshot (it decreases with faster and more distant approaches to a planet), which pushes towards a handful of bodies and multiple passes.

 

FWIW, My back of the envelope calculations are that from a 200 km Earth orbit, you need 5375 m/s for a Hohmann transfer to Venus, and 7236 m/s for a Hohmann transfer to Jupiter. But just getting to orbit tends to cost 8-9 km/s...

 

Edited February 2, 2020 by UmbralRaptor
km/s, not m/s

[The Aziz]

15 hours ago, jevry said:

 why is there no transfer window calculator that can calculate a route from planet A to C via slingshot planet B? it should be possible right?

 

Actually, there is.

 

Takes a while to get it right, but gives a lot of info about gravity assists you need to get to save fuel.

I used it only once in OPM, on a Kerbin-Jool-Sarnus trip, and the assist I got from Jool extended the flight time over 2 years but saved me a lot of fuel indeed, probably because I didn't have to brake so hard when I reached my destination.

[Incarnation of Chaos]

12 hours ago, UmbralRaptor said:

In KSP, rockets are easy while compute is limited. In real life, compute is plentiful and rockets are hard. So going through a bunch of simulations to save a few hundred to a few thousand m/s actually gets you significantly more payload for little effort. There's also an upper limit to how much velocity you can gain from a single slingshot (it decreases with faster and more distant approaches to a planet), which pushes towards a handful of bodies and multiple passes.

 

FWIW, My back of the envelope calculations are that from a 200 km Earth orbit, you need 5375 m/s for a Hohmann transfer to Venus, and 7236 m/s for a Hohmann transfer to Jupiter. But just getting to orbit tends to cost 8-9 km/s...

 

Adding on to this most space probes don't have the DV needed to actually perform their transfer burns and the insertion burn at the intended destination, so gravity assists are used to get them to their destinations in the first place. Also once there they need fuel for manuevers, station keeping and momentum dumping; which quickly adds up.

So with gravity assists you end up being able to reduce the total fuel mass of the probe; which also makes the most of what you have onboard. While also allowing you to use more energy from the launch vehicle/second stage to get it where it needs to be in the first place. If you wanted to go directly to the destination it would require much more fuel mass on the launch vehicle and the probe, but would also be much faster.

So it's all about making the most with our current technologies, and the most of whatever budget the mission has. If SpaceX ever gets Starship off the ground i'm hoping that we'll see some MONSTER probes to the outer planets, and it would also get them there within my lifetime as a bonus!

 

[Fierce Wolf]

Gravity assists are a must! You can save hundreds of Dv adjusting your path wisely. I once braked around Jool using 10 dv in midflight, parking me near my destination.

[jevry]

On 2/1/2020 at 9:56 PM, UmbralRaptor said:

 

FWIW, My back of the envelope calculations are that from a 200 km Earth orbit, you need 5375 m/s for a Hohmann transfer to Venus, and 7236 m/s for a Hohmann transfer to Jupiter. But just getting to orbit tends to cost 8-9 km/s...

huh 5375 m/s? i calculated that from an orbit equal to that of earth to a transfer to venus would cost 2116 m/s (admittedly i already figured somethign has to be wrong since escape velocity is above 2116 m/s), any idea what i missed in the calculation? (for refference i litterally calculated the speed of an orbit, that has an apoapsis of earth's periapsis and a periapsis of venus's apoapsis, at the apoapsis, and subtracted the speed of earth's orbit at periapsis)

 

On 2/2/2020 at 9:53 AM, The Aziz said:

Actually, there is.

 

Takes a while to get it right, but gives a lot of info about gravity assists you need to get to save fuel.

I used it only once in OPM, on a Kerbin-Jool-Sarnus trip, and the assist I got from Jool extended the flight time over 2 years but saved me a lot of fuel indeed, probably because I didn't have to brake so hard when I reached my destination.

im defenitely going to try and see if i can make this work with RSS. thanks!

