Author Topic: SpaceX Starship/Super Heavy Engineering General Thread 3  (Read 348227 times)

Offline Ace

  • Full Member
  • **
  • Posts: 215
    • Going to Mars with SpaceX
  • Liked: 175
  • Likes Given: 86
Re: SpaceX Starship/Super Heavy Engineering General Thread 3
« Reply #60 on: 08/26/2019 12:47 am »
The 85t dry weight is probably from dear moon or another musk presentation.
1100t propellant the same.
engine numbers are recent from raptor testing at mcgregor.

I just re-watched Dear Moon; they aren't from that one.

It's from IAC 2017 presentation.

Thanks.

That was before the switch to stainless. A couple of those numbers are known to have changed:

Ship Length 48m --> now 55m
Body Diameter 9m --> unchanged
Ship Dry Mass 85t --> ??
Propellant Mass 1100t --> ??
Max Ascent Payload 150t --> now 100t
Typical Return Payload 50t --> ??

My assumption is that the increase in length was primarily to provide room for more fuel. However, they've also increased the payload volume. Musk said he thought they'd end up with around 1100 m3.

I'm interested in developing a more accurate estimate / understanding of how many launches will be needed to completely refuel Starship. The oft-stated "5" seems ... optimistic.

Offline ZChris13

That was before the switch to stainless. A couple of those numbers are known to have changed:

Ship Length 48m --> now 55m
Body Diameter 9m --> unchanged
Ship Dry Mass 85t --> ??
Propellant Mass 1100t --> ??
Max Ascent Payload 150t --> now 100t
Typical Return Payload 50t --> ??
It would be more accurate to say that the Max Ascent Payload has changed from 150t to 100t+?, which was the last value given. I'm speculating that they hadn't nailed down the dry weight and weren't confident to more than one significant figure.
edited for clarity
« Last Edit: 08/26/2019 02:18 am by ZChris13 »

Offline Ace

  • Full Member
  • **
  • Posts: 215
    • Going to Mars with SpaceX
  • Liked: 175
  • Likes Given: 86
Re: SpaceX Starship/Super Heavy Engineering General Thread 3
« Reply #62 on: 08/26/2019 03:38 am »
Given the above numbers, we can make a few estimates:

85 t @ 48 m --> 1771 kg/m
48 m vs. 55 m --> 15% longer

Estimating that mass scales linearly with length:
85 t dry mass + 15% = 97.4 t

Assuming stainless is a little lighter, say 5%:
97.4 - 5% = 92.5 t dry mass using stainless

Next, on the rocket equation side:

Given deltaV: 6900 m/s
Exhaust velocity (Ve) = Isp * g = 380 sec * 9.81 m/s^2 = 3728 m/s
Required Full to Empty mass ratio = e^(deltaV / Ve) = 6.37

First, solve for fuel mass using the estimated dry mass and the given cargo mass:

92.5 t dry mass + 100 t cargo = 192.5 t dry mass with cargo
Fuel mass = (Mass ratio * Dry mass with cargo) - Dry mass with cargo
(6.37 * 192.5 t) - 192.5 t = 1033 t fuel mass

We can also fix fuel mass at, say, 1100 t, and solve for cargo mass:

Cargo mass = (Fuel mass - ((Mass ratio - 1) * Dry mass without cargo)) / (Mass ratio - 1)
(1100 - 5.37*92.5) / 5.37 = 112 t cargo mass

The latter direction may make more sense; I would be surprised if the longer Starship structure didn't allow for at least as much fuel as the shorter one did.

If we assume Starship arrives in LEO with a 7% fuel reserve for Earth EDL, and use the second set of numbers above for cargo and fuel mass:

1100 t * 93% = 1023 t refueling requirement

Assuming the entire cargo mass can be used for fuel in a refueling tanker:
1023 t / 112 t cargo per trip = 9.1 trips to completely re-fuel

Getting to Mars from LEO using a "straight-in" aerobraking-type approach only takes roughly 4.8 km/s dV, which means the total fuel requirements will be less for the nominal payload -- although of course that also means more payload could be delivered given the maximum fuel load, so it may not affect the number of refueling trips, except perhaps in early missions where cost may be a larger factor.

