Described herein is a concept (only) for a
facility in low Earth orbit for the assembly and fueling of interplanetary
vehicles requiring hyperbolic departure (and arrival), in particular those associated with space-tug
assisted departures and arrivals. Such a
facility need not be a 1-to-2-decade long international project to build, if it is docked together out of modules that
fit within the payload spaces of the current launcher fleet! That should be easily achievable for a “clean
sheet of paper” design like this!
For lunar missions,
the departure from LEO is not hyperbolic, although it is elliptic at very-near-escape
perigee speeds! Depending upon the
choice of the extended departure (and arrival) ellipse, the LEO departure velocity requirement for a
lunar mission can be reduced to near-zero,
with the space tug assuming most or all of that velocity requirement
just getting the craft onto the ellipse.
No calculations have been made, these results are concept only, as is perfectly reasonable at this early stage! The basic design concept has two core
sections, one made of pressurized
modules docked together, and the other a
truss core to which a multitude of propellant tanks are attached.
Attached to one end is the Power and Propulsion Section, where solar electricity is made, stored,
and distributed. This section
includes propulsion sufficient to address the needs for countering orbital
decay, conducting debris avoidance, and performing end-of-life safe disposal.
The pressurized-core section is the Vehicle Assembly and Refueling
Section, where interplanetary vehicles
are assembled from modules, mated to
space tug vehicles as appropriate, and
fueled-up from the propellant depot section for the relevant missions. Many remote-operated arms similar to those used
at the International Space Station (ISS), and previously on the old Space Shuttle, are installed to make vehicle assembly and
handling operations as safe and easy, as
is possible. This is a manned
microgravity facility, probably manned by
rotating crews, as is the ISS.
The truss core section has multiple propellant tanks
attached to it, with the propellant feed
lines and electric power lines housed inside the truss. This is the Propellant Depot Section, presumed to be kept supplied by
unspecified tanker flights up from Earth. It would have both cryogenics and
storables, to supply a variety of
on-orbit needs. Its capacity is also
easily expandable.
The concept for the Vehicle Assembly and Fueling Section
is sketched in Figure 1. The
concept for the Power and Propulsion Section is sketched in Figure 2. The concept for the Propellant Depot
Section is sketched in Figure 3.
All figures are at the end of this article.
This kind of a facility would be easiest to keep supplied, if located in a low-inclination eastward Earth
orbit. That presumes vehicle modules, propellants,
and supplies are shipped up from the surface. It would clearly be useful in any event, but it is an essential enabling item for
making use of reusable space tugs for elliptic departure and arrivals, as described elsewhere in Reference 1.
Vehicle Assembly and Fueling Section
As the sketches in the first figure indicate, this facility is built up from many
cylindrical modules docked together, and
each is to be small enough to fit in the payload spaces of the existing
launcher fleet. Some of these are
oriented along the section axis, and the
others are perpendicular to it, but all are
in one plane. These could be either
hard-shell modules, or inflatables with
hard structural cores, or a mix of both
types. That choice remains unspecified, at this time.
The modules along the core axis provide much crew living
space, lots of storage space for life
support and other supplies, airlocks for
space-walk activities, plus any
equipment for Earth observations (potentially replacing those functions after
the ISS is decommissioned). These modules
would be equipped with external cradle mounts, to help hold the vehicles being
assembled, thus freeing up the arms for
other tasks that are part of the assembly process. Some of the hatches should be closed, when the modules are not in use by the
crew.
The modules perpendicular to the core provide the spaces for
the arm operators to work. The arms are
affixed to the module ends. These modules
need large windows, by which the arm operators
can see their workpieces in order to work.
Assembly work areas are disposed along this section, on two opposite sides. It should be able to handle a busy traffic
load, if arranged in this way.
Per Reference 2,
I am suggesting that this section’s internal atmosphere follow the “Rule
of 43”, that being a two-gas oxygen-nitrogen
system, at 43 volume percent
oxygen, and 43% of a standard atmosphere
total pressure. This is very close to
the best atmosphere that I found (which was 43% oxygen, 43.5% of an atmosphere total pressure), and it is easier to remember! It has the same oxygen concentration (as mass
per unit volume) as sea level Earthly air at 70 F, so the “predicted fire burn rate danger” from
the usual Arrhenius overall-chemical rate equation, is no worse than that down here on Earth, at sea level on a warm day.
Further, the “pre-breathe
criterion” allows no pre-breathe requirement be imposed for donning pure-oxygen
space suits, of helmet pressures down to
as low as only 3.002 psia (155.2 mm Hg)!
That criterion says the ratio of nitrogen partial pressure to suit
oxygen helmet pressure, may not exceed
1.2, in order to avoid the nitrogen blow-off time otherwise required. (The absolute minimum tolerable suit pressure
for functional cognitive capability is 2.675 psia (138.3 mm Hg), before applying a 10% leak-down factor. The cognition margin is very slightly
negative once leaked down.)
