Quote from: Paul451 on 01/04/2020 06:28 amHow is cutting apart the tanks and welding new ring sections easier than removing some engines to be flown back on a (now empty) supply flight?Stretching the tanks requires only linear cuts and welds of sheet stainless steel. If you can cut and weld a straight line, you can perform that operation. Removing engines, however, means you need to detach:- Various bolts designed for ground operations- All the plumbing and wiring associated with an engine (multiple feedlines, purge lines, pressurisation lines, hydraulic lines, sensor cabling, etc)- All the exterior panels and fasteners needed to get at those fasteners and plumbing (as the engines are buried inside the skirt rather than exposed at the end of a bulkhead)And do all of that in microgravity, and either in spacesuits or with telemanipulation. And then you need to plug up all the holes in the plumbing you've just made. which itself means the characteristics of that plumbing has changed, e.g. the LOX and LCH4 manifolds sized for 6 engine feed lines now has to feed one off-centre engine without any issues like cavitation or entrainment.And you're stuck with a singe off-centre engine.On the other hand, launching with a stubby tank and a single engine means all these changes can be made on the ground prior to launch.Quote(Especially if the rings are carried inside a ship of the same diameter, which means they would be segments of rings that need to be first welded into rings before those rings can welded into the extendo-Starship.Again, more linear welds. Depending on mass treadeoffs, it may make sense to pre-cut ring sections and just nest them a few thick (leaving a large internal void unused), or to lift tightly coiled continuous rolls along with a ring-roller to unroll them into large diameter ring and cut to size in orbit. I suspect pre-cut sections would be easier simply to avoid as much microgravity/vacuum operation as possible. QuoteHow many freakin' extra rings are you installing if you are more than doubling the tank volume?A 'full size' Starship currently uses 16 'tank rings' (longer overall due to extended base skirt and nose volume). Each ring is roughly 1.4 tons if 3mm thick (could be thinner for a microgravity-only Starship, but current rings are about this thickness so we'll use this for simplicity). With Starship's projected 100 ton LEO capacity, and assume 1/2 of that is dedicated to support equipment (welding equipment, crew support, etc) that leaves you with capacity for an additional two Starships' worth of tank length that can be lifted in one go. Drop down to 1mm sheet (serious limits imposed on maximum acceleration) and you have tanks over 6 times the length (and 6 times the volume) of a stock Starship.Or in other words, the modded stubby Starship gives you a hair over 2x a single starship tank capacity for 2x Starship launches (or 4x the capacity for 3 launches, 6x for 4 launches, etc) using 3mm sheet. Or a whopping 6x the capacity for 3 launches using 1mm sheet. And you only tie up one Raptor engine idling in orbit rather than 6.
How is cutting apart the tanks and welding new ring sections easier than removing some engines to be flown back on a (now empty) supply flight?
(Especially if the rings are carried inside a ship of the same diameter, which means they would be segments of rings that need to be first welded into rings before those rings can welded into the extendo-Starship.
How many freakin' extra rings are you installing if you are more than doubling the tank volume?
After Mars, what’s next? Moons of Jupiter?Elon Musk ✓ @elonmuskCeres, Callisto, Ganymede & Titanhttps://twitter.com/elonmusk/status/1170983609492103168
Pics of the nosecone and
In this thesis an analysis of the suitability of welding processes for space applications was performed. We concluded that defocused EBW [electron beam welding] is the best for most EVAs [extravehicluar activity], and GTAW [gas tungsten arc welding] is the best for most IVAs [intravehicular activity].An analysis of feasibility of using earth bounded NDT [non-destructive testing] techniques in the space environment was performed. We concluded that x-rays are well suited for EVAs and ultrasound for IVAs. Methods involving liquids are not suitable for EVAs.A mathematical model was used to analyze the generation of humping [beading] in microgravity. Space conditions are more prone to humping generation than earth conditions.
