Moonshot rocket
This is initial proposal for a crewed lunar program. It is to be based on a large, partially reusable rocket, that is relatively simple to manufacture, placing lightweight two-person spacecraft (lander) in low lunar orbit (LLO). Spacecraft consists of one engine and only single pressurized crew compartment (but cosmonauts have individual pods with life-support that would keep them alive even after depressurization event). Propellant tanks needed to leave LLO, along with heat shield are stored in orbit while spacecraft lands on the lunar surface. Tanks needed for initial phase of descent are discarded before touchdown.
Launch vehicle
Launch vehicle is a five-stage rocket able to put at least 18.5 tons of payload into low lunar orbit (it has ∆v of 14 km/s). Gross mass of the rocket is 1 793 tons, while dry mass is 257 tons. Each stage consists of 3 spherical tanks that decrease in diameter the higher they are placed, and are able to fit inside conical frustum. Conically-shaped rocket has its center of pressure located more backwards, making it more aerodynamically stable. This is similar to setup to the Soviet N1 rocket, but I would like to use LH2/LOX propellants in all the stages (just like Delta IV Heavy does). Two lower tanks hold LH2, while upper one stores LOX. General idea is to make extremely large rocket (it has a height of 139m without payload, 14.3m diameter at the bottom [excluding some external hardware] and diameter of 3.6m at the top) that uses propellants able to provide high specific impulse. During calculations of ∆v I assumed quite conservative values of effective exhaust velocity (3 192 m/s for first stage, 3 877 m/s for everything else), so this launch vehicle may perform even better if even marginally decent engines are used. You can find the calculations in the “rocket_o.ods” file.
Spherical tanks are lightest possible pressure vessels which is important because I would like to make this rocket big semi-dumb booster that stores propellants at 500 kPa pressure, reducing turbopumps requirements and chances of cavitation. Tanks could be tiled by “pentagons” forming “regular dodecahedron”. To form a “pentagon” flat sheet of metal is transformed into spherical cap by metal spinning technique and then unnecessary material is removed. “Pentagons” are then joined with the use of friction stir welding. LH2 tanks are connected by hyperboloid of one sheet that smoothly transforms into the spheres (details are contained inside “hyperboloids” folder). Inside the tanks baffles preventing propellant sloshing and vortex creation must be installed.
Long propellant lines should have multiple valves across the whole length of the line, so hammering during engine shutdown cannot occur (this might have been the cause of N1 7L launch failure) and increase of hydrostatic pressure resulting from the acceleration can be mitigated. Propellant lines, wherever possible, should go downwards, without any making any “U” shapes. Verniers will be constructed similarly to the ones in RD-108, so in places where liquid must go down, then up, and then down again, valves could be installed through which gas can escape before launch (some of them could be integrated with small electric pumps that recirculate LH2 and LOX and prevent these liquids from boiling in the propellant lines).
Propellant tanks are suspended inside gridshell (lattice-shell) consisting of “geodesic domes” next to the tanks that transform smoothly into hyperboloid structures (calculations are contained inside “hyperboloids” folder). You can see cross-section of this concept in the “d_07.pdf” file. To make this large gridshell made up of triangular elements I propose that rings forming parts of the structure could be made by centrifugal casting process and then welded together.
Outer “skin” of the rocket is a lateral surface of a cone and should be made from thin sheets of metal. Preferentially it it should be reinforced by stiffening ribs and form an isogrid (some cheaper technique that adds material should be used, orthogrid may also be acceptable). It is is especially important in the cases of lower parts of the stages, as outer ”skin” there could be used to support some weight. Space between tanks and outer “skin” should be filled with thermal insulation.
