Multi core launcher
The « Shuttle in parallel » configuration is similar to that considered in our first concept, when SpaceX seemed to favor the three core formula, “à la” Falcon Heavy. Except that, if we refer to the orders of magnitude of the single core launcher, each module should be equipped with 20 engines and lift some 2,500 tons of propellant.
An 8.40 m diameter core (like that of the SLS) would allow building a MCT launcher without requiring any change of scale in the production means.
At this stage, since this investigation of a multi core, prompted mainly by the search for a better ballistic coefficient for the shuttle, is motivated too by looking for some cost saving in the industrial production which would thus be less affected by a gigantic scale, it seemed interesting to push the logic to its end, wondering whether it would not be possible to design a MCT which would not have 3 cores but 5, i.e. 3 under the shuttle and 2 above. The attractiveness of this architecture, which certainly presents challenges of load distribution and aerodynamics, is that it would allow using cores of 8.40 meters diameter each, which was a standard size for NASA for its Space Shuttle tanks, and is still now, with the central body of the SLS heavy launcher. This diameter is in effect sufficient to lodge 12 Raptor in the propulsion bay, according to a ternary symmetry.
In the case of economic return of the classic scheme (§2.1.1), the required performance, with a payload of 85 tons, is 661 tons in LEO (at 300 km); it is 691.5 tons for the « immediate » return, even taking into account a payload reduced to 82 tons. These performances can be obtained with a multi core launcher made of 5 cores of equal size (diameter 8.40 m; height with ogive 64 m) each equipped with 12 Raptor engines of 270 tons ground thrust, i.e. a total of 60 (versus 61 for the single core; the thrust and take-off mass are thus substantially identical). The payload in LEO in the « immediate return » case is achieved thanks to the larger mass of propellant in the shuttle that can be dedicated to the insertion into LEO orbit.
Take-off and the initial phase, until a time to be optimized, are performed with the 60 engines at full thrust, the gravity loss rate being maximum. Then the central module is set at a minimum thrust regime, the 4 other modules continuing to operate as a first stage. In the table of launcher data, this moment is characterized by the percentage of propellant of stage 2 (the central core) burned together with stage 1. We considered the value to be 20%, corresponding to about 45 s, when the gravity loss rate begins to decrease.
After the 4 modules of the 1st stage are burnt out, they are separated and start operating a recovery return, for which they have retained 6.5% of stored propellant; this low percentage is sufficient to provide these modules a ΔV closed to the speed reached (theoretical ΔV of the stage less most part of the losses).
The central core, put on to full capacity again, acts, then, as a 2nd stage, retaining 18% of its mass of propellant for its return to Earth.
Eventually, after separation and rotation to acquire the right attitude, the shuttle having already opened its 6 engine doors and deployed its divergent extensions, ignites its engines to deliver the final thrust for getting into a parking orbit.
Both shuttle versions sensibly differ in their mass budget; the « immediate return » version has a dry mass of 11% and a takeoff mass from Mars increased by 68%, due to the ΔV effort required, which are really at the limit of possibilities for a single stage equipped with 380 s Isp engines! Planning so close to the capability limits is dangerous. Of course, some items in our mass budget could prove to be pessimistic, the “heaviest” (structures and thermal protection) have only been extrapolated from documented cases: Space Shuttle, X33 and, to a lesser extent, DRA5. But project experience shows that it is generally the opposite which occurs…
In the two tables hereunder you will see the characteristics of the launchers adapted to both kinds of shuttle considered; they are almost identical but for the contribution to launching the shuttle into LEO (and the mass), along which their operational characteristics do differ.