Offline Twark_Main

  • Senior Member
  • *****
  • Posts: 2301
  • Technically, we ALL live in space...
  • Liked: 1341
  • Likes Given: 885
Re: SpaceX Starship/Super Heavy Engineering General Thread 3
« Reply #63 on: 08/26/2019 04:15 am »
That was before the switch to stainless. A couple of those numbers are known to have changed:

Ship Length 48m --> now 55m
Body Diameter 9m --> unchanged
Ship Dry Mass 85t --> ??
Propellant Mass 1100t --> ??
Max Ascent Payload 150t --> now 100t
Typical Return Payload 50t --> ??
It would be more accurate to say that the Max Ascent Payload has changed from 150t to 100t+?, which was the last value given. I'm speculating that they hadn't nailed down the dry weight and weren't confident to more than one significant figure.
edited for clarity

Musk clarified on twitter that 150 tonnes was the payload to a "reference" low Earth orbit that other providers use to calculate their payload capacity (200x200 km). 100-125 tonnes was the payload to a useful orbit, eg the Starlink deployment orbit.

https://twitter.com/elonmusk/status/1149571338748616704
"The search for a universal design which suits all sites, people, and situations is obviously impossible. What is possible is well designed examples of the application of universal principles." ~~ David Holmgren

Offline Twark_Main

  • Senior Member
  • *****
  • Posts: 2301
  • Technically, we ALL live in space...
  • Liked: 1341
  • Likes Given: 885
Re: SpaceX Starship/Super Heavy Engineering General Thread 3
« Reply #64 on: 08/26/2019 04:20 am »
Given the above numbers, we can make a few estimates:

Snip

If we assume Starship arrives in LEO with a 7% fuel reserve for Earth EDL

That was the number quoted for Super Heavy, not Starship. Iirc video analysis of the DearMoon Earth EDL simulation showed the landing used ~8 tonnes (someone please correct me here), which is far from the 77 tonnes assumed above.
"The search for a universal design which suits all sites, people, and situations is obviously impossible. What is possible is well designed examples of the application of universal principles." ~~ David Holmgren

Offline Ace

  • Full Member
  • **
  • Posts: 215
    • Going to Mars with SpaceX
  • Liked: 175
  • Likes Given: 86
Re: SpaceX Starship/Super Heavy Engineering General Thread 3
« Reply #65 on: 08/26/2019 04:40 am »
Musk clarified on twitter that 150 tonnes was the payload to a "reference" low Earth orbit that other providers use to calculate their payload capacity (200x200 km). 100-125 tonnes was the payload to a useful orbit, eg the Starlink deployment orbit.

I think "true useful load to useful orbit" is consistent with the sense of the term "payload" as I've used it.

Once we have better numbers for Super Heavy, the payload, fuel mass and refueling estimates could be refined quite a bit (what it takes to get Starship to LEO depends on how much dV SH can provide before staging).

Offline Ace

  • Full Member
  • **
  • Posts: 215
    • Going to Mars with SpaceX
  • Liked: 175
  • Likes Given: 86
Re: SpaceX Starship/Super Heavy Engineering General Thread 3
« Reply #66 on: 08/26/2019 05:03 am »
Given the above numbers, we can make a few estimates:

Snip

If we assume Starship arrives in LEO with a 7% fuel reserve for Earth EDL

That was the number quoted for Super Heavy, not Starship. Iirc video analysis of the DearMoon Earth EDL simulation showed the landing used ~8 tonnes (someone please correct me here), which is far from the 77 tonnes assumed above.

It makes sense that the needed reserve could be less than SH, given that Starship should be able to bleed off much more energy through aerobraking than SH can. Worst case, if they need to use 100% of the onboard fuel to get to orbit, 1100 / 112 = 9.8 launches, or roughly 1 more launch than the 7% reserve case. I'm actually hoping Super Heavy will provide enough dV that SS will be able to use a relatively low parking orbit for refueling, allowing it to retain a larger fraction of its total fuel (current guess: ~250 km orbit, which would take ~9 km/s total dV).

What I saw on the Dear Moon "true physics" simulation was that the engines fired for ~16 seconds just before landing. One could estimate fuel use during that time, but of course SS mass and engine layout have changed since then, too, which would affect the numbers.