Thrust structure consists of two large rings located bellow bigger LH2 tanks connected by great-circle arcs (details can be seen in “d_07.pdf” file). Outer ring is connected directly to the bottom of the spherical-hyperboloid gridshell. Combustion chamber/nozzle assemblies (I want to reuse classic Soviet design of singe turbopump supplying propellants to the multiple combustion chambers, some of them being parts of verniers) are connected to the rings by the heavy-duty struts. In cases of low-lying vernier thrusters I would like to place hyperboloids between them and the struts. Drawings in “d_07.pdf” file are just the representation of general idea, and in reality more connections between various parts of the rocket would be needed.
Most of the rocket would be made out of aluminum alloys. Aluminum 2219 could be used in propellant tanks and other components that will be in contact with liquid hydrogen or oxygen (it was used in Space Shuttle Standard Weight Tank and SLS core stage). SSA000 and SSA018 alloys seem promising at first glance, but they contain large amounts of magnesium which may increase the embrittlement and they were not tested at temperatures close to absolute zero. Titanium alloys can degrade in presence of hydrogen as well. Austenitic stainless steels are good choice for a material that has contact with LH2, however they have low specific strength.
First stage of the launch vehicle is designed to be recoverable and reusable. It is to be accomplished in similar manner to how Falcon 9 lands. There are four grid fins located on the side of the rocket, but they are mounted at the bottom of the rocket during launch (just as it was done in N1). After the main burn is completed, when stage is in low-pressure environment, grid fins are unfastened (along with the motors that control them) and travel to the top of the stage on a long linear slide and are firmly attached to the stage there. Central engine can be used to hoverslam, but it can control roll as well as pitch and yaw because is made out of 4 “verniers”. There are 12 landing legs in my design. 4 of them extend in linear fashion from the bottom of the stage and are supposed to support most of the vehicle’s weight, while 8 of them are located on the outside and are similar to the ones found in Falcon 9. Their job is to prevent vehicle from tipping over after landing. First stage of my rocket is only 5% longer than first stage of Falcon 9 but its bottom has 15 times more area while empty weight is 6.6 times larger for my first stage. Diameter of my stage is decreased at the top, so its shape is more similar to a reentry capsule. Bottom of the stage is more flat, just like in in rockets derived from R-7 Semyorka. All of those features should improve stage’s performance during atmospheric reentry (but large frontal also area increases drag during ascent).
Second stage potentially also can be reused, but it would be more challenging as it would achieve velocity of 3-4 km/s, so for now, I did not not included this concept in my calculations and drawings. It would require it to be heavier and not use all of its propellants on accelerating upper stages. If more efficient engines than the ones that I assumed in my calculations are used and/or dry mass of the stages can be reduced, reuse of the second stage is worth contemplating. Better thermal protection and passing of coolant through sensitive components would be needed. Retractable grid fins would have to be installed at the top.
In case that a country that does not have good location to build a conventional cosmodrome would like to build and launch those rockets I propose that they should be vertically stacked on top of a “mobile launch platform” that sits atop semi-submersible marine vessel equipped with additional ship stabilizing gyroscopes that is able to travel near the equator. Transporting tallest rocket ever on top of the vessel traversing ocean for weeks might be a challenge, but there are precedents for sticking long vertical objects on top of the ships (I’m talking about masts of sailing vessels).
When launch vehicle arrives at the equator, another ship docks with “mobile launch platform”. It is equipped with a nuclear reactor and hardware necessary to produce LH2 and LOX. Fueling of rocket takes takes 24-48 hours. During that time ship’s company evacuates to a safe distance (accidents sometimes happen). In case of a crewed flight, cosmonauts return to the main ship on a speedboat and quickly enter the capsule. First (and possibly second) stages land on a automated floating landing platform.
It is important to make mass production of the large components of the rocket, their assembly into stages and final stacking centralized and located in one place. Scott Manley attributes failure of projects like xEMU or Ariane 6 to delegation of various task to multitude of different organizations that are separated by vast distances and have competing interests. There is a Polish proverb: “gdzie kucharek sześć, tam nie ma co jeść” Its translation into English is: “where there are six (female) cooks, there is nothing to eat”. English equivalent is: “too many cooks spoil the broth”.
