SpaceX Starship : Texas Prototype(s) Thread 8 : Discussion

[OTV Booster]

his is one of the reasons I was hoping to see header tanks moving out of the fairing section - to reduce complexity and increase flexibility so that the fairing itself is treated more as the payload or “client” of the tank/engine section.

Exactly if we see the users guide this is clear that lower section is the same for all serial numbers and upper section is build according to the prupose the Starship is build. And Elon said it clearly in his tweet. That realy smart and it's logical with the logic service module/command module or payload or properlant tank module.

It's not just the header tanks, What you are referring to as the "fairing section" also includes forward fins and actuators (and all associated electrics, electronics maybe hydraulics, etc). Unless they radically alter the reentry aerodynamics of SS the fairing section will always be an integral part of the vehicle.

That's not to say that there isn't a possible "expendable mode" of using the recoverable SH and just the tanks and engines of SS in an expendable second stage. That would allow for a customized payload bay probably enclosed in the usual fairing.

As long as each upper section has all the extra goodies there is no reason to limit a change out to an expendable only, IIUC. For that matter, I doubt a tank section could successfully do EDL without an upper section. If a mission were intended to be expendable the whole SS is along for the ride. Which also means that none of the fin system is needed, front or rear.

Because of the early build out and testing we’ve been privileged to watch (thankee BC crew) we’ve become accustomed to thinking of the tank section as the SS and the as yet unmounted upper section as an extension of the tanks. The truth is he other way around. The upper section is the point of the whole exercise, the tanks are a pesky but necessary extension that lets the upper section do it’s job.

Looking at it this way, they will rarely swap out an upper section. They will swap out tanks as necessary. As per Elon, it looks like they’ll have separate numbering. I expect to eventually see a couple unmated tanks at each launch location as running spares.

Operational flexibility is a good thing. To this end we may also see some spare upper sections of different variants waiting to be mated as needed. The crewed module itself will probably have multiple variants and be expensive enough that spares will be rare. All this is way down the line.

Phil

[frim]

Makes me feel less confused regarding nosecones. I suspected this would be the case when they kept working on the one made way back in SN1 days.

Elon on nosecone allocation per Mary's tweet

https://twitter.com/elonmusk/status/1254012962160992262

It seems like nosecone production is a separate team that has been on its own schedule for a while. Maybe they just saw a chance to sneak into the high bay during the SN4 testing lull.

[BZHSpace]

Makes me feel less confused regarding nosecones. I suspected this would be the case when they kept working on the one made way back in SN1 days.

Elon on nosecone allocation per Mary's tweet

https://twitter.com/elonmusk/status/1254012962160992262

It seems like nosecone production is a separate team that has been on its own schedule for a while. Maybe they just saw a chance to sneak into the high bay during the SN4 testing lull.

Each tent will have prupose one for bulkheads one for rings and the last one for nose cone.

[samgineer]

How about hexagonal tiles on nosecone? It is impossible with one kind of tile unless there is some special nosecone geometry I don't know about?

Hex is pretty universal on plain nose, it will be mostly without special tile shapes. Flaps and top of nosey will definitely need more specific tile shapes but if they plan to run mass production it is little issue. For maybe 90-95% of surface is hex more than good shape.

[xvel]

How about hexagonal tiles on nosecone? It is impossible with one kind of tile unless there is some special nosecone geometry I don't know about?

Hex is pretty universal on plain nose, it will be mostly without special tile shapes. Flaps and top of nosey will definitely need more specific tile shapes but if they plan to run mass production it is little issue. For maybe 90-95% of surface is hex more than good shape.

tiling non zero gaussian curvature (which nosecone is) with hexagons IS impossible

[tssp_art]

his is one of the reasons I was hoping to see header tanks moving out of the fairing section - to reduce complexity and increase flexibility so that the fairing itself is treated more as the payload or “client” of the tank/engine section.

Exactly if we see the users guide this is clear that lower section is the same for all serial numbers and upper section is build according to the prupose the Starship is build. And Elon said it clearly in his tweet. That realy smart and it's logical with the logic service module/command module or payload or properlant tank module.

It's not just the header tanks, What you are referring to as the "fairing section" also includes forward fins and actuators (and all associated electrics, electronics maybe hydraulics, etc). Unless they radically alter the reentry aerodynamics of SS the fairing section will always be an integral part of the vehicle.

That's not to say that there isn't a possible "expendable mode" of using the recoverable SH and just the tanks and engines of SS in an expendable second stage. That would allow for a customized payload bay probably enclosed in the usual fairing.

