Control cost or go home

Terraformer · 2012-05-17 09:37:09

Erm, spinning it at 12RPM cuts your potential crew down to maybe nothing, because at those rates very few people would be able to operate a spacecraft.

Rune · 2012-05-17 12:01:45

That is simple, Impaler? Wow. Different definitions of simple. I mean, I would say taking a mars/moon hab, a couple tethers, and using the propulsion stage as counterweight is just a bit simpler. As in, the only new design is the tether system. Hell, you can even land the hab later, and that way you don't have to take out of the design the landing systems. And you need to design a hab if you are going to Mars, right? Always good to have one radiation shelter with you, even when you are on the surface. I mention tether because some versions of the ship may turn out quite compact, and there is a limit to the rpm a human can take and be comfortable. That is, not nauseated all the time. Conservatively, I think, that is around 3rpm.

Oh, and you don't need to identical counter-rotating sections. Have you heard of IMU's (inertial maneuvering units)? Big-ass gyroscopes for attitude control. Really any kind of flywheel would do, and the smaller the lower the friction losses. But having angular momentum stored in your ship itself by spinning it whole seems even easier, since it adds only the fuel required for a few spin-despin maneuvers. And have you considered what a spinning atmospheric seal rated for vaccum and 2+years of operation under stress with minimal losses would look like? I wouldn't like to have to design that.

And before someone points out a spinning ship is difficult to control, you can provide course correction on such a tethered ship... with ion thrusters, even. The reference I take for that is on Atomic Rockets' section in TransHab, where he links to a very enlightening study about a Transhab redesigned for 1G at the end of an extensible, flexible boom. The whole thing is like less than 40mT, supplies for a reference Mars mission using ion propulsion included. Go look it up, hurry! (For the lazy ones that don't care to use google.)

GW, to reuse a significant fraction of the ship, and not go throwing advanced in-space engines and tanking around, you really need to have a nuclear in-space stage for a Mars-and-back ship, unless you have refueling stations there, and that supposes either a lot of expendable missions to fill the station, or working ISRU by the time you send the first mission (which I always thought was stupid). So not only the lander. In fact I see as easier to make ISRU work for a lander and get it in a single stage using methane/LOX rockets. With proper redundancy (a fueled lander waiting by the time the astronauts land in the now empty one), and validating the method with a robotic sample return mission (perhaps in the orbital test flight of the reusable mars transfer nuclear stage?), I think it could be good enough. Lowers the required T/W on the NTR so you can build it shielded enough to convince a greenie it is safe to launch. Also, smaller, so less nuclear fuel and megawatts in there.

That being said, with enough Falcon Heavies you could fill a Mars depot from earth with ~10mT loads at a reasonable price (14mT on mars transfer, some expendable tanker using soyuz tech... you get the idea). Pick storable fuels, maybe augment them with supplemental ion/arcjet systems for the coast period (again, you can use a continuous thrust system on a spinning ship), and you can actually do an all-chemical mission with currently validated technology. Look at that! You also get very busy launch windows, I might add, which is a kind of complexity in its own way. To return a ~50mT Hab from High mars orbit would take... say between 5 and 10 fueling flights? Just order-of-magnitude approach to pick architecture, I don't want to get into details in case I lose sight of the whole picture. I throw in arcjets because I've been looking at them recently and it seems all very nice: all electric, using any gas, and NTR-like isp (so NTR-like thrust for the same power). For course corrections and a small isp boost it looks nice, but it is just a small personal twist, not the central point I want to make.

Rune. Which is, I think, KISS. So maybe no arcjets, and all chemical. Wouldn't that be old school?

Rune · 2012-05-17 12:34:15

Oh, and in case it isn't clear (I don't know why it would be since I barely mention it at all!). I also think the lander is the pacing item for any Mars mission, for sure. Radiation we know what to do, propulsion we know what to do, artificial gravity we know what to do (it's untested, but as long as Newton works, I will trust a spinning ship to do the job). It's the ship capable of landing on Mars and taking off again that we have never built.

Can it be done? I think so, even with chemical engines as I said before. I mean, at most (during launch from Mars) we are asking about 6km/s out of it. That is about Deimos orbit from the planet surface, a sufficiently high orbit for pretty much any MTV to park in. Using your "standard" Methane/LOX engine that every mars mission assumes, even when there is none built at this point (though apparently LH2/LOX turbo-machinery works with little modifications, and there are plenty of tests), at say 350s isp (easy for methane with just a big nozzle) that is a mass ratio of 5.75. A tad more than 17% of launch weight for payload, structure, engines and everything else. When you compare it to the reusable SSTO's we are contemplating on building here, and take into account LCH4/LOX is a fairly dense, only mildly cryogenic combination, it doesn't sound that ridiculous. Yeah, it would be nice if the thing aerobraked on it's own before a short powered landing to reuse it for more than one mission and transport it to mars (mostly) empty, but it is not really necessary if you can't crack the heatshield mass issue, you could ditch it before landing and use the landers only for a single ascent-descent mission. Only the working ISRU unit, and as I said before, you can validate that with a sample return mission.

As for the architecture to use it, something like DRM (mars semi-direct, by it's common name). Redundancy by having a prepositioned fueled lander waiting the one bringing the astronauts in, which stays to be used on mission two. Simple and robust, and if you can reuse them you can get a lot of missions out of every couple of landers, since chemical engines are easier to design for a lot of restarts. We NTR fans should remember a NERVA, for example, was designed for mere 10 restarts after which it would have burned out enough of its core to be disposed of.

