What do you think is the best method to transfer Methalox fuel from one ship to another in space?Some options to separate liquid from gas in the tank could be: spinning ships, accelerating ships, making use of a gravity gradient induced by a planet? magnetic force (liquid Oxygen is attracted by magnets), creating an internal vortex, etc.
Do you envision the rotation to settle the fuel at the bottom or at the side of the rocket?
The paper chose rotation as the method for settling propellants, but now I would favor acceleration.
Two cyndrelical rockets docking belly to belly, then introducing a minor (sub-1 RPM) rotation to settle the tanks. Within a quarter rotation, all propellant in both tanks should be interacting with a tank wall and gaining the effects of centrifical force, pooling on the side of the tanks opposide the docking port. Then just pump fuel and oxyser into the opposite pressure vessel.
Quote from: rakaydos on 08/05/2017 05:32 pmTwo cyndrelical rockets docking belly to belly, then introducing a minor (sub-1 RPM) rotation to settle the tanks. Within a quarter rotation, all propellant in both tanks should be interacting with a tank wall and gaining the effects of centrifical force, pooling on the side of the tanks opposide the docking port. Then just pump fuel and oxyser into the opposite pressure vessel.But then you have a shifting center of mass. Eventually you have a mostly full spacecraft and a mostly empty spacecraft docked together with a center of mass mostly in the full one. It might work but it might put a big strain on the docking interface.
Rotation could be a little tricky when your center of mass is constantly changing.
But then you have a shifting center of mass. Eventually you have a mostly full spacecraft and a mostly empty spacecraft docked together with a center of mass mostly in the full one. It might work but it might put a big strain on the docking interface.
Quote from: Ionmars on 08/05/2017 04:03 pmThe paper chose rotation as the method for settling propellants, but now I would favor acceleration.If you're already planning to be accelerating towards a destination then this method would make a lot of sense. And it's not like rotational methods don't use any fuel, since they have to start the rotation of the two spacecraft at the very least, and depending on the fueling methods and docking systems they might have to de-rotate before undocking too. Spinning up and spinning down would consume some of the fuel that otherwise could have gone towards acceleration.I'm not sure what the percentage of use cases are where that would be preferable, but it's worth remembering as an option.Nice job on the AIAA paper too - it's one thing to talk about something on a forum, but it's a whole lot harder to write a paper with the intent to be published on AIAA.
Quote from: Coastal Ron on 08/05/2017 07:19 pmQuote from: Ionmars on 08/05/2017 04:03 pmThe paper chose rotation as the method for settling propellants, but now I would favor acceleration.If you're already planning to be accelerating towards a destination then this method would make a lot of sense. And it's not like rotational methods don't use any fuel, since they have to start the rotation of the two spacecraft at the very least, and depending on the fueling methods and docking systems they might have to de-rotate before undocking too. Spinning up and spinning down would consume some of the fuel that otherwise could have gone towards acceleration.I'm not sure what the percentage of use cases are where that would be preferable, but it's worth remembering as an option.Nice job on the AIAA paper too - it's one thing to talk about something on a forum, but it's a whole lot harder to write a paper with the intent to be published on AIAA.I agree with your analysis. In the case of the large depot, it consists of a lightweight framework with up to six spaceships attached to it, all in parallel with one another. Imagine another spacecraft approaching from "behind," intending to attach itself to an empty berth (parking space). As it approaches, it sees the framework and ships (the depot) as revolving around their common center. In order to park, it needs to match the motion of the empty berth. So it begins to spiral towards the depot whereby the radius of the spiral is a little greater than the radius of the spinning berth. Its rotational speed also matches the rotational speed of the depot. It comes to hover just "above" the empty berth. Now it must also revolve on its own axis slowly, so that its dorsal surface with fuel connectors is always facing the berth. Then it can connect to the berth framework.This maneuvering is a complex choreography using GNC thrusters. More important, it requires energy (fuel) that could be applied to a better use. So I now think it would be better to not rotate the depot, but allow spaceships to approach and park in a simpler manner. When it is time to transfer fuel, a slight acceleration (10E-5 g) would be applied to the depot, including the ship providing fuel and the ship receiving fuel. The direction of acceleration is important because we want to steer the depot towards a higher orbit that contributes toward TMI for all ships going to Mars. The others (presumably tankers) would return to Earth for reuse.