As long as each upper section has all the extra goodies there is no reason to limit a change out to an expendable only, IIUC. For that matter, I doubt a tank section could successfully do EDL without an upper section. If a mission were intended to be expendable the whole SS is along for the ride. Which also means that none of the fin system is needed, front or rear.

Because of the early build out and testing we’ve been privileged to watch (thankee BC crew) we’ve become accustomed to thinking of the tank section as the SS and the as yet unmounted upper section as an extension of the tanks. The truth is he other way around. The upper section is the point of the whole exercise, the tanks are a pesky but necessary extension that lets the upper section do it’s job.

Looking at it this way, they will rarely swap out an upper section. They will swap out tanks as necessary. As per Elon, it looks like they’ll have separate numbering. I expect to eventually see a couple unmated tanks at each launch location as running spares.

Operational flexibility is a good thing. To this end we may also see some spare upper sections of different variants waiting to be mated as needed. The crewed module itself will probably have multiple variants and be expensive enough that spares will be rare. All this is way down the line.

Phil

Actually you and I may be in agreement on this.

"As long as each upper section has all the extra goodies there is no reason to limit a change out to an expendable only"

Well they're not really "goodies"; they are essential. But I think you know that and what you're really saying as long as you've preserved the integrated flight characteristics of SS, including CG, upper fins, heat shield, etc , then it doesn't have to be expendable. I agree.

For that matter, I doubt a tank section could successfully do EDL without an upper section. If a mission were intended to be expendable the whole SS is along for the ride. Which also means that none of the fin system is needed, front or rear.

Of course a tank section can't do an EDL - it is not part of an integrated vehicle, has no known flight characteristics and is unable to provide control through EDL, etc. But if you wanted to do a customized "upper section" with a traditional payload enclosure (maybe an oversized habitat section that needed to be larger than 8M in diameter) it would no longer be capable of EDL but it might still make sense as part of an expendable second stage using the engines and tanks of the "lower section".

My whole point is that doing an EDL requires an integrated flight vehicle that is aerodynamically and structurally capable of surviving the stresses of EDL.

Edit: clarify expendable needs

[docmordrid]

How about hexagonal tiles on nosecone? It is impossible with one kind of tile unless there is some special nosecone geometry I don't know about?

Hex is pretty universal on plain nose, it will be mostly without special tile shapes. Flaps and top of nosey will definitely need more specific tile shapes but if they plan to run mass production it is little issue. For maybe 90-95% of surface is hex more than good shape.

tiling non zero gaussian curvature (which nosecone is) with hexagons IS impossible

Couldn't this be solved by adding pentagon tiles, the soccer ball solution? Heptagon?

[tssp_art]

I have some serious doubts about a 'completely modular' Starship, with a common fuel section and interchangable payload seciton.

While the user's guide seems to imply the payload section will be mated with the booster while on the pad IIRC, I think this procedure only really makes sense for a cargo variant, if at all.

A manned SS is going to be orders of magnitude more complex, with many redundancies in place. As some folks have mentioned, even mating a 'simple' payload section includes mating the header tank lines, and likely electronics, and hydraulics, and this would certainly take things from the realm of reusable to Shuttle-style refurbishment.

However I think the strongest argument comes in the form of the tanker variant. Why have an SS fly a fuel tank in its nose cone, rather than extending the tanks internally to take up all the available volume? SpaceX is already going to be making a stretched tank for the SH, they're already working on making the most of the cone volume by using the tip as a tank wall, so what's stopping them from putting the common bulkhead at the height where the top bulkhead currently sits, and using every CM^3 of volume available for fuel? To me it seems like a simpler solution than trying to make a separate LOX/CH4 tank with its own bulkheads that fits in a nosecone and can be mated to a 'common' thrust section. They'd need to run plumbing from such a tanker payload bay to the rear-mounted refueling adapters on every thrust section for it to be truly modular, meaning each of the other variants would take a mass reduction.

A totally modular payload section seems overcomplicated and inefficient to me. I could see simple cargo SS's to have a removable/interchangable payload section, maybe, if they can figure out how to do the various plumbing/electrical connections reliably, and even that's a pretty big if.

I think it's much more likely that we see some specialized SS's begin development after a prototype makes orbit.

My thinking is not interchangeability like a customer payload, but rather as two fairly independent modules that are built concurrently and then mated on final assembly as required by current mission requirements.

