Rethinking the orbital bus plan | Mission Planning Workgroup | Forum

Having slept on it, this now feels less to me like a problem with the orbital bus concept and more like a problem with the low energy transfer of the LL.  This approach saves the lander about 20% of the fuel budget of a direct-to-moon trajectory.  If the lander weighs about half the mass of the CM, then this is only 10% of the overall CLLARE fuel budget.

Now, of course, we should try to save fuel where we can, but this 10% saving is not free.  It comes at the cost of a 5 month transit time and a very large maximum distance from Earth during travel.  Because of the 5 month transit time we cannot use cryogenic fuels on the lander, which will force us to use either hypergolics or hybrids.  Hybrids on the lander would, I think, be a very awkward configuration (considering the need for long and thin grains).  Hypergolics tend to be positively nasty things.

Even more to the point, hypergolics have a lower energy density than cryogenics.  A 10% reduction in total energy required combined with an increase in kg of fuel per joule of energy (for both the LL's and the CM's orbital bus – if one of them is hypergolic then both have to be to enable the fuel transfer in lunar orbit before burning for home) might actually make the whole deal a bad idea overall.

As a less important point, the fact that the low energy transfer takes the LL so far away from Earth means that if we wanted to stay in radio contact with it, it would need a pretty powerful transmitter on it, even more powerful than the CM.  On a direct-to-moon trajectory, with the LL presumably coupled to the CM somehow, we would only need the one powerful transmitter on it.  Although, this situation is a little more complicated.  We will need to send data to Earth both from the CM during the trip to and from the moon and from the LL on the surface.  The LL will either need its own comms gear powerful enough to get to Earth or smaller, cheaper, weaker gear which talks to the orbiting CM which relays the signal back to Earth.  The problem with this is periodic radio blackouts as the orbiting CM relay passes behind the moon from the perspective of the LL.  Now, if we did take the low energy route, the LL would need a very powerful transmitter on it for telemetry back to Earth during its trip.  If we were clever, we could separate the LL from its orbital bus in such a way that this transmitting equipment remained on the orbital bus.  If we could get the orbital bus into an orbit where it was directly opposite the CM, and use its comms gear as a second relay, the LL would have a connection to at least one of the two orbiting relays most of the time.  But there are probably much easier solutions to the LL comms problem than this.

This kind of feels like a major shake up of our mission plan brewing. :/

Rocket-To-The-Moon said:

This isn't necessarily a bad thing. I'd rather we spend the time now to optimize the mission profile for a lowest cost approach instead of heading down a more expensive path.

Agreed, it isn't a bad thing – perhaps bad timing though if we're hoping to attract an influx of new members soon.

Rocket-To-The-Moon said:

We would still launch the lander and CM on separate launch vehicles (would we?)

I think it would best not to commit to an answer on this too much until we have figured out exactly what our overall plan will be propulsion wise.

Rocket-To-The-Moon said:

Doing this would allow us to fly only one engine (and one fuel supply system)…the lander's. The engine would perform all burns to include the lunar descent and ascent. In essence, the lander would be the bus; a bus with legs for the landing.

I am not too sure about this idea.  We would need a large enough engine and tanks do a TLI burn for the entire CM/LL bundle.  This would be rather a large assembly and carrying it all the way down to the moon and back up again could be quite wasteful – imagine Apollo dragging its SM down to the moon and back!  Or have I missed something?

Rocket-To-The-Moon said:

This is a thread that I hope receives the attention that it deserves.

Me too!

I think it is important that, when considering possible mission plan tweaks to account for the issues raised here, we try to reduce the complexity/criticality of the lunar orbit rendezvous after lunar ascent.  The old plan of transferring fuel during EVA felt extremely risky and naive to me.  If we can't avoid it then we'll just have to give it our best shot but if by doing things differently we can reduce or remove that sort of messing about in orbit, it would be good for the project as a whole.

I can certainly see the appeal of a single engine solution from a cost/simplicity point of view.

