Reusable earth departure stages

[Hop_David]

One of my favorite fantasies is staging platforms at EML1 and 2. Outbound spacecraft could go to these places to stock up on propellent, water (for drinking as well as radiation shielding), and air. These staging platforms could be supplied by volatiles from the lunar poles and/or carbonaceous asteroids parked in lunar orbit.

I imagine a spacecraft with an Earth Departure Stage (EDS) departing EML2 for a deep perigee via the 9 day Farquhar route. ~11 km/s at perigee confers a big Oberth benefit for a TMI burn or trans asteroid burn. After TMI, the EDS disengages, turns 180º and then does a braking burn. Slowing down to just below escape velocity would put the EDS on an ~ 9 day orbit. The 3rd perigee would be 27 days later and thus back in the moon's neighborhood. Then it could return to EML2 to get ready to send another spacecraft on its way.

But I was looking at a near parabolic orbit with a 400 km perigee. Precious little time is spent in the neighborhood of perigee. In the attached graphic a region of the ellipse is divided into 16 time increments, each increment about 3.4 minutes. The ship would spend about 54 minutes in the neighborhood portrayed.

Is there enough time to accelerate .6 km/s, disengage, turn 180º and then decelerate .6 km/s?

[savuporo]

The ship would spend about 54 minutes in the neighborhood portrayed.
Is there enough time to accelerate .6 km/s, disengage, turn 180º and then decelerate .6 km/s?

An RL-10 powered stage normally has a total burn time of 14-20 minutes. So, yes ?

[e of pi]

At 1G, conducting a 600 m/s burn would take only about a minute. 54 minutes is thus plenty of time for even a relatively low-thrust system (excepting ion drives and such, obviously) to make a 600 m.s burn, drop the payload, and brake back down those 600 m/s. Heck, at even a tenth of a G, 5 minutes would be more than enough. How long would flipping take after dropping the payload? I'm not entirely sure, but I definitely suspect much less than 45 minutes.

[jongoff]

But I was looking at a near parabolic orbit with a 400 km perigee. Precious little time is spent in the neighborhood of perigee. In the attached graphic a region of the ellipse is divided into 16 time increments, each increment about 3.4 minutes. The ship would spend about 54 minutes in the neighborhood portrayed.

Is there enough time to accelerate .6 km/s, disengage, turn 180º and then decelerate .6 km/s?

As others have noted, I would think so. It'll depend strongly on your system's T/W ratio though. The T/W ratio of the stage after separation (during the breaking burn) should be much higher. For instance, with a Dual Engine Centaur stage and a 60-ish tonne payload being slung on say a TMI trajectory, your T/W ratio is down around 0.25 for the departure burn (taking roughly 4min for the burn). But once you've staged, your T/W ratio is probably >1 (my super quick BOTE is saying you'd have a burnout mass on the Centaur of <9tonnes, which would give a T/W of ~2.5 for the Centaur, meaning it would only take it 24 seconds for the retro burn.

~Jon

[john smith 19]

The Project Troy study by Reaction Engines for a Mars mission utilizes a reusable EDS.

The OP narrows the field to EML1 and 2.

The study can be found here.

http://www.reactionengines.co.uk/tech_docs/mars_troy.pdf

It's a strategy that saves a fair bit of hardware.

[Hop_David]

But I was looking at a near parabolic orbit with a 400 km perigee. Precious little time is spent in the neighborhood of perigee. In the attached graphic a region of the ellipse is divided into 16 time increments, each increment about 3.4 minutes. The ship would spend about 54 minutes in the neighborhood portrayed.

Is there enough time to accelerate .6 km/s, disengage, turn 180º and then decelerate .6 km/s?

As others have noted, I would think so. It'll depend strongly on your system's T/W ratio though. The T/W ratio of the stage after separation (during the breaking burn) should be much higher. For instance, with a Dual Engine Centaur stage and a 60-ish tonne payload being slung on say a TMI trajectory, your T/W ratio is down around 0.25 for the departure burn (taking roughly 4min for the burn). But once you've staged, your T/W ratio is probably >1 (my super quick BOTE is saying you'd have a burnout mass on the Centaur of <9tonnes, which would give a T/W of ~2.5 for the Centaur, meaning it would only take it 24 seconds for the retro burn.

~Jon

Thanks! Do you know the dry mass and propellent mass a Dual Engine Centaur? Newtons? If I knew those things I believe I could do my own BOTEs.

A few things still unknown:

Time it'd take to flip 180 for the braking burn.

How much delta V it would take to return to EML2. I believe it's doable to time the third apogee to be in the moon's neighborhood. But the route from an apogee in the moon's neighborhood to a halo around EML2 still hasn't coalesced in my imagination.

[RanulfC]

Thanks! Do you know the dry mass and propellent mass a Dual Engine Centaur? Newtons? If I knew those things I believe I could do my own BOTEs.

