Vacuum only engines - possible to catch a spacecraft on a ballistic arc?

[SomeGuy12]

I thought of a maneuver that might be interesting. Let's suppose you had an interplanetary spacecraft with a powerful, high thrust, high ISP engine that only works in vacuum. Examples : a high efficiency NERVA engine. Reason it would only work in vacuum is that the propellant flow isn't enough to shed the waste heat from the nuclear reactor, so the engine would need a massive heat radiator. If the heat radiator were a droplet radiator or a lightweight conventional radiator, even the slightest impingement of atmosphere would destroy it. Or an open exhaust fusion engine would presumably be the same way - the tiniest bit of cold atmospheric gas would contaminate the fusion reaction and halt it.

So you've got your interplanetary mother ship in a circular orbit. Explorers are down on the planet and want to get back to the mother ship. You have balls of steel and perfect engineering skillz so you are confident your equipment will not fail.

So you undock the engine from the rest of the mother ship, leaving the main propellant tanks, habitation and laboratory modules, etc in orbit. The engine module is just the heat radiator and enough propellant tankage for this maneuver, and the power take-off module that extracts power from the nuclear engine. (with a fusion engine, you'd collect power from the electrically charged particles from aneutronic fusion, you'd have a sterling engine if it were a NERVA fission engine)

The ascent vehicle has enough delta V to reach apogee above the atmosphere, but will swiftly fall back down. Your engine module does a big braking burn and essentially intersects the ascent vehicle at the top of it's ballistic arc. You dock instantly (the ascent vehicle is computer controlled and aligned the right way) and then the engine module does a high ISP burn to get back into orbit. With the engine module + ascent vehicle mass, in vacuum you develop enough thrust to be greater than the gravity of the planet in question.

Reason to do this is to be a propellant miser - a heavy vacuum-only nuclear engine is going to be more efficient than anything an ascent vehicle could have on it.

Reasons not to do this - on many possible approach trajectories, the nuclear engine will pass right by the crew compartment of the ascent vehicle. A NERVA engine or a fusion engine at full burn would be just as dangerous as standing next to the live reactor core after the Chernobyl reactor exploded - lethal dose in seconds.

If the maneuver fails, the ascent vehicle would fall back down. On somewhere like Mars, it would mean death for the crew as it wouldn't have enough propellant to stop.

Would it work?

- - - Updated - - -

Oh, I thought of a few tweaks to make it more practical. You'd send the ascent vehicle up with enough fuel so that if the maneuver does fail, it can safely land. You'd "catch" it and transfer propellant to the tankage of the engine module, then drop it. This would be a way to refuel the mothership. You'd send the ascent vehicle up unmanned to avoid the radiation problem mentioned before. The times you have crew onboard, you would do a more complete burn with the ascent vehicle to put it all the way into orbit.

[Hcube]

So you mean a sub-orbital RDV then orbital injection with the mothership to use the engine that has the best Isp ? Well... why not ? but it sounds very risky and also not very cost/mass efficient : the whole mecanism to attach/detach the engine package would probably be very heavy and complex, with a lot of valves ! while the gains from using a smaller lander and have a better Isp for a single burn are probably not huge...

[Nibb31]

The docking is where it goes wrong. To dock means only two possibilities:

- The relative speed between the two vehicles must be zero, so either your ascent vehicle has to accelerate to the same speed as the descent vehicle, or the descent vehicle has to match the speed of the ascent vehicle. Either way, you have gained nothing in terms of energy.

- The relative speed between the two vehicles is high, so the amount of energy that needs to be absorbed by the docking system is equal to the energy differential between the two vehicles, so if the ascent vehicle is near zero velocity at its apogee and the descent vehicle is at 4000m/s, then you have to absorb that energy differential without being blown to pieces. Tough.

[shynung]

The docking is where it goes wrong. To dock means only two possibilities:

- The relative speed between the two vehicles must be zero, so either your ascent vehicle has to accelerate to the same speed as the descent vehicle, or the descent vehicle has to match the speed of the ascent vehicle. Either way, you have gained nothing in terms of energy.

