Subject: Space-tech Digest #60 Contents: Edmund Hack Re: the other 99.4% of space shuttle launch costs? Dominic Herity Re: the other 99.4% of space shuttle launch costs? Henry Spencer Re: the other 99.4% of space shuttle launch costs? Joe Pistritto Re: the other 99.4% of space shuttle launch costs? Joe Pistritto Re: the other 99.4% of space shuttle launch costs? Henry Spencer Re: the other 99.4% of space shuttle launch costs? Joe Pistritto Re: the other 99.4% of space shuttle launch costs? Dominic Herity Re: the other 99.4% of space shuttle launch costs? Tim Eisele booster scaleup Henry Spencer Re: booster scaleup Bill Davidsen Re: booster scaleup ------------------------------------------------------------ Date: Thu, 10 May 90 16:31:15 PDT From: Edmund Hack X-Vmsmail-To: AMES::"space-tech@cs.cmu.edu" Subject: RE^2 99.4% of STS costs, etc. To: space-tech@CS.CMU.EDU >henry@zoo.toronto.edu >> B. Yamauchi @ U of Rochester >>Well, this may be true in reference to maintenance, but the EVA retreiver is >>designed to be autonomous and not teleoperated. This is a free-flying robot >>designed to recover objects that fall off of Freedom... >One must distinguish between design and actual usage. I'd bet almost any >amount of money that the retriever will be teleoperated when retrieving >anything valuable, at least after the first time it's tried. The shuttle >orbiter was designed to make an autonomous landing (except for lowering the >gear), but has never once been flown that way. The difficulty of NASA EVA >operations in recent times has been directly proportional to the degree >of automation involved. (Compare the Solar Max repair with the Syncom >repair.) Pork-barrel robotics-is-the-wave-of-the-future propaganda aside, >the automated stuff simply will not be trusted with anything vital. With >good reason. My department at LESC is responsible for the control and planning software for EVA Retriever (EVAR) and I was on the project in its early stages. It was going to have a teleoperation mode at one time, but for cost and political reasons it was dropped. EVAR is totally autonomous (except for an emergency kill and a few other simple safety commands). However, it is not yet baselined for use on Freedom, pending a flight experiment on a STS mission (which has not yet been funded, but has made the short list of new flight experiements for the last 2 fiscal years). It will be capable of recognizing and "fetching" tools and astronauts that fall off of the Station. There is some reluctance on the part of flight crews to allow a robot to grab them (which I understand). We expect it to take a couple of more years to mature the software, especially the vision and grasp planning parts. The rest is easy (by comparison). On the use of robotics in STS processing: Lockheed Shuttle Operations Co., which does all STS processing, except payload checkouts in the O&C building, has several systems in various stages of development. The logistics system has a set of robots to fetch parts from the warehouse - supposedly the biggest such system in use. In addition, the problem of cutting and fitting thermal tiles by robot is in active investigation, and may be on line in a few years. One of my staff is supporting trade studies underway by NASA on the use of Humans/Automation/Robotics/Telerobotics for Planetary Surface Systems for the Space Exploration Initiative. The bottom line on Robots/Telerobots is this: an autonomous robot is useful for repetitive and routine tasks like piling up regolith for an oxygen factory on the moon. Telerobots win for few of a kind and dangerous operations (changing out parts of nuclear reactors). Humans win in some situations. Automation wins big for process control, monitoring and diagnosis. Edmund Hack hack@lock.span.nasa.gov Intelligent Systems Dept. Lockheed Engineering & Sciences Houston, TX Note: These are my opinions, not NASA's, LESC's or any known bureaucrat's. Also note: I'm on vacation 5/12-21, blowing bubbles along the reefs of the island of Bonaire. I'll be glad to reply to my reply then. ------------------------------ Date: Fri, 11 May 90 17:32:00 BST From: dherity%cs.tcd.ie@vma.CC.CMU.EDU Subject: Re: What's the other 99.4% of space shuttle launch cost? To: space-tech%cs.cmu.edu@vma.CC.CMU.EDU "Joseph C. Pistritto" says : > Orbiter Cost/Mission: $25000K (no interest) > $63000K (at 9%) > Support Staff: $16000K > Facilities maint.: $ 500K (real lowball estimate) > Ancillary Equipt: $ 250K > > Things I can't estimate now: > Cost of external tank ??????? > Avg. cost of SRBs ??????? > > Total: $41750K ($41million/mission) > (with avg 40K payload) ~$10K/mission/pound > (with max 60K payload) ~$7K/mission/pound