delltaV and maneuvers of space missions

[king of nowhere]

ever since I started playing this game and I got some skills at navigation, I started wondering about exact trajectories of actual space missions.

How much remaining deltaV did the space shuttle have once in orbit? how do they deal with inclination on a mars mission? did the cassini mission use a gravity assist from titan for capture around saturn? and if not, why not? how much did it cost for the rosetta mission to rendez-vous with its target? do they use an approach more efficient than mine to approach mercury?

and all those things are never, ever discussed anywhere. too technical, i guess. but even sites where they do give quite technical details on the machinery and instruments are not technical on the issue of maneuvers. if you're lucky you can see a trajectory, but you never see any kind of discussion like "here we made a burn 450 m/s prograde, 30 m/s antinormal".

anyone knows if/where I can find sources on the topic?

[KSK]

NASA probably have what you’re looking for somewhere - for the US missions at any rate.

It’s not precisely what you asked for but the Apollo flight journals get pretty technical. They’re based on actual communications between the crews and Mission Control, so they include burn details and suchlike.

https://history.nasa.gov/afj/

And yeah - I don’t play KSP these days but one of the best things about having played it, is that I can read something like the Apollo flight journals and have a good visual feel for what’s going on, even if the maths behind it all is rather out of reach!

 

Edited October 27, 2022 by KSK

[Beamer]

This is a pretty good document to start with. It has tons of info on fuel budgets, thrust vectors, burn duration and acceleration, etc. Even fun stuff like how much money it cost Note that most of it is in ft/s & ft/s^2, a sign of its age I guess: Apollo By The Numbers (pdf document). The more interesting stuff starts at table 23 or thereabout.

 

[RCgothic]

This is how much DV it generally takes to go places in the solar system. 

[sevenperforce]

5 hours ago, king of nowhere said:

ever since I started playing this game and I got some skills at navigation, I started wondering about exact trajectories of actual space missions.

anyone knows if/where I can find sources on the topic?

All of this info can be found online if you know what to search for, but I can also answer these questions (mostly) off the top of my head.

5 hours ago, king of nowhere said:

How much remaining deltaV did the space shuttle have once in orbit?

The Shuttle's OMS system contained about 300 m/s of total dV depending on the amount of payload it carried, but the overall launch profile dropped it off in a slightly-suborbital trajectory in order to allow the external tank to re-enter passively. It used about a third of this to raise its orbit after jettisoning the external tank and about third of this to lower its orbit back into the atmosphere at the end of the mission. The remaining third was used for on-orbit maneuvering and for attitude control during the first phases of re-entry. 

5 hours ago, king of nowhere said:

how do they deal with inclination on a mars mission?

The inclination of Mars is less than 2° off from the inclination of Earth. The two orbital planes intersect along a line, and any transfer trajectory between the two bodies will cross this line at one point or node. So typically in a Mars mission, the vehicle perform a very small mid-course correction at whatever place along the transfer orbit crosses that line. It's not the most efficient way to change inclination, but since it is such a small correction it's not that big of a deal and is typically performed with the basic RCS system on the vehicle.

Every now and then, the orbital planes and the optimal transfer trajectory will line up such that the node is either at Mars or at Earth. If it's at Mars, then you can just use your atmospheric capture and you never need to correct; if it's at Earth, you perform that correction along with your transfer burn to take advantage of Uncle Oberth. 

5 hours ago, king of nowhere said:

did the cassini mission use a gravity assist from titan for capture around saturn? and if not, why not?

No, it didn't -- Saturn's gravity well is extremely deep and so performing a capture burn as close as possible to Saturn's surface was the more efficient option. Titan is so small compared to Saturn that a gravity assist off of Titan would not have been sufficient to provide a capture.

KSP is extremely scaled down. Jool is only about 100 times more massive than Tylo, while in real life Saturn is more than 4,200 times more massive than Titan.

5 hours ago, king of nowhere said:

how much did it cost for the rosetta mission to rendez-vous with its target?

It took about 780 m/s of dV, spread over 10 different burns as it got closer and closer.

5 hours ago, king of nowhere said:

do they use an approach more efficient than mine to approach mercury?