And Terraformer, you beat me to it by seconds.

Rune. There, I finally talked about Mars landers. About time.

Last edited by Rune (2012-05-17 12:36:14)

Impaler · 2012-05-17 12:35:00

12 rpm is what you need for Earth Gravity but nearly all centrifuge research agrees/anticipates that significantly less then normal Earth gravity would be adequate to preserve health which is why I gave figures for Mars g. From what I've found 6 rpm seems to be sighted commonly as a speculative maximum comfortable value, in a large inflatable that would give ~25% normal gravity. But keep in mind that everything we know nearly nothing about intermediate gravity because no Earth based centrifuge can simulate it being always in a 1g field they can only explore the >1g range. Their dose indeed seem to be a consensus that short ~2 meter radius centrifuges operating in the 1g range produce too much head-foot gradient and too much Coriolis force to be comfortable and would only be useful for exercise and not habitation, but 6 meter radius is in my opinion worthy of investigation. Also keep in mind that the difficulty of adapting to the centrifuge only needs to be equal or less then that a 0g environment which typically takes several days to get over.

Terraformer · 2012-05-17 13:02:18

Well, I've read a figure of 10RPM before, but I seem to recall 4RPM as a good figure to work with. According to this calculator, that's enough with a 15m radius to produce ~1/4g.

Though, the possibility of having an inbuilt centrifuge for exercise can not be ignored. I'm thinking of bicycles set into a ring, so that when the astronaut cycles, the will generate the acceleration. Send a mass rotating in the opposite direction to deal with the angular momentum, even have two cyclists oppose each other. Or they could run, either way it's a possibility.

Impaler · 2012-05-17 13:43:33

Rune: A centrifuge operating inside of a pressure vessel is certainly simpler then the Tether approach.

We know how to make a spinning wheel with literally zero engineering challenge
Centrifuges can be accelerated/decelerate with only electrical power any time we choose and manual breaks-pads that guarantee our ability to decelerate
The exterior of the vessel is spin free meaning we can attach delicate structures like Solar Panels or make EVA's
Likewise the external star-field is stationary for navigation

A tether on the other hand requires many things in addition to a strong steel cable which I will grant is not a challenge, you need to expending propellent to spin up and you can't do this until your booster engine/tank is expended meaning you just committed to a 6-month flight without having validating your artificial gravity. The now Empty engine should be much less massive then your habitat so your tether length is multiplied by a large amount to get the needed g in the hab but now your booster is under several negative g's and you probably need to engineer it differently because rockets are generally designed to take compressive loads. Unless your deploying solar panels from the center of rotation (some scenarios do but its more complexity) they will be under g and your standard panel is way too delicate for that. Also your going to need a communication umbilical with the RCS in the booster so that you can synchronize the acceleration and deceleration thrust. You would need to have a system to sever the cable at the end of the transit stage or if deceleration fails. Basically tethers don't have much margin for error ware centrifuges have copious margins and would be easier to develop and validate on space stations ware we can see how human health is effected by intermediate gravity before sending them on long duration missions.

Also I think your misunderstanding the design I'm proposing from the comment about rotating airlocks. The entire exterior of the inflatable including its end-cap airlocks and docking ports are non-rotational. Rather then the central core being one rigid continuous tube it has 2 segments and 3 rotational cuffs, one cuff on each end ware it mates to the end-caps and one in the center connecting the segments. All are inside the pressure vessel and accessible for maintenance. The motor/brake would be in the central cuff to spin the segments in opposite directions without imparting spin to the whole vessel. Yes it is a bit like the idea of a fly-wheel/gyroscopes in a commercial satellite but by splitting the habitat into 2 counter-rotating sections of equal mass their is no wasted mass in something that not habitable. Friction is bad in an IMU cause it torques the satellite but in a counter rotational system so long as the friction is balanced between the segments their is no torque and your motor just needs to compensate for friction loss to maintain the desired spin.

Last edited by Impaler (2012-05-17 13:48:27)

Rune · 2012-05-17 14:05:18

Ah, ok, you are advocating something with a catastrophic failure built-in. No worries . What is the relative speed of the rotating hard surface respective to the inflatable shell? And here I am thinking of a kevlar line with cables embedded on it (or, you know, go wild and use wi-fi) released in case of emergency by explosive bolts. 'Cause rocket stages can explode, for example, and you asked for it. With independent control authority in the spent departure stage, the unspent return stage, and the Hab, all of them able to talk to each other. If it is a single stage it is partially full, BTW, and with all the redundancy expected of such hardware. I must be crazy, right?

Have you looked at the .pdf on the atomic rocket site I linked to? It's a fancier alternative with an electric behemoth of a ship that doesn't stop spinning unless it's either at earth or mars orbit and some other vehicle has to dock. The total propellant required to spin or despin is a whooping 200-400kg depending on the attitude control system used. That is, 1-2mT for a huge electric ship with a multi-megawatt nuclear reactor in a full mars mission. Everything designed for 1G except the thrusters, which are kind of in the middle, and it turns out that simplifies building things like radiators or booms. Not least because they work at Earth surface, too. You might like it, seriously.

Rune. If you don't like to work out the dynamics of a tether system 'cause, I dunno, you are feeling lazy, you could always use a rigid extensible boom. As GW said, it's not that long, not that much mass.

