A single Falcon 9. Existing rocket, in mass production, that can take 2000 kg to TLI.
2 people at a time, each with a separate 1000 kg descent stage
Each descent stage consists of:
~550 kg N2O4/hydrazine bipropellant
~150 kg of tankage/structure/etc.
~60 kg of astronaut
~115 kg of EMU suit
~50 kg of “enhanced life support” (extra CO2 scrubbers, water supplies, food, etc.)
~50 kg for maneuvering thrusters (though the descent is largely controlled by a gimballed engine, or array of thrusters).
~25 kg for radios, batteries, etc.
Flight is largely manual, though helped out by gyroscopes. Few instruments on board–a radar altimeter, maybe. Things are mostly directed from the ground. The ride isn’t pleasant–the lucky souls get the same water and fruit bars that current astronauts get. The only waste disposal is a diaper. But people have survived far worse conditions for 3 days.
If the goal is to get at least one warm body to the surface of the moon, I posit that something like this has a reasonable chance of succeeding and could be done within months.
NASA actually studied designs not too far from this as emergency escape vehicles. Wikipedia has a nice page on them. Obviously they couldn’t be used strictly as designed, but probably they already worked through some of the early design difficulties.
If the plan is just to put a body on the Lunar surface without any concern for live return or even high probability of success, the plan you lay out is more or less valid, save for the mass estimates for structure, engines, propellant, life support, et cetera, need to be at least doubled. However, what you’d end up with is an astronaut with very limited options for landing sites, with perhaps a few days of life support, and no useful tools or instrumentation. This would still take at least a couple of years to be done with any degree of confidence that it would be successful, as even this minimalist system will still require integration and testing of various component subsystems. Of what use this could possibly be to anyone I can’t even begin to guess, but it doesn’t represent any kind of realistic crewed lunar program.
I generally agree with Stranger On A Train’s assessment of the hypothetical situation… OTOH according to the Space Launch System Wikipedia page the proposed initial uncrewed test is in 2017 (3 years), with a visit to an asteroid in lunar orbit by 2021 (7 years) which makes it sound an awful lot further along than I’d previously imagined. Since turtlescanfly is desperate to plug the meter on the LRV (or something, I forget), it seems we could make a slight adjustment to the destination? My Kerbals’d be willing to do it.
To paraphrase George E.P. Box, all schedules are wrong but some are useful. Just about the first thing you learn in engineering when you get to the level where you start interpreting milestone schedules instead of just failing to adhere to them is that such schedules represent some program manager or vice president’s “vision” of the program which has no direct connection to reality as you and I know it, and yet, once laid on paper are treated as carved in basalt, only to be revised once it is absolutely clear to everyone that not only has the program blown past the milestones without meeting the criteria but in fact the supposed wind tunnel model is actually a cast wax figure spray-painted silver (not lying, I’ve actually seen it); the “fit check” hardware consists of painter’s tape and cardboard (I swear it really happened), and the “protoflight development article” consists of a hollow shell with no avionics, plumbing, or functional propulsion (I could not make this shit up if I tried). And these aren’t from fly-by-night outfits consisting of one eccentric genius and two unpaid interns working out of a garage in the middle of West Texas; we’re talking the major aerospace contractors which are frequently known by their two letter acronyms, billion dollar contracts and army of lobbyists. Norman Augustine, of the Augusting Committies and his eponymous Laws with which most people working in aerospace are aware and often quote, has many things to say about schedules, and almost none of it is good.
Regarding the SLS specifically, this is an example of how using “off the shelf” technology doesn’t necessarily save development time or reduce risk. Sure, the SLS is using the RS-25 Space Shuttle Main Engines (SSMEs) for the core stage and five segment Shuttle Solid Rocket Boosters (SRBs) with the venerable RL10 (used on Centaur and the Saturn S-IV) for the Earth orbital missions or the uprated J-2X for Earth departure missions, but none of these are used in the condition or exposed to the environments they’ve seen in previous service, and in the case of the SRBs, require major redesigns to accommodate the longer length, different mounting configuration, and higher propellant load. (On the other hand, NASA appears to have abandoned plans to recover and refurbish the motors, which is probably the smart move given the negligible flight rate of 1-2 missions a year.) The RS-25 engine controller is exposed to environments considerably higher than seen on the Shuttle, and the J-2X has long had a problematic development history from back when it was planned for use on the now-cancelled Ares rocket system. In fact, pretty much the only part of that system that doesn’t have to be essentially requalified or substantially redesigned is the RL10B-2, which itself has had a problematic operational history and is questioned by some as being suitable for a crewed vehicle due to the inherent complexity and large number of critical failure modes of the design. Currently the SLS development is being so pushed forward that the first crewed flight will occur prior to the flight test of the abort launch system, which is an indication of just how insanely disconnected the schedule actually is.