I don't know what your approach is to approach Mercury, but all missions to Mercury -- Mariner 10, MESSENGER, and the ongoing BepiColombo -- have used at least one planet flyby to reduce the propellant needed to reach the innermost planet. Mariner 10 used a single Venus flyby but it never entered Mercurian orbit, instead doing all of its observation during three Mercury flybys. MESSENGER used one Earth flyby, two Venus flybys, and three Mercury flybys before finally performing its Mercurian orbital insertion burn. BepiColombo has already done one Earth flyby, two Venus flybys, and two Mercury flybys, and will do four more Mercury flybys before finally entering a polar orbit around Mercury with minimal propellant use.

[king of nowhere]

52 minutes ago, sevenperforce said:

All of this info can be found online if you know what to search for, but I can also answer these questions (mostly) off the top of my head.

well, clearly I don't know what to look for.

Quote

The Shuttle's OMS system contained about 300 m/s of total dV depending on the amount of payload it carried, but the overall launch profile dropped it off in a slightly-suborbital trajectory in order to allow the external tank to re-enter passively. It used about a third of this to raise its orbit after jettisoning the external tank and about third of this to lower its orbit back into the atmosphere at the end of the mission. The remaining third was used for on-orbit maneuvering and for attitude control during the first phases of re-entry. 

My doubt about that is that the shuttle put hubble into orbit, and hubble is a lot higher than the iss. 300 m/s are not enough to raise its orbit to 500x500, especially not if you also have to pay for reentry.

unless that the external tank had enough of an extra fuel to support the shuttle entering a 500x50 orbit before being dropped? so it means in regular missions it was dropped with some fuel still in it?

Quote

The inclination of Mars is less than 2° off from the inclination of Earth. The two orbital planes intersect along a line, and any transfer trajectory between the two bodies will cross this line at one point or node. So typically in a Mars mission, the vehicle perform a very small mid-course correction at whatever place along the transfer orbit crosses that line. It's not the most efficient way to change inclination, but since it is such a small correction it's not that big of a deal and is typically performed with the basic RCS system on the vehicle.

Every now and then, the orbital planes and the optimal transfer trajectory will line up such that the node is either at Mars or at Earth. If it's at Mars, then you can just use your atmospheric capture and you never need to correct; if it's at Earth, you perform that correction along with your transfer burn to take advantage of Uncle Oberth. 

in my own rss experience, that small inclination still costed nearly 1000 m/s to correct. it's not a small amount by any mean. I myself used a slower trajectory, launching from earth when earth was passing through a planar node; then I waited a couple orbits in that parking orbit until I intersected Mars. This way I saved those 1000 m/s - I needed a few hundred m/s to syncronize my orbit for a Mars encounter, but still a large saving. But this two-times trajectory is never used in reality.

what sparked my interest was exactly that I was surprised by how expensive that plane change was. And that none of the Mars missions ever mentioned that some transfer windows are cheaper than others, even though I know that if the window happens on a planar node you don't need the plane change and you can save a lot of deltaV. Generally speaking, in all my interplanetary transfers I went through great hassles to avoid paying for plane changes, because the cost was measured in km/s. and none of the planets are nicely aligned. I even named a subchapter in my mission report "there is no such thing as a simple hohmann transfer in rss" because of that.

Speaking of atmospheric capture, even though the game is a lot more forgiving than reality in regard to thermal control, I found Mars too fast for aerocapture. Ok, I could perform the injection burn in the high atmosphere and save perhaps 100 m/s, but that's it. I could then slowly lower apoapsis with multiple passages, but I had to pay the cost of injection with rockets. A mars lander has a thermal shield to protect it, but I don't expect a mars satellyte to be able to aerocapture.

On the other hand, Titan was extremely convenient for aerocapture. being much smaller, I got much slower at periapsis and I could aerobrake without burning. But then, I was coming in from Jupiter, or at most Ceres, so my intercept speed was a lot smaller than it would have been coming from earth; so I don't expect our real probes to be able to aerobrake at titan.

Quote

It took about 780 m/s of dV, spread over 10 different burns as it got closer and closer.

so little? Once more, this is a lot less than my own rss experience. When going to any target in solar orbit, even with the best hohmann transfers, intercept speed is going to be high. ok, its perihelion is close to earth, so with the right gravity assist you can eject into a very similar orbit, with the same aphelion, you only have to slightly raise periapsis. and I suppose they picked the comet specifically because it was in the right orbit. but still, it's impressively low.