......Another method could be to temporarily rotate the fuel inside the large depot tank with a small electric motor connected to a mechanical stirrer inside the tank.If there's no rotation of the fuel needed, the rotational energy could be converted back to electric energy trough regenerative braking.
Quote from: Peter.Colin on 08/05/2017 08:49 pm......Another method could be to temporarily rotate the fuel inside the large depot tank with a small electric motor connected to a mechanical stirrer inside the tank.If there's no rotation of the fuel needed, the rotational energy could be converted back to electric energy trough regenerative braking.It is not intuitively obvious to me that this would work. The interior motor would begin a rotational motion of the fluid, but it seems to me that its energetic action would also cause some droplets to spray into the empty portion of the tank, adding to fuel dispersion as well as conglomeration.
I believe the easiest and least complex way to do this is by designing the tanker with a separate tank for filling the spacecraft (ie not shared with propulsion tank). This tank is made of some ALU compound and is shaped like a cylinder with a piston inside the tank, not pressurised and with some kind of drivebar in the middle for driving the piston. Motor and mechanics are sealed inside the tank, eg no movable parts that go into the tank. Piston ring seals around and a pressure valve for backfilling the other side of the piston with gas to prevent under-pressure.The methane is easy as it doesn't burn without oxygen (like a car gasoline pump floating in petrol). The oxygen is perhaps a bit harder due to fire hazard.The vehicle being filled requires no such device.Then it's a simple matter of matching orbit, connect a hose and squeezing the content over to the other vehicle.I believe rotation or acceleration will complicate matters more than wished.Edit: Actually a much simpler design would simply be to leave the piston loose. Just push it down using pressurised methane or oxygen in gas form from the rear side (from boil-off). No center pole or motor needed.
The problem I see with this is rockets tanks are usually not empty. The are stringers and other supports inside otherwise the wall thickness would have to increase to give the tank enough rigidity.
Quote from: jak Kennedy link=topic=43522.msg1710332#msg1710332 The problem I see with this is rockets tanks are usually not empty. The are stringers and other supports inside otherwise the wall thickness would have to increase to give the tank enough rigidity.Good point.Hence separate tanks for fuel to be transferred on top of the craft and non-pressurised should help reduce the load. Cylindrical shape will increase strength of outer wall, with the biggest weakness being the bottom, so it should probably be rounded in bottom (even if not all fuel can squeeze out). Bottom will have to sustain ~200 ton at 5-10G, which can be reinforced from below.
Thrusting is the way, that was determined long ago.
As far as I know transfer of cryogenic propellants in orbit has never been done before so the engineering is not well established.
Quote from: DreamyPickle on 08/07/2017 08:27 amAs far as I know transfer of cryogenic propellants in orbit has never been done before so the engineering is not well established.Settling of cryogenic propellants on orbit has been done though.
Quote from: savuporo on 08/07/2017 08:33 amQuote from: DreamyPickle on 08/07/2017 08:27 amAs far as I know transfer of cryogenic propellants in orbit has never been done before so the engineering is not well established.Settling of cryogenic propellants on orbit has been done though. exactly
Is thrust ullage for the entire duration of a multi kiloton fuel transfer really easier than burning up (and down) a minor spin ullage?From first principles, I would think that spin ullage would be a better fit for slower, longer duration fuel transfers, (thrust the whole time vs thrust once to spin up and once to spin down) which due to square/cube on transfer pipes should scale better to larger craft.
Quote from: rakaydos on 08/07/2017 03:49 pmIs thrust ullage for the entire duration of a multi kiloton fuel transfer really easier than burning up (and down) a minor spin ullage?From first principles, I would think that spin ullage would be a better fit for slower, longer duration fuel transfers, (thrust the whole time vs thrust once to spin up and once to spin down) which due to square/cube on transfer pipes should scale better to larger craft.Thrusting is easier. No attitude changes, no additional stresses, no worry about changing moments of inertia, etc.Boil off gases can be used for the thrusting.
You do realize that for maintaining the same acceleration constant for longer time you need to throttle up the trusters exponentially.
The thrusters can be shut off while spinning.
