totm nov 2023 SpaceX Discussion Thread

[sevenperforce]

On 6/20/2020 at 12:25 PM, Nightside said:

Just because the ship will return to Earth doesn't mean the crew needs to land with it. I didn't forget, I guess I am assuming that Mars and Earth are just different EDL problems, but you could be right that Mars is just an easier version of Earth (assuming you were going to land on rockets in either case).

How else do you propose people getting back from Mars?

People have to be able to get back from Mars. Thus the ship that goes from Mars to Earth has to be able to land on Earth. Thus the ship that lands on Mars has to be able to land on Earth.

Mars is not just an easier version of Earth, no. Which is why the solution remains complex -- 4 flaperons.

[kerbiloid]

16 minutes ago, sh1pman said:

I'll bookmark this post, let's see how many years it will take them to reach that rate. 

Don't forget to move it into your next browser, every time buying the next device.

[sevenperforce]

On 6/20/2020 at 1:05 PM, kerbiloid said:

 You can use them to deorbit. Delta-V is comparable, LES ~300 m/s, deorbit from ISS ~250 m/s.

They are less efficient, but still better than save some mass but lose the ship.

Just don't turn them on all at once.

That's actually rather challenging. SRBs don't burn with perfect rate consistency like they do in KSP. If you have clustered SRBs then you need gimbal on them to correct for thrust variations. Gimballed SRBs are more failure points than you want on your escape system.

On 6/20/2020 at 2:33 PM, CatastrophicFailure said:

IIRC, Musk said a while back that Starship would have enough thrust itself to abort from Superheavy off the pad, they’d just light the vacuum Raptors, too. There would be a loss of efficiency but that’s not really a concern at that point. I assume the design is such that the engines could tolerate the underexpansion, at least for a time. This would happen in a fraction of a second, too. 

The vacuum engines will be dual-bell so they have controlled flow separation -- the problem is overexpansion, not underexpansion.

But even with vacuum Raptors running, it still doesn't push enough thrust for a healthy abort.

[kerbiloid]

1 minute ago, sevenperforce said:

That's actually rather challenging. SRBs don't burn with perfect rate consistency like they do in KSP. If you have clustered SRBs then you need gimbal on them to correct for thrust variations. Gimballed SRBs are more failure points than you want on your escape system.

When the LES is powered with several fuel charges, same nozzles can be used without clusterization and provide several kilometers accuracy.
So, if the flight is normal, they can adjust their landing point with the winglets. And if the flight is abnormal, the accuracy plays no role.

[wumpus]

12 hours ago, SOXBLOX said:

Sure. Who's paying for it, though? SLS is government funded. As SpaceX fans are so fond of pointing out, Starship isn't.

As of April, NASA is paying for Moonship (or whatever they call the Starship that lands on the Moon) development.  Presumably construction, testing,  and plenty of launches if the program lives that long (I suspect 1-2 landers are expected to be culled early, and Starship looks like a comically overpowered lander).

Don't forget that Falcon 9 is sucking up most of the money spent on commercial space launches (US DoD still loves its military-industrial complex partners, presumably now including Northrup-Grumman, Russia and China also like their own rockets for both military and commercial use).  That is a lot of money, and Spacex isn't paying a lot of money to develop Falcon 9 anymore.

1 minute ago, sevenperforce said:

That's actually rather challenging. SRBs don't burn with perfect rate consistency like they do in KSP. If you have clustered SRBs then you need gimbal on them to correct for thrust variations. Gimballed SRBs are more failure points than you want on your escape system.

I'd like to believe that if your SRBs are supplementing a liquid rocket, then the liquid rocket can provide thrust variation (either by gimballing or with multiple engines selectively throttling).  At some point, that's probably not enough, especially if your SRBs are providing nearly all the thrust and are positioned further from the center of mass (such as the Shuttle).

[sevenperforce]

12 hours ago, CatastrophicFailure said:

Well, if they’re ever going to do P2P, Starship will have to have a TWR greater than 1.

P2P is still planned to use Superheavy.

I think that if they were doing P2P they should just pack Starship with sea level engines. Slightly reduces the available landing sites but you can still reach almost anywhere in the same timeframe.

12 hours ago, CatastrophicFailure said:

IIRC in normal operation Starship would only use the vacuum engines during ascent, the sea level engines would be used for landing. 

No, Starship uses the sea level engines as well immediately after staging, both because you otherwise suffer from gravity losses due to low thrust and because the vac engines don't gimbal. The sea level engines shut off midway through Starship's orbital insertion burn and differential thrust is used thereafter.