| ITEMS | VALUE |
| % propellant for recov. stage1 | 0.061 |
| % propellant for recov. stage2 | 0.180 |
| gIsp stage3 (380 s) | 3.730 |
| gIsp stage2 (363 s) | 3.561 |
| gIsp stage1 (352 s) | 3.453 |
| Final mass stage3 | 660.8 |
| propellant mass stage3 | 400.0 |
| % prop. stage 2 burnt during phase 1 | 0.200 |
| ratio mass stage1/stage2 (full) | 4.000 |
| Initial mass stage3 (shuttle) | 1060.8 |
| deltaV stage3 | 1.766 |
| initial mass composite1 | 13500.0 |
| structural index single core | 0.050 |
| total mass single core | 2487.8 |
| Propellant mass single core | 2369.4 |
| propellant mass (useful) stage2 | 1554.3 |
| initial mass composite2 | 2773.6 |
| deltaV stage2 | 2.994 |
| propellant mass stage1 | 9344.3 |
| deltaV stage1 | 4.068 |
| deltaV total | 8.828 |
| Overall losses, grav. & drag – Earth rotation | 1.100 |
| FINAL SPEED | 7.728 |
| deltaV for recov. stage2 | 5.269 |
| deltaV for recov. stage1 | 2.754 |
| composite n: whole ship upon operation of stage n | ||
| Orbital speed at 300 km : 7.725 km/s | Masses in metric tons | |
“delayed” return flight case
| ITEMS | VALUE |
| % propellant for recov. stage1 | 0.065 |
| % propergol for recov. stage2/td> | 0.180 |
| gIsp stage3 (380 s) | 3.730 |
| gIsp stage2 (363 s) | 3.561 |
| gIsp stage1 (352 s) | 3.453 |
| Final mass stage3 | 691.5 |
| propellant mass stage3 | 616.5 |
| % prop. stage 2 burnt during phase 1 | 0.200 |
| ratio mass stage1/stage2 (full)/td> | 4.000 |
| initial mass stage3 (shuttle | 1308.0 |
| deltaV stage3 | 2.377 |
| initial mass composite1 | 13500.0 |
| structural index single core | 0.050 |
| total mass single core | 2438.4 |
| propellant mass single core | 2322.3 |
| propellant mass (useful) stage2 | 1523,4 |
| initial mass composite22 | 2947.5 |
| deltaV étage2 | 2.590 |
| propellant mass stage1 | 9119.6 |
| deltaV stage1 | 3.887 |
| deltaV total | 8.854 |
| Overall losses, grav. & drag – Earth rotation | 1.100 |
| FINAL SPEED | 7.754 |
| deltaV for recov. stage2 | 5.269 |
| deltaV for recov. stage1 | 2.876 |
« Immediate » return flight case
Embarquement pour Mars – Préface Thomas Pesquet
Webinar #2 APM sur STARSHIP disponible en replay
RSS
A very interesting analysis. But I sincerely doubt SpaceX would go the route of sandwiching their passengers in amongst all that rocket fuel. So far everything that they have done has been safety first, re-usability second, cost third, and everything else following on from there.
A single core seems highly improbable, but multiple cores joined like Falcon Heavy (maybe even in a 5, or 9 core config) seems more likely. It maintains their philosophy of doing everything like a production line.
But even multi-core doesn’t preclude the placing of a large craft on the top.
And have you considered that landing of payload my not require the whole thing to return to earth? Payload capsules could be left behind (even maybe acting as shelters) which would reduce the return mass considerably.
I think you need to seriously consider the cost of the relative parts to see which bits they will return and which will get left behind. It’s not all or nothing here.
Anyway, fascinating analysis. Thanks.
hi,
*passengers safety: in any case, as there is no possibility to install a launch escape system, it could prove necessary to ferry the passengers with « taxis » after the mammoth launcher has placed the MCT in LEO.
*placing the craft on the top: as I wrote in the report this poses the dilemma of choosing between a low ballistic coefficient (to improve EDL and thermal conditions) and a lighter vessel (with no wings at all) ; I choose the first option, as I thought it was a condition for a really reusable Thermal Protection. I may be too pessimistic.
*not returning everything on Earth: many people think even not returning the shuttle on Earth and refueling it on LEO… In this case, we certainly come to a less gigantic launcher, but on the cost of multiple launchs and augmented operations complexity.
Thank you for your contribution.
Richard
You start with the assumption that the Mars landing craft must have a minimum habitable volume of 10m^3 per person. Dragon v2 has a habitable volume of 10m^3 for 7 people. Why would the Mars landing craft be any bigger than necessary to accomodate 100 people crammed together in their seats? This landing craft could be a capsule shape. I think the MCT architecture will go something like this… 100 person capsule is launched into LEO where it is docked with (3) BA-330’s and a Mars TMI stage. With a 6 month voyage to Mars, everyone will have about 10 m^3 of space. Once at Mars, the capsule separates and performs EDL. The BA-330’s head back to Earth on a free-return trajectory. A capsule on Mars could launch and rendevous with the BA-330’s if anyone wants to go home.
That makes a lot of sense. I think a permanent spacecraft (as in, a spacecraft that never reenters the atmosphere), is very likely. The use of BA330, if bigelow is successful, is almost a given. Isn’t 10m^3 a little cramped for a month long voyage ? I’m trying to remember how big a dragon is, and I wouldn’t like to live inside one for month. But space is a hard place, so it might do.
Also consider that this is 10m^3 of 3D space in microgravity. Every wall and ceiling is another potential floor. Take a look above you at the ceiling (assuming you’re indoors) and imagine what you could do with all that unused real estate if gravity weren’t in your way. With that in mind 10 m^3 might even be too much room.
10 m3 is not for EDL but for the long transfer journey. It’s very minimum ; 15 m3 preferable (and possible in this design).
Landing 100 pax / 100 T total P/L within a capsule is not feasible ; the b allistic coeff would be very high, with inefficient deceleration and very high thermal conditions. You need to present a large surface to the atmosphere. Furthermore, you need another different vehicle to come back !… The idea of this study was to have only one piece of HW (the shuttle) and to not throw away anything (fully reusable). That said, this principle leads to an over-sized launcher, and raises the question of the reusabillity period (52 months except if it can afford an « immediate » (opposition type) return.
Your proposal of a BA330 on a free return trajectory is seductive. But it looks like the cycler concept, which Elon Musk has explained why it will not be his choice.