Quote from: Peter.Colin on 08/07/2017 06:09 pmYou do realize that for maintaining the same acceleration constant for longer time you need to throttle up the trusters exponentially.There is no need for the same acceleration rate, just a constant acceleration.But your point is wrong anyways. There is no mass loss or gain. The "system" is the two spacecraft and they are joined and the mass transfer is internal to the "system". So the same thrust is the same acceleration.Quote from: Peter.Colin on 08/07/2017 06:09 pmThe thrusters can be shut off while spinning.No, they can't. A rotating system will have changing moments of inertia and likely require a need for additional thruster firings for control.
Same trust is not same acceleration (= not same g-force)An object moving twice as fast has four times the energy.going from 50km/h to a 100km/h takes 3 times more thrust than than going from 0km/h to 50km/h.This adds up exponentially.
Also a rotating craft would need additional thruster firings to stop the rotation after the transfer is complete.
It seems like we are trying to solve a problem that has already been solved, as fuel transfer from two docked craft does not seem that different from fuel transfer to the engines.
Quote from: Basto on 08/07/2017 06:46 pmAlso a rotating craft would need additional thruster firings to stop the rotation after the transfer is complete.While that certainly could be true, it's possible that you could create fueling systems that undock while still rotating.QuoteIt seems like we are trying to solve a problem that has already been solved, as fuel transfer from two docked craft does not seem that different from fuel transfer to the engines. I agree. We have enough history on this already to understand some of the solutions, and I'm sure there will be more than one way to do this safely and fuel efficiently.
The 69 longituninal slow spin seems to me like the most logically sound approach for long duration fuel transfers. 5 seconds RCS to spin up and spin down, vs 5 hours of ullage thrust, stable rotational moment with a predictable drainable point in all tanks.
Quote from: rakaydos on 08/07/2017 09:31 pmThe 69 longituninal slow spin seems to me like the most logically sound approach for long duration fuel transfers. 5 seconds RCS to spin up and spin down, vs 5 hours of ullage thrust, stable rotational moment with a predictable drainable point in all tanks.That is neither logical nor sound. You don't know the actual configuration of the vehicle, size of thrusters, location of the tanks, the rotation moment is not stable. The transfer of propellants will involve venting, which will be a free source of propulsion. Which makes thrusting more "logical". It more "sound" because all the previous data.
Can anyone give some numbers on the efficiency of using acceleration? What would be the delta v over the entire operation? Is this delta-v wasted, or a tiny nudge in the direction you were going anyway? Is it totally negligible?Until someone gives numbers, for all I know it could be as small as moving the length of the tank during the refueling operation. Some simple numbers might make discussion of spinning and so on obviously not worth the bother.
Quote from: Jim on 08/07/2017 10:21 pmQuote from: rakaydos on 08/07/2017 09:31 pmThe 69 longituninal slow spin seems to me like the most logically sound approach for long duration fuel transfers. 5 seconds RCS to spin up and spin down, vs 5 hours of ullage thrust, stable rotational moment with a predictable drainable point in all tanks.That is neither logical nor sound. You don't know the actual configuration of the vehicle, size of thrusters, location of the tanks, the rotation moment is not stable. The transfer of propellants will involve venting, which will be a free source of propulsion. Which makes thrusting more "logical". It more "sound" because all the previous data.Should venting occur, which I doubt, the vent opening could be placed to slightly increase the 69 latitudinal spin.This should increase the settling force a little.Even if the center of rotation changes towards the being filled ship or tanker, the propellant tanks still maintain a stable and predictable drainage and filling point.
If two craft with dramatically changing mass and centers of mass are connected passively at a single point the stresses on the point would be huge. Now you have to design the whole ship around a new loading problem.
Quote from: intrepidpursuit on 08/07/2017 11:40 pmIf two craft with dramatically changing mass and centers of mass are connected passively at a single point the stresses on the point would be huge. Now you have to design the whole ship around a new loading problem.Just as a reminder, it appears that the approach SpaceX plans to use with the ITS is to attach two spacecraft side by side and then rotate them. With such a technique you don't have to worry where the center of rotation is, although you would have to use pumps to transfer liquids.