[magnemoe]

On 6/20/2020 at 7:19 PM, tater said:

Seems like the failure mode of a really bad reentry is already LOM/LOC. We have to assume some baseline reentry profile, then a failure. So a TPS failure, skydiver failure (like the hydraulics on that offshore F9 booster landing), etc. At that point if the entry was survivable (likely a g-load issue on the squishy bits inside), then they can eject the pod instead of landing the wrecked SS. For an ascent abort, I'd imagine the critical bit is aerodynamic forces and g on the crew.

One major problem during accent is that Superheavy is RTLS so launch angle is steep, the ballistic trajectory will give an very high g load during an ballistic reentry. 
As I understand Starship will use its surface engine a short time after separation but they will be shut down probably one by one. 

Note that the steep angle is only an problem for an escape pod not for starship as it has engines and will do its own RTLS, it however will have to burn off most if its fuel and oxidizer before landing, say superheavy fails shortly after takeoff you will you have the curious sight of starship hovering to burn off fuel. 
If it looses one of its vacuum engines the atmospheric can take over but they has lower isp, an fail shortly after separation and its an RTLS, deeper into the burn you have two options, cross the Atlantic and land in Africa / Europe, one other option is to abort to orbit and refuel so you can reach objective or land. 

This is also true for the cargo version. 

[RCgothic]

Superheavy pad is under construction. I think they may be doing a raised pad because the ground is too soft for an underground flame diverter.

 

[CatastrophicFailure]

2 hours ago, sh1pman said:

I'll bookmark this post, let's see how many years it will take them to reach that rate. 

And that would be an average, too. I would expect rapid-fire flights every eight hours, or something like that, when they’re building up fuel for a lunar or mars trip. What’s it need, like a dozen tankers in fairly short order? Outside that it could be a more reasonable “launch every few days.” 

[RealKerbal3x]

8 minutes ago, RCgothic said:

Superheavy pad is under construction. I think they may be doing a raised pad because the ground is too soft for an underground flame diverter.

[snip]

I wonder how big it'll actually be. That space doesn't look big enough for an LC-39 sized pad, so maybe it'll be closer to the pad shown in the launch animation.

[tater]

18 minutes ago, CatastrophicFailure said:

And that would be an average, too. I would expect rapid-fire flights every eight hours, or something like that, when they’re building up fuel for a lunar or mars trip. What’s it need, like a dozen tankers in fairly short order? Outside that it could be a more reasonable “launch every few days.” 

Bring back the bookmark after a flight every 36 hours for a solid year.

Yeah, lunar SS will require a bunch of retanks, but those will be entirely driven by phasing concerns. If a dogleg or other rendezvous prop expenditure exceeds boiloff until the next flyover, they wait for the next flyover. Regardless, that's a dozen flights, so only 258 more to go at the same (or higher) cadence to get 270 in a year.

I'm pro-SS (because who wouldn't want it to work?). I think if costs are low per launch then they can replace F9 flights with SS 100%. Even with Starlink they don't have a need for near daily flights (not to mention they then need to make Starlinks at literally hundreds per day). Testing is one thing, and using their own payloads to offset testing costs is great. But if the goal it to test fly to demonstrate crew safety... yeah, that;s gonna take a while.

[Nightside]

3 hours ago, sevenperforce said:

How else do you propose people getting back from Mars?

People have to be able to get back from Mars. Thus the ship that goes from Mars to Earth has to be able to land on Earth. Thus the ship that lands on Mars has to be able to land on Earth.

Mars is not just an easier version of Earth, no. Which is why the solution remains complex -- 4 flaperons.

Well, another crew transfer in LEO.

Obviously this means the returning ship would have to circularize in LEO rather land directly, which would be less efficient, and certainly adds mission complexity, which carries its own risks. 

The upside is that the crew will land on a recently maintained ship, purpose built for Earth EDL rather than having them land on a ship that has been blasted by Martian dust and deep space radiation for several years. Besides, I'm still not convinced that Mars EDL is just an easier version of Earth.

If Starship launches are as cheap as planned, then having a few extra launches won't be a problem.

Listen I'm all for getting a Tintin rocket that can go everywhere, but I think that will be a much larger engineering challenge - and gets into "a camel is a horse designed by a committee" territory where you end up with something is mediocre at everything (Nothing against camels! They are great at being camels). 

15 minutes ago, tater said:

I'm pro-StarShip (because who wouldn't want it to work?). 

...Astronomers. But this time they have no excuse for not taking SpaceX seriously on their plans.