I agree with the concept of two fairly independent modules for the cargo and crew variants. But not for the tanker version as the previous poster (inaccurate_reality) pointed out. I think once the prototypes have completely vetted the (for example) welding processes it is much easier to add a few "rings" to each of the tanks (while removing the same number of rings from the payload area to preserve aerodynamics, CG, etc) than it is to add new tanks to the payload section. Plus looking at the mission plans, it looks like there may be at quite a few tankers needed.

[gaballard]

Per Elon’s comment on the nose cones on the update thread, I wouldn’t doubt if even the production nose cones themselves will have their own serial number series, and will be agnostic as to which tank they will be used on. It sounds like Starship will be a modular system that can have any number of nosecones used based on payload, type or other factors.

Would not surprise me if the nose cone and front payload bay section is a bolt-on joint to the aft section...
The VAB there is not tall enough to weld them together (IMHO)
For now... they will build/weld fronts and backs side by side in the VAB and stack them and add fins outside for now...
All those joints are likely bolted (IMHO)

It’s explicitly stated in the Starship user guide that the payload sections will be modular and will be attached at the pad

[tssp_art]

How about hexagonal tiles on nosecone? It is impossible with one kind of tile unless there is some special nosecone geometry I don't know about?

Hex is pretty universal on plain nose, it will be mostly without special tile shapes. Flaps and top of nosey will definitely need more specific tile shapes but if they plan to run mass production it is little issue. For maybe 90-95% of surface is hex more than good shape.

tiling non zero gaussian curvature (which nosecone is) with hexagons IS impossible

Couldn't this be solved by adding pentagon tiles, the soccer ball solution? Heptagon?

That's quite a clever thought. I have never understood the math that generates custom geodesic dome shapes (even asymmetrical ones) from standardized hexagons and pentagons. But some of the posters here seem quite able mathematicians who might be able to extend your idea.

[OTV Booster]

I believe that SpaceX is already working on SN6 and could have a ring or two already built, and is most likely going to have the new welds that Musk has been talking about. I think production of these are solid right now for the rest of the year, and Raptor production is able to meet this demand due to low amount of raptors being used for each ship. I’d say for every complete SN they are already working on two SNs higher. (Completed SN6 means they are working starting work on SN8)

Right on, but maybe too conservative. The production process is progressing right along with the ship itself. We went from MK1, built as a one off to SN1. IIRC SN2 started shortly before SN1 was tested.

We now have SN4 ready for testing with SN5 building and SN6 starting to build. We’ve gone from a production line one deep to a production line three deep.

By the end of the year we may see a production line five or more deep. We’ve also seen output very slightly increase. There are two processes in tension here. They are getting better and faster in doing what they do but, assuming SN4 is reasonably successful, each build will have a higher part count and be more complex.

Until orbital flight as achieved I’d expect output to stay roughly constant but the pipeline to get deeper. Once orbit is achieved the basic design will be fairly well locked down for the ship with the production line moving in that direction. That’s when production rate should start to pick up. Once the production line is finalized, we and Elon will know how many are in the pipeline at any time and a have a fairly good handle what the potential output rate actually is.

Right now we’re seeing a build rate of ~17/year. If this can be tripled (very reasonable) there will be one a week or 50+/year. Double it again (maybe) and the goal of 100 a year is met.🚀

Phil

[OTV Booster]

I have some serious doubts about a 'completely modular' Starship, with a common fuel section and interchangable payload seciton.

While the user's guide seems to imply the payload section will be mated with the booster while on the pad IIRC, I think this procedure only really makes sense for a cargo variant, if at all.

A manned SS is going to be orders of magnitude more complex, with many redundancies in place. As some folks have mentioned, even mating a 'simple' payload section includes mating the header tank lines, and likely electronics, and hydraulics, and this would certainly take things from the realm of reusable to Shuttle-style refurbishment.

However I think the strongest argument comes in the form of the tanker variant. Why have an SS fly a fuel tank in its nose cone, rather than extending the tanks internally to take up all the available volume? SpaceX is already going to be making a stretched tank for the SH, they're already working on making the most of the cone volume by using the tip as a tank wall, so what's stopping them from putting the common bulkhead at the height where the top bulkhead currently sits, and using every CM^3 of volume available for fuel? To me it seems like a simpler solution than trying to make a separate LOX/CH4 tank with its own bulkheads that fits in a nosecone and can be mated to a 'common' thrust section. They'd need to run plumbing from such a tanker payload bay to the rear-mounted refueling adapters on every thrust section for it to be truly modular, meaning each of the other variants would take a mass reduction.