Quote from: Coastal Ron on 08/08/2017 12:16 amQuote from: intrepidpursuit on 08/07/2017 11:40 pmIf two craft with dramatically changing mass and centers of mass are connected passively at a single point the stresses on the point would be huge. Now you have to design the whole ship around a new loading problem.Just as a reminder, it appears that the approach SpaceX plans to use with the ITS is to attach two spacecraft side by side and then rotate them. With such a technique you don't have to worry where the center of rotation is, although you would have to use pumps to transfer liquids.Side by side, yes (as seen in video), but where do you get the rotating from? Sideways thrusting would work. (Spinning it also adds a complication of working against the spin gravity)
Quote from: Coastal Ron on 08/08/2017 12:16 amQuote from: intrepidpursuit on 08/07/2017 11:40 pmIf two craft with dramatically changing mass and centers of mass are connected passively at a single point the stresses on the point would be huge. Now you have to design the whole ship around a new loading problem.Just as a reminder, it appears that the approach SpaceX plans to use with the ITS is to attach two spacecraft side by side and then rotate them. With such a technique you don't have to worry where the center of rotation is, although you would have to use pumps to transfer liquids.I don't remember rotation being explicitly stated or implied. Is that just an assumption you made or am I not remembering something? "Attaching two spacecraft side by side" seems like the easiest way to attach them no matter what you're going to be doing, so I think that continually asserting that rotation is obvious by this orientation is a bit off.If ullage thrusting is used, a pump near the bottom of the tanker is all that's needed, and is likely less weight and almost definitely less complexity than docking the spacecraft in some strange nose-nose or nose-engine orientation to allow for "passive" refueling via ullage thrusting.
Obviously RCS would be used, has the term "ullage thrusting" applied to main engines ever in anything?Why are no pumps needed? It'll either have to fill the tanks from the bottom or run up a tube along the side to fill from the top. Why doesn't the ullage thrusting cause head pressure which fights against the "passive filling" at some point in the transfer process?The bottom drilled overflows in my fish tanks won't push water through vinyl hoses if I grab one and hold it's end above the water level in the tank, so what gives?
Quote from: Req on 08/08/2017 12:49 amObviously RCS would be used, has the term "ullage thrusting" applied to main engines ever in anything?Why are no pumps needed? It'll either have to fill the tanks from the bottom or run up a tube along the side to fill from the top. Why doesn't the ullage thrusting cause head pressure which fights against the "passive filling" at some point in the transfer process?The bottom drilled overflows in my fish tanks won't push water through vinyl hoses if I grab one and hold it's end above the water level in the tank, so what gives?It does if you boil the tank. The transfer is driven by pressure differential. The ullage thrusting is measured in fractions of a micro-g. It's completely irrelevant for anything other than settling - it does not cause any significant flow.
Side by side, yes (as seen in video), but where do you get the rotating from? Sideways thrusting would work. (Spinning it also adds a complication of working against the spin gravity)
Also, no pumps are needed. Just pressure.
Of course you have to worry about where the center of rotation is. The stress is there exists no matter how the vehicles are docked.
... Is it completely apparent/proven that maintaining that pressure differential is preferable to some pumps in terms of mass/complexity/risk/cost/etc at the desired flow rates?
Quote from: KelvinZero on 08/07/2017 10:52 pmCan anyone give some numbers on the efficiency of using acceleration? What would be the delta v over the entire operation? Is this delta-v wasted, or a tiny nudge in the direction you were going anyway? Is it totally negligible?Until someone gives numbers, for all I know it could be as small as moving the length of the tank during the refueling operation. Some simple numbers might make discussion of spinning and so on obviously not worth the bother.According to this paper, it doesn't look that bad: http://www.ulalaunch.com/uploads/docs/Published_Papers/Extended_Duration/SettledCryogenicPropellantTransfer.pdf ~10 lb/hr with a 100 mT hydrogen stage at 10^-5 g. The BFS is a lot bigger and uses a different prop, but you are still looking at only hundreds of pounds per hour. Any sort of grappling system and spin up/down propellant is going to weigh more than that.