It is probably more feasible to human-rate the Delta IV or Falcon Heavy and perform Earth orbit rendezvous before the SLS is ready for crewed flight.
Something else to be considered is Apollo was a crash program. Round the clock shifts for manufacturers, individuals working enough to cause countless divorces and nervous breakdowns.
Substantial risks being taken - both knowingly and unknowingly. We’re not talking a difference between a leisurely development program and this mythical emergency.
Not alone, I am an inside/outside mechanic but while I do feel confident in my abilities to work in a nuke plant [as that was once my profession] I have never worked in airframe so I couldn’t whip off a B-52. However, the crew at Boeing could whip one off if Boeing decided for some reason we really absolutely NEED a b-52 for some reason. They happen to have the really big facility with all the goodies needed to pop out a really large airplane if they need one. I could probably turn out a miniature RF B-52 with a bit of work and one of the smaller hobbiest mills and lathes each set up in my barn. Well, I would also like my old blast-n-peen and my old cabinet blaster with a softer media for helping finish off many of the parts. I think I would weld rather than rivet, I hate riveting small projects.
I wouldn’t consider 12 months really rushing something like a ‘prototype’ one off B-52. There are actually aircraft manufacturers that will whip off a one off custom airplane. Hell, there are hobbyists who actually whip off personal aircraft for the hell of it. One guy out there made himself the cutest little single person jet. Otherwise the FAA wouldn’t have a procedure for authorizing one off aircraft. Seems the world is just really wanting someone to whip off a new B-52 …
Pffsh, bullpucky. A shop given materials and blueprints can make anything they can shove metal in a mill or lathe and turn out. What the hell do you think a prototype shop does, drink coffee and watch that motorcycle customizing show all day? They get handed a project, they make the project, they do the paperwork to get paid and they turn the project over to whomever ordered it. If you really wanted the left hand wing mounting widget, give a shop the plans, the finances to get the materials and pay the employees, and you will get a left hand wing mounting widget strapped to a pallet arriving in your hangar in a week or however long it takes to crank one out. It is setting up a factory to crank the widgets out that takes a year to set up.
So please, do not tell me ‘we have forever lost the tech’ to do something. The tech is still around, the blueprints are still around, we could dust them off and send them out to be made and build the old tech if someone wanted to spend the money to get it done. How the hell do you think the old crap in museums gets worked upon? The Smithsonian has little gnomes working in little gnome workshops tap tap tapping out little aircraft parts?:dubious:
Your machine shop qualified to machine beryllium or titanium to aerospace qualification? How about chemical milling? Do they have a 3.6 x 12.2 metre bed mill? How about a power brake capable of handling sheet stock 1.9 centimeters thick and 13.4 metres long? How about taking a five ton 3.4 x 8-metre sheet of aluminium, rolling it into a section of a cylinder, heat treating it, and then milling it down to a complex shape weighing one ton? Forming 0.8mm thick sheets, and welding them into a dome 6.8 metres in diameter, with welds that are capable of holding back liquid hydrogen?
There were many parts in the Saturn V where there was only one facility in the country capable of fabricating the part. Many other parts were so complex in fabrication they they essentially had to build new machinery to make them.
The answer is that your local jobbing machine shop would not be capable of machining almost any part you could name. If they could, the part would be, almost by definition, too heavy, and thus would have been redesigned early in the project. The Saturn was pushing every technological barrier to the limit to build a machine that could just manage to meet the requirements. The sheer size, and the technological complexity were at the edge of what we were capable of.