Quote

I don't know what your approach is to approach Mercury, but all missions to Mercury -- Mariner 10, MESSENGER, and the ongoing BepiColombo -- have used at least one planet flyby to reduce the propellant needed to reach the innermost planet. Mariner 10 used a single Venus flyby but it never entered Mercurian orbit, instead doing all of its observation during three Mercury flybys. MESSENGER used one Earth flyby, two Venus flybys, and three Mercury flybys before finally performing its Mercurian orbital insertion burn. BepiColombo has already done one Earth flyby, two Venus flybys, and two Mercury flybys, and will do four more Mercury flybys before finally entering a polar orbit around Mercury with minimal propellant use.

starting from phobos because my mothership was refueling there, I took an earth flyby (1500 m/s) to get into a venus intercept. From venus I took a couple more flybys to lower my orbit. there I faced a conundrum; the cheapest mercury injection is by meeting mercury at perihelion, so that you have less of an aphelion lowering to pay for. But on the other hand, the cheapest injection is also on a planar node, because you can use venus gravity to pay the plane change. But the planar nodes are shifted by almost 90° from the aphelion-perihelion axis, so I had to pick one. In the end I decided to meet the innermost planet at perihelion, and I had to pay almost 3 km/s for a plane change in solar orbit. and there were still over 2 km/s of intercept speed, for a total cost of roughly 7 km/s from phobos to mercury orbit.

Keeping the mothership in solar orbit and using one of its subships as an impromptu additonal stage I managed to land on mercury and return to phobos, but I was left with the feeling that I could have been more efficient in the approach. that's why I was specifically interested in that part. real mercury orbiters used ion engines, so it's going to have a different profile, but they still save as much as they can by gravity assists, and there I wanted to see if they could do it better than I did.

[JoeSchmuckatelli]

1 hour ago, RCgothic said:

This is how much DV it generally takes to go places in the solar system. 

I look at these and have to remind myself that they're based (AFAIK) on an 'optimum' relationship between the orbits of the Earth and destination.  i.e. you can't launch at just any old place in Earth's orbit and hope to get to Jupiter for 3360 m/s, the planets need to be in the correct alignment, like the hands on a clock where the big hand points at the target and the little hand at the Earth.  When those are aligned correctly, you have a launch window.  (Which reminds me; are there simple maps that show these?)

My question: is there a 'fan' of possible direction changes from just an intercept?  I'd guess that would be determined by the mass of the planetary body, where Jupiter would give you a pretty large fan, but Mars would be considerably constrained.  Any easy to understand rule about this?

Finally - if you wanted to do a 'whip around the planet and return' mission, should I read the above as needing some percentage of deltaV of the 'Low Orbit' requirements?  

[RCgothic]

Oh yes, those are typical DV requirements during transfer windows. Should have made clear. If anything they're slightly conservative, giving a little margin. Gives a little leeway over when to launch and inefficient manoeuvres en-route. For return trajectories you can hit pretty much any periapsis off the "intercept* figure.

[king of nowhere]

I used the deltaV map when i was less experienced, but now i find it of little use. for anything further away than mars or venus i'd use gravity assists, and the injection values are often not meaningful. for example, if I wanted to visit europa i wouldn't inject into a low jupiter orbit, circularize, and then go to europa. I'd inject directly into europa, or else I'd find some compromise between haviing enough oberth effect from jupiter but reducing intercept speed.

even for a normal mission, what you actually have to pay for inclination can be very different.

[sevenperforce]

5 hours ago, king of nowhere said:

well, clearly I don't know what to look for.

As you learn and research more about the history of space travel and the mechanics of interplanetary missions, you'll get a better idea of what to search for.

I don't mean this (at all) to be rude or condescending. I have been actively and aggressively researching space travel and the history of spaceflight for over a decade, I have a history degree, and I teach graduate-level research. But I promise you, it's not that hard to pick up.

5 hours ago, king of nowhere said:

My doubt about that is that the shuttle put hubble into orbit, and hubble is a lot higher than the iss. 300 m/s are not enough to raise its orbit to 500x500, especially not if you also have to pay for reentry.

unless that the external tank had enough of an extra fuel to support the shuttle entering a 500x50 orbit before being dropped? so it means in regular missions it was dropped with some fuel still in it?

Hubble orbits at an altitude of approximately 535 km; ISS orbits at an altitude of approximately 408 km. 535 km seems like it's 31% higher than 408 km, right? That's a lot. But the energy of an orbit is actually measured with reference to the distance from the center of the primary. The radius of Earth is 6,378 km, and so the difference in altitude between a 535 km orbit and a 408 km orbit is actually just 1.78%. That's not much.