Well, that's a pretty comprehensive answer to my question. By the way, the last two figures should be $1K/mission/pound and $700/mission/pound respectively. If we count in the $63M interest (we're interested in REAL cost), we get $2619/pound average and $1745/pound optimum. These figures are in the ballpark of actual charges. In terms of cost, the villian of the piece is the orbiter. If it cost less, carried more and turned around faster, this would have a dramatic impact on the figures. An unmanned orbiter replacement consisting of engines, control, structure, an expendable fairing to cover the payload, heat shield and parachute (instead of wings) would meet these criteria. The orbiter weighs 68 tonnes empty, of which less than 10 tonnes are engines and carries a 30 tonne payload, grossing 98 tonnes. The structure, control, parachute and heat shield of the suggested vehicle should come in under 18 tonnes, leaving a 70 tonne payload capacity. The main difference between this and the shuttle C proposal is that it returns to earth itself. (I figure the most expensive part of shuttle C would be using a shuttle to retrieve it!) Suppose this vehicle cost $1 billion (daft guess) and had a one month turnaround (another daft guess, but plausible). Plug this and a 70 tonne payload into Joseph Pistritto's calculations and you come out with a cost of $222/pound as opposed to $1745/pound. Can anyone do a better analysis than this ? What has been published about shuttle C costs ? ------------------------------ From: henry@zoo.toronto.edu Date: Mon, 14 May 90 15:51:32 EDT To: space-tech@CS.CMU.EDU Subject: Re: What's the other 99.4% of space shuttle launch cost? > The main difference between this and the shuttle C proposal is that it > returns to earth itself. (I figure the most expensive part of shuttle C > would be using a shuttle to retrieve it!) Unless something has changed without me hearing about it, there is no such retrieval in the Shuttle-C proposal. The intent is to use engines that are at the end of their useful lives (which are much shorter than originally intended!), and simply throw them away. The most expensive part of the Shuttle-C proposal is all the other stuff that is being bundled in with it. Somebody either wants it to fail, or is so confident it will succeed that they want to give assorted other things (e.g. upgrades to shuttle support facilities) a free ride on it. The former sounds more likely to me. Henry Spencer at U of Toronto Zoology uunet!attcan!utzoo!henry henry@zoo.toronto.edu ------------------------------ Subject: Re: What's the other 99.4% of space shuttle launch cost? To: henry@zoo.toronto.edu Date: Tue, 15 May 90 9:09:02 MESZ From: "Joseph C. Pistritto" Cc: space-tech@CS.CMU.EDU Mailer: Elm [revision: 64.9] > Unless something has changed without me hearing about it, there is no > such retrieval in the Shuttle-C proposal. The intent is to use engines > that are at the end of their useful lives (which are much shorter than > originally intended!), and simply throw them away. Well gosh, those engines are going to end up in LEO right? So why not collect them up there, then reuse them (either as parts or as complete assemblies for launching lunar/Mars missions). Those things are awful heavy to throw away once you've just about got them in orbit anyway. (of course, I think the same thing about external tanks, so just color me 'cheap'). -- Joseph C. Pistritto (cgch!bpistr@chx400.switch.ch, jcp@brl.mil) Ciba Geigy AG, R1241.1.01, Postfach CH4002, Basel, Switzerland Tel: +41 61 697 6155 (work) +41 61 692 1728 (home) GMT+2hrs! ------------------------------ Subject: Re: What's the other 99.4% of space shuttle launch cost? To: cs.tcd.ie!dherity@vma.CC.CMU.EDU Date: Tue, 15 May 90 9:29:08 MESZ From: "Joseph C. Pistritto" Cc: space-tech@CS.CMU.EDU Mailer: Elm [revision: 64.9] > > Total: $41750K ($41million/mission) > > (with avg 40K payload) ~$10K/mission/pound > > (with max 60K payload) ~$7K/mission/pound > > Well, that's a pretty comprehensive answer to my question. By the way, > the last two figures should be $1K/mission/pound and $700/mission/pound > respectively. Oops, I goofed. Sorry about that. > If we count in the $63M interest (we're interested in REAL > cost), we get $2619/pound average and $1745/pound optimum. These figures > are in the ballpark of actual charges. Yes, but they're too low. SRBs are probably a few million per mission, just as a guess, and you need two of them. The external tank costs something too. > > In terms of cost, the villian of the piece is the orbiter. If it cost less, > carried more and turned around faster, this would have a dramatic impact > on the figures. It certainly would. Now you figure out how to turn around SSMEs in less than 30 days, then go straight to the head of the class. Those things are *COMPLEX*, and one little step left out has at least a theoretical possibility of wasting the entire stack! Not to mention all those other little things that go wrong, like APUs. If we were trying to make money (or not lose it), with this thing, what do you think it'd cost to have a stack sitting on the pad for 2 to 3 weeks extra to replace a Freon line, for instance... > An unmanned orbiter replacement consisting of > engines, control, structure, an expendable fairing to cover the payload, > heat shield and parachute (instead of wings) would meet these criteria. > The orbiter weighs 68 tonnes empty, of which less than 10 tonnes are engines > and carries a 30 tonne payload, grossing 98 tonnes. > The structure, control, parachute and heat shield of the suggested > vehicle should come in under 18 tonnes, leaving a 70 tonne payload capacity. > The main difference between this and the shuttle C proposal is that it > returns to earth itself. How? the heavy part of the shuttle is the 'wings' and all the aerodynamic structures to support them. Better make it entirely expendable. If you want to reuse the engines (probably a good idea), then mount them in removable pods at the back, carry them to orbit near the space station, then have an astronaut detach them, then stuff them into a shuttle going back down. (Payload space is real AVAILABLE going from orbit to the ground). > (I figure the most expensive part of shuttle C > would be using a shuttle to retrieve it!) The only part worth retrieving at all is the engines. And they might be better used on orbit anyway. We would need an on-orbit pressurized service structure, however to make this cost effective. How about a big (50m or so) kevlar ball? That way your astronauts could work in a shirtsleeve environment while servicing things. What we need is a space mechanic! > Can anyone do a better analysis than this ? What has been published > about shuttle C costs ? > Probably almost anyone at NASA, for instance. I know that the 'Space Shuttle Operating Manual' or something like that has a section on costs. Also check the filings NASA makes with Congress (probably in the Congressional Record). I bet they're some gross level cost data in there. I believe the OTA made a study as well, but I couldn't be sure of it. Possibly some data in the Challenger commission report as well, I know they analyzed several of the manufacturing steps. -- Joseph C. Pistritto (cgch!bpistr@chx400.switch.ch, jcp@brl.mil) Ciba Geigy AG, R1241.1.01, Postfach CH4002, Basel, Switzerland Tel: +41 61 697 6155 (work) +41 61 692 1728 (home) GMT+2hrs! ------------------------------ From: henry@zoo.toronto.edu Date: Tue, 15 May 90 13:13:54 EDT To: space-tech@CS.CMU.EDU Subject: Re: What's the other 99.4% of space shuttle launch cost? > > ... in the Shuttle-C proposal. The intent is to use engines > > that are at the end of their useful lives (which are much shorter than > > originally intended!), and simply throw them away. > > Well gosh, those engines are going to end up in LEO right? So why not > collect them up there, then reuse them (either as parts or as complete > assemblies for launching lunar/Mars missions). Those things are awful > heavy to throw away once you've just about got them in orbit anyway. Yes, they're heavy, but so what? At the end of their useful lives they are so much junk, not worth saving unless you've got a use for orbiting metallic junk. Dismantling or refurbishing them in orbit is going to be a daunting job, and I don't think very many of the parts will be useful for other purposes. Actually, there *is* a use for orbiting metallic junk, if it's in just the right place. The reboost fuel requirements for maintaining a space station's orbit are independent of its mass; heavier stations are harder to reboost but don't need it as often. (Roughly speaking, the requirement is that the average rocket thrust, averaged over the whole period between reboosts, equals the average air drag, so mass does not enter the picture.) Providing it does not increase the frontal area significantly, or cause difficulties with attitude control etc., accumulating massive non-outgassing junk at the space station is an excellent idea. The microgravity people would much prefer vigorous reboosts at long intervals to little ones all the time. Henry Spencer at U of Toronto Zoology uunet!attcan!utzoo!henry henry@zoo.toronto.edu ------------------------------ Subject: Re: What's the other 99.4% of space shuttle launch cost? To: dherity@cs.tcd.ie