The key to making it work may be some sort of very clever fuelling system.  Maybe two sets of fuel tanks, a large set and a small set.  The lander engine is originally connected to the large set (which does TLI, lunar capture, lunar escape).  As part of undocking the LL from the CM before ascent, this connection is broken and a connection to the smaller set (which does purely lunar ascent and descent) is made.  After ascent the large tank is reconnected and the smaller tanks are left behind.

Sounds good as a high-level description like this but figuring out how it all goes together could be tricky.  It might be worth while looking into how NASA planned to use some of its super light landers, with regards to how they were docked with the CSM.

Goodnight!

Advantages of the one (lander) engine plan:

• Obviously, simplicity and cost effectiveness.  We only need to design, build and test one engine for our non-booster propulsion needs (vs 2 kinds of engine and 3 physical engines for the old plan or 2 for Apollo's plan).
• The lander needs to redock with the CM before heading for home, solving the problem of what to do with it and resulting in less trash left on the lunar surface.

Disadvantages:

• We need the LL to be able to properly and securely dock with the CM (at least once in lunar orbit, possibly also in LEO if we launch the CM and lander separately).
• Transfer from the CM to the LL is made a little trickier because the LL needs to be attached to the underside of the CM.
• We'll need a very clever, possibly complicated fuelling arrangement.

I don't think this is an unworkable proposal, and the first advantage is a big one (and quite in keeping with our Design Philosophy).  We just need to throw some brain power at it.  Unfortunately now is about the time when brain power will hit something of a low due to holidays.

For the sake of keeping things clear on the Wiki:

• The original plan of doing a low energy transfer is now called "mission plan alpha".
• The currently brewing plan of transporting the CM and LL together using a single engine is called "mission plan bravo".
• The Wiki page "CLLARE mission overview" includes (via Mediawiki's templating facility) mission plan alpha for now but with a note at the top stating "Various improvements to both plans will continue with time, and further plans may be proposed. The actual plan to be flown will be the first plan developed which the CSTART community considers to best satisfy the simultaneous goals of simplicity, low cost and safety/reliability", which seems reasonable.  We can change which plan is included by the overview page over time so that it always points to the current favourite.

I've done a few quick TLI simulations using a modified version of the code I used to generate the preliminary flight plans over the NGW forums.

You can get from a 200km circular parking orbit onto a trajectory that brings you near enough the moon for lunar capture by applying a 0.75 m/s/s tangential acceleration constantly for about 90 minutes.  This is a total delta v of 4275 km/s or about 135% of the sort of delta v you can get away with using an impulsive TLI burn.  So this approach is less efficient however:

• From the Wikipedia article on Hohmann transfers: "Such a low-thrust maneuver requires more delta-v than a 2-burn Hohmann transfer maneuver, requiring more fuel (for a given engine design). However, if only low-thrust maneuvers are required on a mission, then continuously firing a very high-efficiency, low-thrust engine might generate this higher delta-v using less total mass than a high-thrust engine using a "more efficient" Hohmann transfer maneuver. This is more efficient for a small satellite because the additional mass of the propellant, especially for electric propulsion systems, is lower than the added mass would be for a separate high-thrust system".  Whether or not this will actually be the case in our scenario I do not know.
• I am fairly certain that it is possible to do more efficient low-thrust transfers to the moon than this continuous tangential burn approach, which wastes a lot of energy.

At any rate, I am now feeling even more confident that this plan is feasible from an astrodynamics point of view.  The engine design perspective I remain unsure about, I've had a hard time finding data on long-duration rocket burns.

I do not think that what you've written above is the correct explanation for the decreased efficiency of the low thrust burn but I am only just starting to slowly wrap my head around some of the concepts involved here so I'm not too confident.

First thing to clarify:

because the thrust vector must constantly be angled toward the Earth to help counteract the centrifugal force.

When I talked about a 90 minute tangential burn earlier, by tangential I meant at a tangent to the Earth's gravitational pull, that is, in the direction of the spacecraft's travel, not tangential to the spacecraft's travel, that is, toward the Earth.  Applying thrust along the Earth-craft line is never a good idea because part of the thrust is wasted fighting against gravity.  Directing thrust tangentially to gravity eliminates any gravity loss.

