Thrust to weight ratio of starship V2 with 6 vacuum engines on the second stage

[InterestedEngineer]

One can't add 500t to the Starship (Fuel + cargo + rings) and not add fuel to the booster, the rocket equation doesn't work like that.

The current deltaV of the booster is about 3.5km/sec.   That's a mass ratio (M_r) of 2.7

If you add 500t of payload to the booster (aka fuel and payload and rings for Starship), to get the same deltaV, the booster needs 850t of fuel -- (M_r-1) times payload increase

There's no place for that fuel to go without adding rings.  Each ring adds 100t of fuel capacity.   So ~8 more rings.

You don't need to get the same deltaV, you can just have stage separation earlier. So the ship would do most of the work to reach orbit.

earlier stage separation also means that, for a RTLS flight path, the boostback burn requires less deltaV, which reduces the booster propellant reserves needed at stage separation.   This is likely only a small benefit, though.

I'm a little confused as to why they had so much more deltaV for Booster vs. Falcon-9.

Possibly the mass ratio of Starship is less than that of Falcon upper stage?

Falcon-9 MECO is ~2km/sec, so about 3km/sec deltaV.

Booster-7's MECO was supposed to be about 2.7km/sec (still looking for the source), so net 3.7km/sec deltaV. 

That requires ~6km/sec deltaV out of Starship, or a mass ratio of 5.3, which is pretty small really, the nominal mass ratio for Starship is 5.8 for a 100t payload and a 150t (wet) Starship at SECO.

If we cut Booster down to 3km/sec deltaV (2km/sec MECO), that's a M_r of 2.4, but that increases Starship's mass ratio to 7, which is not doable with today's Starship configuration, it requires another 300t of fuel or 3 more rings.

I really need to make a spreadsheet to analyze the corners here.

I note TWR is a completely different problem which also needs to be solved.  the 250t "booster" (limited throttle) version would probably help.

[InterestedEngineer]

I created a spreadsheet to estimate  the trajectory of Booster + Starship, calculate the fuel used, etc.

I feel like I'm missing ~300m/sec of deltaV.

But it looks like the booster does not need to be extended.  The model doesn't show any improvement in trajectory for 9 v 6 engines, just an improvement in TWR and less burn time for Starship.

My guess is I'm missing some gravity loss for the Starship (aka second stage), but I'm not seeing it.  its initial apogee after SECO is 121km and it's a cheap Hohmann transfer to get to 200km orbit.  Probably not modeling that correctly.

https://docs.google.com/spreadsheets/d/1bW0qWPjSl85lYLOwO9m6j9l71uWHWRSR1gMFtuR2bb0/

[hkultala]

Possibly the mass ratio of Starship is less than that of Falcon upper stage?

Falcon-9 MECO is ~2km/sec, so about 3km/sec deltaV.

Please do not mix and compare theoretical delta-vs versus real staging speeds. Even when you try to add compensation for the losses, as your compensation factors are way off.

MECO for Falcon 9 is typically at about 2.25 km/s for barge landings, and there is about 3.5 km/s of theoretical delta-v, and this is for barge landing. The losses and re-entry burn, landing burn etc for falcon first stage are clearly more expensive than you think.

And for return to launch site landings, the MECO is typically clearly below 2 km/s. Over 1.5 km/s for losses and recovery for the booster.

Booster-7's MECO was supposed to be about 2.7km/sec (still looking for the source), so net 3.7km/sec deltaV.

That requires ~6km/sec deltaV out of Starship, or a mass ratio of 5.3, which is pretty small really, the nominal mass ratio for Starship is 5.8 for a 100t payload and a 150t (wet) Starship at SECO.

If we cut Booster down to 3km/sec deltaV (2km/sec MECO), that's a M_r of 2.4, but that increases Starship's mass ratio to 7, which is not doable with today's Starship configuration, it requires another 300t of fuel or 3 more rings.

You are overexxagerrating the difference in lower stage delta-v from upper stage weight increase.

And the flyback becomes much cheaper with earlier staging, though gravity losses also increase when upper stage has more mass.

[InterestedEngineer]

Please do not mix and compare theoretical delta-vs versus real staging speeds. Even when you try to add compensation for the losses, as your compensation factors are way off.

MECO for Falcon 9 is typically at about 2.25 km/s for barge landings, and there is about 3.5 km/s of theoretical delta-v, and this is for barge landing. The losses and re-entry burn, landing burn etc for falcon first stage are clearly more expensive than you think.