Engines
As for the engines, in first and second stages there are five turbopumps (per stage). The one in the center supplies propellants to four inner vernier thrusters (4.01 MN of thrust, 1st stage). Each of the remaining turbopumps is connected to the two large stationary combustion chamber/nozzle assemblies and one outer vernier (5.64 MN of thrust, 1st stage). Each of the remaining stages uses only one turbopump. I actually did not think too much about details of turbopumps, but I did notice that engines using LH2/LOX usually either have separate turbopumps for the LH2 and LOX (RS-68 and J-2) or a fuel pump directly coupled to a turbine with oxidizer pump connected to the main shaft via transmission (HM7B and RL10). Nevertheless, Soviet RD-0120 engine used single-shaft turbopump and I would like to use this “simple” configuration, but working in even simpler gas-generator cycle. To make the design simple, cheap and reliable I would like to achieve chamber pressures of only 4-5 MPa. Pressure at the output of the pumps needs to be 30-50% higher than that.
When it comes to the construction and cooling of the combustion chamber/nozzle assemblies I would like to take a novel approach based on free-surface jet impingement (drawing can be found inside “engine” folder). Assembly consists of 3 walls. Inner one is made of copper alloy, either containing 3% silver and 0.5% zirconium (NARloy-Z used in RS-25) or ~3% chromium (typical for Soviet engines). Middle and outer walls are made of high-strength material (most likely decent aluminum alloy) compatible with LH2 that fills the void between them. Middle wall is perforated and jets of LH2 shoot out from it and impinge onto inner wall, cooling it in the process. Most of the heat is absorbed by boiling of the fuel, so holes have to be spaced in a way that equally distributes liquid throughout copper alloy surface. Some fins could be milled into copper alloy to allow heat transfer into gas that carries some unvaporized fuel droplets, best place for them would be top of combustion chamber where heat flux through the wall is still quite low and fuel moves fast. Some protrusions in the form of thick fins (they need to be well cooled) that reinforce inner wall may be beneficial, especially around the throat, allowing inner wall to better withstand thrust produced. Inner wall could be covered in orthogrids (rings would have to have holes in them that allow propellant to pas through). Walls could be made by centrifugal casting, but middle and outer wall have to be made out of at least two parts that are then welded together (unless they do not become narrow around the throat). Walls could be connected together by protrusions coming out of them, increasing their strength and resistance to buckling (this is combination of inflatable mattress and keyboard mounting in cheap laptops technologies). “Large” diameter rings would protrude inward from the outer wall, smaller rings would protrude outward from the middle wall and yet smaller-diameter cylinders would protrude from the inner wall. All of the protrusions must have large clearances that will allow them to fit together. Unfortunately this system will increase amount of individual sections from which outer and middle walls are made of, because walls are not flat, so angle of protrusions changes, preventing simple installation of large sections. All of the walls have to be joined at the top, bottom and the circular protrusions, possibly by dissimilar friction stir welding. Empty spaces between protrusions may have to be filled with thin rings or powdered metal before welding (it might be necessary to bind the powder in some way before that). Or not, as friction stir welding will leave a hole that has to be later repaired (alternative methods). Manufacture of the rocket engine with the process described above will not require brazing of different materials in order to form internal cooling channels.
If possible, bottoms of the nozzles should be radiatively cooled (most likely application of this would be engines of upper stages that are optimized to work in the vacuum). If radiative cooling is not sufficient, small tanks with water and separate H2O turbopumps could be used to provide film cooling for some parts of the nozzles. Water could also be used to reduce temperature of gas entering turbines and to cool other sensitive components (H2O’s heat of vaporization is 2257 J/g, while for H2 it is 447 J/g, so even if water would be used to cool demanding parts of the engines and then dumped without providing any thrust, gross mass of the rocket would increase by only 2.49%). Water entering gas generators may have to be heated before entering them (by their exhaust in steady-state operation) in order to prevent freezing. Exhaust from turbine could be used to cool lower parts of nozzles as well.
