SuperHeavy engine bay w/39 engines

[neutrino]

I made an “engineering” model (AKA PowerPoint drawing) for a different SuperHeavy engine bay configuration which now includes 39 engines (vs. the current 33): two rings with 34 fixed engines and a central cluster of 5 gimbaling engines. In this simple drawing, it appears that the gimbal range will be only very slightly reduced relative to the current configuration, and my non-professional hunch is that one does not actually need the full control authority of 13 (!) engines – 5 will still be just fine. Apart from the obvious benefits (in NSF-speak: "more engines is more better") there are some additional interesting points:

- Most launches - by far - will be using the tanker variant, and this variant will be mass-limited, not volume-limited, so a more powerful booster is a real plus.

- Having >60% fewer gimbaling engines also means a corresponding reduction in power needed for TVC - so less battery weight.

- Having 70% more *fixed* engines: at some point these were supposed to have more thrust than the gimbaling engines ==> even more thrust down the road.



What do you think?

[catdlr]

Where have you been, lurking since 2010? My! You've started several threads, one that garnered 124K views, yet no one has ever welcomed you to the Forum. Welcome. Hope you get a good response. I do remember Elon posting a tweet that he was interested in a few more engines, you know, trying to get to that magic "42".

Things to consider:
Plumbing
Thrust distribution
Puck size and complexity
additional helium capacity to start the additional engines
Engine shielding (which may not be as a concern with V3 raptors as V2 are)
V3 Raptors performance may allow fewer engines required.

There will be other comments and considerations upcoming.  Good luck with your question.

[neutrino]

Thank you, catdlt, for the warm welcome!

Getting to the "magical 42" is easy: by adding three more Rvac engines to the ship -- as was already hinted by Elon -- one gets 33 + 9 = 42 engines. But this is just for fun. As for your other comments, let's see:
- Plumbing: I trust SpaceX on this one. These things are generally solvable after a few iterations.
- Thrust distribution: I believe this is actually better with the new configuration: the thrust, on average will be closer to the load-bearing cylindrical outer shell.
- Puck size and complexity: from the above point - it will be either smaller and thinner, or double-zoned - with slightly different zones for each of the two inner rings.
- additional helium capacity to start the additional engines: thank you - another plus for the new design: LESS helium will be needed onboard, as all fixed engines are only ground-lit -- and fewer engines will be re-lighted in this design.
- Engine shielding: engine spacing is no worse than the current spacing of the outer ring.
- V3 Raptors performance may allow fewer engines required: I got this one comment already, but I still don't understand: why is this an advantage? as I wrote: most launches will be mass-limited, and so a more powerful booster will be a huge advantage.

A note: a 33->39 engine change means total thrust increase of ~18% to the total thrust and a tiny addition of <10tons (~0.2%) to the weight,Therefore the thrust-to-weight ratio will increase from ~1.5 to ~1.77. The vehicle's upwards acceleration will increase from ~0.5g to ~0.77g -- an increase of 54% (!) since the total drag on the vehicle did not change.  Obviously, a better use of this additional power is not to accelerate more quickly but to accelerate with more fuel and payload -- but what I wanted to say is that this is not a minor upgrade -- it is a major rocket version.

[kevin-rf]

One thing that may or may not be an issue is the second ring may be currently spaceed inboard like it currently is for thermal reasons.  Reducing the spacing and pushing it up against the outer engine ring might cause some thermal issues that would also need to be solved.

But,  who knows...

[Brigantine]

- Having >60% fewer gimbaling engines also means a corresponding reduction in power needed for TVC - so less battery weight.

I'm not sure that's right. A reduction in force, over an increased distance, in the same time achieves the same control authority with the same power

Also, you might want some of the ring of 14 to be re-lightable for the boostback - but that's not so difficult

[catdlr]

- additional helium capacity to start the additional engines: thank you - another plus for the new design: LESS helium will be needed onboard, as all fixed engines are only ground-lit -- and fewer engines will be re-lighted in this design.

Neutrino,
The current setup includes 20 fixed engines on the outside, each directly connected through one OLM-QD valve configured with a booster hold-down bracket. Adding 12 more engines will require installing 12 additional QDs and finding a way to fit them in and connect them to the engine heads that are several feet inside, will create a significant engineering hurdle and could be problematic if one or several get stuck when pulling back on liftoff. Additionally, there may be complications if two engines share a single QD plumbing, particularly regarding the engine starting sequence. As a solution, I suggest using onboard helium to start the row of 12 engines, especially if they are needed later for boostback and maybe landing. This approach avoids unnecessary complexity. If it works now why change it?