Georgi Shonin and Valeri Kubasov, Russian cosmonauts who crewed the Soviet Soyuz 6 mission to space in 1969, were the first to experiment with welding in space, using a versatile tool known as the Vulkan. Shonin and Kubasov were also the first to experience just how dangerous space repairs can be. The pair tested three welding processes to see how they would function in a depressurized environment: electron beam welding, low-pressure compressed arc welding and arc welding with a consumable electrode.During the test, Kubasov almost burned through the hull of the Soyuz 6’s living compartment, a mistake that would have hurled the pair into space without spacesuits to face the final 30 seconds of their lives. Fortunately, the hull remained intact, but with a warning about the harshness and complexity of welding in space. The weld quality of the titanium, aluminum alloy and stainless steel samples was comparable to that found on Earth.
Quote from: rakaydos on 11/03/2019 03:55 pmQuote from: edzieba on 11/02/2019 11:28 amQuote from: Twark_Main on 10/31/2019 02:35 amQuote from: edzieba on 10/30/2019 02:24 pmQuote from: Twark_Main on 10/26/2019 04:52 pmWhy does this modification have to be done in orbit? Just stretch the tanks on the ground and launch them partly filled. Far far less R&D and risk than inventing on-orbit construction.To launch the absolute minimum pass possible, in order to attach as few engines as possible. Ideally you would only launch with one Raptor (or zero and get to orbit on the Methox RCS thrust alone but that may be unfeasible while still recovering Superheavy). Launching a stretched version not only requires the same or greater number of Raptors to be mounted, it also means a stack with an extremely fine aspect ratio, so more simulation needed (plus potentially additional stiffening hardware). A sufficiently short starship with lower thrust may even get away with less stiffening members than the stock configuration.By launching a 'stubby' Starship with as few engines as possible, then adding lightweight rings in orbit (carried by a second launch), you have as few 'dead' engines sitting in orbit as possible (both wasted money and wasted mass), andwithout the need to mess with engine plumbing while in orbit to try and remove and recover them.The cure here is worse than the disease. You're trying to reduce launch mass, but then proposing a system that require at least two launches, plus on-orbit assembly (at unknown cost). Why save upmass if it doesn't actually save launches?Reducing the TWR isn't good for mass-to-orbit efficiency. If it were more efficient to use just one Raptor engine, don't you think SpaceX would do that?Just send up a few $200k Raptors and be done with it. Lower cost, and lower development risk too.You misunderstand: the goal is not to reduce launch mass, it's to reduce dry mass. That is achieved with as few engines as possible and as light tank walls as possible.Applying the constraints of not modifying anything in orbit other than basic linear cuts and welds in Stainless Steel (e.g. no touching engines, RCS, etc) that means the fewer engines you launch with, the lower your dry mass. To reach orbit with fewer engines means needing a lower launch mass (because if you have a TWR below unity then you may not achieve orbit in the first place without an extreme lofted booster trajectory as with Vulcan/Centaur, which expends your booster). The low launch mass is a consequence of optimising for dry mass, not the goal. The reason for launching a 'stubby' Starship and then a second carrying rings & crew to expand it in orbit, is that after those two launches you have a single Starship with a propellant capacity dramatically above that of a pair of 'stock' Starships in orbit (plus one recovered Starship to re-use), as well as a dry mass (and thus propulsive efficiency to move your depot to where it is needed) dramatically below that of a pair of Starships.So what destinations does this open up?Earth orbit, specifically plane changes. A high-capacity tanker with high propulsive efficiency can move itself to the desired plane for the next Starship launch.