If anyone knows of recent / more accurate estimates of Earth EDL fuel reserve requirements for Starship, I would appreciate a pointer.

EDIT: From another thread here, estimated fuel flow rate is 525 kg/s per engine. Assuming only the middle 3 engines with SL nozzles are used in the landing, and that they're always firing at full thrust, that would be 525 kg/s * 3 engines * 16 sec = 25200 kg = 25.2 t. There would very likely be some margin on top that.

« Last Edit: 08/26/2019 09:23 am by Ace »

Offline Ace

  • Full Member
  • **
  • Posts: 215
    • Going to Mars with SpaceX
  • Liked: 175
  • Likes Given: 86
Re: SpaceX Starship/Super Heavy Engineering General Thread 3
« Reply #67 on: 08/26/2019 05:19 am »
Getting to Mars from LEO using a "straight-in" aerobraking-type approach only takes roughly 4.8 km/s dV, which means the total fuel requirements will be less for the nominal payload -- although of course that also means more payload could be delivered given the maximum fuel load, so it may not affect the number of refueling trips, except perhaps in early missions where cost may be a larger factor.

Quick follow-up on the point above:

Changing dV from max achievable of 6.9 km/s to the ~4.8 km/s required to reach Mars, while keeping payload at 100 t, results in a fuel requirement of 505.2 t. Ignoring Earth EDL reserve requirements for the moment:

505.2 t / 100 t per refueling ship = 5.05 launches

Which is consistent with the number "5" that's often tossed around....

4.8 km/s is only an estimate, of course. The exact dV requirement (and therefore required fuel mass) varies by both the biennial launch window and the specific date chosen within that window, so it's good to have plenty of margin.

Offline RobLynn

  • Full Member
  • ****
  • Posts: 695
  • Per Molestias Eruditio
  • NZ
  • Liked: 482
  • Likes Given: 213
Re: SpaceX Starship/Super Heavy Engineering General Thread 3
« Reply #68 on: 08/26/2019 08:59 am »
Musk has tweeted about sticking up to 10 engines on SS for point to point.  With 7x250tonne booster raptors and 3x 200tonne [edit oops, 200 tonne not 300] SL gimballing raptors that would be about 2350tonnes thrust, and should allow filling entire volume of SS to create a tanker version with something like another 30tonnes fuel payload above and beyond the normal 6 engine SS (assuming maintain close to 3% payload fraction).  Bonus staging at lower velocities for easier SH return to launch site.

Any reason not to do this given they will already have a balanced re-entry aero-solution for the heavier tail end of the point to point SS? 
« Last Edit: 08/26/2019 06:17 pm by RobLynn »
The glass is neither half full nor half empty, it's just twice as big as it needs to be.

Offline speedevil

  • Senior Member
  • *****
  • Posts: 4406
  • Fife
  • Liked: 2761
  • Likes Given: 3369
Re: SpaceX Starship/Super Heavy Engineering General Thread 3
« Reply #69 on: 08/26/2019 10:24 am »
Given the above numbers, we can make a few estimates:

I'd suggest 'can we' would be the more appropriate order, and the answer is probably not.

Pretending that there was ever one unchanging design and overinterpreting every number ever released about it unquestioningly is not a productive task.

For example - you can't square the "100+ tons" to orbit as meaning much else than 150 tons, if you consider the required delta-v for a lunar flyby and still believe the 85 ton number, and think the lack of a tanker in the Dear Moon presentation was intentional.

This is especially true for trying to do calculations with numbers derived from multiple epochs.

Fortunately, the architecture often makes it reasonable to not care too much.

If taking 9, not 7 trips to retank is a major deal-breaking issue, things have gone horribly wrong.

Even if the tanker is required to propulsively brake and return, an extra 500m/s can be gotten if your SS won't quite make whatever delta-v you want with an extra tanker flight.

It matters if you are costing things like mass transit passenger flight affordability to the general public - $10K vs $20K to the moon matters.

It does not matter a tiny bit, if you are comparing with (for example) Artemis, at $30B, for ~12 tons over 4 landings to the lunar surface.
Arguably, 30 nominal 150 ton launches to LEO gets you 4 landings of 100 tons each, for of the order of 1% of the Artemis budget.