The canonical reference - worth a read before opining on how easy it should be to just send the plans down to the shop: Stages to Saturn.
You write about “shov[ing] metal in a lathe” and “blueprints” (a colloqual term no engineer working in the aerospace industries uses) as if a rocket launch vehicle is all structure that you bolt, rivet, or weld together and as long as all of the physical dimensions are adequate it is ready to go. There are many manufactured items for which this may be adequate, such as plow blades, truss towers, or Tonka trucks. But for complex devices there is far more than just he structure and physical dimensions, and rocket launch vehicles are both particularly complex and full of critical, unrecoverable failure paths requiring tightly controlled integration and testing, not to mention the precise and reliable functioning of propulsion systems, guidance and control avionics, and as I’ve addressed at length, the software which is required to operate the system. The “old crap in museums” is not maintained at or restored to a flyable conditioned or in any way qualified for operation, nor are museum workers experienced in integrating a vehicle for flight; they just know how to assemble the structural elements to make look right (sort of; every time I see rockets in museums on Redstone or WSMR I see major problems).
If building and flying space launch vehicles were as trivial as you suggest, everybody would be doing it and cost to orbit would be only marginally more than the manufacturing and propellant costs, which are a small fraction of the total launch to orbit costs.
There’s a funny take on this in the newest (May 2014, on the cover,) issue of Analog Magazine: “Our New Overlords,” by Jerry Oltion. In that story, the time limit is only two months, and the stakes are enormous: membership in a Galactic federation, access to alien technology, and effective power of the earth by lobbying the visiting aliens. (They noticed the lunar landings in the 60s and figured that we’d be a real space travelling civilization by now, so we need to get back to the Moon to prove that we’re still up to it.)
Of course, things don’t go well, especially when different factions start sabotaging each other’s spaceships…
If we want to build something that LOOKS like a Saturn V, or a LEM, or the command module, well, Adam and Jamie can whip that shit out in a few months. But they can’t just whip shit up that has the same weight and strength and the dozen or so other parameters that each piece has to meet for a successful launch.
If it doesn’t have to hold liquid hydrogen, or gaseous hydrogen, or 6000 F burning hydrogen, then you can make it out of aluminum foil and paper mache. This is what they do at the Smithsonian.
People somehow think that, since we did it back in the Apollo days, it would be super-easy to do today, and the only reason we don’t have vacation resorts on the Moon is because NASA is a bunch of pussies.
But in fact, the Apollo missions were almost exactly the scenario were are talking about here. Put a man on the moon (and bring him back), money no object, danger no object, Congress and the American people fully on board to pay any price, bear any burden, meet any hardship, yadda, yadda. And the point was to show the Soviets and the world that nuclear combat wouldn’t favor the Russkis , we had tricks up our sleeve too.
It is really really really hard to put tons of metal and human protoplasm into space. And the moon is a lot farther away than people think, LEO satellites and space stations and shuttle launches and space telescopes are really really close to the Earth. And while you’re already halfway there in terms of delta-v, you still have to haul up the goddam propellent to get the other halfway to the moon to LEO.
These are hard physical limits constrained by the distances and velocities needed and the specific impulse of your rocket engine that cannot be overcome by fancier computers and better materials. We still have to put skyscraper’s worth of liquid hydrogen and oxygen in a pile and light that shit on fire and have a couple guys sitting on top praying the whole thing works. There are no alternatives, the chemical reactions that drive this mess have not got any more or less energetic in the last few decades. This is the problem, not the EPA or unions or bureaucratic inertia.
As you say, current rocketry is very close to the physical limits of the chemicals it uses. But it is orders of magnitude away from the economic limits. Space isn’t expensive because of the fuel–that part is incredibly cheap. It’s not even expensive because of the other materials, which are also fairly cheap. It’s expensive because it’s not a commodity item like a car.
Also, I disagree that “danger was no object” for Apollo. The very nature of the space race required very high reliability, especially for the US since we couldn’t effectively cover up our failures. Not at all the same as some “aliens will blow up the Earth if we don’t get to the moon in 3 months” type scenario.