As a further bonus,
the proposed oxygen partial pressure is the same as that at about 2500 m
altitude, so there should be no
long-term hypoxia risks, or even any reproductive
health risks for female crew, based on
centuries of human experiences living up to that altitude, but not above it.
Power and Propulsion Section
Most likely, the
“best” propulsion choice for this application is a hypergolic storable
bi-propellant system, pressure-fed for
the greatest engine simplicity and reliability.
Tanks would be bladdered, with
inert gas (helium) expulsion at effectively the feed pressure to the
engines. Propellants would likely be
nitrogen tetroxide (NTO) oxidizer and monomethyl hydrazine (MMH) fuel, although the other hydrazines could also
serve, which include plain
hydrazine, unsymmetrical dimethyl
hydrazine (UDMH), and Aerozine-50 (a
50-50 blend of plain hydrazine and UDMH).
These tanks would need a thin layer of insulation topped with a very
reflective aluminum foil, plus electric
tank heaters to prevent freezing while shadowed.
There is a core module to this section that connects to the
rest of the station on one end, and the
engines and their propellant tanks on the other. It would have multiple “fins” mounted to the
sides, some being waste heat
radiators, the others being solar
photovoltaic panels. There would be
controls, batteries, and distribution switching equipment inside, plus a docking module for capsules bringing
crew and supplies. This core module is
pressurized for easy access, but the
hatch into it should be closed, when crew
are not working in there.
Propellant Depot Section
This section has a modular truss core containing the
multiple types of propellant feed lines,
and the necessary power lines.
The propellant tanks are mounted to its periphery, as sketched in the third figure. There are basically two types of propellant
tanks, those equipped to store and
deliver cryogenics, and those equipped
to store and deliver storables. Each
propellant species must have its own line fittings, not interchangeable, so as to prevent incorrect hookups (which
would most likely be disastrous). There
are no pressurized modules in this section.
For the storables (which includes rocket-grade kerosene RP-1), the bladder in the tank provides the means to
transfer propellant, driven by inert gas
pressure that everts the bladder, as
indicated in the third figure. These
tanks will also need some insulation topped by reflective foil, and some in-tank heaters, much like the tanks on the Power and
Propulsion Section. The difference is
that the inert gas pressure can be lower,
just enough to drive the transfer,
and not at all far above the level to prevent “hot room temperature” boiloff.
The cryogenics are different, in that there are no feasible bladder
materials that could survive eversion at cryogenic temperatures. These have to be metal tanks with no
bladders, although they do need a layer
of insulation topped by reflective aluminum foil, plus cryocooler equipment.
In zero gee, the
propellant will initially be free-floating globules, eventually settling into a thick film coating
the entire inner surface of the tank with a vapor void up the core, but with no pressure other than enough inert
gas pressure to stop boiloff. The
slightest touch causes the thick film to break up into free-floating globules
again. Up to now, the only way to control this behavior into a stable
pool from which a pump can draw suction, was to use thrusters to accelerate the
vehicle. You can’t do that with tanks
on a space station whose orbit you do not want to change.
I had previously come up with the spinning tank concept to
fling the propellant to the outer wall by centrifugal force. From there,
a pump could draw suction from openings along the tank sides instead of
one end. This was conceptualized as the
vehicle docking with the tank, in turn
undocked from the station. The docked
vehicle and tank would move away to a safe distance and then spin-up in
“rifle-bullet” mode, to fling tank
contents to the outer walls. Then pumped
transfer could take place, followed by
de-spin, then redocking the tank with
the station, and finally undocking the
vehicle from the station’s tank. While
this would work, it does involve
multiple docking operations, and
spin-up/de-spin of some massive objects.
But it was better than trying to store spinning tanks on the station. This concept was described in Reference 3.
I have since revised the concept to just spinning the
propellant inside the stationary tank,
by means of moving vanes inside the tank. The suction pickups remain on the outer
periphery. If you use a pair of
counter-rotating vane sets inside the tank,
separated by a perforated baffle,
you can avoid gyroscopic forces being applied to the station. This
concept is shown in the third figure.
There are no dock/undock operations associated with a
propellant transfer by this means. The
vanes can be spun by a coaxial counter-rotating shaft assembly, entering one end of the tank through a gland
seal, with the drive motor left
accessible for repairs and replacements.
You just spin up for the transfer.
Otherwise, nothing moves. This is also the least mass to spin up, reducing the energy requirements for
spin-up/de-spin, and eliminating any and
all thruster operations.