According to this paper, it doesn't look that bad: http://www.ulalaunch.com/uploads/docs/Published_Papers/Extended_Duration/SettledCryogenicPropellantTransfer.pdfThis 10^-5 g is sufficient for settling and pumping out liquid.Pumping out liquid infers almost no flow to the liquid in the vessel, pumping in liquid does.For settling pumped in liquid there is no data, and that's probably why they purpose to do an cryogenic transfer experiment.Suppose 10^-2 g is the limit for keeping pumped in cryogenic Methalox settled (viscosity is important for every pumped in liquid, to slow down flow).Than:Acceleration of 10^-2 g for 5 hours is a delta V of 1.8 km per second.if this is done by thrusters with low ISP the propellant loss is significantly more than with high ISP Raptor engines.Until the lower g limit for pumping-in cryogenic Methalox is comfimed, discussions about rotation vs linear acceleration are relevant.
Quote from: Req on 08/08/2017 02:28 am... Is it completely apparent/proven that maintaining that pressure differential is preferable to some pumps in terms of mass/complexity/risk/cost/etc at the desired flow rates?Yes. For two reasons:1) the tanks must be made to withstand a fair amount of pressure to survive launch. Just sitting on the pad, a 10 m tall water tank will have 1 atmosphere of hydro-static pressure at the bottom (i.e. the weight of the water), in addition to the pressure at the top. Rocket tanks are also pressure stabilized, which means that extra pressure is applied to add structural rigidity. I suspect the flight pressure is around 1 atmosphere, before adding the hydro-static pressure and the ambient outside pressure (the ambient pressure isn't felt on the pad since it's balanced, but the tank feels the lack of it in space). If you listen to the technical announcers during a launch, they'll usually call out tanks being at flight pressure, just minutes before ignition. 2) tanks pressure must be actively managed on-orbit, via venting, to control propellant temperature (conversely pressure becomes a side-effect of temperature control, if a cryo-cooler is used). For a liquid stored at it's boiling point, the boiling temperature is a direct function of the pressure (between the limits of T_freeze and T_critical). So in order to sub-cool the propellants, they must be stored at somewhere between 0 and 1 atmosphere of pressure. Of course larger pressures can be applied for a short time, but the propellant will gradually warm up (unless a cryo-cooler or really good passive cooling is used).So you've probably got over 1 atmosphere of pressure differential to work with, without adding any additional equipment or cost (just software). If your ullage acceleration is 1000 ugees, you can make the propellant flow uphill for about 29000 feet, with no pumps.Note that as the propellant is transferred from the higher pressure tanks into the receiving tank with lower pressure, some "ullage" gas must be vented in the receiving tank to make room for the incoming propellant. If you want to do no-vent propellant transfers, then you either have to use a cooler to condense the extra gas in the receiving tank, or pump it into the sending tank (thru more pipes, valves, and connectors). Remember that ullage gas is 200x less dense than LOX, so throwing away a tank of ullage gas for each transfer is really only wasting 0.5% of your propellant.
Gotcha. Is it completely apparent/proven that maintaining that pressure differential is preferable to some pumps in terms of mass/complexity/risk/cost/etc at the desired flow rates?
Quote from: Peter.Colin on 08/08/2017 06:51 amAccording to this paper, it doesn't look that bad: http://www.ulalaunch.com/uploads/docs/Published_Papers/Extended_Duration/SettledCryogenicPropellantTransfer.pdfThis 10^-5 g is sufficient for settling and pumping out liquid.Pumping out liquid infers almost no flow to the liquid in the vessel, pumping in liquid does.For settling pumped in liquid there is no data, and that's probably why they purpose to do an cryogenic transfer experiment.Suppose 10^-2 g is the limit for keeping pumped in cryogenic Methalox settled (viscosity is important for every pumped in liquid, to slow down flow).Than:Acceleration of 10^-2 g for 5 hours is a delta V of 1.8 km per second.if this is done by thrusters with low ISP the propellant loss is significantly more than with high ISP Raptor engines.Until the lower g limit for pumping-in cryogenic Methalox is comfimed, discussions about rotation vs linear acceleration are relevant.You are assuming that the ullage thrusters burn continuously for 5 hours. This is totally unnecessary. They only pulse as required to keep the prop settled. Note also, it's only necessary to keep the prop settled in the tanks of the tanker during the transfer to ensure that it remains over the outlet. It doesn't matter if the prop is not settled in the receiving tank. It can be settled later after the spaceship and tanker have separated.Also, cryogenic prop has already been kept settled for the order of five hours on Delta IV GSO missions. I think all the basic technology for the mass transfer of cryogenic prop is already in place. It requires development of course. But I don't see any obvious showstoppers. And I don't see the need for rotation.