Quote from: edzieba on 11/02/2019 11:28 amQuote from: Twark_Main on 10/31/2019 02:35 amQuote from: edzieba on 10/30/2019 02:24 pmQuote from: Twark_Main on 10/26/2019 04:52 pmWhy does this modification have to be done in orbit? Just stretch the tanks on the ground and launch them partly filled. Far far less R&D and risk than inventing on-orbit construction.To launch the absolute minimum pass possible, in order to attach as few engines as possible. Ideally you would only launch with one Raptor (or zero and get to orbit on the Methox RCS thrust alone but that may be unfeasible while still recovering Superheavy). Launching a stretched version not only requires the same or greater number of Raptors to be mounted, it also means a stack with an extremely fine aspect ratio, so more simulation needed (plus potentially additional stiffening hardware). A sufficiently short starship with lower thrust may even get away with less stiffening members than the stock configuration.By launching a 'stubby' Starship with as few engines as possible, then adding lightweight rings in orbit (carried by a second launch), you have as few 'dead' engines sitting in orbit as possible (both wasted money and wasted mass), andwithout the need to mess with engine plumbing while in orbit to try and remove and recover them.The cure here is worse than the disease. You're trying to reduce launch mass, but then proposing a system that require at least two launches, plus on-orbit assembly (at unknown cost). Why save upmass if it doesn't actually save launches?Reducing the TWR isn't good for mass-to-orbit efficiency. If it were more efficient to use just one Raptor engine, don't you think SpaceX would do that?Just send up a few $200k Raptors and be done with it. Lower cost, and lower development risk too.You misunderstand: the goal is not to reduce launch mass, it's to reduce dry mass. That is achieved with as few engines as possible and as light tank walls as possible.Applying the constraints of not modifying anything in orbit other than basic linear cuts and welds in Stainless Steel (e.g. no touching engines, RCS, etc) that means the fewer engines you launch with, the lower your dry mass. To reach orbit with fewer engines means needing a lower launch mass (because if you have a TWR below unity then you may not achieve orbit in the first place without an extreme lofted booster trajectory as with Vulcan/Centaur, which expends your booster). The low launch mass is a consequence of optimising for dry mass, not the goal. The reason for launching a 'stubby' Starship and then a second carrying rings & crew to expand it in orbit, is that after those two launches you have a single Starship with a propellant capacity dramatically above that of a pair of 'stock' Starships in orbit (plus one recovered Starship to re-use), as well as a dry mass (and thus propulsive efficiency to move your depot to where it is needed) dramatically below that of a pair of Starships.So what destinations does this open up?
Quote from: Twark_Main on 10/31/2019 02:35 amQuote from: edzieba on 10/30/2019 02:24 pmQuote from: Twark_Main on 10/26/2019 04:52 pmWhy does this modification have to be done in orbit? Just stretch the tanks on the ground and launch them partly filled. Far far less R&D and risk than inventing on-orbit construction.To launch the absolute minimum pass possible, in order to attach as few engines as possible. Ideally you would only launch with one Raptor (or zero and get to orbit on the Methox RCS thrust alone but that may be unfeasible while still recovering Superheavy). Launching a stretched version not only requires the same or greater number of Raptors to be mounted, it also means a stack with an extremely fine aspect ratio, so more simulation needed (plus potentially additional stiffening hardware). A sufficiently short starship with lower thrust may even get away with less stiffening members than the stock configuration.By launching a 'stubby' Starship with as few engines as possible, then adding lightweight rings in orbit (carried by a second launch), you have as few 'dead' engines sitting in orbit as possible (both wasted money and wasted mass), andwithout the need to mess with engine plumbing while in orbit to try and remove and recover them.The cure here is worse than the disease. You're trying to reduce launch mass, but then proposing a system that require at least two launches, plus on-orbit assembly (at unknown cost). Why save upmass if it doesn't actually save launches?Reducing the TWR isn't good for mass-to-orbit efficiency. If it were more efficient to use just one Raptor engine, don't you think SpaceX would do that?Just send up a few $200k Raptors and be done with it. Lower cost, and lower development risk too.You misunderstand: the goal is not to reduce launch mass, it's to reduce dry mass. That is achieved with as few engines as possible and as light tank walls as possible.Applying the constraints of not modifying anything in orbit other than basic linear cuts and welds in Stainless Steel (e.g. no touching engines, RCS, etc) that means the fewer engines you launch with, the lower your dry mass. To reach orbit with fewer engines means needing a lower launch mass (because if you have a TWR below unity then you may not achieve orbit in the first place without an extreme lofted booster trajectory as with Vulcan/Centaur, which expends your booster). The low launch mass is a consequence of optimising for dry mass, not the goal. The reason for launching a 'stubby' Starship and then a second carrying rings & crew to expand it in orbit, is that after those two launches you have a single Starship with a propellant capacity dramatically above that of a pair of 'stock' Starships in orbit (plus one recovered Starship to re-use), as well as a dry mass (and thus propulsive efficiency to move your depot to where it is needed) dramatically below that of a pair of Starships.