Last edited by Rune (2012-05-17 14:09:17)

Impaler · 2012-05-17 15:20:47

At 6 rpm the tangent velocity is only 4 meters per second, and note that Bigelow/Transhab inflatables are made of kevlar fabric and when inflated are literally rock-hard AND their is a water bladder inside of that ranging from 5-20 cm thick depending on how much Radiation shielding you want and how many water delivery trips your willing to make. Even if the entire centrifuge suddenly detached from the hub it would first 'splash down' into the bladder and if it managed to rub against the inner wall at all it would be over a very broad area, their would never be enough angular momentum in the system to cause a breach as each rotating element is in the range of 25-30 mt (The Hab gets loaded up with ~50 mt of equipment plus a bit for the hub, spokes, deck plates and then split evenly in half). With spin that gives 25% of Earth Gravity and assuming ALL the mass is at the edge the tensile load on the spokes is only equal to ~6000 Newtons so 60 spokes each carrying a load of 100 N would hold the whole thing together with trivial ease, it could be done with strong wire but I'd go with telescoping carbon fiber poles mainly so you could use it as a frame for wall partitions and equipment. Similarly the floor panels are just curved carbon-fiber panels about a meter across attached with hinges.

GW Johnson · 2012-05-17 19:15:15

Just for the record, the spinning ship I was proposing was not cable-connected. It was a stiff structure of docked modules, of moderately high overall L/D. Typically, these ship designs were between 100 and 300 meters long. They would resemble the "2001" movie's "Discovery" that was spinning end-over-end when it was salvaged in the movie "2010".

Parallel-stacks of modules could be docked and connected together axially and laterally, if a single-module stack was too long to be dynamically stable and structurally sound. No cables, no fancy space trusses (which do not contain propellant or life support!!!), no reconfigurations to do anything. Just spin it up and coast a while (could be thrusters or a flywheel, not sure which might be lighter, but either could be made to work right now). De-spin to make maneuvers, so the dynamics and control are clean and uncomplicated. Simple. Clean. Very do-able. Right now.

The spin rate criterion is a fuzzy boundary. People with extensive military flight training in high-performance aircraft can tolerate spin rates above 10 rpm for quite a while, usually. Not always. Getting sick from spin is both training-dependent and exposure time-dependent. The astronaut who can take 12 rpm for an hour in a centrifuge would quite likely get sick, even incapacitated, if spun that fast for days, much less months. Less training, lower short-term tolerance, too, much less the long-term tolerance. For "civilians", 4 rpm seems to be tolerable for very long intervals, typically. That's why I picked it. Maybe 2 or 3 is better. That's an experiment needed to be done right here on Earth right now. Any of those numbers (2 to 4 rpm) are tolerable in some very practical designs.

I had sizes and numbers worked out for 30 metric ton modules (before Spacex revised it to 53 tons), with the human vessel powered by a gas core NTR, as the baseline in my original paper at the Mars Society convention last year in Dallas. The lander-plus-landing propellant vessels had solid core NTR's which pushed those vessels to Mars, then got used again as single-stage reusable landing boats fueled from orbit. These landing boats were 60 ton max weight, with a 6 ton payload, at 20% structure, and 70% LH2 propellant. These were good for a two-way trip, fully loaded both ways, rocket-braking-only descent (no benefit from aerobraking), and 30-degree out-of-plane both ways.

I did a revision of this mission design for using solid NTR instead of gas core on the manned ship, which made it quite a bit larger. I had to "stage" by leaving several of the empty fuel tanks in LMO, but still was able to slowboat it two-ways Hohmann transfer without ditching anything at all into deep space. (The gas core version was a fast-trip at 75 days one way.) I did this revision because solid core NTR was ready-to-fly in 1973, while gas core needed some development years. The right team might get one working in 10-20 years; the wrong team? Never.

I haven't checked it for all-chemical, but you'd have to stage a lot more by ditching most of the empty tanks into deep space. But you could still reuse the habitat, engines, and those tanks not staged off by the time of return-arrival in LEO. Myself, I think the NTR is just a better deal, especially if maximum re-usability is a design requirement (and it should be from the get-go!!).

Gas core would take longer to develop, and looks more like a later upgrade to me now. At high payload fraction, and simultaneously high structural fraction (for tough-as-an-old-boot reusability), I rather think the solid core NTR makes more sense, especially for a single stage lander item. It is really hard to beat 900-1000 sec Isp to get you practical mass ratios. I'd love to have a nuclear light bulb gas core at about 1300 sec for the lander, but that would be some years' worth of development away. NERVA could be resurrected by the right team in under 5 years. The gas core in my original paper was a design projection for an open-cycle machine of 6000 sec, but penalized by a waste heat radiator requirement.

I've got some illustrations and numerical data for the all-solid core NTR machines worked out and posted over on "exrocketman". See "Mars Mission Second Thoughts Illustrated" dated 9-6-11. You'll have to scroll down or navigate down by date and title. The site is http://exrocketman.blogspot.com

Not only are the performance and size numbers posted, but the rationales for why I chose what I chose are there. There's a version of the original paper posted there as well, dated 7-25-11. The discussions and rationales for why maximum self-rescue capability is required, and why we ought to make multiple landings in the one trip, are there, and I think those are very good rationales. Those are all things we did not do during Apollo, but we can now, and we should.