And access to space isn’t a commodity item like a car because rocket propulsion is fundamentally unlike operating a car or any other terrestrial activity. If the alternator on stops working, you don’t suddenly lose control of the vehicle; the power just flickers and you roll to a stop. If you run out offuel, it doesn’t turn around and crash into the road; you just pull off and call AAA or Better World. If you take the wrong exit or let go of the wheel for a minute, it doesn’t spin wildly out of control and explode by the onboard ordnance termination system. If the air conditioning stops working, you don’t freeze or boil or suffocate within minutes; you just roll down the window and let the breeze flow through. There is almost nothing that can fail on a car that will instantly result in catastrophic failure; there is almost nothing on a propulsion or guidance system that can fail without resulting in catastrophic failure.
Going to the Moon isn’t anything like throwing a backpack in the car for an impromptu weekend hiking trip, or driving across North America, or even going on an Antarctic expedition to the South Pole. It is an effort that requires coordinating the efforts of thousands of people performing highly complex intellectual and technical tasks, amny of which have to occur in a specific sequence and produce products with tightly controlled properties, which then have to be checked and checked again to assure that nothing of critical importance, not a single missing fastener or avionics board of which could cause loss of mission is out of place. And while there are certainly ways in which the effort and costs of launches could be reduced, no foreseeable technology will make orbital space launch as routine and safe as a trip to the hardware store.
That would be relevant if stuff on cars actually failed all the time. We could build unreliable cars (and used to), and in that case you’d be right that it’s not so bad compared to rockets because most of the time people don’t die as a result. But the fact is that modern cars are fantastically reliable even though the stakes are much lower and they are built so very cheaply.
Of course, this is because it’s easier to make reliable commodities, because the experience is amortized across many more objects. Hondas are more reliable than any hand-built supercar you can name.
The same could be said of commercial flight. It also requires fantastic coordination between thousands of people. And while airliners do have more redundancy than rockets, I again reiterate my previous point: it is exceptionally rare that anything fails. A stray bolt can kill a turbojet as easily as a rocket engine. That an airliner can typically recover from a single-engine failure and a rocket typically can’t is irrelevant to the fact that despite the enormous complexity, these failures simply don’t happen with any regularity.
No, it’ll never be quite as simple as a trip down the road to the chemist’s, or a flight to Australia, but I think it’s crazy to suggest that no commoditization is possible.
Wikipedia says that solid rocket boosters have a 1% failure rate (Solid rocket booster - Wikipedia). That seems huge. Why is that?
I’ve read that solid propellant tends to be simpler (and one would think, more reliable) than liquid propellant. Does that mean that liquid propellant is even more risky?
It depends very much on the particulars and your definition of risk. Solids can’t be shut down and any failure is likely to be catastrophic. Liquid engines can be shut down and for some rockets, under some stages of flight, and with some failure modes, a failure doesn’t guarantee loss of the vehicle or mission failure. A Falcon 9 was able to complete the primary mission after a single-engine failure; it’s difficult to imagine a similar result if the rocket used strap-on solids instead and experienced a failure in one of them (even assuming the damage stayed localized).
You bandy about the term “commoditization” as if as long as we have enough volume the absolute number of failures is irrelevant, which is exactly the approach taken by the automotive and, to a certain extent, commercial aerospace, industries. But in reality, components fail on both automobiles and aircraft with great regularity (i.e. on a daily basis). That only a modest and fiscally acceptable number of catastrophic failures occur due to x thousands of units is irrelevant; on launch systems, a failure of some critical components is often an unrecoverable failure of the vehicle itself, and even the most wildly optimistic evaluations of launch rates only have a few hundred vehicles per year being launched to orbit, such that a failure of even a handful of flights is a substantial financial impact. An airliner which has three or four engines can successfully with the failure of even one or two, or loss of primary hydraulics, or avionics, et cetera, in no small part because an aircraft–even a modern fly-by-wire aircraft–can be controlled by a human pilot.
A rocket launch vehicle, on the other hand, has no human pilot; the necessary response to inputs is so quick that no human pilot could guide it, so any systematic failure the guidance system or propulsion system failure means unrecoverable failure of the vehicle. Rocket launch vehicles will never be “commoditized” the way automobiles or even airliners are, just owing to the high criticality of failures regardless of the relative frequency of component level failures.