Date: Tue, 15 May 90 18:38:01 MESZ From: "Joseph C. Pistritto" Cc: jcp@cgch.uucp, space-tech@CS.CMU.EDU Mailer: Elm [revision: 64.9] about SSME's heat shield, parachute, etc. This is in actuality very close to the ides of SSX, (except you replace the SSMEs with something simpler, like RL-10s). With SSX, you have a big capsule (sort of the same shape as the apollo capsule, just lot's bigger), and a whole bunch of liquid fuelled engines. You run the engines going up and coming down (although less so). You make the whole thing fairly rugged, and include extra engines so that a shutdown of 1 or 2 doesn't blow the whole mess. You can even test it sub-orbitally. It's a one piece vehicle (no expendable parts). Yes, it's heavier than the shuttle because you can't throw stuff like SRBs and external tanks away, but all that costs you is fuel, and as we started the discussion realizing, fuel is cheap. SDIO has an RFP on the street for this *NOW* (came out about 2 weeks ago in the Federal Register). -jcp- -- Joseph C. Pistritto (cgch!bpistr@chx400.switch.ch, jcp@brl.mil) Ciba Geigy AG, R1241.1.01, Postfach CH4002, Basel, Switzerland Tel: +41 61 697 6155 (work) +41 61 692 1728 (home) GMT+2hrs! ------------------------------ Date: Tue, 15 May 90 14:08:44 BST From: dherity%cs.tcd.ie@vma.CC.CMU.EDU Subject: Re: What's the other 99.4% of space shuttle launch cost? To: "Joseph C. Pistritto" Cc: space-tech%cs.cmu.edu@vma.CC.CMU.EDU >It certainly would. Now you figure out how to turn around SSMEs in less than >30 days, then go straight to the head of the class. Those things are *COMPLEX* , >and one little step left out has at least a theoretical possibility of >wasting the entire stack! Not to mention all those other little things that >go wrong, like APUs. My point is that with a simple reusable booster, there are less 'little things'. Factors improving turnaround time are : There's a LOT less to maintain. Corners can be cut because it isn't manned. It returns to earth immediately after launch. >> The main difference between this and the shuttle C proposal is that it >> returns to earth itself. >How? the heavy part of the shuttle is the 'wings' and all the aerodynamic Remember Mercury, Gemini, Apollo, etc? A heat shield and parachute are all you need. Early '70s technology could aim an Apollo capsule within about five miles from the far side of the moon. Why would you need a gliding range of 1100 miles, as per the shuttle, coming from LEO? The shuttle's heat shield has to be tiled because it flexes when flying. If you get rid of the wings, you vcan replace the tiles with a single piece heat shield which should be reusable at least a few times. Replacement isn't labour intensive. >structures to support them. Better make it entirely expendable. If you want >to reuse the engines (probably a good idea), then mount them in removable pods >at the back, carry them to orbit near the space station, then have an astronaut >detach them, then stuff them into a shuttle going back down. (Payload space >is real AVAILABLE going from orbit to the ground). This all assumes you put up a shuttle in addition every time you launch - an extra $100M+? for an extra 30 tonnes. This will wreak havoc with your cost savings. I would hope that the reduced cost would create a market much larger than that for the shuttle, while reducing demand for the shuttle. So free rides down on a shuttle aren't available. >The only part worth retrieving at all is the engines. And they might be >better used on orbit anyway. We would need an on-orbit pressurized service >structure, however to make this cost effective. How about a big (50m or so) >kevlar ball? That way your astronauts could work in a shirtsleeve environment >while servicing things. Engines in orbit ain't worth a damn without fuel. And you get fuel by sending up more vehicles and thus more engines. And if it's expensive to service them on the ground, it will be a LOT more expensive to service them in orbit. No. The answer is replacing the orbiter with a bunch of SSME's, a heat shield and a parachute. ------------------------------ Date: TUE, 15 May 1990 08:35:41 EST From: Tim Eisele To: space-tech@CS.CMU.EDU MMDF-Warning: Parse error in original version of preceding line at CS.CMU.EDU X-Mailer: VM/Netnote V1.05 Subject: booster scaleup This is kind of tenuously related to the discussion of shuttle costs, but here goes anyway: What kind of economies of scale do you get by making boosters larger, in terms of dollars per pound orbited? For that matter, how small can a rocket be and still put some arbitrarily tiny (say 10 grams) object into orbit? And what is it that puts the lower limit on the size? (air