Maybe I wasn't clear, I always do calculations in terms of conic sections, so when I say deltaV for MECO I'm including only the mass ratio for that part of the flight, not landing fuel.   So the mass at MECO includes the burn-back fuel, reentry burn fuel, landing fuel, the dry mass of the booster, as well as the fully fueled upper stage of course.

there's about 1km/sec of gravity losses, which is not trivial to calculate.  That and the angle of the vector above the earth is the only real "compensation" factor I'm guessing at.

Detailed spreadsheet here:

https://docs.google.com/spreadsheets/d/1bW0qWPjSl85lYLOwO9m6j9l71uWHWRSR1gMFtuR2bb0

[hkultala]

Please do not mix and compare theoretical delta-vs versus real staging speeds. Even when you try to add compensation for the losses, as your compensation factors are way off.

MECO for Falcon 9 is typically at about 2.25 km/s for barge landings, and there is about 3.5 km/s of theoretical delta-v, and this is for barge landing. The losses and re-entry burn, landing burn etc for falcon first stage are clearly more expensive than you think.

Maybe I wasn't clear, I always do calculations in terms of conic sections, so when I say deltaV for MECO I'm including only the mass ratio for that part of the flight, not landing fuel.   So the mass at MECO includes the burn-back fuel, reentry burn fuel, landing fuel, the dry mass of the booster, as well as the fully fueled upper stage of course.

there's about 1km/sec of gravity losses, which is not trivial to calculate.  That and the angle of the vector above the earth is the only real "compensation" factor I'm guessing at.

Detailed spreadsheet here:

https://docs.google.com/spreadsheets/d/1bW0qWPjSl85lYLOwO9m6j9l71uWHWRSR1gMFtuR2bb0

"boost back horizontal delta-v" of only 100 m/s (6 km/min) does not seem reasonable. Clearly faster horizontal flyback velocity is needed to get back to the launch site in the time available, unless staging very very early at very vertical angle.

And this is then worse if staging later

[InterestedEngineer]

"boost back horizontal delta-v" of only 100 m/s (6 km/min) does not seem reasonable. Clearly faster horizontal flyback velocity is needed to get back to the launch site in the time available, unless staging very very early at very vertical angle.

And this is then worse if staging later

Thanks for the feedback.

What time limit?  I think you mean before the vertical velocity causes it to hit the ground.   I can see i need a more complicated trajectory calculation for the boost-back velocity.

B26 has the angle, which after examining Falcon-9 I set at 30 degrees.   That horizontal velocity is canceled in line 33.

I"ll work on it, thanks for finding!

[InterestedEngineer]

Please do not mix and compare theoretical delta-vs versus real staging speeds. Even when you try to add compensation for the losses, as your compensation factors are way off.

MECO for Falcon 9 is typically at about 2.25 km/s for barge landings, and there is about 3.5 km/s of theoretical delta-v, and this is for barge landing. The losses and re-entry burn, landing burn etc for falcon first stage are clearly more expensive than you think.

Maybe I wasn't clear, I always do calculations in terms of conic sections, so when I say deltaV for MECO I'm including only the mass ratio for that part of the flight, not landing fuel.   So the mass at MECO includes the burn-back fuel, reentry burn fuel, landing fuel, the dry mass of the booster, as well as the fully fueled upper stage of course.

there's about 1km/sec of gravity losses, which is not trivial to calculate.  That and the angle of the vector above the earth is the only real "compensation" factor I'm guessing at.

Detailed spreadsheet here:

https://docs.google.com/spreadsheets/d/1bW0qWPjSl85lYLOwO9m6j9l71uWHWRSR1gMFtuR2bb0

"boost back horizontal delta-v" of only 100 m/s (6 km/min) does not seem reasonable. Clearly faster horizontal flyback velocity is needed to get back to the launch site in the time available, unless staging very very early at very vertical angle.

And this is then worse if staging later

The CRS10 had a horizontal velocity at apogee of about 480m/sec.  So I increased the value to about 460km/sec.

After trying to do a conic sections on a boostback and re-entry path.  Whew!   Very hard to get right.

It might be less, as Booster has more horizontal "surfing" capability than Falcon-9.

Also Booster can go higher because it can handle higher reentry velocities, and getting some vertical acceleration is free if you keep the angle under ~15% to minimize cosine losses.  So there's a bit of a higher boostback parabola than Falcon-9

But yes, 100m/s was too little.  Thank you for the correction!

Note net recovery fuel is 8.7% of total.  Boost back to liftoff for Falcon-9 is close to 10%, so I'm assuming an improvement due to no reentry burn and more horizontal surfing capability for SH Booster.