Too many are still focused on another Apollo-style flag-and-footprints stunt. That's no reason to go to Mars. It's just too far, too dangerous, and nobody is racing anybody else this time. Exploration is the real reason to go to Mars. And that word requires a proper definition, one based on about 500 years of history.

GW

Last edited by GW Johnson (2012-05-17 20:16:20)

Rune · 2012-05-18 05:52:48

I knew your ship was rigid, GW, but you also tend to over-kill the requirements a bit elsewhere, in my humble opinion. I mean, when I see a 6000s gas core talked about, it is to go to the outer planets or something! You could go LEO-LMO-LEO propulsively with a solid core in a single stage (and no heatshield, just maybe an aerobraking period at the end of each leg to lower and circularize the parking orbits). I know 'cause I run the numbers elsewhere, and using methane as fuel. If you tweaked the architecture and left from L1/2, and waited on Mars in a very high orbit (like near Deimos), even better if it is highly elliptical, so your lander doesn't need lots of delta-v to land, only to take off, then a couple of chemical stages or maybe even one and a bit of aerobraking will do. Higher fuel mass, of course, and I haven't worked out proper numbers, but we have some first stages with mass ratios of about 20. When I try to work out masses and volumes, and since I assume a reusable nuclear stage would actually be, you know, reused (and probably tested while you are at it) to position cargo and Hab elements on the ground, maybe even backup landers, in previous unmanned missions, my MTV turns out "a bit" smaller, lighter, and much more compact. I mean, a ~50mT Hab (a transhab is 3 stories high) with a ~25mT mostly unfueled lander pushed by either a nuclear methane stage or a storable chemical return stage does not look like it would be +100m long, even at low diameters. That stack would need something else to spin to meaningful G, and would be, overall, less than 1000mT total weight depending on propulsion (say two MR3 chemical stages at worst?), probably much less if that propulsion is nuclear, getting IMLEO into the low hundreds of tons. And by sending previous unmanned cargo missions you can get as much surface equipment, and as many "landing sites" as you need.

Tough if we go with Impaler's 0.25G as an acceptable thing (I wouldn't), even that small length would be enough. And easier to boost to mars and refuel when it gets back that something that multiplies several times your IMLEO just in unnecessary shielding. You do know a few cms of water are good against solar flares only, right Impaler? Cosmic rays go through anything not measured in tens of tons per square meter. That is why everybody talks about smaller radiation shelters.

So really, you do not need lots of complex hardware and billions of dollars to go to Mars, like the thread is all about. BTW, RobS, I agree with most of what you said at the beginning of it, even if I have some different ideas about how to go about it . If you are smart, and don't get seduced by "what would be very cool to have", then you actually need to develop only a few relatively simple items (say ~4: manned lander, cargo delivery lander, hab and in-space stage), most of them very technologically mature already. I am starting to believe you don't even need a costly nuclear rocket development program... and it is me saying that, you know, the nukie.

Rune. Though it would be the first "cool thing to have" I would add, since it totally makes it doable with single stage reusable vehicles.

Note: Throughout the post, I am referring to the literal definition of aerobraking, as opposed to aerocapture or aerodynamic reentry deceleration. That is mostly free in mass since it requires no heatshield.

Last edited by Rune (2012-05-18 07:29:03)

GW Johnson · 2012-05-18 08:14:14

Hi Rune! You're right. I do tend to overkill the requirements a bit. That's because I'm a suspenders-and-belt-and-armored-codpiece sort of guy. There's absolutely nothing more expensive than a dead crew.

I used three overriding requirements on my design (1) a "way out" at every phase, (2) more than a dozen separate landings all over the planet in the one trip, and (3) throw nothing away into deep space. That first item drove me to carry enough propellant on the manned ship to return home, in case rendezvous failed with the other assets in LMO. Plus, I defined what "exploration really had to mean, so my mission design approach ended up looking nothing at all like the one launch/one mission, and one mission/one landing approaches that went into Apollo.

I'm not very knowledgeable about the energy saving orbits y'all have been discussing. Mine were just circular LEO, circular LMO, classic Hohmann transfer for slowgoats, and near straight-line "shots" for the fast trip GCR thing. I did back off of near-term availability of fast-trip GCR from the original paper, and went all solid-core NERVA slowboat in the revised design. It was the manned ship's requirement for return propellant that mainly drove its size. My habitat was larger than NASA's Transhab, but a bit smaller than Skylab.

Still, all 4 vehicles were well under 1000 tons each, as assembled. Around 3000 tons of modules for the whole lot. Launched at shuttle prices, that's not affordable. Launched at Falcon-Heavy prices, that's amazingly cheap, compared to most other proposals I've heard, especially on a mission price/number of landings ratio (I had 16 separate landings). There's plenty of room for launch cost to program cost ratios in the 10-30% range.

If somebody more familiar with energy-saving orbits was to take my basic suspenders-and-belt-and-armored-codpiece design, and skinny it down with clever orbital mechanics, my guess is things would shrink by around a factor of 2. I just don't know how to do that.

The landers could use some real attention, too. I assumed no benefit from aerodynamic drag on descent, just rocket braking all the way down, and rocket thrusting all the way up. I did include 30-deg plane change at LMO velocities. If drag could save fuel on descent, you could buy even more plane change and cover even more of Mars on the one mission.

The numbers I used are over there on "exrocketman", including the original paper 7-25-11, and the final revised version 9-6-11. Be my guest.