You forgot about in orbit refueling,
as per Musks comment:
« I mean, if you do a densified liquid methalox rocket with on-orbit refueling, so like you load the spacecraft into orbit and then you send a whole bunch of refueling missions to fill up the tanks and you have the Mars colonial fleet – essentially – that gets built up during the time between Earth-Mars synchronizations, which occur every 26 months, then the fleet all departs at the optimal transfer point. »
You are righ ; I get a lot of commentaryabout that. But you have to understand that I wanted to look to this extreme solution of only one launcher. As you saw, the result is decouragingly gigantic. So yes, refueling could be a solution, but it would also be more difficult to get low travel price if you have a « bunch » of launches! Also, transefering cryogens in microgravity is not simple ; I would rather consider plugging interchangeable tanks.
Thanks!
Interchangeable tanks have several advantages; they avoid the complications of zero-g pumping, they could be collected in orbit while the expensive parts (the engines) of the tanker returns for more, and if removable at martian gravity, they could be trucked between the landing pad and the fuel factory, allowing these to be located safely apart from each other.
Richard,
I’ve seen a few attempts at modeling this before, but yours is particularly impressive and the first to show the effects of the MCT’s dry mass. This is particularly useful for me, as we have no lack of people at Nasaspaceflight.com convinced that the most logical thing to do is to fuse the second stage and the MCT lander together. The argument they make is based on economics and not mass optimization, and I believe more than a few would change their minds if they could see how heavy the MCT would need to be to make it back from Mars. You should mention to your translator, Pierre, that the term « ton » in North America means 2,000 pounds, not 1,000 kilograms. To avoid any confusion I suggest using the term « tonne », which is the correct term in all English-speaking countries for the unit you mean.
It was mentioned that your reason for this approach was the fact that “I wanted to look to this extreme solution of only one launcher”. While I appreciate the simplicity of single reusable launch to Mars, there are a number of issues besides the fact that Spacex mentioned in-orbit refueling.
–Explosion Risk: A 2,750 tonne Soviet N-1 moon launcher exploded seconds after lift-off, resulting in a 7 kiloton blast that was the largest non-nuclear explosion in human history. You propose a 13,700 tonne rocket, which would mean you would have to sell the public on allowing a rocket able to explode with a force over 2X that of the Fat Man atomic bomb to launch anywhere near them. A smaller rocket would attract less public opposition and be far less dangerous and costly in event of failure.
–Engine Count: The only LV in history to have 30 engines failed in all 4 of its launches, though obviously Spacex are working on a 27-engine rocket. You propose that the risk of a 61-engine first stage is acceptable. I think the risk of engine failure leading to a catastrophic failure is too high, even with the more benign nature of the full-flow staged combustion cycle. A more realistic alternative is a 27-Raptor first stage, which would still enable a 5200 tonne RLV capable of pulling off your Mars mission in 3 launches.
–Economics: Your rocket’s first stage is 20 meters in diameter while the second stage is 12 meters. This requires two sets of tooling, which would add enormous cost. The ideal solution would be two stages with the same diameter, which would let you build them with the same set of tooling. This would make your RLV much less expensive.
–Lack of LAS: If you are going to launch 80-100 people at a time, chances are the US government will insist on there being some way to save the crew if the rocket were to explode at launch. You should probably add LAS engines to reflect this reality.
One possible reason why Spacex’s figures are so hard to attain is because you have used an entirely conventional approach. You probably did not hear that Spacex are considering solar electric propulsion in addition to the chemical engines to up payload and trim transit time. Some calculations on our site showed that the MCT could plausibly generate more than twice the power of the ISS, which would make this a viable option. Hope that helps and thanks again for the superb work!
Why assume the lander and rocket would transport 100 at a time? I imagine a large space station that you could attach many landers to and load it up over many launches to fill this colossal space station and use it as a back and forth transport ship
I heard Musk say that they will probably use a vehicle in constant transit within the Earth-Mars system (transporter) while using tanker launches to refuel the cycler. Can`t remember the source.
I haven`t heard anything regarding EDL but their current interest in hypersonic reentry profiles would indicate that whatever tech they use to land, will not be centered around inflatable decelerators nor cranes (due to high mass of the vehicles)
I guess we`ll have to hold our breath until September 2016 when the architecture is unveiled.
Just happened across your site and I really enjoyed reading your analysis and proposals.
My own opinions are :
1. Booster necessary to lift heavy hardware to Earth orbit.
2. Earth – Mars (maybe Moon use too) Ferry, the MCT
3. Mars SSTO for transfer of material and people between orbit/ground. CH4+O2 engines
4. Nuclear plant on Mars for in situ electrical and CH4 generation.
5. Some kind of reusable entry skin/bladder/container with engines that once the payload is extracted it is folded up and shipped back to Earth for refilling. Very low volume/mass.
Also the first few rockets would carry a constellation of combined Weather, Communication and GPS satellites to place in orbit around Mars.
With the payloads being talked about the MCT could probably deliver all this in once shot.
Investigate a Phoboa and Diemos tankage / rendezvous location as well.