Quote from: edzieba on 10/30/2019 02:24 pmQuote from: Twark_Main on 10/26/2019 04:52 pmWhy does this modification have to be done in orbit? Just stretch the tanks on the ground and launch them partly filled. Far far less R&D and risk than inventing on-orbit construction.To launch the absolute minimum pass possible, in order to attach as few engines as possible. Ideally you would only launch with one Raptor (or zero and get to orbit on the Methox RCS thrust alone but that may be unfeasible while still recovering Superheavy). Launching a stretched version not only requires the same or greater number of Raptors to be mounted, it also means a stack with an extremely fine aspect ratio, so more simulation needed (plus potentially additional stiffening hardware). A sufficiently short starship with lower thrust may even get away with less stiffening members than the stock configuration.By launching a 'stubby' Starship with as few engines as possible, then adding lightweight rings in orbit (carried by a second launch), you have as few 'dead' engines sitting in orbit as possible (both wasted money and wasted mass), andwithout the need to mess with engine plumbing while in orbit to try and remove and recover them.The cure here is worse than the disease. You're trying to reduce launch mass, but then proposing a system that require at least two launches, plus on-orbit assembly (at unknown cost). Why save upmass if it doesn't actually save launches?Reducing the TWR isn't good for mass-to-orbit efficiency. If it were more efficient to use just one Raptor engine, don't you think SpaceX would do that?Just send up a few $200k Raptors and be done with it. Lower cost, and lower development risk too.
Quote from: Twark_Main on 10/26/2019 04:52 pmWhy does this modification have to be done in orbit? Just stretch the tanks on the ground and launch them partly filled. Far far less R&D and risk than inventing on-orbit construction.To launch the absolute minimum pass possible, in order to attach as few engines as possible. Ideally you would only launch with one Raptor (or zero and get to orbit on the Methox RCS thrust alone but that may be unfeasible while still recovering Superheavy). Launching a stretched version not only requires the same or greater number of Raptors to be mounted, it also means a stack with an extremely fine aspect ratio, so more simulation needed (plus potentially additional stiffening hardware). A sufficiently short starship with lower thrust may even get away with less stiffening members than the stock configuration.By launching a 'stubby' Starship with as few engines as possible, then adding lightweight rings in orbit (carried by a second launch), you have as few 'dead' engines sitting in orbit as possible (both wasted money and wasted mass), andwithout the need to mess with engine plumbing while in orbit to try and remove and recover them.
Why does this modification have to be done in orbit? Just stretch the tanks on the ground and launch them partly filled. Far far less R&D and risk than inventing on-orbit construction.
If that Starship is launching to the Moon, Mars, or just a higher Earth orbit, having a tanker waiting for it means you can launch directly into the correct inclination, load propellant, then burn to the insertion orbit directly. Having Starship need to meet a tanker in the tanker's orbit means Starship plus the payload need to plane-change, wasting propellant.
Meeting the tanker for any given launch attitude means you can also accumulate 'free' propellant in orbit. Any Starship that is not using its maximum payload in actual payload capacity can 'excess' propellant, which it can drop off into the waiting tanker.
With very high capacity and high propulsive efficiency it can also provide a fail-safe option for early human Mars exploration. It would allow a full Starship's propellant load to by transported to Mars and by ready to go in Martian orbit in the event of an ISRU failure. That means contingency propellant carried by a human expedition need only be sufficient to reach low Mars orbit rather than perform a full return (which may not be otherwise possible without multiple 'normal' tankers sent to combine their loads).
And even if it never moves under its own power again, having an extremely large tanker also aids in multi-Starship BEO operations. You can perform multiple propellant loading flights over an arbitrary period of time to accumulate multiple 'normal' Starships' worth of prop load, then when your TLI/TMI/TJI/etc window opens you can launch multiple Starships at the same time with your waiting propellant independent of tanker launches. The benefits scale with simultaneous departures: e.g. for a three-Starship departure, you either have three returnable Starship tankers locked up in orbit waiting to be used, or a single dedicated triple-capacity tanker.
If the tanker is not required to move beyond RCS for compliant rendezvous, it can be extended using thin sheet stainless to truly absurd lengths with very little concern for buckling loading.
The core idea is to take the same capabilities the already proposed dedicated Starship tanker variant affords, and quantitatively increase them by performing the most basic in-orbit modifications possible (linear cuts & welds).
