Right Side Up: A History of the Space Transportation System

Chapter 1: Preflight
“In short, the Space Shuttle is so inefficient because it is built upside-down.”


Chapter 1: Preflight


Technicians swarmed around the gleaming white delta-winged shape, mostly around the nose and tail, but some at strategic points along the length, at the engine bays, the landing gear wells, and the control surfaces on the aft side of the wings. The ship’s gleaming white aluminum skin was inspected, with sections yellow or browned with use cleaned and checked. The more resistant titanium armor on the belly, the blunt nose, and the wing leading edges was checked as well. The mighty F-1B engines were inspected and, where needed, swapped out for maintenance. Though they were rated for many more flights, this was to be the highest-profile mission yet in the Space Shuttle program—no one at Boeing or at NASA wanted to take chances now.

Several long gray cables trailed from two boxes embedded into the walls of the flight deck to the hangar floor, where they plugged into a console atop which sat a bulky CRT monitor. Green text on a black background reflected off an engineer’s glasses as he inspected the stored flight data from the last test flight and as his teammates checked that the computer, responsible for the fly-by-wire actuation of the control surfaces, measurement of fuel levels, and the limited life-support capacity of the flight deck, and countless other systems, responded properly to simulated inputs. The comparatively modern IBM AP-101 was a massive leap over the core rope that had graced the Apollo Guidance Computer, and enabled a lot more functions to be off-loaded to the vehicle--and given the flight regime for which it was designed, that was necessary.

Behind an access panel between the LOX tank forward bulkhead and the flight deck hatch, a technician ran a very careful low-power test of one particular circuit, the one that controlled the pyrotechnics that fired the escape pod. A far cry from the launch-abort towers that had protected the Apollo astronauts, but still far better than the ejection seats with which the Gemini crews had had to make do, this system ensured the survival of the crew should the worst happen. This was something the technician was unable to forget, with her supervisor looking over her shoulder at the multimeter in her hand, and with a poster of Snoopy in an orange flight suit reminding her that “Mission success is in YOUR hands!” hanging on the hangar wall. The results checked out, verified by the supervisor with a little help from his pocket calculator--a new model, with an LCD display—as far as they could tell, this system was good to go. The supervisor checked that particular circuit off of the dot-matrix checklist on his clipboard, and they moved on to testing the redundant and triple-redundant back-ups. This particular access point was located near the top of the vehicle--by the light filtering in through open access panels all around, the technician could just make out the yellow-painted bulk of the LOX tank’s forward bulkhead, and the small propellant tanks that fed the separation motors and reaction-control thrusters. Even with those here, the nose of the vehicle was a cavernous void--a vestige of the original design scheme, which had called for the nose to retract back into that void.

Around the hangar, similar inspections ultimately yielded the same results. All five engines were flight-worthy, the control surfaces demonstrated exactly the desired range of motion, the hydraulic actuators that controlled the covers over the jet engines performed as expected, landing gear wheels rotated freely, and the dials on the flight deck were all illuminated perfectly. The last hatches were dogged shut, umbilical cables pulled out, and access ladders wheeled away as an airport tow truck with a bright-red NASA worm on its front and sides rolled in. Pinned securely to the truck, the spacecraft left the fluorescent lighting of the hangar for Florida’s brighter morning sun, the massive American flag painted on each side of her fuselage breaking up her otherwise clean white appearance.

RS-IC-102, “Constitution,” had a date in the VAB.


The transition from a Saturn V first stage to the reusable booster of the Space Transportation System seems obvious and natural in hindsight, almost two decades removed from the birth of either system. However, the Reusable Booster had a much more complex history than many assume, and a close study of the complex origins of the idea illustrates how the most “optimal” design in aerospace can depend on a variety of definitions. Marshall Space Flight Center funded the first studies of what eventually became RS-IC in 1962, with the publication of a study titled “50- to 100-Ton Payload Reusable Orbital Carrier.” Though previous studies had found that retrofitting the S-I stage of the Saturn I and IB with a flexible and deployable wing would be impractical, this study concluded that the much larger S-IC on the Saturn V had more room for improvement. This study envisioned an S-IC modified with landing gear, sharply-swept delta wings with large vertical tips, a flight deck, and modest thermal insulation to protect the booster from the heat of sub-orbital reentry. Boeing developed the design in more detail as the “Model 922,” studying several variants. In the most powerful of these, the Model 922 booster would be paired with an unmodified Saturn V second stage, retaining its full lifting power. This pairing, the Model 922-104, produced a booster that returned the first stage while losing only 20% of its lift capacity. Though these studies were not pursued in the early 1960s (all of NASA’s attention going to getting S-IC and the other parts of the Apollo-Saturn system flying at all), it did plant the first seeds of the flyback first stage in the minds of Marshall and Boeing engineers.