I don't think there's been much "official" information on the DEC, these websites might help though:
http://alternatewars.com/BBOW/Boosters/Centaur/Centaur_GIW.htm
http://alternatewars.com/BBOW/Boosters/Centaur/Centaur.htm
http://www.ulalaunch.com/uploads/docs/Published_Papers/Upper_Stages/TheCentaurUpperStageVehicleHistory.pdf

A few things still unknown:

Time it'd take to flip 180 for the braking burn.

Less than a minute I suspect

How much delta V it would take to return to EML2. I believe it's doable to time the third apogee to be in the moon's neighborhood. But the route from an apogee in the moon's neighborhood to a halo around EML2 still hasn't coalesced in my imagination.

IIRC some of the early Mars missions (VonBraun? VSI?) called for reusable boosters that would return to Lunar orbit for refueling. I seem to recall that both nuclear and chemical boosters were discussed but my search-fu has failed me in finding a reference at this point.

Randy

[metaphor]

One problem would be that the EDS's engines would be pointed exactly at the departing payload for the braking burn.  So you would need to wait long enough for the stage to move away from the vicinity of the payload so it would be safe to start its engines.  That might take a few minutes depending on the strength of your RCS thrusters.

[john smith 19]

Note the key feature of Project Troy is the use of a resonance orbit in which the "departure" orbit is a sub multiple of the orbit around the planet or the Moon and is slightly below escape velocity, leaving the payload to do the last few m/s.

This in fact requires no braking thrust from the EDS.

[JasonAW3]

Why have the EDS detach and return back to the EML-2 point?

A Mars bound craft is going to need a midcourse correcction burn, and, if you weree using something like a Fusion engine, you'd want to do a continious acceleration out, a midcourse flip, and a continious decelleration to a safe atmospheric or orbital entry speed for the craft going to Mars. Once that speed has been achieved, detach the craft and begin accelletating the EDS back to Earth with a midcourse flip and correction, and decellerate it into EML-2 at that point for refurbishment and refuel.  In the meantime, the next EDS is launched before the first one has returned to Earth.

If an abort is declared prior to the detachement, they simply repeat the same procedure back to Earth.  If they abort to orbit, they wait there until the next EDS arrives and match velocities with it and dock after the EDS has dropped of it's payload.

In this situation, I'm estimating a round trip from Earth to Mars as 3 to 6 months, depending on orbital dynamics and at least 3 to 4 EDS units.  By using them to drop off supply payloads prior to launching a manned mission, the control and use procedures can be worked out, as well as Man Rating the EDS, prior to a manned mission.

[Nilof]

Well, if you reuse the EDS you can have it accelerate several payloads into Mars transfer orbit in one window. You can use it for sending stuff to the moon or to near earth asteroids as well.

[jongoff]

Well, if you reuse the EDS you can have it accelerate several payloads into Mars transfer orbit in one window. You can use it for sending stuff to the moon or to near earth asteroids as well.

My inner manufacturing engineer is a big fan of getting more "inventory turns" on your expensive hardware than once every two years. It would be interesting to see if you could find a way to enable multiple Mars departures in a single launch window with a single reusable EDS...

~Jon

[cordwainer]

I am guessing John when you say "resonance" orbit you mean low energy transfer orbit. While optimally you wouldn't need braking for the EDS for Mars you would need some braking for the Orbital Crew Module or payload that you would be sending. Albeit it wouldn't need much retro-thrust and such braking could be down gradually during the cruise phase using a low thrust/high Isp form of propulsion. As for the Moon it would probably be better to build the EDS and payload as one vehicle with two engine modules 180 degrees apart on booms, thus doing away with the need to flip the craft. You would have to have some method other than nose-in docking for refueling from a fuel depot, but if you have sufficient depot infrastructure there is no reason you couldn't build a reusable Cis-lunar shuttle.

[cordwainer]

Where they discuss resonance orbits is in relation to the OBO to minimize the energy and braking needed by the Skylon's to the departure window to meet up with the OBO.  The braking needed for the EDS to meet up and refuel with the OBO would be minimal, nothing that could not be achieved with simple RCS systems. Then the OBO maneuvers around the returning EDS to allow it to overtake the OBO upon return, which optimally doesn't need any braking. Still it is a somewhat risky maneuver.

[john smith 19]

Where they discuss resonance orbits is in relation to the OBO to minimize the energy and braking needed by the Skylon's to the departure window to meet up with the OBO.  The braking needed for the EDS to meet up and refuel with the OBO would be minimal, nothing that could not be achieved with simple RCS systems. Then the OBO maneuvers around the returning EDS to allow it to overtake the OBO upon return, which optimally doesn't need any braking. Still it is a somewhat risky maneuver.

Wouldn't that apply to any reusable departure stage?

[MP99]

After TMI, the EDS disengages, turns 180º and then does a braking burn. Slowing down to just below escape velocity would put the EDS on an ~ 9 day orbit. The 3rd perigee would be 27 days later and thus back in the moon's neighborhood. Then it could return to EML2 to get ready to send another spacecraft on its way.

When you intercept the Moon, you then need to perform further burns to target EML. That could be a problem 30+ days after departure, due to boiloff.