First of all, my own accounting of propulsive failures of SRBs is well less than 0.5%, and most of those are due to damage in handling or system failures (e.g. a failure to consider the effect of system effects rather than the propulsion system by itself). In fact, the only failure of a space launch class SRB I can think of the last thirty years that was due to a design flaw is the Titan 34D-9 flight, in which the propellant grain peeled back from the case due to tolerance stickup and resulting burn through. Even in Challenger with the venting field joint, the SRB itself functioned fine (and in fact survived a dramatically beyond design condition loading of tumbling end over end for 37 seconds after vehicle breakup). The Minuteman I Stage 2 (M56) motor used on the Aries rocket started failing due to (alleged) nozzle liner ejection or separation, but that appeared to be an aging problem for a motor that was operated decades beyond design life. That being said, solids propellant motors, while having fewer components than liquid propulsion systems, are nowhere as “simple” as people perceive them to be, and are highly sensitive to apparently small design features which can readily fail due to bad design practices (as has happened on several programs with which I’ve been involved) but that generally becomes apparent in a rigorous ground static fire program before anything actually flies.
Liquid propellant systems are more complicated and there are far more opportunities for the propulsion system to fail, especially in the case of upper stages that have to restart or dramatically throttle, but most failures are not actually the propulsion system itself but are in avionics, excess environments, or the dreaded software failures (which despite taking up the vast majority of the testing budget, can never be tested enough). In general, due to the unavoidable criticality of failures and the inability too provide the kind of true redundancy capable with any terrestrial system, including aircraft and automobiles, the probability of failure is orders of magnitude higher per unit of operating time than any other engineered system. Flying a rocket is like taking a skyscraper and turning it into an ocean liner; it is almost impossible and highly absurd, not to mention an enormous waste of effort to achieve any kind of rational goal. But everybody likes GPS, Google maps satellite view, and doppler radar weather predictions that are offered by and only by satellite surveillance, and that requires launching rockets.
But all that being said, a 1% realized failure rate would be enviable for almost any mature rocket launch system. There are very few operating systems that can meet or exceed that degree of realized reliability, and even fewer that can meet the predicted reliability at that level. There are just too few many ways to fail and too few to succeed to do much better than that, and the only systems which can approach that value are based upon systems which have been rigorously tested and refined at enormous expense, all at the need to provide "strategic response; i.e. they are ICBMs designed to meet nuclear surety and strategic availability requirements which are dramatically greater than commercial launch reliability expectations.
Of course! I don’t see how any sane industry could possibly behave any differently. Failures do happen and people die. But on an individual level, cars and airplanes are extremely reliable.
Again, I say that’s irrelevant because aircraft engine failures simply don’t happen that frequently. If there were, say, a 1% chance on any given flight that an engine failed, you would have a excellent point and I would thank my stars that all commercial aircraft are designed to land safely with a single engine loss, and be very glad that I’m not on a rocket. But turbojet engines are nowhere close to that unreliable. They don’t fail at a hundredth that rate, and I doubt they fail at even a ten-thousandth that rate.
Or, let me put it this way: suppose you took all the major systems on a plane (engines, landing gear, hydraulics, control surfaces, etc.) and wired them up to explosives such that the whole damn plane blew up if you lost any one of them. Do that, and commercial flight would still be among the safest and most reliable kinds of transport.
All that said, I do actually agree with the point that the limited number of flights is a real problem. Commoditization requires failure, because there’s no replacement for direct experience. And that means lots and lots of flights, so that you can amortize your learning experiences. It may just be that there will never be enough demand for rockets to reach that kind of virtuous cycle. But who knows; if SpaceX succeeds with their reusability, we might get past the necessary cost threshold.
First of all, the rocket launch business is not a “sane” industry. It is a business that goes about trying to perform the almost impossible task of lofting a payload from a stationary ground position to a speed and altitutde where it is literally falling above the horizon without exposing it to destructive levels of loads and environments, all done using systems that are both operating at the highest possible performance combined with the lowest allowable margins and with very limited possible redundancy and a not insignificant amount of guesswork about the actual conditions the vehicle would experience through the wide range of environments it is exposed to in its short lifetime.