90% of welding is not welding, it's fixturing and cleaning. Vacuum minimises contamination concerns (no oxygen to oxidise, no airborne contaminants to be deposited),
and fixturing can be MUCH lighter weight in microgravity as you do not need to fight against gravity to hold everything in place and undistorted. Without flux to form slag or purge-gas to blow melt around, even welds with filler would be as clean or cleaner as autogenous welds are in atmosphere. Any fixtures only need to be sufficiently stiff to counter forces from the welding process itself (minimal from the weld head itself, thermal asymmetry would need to be taken into account by e.g. stitch-welding as it does on the ground) and movements of the two halves from external influence (tidal forces depending on alignment, atmospheric drag, etc). These are much lower than the forces applied by gravity, and the thermal effects in particular will be in-place with the parts being welded.
With ring segments pre-cut to size before launch, the existing tank segments can be used to locate and fix subsequent rings to the tank. A single-ring-width fixture that can extend/retract one ring width allows you to 'walk' the fixture down the tank without the two halves ever ending up completely free-floating, and can provide a track for a welder to run along (as ring welders already do). The fixture can be attached via resistance spot-welds (leaving a trail of discarded 'feet' down the tank) or magnetic clamping with semi-free-floating partner magnetic surfaces inside the tank (to be removed on closeout), or via effectively a large collet around the entire circumference. Spot-welding seems the simplest and lowest mass, if visually inelegant.Extending the downcomer is a similar process to extending the tank alls, but with the smaller diameter you can skip the vertical seam weld and just perform the ring welds, bringing up pre-cut pipe segments. Access to the interior to pre-stage all required parts and equipment can be done when the tank is opened for lengthening, the downcomer segments installed as the tank is grown, and the pipe joining rig removed prior to the final tank segment being installed. This minimises the number of times a spacesuited human needs to squeeze through the tank inspection hatch (and ideally final internal inspection could be performed with a long non-manipulation arm (lightweight) or even a free-floating or semi-tethered camera (with enough tank open time to allow nitrogen to disperse from any cold-gas thruster use), allowing complete avoidance of having to squeeze a spacesuited human through the tank inspection hatch. It also minimises the hands-on interaction required for operation, mainly limited to initial install of the fixture, replenishment of pre-cut rings and feedstock, and repositioning and removal of the fixture. With a sufficiently long teleoperated arm, the replenishment operations could be conducted without any human hands involved further reducing risk. This is compared to engine removal where hands-on operation in a confined space is extensive (likely needing to shove at least two spacesuited humans up in there along with a full complement of Raptors in order to get the first one out), chance for damage (to both suits and engine bay components) is high, and a large number of specialised hand tools are required. Even using an set of arms for teleoperation still requires a large array of tools, along with extensive simulation to avoid clearance issues.
We're a long, long way from EVA being routine. Removing an engine in-orbit would be far more of a challenge than the Hubble servicing missions, and those were hardly a walk in the park.
And even if engines were a simple 'unplug and go' item, it would still be less effective than a tank-lengthening mission in terms of available in-orbit propellant capacity per launch.
The LEO environment is full of atomic oxygen, so there definitely will be oxidation concerns.
And in space anything volatile expands and re-condenses on nearly surfaces, so there definitely will be contamination concerns.
The thermal expansion caused by 400 °F between sunlight and shadow is nothing to sneeze at. It won't be "in-place" if parts are differently shaded (eg by fixtures or tooling), different thicknesses (meaning different time constants as they heat/cool over the day/night cycle), different surface wear affecting the optical properties, etc.
And how is all of this less R&D than simply stretching the tank and re-certifying the new design for launch (which you claimed was so problematic)?
As you said, this plan doesn't reduce launch mass (since you have at least two launches plus in-space fabrication costs per tanker).
Simply launching an extra load of propellant with that second launch would easily overwhelm any plausible reduction in effectiveness from the slight increase in dry mass (a measly ~5 tonnes).
Quote from: Twark_Main on 02/12/2020 02:42 amThe LEO environment is full of atomic oxygen, so there definitely will be oxidation concerns.At sufficiently low densities that they will be less of an effect than oxygen mixing into the shielding gas at 1ATM.