I believe that the reason for the decrease efficiency seen here is the counter-intuitive Oberth effect, whereby you can achieve a greater total increase in orbital energy (kinetic energy plus gravitational potential energy) by doing a given delta-v burn close by to a planet (when your energy is mostly kinetic) than you can by doing the same delta-v burn further away from the same planet (when your energy is mostly potential).  When you apply a long-lived low thrust, the thrust is spread out over an eccentric flight path (see the image below from my simulation), which means that a lot of the thrust is delivered further away from the planet and hence does less to increase your overall energy.  Short lived, high thrust burns happen more or less the same distance from Earth.

This is why I think that the total delta-v required to get to the moon via low thrust that I gave above is higher than optimal.  Doing a series of low thrust burns at the perigees of increasingly eccentric orbits will end up using less delta-v overall than a long, continuous burn.  Of course, it will also increase travel time somewhat, since we waste some time waiting for half-orbits to pass (however, this does give us considerably more abort options).  Decreasing delta-v removes mass in the form of fuel and oxidiser, but extending total travel time adds mass in the form of oxygen, water and food.  It's a question of figuring out the trade-off point.

Another thing to bear in mind is that constantly applying a thrust tangentially to gravity will require rotating the spacecraft at the correct rate using RCS.  Without this approach, doing a long-lived low-thrust burn would be inefficient in that as the burn progressed, more and more of the thrust vector would be parallel with (and hence fighting) Earth's gravity.  This consideration may be another reason that Wikipedia pages etc. say that low-thrust maneuvers are less efficient than high ones.

Even if it is less efficient from a propellant standpoint this approach will reduce complexity a great deal.

Yes, and I think we should definitely be willing to sacrifice performance for the sake of simplicity, reliability and cheapness.

I remember Gary saying that firing a canon (ie. instantaneous acceleration) is the most efficient way to get to orbit. Perhaps the same holds true for orbit changes.

This is definitely true (i.e. impulsive delta-v is more efficient than anything else for orbit changes), I have read it in more than one place.

I haven't had any luck so far finding a precedent for chemical rocket engines burning more than 10 minutes or thereabouts.  One option to consider is simply building our nozzles out of something with a high enough melting point that they can survive a 10 minute burn and then replacing any longer burns in our flight plans with a series of 10 minute burns with cool-off periods inbetween (this would require designing the nozzle to maximise its efficiency at radiating away heat – adding fins to the outside would help with this).  The downside to this approach is that it requires an extremely reliable ignition system, which pushes us back in the direction of hypergolics.  A possible compromise would be to not turn the engine off after a 10 minute burn but throttle it right down to the point where the nozzle can still cool somewhat.

I have done some more simulations for low thrust transfers and have managed to confuse myself quite thoroughly.  By my understand it should be more efficient to do several short low-thrust burns at the perigees of increasingly eccentric orbits than to just do one long low-thrust burn with the same total delta-v (I am even sure I have explicitly read this somewhere), but my simulations show that this is less efficient.  I am not sure if there is a problem with my simulator code, my understanding of orbital mechanics, or both.

I might put this stuff on the back burner for a bit.  I got a gift card for a large chain bookstore for Christmas, and they have a really excellent looking astrodynamics text book at a good price so I'm going to buy it and knuckle down a bit until I really understand this stuff.

Of course, anybody else should feel free to do more investigation of this low-thrust option.  I do feel like we can make it work if we just know what we are doing.

I will spend the rest of the time before my announced slowing-down trying to make sure CSTART is as accessible as possible to newcomers.

Another thought: we could always, ala Apollo, have the final stage of the launch vehicle perform TLI.  Then the only "big burns" the lunar lander's engine would only need to do would be lunar capture and lunar escape, which would probably last 20-30 minutes instead of 90-100.  Or we could have the final LV stage do some of TLI and the LL engine do the rest.