Edited June 22, 2020 by Nightside

[tater]

1 minute ago, Nightside said:

...Astronomers

That's Starlink

 

[Nightside]

2 minutes ago, tater said:

That's Starlink

 

Starlink's are tiny with a small reflective solar panel. How much light will a fleet of massive shiny ships constantly going up and down give off?

[RCgothic]

9 minutes ago, Nightside said:

Well, another crew transfer in LEO.

Obviously this means the returning ship would have to circularize in LEO rather land directly, which would be less efficient, and certainly adds mission complexity, which carries its own risks. 

The upside is that the crew will land on a recently maintained ship, purpose built for Earth EDL rather than having them land on a ship that has been blasted by Martian dust and deep space radiation for several years. Besides, I'm still not convinced that Mars EDL is just an easier version of Earth.

If Starship launches are as cheap as planned, then having a few extra launches won't be a problem.

Listen I'm all for getting a Tintin rocket that can go everywhere, but I think that will be a much larger engineering challenge - and gets into "a camel is a horse designed by a committee" territory where you end up with something is mediocre at everything (Nothing against camels! They are great at being camels). 

...Astronomers. But this time they have no excuse for not taking SpaceX seriously on their plans.

Pretty sure Mars return requires aerobraking as there isn't enough DV to do SSFMTLEO.

That requires functioning aero surfaces, so you don't eliminate the main point of failure of a crewed landing that way.

[sevenperforce]

3 minutes ago, Nightside said:

Starlink's are tiny with a small reflective solar panel. How much light will a fleet of massive shiny ships constantly going up and down give off?

Not enough to matter because they are almost all in a single orbital plane -- which will incidentally be the same as the orbital plane of the ecliptic, which is worst for astronomy anyway.

9 minutes ago, Nightside said:

Well, another crew transfer in LEO.

Obviously this means the returning ship would have to circularize in LEO rather land directly, which would be less efficient, and certainly adds mission complexity, which carries its own risks. 

The upside is that the crew will land on a recently maintained ship, purpose built for Earth EDL rather than having them land on a ship that has been blasted by Martian dust and deep space radiation for several years. Besides, I'm still not convinced that Mars EDL is just an easier version of Earth.

Listen I'm all for getting a Tintin rocket that can go everywhere....

Propulsive circularization in LEO is a nonstarter. You can go single-stage from Mars to Earth entry interface, but you cannot go single-stage from Mars to propulsive entry in LEO.

I don't think this is driven by a "Tintin rocket" approach. They have already shown they are willing to do a lunar redesign. This is driven by the need for these specific vehicles to land on both Mars and Earth.

I suppose you could do aerocapture in LEO with your Mars vehicle. 

No one is suggesting Mars EDL is easier than Earth. 

[tater]

22 minutes ago, Nightside said:

Starlink's are tiny with a small reflective solar panel. How much light will a fleet of massive shiny ships constantly going up and down give off?

I'll let you know when I see a fleet of Starships.

Honestly, if we ever become truly spacefaring, this is just the new normal, and if you want unfettered sky, leave Earth.

 

[Silavite]

2 hours ago, sevenperforce said:

That's actually rather challenging. SRBs don't burn with perfect rate consistency like they do in KSP. If you have clustered SRBs then you need gimbal on them to correct for thrust variations. Gimballed SRBs are more failure points than you want on your escape system.

Thrust termination ports (blow-out panels) could provide control for the exact amount of impulse necessary.

Spoiler

 

 

[sevenperforce]

2 hours ago, Silavite said:

Thrust termination ports (blow-out panels) could provide control for the exact amount of impulse necessary.

But that won't help you if you have a cluster of motors and one is burning faster than the others.

If you have a central gimbaled liquid engine for control, it's fine.

[Codraroll]

4 hours ago, tater said:

Bring back the bookmark after a flight every 36 hours for a solid year.

Yeah, lunar SS will require a bunch of retanks, but those will be entirely driven by phasing concerns. If a dogleg or other rendezvous prop expenditure exceeds boiloff until the next flyover, they wait for the next flyover. Regardless, that's a dozen flights, so only 258 more to go at the same (or higher) cadence to get 270 in a year.

I'm pro-SS (because who wouldn't want it to work?). I think if costs are low per launch then they can replace F9 flights with SS 100%. Even with Starlink they don't have a need for near daily flights (not to mention they then need to make Starlinks at literally hundreds per day). Testing is one thing, and using their own payloads to offset testing costs is great. But if the goal it to test fly to demonstrate crew safety... yeah, that;s gonna take a while.

I could see a scenario where a few hundred launches per year could be feasible or even necessary. It goes well into speculative fiction, however, and may not be entirely realistic, but entertain the idea for a second:

Imagine if starship is as inexpensive to launch and recover as SpaceX is hoping for. Suddenly, the cost per kg to orbit becomes very low, and the rocket can carry large volumes as well.