I also fixed the problem that the boost-back doesn't need to boost back against earth's rotation speed.

https://space.stackexchange.com/questions/20246/what-is-this-debris-on-the-crs-10-reentry-at-t630/20275#20275

[Sarigolepas]

I'm a little confused as to why they had so much more deltaV for Booster vs. Falcon-9.

Possibly the mass ratio of Starship is less than that of Falcon upper stage?

Falcon-9 MECO is ~2km/sec, so about 3km/sec deltaV.

Booster-7's MECO was supposed to be about 2.7km/sec (still looking for the source), so net 3.7km/sec deltaV. 

That requires ~6km/sec deltaV out of Starship, or a mass ratio of 5.3, which is pretty small really, the nominal mass ratio for Starship is 5.8 for a 100t payload and a 150t (wet) Starship at SECO.

If we cut Booster down to 3km/sec deltaV (2km/sec MECO), that's a M_r of 2.4, but that increases Starship's mass ratio to 7, which is not doable with today's Starship configuration, it requires another 300t of fuel or 3 more rings.

I really need to make a spreadsheet to analyze the corners here.

I note TWR is a completely different problem which also needs to be solved.  the 250t "booster" (limited throttle) version would probably help.

It's pretty much the opposite, the booster goes far slower than a falcon 9 booster, which helps it to land back on the launch pad with minimal payload loss compared to a droneship landing. They also don't need a reentry burn with superheavy.
That's why TWR matters because early staging means way more gravity losses (same issue with asparagus staging).

Starship can be refilled in orbit so it's only designed to reach LEO, so the weight of the ship is not an issue, it's quite the opposite in fact since the ship can be considered as part of the payload since it can be used as a space station, a spacecraft, a fuel tank, a moonbase or just a third stage after being refilled. A falcon 9 second stage however is just dead weight once it has burned all of it's propellant, which is why it's designed to be so light.

I will go as far as to say that since the ship and booster are made in the same megabay from a manufacturing standpoint it would make sense to make the ship as big as the booster, you would have the same advantage as an SSTO, which is that you are launching a whole rocket to space and that's insanely useful if it doesn't end up as space trash.

[Sarigolepas]

I created a spreadsheet to estimate  the trajectory of Booster + Starship, calculate the fuel used, etc.

The ship with 6 vacuum engines will be 10 meters taller and have more fuel, so I'm expecting it to need more delta-v to reach orbit and to have more gravity losses because of how early it will separate from the booster.
My guess is slightly heavier but 50% more thrust so overall slightly more TWR just like for the booster.

[whitelancer64]

The CRS10 had a horizontal velocity at apogee of about 480m/sec.  So I increased the value to about 460km/sec.

*snip*

460 km/s is over 15 times faster than the Earth orbits the Sun. It's nearly the escape velocity for the Galaxy.

[InterestedEngineer]

The CRS10 had a horizontal velocity at apogee of about 480m/sec.  So I increased the value to about 460km/sec.

*snip*

460 km/s is over 15 times faster than the Earth orbits the Sun. It's nearly the escape velocity for the Galaxy.

What's 3 orders of magnitude amongst typos?

[InterestedEngineer]

I created a spreadsheet to estimate  the trajectory of Booster + Starship, calculate the fuel used, etc.

The ship with 6 vacuum engines will be 10 meters taller and have more fuel, so I'm expecting it to need more delta-v to reach orbit and to have more gravity losses because of how early it will separate from the booster.
My guess is slightly heavier but 50% more thrust so overall slightly more TWR just like for the booster.

copy the spreadsheet and try it.

It's pretty much all driven off of "these the main controls" cells.

TWR of booster being one of them.  It derives the maximum Starship payload from that.


If I increase the booster engines to 250t thrust I get 180t of payload on Starship and two extra rings (3.64m) and booster one extra ring (1.82m)

If I do that AND drop the TWR I 200t on payload on 3 extra rings (5.5m).  Additional gravity losses was 200m/sec so probably shouldn't drop TWR.  Also had to add 3 rings to the booster (another 5.5m).

So at least by my model, one adds equal amounts of fuel (and rings at about 100t fuel per ring) to the booster and the starship

[catdlr]

With the recent announcement from Elon of a Version-2 Starship, is this thread the place to continue discussion or is this thread dead since its last post was in May?  What were the takeaways from this discussion on a proposed Version 2?

[tyrred]

This seems be as good a place as any.

I see current rumors of:

Up to 10 meter stretch of StarShip, stretch OLIT as well.

No stretch of SuperHeavy, however something needs to be be done by about staging.

Increase to 6 Raptor Vacuum engines on StarShip, no big surprise there.