drag, mass of the avionics, or what?) Presumably there will be some size of unmanned, expendable launcher where the cost per pound will level out, but will this occur at about the scale of a Delta, or somewhere over the size of an Energia? Tim Eisele Michigan Tech University TCEISELE at MTUS5.BITNET ------------------------------ From: henry@zoo.toronto.edu Date: Tue, 15 May 90 12:24:15 EDT To: space-tech@CS.CMU.EDU Subject: Re: booster scaleup > ... What kind of economies of scale do you get by making > boosters larger, in terms of dollars per pound orbited? ... For current US boosters, the big cost is the "standing army" of staff. The size of that army scales strongly with how ambitious your technology is, but only very weakly with size. A Titan IV doesn't need a lot more babying than a low-end Delta. So to a very sloppy first approximation, cost per ton is inversely proportional to size. It's not hard to show that there are advantages of scale in the technology as well, but the bureaucratic aspects currently dwarf them. > how small can a rocket be and still put some arbitrarily tiny (say 10 grams) > object into orbit? And what is it that puts the lower limit on the > size? (air drag, mass of the avionics, or what?) Many things scale non-linearly, and I'm not sure offhand what the biggest problem would be, but one obvious one is that things like guidance systems basically don't scale at all. Henry Spencer at U of Toronto Zoology uunet!attcan!utzoo!henry henry@zoo.toronto.edu ------------------------------ Date: Tue, 15 May 90 13:30:01 EDT From: davidsen@crdos1.crd.ge.com To: space-tech@CS.CMU.EDU Subject: Re: booster scaleup Reply-To: davidsen@crdos1.crd.ge.com Sender: mnr@DAISY.LEARNING.CS.CMU.EDU One cost limit is acceleration. You can use a lot less fuel boosting at ten G than 3, but some cargo can't take it. Use of wings instead of heat shield is another fixed cost, since you have to push them into orbit. Date: Wed, 16 May 90 09:54:46 BST From: dherity%cs.tcd.ie@vma.CC.CMU.EDU Subject: Re: booster scaleup To: space-tech%cs.cmu.edu@vma.CC.CMU.EDU > One cost limit is acceleration. You can use a lot less fuel boosting > at ten G than 3, but some cargo can't take it. Use of wings instead of A rocket produces a thrust per unit mass of 'a'. To counter the earth's gravity, it must produce a vertical component of 'g-v^2/r' when travelling horizontally at speed 'v', where 'r' is the earth's radius and 'g' is the acceleration due to gravity. So the acceleration horizontally is 'dv/dt = sqrt(a^2-(g-v^2/r)^2)'. So the time needed to reach orbital velocity 'vo' is integral from 0 to vo of 'dv/sqrt(a^2-(g-v^2/r)^2)'. If it didn't have to counter gravity, it would accelerate to orbital velocity 'vo' in time 'vo/a'. So the effect is of having to reach a higher velocity. The ratio is 'integral/(vo/a)'. Substituting 'g=vo^2/r', 'x=v/vo', 'A=a/g' in the above simplifies the equation to 'ratio = integral from 0 to 1 of (dx/sqrt(1-(1-x^2)^2/A^2))'. Solving this integral proved beyond me, so I made the simplifying assumption that 'a>>g => A>>1', which is reasonable for 'A>=3'. (I confirmed this by solving the above integral numerically for A=3 and comparing the exact to the approximate result.) This permits the approximation : '1/sqrt(1-(1-x^2)^2/A^2) = 1 - 1/2/A^2*(1-x^2)^2', so the integral solves to : 'ratio = 1 + 4/15/A^2'. Therefore, for A=3, ratio = 1.03, which means that if a rocket starts from the ground at 3*g and goes to 'vo', it must expend the fuel required to reach '1.03 * vo'. Similarly, for A=10, ratio = 1.0027. If vo = 7800m/s, A=3 has a 'penalty' of 0.03*7800m/s = 234m/s and A=10 has a penalty of 0.0027*7800 = 21m/s. If H2/O2 fuel is used, with an exhaust velocity of 4400m/s, these velocity penalties translate to mass penalties of 5.3% and 0.47% respectively. That is to say, a rocket with 3g acceleration must add 5.3% to its mass in extra fuel to counter gravity. H2/O2 engines deliver 80 times their own weight in thrust. (This figure is based on the Space Shuttle main engine, and s believed to be typical.) So to get 10g acceleration, you must add 10/80 = 12.5% to your rest mass for engines. For 3g acceleration, add 3.75%. So, to counter gravity, a 3g rocket must have 5.3% + 3.75% = 9.05% extra mass and a 10g rocket must have 0.47%+12.5% = 13% extra. So a 3g rocket comes out smaller. This ignores the extra structural mass of a 10g rocket. Also, extra engine mass is a bigger penalty than extra fuel mass because the engines must be carried all the way to vo. But I ignore this bacause the above maths are bad enough already :-). ------------------------------ End of Space-tech Digest #60 *******************