A totally modular payload section seems overcomplicated and inefficient to me. I could see simple cargo SS's to have a removable/interchangable payload section, maybe, if they can figure out how to do the various plumbing/electrical connections reliably, and even that's a pretty big if.

I think it's much more likely that we see some specialized SS's begin development after a prototype makes orbit.

I agree on specialized variants. Especially the tanker and I agree on stretched tanks but not the whole length. There are some mass trade off here. If payload to LEO is 150t the tanks can be stretched enough to hold that much more. That allows the four rings of the cargo section to be cut back to a lesser number. Each ring removed cuts mass ~1.6t allowing that much more propellant. This we might be offset by larger front fins needed because of the shortened lever arm they’ll have back to the CG. This ship will be a high use workhorse with optimization, IMO, justified.

It also leads to a full length or longer full tanked, and unfinned variant as an orbital depo to cut down the number of refuelings a cargo or crewed ship will experience. This has been much discussed and getting OT.

The difficulties of plumbing, electrical and data in swapping tank sections, while moderately time consuming, offers no real technical challenges. It’s all well known stuff. I doubt the mass penalty would reach 500kg. Heat shield continuity will, IMO, be the hard but not insurmountable issue. Swapping out the tank section is not something that will be done on a regular basis.

Phil

[dnavas]

Right now we’re seeing a build rate of ~17/year.

One caution: flappies and heat shield have yet to be integrated into builds, which will affect build rates.

[Nilof]

How about hexagonal tiles on nosecone? It is impossible with one kind of tile unless there is some special nosecone geometry I don't know about?

Hex is pretty universal on plain nose, it will be mostly without special tile shapes. Flaps and top of nosey will definitely need more specific tile shapes but if they plan to run mass production it is little issue. For maybe 90-95% of surface is hex more than good shape.

tiling non zero gaussian curvature (which nosecone is) with hexagons IS impossible

Couldn't this be solved by adding pentagon tiles, the soccer ball solution? Heptagon?

That's quite a clever thought. I have never understood the math that generates custom geodesic dome shapes (even asymmetrical ones) from standardized hexagons and pentagons. But some of the posters here seem quite able mathematicians who might be able to extend your idea.

Those work because geodesic domes aren't spheres but polyhedra, the curvature is along their edges, so all they need to do is find spherical polyhedra. And they use triangles, which are more flexible than hexagons.

To tile a curved surface, the tiles need to be curved along the surface as well. There's both extrinsic and intrinsic (gaussian) curvature that needs to be handled.

Of course, if Starship moved towards a biconic nosecone, they'd have zero problems with gaussian curvature and they could be tiled with hexagons. You'd only need the nosecone tiles to be curved along the principal axis. But it looks like SX is going for a Von Karman nosecone, which is harder to tile. One option is to simply accept large gaps between tiles due to the angles not matching, and bolting on a second layer of tiles on top of the first layer.

[xvel]

How about hexagonal tiles on nosecone? It is impossible with one kind of tile unless there is some special nosecone geometry I don't know about?

Hex is pretty universal on plain nose, it will be mostly without special tile shapes. Flaps and top of nosey will definitely need more specific tile shapes but if they plan to run mass production it is little issue. For maybe 90-95% of surface is hex more than good shape.

tiling non zero gaussian curvature (which nosecone is) with hexagons IS impossible

Couldn't this be solved by adding pentagon tiles, the soccer ball solution? Heptagon?

something like this, tiling with minimal number of tile types, but current nosecone geometry seems bad for this, that's an interesting topic, is there a mathematician here?

[OTV Booster]

Per Elon’s comment on the nose cones on the update thread, I wouldn’t doubt if even the production nose cones themselves will have their own serial number series, and will be agnostic as to which tank they will be used on. It sounds like Starship will be a modular system that can have any number of nosecones used based on payload, type or other factors.

Would not surprise me if the nose cone and front payload bay section is a bolt-on joint to the aft section...
The VAB there is not tall enough to weld them together (IMHO)
For now... they will build/weld fronts and backs side by side in the VAB and stack them and add fins outside for now...
All those joints are likely bolted (IMHO)

This is one of the reasons I was hoping to see header tanks moving out of the fairing section - to reduce complexity and increase flexibility so that the fairing itself is treated more as the payload or “client” of the tank/engine section.

Not arguing for or against the nose header but I’d expect unmating to be an executive level decision and not undertaken lightly. It will be beyond routine maintenance and if plumbing needs reconnect and testing, well that’s the cost of doing business. Added hassle on one hand, flexibility on the other

Phil

SpaceX will need fast connect propellant and data lines for mating Starship to Superheavy on the pad, doing the same thing for mating Starships propulsion and payload segments would only require reusing basic designs already needed anyway.