Ship forward flaps moving leeward possibly to 120° and forward, or aft, ymmv.

Raptor 3's will be the special sauce.

As per usual, I won't hold my breath for any particular architecture, but watch nevertheless with curiosity.

Ideally, I would love to see them try everything.

But alas, time is of the essence.

And money, of course.

[Asteroza]

Is it safe to say this stretch will get basic SSv2 to the 1500t propellant range? There's noises in the lunar starship threads that 1200t is sorta hurting the CONOPS of the whole LSS, but 1500t gets things down to VLEO refueling simplifying a lot.

[InterestedEngineer]

Is it safe to say this stretch will get basic SSv2 to the 1500t propellant range? There's noises in the lunar starship threads that 1200t is sorta hurting the CONOPS of the whole LSS, but 1500t gets things down to VLEO refueling simplifying a lot.

The whole 1500/1200 CONOPS problem for a fuel depot is a problem that doesn't exist.

For each 1.82m ring there's another ~113t of Methalox capacity.

The cargo section has 5 such rings, therefore a Starship whose only cargo is Methalox has capacity of an additional 565t of Methalox (1750t IOTW) simply by moving headers around.

No stretch needed.

Now there may be other reasons to stretch the Starship, such as wanting 1400-1500t of fuel in addition to cargo.

[TomH]

One can't add 500t to the Starship (Fuel + cargo + rings) and not add fuel to the booster, the rocket equation doesn't work like that.

The current deltaV of the booster is about 3.5km/sec.   That's a mass ratio (M_r) of 2.7

If you add 500t of payload to the booster (aka fuel and payload and rings for Starship), to get the same deltaV, the booster needs 850t of fuel -- (M_r-1) times payload increase

There's no place for that fuel to go without adding rings.  Each ring adds 100t of fuel capacity.   So ~8 more rings.

All good points. It is possible to stage lower and slower, although they already are staging relatively low and slow to begin with. Moving on to Raptors (V2.1 or V3, whatever the nomenclature) with higher thrust would help compensate for reduced T/W early in the profile, although they will burn through the prop faster. Also, waiting longer before throttle down would partially help compensate for the increased gravity losses.

[ZachF]

The big reason for switching to 6 RVacs is IMHO more so that they can increase expansion ratios (and thus ISP) rather than just increasing thrust. I wouldn’t be surprised if the future RVac actually has *less* thrust and from pinching the throat to increase ER.

Another thing I wouldn’t be surprised to see is an adjustable mix ratio. Going slightly more fuel rich as the burn progresses to increase ISP.

[EstorilM]

Considering Elon continues to push the Raptor design (makes me nervous) yet 2.0 performed more-or-less flawlessly, I have faith that even a minimal increase in thrust on subsequent (and certainly real missions) will offset the weight added by Starship 2.0 - at least when you multiply the additional thrust x33.

After staging, Starship is on its own with all that additional mass though - we already know 3 sea-level raptors is more than enough to land it, even with 2.0. IMO the 6 r-vac's are needed to achieve the necessary delta-v for various flight profiles, not just LEO.

A fully-fueled tanker Starship 2.0 may have been cutting it close with delta-v in certain situations previously.

Another consideration that Elon and the engineers are probably looking at is engine-out performance and control. So far, we've seen that asymmetrical thrust and the subsequent offset gimbaling doesn't seem to be enough in many scenarios.

When you're looking at ~16+ flights to refuel a tanker Starship for a deep space mission, reliability is absolutely everything. I don't think their current 3-engine r-vac layout would allow for a SINGLE engine failure on ascent. The r-vac position (far outside center of vehicle) and only having 3 of them, just seems like a nightmare in an engine-out scenario.

Then again the Shuttle did just fine on STS-51F with one of three engines out, so who knows. With the tank attached and the huge mass of the Shuttle itself, perhaps asymmetrical thrust wasn't as much of an issue (SSME's have decent gimbal authority as well.) Then again the r-vac's don't gimbal, so again - it would likely impact its ascent performance significantly as the sea-level raptors tried to correct.

In contrast, losing 1-of-6 with a slight gimbal correction and a longer burn time is almost negligible - that'll work every time.

Perhaps that was a significant factor in their design choice? 

One other quick thought is that 6 r-vac's also allows for hot-stage flexibility - ie. separation only using perhaps 4 r-vacs (MUCH more gentle than 6 engines on V1, but enough to offset the booster thrust - even with the ship's fuel mass included) NONE of which impinge directly onto the center of the hot stage ring. Light the sea-level raptors and remaining r-vac's after booster flip is complete and it's clear of the Starship thrust zone.