How do you vent gas if the prop isn't settled in the receiving tank?
Quote from: Peter.Colin on 08/08/2017 03:46 pmHow do you vent gas if the prop isn't settled in the receiving tank?it is settled and vented
Will the prop be offloaded from the same tanks that provide fuel and oxidizer to the tanker's own engines (like siphoning gasoline from one car's tank to another) or would the prop which is to be delivered be in separate tanks? (E.g. when diesel fuel is delivered to a fuel station, the semi tractor draws from its own tanks and the diesel payload is separate in the trailer.)If the answer above is that it will be in a separate payload tank and if multiple tanker launches are already required, this begs a new question. Would it be more efficient to bring fuel and oxidizer on every flight (thus requiring two payload tanks, two sets of pumps/transfer lines, two pumping events, etc. on every flight, or would it be better to have specialized and separate fuel and oxidizer tankers, which would mean only one payload tank, one set of pumping equipment, one pumping event, etc. per flight?It seems that separate specialized tankers might require fewer separate transfer events and potentially be less risky. The tankers themselves would be a simpler design than a dual payload tanker. OTOH, it would mean different loading procedures on the ground and tankers which may be different sizes to accommodate liquids of differing densities.A tanker which delivers both fuel and oxidizer on each flight would mean only one method of loading for each launch, one tanker design, one docking method, etc., rather than two.Thoughts?
So Musk confirmed tanker and spaceship will mate end-to-end and use ullage thrusters to settle the propellant down into the empty tanks of the ship.He said the plan is to reuse the existing plumbing from the booster to transfer fuel. But in the image here, if say CH4 is on the left and O2 is on the right, when the ship and tanker dock end-to-end, then those connections don't line up. They might if the tanker rolled over, but that's not what's shown.Discuss.
You mean the fill and vent lines? And they're bidirectional? I have no idea if the valves can support that or not.
Quote from: Norm38 on 09/29/2017 04:09 pmYou mean the fill and vent lines? And they're bidirectional? I have no idea if the valves can support that or not.They are called fill and drain lines. The tanks have to be emptied for scrubs.
Quote from: Jim on 09/29/2017 06:25 pmQuote from: Norm38 on 09/29/2017 04:09 pmYou mean the fill and vent lines? And they're bidirectional? I have no idea if the valves can support that or not.They are called fill and drain lines. The tanks have to be emptied for scrubs.Got it. So when the separate ships are fueled for launch on Earth, are they filled from the top or the bottom? From John Alan's sketch, they will fill from the bottom due to ullage thrust.
....This assumes autogenous pressurisation works, and the engines are OK with starting up with less than nominal pressure if you're transferring a lot of fuel, or some autogenous pressurisation device that works without the engines being on.I was idly wondering in the context of my silly speculations on using BFS SSTO, refuelling suborbitally, and if it only takes two minutes, plus a couple of minutes to rendevous from close flight, this might be quite plausible.
Quote from: speedevil on 03/01/2018 09:43 pm....This assumes autogenous pressurisation works, and the engines are OK with starting up with less than nominal pressure if you're transferring a lot of fuel, or some autogenous pressurisation device that works without the engines being on.I think it would much longer than two minutes. Propellant would only flow at that rate if the main engines were running and the flow was being driven by their turbopumps.
....This assumes autogenous pressurisation works, and the engines are OK with starting up with less than nominal pressure if you're transferring a lot of fuel, or some autogenous pressurisation device that works without the engines being on.
The tanks won't be at zero pressure. Main tank pressures are about 30-45 psi. Receiving tank could be dropped 10 to 20 psi below that. Once transfer is complete, pressure can be raised back up using warm GOx and GCH4. Main engines would never be operated until design tank pressures have been reestablished. Small pressure fed vernier motors would use a different tank system, probably pressurized to a few hundred psi.John
When dealing with tanks partially filled (separately, of course) with super-cooled liquid methane and with LOX, at what temperature and pressure does each freeze into a solid? I get the feeling that it's a sliding scale, depending on the amount of fluid vs. gas in each tank.I'm assuming that the freezing point goes higher as pressure decreases. At a vacuum and super-cooled temps, I'd think you'd flash-freeze a lot of each liquid...