Quick disconnects always seem to be a lot heavier and much much more expensive than manual. A service station air quick disconnect is quite a bit heavier than a compression or flair fitting for one example. Demating, IMO just won’t be a routine thing.

Phil

[Nilof]

I definitely think we'll see a lot of automation happen for the Starlink and Raptor assembly lines, which definitely could end up reaching production volumes comparable to the car industry.

For the Starship hull assembly, probably not.

[Dr.Unhold]

I have some serious doubts about a 'completely modular' Starship, with a common fuel section and interchangable payload seciton.
However I think the strongest argument comes in the form of the tanker variant. Why have an SS fly a fuel tank in its nose cone, rather than extending the tanks internally to take up all the available volume? SpaceX is already going to be making a stretched tank for the SH, they're already working on making the most of the cone volume by using the tip as a tank wall, so what's stopping them from putting the common bulkhead at the height where the top bulkhead currently sits, and using every CM^3 of volume available for fuel? To me it seems like a simpler solution than trying to make a separate LOX/CH4 tank with its own bulkheads that fits in a nosecone and can be mated to a 'common' thrust section. They'd need to run plumbing from such a tanker payload bay to the rear-mounted refueling adapters on every thrust section for it to be truly modular, meaning each of the other variants would take a mass reduction.

A totally modular payload section seems overcomplicated and inefficient to me. I could see simple cargo SS's to have a removable/interchangable payload section, maybe, if they can figure out how to do the various plumbing/electrical connections reliably, and even that's a pretty big if.

I think it's much more likely that we see some specialized SS's begin development after a prototype makes orbit.

I am not sure stretching the tanks does work out that easily. If you stretch the LOX tank, the CH4 header tank will move up, too. And that would change the weight distribution, which is problematic for the flip. (At least it would require a different software profile, which adds complexity.)

That said, I am equally not convinced they have a modular design. There is simply no evidence yet to speculate in this direction either.

[SkyRate]

tiling non zero gaussian curvature (which nosecone is) with hexagons IS impossible

Only if you limit yourself to regular hexagons.
Of course, as you approach the nose they will sooner or later have to become non-flat, but the 2D projection could still be very close to hexagonal. Guess it depends on the gap-filling approach. (I assume we do not know whether the test articles are flat or follow the 9 m diameter curve.)

[cdebuhr]

I have some serious doubts about a 'completely modular' Starship, with a common fuel section and interchangable payload seciton.
However I think the strongest argument comes in the form of the tanker variant. Why have an SS fly a fuel tank in its nose cone, rather than extending the tanks internally to take up all the available volume? SpaceX is already going to be making a stretched tank for the SH, they're already working on making the most of the cone volume by using the tip as a tank wall, so what's stopping them from putting the common bulkhead at the height where the top bulkhead currently sits, and using every CM^3 of volume available for fuel? To me it seems like a simpler solution than trying to make a separate LOX/CH4 tank with its own bulkheads that fits in a nosecone and can be mated to a 'common' thrust section. They'd need to run plumbing from such a tanker payload bay to the rear-mounted refueling adapters on every thrust section for it to be truly modular, meaning each of the other variants would take a mass reduction.

A totally modular payload section seems overcomplicated and inefficient to me. I could see simple cargo SS's to have a removable/interchangable payload section, maybe, if they can figure out how to do the various plumbing/electrical connections reliably, and even that's a pretty big if.

I think it's much more likely that we see some specialized SS's begin development after a prototype makes orbit.

I am not sure stretching the tanks does work out that easily. If you stretch the LOX tank, the CH4 header tank will move up, too. And that would change the weight distribution, which is problematic for the flip. (At least it would require a different software profile, which adds complexity.)

That said, I am equally not convinced they have a modular design. There is simply no evidence yet to speculate in this direction either.

Consider the difference in mass distribution between an empty "chomper" and a crewed "people hauler" with all its ECLSS and other stuff needed to keep the biologicals healthy and happy in space.  Or the difference between a cargo hauler returning to earth empty vs. the same craft returning with up to 50 tons of something-or-other.  Being able to deal with variable weight distribution isn't a problem to be avoided - its a core capability.  Stretching the tanks seems like a relatively minor issue by comparison.

If you want to minimize issues though, there is a relatively simple solution here ... if the CH4 header is moved forward enough to screw with trim, just pull the LOX header back to compensate.