BTW, at that convention, I saw another paper that really intrigued me vis-a-vis refueling the reusable transit ship in LEO. There's a group with a light gas gun already launching small test articles for USAF at around half orbital speeds. They presented design projections for a scaled-up gun that could shoot loaded propellant tanks to LEO. Hard tank plus a small solid motor to circularize. Their estimate of cost per pound delivered was amazingly low: under $100/pound. Until we have propellant-making infrastructure in space, that's an amazingly practical way to refuel reusable things in LEO for subsequent missions (including emplacing said infrastructure).

GW

Impaler · 2012-05-18 22:37:55

Keep in mind that the Mission even if you use off the shelf commercial launchers still requires the vehicle development cost

Initial Vehicle Mass LEO * (Construction cost per kg + Launch cost per kg) = Marginal cost
(Initial Vehicle Mass LEO * Development cost per kg) + (Marginal cost * Mission count) = Total cost

Many of the advanced probe/rover/telescopes are more expensive to Construct per kg then they are to launch. Development costs scale with the size of the vehicle so developing a 1000 ton vehicle is going to be HUGE even if it was free to put it in LEO. To really drive down mission costs we need to apply the same mass-production principles to the vehicles as well as the launchers, low launch costs should be used first and foremost to reduce the development and construction per kg of the final vehicle by reducing the extremely high performance per kg demanded from mass budget. It's would be trivially easy to design a vehicle that could accomplish the goals of most of our current scientific missions if we doubled or quadruple the mass budget of the vehicle. Of course at current launch costs it's appropriate to put intensive engineering effort into getting high performance per unit mass, if launch cost comes down an order of magnitude we should see an order of magnitude reduction in the appropriate engineering intensity per kg in the vehicle and lowering substantially the target performance per kg.

GW Johnson · 2012-05-19 07:24:26

Hi Impaler!

Neat to see somebody with a real costing model. I never had one. All I did was sum up direct launch costs and use a jigger factor in the 10-30% range for launch/program costs.

I will say this: I cheated, my mission design doesn't quite fit the implied assumptions in your costing model. You don't have to develop 1000's of tons of new vehicle, just several dozens worth. Just a short series of fairly small modules, and one lander design.

I used a very modular design: a habitat module (3 closely-related modules, actually: same shell, just a different trick-out inside), a common propellant module, and an engine cluster "module". The crew return was to be two slightly modified Dragons tricked out with extra propellant in the trunk. The fourth development item was the landing boat vehicle, which is the toughest of the bunch. It plus a bunch of propellant modules was the unmanned-asset ship.

All these modules as I originally worked it out were 30 tons, and pretty close to the largest payload shroud dimensions for Falcon-Heavy, although I assumed they would ride up "naked". The manned and unmanned ships were mostly just a stack-up of a whole bunch of these identical propellant modules all docked together, then plumbed up and wired. Given today's projections for Falcon-Heavy, I'd redesign around 50-53 ton modules, instead of the 30-ton projection I was using then.

That's quite a bit different approach as compared to everything we have done before, but every single piece of it is based on things we have done before. By the time we would actually start doing it, Falcon-Heavy and manned Dragon will also fall in the "we've done this before" category.

GW

Impaler · 2012-05-19 18:51:12

Well I wouldn't say my cost model is really all that much, it's just an attempt to acknowledge that we need to pay for development costs in addition to launch. I've read that most vehicles (military jets, boats, automobiles etc) have industry average development costs per kg but I don't have any of the numbers, it would just seem prudent to use such numbers when coming up with reference mission architectures. Naturally if you develop a modest sized vehicle but then 'marshal' them in orbit with a TMI stack your breaking the 1:1 relationship between Initial mass in LEO and developed mass (the norm for every mission to date) and potentially saving a lot of development cost AND manufacturing costs (assuming you make a bulk purchase from the manufacturer).

The reasonable launcher masses to build around would seem to be the 10 (Soyuz/F9), 20 (Titan-IV/Proton/Aryan-V), 50 (F9H), anything above that's speculative and as you've said are unnecessary. I think we should also consider mixing vehicles particularly the 20 and 50 because it allows many more of our existing launchers to partake in the marshaling process and on a world-wide basis. If only F9H is used you mandate a very intense launch passe for a single company Space-X.

GW Johnson · 2012-05-20 12:55:00

"If only F9H is used you mandate a very intense launch passe for a single company Space-X."

True enough.

I just picked a model and ran with it, just to see where it led. It led to Mars, Venus, Mercury, and the NEO's, long before 2030.

But it also requires a complete paradigm shift on the part of whatever government agency or agencies might be involved (and whatever contractors they might use). "Business-as-usual" will accomplish the same nothing we have seen for 40+ years, in terms of manned spaceflight beyond Earth orbit.

My unique specialty when I was employed in aerospace/defense was figuring out what is actually possible to do, and precisely how to do it. In a lot of cases, I actually did it. Not alone, of course.

GW

SpaceNut · 2012-05-20 18:00:26

Even when building in 20T chunks or more in LEO for a mission that might amass only 400T it is still alot when spread over a 2 year cycle leaving 7 weeks for a launch window to make it happen. I am not sure of the build from start to finish of the Space x product but for the Delta & Atlas which are on the order of 18 months that is still going to need alot of ramping up on the production level in order to just get the launch vehicles that we need over a short period of time. Then lets not for get the important part which is each payload being designed from scratch which when we look at satelites are on the 6 year cycle.....Just look at the ISS for an example of how hard it is to make such a feat happen....