However, if you're prepared to wait, the EDS could follow a weak stability boundary orbit, which takes ~3 months, but requires nothing more than a few m/s of thruster to insert to EML.

I suspect part of the issue with this is whether the vector on the TMI burn lines up anywhere close to that needed for the return orbit.

Cheers, Martin

[Nilof]

My inner manufacturing engineer is a big fan of getting more "inventory turns" on your expensive hardware than once every two years. It would be interesting to see if you could find a way to enable multiple Mars departures in a single launch window with a single reusable EDS...

~Jon

I worked on some numbers on the "SpaceX FX/FXX/BFR Speculation Thread" for Mars transfers for the case where the MCT was an upper stage of an FXX-class rocket. I came up with some numbers for what I called a "Mars toss" mission:

MCT: 100 tons payload to Mars, has a 75 tons dry mass, and contains 775 tons of propellant when fully loaded. Has an Isp of 340 and a Delta-V of ~5.6 km/s with the full payload, 6.5 km/s with a 50 ton payload, 7.2 km/s with a 25 ton payload, and 8 km/s with no payload.

[...]
Mars toss transfer missions: gets fully refueled at LEO, places a 100 ton payload in a Mars transfer orbit with a 4 km/s burn, separates from the payload with ~200 tons of propellant(so ~4km/s delta-v) left. Slows down and lands at the launch site or docks with the depot within a day or two for a second mission.

The basic idea is that you can get back to the depot rather quickly if it's in LEO, using a high T/W transfer stage that can place a payload in a hyperbolic orbit and then quickly brake into an orbit that will put it back at the depot within a small integer multiple of the depot's orbital period.

With that said, Mars transfers do chug through propellant, as in roughly 7-8 tons of propellant per ton of payload for Kerolox or Hypergolics. However, stage reuse doesn't change that much if your stage has a decent mass ratio. For multiple launches, the large depots needed would be the bottleneck either way.

[jongoff]

My inner manufacturing engineer is a big fan of getting more "inventory turns" on your expensive hardware than once every two years. It would be interesting to see if you could find a way to enable multiple Mars departures in a single launch window with a single reusable EDS...

~Jon

I worked on some numbers on the "SpaceX FX/FXX/BFR Speculation Thread" for Mars transfers for the case where the MCT was an upper stage of an FXX-class rocket. I came up with some numbers for what I called a "Mars toss" mission:

MCT: 100 tons payload to Mars, has a 75 tons dry mass, and contains 775 tons of propellant when fully loaded. Has an Isp of 340 and a Delta-V of ~5.6 km/s with the full payload, 6.5 km/s with a 50 ton payload, 7.2 km/s with a 25 ton payload, and 8 km/s with no payload.

[...]
Mars toss transfer missions: gets fully refueled at LEO, places a 100 ton payload in a Mars transfer orbit with a 4 km/s burn, separates from the payload with ~200 tons of propellant(so ~4km/s delta-v) left. Slows down and lands at the launch site or docks with the depot within a day or two for a second mission.

The basic idea is that you can get back to the depot rather quickly if it's in LEO, using a high T/W transfer stage that can place a payload in a hyperbolic orbit and then quickly brake into an orbit that will put it back at the depot within a small integer multiple of the depot's orbital period.

With that said, Mars transfers do chug through propellant, as in roughly 7-8 tons of propellant per ton of payload for Kerolox or Hypergolics. However, stage reuse doesn't change that much if your stage has a decent mass ratio. For multiple launches, the large depots needed would be the bottleneck either way.

Nilof,

One challenge that I think Hop is hitting on is that if you do a direct from LEO departure burn to Mars, by the time you're done with the toss, you're going to be pretty far from periapsis, and may thus pay a pretty heft propellant "fine" to keep the stage. I would think that if you wanted to recover the stage, your best bet would be first to boost to a highly elliptical earth orbit (or come in from L1/L2), and only do the last couple hundred m/s of the departure burn when you're back approaching periapsis.

I'm doing a paper right now that's vaguely related (it's looking at a departure burn method to enable departures to arbitrary BEO destinations using a LEO depot), if we actually get it written this year, it might be good to do something like this as a follow-on.

~Jon

[Nilof]

Well, when I ran the numbers, my main conclusion was that the ~4 km/s Mars burn is small enough that there's lots of margin in the delta-v budget for slowing down.

In the example above I assumed a rather mediocre mass ratio(~11) for the upper stage because I assumed that MCT would have a heat shield. If it is a reusable transfer stage and nothing else, you can assume a mass ratio of ~16 for a Kerolox stage, and you get significantly more than 4 km/s available for a braking burn. So even if the timing on the first braking burn isn't ideal, braking down the stage is still quite practical.

[cordwainer]

True, any EVA docking would be risky.  If that is what your referring to, then yes. I think it comes down to designing appropriate docking, guidance and EDS capture technology with sufficient safeguards and redundancy to the OBO. I think you would want two OBO's in trailing orbit of one another in case you fail to capture on the first attempt or if something catastrophically goes wrong with one of the OBO's.