In 1965, Congress began trimming NASA’s budget, which by that point had exceeded $5 billion per year. Smelling a coming storm on the wind, Marshall Space Flight Center and the prime contractors on the Saturn V (Boeing, North American, and Douglas) began studying lower-cost variants of the Saturn system, in order to keep it in service even in the face of future budget cuts. Boeing’s studies were the most wide-ranging, covering Saturn variants from the smallest (~20 tonnes to LEO) to the largest (over 200 tonnes to LEO) capacities. Of most interest to MSFC at the time, however, was Saturn INT-22, a combination of a winged S-IC with a reduced-cost S-IVB to yield a launch vehicle of 45 tonnes capability for a significantly lower cost per-launch than either the Saturn V or Saturn IB. A particularly revolutionary innovation in this study was the concept of “propellant ballasting.” By carrying more propellant than strictly necessary for lower-end payloads, and burning it off in a second post-staging burn of the first stage, reentry velocity could be reduced considerably for smaller payloads (like those needed to service a space station), extending stage life. Indeed, with sufficient ballasting, a payload of 25 tonnes could be delivered with such minimal heating on the booster that the existing aluminum skin of the S-IC would suffice for thermal protection. Though the INT-22 study did not become NASA’s official policy, it was favorably-enough received at MSFC to become the assumed baseline booster for post-Skylab space station programs, and featured prominently in Apollo Extension Series (later Apollo Applications Program) studies.

One should not be fooled by the prominent wings on the INT-22 first stage—this vehicle was not a shuttle, or at least not The Shuttle as that term was understood by NASA in the late 1960s. Shuttle was supposed to be a complete break with the Apollo Program, a fully-reusable, two-stage-to-orbit system propelled by high-thrust staged-combustion hydrogen-burning rocket engines. Even at Boeing and Marshall, this understanding of the plan for the 1970s was inherent in their plans for INT-22—it was to be an interim solution, providing for early Space Stations until Shuttle came into service around 1977. The economic justification for putting wings on the S-IC assumed that the system would be phased out by 1980. When funding for a second run of Saturn components did not materialize by the end of the Johnson Administration, Boeing turned away from INT-22, and instead turned its focus to the two-stage Shuttle. The termination of the Apollo Applications Program and the shifting of focus at NASA from Space Stations to a reusable Space Shuttle in 1969 would have sent INT-22 to join NERVA, X-20, and Project Orion on the heap of space might-have-beens, were it not for a surprise decision by NASA in summer of 1970 to take a second-look at alternative Space Transportation System architectures.

NASA’s Space Shuttle contracting process was divided into four Phases--A, B, C, and D. Phase A consisted of preliminary studies to determine the technical feasibility of an approach to the Shuttle problem. Phase B consisted of detailed studies and preliminary design, while C and D covered test articles and final development, respectively. NASA selected two companies to receive Phase B contracts in May, 1970, North American Rockwell and McDonnell-Douglas, deeming their proposals the strongest. Grumman Chairman Lew Evans, however, raised a massive complaint to Tom Paine’s office, strongly condemning NASA’s preferred Shuttle architecture and blaming Grumman’s loss on weak support from New York’s senators and accusing NASA of playing favorites with North American. Though he was unsuccessful in winning Grumman a Phase B contract at that time (and arguably contributed to the rift that had always existed between Grumman and NASA executives), Evans was persuasive enough, and Grumman’s proposal good enough, for NASA to finance studies of alternative Shuttle architectures. Grumman won the largest of these contracts, but lacked experience with large booster development, and so reached out to Boeing for a collaborative approach.