Another fun cap on pricing is looking at the cars at Cocoa and BC, multiplying by $100K/year.

This comes out to of the order of $20M/year.

It doesn't seem a huge stretch to imagine from the progress that we are seeing that the shell of a SS with no aerofeatures could come in at under $4M, plus $1M for engines if the '500/year' production line comes to life.

It then becomes reasonable to imagine that even a fully expended SS+SH, with 300 tons to orbit, might cost under $30M.

We will have a better handle on this when we see how long it takes to outfit the prototypes.

Pretending that counting the numbers of rockets on a slide from 2017 means much now, when the whole costings of the project have been comedically revamped makes as much sense at this point as imagining SLS is going to fly 4 times a year.

Offline meekGee

  • Senior Member
  • *****
  • Posts: 12460
  • N. California
  • Liked: 11350
  • Likes Given: 1288
Re: SpaceX Starship/Super Heavy Engineering General Thread 3
« Reply #70 on: 08/26/2019 12:21 pm »
Given the above numbers, we can make a few estimates:

Snip

If we assume Starship arrives in LEO with a 7% fuel reserve for Earth EDL

That was the number quoted for Super Heavy, not Starship. Iirc video analysis of the DearMoon Earth EDL simulation showed the landing used ~8 tonnes (someone please correct me here), which is far from the 77 tonnes assumed above.

It makes sense that the needed reserve could be less than SH, given that Starship should be able to bleed off much more energy through aerobraking than SH can. Worst case, if they need to use 100% of the onboard fuel to get to orbit, 1100 / 112 = 9.8 launches, or roughly 1 more launch than the 7% reserve case. I'm actually hoping Super Heavy will provide enough dV that SS will be able to use a relatively low parking orbit for refueling, allowing it to retain a larger fraction of its total fuel (current guess: ~250 km orbit, which would take ~9 km/s total dV).

What I saw on the Dear Moon "true physics" simulation was that the engines fired for ~16 seconds just before landing. One could estimate fuel use during that time, but of course SS mass and engine layout have changed since then, too, which would affect the numbers.

If anyone knows of recent / more accurate estimates of Earth EDL fuel reserve requirements for Starship, I would appreciate a pointer.

EDIT: From another thread here, estimated fuel flow rate is 525 kg/s per engine. Assuming only the middle 3 engines with SL nozzles are used in the landing, and that they're always firing at full thrust, that would be 525 kg/s * 3 engines * 16 sec = 25200 kg = 25.2 t. There would very likely be some margin on top that.
If the fueling orbit is lower, you can get more fuel in per refueling run, but then burn that fuel otw to Mars.

If you're using a final refueling run in HEO, then I think it all cancels out.
ABCD - Always Be Counting Down

Offline speedevil

  • Senior Member
  • *****
  • Posts: 4406
  • Fife
  • Liked: 2761
  • Likes Given: 3369
Re: SpaceX Starship/Super Heavy Engineering General Thread 3
« Reply #71 on: 08/26/2019 12:44 pm »
EDIT: From another thread here, estimated fuel flow rate is 525 kg/s per engine. Assuming only the middle 3 engines with SL nozzles are used in the landing, and that they're always firing at full thrust, that would be 525 kg/s * 3 engines * 16 sec = 25200 kg = 25.2 t. There would very likely be some margin on top that.
This is a large overestimate.
From memory, the landing burn begins at ~100m/s and ~1km up or so, and lasts around 20 seconds.
This puts the maximum deceleration at ~5m/s^2, meaning that for it to hit the ground at 0m/s with three engines firing would put the required mass at somewhere over 300 tons.

For landing empty at least, one engine is completely reasonable, which puts it at closer to 7 tons.

Offline Negan

  • Full Member
  • ****
  • Posts: 675
  • Southwest
  • Liked: 197
  • Likes Given: 502
Re: SpaceX Starship/Super Heavy Engineering General Thread 3
« Reply #72 on: 08/26/2019 10:52 pm »
How will SS crewed pressure vessel section manufacturing compare to Dragon's pressure vessel manufacturing?