Why would they NOT use turbopumps to accelerate fuel transfer? Spend a little fuel via turbopump, save a bit of fuel in ullage thrust, save time because the pumping is faster, and hot partially burned methalox makes for more efficent ullage.
Quote from: rakaydos on 03/02/2018 06:44 pmWhy would they NOT use turbopumps to accelerate fuel transfer? Spend a little fuel via turbopump, save a bit of fuel in ullage thrust, save time because the pumping is faster, and hot partially burned methalox makes for more efficent ullage.Tank pressure is more than enough to empty the tanks in minutes. How do you think the booster tanks empty in less than 2 minutes during ascent? Hint: the turbopumps are not "sucking" it out...
As the liquid flows from the tanker why would you not just transfer gas into the tanker from the BFS? I thought there were 4 connections between tanker and BFS. 1 supply and 1 return line for o2 and methane.
Quote from: Steve D on 03/02/2018 08:38 pmAs the liquid flows from the tanker why would you not just transfer gas into the tanker from the BFS? I thought there were 4 connections between tanker and BFS. 1 supply and 1 return line for o2 and methane.You would need a modestly large pump - of the order of 100kW or so power in order to maintain the pressure difference while pumping the return gas.And of course a power system for this to run from.
(Takes a while to wait for the orbit to pass over the launch site again while in the proper position)
Quote from: envy887 on 03/02/2018 06:49 pmQuote from: rakaydos on 03/02/2018 06:44 pmWhy would they NOT use turbopumps to accelerate fuel transfer? Spend a little fuel via turbopump, save a bit of fuel in ullage thrust, save time because the pumping is faster, and hot partially burned methalox makes for more efficent ullage.Tank pressure is more than enough to empty the tanks in minutes. How do you think the booster tanks empty in less than 2 minutes during ascent? Hint: the turbopumps are not "sucking" it out...In anything except a pressure fed rocket, yes, in fact, the turbopumps ARE sucking it out.
Quote from: Lars-J on 03/03/2018 07:27 am(Takes a while to wait for the orbit to pass over the launch site again while in the proper position)Try about a day before they come back over the launch site. The ground track of the vehicles moves X degrees for every orbit. If you site is at the poles every orbit puts you over it. At the Equator its once a day. Everywhere else is somewhere in between. Whatever site you want to land on (or rather get back to) from a given orbit depends on your cross range. The more orbits away from launch you are (to a maximum 180degress ) the more cross range you need. The design goal to return to launch site in 1 orbit (which is the easy case) consumed 10s of 1000s of wind tunnel hours in Shuttle design. This also brackets your propellant transfer time and hence rate. Maximum rate is about 225tonnes in < 90 minutes if BFS has the designed cross range of the Shuttle. Or you ignore designing for RTLS in one orbit and just wait till it comes back round, which gives you one less design constraint for BFS to meet. If other factors lead to a design with that much cross range, great. If not it takes as long as it takes.
Quote from: Lars-J on 03/03/2018 07:27 am(Takes a while to wait for the orbit to pass over the launch site again while in the proper position)Try about a day before they come back over the launch site. <snip>This also brackets your propellant transfer time and hence rate. Maximum rate is about 225tonnes in < 90 minutes if BFS has the designed cross range of the Shuttle.
Quote from: rakaydos on 03/02/2018 08:41 pmQuote from: envy887 on 03/02/2018 06:49 pmQuote from: rakaydos on 03/02/2018 06:44 pmWhy would they NOT use turbopumps to accelerate fuel transfer? Spend a little fuel via turbopump, save a bit of fuel in ullage thrust, save time because the pumping is faster, and hot partially burned methalox makes for more efficent ullage.Tank pressure is more than enough to empty the tanks in minutes. How do you think the booster tanks empty in less than 2 minutes during ascent? Hint: the turbopumps are not "sucking" it out...In anything except a pressure fed rocket, yes, in fact, the turbopumps ARE sucking it out.You are both right. Tanks are at about 30 psi in orbit so they would empty out quite fast on there own. The pumps do suck the pressure down below 30 psi, but not so much that the propellant would boil / cavitate.John
The big reason you need the pumps is that you're topping off a partially filled tank that's also already at 30 PSI.