[neutrino]

If it works now why change it?

This is not quite the SpaceX general approach, aint it?   

And to your point(s): these are all relatively small technical details - all of which can and should be addressed in any version of SuperHeavy, not just the one proposed above  (e.g. the QD-OLM should be reliable no matter what). Using the on-board Helium supply can indeed be a quicker way to jump-start this version -- no OLM modifications will be needed. The exact routing of pipes and cables is well beyond the scope of this PowerPoint drawing ...

[Okie_Steve]

If the tanker variant is indeed mass not volume limited then how many more engines would it take to change that? Ignoring engine thrust upgrades and the unknown added mass in fuel lines etc for a first approximation.

[tbellman]

If the tanker variant is indeed mass not volume limited then how many more engines would it take to change that? Ignoring engine thrust upgrades and the unknown added mass in fuel lines etc for a first approximation.

Rough estimate follows.

Assumptions:
• SuperHeavy dry mass: 200 tonne
• SuperHeavy propellant capacity: 3400 tonne
• Ship dry mass: 100 tonne
• Normal ship propellant capacity: 1200 tonne
• Normal Starship payload capacity: 100 tonne
• Bulk density of subcooled methalox: 900 gram/litre
• Cargo ("nosecone") volume: 1000 m³

A fully loaded ship would thus mass 1400 tonne, and the full Starship stack is 5000 tonne.  (By "normal" Starship, I mean one that is designed to carry e.g. satellites to orbit, or cargo to the Moon or Mars, i.e. not a tanker or a depot.)

Making a tanker by extending the tanks of the ship fully into the nose cone would increase its propellant capacity by 1000m³ * 900g/l = 900 tonne.  The ship's mass would then be 100t+1200t+900t = 2200t, and the entire stack's mass would be 200t+3400t+100t+1200t+900t = 5800t.

5800t is a 16% increase over 5000t.  This actually matches fairly well with the 18% increase in thrust you would get be going from 33 to 39 engines on SuperHeavy.

For the ship, 2200t is a 57% increase over 1400t.  That's reasonably close to the increase in thrust you get by going from six engines to nine, which SpaceX is already planning.

[hkultala]

If the tanker variant is indeed mass not volume limited then how many more engines would it take to change that? Ignoring engine thrust upgrades and the unknown added mass in fuel lines etc for a first approximation.

Rough estimate follows.

Assumptions:
• SuperHeavy dry mass: 200 tonne
• SuperHeavy propellant capacity: 3400 tonne
• Ship dry mass: 100 tonne
• Normal ship propellant capacity: 1200 tonne
• Normal Starship payload capacity: 100 tonne
• Bulk density of subcooled methalox: 900 gram/litre
• Cargo ("nosecone") volume: 1000 m³

A fully loaded ship would thus mass 1400 tonne, and the full Starship stack is 5000 tonne.  (By "normal" Starship, I mean one that is designed to carry e.g. satellites to orbit, or cargo to the Moon or Mars, i.e. not a tanker or a depot.)

Making a tanker by extending the tanks of the ship fully into the nose cone would increase its propellant capacity by 1000m³ * 900g/l = 900 tonne.  The ship's mass would then be 100t+1200t+900t = 2200t, and the entire stack's mass would be 200t+3400t+100t+1200t+900t = 5800t.

5800t is a 16% increase over 5000t.  This actually matches fairly well with the 18% increase in thrust you would get be going from 33 to 39 engines on SuperHeavy.

For the ship, 2200t is a 57% increase over 1400t.  That's reasonably close to the increase in thrust you get by going from six engines to nine, which SpaceX is already planning.

with 230 tonnes per engine (raptor 2) , SuperHeavy already has about 7600 tonnes of thrust. Even with liftoff thrust of 5800 tonnes, this is still T/W ratio of 1.3.

And, Raptor 3 is probably getting 269 tonnes of thrust per engine. With 5800 tonnes of liftoff mass, 33 engines then means liftoff T/W ratio of 1.53, which is plenty.

There is no need to add first stage engines. Gravity losses will be greater, but total payload to orbit will still be clearly greater.

The engines are what are expensive. The tank volume is cheap, and fuel is cheap. There is no point of adding more expensive engines to optimize using the cheap tank space and cheap fuel more efficiently.