First stage engines will require small nozzle expansion ratio. In the case of 4 MPa chamber pressure, expansion ratio must be be bellow 20:1, otherwise flow will undergo separation during sea-level operation. Nozzle shaped like the one used in RS-25 allows engine to work even when flow is grossly overexpanded. You can perform some basic calculations in “nozzle_4.ods” spreadsheet.
Initially I considered combustion chamber/nozzle assemblies with additional inlets for gaseous hydrogen (GH2) separated from main hydrogen flow by reed valves and a sleeve valve before injector section (this design can be seen in “engine_5 (gh2_inlet_included).pdf”). For initial starting of the engine, boilers would supply GH2. Boilers in the first stage of the rocket would in most extreme scenario have to provide around 633 MW of heat (power needed to vaporize all hydrogen needed for full-thrust operation, extrapolation from Merlin 1C cooling system heat absorption) or store hundreds of cubic meters of already vaporized hydrogen (engines need 1250 kg of H2 per second during operation at full thrust). Hundreds of megawatts may seem like much, but combustion of all the oxygen supplied to the engines releases 106 GW of heat (full-thrust, Poland on average consumes 147 GW of primary energy). Turbine of RD-170 produces 170 MW of mechanical power and 3.52 of those engines would be needed to provide thrust for the first stage of my rocket. Obviously usage of extreme boilers should be avoided. Heat already present in the walls of the engine may be enough to initially vaporize some portion of LH2, as was the case in the RS-25 (as described in this document, it is worth reading to get general grasp on difficulties stemming from LH2 usage in rockets [spaces through which H2 will flow have to be purged with He in order to prevent solidification of air, LOX channels must be purged with pure N2, propellants have to be constantly recirculated before the launch to prevent formation of gas pockets]). Unfortunately, exploitation of this phenomena (boiling of LH2 in contact with room temperature metal) resulted in large oscillations and slow start-up sequence. But in my engine LH2 would flow not into narrow cooling channels but into large volume around combustion chamber and nozzle, so flow of LH2 would (probably) not change direction after initial opening of the valve. 25 tons of copper (estimated amount in 1st stage engines, specific heat 0.385 J/(g*K) ) absorbing full-thrust heat (633 MW) would increase temperature at the rate of only 66 K per second (or cool by that amount if it would lose heat to the boiling LH2).
Nevertheless small boilers may be beneficial as GH2 and GOX produced by them could be used for various purposes, like powering gas generators or small RCS thrusters. I imagine that 12 thrusters would be located on the sides of the stages that are to land back on Earth or will be performing orbital maneuvers. 4 clusters consisting of 2 horizontal (meaning that they can release jets in the direction of the horizon when rocket is at the launchpad) thrusters placed back-to-back would be located at the bottom of the stage and separated by 90 degrees. Above each of those clusters, at the top of the stage, 4 thrusters would be located (separation between then would be again 90 degrees) that have their backs (combustion chambers) pointing in the direction of rocket’s axis of symmetry. This setup would allow control of 3 axes of rotation (roll, pitch and yaw) as well as translations in 2 perpendicular directions. Additional thrusters have to be placed at the bottom surface of the stage to act as ullage and separation motors, and to provide translation control in the third direction. Liquid propellants located at the bottoms of the boilers (close to low-temperature inlets) could be used to cool turbopump's bearings (that have to be mounted into elastic damper supports, note that those bearings may require frequent replacements if they do not have proper coatings). Boilers would be primed before launch by small electrical pumps. One has to be wary of liquefaction possibility that passing of cool GH2 or GOX through valve poses (Joule–Thomson effect).