Early in the Shuttle program, each launch would go to the same approximate apogee with the same suborbital perigee, and then Shuttle would use its OMS to circularize (that is, bring its perigee up to its apogee), and then Shuttle would perform a series of phasing and orbit-raising burns to get to wherever it wanted to go. However, as NASA's experience with the Shuttle grew (and the mass of the external tank decreased due to materials improvements) they started to do "direct ascent" launches where the external tank and SSMEs would burn to the final target apogee and Shuttle would only need to do a single burn to raise orbit and phase for the destination. That's much more efficient (although the external tank's suborbital perigee was generally lower and thus would impact a few hundred miles west of the original set of launches).

How familiar are you with Hohmann transfers? There are so many huge differences in how you can do an efficient transfer.

5 hours ago, king of nowhere said:

in my own rss experience, that small inclination still costed nearly 1000 m/s to correct. it's not a small amount by any mean. I myself used a slower trajectory, launching from earth when earth was passing through a planar node; then I waited a couple orbits in that parking orbit until I intersected Mars. This way I saved those 1000 m/s - I needed a few hundred m/s to syncronize my orbit for a Mars encounter, but still a large saving. But this two-times trajectory is never used in reality.

The average mid-course correction for a Hohmann transfer between Earth and Mars is between 92 and 103 m/s of Δv.

My guess is that your trajectories are not optimal. Nothing to be ashamed of, of course. This is actual rocket science.

5 hours ago, king of nowhere said:

Speaking of atmospheric capture, even though the game is a lot more forgiving than reality in regard to thermal control, I found Mars too fast for aerocapture. Ok, I could perform the injection burn in the high atmosphere and save perhaps 100 m/s, but that's it. I could then slowly lower apoapsis with multiple passages, but I had to pay the cost of injection with rockets. A mars lander has a thermal shield to protect it, but I don't expect a mars satellyte to be able to aerocapture.

Real-world heat shields are many many times tougher than the heat shields in KSP.

No Mars satellite has ever aerocaptured into orbit. The Mars Reconnaisance Orbiter performed a propulsive capture followed by many successive aerobraking passes, but those passes were highly controlled because propulsive capture had already been completed.

5 hours ago, king of nowhere said:

6 hours ago, sevenperforce said:

12 hours ago, king of nowhere said:

how much did it cost for the rosetta mission to rendez-vous with its target?

It took about 780 m/s of dV, spread over 10 different burns as it got closer and closer.

so little? Once more, this is a lot less than my own rss experience. When going to any target in solar orbit, even with the best hohmann transfers, intercept speed is going to be high. ok, its perihelion is close to earth, so with the right gravity assist you can eject into a very similar orbit, with the same aphelion, you only have to slightly raise periapsis. and I suppose they picked the comet specifically because it was in the right orbit. but still, it's impressively low.

Keep in mind that the Rosetta mission made approximately 10 orbits of the sun with multiple planetary flybys before completing its rendezvous with the comet. This was nothing even close to an Hohmann transfer. It was many many times more efficient.

5 hours ago, JoeSchmuckatelli said:

If you wanted to do a 'whip around the planet and return' mission, should I read the above as needing some percentage of deltaV of the 'Low Orbit' requirements?  

A free-return trajectory around a body is usually only a few percent more (in Δv costs) than a flyby.

Edited October 27, 2022 by sevenperforce
stinking formatting

[Ultimate Steve]

Adding to the Mars inclination discussion, you will generally not correct your inclination with respect to Mars to zero midcourse, rather, you will alter your Earth ejection to place the ascending or descending node at the intersection point. You can do this by including a normal or anti normal component to the ejection burn (likely accomplished by launching into an inclined orbit to save on fuel) or by burning prograde a bit more to increase the aphelion  so that a node lines up with the intersection point, although in the latter case you have to spend more fuel to capture.

The orbit will still be inclined but you can combine the plane change burn and your capture burn that way, and because of Pythagoras and possibly oberth it ends up cheaper than doing it in the middle of nowhere.

[K^2]

43 minutes ago, Ultimate Steve said:

Adding to the Mars inclination discussion, you will generally not correct your inclination with respect to Mars to zero midcourse

For manned missions, all of that above, absolutely. You generally don't want to shoot a manned mission into the empty space with precise point in time where their engines have to work or its curtains. Using Mars' gravity for a capture greatly increases your options, as you can achieve aero-capture using your maneuvering thrusters if your mains are out, etc. Redundancy is usually of higher priority than the absolute efficiency.