SpaceNut · 2012-05-20 18:46:32

Something else to keep in mind What is an Astronaut's Life Worth?: An Interview with Robert Zubrin....

There are links for video but you can go to the one provided to listen but here is a quote:

"You're saying that you're going to give up four billion dollars to avoid a one in seven chance of killing an astronaut, you're basically saying an astronaut's life is worth twenty-eight billion dollars," says astronautical engineer and author Dr. Robert Zubrin.

So even if the cost were contained to that figure we still are saving for a very far off time with the ability of groups to aquire funds so unless there is government funding included Mars is a very long ways off.....

Impaler · 2012-05-21 09:26:53

SpaceNut wrote:

Even when building in 20T chunks or more in LEO for a mission that might amass only 400T it is still alot when spread over a 2 year cycle leaving 7 weeks for a launch window to make it happen. I am not sure of the build from start to finish of the Space x product but for the Delta & Atlas which are on the order of 18 months that is still going to need alot of ramping up on the production level in order to just get the launch vehicles that we need over a short period of time. Then lets not for get the important part which is each payload being designed from scratch which when we look at satelites are on the 6 year cycle.....Just look at the ISS for an example of how hard it is to make such a feat happen....

Designing a multitude of unique 20mt pieces to assemble in LEO would be terrible from a design standpoint, that's not at all what I'm proposing when I say we should think with the 20mt launchers in mind. Your going to want simple, interchangeable blocks like fuel tanks/boosters to be the 20mt pieces. Group 5-6 of these together and you have your TMI booster stage to send the large 50mt pieces on their way rather then using boosters that are the same mass as the cargo. A 20mt booster is nothing special and will require little development cost, then they can be produced in large quantities and a large number of launchers can handle them. The launch schedule becomes much more flexible because their are a half dozen nations and launch facilities that have that capability and they individually only have to achieve a modest launch pace. I'd actually start by launching the larger 50mt Mars bound payload first because its by nature capable of LONG endurance in space ware as boosters have boil-off issues. Then all that's needed is a total of 5-6 successful booster launch (schedule and prep 7-8 for scrubs/failure contingency) and docking over 3-6 weeks and the stack is complete and ready for TMI. Scrubbed boosters just get re-inserted in the launch manifest of your next vehicle.

Also this is the only politically viable means of getting an international mission done, National aerospace companies/facilities/manufactures have got to get a slice of the launching action just like they will want some of the hardware development and crew action. Many nations are looking to acquire larger launchers in the 20mt range, it's reasonable to anticipate China, India, Japan and Brazil having robust reliable launchers of that caliber in the 2030 time-frame to complement the existing Atlas V, Proton and Aryan 5. Commercial satellite launch is and will remain the fundamental downward price driver on Launch vehicles and 20mt is the maximum viable size for current satellite launchers to be competitive (launching to GTO with a pair of Satellites). If something new comes along to change the market and make 50+ the ideal place for the market then that would be great but we should assume that's not going to happen and 50mt will be government only market. Manned-Moon ambitions on the part of Russia and China will certainly require them to develop launchers of this class but I doubt they will build anything larger as the mechanics of vehicle assembly and marshaling are deeply ingrained in the Russian engineering culture and China is basically cloning Russian designs.

Rune · 2012-05-21 10:01:07

Quite true, Impaler. While having a ~50mT launcher is nice to loft the big things in one piece (let's say Hab, propulsion module, and lander(s)), it makes a lot of sense to use the existing 10-20mT launchers that most nations do already have to launch the heavy dividable stuff like fuel. Just one... call it small discrepancy. I wouldn't launch propulsion stages, I would launch tankers with a standard docking interface. First, most of the countries that could pitch in already have ships that could be upgraded into these things (part of the technological/industrial pie to split). Second, a big tank can launch empty attached to a small heavy engine with no problems, especially if the tank is made to hold dense, storable propellants like methane. It also makes sense to launch just one tank, to minimize plumbing and interfaces in the final thing and make it mass efficient (I think ISS taught us that with its $100B price tag). Third, every country could design it's own, or use several sizes, and use them for other things, and that makes the industry that much bigger and sustainable. A nice bonus.

Rune. Good candidate for ISS-style barter agreements, too: You build the lander and provide X mT of fuel, you the Hab and Y mT, you launch the big stuff, you the crew... you can build on everyone's specialties.

GW Johnson · 2012-05-21 11:57:22

ISS was built in 25 ton chinks with a launcher that cost 27,000/pound, when loaded to 25 tons (more cost if lightly loaded, all the same launch cost). With today's launchers, we are closer to $2500/pound at 25 tons. Launched and built today, all other things being equal, ISS would cost closer to $10B than $100B.

Standardized tankers is not very far at all from my standard stack-up of modules. It's all docked, wired, and plumbed. What's the difference? We have a lot more latitude in spacecraft design now than we ever did before ISS.