I put the oxygen (LOX) tanks closest to the Vehicle Assembly
and Fueling Section, because that is the
largest species volume being used, and
that shorter length minimizes the power line losses for the motors powering the
liquid spin. As the third figure
shows, I put the cryogenic fuels
hydrogen (LH2) and methane (LCH4) adjacent to the oxygen, because their volumes are also large, to reduce transmission line losses a bit
further. I separated the NTO from the
MMH with the storable RP1, in order to
minimize the possibility of spilled hypergolics coming into contact, even in vacuum. That is a crucial safety consideration!
The truss can be extended further, with additional tanks installed, either for other propellant species, or for additional capacity, or both.
This is because there is no other section to the station beyond the
Propellant Depot Section.
I think this “spin the propellant inside stationary tanks”
concept may be easier to develop and implement than the alternative “each tank
is its own syringe” concept, because (1)
the rotating-shaft gland seal technology already exists, even for cryogenics, (2) the required piston seal concepts and
associated leakage recovery concepts for the “syringe” do not yet exist for
cryogenic fluids, and (3) the
tank-and-equipment masses and dimensions would be lower: vanes and motor vs a syringe piston and its driving
equipment. The hardware has to ride up
to LEO inside existing payload spaces,
after all!
Conclusions
#1. A combined
vehicle assembly facility and propellant depot in LEO could enable all sorts of
interplanetary missions very effectively,
and even missions to lunar orbit,
plus replace the Earth-observation functions that will likely cease for
a while when the ISS is decommissioned.
#2. This type of LEO
facility is an enabling item to put an effective space tug operation into
effect, that uses elliptic departures
and elliptic arrivals, to reduce the
velocity requirements of interplanetary (and lunar) vehicles.
#3. This kind of
fueling operation could use “spin-the-fluid-in-a-stationary-tank” to reduce the
overall energy requirements of propellant transfer, eliminate any need for the use of ullage
thrusters, and also eliminate many
dock/undock operations.
#4. All the other
technologies required to build this thing already exist.
References (use date and title in the archive tool on the
left, to access quickly):
#1. G. W.
Johnson, “Tug-Assisted Arrivals and
Departures”, posted to “exrocketman” 1
December 2024.
#2. G. W.
Johnson, “Refining Proposed Suit and
Habitat Atmospheres”, posted to
“exrocketman” 2 January 2022.
#3. G. W.
Johnson, “A Concept for an On-Orbit
Propellant Depot”, posted to
“exrocketman” 1 February 2022.
Figures:
Figure 1 – Concept Sketches For the Vehicle Assembly and
Fueling Section
Figure 2 – Concept Sketches For the Power and Propulsion
Section
Figure 3 – Concept Sketches For the Propellant Depot Section
Update 5-4-2025:
Conversations with a friend led
me to understand that what I have in mind for the vane-equipped tank may not be
readily apparent to the reader. Please
see the sketch in Figure A below,
as you read the following more detailed descriptions.
There are just two sets
of vanes inside the tank, mounted on shafting that causes them to
counter-rotate. Their tips spin circumferentially, but in opposite
directions (which avoids applying gyroscopic forces to the depot
station). There is a perforated baffle between the two sets of vanes so
that the two volumes of fluid which are affected by the vanes, also rotate circumferentially, independently in each section.
Yet the baffle is
perforated, so that the radially-measured
levels of the fluid, flung out to the
tank wall, are equal in the two
sections. It is one shaft, with a gland seal at one tank
head, and an internally-mounted bearing at the other.
There is a gear box near the middle that makes the shafting turn in opposite
directions in the two sections. There are probably 4+ vanes in each
section, mounted to the shafts. If you forgo “instant”
response, these vane and shaft
assemblies can be fairly lightweight construction.
The propellant pickup is
along one side of the tank, not one or the other end head, since the centrifugal forces will fling the
propellant to the outer cylindrical wall, forming a big hollow
cylindrical "form" in each of the two sections, as wetted to
the local outer wall. These propellants are moving in opposite directions
circumferentially in the two sections, induced to do so by the
spinning vanes. But that circumferential motion does not really affect
the suction drains along the tank cylindrical wall!
You spin the vanes to
withdraw propellants, but you need no spin to pump propellant into the
tank. I put the drive motor outside the propellant tank for its safety
(remembering the in-tank stirring-fan device that caused the explosion on
Apollo-13), and for easy maintenance and repair. Cryogenic gland
seal technology already exists, in
rocket engine turbopumps. The vane shaft
motor and the propellant withdrawal line are on the end that attaches to the
core structure of the orbital propellant depot space station. All the power lines and fluid delivery piping
is inside that core.
This is a heavier
solution than ullage thrusters, so this is definitely only for a
propellant depot in orbit (where you do not want to disturb the orbit with
ullage thrust), not the vehicles it is supposed to supply with
propellants on orbit.