- The relative speed between the two vehicles is high, so the amount of energy that needs to be absorbed by the docking system is equal to the energy differential between the two vehicles, so if the ascent vehicle is near zero velocity at its apogee and the descent vehicle is at 4000m/s, then you have to absorb that energy differential without being blown to pieces. Tough.

This is basically correct. From what I learned on KSP, docking is a devilishly hairy thing to do. RDV and matching velocities, and the problem of maneuvering the ports to mate each other, with both ships being in a stable orbit, is already complicated. Doing all this stuff while being in what is basically an ICBM trajectory (high apogee suborbital, right?), along with all the tight time constraints, just brought it to another level of difficulty.

[Dkmdlb]

He's not talking about ramming into a sub-orbital payload at orbital speed, he's talking about doing a retrograde burn with a high efficiency engine to meet the sub-orbital payload at the top of its arc and at the same speed, at which point the payload is picked up and accelerated to orbital velocity using the high efficiency engine.

"The ascent vehicle has enough delta V to reach apogee above the atmosphere, but will swiftly fall back down. Your engine module does a big braking burn and essentially intersects the ascent vehicle at the top of it's ballistic arc."

[shynung]

I understand that. What I'm pointing out is that doing all the docking movements while being in a suborbital trajectory puts a time limit on the docking maneuver.

[Nuke]

i once managed to rescue a ship on a high (300ish km iirc), but planet intersecting trajectory when its reactor overheated and blew up (i was using one of the nuclear-electric mods) making the engines useless. this happened before the circularization manuver and so the ship was pretty much doomed. fortunately i had another nuclear electric ship in a lower, circular orbit. with a great feat of rendezvoodoo i managed to plot an intersect of the two orbits just minutes before it hit atmo. this was followed by one of the quickest and dirtiest dockings i had ever pulled off. that was followed by a radial burn to push perigee far enough out of kerbin to not die. there was a little unintentional aerobraking involved but it worked. it was nerve wracking and took 3 reloads to pull off correctly.

doing similar with a sub orbital vehicle and an engine module would be even more difficult. you would need enough extra fuel on board to abort the maneuver and re-orbit the engine module should something go wrong. however the whole thing could be planned well in advance. one thing you couldnt begin the de-orbit burn until the sub orbital module module and engine module are right on top of each other, posing a collision hazard at high relative velocity (compounded with the fact that you are actually wanting to get as close as possible). you then need to bleed off the difference in velocity very quickly, followed by a quick and dirty approach and docking. i think the current school of thought on docking at nasa is slower is better, and this would be the complete opposite of that way of thinking.

[Scotius]

If anything goes wrong with the lander, you will lose it. But if engine module fails, you will have a total failure of the mission - lander and engine will be lost, and crew will be stuck in orbit until you send another engine. It will complicate the overall mission, make it more risky... And if you have the technology to pull off something like that, you probably are advanced enough to not care much for some extra dV for the lander.

[SomeGuy12]

The docking is where it goes wrong. To dock means only two possibilities:

- The relative speed between the two vehicles must be zero, so either your ascent vehicle has to accelerate to the same speed as the descent vehicle, or the descent vehicle has to match the speed of the ascent vehicle. Either way, you have gained nothing in terms of energy..

Correct, you don't gain in energy. However, if your NERVA engine has an ISP of 1200, or you have a fusion engine with an ISP of 10k or more in high thrust mode, you gain a lot of propellant in theory. You lose the propellant to decelerate the engine module down to meet the lander, and vice versa, but it could result in a net gain if you had a hugely more efficient engine.

Energy isn't the problem with nuclear engines - plenty of energy from the mass deficits. (the matter going to energy).

If you have the technology to pull something like this off, you still want to conserve rocket fuel - it's still limited even if you have insanely good technology.