Multi-stage rockets are inherently unlike any other highly engineered system insofar as it operates only for at most a few tens of minutes by design under conditions which cannot be fully simulated in ground testing, are designed for the minimum allowable margins (some times as little as 5% below the limit failure condition), and undergo many operations in which redundancy is simply not possible, e.g. it is impossible to have a fully redundant staging system; if you fail to recovever command authority after staging, no amount of redundant flight computers or multiple engines or anything else is going to save you. And that is just one of many examples of events which no design can make redudant. By comparison, the conditions, environments, and possible failure modes seen in other highly engineered systems such as commerical turbofan engines, while certainly sophisticated and built to tight tolerances, are benign compared to rocket propulsion systems, and unlike turbofans, which can be tested on the ground for hundreds of hours of operations under high stress conditions to get an estimate of component failures which high statistical confidence, rocket engines can only be static fired for a few hundreds of seconds (just by virtue of running out of propellants) and under conditions that are only very roughly similar to flight-like conditions. Gaining the same amount of running time with a rocket system as one can get with an air breathing commercial turbofan or turbojet would take many orders of magnitude more cost and effort. Planes and automobiles may fail statistically less often than rocket systems, but they can also tolerate faliures without catastrophic loss for the vast majority of failures, whereas most significant failures of the propulsion system will result in vehicle destruct, and even the systems which can be built with redundancy (flight computers and other avionics) may experience system and limit failures which negate the supposed reliability benefits of redundant paths.
I’d also like to readdress your assertion that “commoditization” (as you refer to it) and reusability will inevitably result in greater reliability and cost reductions. This has long been an assumption that has fed into analyses and trade studies which have been used to justify efforts to develop single-stage-to-orbit and partially or fully reusable vehicles. But in fact the empirical evidence all points in the other direction; that a desire for reusability results in reduced reliability and higher operating costs due to the requirements to assure extended service life and the risks of uninspectible damage or degredation (e.g. both flight and age related degradation of Shuttle thermal tiles and forward RCC panels). Reusable single stage to orbit (RSSTO) and two stage to orbit (RTSTO) vehicles, by virtue of having to carry so much mass to orbit or extra propellant to allow for return to a recovery site (as SpaceX and Blue Origin plans to do with the Falcon and New Shepherd, respectively) result in a substantial reduction in payload capacity to the point with most proposed SSTO designs that they would not be able to carry a useful payload to orbit without going to almost absurdly low mass fractions necessitating the use of advanced lightweight composites operating with minimum possible margins, which again reduces reliability, especially in the case of reuse where material properties will degrade due to exposure to the high radiation space environment and high temperature reentry conditions. Reusing components like engines and avionics requires both testing and likely refurbishment, which are the major costs in building up a vehicle anyway, so while you may not have to manufacture the stage structure or combustion chamber, you still incur the majority of the costs that go into making individual systems into a flyable rocket. And regardless of what you’ve heard about how such-and-such vehicle is propulsively redundant insofar as being able to tolerate loss of one or more engines, this is only true in a categorical sense, i.e. it can only accomplish flight and insertion requirements after losing an engine after a certain point in flight or by carrying excess margin, or otherwise inserting into a lower energy or less precise orbit. The performance margins on rocket launch vehicles are just too small to achieve even a modest degree of redundancy much less tolerate failure without signficant degradation.
The actual way to reduce launch costs and improve practical reliablity is to accept lower performance, reduce design redundancy to only high value conditions, test and refine systems to where the realized reliability is high enough that elaborate design mitigations and acceptance testing & inspection (ATP) are unnecessary, and make reusability a secondary and long term goal versus driving out costs in manufacturing, processing, and integrating expendable stages. I suppose this is what you would call “commoditization”, i.e. driving the design to be simple and cheap enough to produce in a serial, production line fashion. So far, nobody–not even the darling favorite of the space enthusiast community–has demonstrated a path toward a low precision design/minimal ATP/minimum processing labor to the point that launch costs will drop and launch rates will increase by orders of magnitude to compare to anything like commercial air travel or operating an automobile.