GW

Impaler · 2012-05-21 12:12:16

Agreed tanks without engines would be ideal, I was being somewhat conservative by omitting the obvious next step which is an International Fuel-Depot either in LEO or EML1. That would provide the ability for virtually any size of launcher and thus any Nation/Company to participate in the fuel marshaling process. Then you launch your partly-full/empty booster to the Depot, fill up and mate the payload too it for a single TMI burn. That essentially eliminates all your multiple launch

This architecture has a lot of advantages, first it's usable for virtually ANY mission beyond earth orbit and their are definitely going to be scientific missions from many nations over the coming decades that would justify such an effort without needing to appeal to near generation long time horizons for the 'payback'. Second it's building off of technologies that are already being developed like (Non Russian) Autonomous Rendezvous and Docking, Cryogenic fuel storage and Cryogenic Fuel transfer. Third it's attractive for nations that have small-medium sized launchers to buy from the Depot to allow them to conduct missions (using only domestic vehicles and launch sites) that are on par with those of the big players, because fuel purchase is less likely to be seen as 'dependence' on foreign technology/governments/corporations (unlike for example buying a ride for an Astronaut to the ISS in a Soyuz) their will be much less of a political barrier to this kind of 'asynchronous' cooperation. Forth if the Depot is at EML1 then it makes the continued development of large Solar Electric Propulsion based 'tugs' very attractive as this would be an extremely efficient means of delivering cargo from LEO to EML1.

An international treaty/consortium to manage fuel sale and purchase from the Depot would be necessary of-course but I think many national space programs would be attracted to that because it would allow them to avoid delays and cost over-runs caused when existing bi-lateral partners get budget cuts or scrub missions. US withdraws from several join missions like ExoMars (due to Congresses flagrant disregard for international agreements when cutting NASA's budget) have been dimming prospects for the kind of close cooperation we saw on ISS ware the failure of any government to deliver the agreed funding promptly causes head-aches for all.

GW Johnson · 2012-05-21 15:32:56

Depots or stacked vehicles, no difference. With propellants, there are storables (like hydrazine and nitrogen tetroxide, and kerosene), there are cryogenics (like LOX and liquid methane), and there are more difficult cryogenics that leak really easily (LH2).

It would be easy to build a depot or a stack-up vehicle, out of tank modules full of storables, all plumbed together. These would store for many months, even a few years. But, Isp is down under 300 seconds. No practical way around that.

Cryogenics like LOX and liquid methane have a boiloff problem, although it can be mitigated a bunch by a simple sunshade. But, again, Isp is limited to the 300-400 sec range. Really high-pressure engines with big (vulnerable) bells would fall toward the higher end of the Isp range. With extra propellant to cover boiloff, you might get a couple of years storage out of that.

Hydrogen is the toughie! It leaks extraordinarily easy, even right through the metal wall of a tank. Fittings leak all the time. And the boiloff problem is much worse. But LOX-LH2 at high pressure could get you shuttle engine performance or a tad higher: 450-470 sec maybe. Trouble is, we still have no really good way to store vast quantities of LH2 for a 2 year mission. That technology needs a hard look before we fly it to Mars for a round trip with men. Not that it couldn't be done. But you're going to need some kind of a cryo-cooler, and some very stringent leak prevention engineering.

Or, maybe store it and ship it as water, and just make enough hydrogen from that water to support the next burn. Solar electrolysis module, perhaps?

Solve that storage problem with LH2, and not only is LH2-LOX possible for Mars, so also is LH2 solid core NERVA. That one was cancelled about a year before first flight, demonstrating 3/4 hour burns, restarts, and 900 sec at T/W 3.6. That could be done "right now" (see below about resurrecting NERVA).

But, if we resurrected NERVA, why not try to do what they didn't do back then, and do a water NERVA. In fact, water contaminated with the other volatiles ammonia and methane. That stuff is present as ice all over the solar system. Just mine the ice, melt it, filter out the solids, and use it in your reactor. That's real simple ISRU, right there.

There's an Isp penalty, to be sure, from MW 16-18 instead of 2, but I'd bet the higher heat capacity of the water might allow higher reactor power at the same mass throughput, offsetting some of that Isp penalty. Think 700 Isp and fuel all over the solar system. That's an exciting prospect.

Once you get above the 450 sec level with your Isp at high engine T/W, mass ratios get far more reasonable at lower stage count. At 900-100 sec for NERVA, single stage reusable transit vehicles (orbit-to-orbit) and single-stage reusable "landing boat" vehicles start getting really feasible. These can be assembled from 25 ton modules in LEO, and sent anywhere desired.

We did this NERVA thing before. We could resurrect it and do it again, pretty quickly, I.F.F. (if and only if) the right crowd did this. (Wrong crowd would take forever.) It sure makes one whopping difference in one's mission designs.

The shopping list for enabling and continuing technologies for men in deep space appears to me to be (1) modular vehicle designs docked from 25 ton modules, (2) some sort of cryocooler for long term LH2 storage, (3) a sunshade that could double as meteoroid armor, (4) resurrecting NERVA, and (5) working on a water NERVA for later upgrades.

So, why are they instead doing a gigantic 100-ton launcher when we have some 25 ton launchers already, and a 53-tonner on a short path to readiness? Bah, humbug! Politics plays too big a role in government space programs. Ridiculous.

GW

Last edited by GW Johnson (2012-05-21 15:40:39)

SpaceNut · 2012-05-21 21:37:55

A dumb tank needs to have station keeping thrusters for docking, extended power source capable of waiting for the next dumb tank payload ect... for ever how many you want and all the while the fuel if cryo is boiling off and the important part is mission launch cycles and weather caused plus other stuff interuption to the build process....

I think the quickest cycle has been 30 days between launches....