[vger]

It sounds to me like this would be a very dangerous move for both vehicles. You've seen how difficult docking is even when a ship has "all the time in the world" to complete the maneuver. The precision required is ridiculous and demands a very steady hand. What you're proposing is like shooting a bullet at the mothership and expecting the bullet to stop on a dime with an error margin of only a few inches.

[Shrike99]

I'm sure that Scott Manley did something like this ages back, on the mun, his ascent craft was unable to make orbit, so he slowed the orbital part to match the suborbital ascent craft, docked, and accelerated again.

I will try to track the video down

[Meecrob]

What you're proposing is like shooting a bullet at the mothership and expecting the bullet to stop on a dime with an error margin of only a few inches.

Actually, he is proposing the mothership stop on a dime with an error margin of only a few inches, then accelerate to orbital velocity again.

[SomeGuy12]

Actually, he is proposing the mothership stop on a dime with an error margin of only a few inches, then accelerate to orbital velocity again.

Almost, but not quite. As the mothership's engine and the lander approach each other, both are using various sensors to precisely determine each other's relative velocity and position. (things like laser beams sent back and forth, etc). So you'd be adjusting the flight path of both vehicles on approach, 1000 times a second. Also, it's the mothership's main engine that does this, not the whole mothership - you shed the unnecessary mass for the maneuver.

[Meecrob]

Also, it's the mothership's main engine that does this, not the whole mothership - you shed the unnecessary mass for the maneuver.

Wait so the engine detaches from the mothership, docks with the ballistic capsule, then boosts it back to the mothership (sans engine) then performs two docking maneuvers (capsule and engine)...I don't think they have even started on docking ports complicated enough for instant engine attachments.

Edited July 28, 2015 by Meecrob

[SomeGuy12]

Wait so the engine detaches from the mothership, docks with the ballistic capsule, then boosts it back to the mothership (sans engine) then performs two docking maneuvers (capsule and engine)...I don't think they have even started on docking ports complicated enough for instant engine attachments.

Real life docking ports are a lot better, ironically, than the ones in KSP. Real life ports have things like motor driven bolts that insert and retract, and the attachment is just as secure as if the 2 spacecraft were manufactured in 1 piece. No wobble whatsoever. (well, some flexion in the structure, but that's going to happen even if the joint were not there)

In this case, the engine section would be connected to the rest of the ship by a long girder backbone that the whole ship is attached to. There would be around 8 or so large motor driven bolts at the attachment point. Docking would be computer controlled using thrusters that can provide very slight nudges for perfect alignment. There would also have to be detachment points for the ship's superconductive main power bus, the data lines, etc.

[Nothalogh]

Wait so the engine detaches from the mothership, docks with the ballistic capsule, then boosts it back to the mothership (sans engine) then performs two docking maneuvers (capsule and engine)...I don't think they have even started on docking ports complicated enough for instant engine attachments.

The idea is that the mothership is built as a hab module docked to a robotic tug.

The tug goes and grabs the ascent vehicle and hauls it to the hab, docks the ascent vehicle to the hab, re docks into tug position, and all goes according to plan, either yours or the kraken's

[sgt_flyer]

To perform such a manoeuver, you'll have an extremely tiny launch window (especially if the mothership is not on an equatorial orbit) - if you miss the launch window, you'll need to add a dog leg manoeuver to your launch rocket ascent profile (were you wanted to save delta-V) because having the orbital spacecraft perform such a correction while it also have to slow down is really going to be using a lot of delta-V. (If the spacecraft could still do several orbits to adjust it's path for the correction it's much less problematic) -

the timings necessary for each vehicle are really going to be insane.

For such kind of vehicle, you should always have room to prepare for the unexpected anyway

Edited July 28, 2015 by sgt_flyer

[Beowolf]

I did that above Mun once. Lander was short on fuel, and I knew it couldn't make orbit but wanted to see if it would get high enough to manage a docking. After burning the ascent tank dry and finding the situation was worse than I'd feared, I accidentally quicksaved instead of quickloading...