However, per the
not-to-scale concept sketch in Figure B below, it might be "just the thing" for
the "payload" propellant tanks of a dedicated tanker vehicle sent up
to supply this depot station. Those will be rather small compared to
the rest of the upper stage delivery craft,
in turn small compared to its launch booster. The sketch is not to scale, in order to provide clarity about which tanks
are vane-equipped, and which are not. Ultimately,
this tanker needs to be a fully-recoverable vehicle.
Figure A – Concept Details for Cryogenic Vane-Ullage
Propellant Tank
Figure B – Concept Sketch for Dedicated Tanker Vehicle
Update 5-27-2025:
I extended my investigations into some preliminary design
analysis to determine some “typical” design characteristics for spinning-vane
tanks of the counter-rotating type. I
ran my numbers in metric, since that is
what most space industry people are now using.
It is customary to use rotation speed measured in
revolutions per minute, rpm. The conversion to or from radians per second
looks like this:
ω, rad/s
= (N, rpm)*(2*pi rad/rev)/(60 sec/min)
or N, rpm = (ω, rad/s)*(60
sec/min)/(2*pi rad/rev)
The radial acceleration component “a” felt at the radius tip
for a rotation speed ω is:
a, m/s2
= (R, m)*(ω, rad/s)2 or ω,
rad/s = [(R, m)/(a, m/s2)]0.5
You divide that “a” by the standard value of g = 9.80667 m/s2
to express that acceleration in “standard gees”, as n = a/g,
which is often also customary,
and dimensionless.
The distance s that something falls through, under constant acceleration a, depends upon the square of the fall
time. Radial acceleration is not
constant, but we will ignore that:
s =
0.5*a*t2
The floating spherical globules are everywhere throughout
the tank cross section, so the average
distance through which a globule must fall is R/2. Substituting that into the previous equation
and solving for a fall time, we have:
t =
(R/a)0.5
which is a sort of time constant tconst for the
process of all the globules accelerating outward along the radius. (Note that for axial “ullage”
acceleration, you would use the tank
length L instead of its radius R for a measure of the average fall length s = L/2
in the same equation for time constant.)
The process is quite stochastic, but the usual “rule of thumb” is to wait 3
time constants, and then the process is
pretty much completed. Thus:
tsettle ~ 3*tconst
Figure C shows N, rpm versus n, gees, parametric upon values of R from 1 to 5
meters. There is clearly a wide range of
possible values. The corresponding
settling time estimates versus n, gees,
parametric upon R are given in Figure D just below.
Figure C – Results for N, rpm, Versus n, gees, Parametric Upon R, m
Figure D – Results for tsettle, sec, Versus n,
gees, Parametric Upon R, m
Amazingly enough,
those tsettle values actually formed a tight band of
solutions across a fairly wide range of R from 1 to 5 meters! If one selects n = 0.1 gees, the settling times all fall 3 < tsettle
< 7 sec! So, the design problem is rather conveniently
bounded at about n = 0.1 applied rotational gees, with any conceivable settling times under
about 10 sec!
Applying that revelation,
I plotted the necessary rotation rates N, rpm, versus the tank radii R, m, for only 0.1 radial gee. Those data are given in Figure E, and indicate a bound on the rotation rate as
4 < N < 10 rpm, for all radii from
1 to 5 meters! That rather conveniently
bounds the rotation rates to rather low values.
Tip speeds are also closely bounded as about 1 to just over 2 m/s, also low numbers.
Bear in mind that if you wait about 10 sec, you have all the settling times covered! That is a rather short settling time! The smaller value of 3 sec occurs at the
smaller R = 1 m, so yet smaller tanks
are no problem at all. At this time in
history, it would be difficult to
imagine successfully transporting to orbit a tank much larger than 10 m in diameter, so we have that end of things fairly well covered, also!
And the highest tip speed is less than twice the max flow speed
recommended for liquids in pipes. We
should be okay there!
Figure E – Required Rotation Rates vs Radius at 0.1 Gee
This Figure E and a nominal settling time of 10
seconds pretty well has most conceivable tank sizes “covered” for designing-in
counter-rotating vanes. Such
would require no dock and undock operations to transfer the propellants, and would apply no unwanted forces due to
spin reactions or gyroscopic resistance forces to whatever structure or vehicle
such a tank is mounted upon (precisely because there are counter-rotating sets
of vanes). The pump suction points are
on the tank lateral wall instead of the ends.
So what? The tank itself does NOT
spin! So, there is no balance problem, other than the vanes themselves, when you build one of these tanks. You just spin up the vanes, wait about 10 seconds, then start your transfer pump. How simple is that?
Again, such would
enable cryogenic propellant storage and transfers from an orbiting propellant
depot station, to any vehicle departing
from there. Such could also be the
“delivered payload” tanks of a dedicated supply tanker vehicle, arriving at that same station. The implications for making deep space
missions cheaper with reusability, and
cheaper still by means of space-tug-assisted elliptic departure, are simply staggering!