Quote from: rakaydos on 03/03/2018 03:02 pmThe big reason you need the pumps is that you're topping off a partially filled tank that's also already at 30 PSI.Can you explain why simply venting the receiving tank to 15PSI or so does not work?
Because you spent hundreds of dollars per pound getting what you're venting up there. Better to recirculate it back to the tanker to keep THAT pressurized.
Quote from: speedevil on 03/03/2018 03:13 pmQuote from: rakaydos on 03/03/2018 03:02 pmThe big reason you need the pumps is that you're topping off a partially filled tank that's also already at 30 PSI.Can you explain why simply venting the receiving tank to 15PSI or so does not work?Because you spent hundreds of dollars per pound getting what you're venting up there. Better to recirculate it back to the tanker to keep THAT pressurized.
As far as mating the two vehicles I think a robot arm of sorts could be helpful.
Quote from: rakaydos on 03/02/2018 06:44 pmWhy would they NOT use turbopumps to accelerate fuel transfer? Spend a little fuel via turbopump, save a bit of fuel in ullage thrust, save time because the pumping is faster, and hot partially burned methalox makes for more efficent ullage.This is very much completely incorrect. Partially burned methalox makes for the worst ullage gas and pressurant you can find.What is partially burned methalox? It is largely: [*]unburnt propellant (mostly good, except when it goes in the wrong tank, in which case it's probably a disaster), [*]carbon monoxide which is slightly worse for pressurization than argon, but not as good as nitrogen(this makes it an improvement over both oxygen and methane but your production yield in combustion is likely to be low), [*]carbon dioxide, which will promptly condense to solid dry ice (it is competing for second worse combustion product, for use as pressurizer, it's vapor pressure graph is only slightly better than N2O, but not quite as good as ethane. [*]water, which will promptly condense to solid ice (it is the worst combustion product for pressurization, it's vapor pressure graph is only an improvement over zero pressure). Putting partially combusted propellant back into it's tank is not only counterproductive for filling ullage space, it is bad for propellant quality and vehicle safety. Piping partially reacted gas back to oxidizer tank is almost certainly prone to rapid unscheduled disassembly risk.If that is not a reason not to even consider turbo pumps for propellant transfer, I'm sure there are other reasons
Quote from: Hominans Kosmos on 03/03/2018 08:18 pmQuote from: rakaydos on 03/02/2018 06:44 pmWhy would they NOT use turbopumps to accelerate fuel transfer? Spend a little fuel via turbopump, save a bit of fuel in ullage thrust, save time because the pumping is faster, and hot partially burned methalox makes for more efficent ullage.... Partially burned methalox makes for the worst ullage gas and pressurant you can find....Pretty sure he meant methalox ullage settling thruster, not methalox ullage space pressurant.
Quote from: rakaydos on 03/02/2018 06:44 pmWhy would they NOT use turbopumps to accelerate fuel transfer? Spend a little fuel via turbopump, save a bit of fuel in ullage thrust, save time because the pumping is faster, and hot partially burned methalox makes for more efficent ullage.... Partially burned methalox makes for the worst ullage gas and pressurant you can find....
Quote from: rakaydos on 03/03/2018 03:22 pmBecause you spent hundreds of dollars per pound getting what you're venting up there. Better to recirculate it back to the tanker to keep THAT pressurized.Which has to be set against the mass of the pump.
I was thinking of this thread while I was pondering a filling soda cup at the convenience store yesterday, that was foaming so much it was taking forever. People were staring at me. It's probably a good thing they didn't know what was going on in my head.
Hypothetical: One or more of the main raptor turbopumps has an alternateplumbing connection to be used for refueling.Use: Pump pressurant out of recieving tanks into source tanks to create pressure differential for fuel transfer.How hard would it be to keep ahead of that kind of pressure differential, with fuel/oxygen flowing out of the tubes we have seen?
Other than the extra weight of 10m or so of pipe, there is little reason to fill the receiving tanks from above the liquid level.