* Orbiting mothership was in about the right place for rendezvous, had way too much dV and could pull 2Gs.

* Ships start about 30 km apart, with over 500 m/s closing velocity. Mothership was in a circular orbit at 20km. Lander had launched vertical reaching apo of only 15km. MechJeb said it'd crash into Mun in 3 minutes.

Took me over 20 tries, but I nailed it! My eventual solution involved burning prograde at 2Gs to reach rendezvous faster, flipping to retrograde without shutting down around halfway to remove all horizontal velocity. Then dive for the target. Docking contact was at close to 10m/s, which I was afraid would break both ships, but didn't (thanks to NavyFish for his docking alignment indicator!). Flip again to point up, boost at full which was around 4Gs after burning off so much fuel, and missed the ground by a few meters. I remember when I docked there was under 30 seconds to impact. Luckily both ships were short, squat designs so rotated fast.

Mothership had just enough fuel left to reach a 10km x 5km orbit and await rescue.

So yes, I can confirm it's possible. I'm not eager to do it again, though!

[MockKnizzle]

Just estimating with some quick numbers:

Say we have a mothership in LKO orbiting at 2200m/s, and say we launch our payload on a suborbital booster that can reach an orbital velocity of 1500m/s.

The mothership drive core needs to brake by 700m/s to rendezvous, and then at spend at least another 700m/s to boost back into orbit. So the whole maneuver will require 1400m/s from the drive core, half of that with the added mass of the payload. The question is, is it more economical to add that 1400m/s to the drive core, or to just add 700m/s to the payload ascent vehicle?

If the scenario is lofting payloads from Kerbin, then adding 700m/s to a lifter is pretty trivial and the added complications of a suborbital rendezvous probably aren't worth the marginal benefits. If, however, we're trying to launch a payload from somewhere that requires a lot more dV to orbit (i.e. Eve), then the cost of adding 700m/s to your lifter is much higher. In this sort of scenario, I can see a brake-and-boost orbit insertion being a lot more attractive.

[Nothalogh]

Also, the concept of this proposal reminds me of an HFY story I read once

[Jouni]

There is a similar maneuver that's quite useful in KSP. Assume that you have an interplanetary ship returning to Kerbin. The ship doesn't have enough fuel to slow down, and it can't aerobrake safely either. You launch another ship to orbit, rendezvous with the returning ship at the closest approach, dock with it, and use the new ship to capture the interplanetary ship to a nice low orbit. You have to know how to rendezvous and dock quickly and reliably, but otherwise it's surprisingly easy even with the stock interface.

[Thomassino]

Das Valdez did this in one of his streams on Duna. He got lander onto ballistic trajectory, interplanetary craft with NERVs matched velocities with lander, docked and prograde/radial burned until on a proper orbit. So it is possible.

[FleshJeb]

I tried this on Kerbin in preparation for an Eve ascent. I managed a few good intercepts, but they all ended up in disaster. I even tried armoring a pod in structural plates and ramming it with a Claw ship at <50m/s.

I just don't think it's feasible or efficient to do it in a deep gravity well. The differences in ISP of the two craft would have to be huge. Not to mention the "catcher" needs a high TWR.

[TheBedla]

<spoiler alert, i guess> Wasn't a similar maneuver used at the end of The Martian?

The MAV was a bit short on dV due to higher than expected drag on ascent, and what they did was an impromptu maneuver with the _entire_ interplanetary ship to match velocities with the MAV. Of course they didn't dock, but eva'd over.

[Endersmens]

<spoiler alert, i guess> Wasn't a similar maneuver used at the end of The Martian?

The MAV was a bit short on dV due to higher than expected drag on ascent, and what they did was an impromptu maneuver with the _entire_ interplanetary ship to match velocities with the MAV. Of course they didn't dock, but eva'd over.

Erm. Next time use a spoiler.

...

[noparse]

...

[/noparse]