While the market is slow at first (as nobody has built any payloads of that size and mass yet - why would they, until now?), someone at NASA is quick to see the implications here and manages to sell a proposal to Congress about using this leg up on the market to cement the US' position in space for the foreseeable future. Plans to build the American Space Station, several times larger than the ISS while also several times cheaper, are approved with a quick schedule for construction. Gotta strike while the iron is hot, after all. It is not a capability the US will have to itself for long.

As said, so done. Starship launches several huge modules to low Earth orbit, where teams of astronauts and robots assemble them together to form the largest space station the world has ever seen. Meanwhile, the Chinese are working hard on their own Starship version and announce grand plans for orbital construction as well.

The US decides to press its advantage while it's still there. A bunch of universities pool their resources and seek the government for support, and before long another space station is under construction. After all, the know-how and manufacturing capabilities for space station modules have been developed, and now that the AmSS is complete, they have to choose whether to dismantle the operation or find new customers. The latter choice is made: Orbital University, a joint project between so many actors there aren't even room for all their logos on the hull. Universities all over the world bid for the opportunity to send somebody there on a two-week stint in their labs. Maybe other stations are built as well, even military ones. Having an astronaut training facility in orbit would be revolutionary for future, longer-distance missions, after all.

For all this, Starship would mostly be used to tug modules and materials into orbit, with the crew launched on other launch vehicles (including SLS, because Congress gotta Congress), but now it takes the role as supply ship as well. Between the various stations, a three-digit number of people are in orbit at the same time, and hotel chains are beginning to contract modules as well. The launch cadence capability of the spaceports might be the limiting factors for how often they can launch at this point. Operating these stations is hellishly more expensive than the ISS ever was, but more value is added per dollar spent than before as well. A whole ecosystem of space companies join in on the fun as well. The mission control operations would quickly outgrow the capacity of Houston Space Center, for instance, so maybe another private company starts offering its services. For every astronaut in space one needs a large group of people on the ground, so this could be a sizable sector of the economy before long. Yet other companies may contract a Starship launch or two to send prospecting probes to promising asteroids. Probes not built at JPL or KSC, but at Harry's Spacecraft Shack in Smalltown, America.

 

If, and it's admittedly a very big and uncertain if, Spaceship makes it feasible to send large payloads into low Earth orbit relatively cheaply and easily, all sorts of organizations would begin to build large payloads. That could trigger a gold rush, as every activity in space needs to be supported by dozens on the ground, and with economies of scale, those activities could be cheaper too. The problem right now is that everything space-related is prohibitively expensive, and so there are very few actors involved with it, and so it remains prohibitively expensive. But if the key operation became cheaper, it might clear the log jam and really propel us into that sci-fi future we've considered right around the corner since, well, Sputnik or so. As they probably said about every innovation in space technology since the early sixties: "If we can pull this one thing off, the road lies open". One day it might come true.

[tater]

 

[kerbiloid]

6 hours ago, sevenperforce said:

But that won't help you if you have a cluster of motors and one is burning faster than the others.

No need in a cluster when you have a single expansion chamber with several radial nozzles and several fuel charges/combustion chambers attached to it, like in LES.
Then you just redistribute pressure between the nozzles.

This is not necessary a LES tower, it can be a LES pad.

[tater]

 

[sevenperforce]

16 hours ago, Codraroll said:

If, and it's admittedly a very big and uncertain if, Spaceship makes it feasible to send large payloads into low Earth orbit relatively cheaply and easily, all sorts of organizations would begin to build large payloads. That could trigger a gold rush, as every activity in space needs to be supported by dozens on the ground, and with economies of scale, those activities could be cheaper too. The problem right now is that everything space-related is prohibitively expensive, and so there are very few actors involved with it, and so it remains prohibitively expensive. But if the key operation became cheaper, it might clear the log jam and really propel us into that sci-fi future we've considered right around the corner since, well, Sputnik or so. As they probably said about every innovation in space technology since the early sixties: "If we can pull this one thing off, the road lies open". One day it might come true

I am all about this and this is a future I would like. But I remain highly skeptical about what crewed space activities are actually going to generate wealth.

 

[Xd the great]

Just now, sevenperforce said:

I am all about this and this is a future I would like. But I remain highly skeptical about what crewed space activities are actually going to generate wealth.

Well, if space tourism costs can be lowered to USD 10000 per 1 week/2 week stay, loads of people will pay for it.

Cha-ching.

Also, mining the Moon. For resources.