Quote from: Twark_Main on 02/12/2020 02:42 amAnd in space anything volatile expands and re-condenses on nearly surfaces, so there definitely will be contamination concerns.That could be a concern. Extremely low pressure Helium (or Argon) pressurisation of the tank volume during welding (with even a basic leaky sliding seal around the joint sections to keep loss rates acceptable) would serve to 'blow out' contaminants that would otherwise have the opportunity to settle inside. Or, just allow them to settle on the tank walls (temporarily plug plumbing intakes to simplify cleaning) and run a partial-fill > spin (to centrifuge) > thrust fore-aft (to slosh to the domes) > purge cycle to 'rinse' the walls. With de-facto vacuum purging of all weld surfaces prior to welding, contamination should be minimal to start with.
Quote from: Twark_Main on 02/12/2020 02:42 amThe thermal expansion caused by 400 °F between sunlight and shadow is nothing to sneeze at. It won't be "in-place" if parts are differently shaded (eg by fixtures or tooling), different thicknesses (meaning different time constants as they heat/cool over the day/night cycle), different surface wear affecting the optical properties, etc.This is indeed a consideration. A large deployable sun-shade could be an option, or 'buddy shadowing' using the maintenance Starship, or a sufficiently fast BBQ-roll to normalise heat around the entire diameter (stainless' poor heat conductivity would be a boon here).
Quote from: Twark_Main on 02/12/2020 02:42 amAnd how is all of this less R&D than simply stretching the tank and re-certifying the new design for launch (which you claimed was so problematic)? Because if something needs to be launched, it has structural an aerodynamic constrains it needs to fulfil that then serve no purpose once in orbit. A purely in-orbit manufacture would allow for optimum dry-mass (e.g. balloon tanks with spaced foil insulation), but is a much greater challenge to actually do vs. planetside manufacture. No R&D can research and develop away the laws of physics, a foil-thin stage with an absurd 33:1 fineness ratio is not going to survive launch from the surface but will be perfectly happy in orbit.
Quote from: Twark_Main on 02/12/2020 02:42 amAs you said, this plan doesn't reduce launch mass (since you have at least two launches plus in-space fabrication costs per tanker).It does reduce launch mass (just not as the end goal) as you launch a single-Raptor Starship with smaller tanks.
But as I said, launch-mass reduction is not the goal, in-orbit dry mass reduction is, which is where a single-engine absurd-length Starship excels vs. a stock Starship or a slightly stretched Starship that needs to be launched as a unitary object. QuoteSimply launching an extra load of propellant with that second launch would easily overwhelm any plausible reduction in effectiveness from the slight increase in dry mass (a measly ~5 tonnes).And if you want two loads of propellant ready in orbit, you need: 1) a double-length stretch tanker (dry mass negligible increase)or2) Two Starships, with 10 tons dry mass an 4-12 Raptors acting as extremely expensive ballast.
With each additional Starship (vs. each additional starship-length ring stack) the dry-mass advantage grows.
It would be interesting - though too much time for a lunch-break - to figure out how far a single-engine starship would need to be extended to boost a small mass to the same C3 as a Starkicker (or other kick stage) while retaining sufficient propellant to return itself. Not really a good use of the thing, but a neat idea.
what about an inflatable tanker? No on-orbit assembly, you can test it on the ground (imagine if your welded tanker bursts on the first pressure test; ouch!), it can be spherical which improves thermal and structures, and kevlar has a higher strength-to-weight (and the sphere is more efficient) so the fuel capacity is greater per unit upmass.
Quote from: Twark_Main on 02/14/2020 01:43 amwhat about an inflatable tanker? No on-orbit assembly, you can test it on the ground (imagine if your welded tanker bursts on the first pressure test; ouch!), it can be spherical which improves thermal and structures, and kevlar has a higher strength-to-weight (and the sphere is more efficient) so the fuel capacity is greater per unit upmass.Where round-trip forces are modest, an inflatable tanker system could boost tanker infrastructure efficiency, yes: e.g., cycling between a LEO LOX trawler depot and a lunar base. Some custom lunar spaceport infrastructure would be needed to manage that notional flight configuration. Still, using inflatables, such a system could plausibly offer a full order-of-magnitude improvement over SpaceX baseline propellant-to-payload ratio.