Larger combustion chambers will most likely require baffles. Most efficient way of realizing them seems to be installation of baffle injectors. To achieve fast and reliable ignition I would like to put pilot flame tube inside each baffled area (although nobody seems to be using this configuration, so maybe some problems are inherent to it). At the top of the pilot flame tube spark plug (created by the 2 rods coming out from the opposite sides of the tube) and a single GH2/GOX injector (simple concentric one should do the job) would be present. Initially, almost stoichiometric ratio of fuel and oxidizer would be admitted (wall perforations between concentric injector and spark gap could be used to supply additional GOX during almost stoichiometric operation, gasses for initial operation could be stored at elevated pressure in small tanks), plasma would be created at the spark gap. Typical automotive electric spark is obvious way of generating plasma, but maybe decent bipolar Tesla coil producing electric arc is a better option (brush discharge could possibly be used as well). After mixture of gases starts to burn energetically and propellants in the combustion chamber are ignited as well, flow of GOX is decreased and flow of GH2 increases. This would reduce temperature of a pilot flame to a value that does not melt hardware. Additional compressors may be needed to increase pressure of gases and assure flow of pilot flame into the main combustion chamber, but if the tube gets wider closer to the combustion chamber, pressure would increase as well (tube would form a diffuser that requires high speed of the pilot flame).
Spacecraft
Capsule and service module containing only one LH2/LOX engine that consists of single turbopump and 4 combustion chamber/nozzle assemblies are connected by two long linear actuators located on both sides of the spacecraft. Between the capsule and service module two detachable propellant modules (always located above the engine) and heat shield are initially located. Landing module consisting of extendable legs and some additional equipment that may be used on the lunar surface is attached to the service module.
Propellants (especially LH2) will require constant cooling during long missions. System utilizing reverse Brayton cycle with almost isothermal compression (and possibly Joule-Thomson expansion) could be used. Hydrogen could be used directly as as working fluid or helium with high heat capacity ratio (this increases efficiency and reduces required compression/expansion ratio) in separate loop could be used as well. I am developing similar (but less extreme) cryocooler for the negative emission ammonia engine and I will write more about it at the later date. If system with mechanical turbines and compressors is not feasible, pulse tube refrigerator could be used instead.
After spacecraft is placed in LLO, it detaches from the fifth stage of a launch vehicle and extends landing legs. This extension is achieved by linear actuators located at the top of the legs (2 of them per leg). Legs are then locked in place and small shock absorbing elements extend from the bottom of the legs. Those elements are to be flat, elastic, slightly curved pieces of lightweight material similar to the running blades of Oscar Pistorius.
After that, departure module that consists of heat shield and propellant tanks required to escape from lunar orbit is unfastened from the rest of the spacecraft. Modules are connected to each other with screws that have integrated compact electric screwdrivers and small linear actuators that can be used to make or break connections between the modules. Large linear actuators connecting capsule and service module increase their length and make enough space for the departure module to be able to separate from the rest of the spacecraft using cold gas thrusters. Next, capsule firmly attaches to the descent module.
Later, perilune is reduced and when the spacecraft gets close to the lunar surface, powered descent is initiated. Liquid propellants flow from the descent module and constantly replenish tanks in the service module. When descent module is finally empty, engine stops producing thrust (turbopump keeps spinning), and empty tanks are “ejected”. Final phase of descent is performed with the propellants located inside service module. This is risky maneuver, so at least few unmanned missions must be sent in order to prove high-reliability of all the components and systems before humans are allowed to land in this spacecraft.
After the touchdown, cosmonauts emerge from the capsule’s hatch and go down onto the lunar surface. To do so, they must either use ladder, which would be used in the configuration shown in the drawings, where one of the landing is located bellow the hatch. Alternatively, if legs were rotated by 45 degrees, small platform raised and lowered by 4 winches attached to the hatch could be used as an elevator.
After science is done, landing module is used as a launchpad and spacecraft goes into orbit where it docks with the departure module. After all the modules firmly connect and spacecraft is in the right position, burn is initiated and voyage home begins. Before reentry into Earth’s atmosphere, capsule separates from the rest of the spacecraft, lifting entry is performed, parachutes open and the capsule splashes down into the ocean.