However, the much heavier tanker second stage will probably want to have 9 engines total.

[Brigantine]

What are the aerodynamic constraints to how tall the full stack can be? assuming no change in construction & wall thickness.

Until you hit that limit, you can always go higher chamber pressure and more engines, if you can tame the SH engine bay environment

Though, maybe you need some "v4 Full Height" high bays.

For now, Starship is a partially successful fully expendable suborbital sounding rocket. Anything better than that would be progress.

[thespacecow]

I mean SH used to have 37 engines, so I don't think an increase of engine count from the current 33 is out of question. But I imagine it's a pretty low priority right now.

[Orbiter]

Don't all of the inner 13 gimbal? Do they have enough clearance to gimbal in that configuration?

[RobLynn]

wouldn't they be better off with just two or three gimballing engines on mirror line between the two landing stub/load points so that SH can land an engine down on both load points without needing to lean over.  Fill in rest of space as completely as possible with non gimballing engines.

[Brigantine]

On IFT-1, even 13 gimbaling engines weren't enough to avoid a Kerbal maneuver...

Or was it fine up until the loss of hydraulic pressure?

In any case, one ought to consider the times you might want to steer, before cutting the steering range back too drastically - Though you have some engines (in the ring of 14 13) that have TVC, but whose range only goes towards the inside. There will still be at least some control authority in any particular direction. By the same logic, you should be able to pack central TVC engines quite close, if you commit to always keeping their movements correlated. (admittedly that introduces one single point of failure)

[neutrino]

Making a tanker by extending the tanks of the ship fully into the nose cone would increase its propellant capacity by 1000m³ * 900g/l = 900 tonne.  The ship's mass would then be 100t+1200t+900t = 2200t, and the entire stack's mass would be 200t+3400t+100t+1200t+900t = 5800t.

Did you notice what you just wrote? by adding 6 more engines - no other improvements included - the tanker's performance jumped 10x from 100t to 1,000t (!!) I am sure this is overly optimistic, but it does get the point across well: this proposed configuration may be really useful.

The entire discussion about engine costs is, IMHO, off the point: SpaceX is feverishly working on cheaper and more manufacturable engines, and they are but one part of the total cost -- but they contribute 100% of the performance -- any improvement in a rocket's engines is strongly amplified in the total system's performance. Also, it is difficult to pass by a 10x improvement as non-economical...

[tbellman]

Making a tanker by extending the tanks of the ship fully into the nose cone would increase its propellant capacity by 1000m³ * 900g/l = 900 tonne.  The ship's mass would then be 100t+1200t+900t = 2200t, and the entire stack's mass would be 200t+3400t+100t+1200t+900t = 5800t.

Did you notice what you just wrote? by adding 6 more engines - no other improvements included - the tanker's performance jumped 10x from 100t to 1,000t (!!) I am sure this is overly optimistic, but it does get the point across well: this proposed configuration may be really useful.

No, it did not.  I mentioned nothing about how much propellant would be left over when the ship reached orbit.  I was only comparing thrust-to-mass of the stack and of the ship in this scenario (stretched ship tanks and increased number of engines), with the thrust-to-mass in the current design.

By increasing the mass of the ship without increasing the propellant capacity of SuperHeavy, staging must happen earlier.  The ship then needs to power through more Δv to reach orbit.  This will consume most of the increased propellant load of the ship.

Any increased capacity to orbit comes mostly from the stretched tanks.  The increase in thrust only helps a little bit, by counteracting the otherwise increased gravity losses.  As hkultala mentioned, even without increasing the number of engines, adding 900 tonnes of propellant to the ship (and removing any other cargo) would still leave SuperHeavy with a thrust-to-weight of about 1.3:1, which isn't horrible.

Calculating how much propellant such a tanker would have left over when it reaches LEO is not trivial, as it depends on many factors.  My gut-feeling guess is that it might allow 200-250 tonnes (while a non-tanker, with 1200 tonnes of propellant, would be able to get 100-150 tonnes of cargo to orbit).  But I am not a rocket scientist, so I suggest that you don't put much trust in my gut feelings.

[scaesare]

I mean SH used to have 37 engines, so I don't think an increase of engine count from the current 33 is out of question. But I imagine it's a pretty low priority right now.

Doesn't that configuration make room for 6 of those engines by extending them out into the landing leg enclosures that no longer exist?