Quote from: LMT on 02/15/2020 03:27 amQuote from: Twark_Main on 02/14/2020 01:43 amwhat about an inflatable tanker? No on-orbit assembly, you can test it on the ground (imagine if your welded tanker bursts on the first pressure test; ouch!), it can be spherical which improves thermal and structures, and kevlar has a higher strength-to-weight (and the sphere is more efficient) so the fuel capacity is greater per unit upmass.Where round-trip forces are modest, an inflatable tanker system could boost tanker infrastructure efficiency, yes: e.g., cycling between a LEO LOX trawler depot and a lunar base. Some custom lunar spaceport infrastructure would be needed to manage that notional flight configuration. Still, using inflatables, such a system could plausibly offer a full order-of-magnitude improvement over SpaceX baseline propellant-to-payload ratio. If it cycles between LEO and LLO, the custom lunar spaceport infrastructure goes away.
Quote from: Twark_Main on 02/15/2020 03:48 amQuote from: LMT on 02/15/2020 03:27 amQuote from: Twark_Main on 02/14/2020 01:43 amwhat about an inflatable tanker? No on-orbit assembly, you can test it on the ground (imagine if your welded tanker bursts on the first pressure test; ouch!), it can be spherical which improves thermal and structures, and kevlar has a higher strength-to-weight (and the sphere is more efficient) so the fuel capacity is greater per unit upmass.Where round-trip forces are modest, an inflatable tanker system could boost tanker infrastructure efficiency, yes: e.g., cycling between a LEO LOX trawler depot and a lunar base. Some custom lunar spaceport infrastructure would be needed to manage that notional flight configuration. Still, using inflatables, such a system could plausibly offer a full order-of-magnitude improvement over SpaceX baseline propellant-to-payload ratio. If it cycles between LEO and LLO, the custom lunar spaceport infrastructure goes away.The extra infrastructure could be as simple as an elevated landing ring. Not a significant factor.
...On how in orbit refueling is done. With the current Starship design, it MUST be done under acceleration. You must take that into account for any system you develop. Will that inflatable tank be able to handle the repeated accelerations and decelerations needed? Think of how much more mass you are accelerating and decelerating when you have a propellant load in that depot greater than than one SS can receive.
The acceleration needed to enable starship refueling as planned is minuscule. Just enough to overcome the surface tension of the propellant and allow it to accumulate at the “bottom” of the tank.Some other thoughts about an optimal solution — a special purpose tanker that accumulates the propellant of up to 3-4 regular tankers prior to the outbound Starship launching.— that tanker can stay in orbit, won’t need heat shield, but will need solar arrays, and possibly ion thrusters.— it will be positioned in the plane of the outbound Starship, so the latter won’t have to plane change to meet it.— the volume of propellant the tanker holds only needs to be enough to completely refuel Starship in one go, plus some residual.— after that mission is complete, the tanker can reposition itself to a different orbital plane as needed for the next starship.— at that location, it gets refilled by several regular tankers, then the next Starship launches, gets refueled, and the cycle repeats.
Quote from: Jcc on 02/15/2020 01:30 pmThe acceleration needed to enable starship refueling as planned is minuscule. Just enough to overcome the surface tension of the propellant and allow it to accumulate at the “bottom” of the tank.Some other thoughts about an optimal solution — a special purpose tanker that accumulates the propellant of up to 3-4 regular tankers prior to the outbound Starship launching.— that tanker can stay in orbit, won’t need heat shield, but will need solar arrays, and possibly ion thrusters.— it will be positioned in the plane of the outbound Starship, so the latter won’t have to plane change to meet it.— the volume of propellant the tanker holds only needs to be enough to completely refuel Starship in one go, plus some residual.— after that mission is complete, the tanker can reposition itself to a different orbital plane as needed for the next starship.— at that location, it gets refilled by several regular tankers, then the next Starship launches, gets refueled, and the cycle repeats.This same mission architecture can be done with a regular unmodified tanker. The mass advantage of deleting the heat shield is overwhelmed by the high fuel mass required to change planes. Cheaper (and less upmass) to just land the tanker and re-launch it in the next orbital plane.