Capsule has curved shape, similar to the one used in the Zond 5 or Chang'e 5-T1 missions. This maximizes interior volume, but also reduces possible angle of attack, so this shape might need further optimization.
Inside the capsule two independent “sleeping pods” are located. They should have life support capabilities that would allow safe return to earth in the case of depressurization of main capsule volume. During engine burns cosmonauts lie flat on the bottom of the capsule and they “sit” in the seats located at the back of the pods. Shirt-sleeve environment is always maintained. In front of the pod there is enough space for cosmonaut to straighten themself up. Inflatable mattress can be rolled out on the bottom of the long section located above seat (parallel rods are embedded in its structure and support the weight of cosmonaut). Whole pod can be raised from horizontal to vertical position. Seat has integrated space toilet (Frito’s chair in Idiocracy was similar, but it is still more elegant system than plastic bag taped to the buttocks used during Apollo program).
Above the pods one-piece space suits are located. Both they and the pods have similar hatches that allow docking between them even in vacuum (compact roll-out barrier with parallel rods must be able to seal the hatch during separation of space suit and pod). Space suits are attached to robotic arms that should be able to reposition the suit in lunar gravity. It would be nice if the space suit would contain robotic manipulators that would be able assist with Valsalva maneuver, tears removal, eating, rehydration and waste management.
Spacecraft will require RCS thrusters on both capsule and service module. Main docking system is actually not required to make a moon landing but it would be nice to equip capsule with it, as spent fifth-stages of the rocket could be used as wet workshops. Androgynous docking system could be used, with six or more (even number must be used) conical frustums around circular hatch. Half of the frustums protrude from the spacecraft (male connectors), half of them are recessed into into the capsule (female connections). Frustums are equipped with with series of electromagnets, arranged similarly to a coilgun. When two spacecrafts get close, magnetic force pulls two vessels together, just like it happens in the Kerbal Space Program (it requires precise electronic control). When frustums are fully inside each other, weak mechanical link is formed by extension of a blade from female connector into a slot in the male connector. Then rigid connection is formed by the automated screws mentioned earlier.
During manned missions empty spacecraft should be present in lunar orbit. It would be used as a lifeboat during emergency. Empty landers that do not have the ability to ascent from lunar surface could be used to provide long-term habitation or deliver cargo required to build lunar outpost. Goal of this this outpost besides usual scientific research and industrialization of the moon could be production of art. Today even backward capitalist regimes such as Russian Federation are able to shoot movies in orbit, as was the case with The Challenge. Performance capture studio could be part of the outpost and video games could finally be made on celestial bodies other than Earth. Studio should have decent size that would allow reduced-gravity parkour to be realistically represented (precise data detailing what capabilities of a human body are in conditions differing from those existing on Earth will be collected during motion capture sessions). Obviously games made on the moon should be free and open-source. Jean-Luc Picard would probably not hold spacefaring civilization that is based on acquisition of wealth in high regard.
In case somebody thinks that making a video game on the moon is absurd I will analyze it from a financial standpoint. I estimate that efficient mass production of 100 rockets per year would require 50 000 people. If they earn 20 000 PLN per month it would mean that 12 billion PLN would be required to pay worker’s wages. Let’s say that additional 8 billion PLN is required to cover additional expenses like buying materials or equipment from outside sources. Cost of a single rocket is then 200 million PLN. If 50 dedicated launches are required to make a game, then cost of outer space component of a game is 10 billion PLN. If 500 million gamers would play this game, 20 PLN ($5) would be a price of the space program per gamer. This is less than a price of a movie ticket.
Files:
moonshot_rocket_pre_v1.zip, mirror, SHA256: d57cc676ef43ff6b8681640d3cac032e0e6b65d417c2ba3ee2855bb2f15c9838
Comments
Post a Comment