For an unmanned mission, you can get creative with gravity. An Earth fly-by a year after launch can give you an inclination correction almost for free, greatly reducing the delta-V requirement in exchange for a longer mission and very strict launch window.

[farmerben]

On 10/27/2022 at 12:28 PM, JoeSchmuckatelli said:

I look at these and have to remind myself that they're based (AFAIK) on an 'optimum' relationship between the orbits of the Earth and destination.  i.e. you can't launch at just any old place in Earth's orbit and hope to get to Jupiter for 3360 m/s, the planets need to be in the correct alignment, like the hands on a clock where the big hand points at the target and the little hand at the Earth.  When those are aligned correctly, you have a launch window.  (Which reminds me; are there simple maps that show these?)

My question: is there a 'fan' of possible direction changes from just an intercept?  I'd guess that would be determined by the mass of the planetary body, where Jupiter would give you a pretty large fan, but Mars would be considerably constrained.  Any easy to understand rule about this?

Finally - if you wanted to do a 'whip around the planet and return' mission, should I read the above as needing some percentage of deltaV of the 'Low Orbit' requirements?  

Porkchop plots might be the easiest way to view  these.  

http://www.projectrho.com/public_html/rocket/images/mission/Mars2353PorkChop.png

 

 

 

[king of nowhere]

16 minutes ago, farmerben said:

Porkchop plots might be the easiest way to view  these.  

http://www.projectrho.com/public_html/rocket/images/mission/Mars2353PorkChop.png

How so?

Ok, I see that on the right window the cost is 22 km/s. This says nothing at all about how inclination is handled, or what kind of maneuvers are made. No, it really says nothing.

So it takes roughly 10 km/s to reach Earth orbit, depending on how efficient is the launch. and from LEO, it takes some 4.5 km/s to get a Mars intercept.

Where do the other 7.5 km/s come from?

what is the usefulness of such a plot, except to estimate when to launch?

[K^2]

1 hour ago, king of nowhere said:

what is the usefulness of such a plot, except to estimate when to launch?

Each pixel on the porkchop plot is an entire mission plan with all of the maneuvers. The plot itself is just a handy way to quickly evaluate what your options are. Once you have some launch windows in mind, you start looking at the details.

[farmerben]

The z axis which is usually represented with color contrast shows how much delta V is required to leave and arrive at a specific date.  Many suboptimal trajectories are possible if the ship had delta V to waste.  Inclination is included in the window's delta V calculation.  Which direction your burns should be are not represented in the porkchop plot.  

[king of nowhere]

1 hour ago, K^2 said:

Each pixel on the porkchop plot is an entire mission plan with all of the maneuvers. The plot itself is just a handy way to quickly evaluate what your options are. Once you have some launch windows in mind, you start looking at the details.

yes, because a pc has generated the mission plan.

but without the mission plan, the plot itself is telling me nothing I don't know already.

[Bej Kerman]

19 minutes ago, king of nowhere said:

but without the mission plan, the plot itself is telling me nothing I don't know already.

You don't think it's because the mission plan comes after the plot?

[king of nowhere]

21 minutes ago, Bej Kerman said:

You don't think it's because the mission plan comes after the plot?

no. unless we have radically different ideas of what a mission plan is, that plot comes after the plan.

to make that plot, you need to already know the maneuvers. if you have a deltaV value for your destination, it means you already know you are going to burn at a specific time at a specific angle for a specific deltaV, and you know all the maneuvers needed. in fact, to make that plot you don't need just one flight plan, you need a pc that can simulate all the flight plans for all combinations of days and trip durations.

frankly, I'm not even sure of what's the point of plotting the resulting data like that; barring extreme situations like the martian, where you had to get food on mars before a certain day else mark whatney would die, you are always going to launch at the moment of minimum deltaV expenditure, or within a few days. for distant targets you may want to pick a high energy, faster trajectory, but on mars there's little point in it. so why build a whole plot instead of just telling the pc to give you the best day to leave and arrive?

Furthermore, the only way for the value of 22 km/s to make sense is if it includes landing on mars. and since landing on mars is done by aerobraking, including its cost in the deltaV is misleading. If I want to know how much fuel to load up, I don't want to include the aerobraking value into the mission plan. if I want to know when to turn on the engines, that's again not relevant.

Edited January 8, 2023 by king of nowhere

[farmerben]

Sometimes the porkchop plot shows multiple good windows, some might be particularly narrow.