Offline speedevil

  • Senior Member
  • *****
  • Posts: 4406
  • Fife
  • Liked: 2761
  • Likes Given: 3369
Re: SpaceX Starship/Super Heavy Engineering General Thread 3
« Reply #73 on: 08/26/2019 11:16 pm »
How will SS crewed pressure vessel section manufacturing compare to Dragon's pressure vessel manufacturing?
Potentially much, much simpler.
Continuing the tank up at the same thickness gives you a pressure hull which can do ~3+ bar, for a really good margin of safety at 1 bar.
It will need several interior layers of insulation - a first against the skin rather high temperature, backed by some low temperature insulation - possibly with some thermal mass in there. It's got to cope with a skin temp of several hundred C on the outside for a minute or ten. This pretty much means it will inherently mitigate normal temperature swings from space/sun.

But this is all against a mostly 'flat' surface that is easily worked on, and is trivially fixable to. You can literally pop on a stud good for 100kg or so force in 5 seconds pretty much anywhere on the surface.
(Stud welding)


Offline Ace

  • Full Member
  • **
  • Posts: 215
    • Going to Mars with SpaceX
  • Liked: 175
  • Likes Given: 86
Re: SpaceX Starship/Super Heavy Engineering General Thread 3
« Reply #74 on: 08/27/2019 12:03 am »
Given the above numbers, we can make a few estimates:
I'd suggest 'can we' would be the more appropriate order, and the answer is probably not.

Specifications have deterministic implications when it comes to rocket science. If you're saying estimates are only as good as the inputs, which may not be very accurate at this stage, I agree.

When SpaceX says "this is what we're planning to build," it's interesting to me to look at what that directly implies in terms of various downstream aspects, even knowing that the given specs are subject to change.

Offline Ace

  • Full Member
  • **
  • Posts: 215
    • Going to Mars with SpaceX
  • Liked: 175
  • Likes Given: 86
Re: SpaceX Starship/Super Heavy Engineering General Thread 3
« Reply #75 on: 08/27/2019 01:03 am »
EDIT: From another thread here, estimated fuel flow rate is 525 kg/s per engine. Assuming only the middle 3 engines with SL nozzles are used in the landing, and that they're always firing at full thrust, that would be 525 kg/s * 3 engines * 16 sec = 25200 kg = 25.2 t. There would very likely be some margin on top that.
This is a large overestimate.
From memory, the landing burn begins at ~100m/s and ~1km up or so, and lasts around 20 seconds.
This puts the maximum deceleration at ~5m/s^2, meaning that for it to hit the ground at 0m/s with three engines firing would put the required mass at somewhere over 300 tons.

For landing empty at least, one engine is completely reasonable, which puts it at closer to 7 tons.

Thanks for the refinement.

Good point that empty vs. full ships will need different fuel reserves. Cargo or passenger ships would be full on landing, while tankers would be empty.

Offline Ace

  • Full Member
  • **
  • Posts: 215
    • Going to Mars with SpaceX
  • Liked: 175
  • Likes Given: 86
Re: SpaceX Starship/Super Heavy Engineering General Thread 3
« Reply #76 on: 08/27/2019 01:21 am »
If the fueling orbit is lower, you can get more fuel in per refueling run, but then burn that fuel otw to Mars.

If you're using a final refueling run in HEO, then I think it all cancels out.

The most efficient approach from a total mission delta-V perspective would probably be for the first (cargo / passenger) SH + SS to create as much on-orbit delta-V as it can, and then for tankers to rendezvous in the resulting orbit. How practical that will be in reality is hard to say atm.

Exactly what that first orbit will look like isn't obvious to me. I've been assuming 250 km circular LEO, but I can imagine other orbits that might be better, depending on mission goals and the exact delta-V available with the combined SH + SS, along with other mission-related trade-offs such as orbital period, number of desired tanker launches, and so on.

Offline meekGee

  • Senior Member
  • *****
  • Posts: 12460
  • N. California
  • Liked: 11350
  • Likes Given: 1288
Re: SpaceX Starship/Super Heavy Engineering General Thread 3
« Reply #77 on: 08/27/2019 01:25 am »
If the fueling orbit is lower, you can get more fuel in per refueling run, but then burn that fuel otw to Mars.

If you're using a final refueling run in HEO, then I think it all cancels out.