[Starman4308]

There's definitely something up with @UmbralRaptor's calculations; in RO, I tend to get 3.5-3.8 km/s ejections to Mars/Venus, and 6.7 km/s for Jupiter (both starting from LEO in the plane of the ecliptic).

As to gravity assists in general: there are some that are pretty easy to set up even without KSP-TOT or a lot of patience.

KKJ (and its EEJ analog in RSS): Eject into a 2-year heliocentric orbit 2 years prior to a Jool transfer window. Make a moderate retrograde burn at apohelion to set up a Kerbin slingshot to finish sending you to Jool.

Moonbrakes: Both Ike and Tylo are pretty good for braking into eccentric orbits of their planets. All you need to do here is make a mid-course adjustment burn relatively early during your transfer, target the relevant moon, and tweak a balance of radial/prograde until you meet the moon in the right place at the right time.

Satellite ping-pong: Once in orbit of a planet with a complex system of satellites: if you have an eccentric orbit in the same plane as said moons, you can make small burns at apoapsis to set up flyby after flyby after flyby. It's brilliant in RSS for visiting all the moons of Jupiter/Saturn/Uranus/Neptune.

[Arrowstar]

On 2/3/2020 at 4:45 AM, jevry said:

im defenitely going to try and see if i can make this work with RSS. thanks!

Author of KSPTOT here: Please let me know if you have any questions by posting on the KSPTOT thread!

[jevry]

On 2/6/2020 at 2:27 PM, Starman4308 said:

There's definitely something up with @UmbralRaptor's calculations; in RO, I tend to get 3.5-3.8 km/s ejections to Mars/Venus, and 6.7 km/s for Jupiter (both starting from LEO in the plane of the ecliptic).

As to gravity assists in general: there are some that are pretty easy to set up even without KSP-TOT or a lot of patience.

KKJ (and its EEJ analog in RSS): Eject into a 2-year heliocentric orbit 2 years prior to a Jool transfer window. Make a moderate retrograde burn at apohelion to set up a Kerbin slingshot to finish sending you to Jool.

Moonbrakes: Both Ike and Tylo are pretty good for braking into eccentric orbits of their planets. All you need to do here is make a mid-course adjustment burn relatively early during your transfer, target the relevant moon, and tweak a balance of radial/prograde until you meet the moon in the right place at the right time.

Satellite ping-pong: Once in orbit of a planet with a complex system of satellites: if you have an eccentric orbit in the same plane as said moons, you can make small burns at apoapsis to set up flyby after flyby after flyby. It's brilliant in RSS for visiting all the moons of Jupiter/Saturn/Uranus/Neptune.

i actually tried this because i got a little intimidated by ksptot (appearently you need to adjust settings in the files to get the mod working with RSS.)
it went pretty well, iirc i managed to save about 1300-1400 dV by doing the 2 year earth maneuvre. it did take 30 minutes of fine tuning though.

i am kind of sad i didn't manage to save more dV because 1400 out of a total of 20K or so is still not a lot
and i also noticed something cheeky, by doing such a maneuvre you may have to burn more fuel during the capture burn, is this correct?

not sure if this is a necro....

[Regor]

22 hours ago, jevry said:

i actually tried this because i got a little intimidated by ksptot (appearently you need to adjust settings in the files to get the mod working with RSS.)
it went pretty well, iirc i managed to save about 1300-1400 dV by doing the 2 year earth maneuvre. it did take 30 minutes of fine tuning though.

i am kind of sad i didn't manage to save more dV because 1400 out of a total of 20K or so is still not a lot
and i also noticed something cheeky, by doing such a maneuvre you may have to burn more fuel during the capture burn, is this correct?

not sure if this is a necro....

Are you kidding?! 1400m/s is so good! Also it depends on your trajectory for the capture burn. As a rule of thumb, the closer your orbit is to your target, the less the capture burn will be.