GW Johnson · 2012-05-22 16:00:39

Storables like hydrazine, kerosene, nitrogen tetroxide, and even nitric acid, will last months, perhaps years, without losses. Trouble is under-300 sec Isp. Long storage times solve the launch rates problem, especially if you are launching with more than one outfit's rockets.

Simple cryogenics like LOX can be stored for a fair amount of time with minimal boiloff, if you just reduce the heat load withy a sunshade. That should get you weeks to months. Add a solar-powered cryocooler, and I'd be you could get years out of that. Kerolox Isp 300-330 sec perhaps? LH2-LOX 450-470 Isp perhaps?

Hydrogen is the tough one. Not only is it more subject to boiloff, but more importantly, it leaks like crazy. Fittings are notorious, and gaseous hydrogen will leak riught through the walls of a steel welding gas bottle, on a scale of around 3 months for a 25% pressure drop, typically here on Earth. Sunshield plus cryocooler plus very, very, very careful engineering to eliminate leaks, plus also some extra propellant to cover losses, and I bet we could cover LH2 for 2 years at practical delivery quantities. That makes LH2/LOX and LH2 nukes feasible. A bit of development there, not exactly "shovel-ready", but nothing too difficult there.

Wild idea: ship and store the hydrogen as water instead. Just use a solar electrolysis module to make just-in-time LH2 for just the amounts needed. Requires planning ahead, and requires a non-objectionable trace electrolyte. Salt works, among others. Cycle efficiency is low, well under 50% to be sure. Biggest problem: liquefaction is very energy intensive. We've never, ever done that in space before. The rest, well, I see little real troubles getting that done.

As for thrusters and control smarts on each and every one of a slew of "dumb" tanks: no, not really, not on every tank. Just dock them together. One cluster, flying free. Maybe one propulsive module for the entire depot cluster. One module, not many, with some smarts, and bit of delta-vee for orbit control. And a beacon. That ought to do for any such depot anywhere in the entire solar system.

GW

Last edited by GW Johnson (2012-05-22 16:05:50)

RGClark · 2012-06-13 06:19:44

GW Johnson wrote:

...
BTW, at that convention, I saw another paper that really intrigued me vis-a-vis refueling the reusable transit ship in LEO. There's a group with a light gas gun already launching small test articles for USAF at around half orbital speeds. They presented design projections for a scaled-up gun that could shoot loaded propellant tanks to LEO. Hard tank plus a small solid motor to circularize. Their estimate of cost per pound delivered was amazingly low: under $100/pound. Until we have propellant-making infrastructure in space, that's an amazingly practical way to refuel reusable things in LEO for subsequent missions (including emplacing said infrastructure).
GW

Good point about gas gun launch being doable now for small payloads. Do you have a link to this teams research?
Railgun launch for small payloads is also doable now by scaling up the Navy research with their railguns:

It's real! Navy test-fires first working prototype railgun.
By Kelley Vlahos
Published February 28, 2012
http://www.foxnews.com/scitech/2012/02/ … e-railgun/

Small payload costs to orbit under $100/kg are possible then. When you consider the energy costs to orbit are less than a dollar a kilogram, this is well within the feasibility range.
Propellant costs to orbit at less than $100/kg makes possible a key advantage of SSTO's. If you calculate the delta-v from LEO to the Moon and back, then the total delta-v is less than that required to get to orbit. So if one did have his own private, SSTO vehicle, then with propellant depots, he would also have his own private lunar vehicle. Note this is NOT true of two-stage-to-orbit-vehicles where the upper stage might only get a delta-v of 6,000 m/s or so.
Then this is another reason why the fact that low cost SSTO's are doable now is so important:

The Coming SSTO's.
http://exoscientist.blogspot.com/2012/0 … sstos.html

Bob Clark

New Mars Forums

Announcement

#26 2012-05-17 09:37:09

Re: Control cost or go home

#27 2012-05-17 12:01:45

Re: Control cost or go home

#28 2012-05-17 12:34:15

Re: Control cost or go home

#29 2012-05-17 12:35:00

Re: Control cost or go home

#30 2012-05-17 13:02:18

Re: Control cost or go home

#31 2012-05-17 13:43:33

Re: Control cost or go home

#32 2012-05-17 14:05:18

Re: Control cost or go home

#33 2012-05-17 15:20:47

Re: Control cost or go home

#34 2012-05-17 19:15:15

Re: Control cost or go home

#35 2012-05-18 05:52:48

Re: Control cost or go home

#36 2012-05-18 08:14:14

Re: Control cost or go home

#37 2012-05-18 22:37:55

Re: Control cost or go home

#38 2012-05-19 07:24:26

Re: Control cost or go home

#39 2012-05-19 18:51:12

Re: Control cost or go home

#40 2012-05-20 12:55:00

Re: Control cost or go home

#41 2012-05-20 18:00:26

Re: Control cost or go home

#42 2012-05-20 18:46:32

Re: Control cost or go home

#43 2012-05-21 09:26:53

Re: Control cost or go home

#44 2012-05-21 10:01:07

Re: Control cost or go home

#45 2012-05-21 11:57:22

Re: Control cost or go home

#46 2012-05-21 12:12:16

Re: Control cost or go home

#47 2012-05-21 15:32:56

Re: Control cost or go home

#48 2012-05-21 21:37:55

Re: Control cost or go home

#49 2012-05-22 16:00:39

Re: Control cost or go home

#50 2012-06-13 06:19:44

Re: Control cost or go home

Board footer