The most efficient approach from a total mission delta-V perspective would probably be for the first (cargo / passenger) SH + SS to create as much on-orbit delta-V as it can, and then for tankers to rendezvous in the resulting orbit. How practical that will be in reality is hard to say atm.

Exactly what that first orbit will look like isn't obvious to me. I've been assuming 250 km circular LEO, but I can imagine other orbits that might be better, depending on mission goals and the exact delta-V available with the combined SH + SS, along with other mission-related trade-offs such as orbital period, number of desired tanker launches, and so on.
LEO refuelling is advantageous from a timing perspective for getting the tankers back asap.

Just one of many considerations..

At the end, HEO or L2 top-off, and then Oberth to Mars - I think everyone agrees on that.

(The top-off may be quite substantial btw.  Probably best, again from logistical reasons, to limit it to one tanker's worth)

ABCD - Always Be Counting Down

Offline Keldor

  • Full Member
  • ****
  • Posts: 670
  • Colorado
  • Liked: 789
  • Likes Given: 106
Re: SpaceX Starship/Super Heavy Engineering General Thread 3
« Reply #78 on: 08/27/2019 01:32 am »
If the fueling orbit is lower, you can get more fuel in per refueling run, but then burn that fuel otw to Mars.

If you're using a final refueling run in HEO, then I think it all cancels out.

The most efficient approach from a total mission delta-V perspective would probably be for the first (cargo / passenger) SH + SS to create as much on-orbit delta-V as it can, and then for tankers to rendezvous in the resulting orbit. How practical that will be in reality is hard to say atm.

Exactly what that first orbit will look like isn't obvious to me. I've been assuming 250 km circular LEO, but I can imagine other orbits that might be better, depending on mission goals and the exact delta-V available with the combined SH + SS, along with other mission-related trade-offs such as orbital period, number of desired tanker launches, and so on.

In order to minimize the number of refuelling flights, you want to refuel in LEO, and specifically, as low as you can reasonably go.  This is because the lower you are, the less work you have to do dragging the tankers themselves out to the refueling point.

Obviously, you still need to have enough delta-V to get from your refueling orbit to your destination and complete your mission.  That's why Starship has to refuel in a high elliptical orbit for lunar missions.

So, refuelling in the lowest energy orbit available that still gives Starship the ability to do the mission once it's fuelled up.
« Last Edit: 08/27/2019 01:37 am by Keldor »

Offline Ace

  • Full Member
  • **
  • Posts: 215
    • Going to Mars with SpaceX
  • Liked: 175
  • Likes Given: 86
Re: SpaceX Starship/Super Heavy Engineering General Thread 3
« Reply #79 on: 08/27/2019 01:37 am »
Musk has tweeted about sticking up to 10 engines on SS for point to point.

With 7x250tonne booster raptors and 3x 200tonne [edit oops, 200 tonne not 300] SL gimballing raptors that would be about 2350tonnes thrust, and should allow filling entire volume of SS to create a tanker version with something like another 30tonnes fuel payload above and beyond the normal 6 engine SS (assuming maintain close to 3% payload fraction).  Bonus staging at lower velocities for easier SH return to launch site.

Any reason not to do this given they will already have a balanced re-entry aero-solution for the heavier tail end of the point to point SS?

The only way I can see 10 engines working would be to replace the outer 3 vacuum nozzled-engines with 7 SL-nozzled engines, while retaining the current 3 gimbaled SL engines in the center. Such a vehicle would only be useful relatively low in the atmosphere.

I'm guessing that "point to point" in this context means from one point on the Earth to another -- NY to Japan, for example. I could imagine that a 10 x SL engine SS config might be useful for routes like that, although I haven't run the numbers myself.

However, such a vehicle would not be useful for refueling, since that requires propulsion above the atmosphere, to get into an orbit that allows a rendezvous with the Starship being refueled. Even if it could get to orbit, the total payload would be tiny compared to what's achievable when combined with SH. One way to think about SH is that its main purpose is to lift SS and all of the fuel it contains above the atmosphere.

Tags:
 

Advertisement NovaTech
Advertisement SkyTale Software GmbH
Advertisement Northrop Grumman
Advertisement
Advertisement Brady Kenniston
Advertisement NextSpaceflight
Advertisement Nathan Barker Photography
1