The Grumman/Boeing proposal differed from the first successful Phase B contracts by incorporating disposable liquid hydrogen tanks. Working in close concert with Max Faget and his team at NASA, Grumman engineers under the direction of Tom Kelly proposed to use disposable, external hydrogen tanks to reduce the weight of the orbiter while at the same time increasing its delta-v capability. This allowed the booster to separate from the orbiter at a lower speed, reducing thermal loading on it and bringing the booster back into the flight regimes studied by Boeing for the INT-22 proposal years earlier. Grumman presented this modified Orbiter design at the Manned Spaceflight Center in Houston in November of 1970. By March of 1971, they had successfully persuaded NASA that their approach was the best, and the agency mandated that the previous Phase B winners, North American Rockwell and McDonnell-Douglas, redesign their Orbiters with external tankage. In May of 1971, working again with Max Faget, Grumman took the next logical step and moved the oxygen tanks out of the Orbiter as well, putting all the Orbiter’s propellant in a disposable, belly-slung aluminum tank. The Booster-Orbiter stack was now somewhat lopsided, as the orbiter hung off the side of the stack, but the numbers didn’t lie--it was as close as NASA had gotten to reaching the peak annual spending cap of $1 billion mandated by the Office of Management and Budget.

Boeing’s management at the time was concerned about the company’s ability to survive the greatest aerospace recession since 1945. Between 1968 and 1971, ¾ of the commercial airplane sector of the company was laid off. These lay-offs rippled across the greater Seattle economy--suburban vacancy rates reached 40%, automobile dealerships collapsed for want of buyers, and so many people fled town that local U-Haul agencies ran out of moving equipment. Two real-estate men in Seattle put up a billboard near the airport, showing a lightbulb hanging on a wire, captioned “Will the last person leaving SEATTLE turn out the lights.” The Boeing 747 was not finding buyers fast enough to cover its development cost, and the US Senate was beginning to move against the Boeing 2707 Supersonic Transport; objections to noise and air pollution by the latter were finding sympathetic Senators in many states not tied to aerospace. The Shuttle became seen by some in Boeing management as critical to keeping the lights on.

By moving the Shuttle Booster’s flight regime back into Boeing’s field of expertise, Grumman offered a way for both companies (for Grumman, too, was suffering from the strain of the F-14 Tomcat program) to save their own skins. By leveraging Grumman’s experience in manned spacecraft and Boeing’s experience in both large supersonic vehicles and large booster development, the two companies hoped to give NASA an unbeatable offer--a Shuttle system more conservative than the main Phase B studies, one easier to develop as it used more off-the-shelf technology, and yet one that still achieved all the payload-servicing, station-resupplying, satellite-deploying objectives NASA wanted in a package that was at least 90%-reusable. It was a match made, so to speak, in heaven, that would allow each company to keep the spacecraft and booster capabilities they’d so painstakingly built up over the past decade--or so it seemed.

The honeymoon ended in late summer of 1971. The Reusable Booster, Reusable-but-with-drop-tanks Orbiter architecture got NASA closer than any other to the OMB’s funding cap--but it still peaked at $1.5 billion per year, half a billion dollars more than OMB would endorse. With the appointment of the new NASA Administrator, James Fletcher, the agency finally accepted that it could not develop the entire Shuttle system at once--the booster and orbiter would have to be developed in a phased development system, one at a time. Though Grumman and Boeing were researching very dissimilar products, they became competitors over scarce funding--NASA would either buy Grumman’s Orbiter, Boeing’s booster, or neither, but it certainly would not buy both at once.

The Space Shuttle Decision, by August of 1971, was reaching its endgame. At this time, on the recommendation of President Nixon’s science advisor, Edward David, a new panel, chaired by Alexander Flax, President of the Pentagon think-tank, Institute for Defense Analysis, was convened to independently analyze the Space Shuttle program. During the summer and autumn of 1971, this panel would meet once a month, meeting with NASA and with the Shuttle contractors. It was during these months that Boeing and Grumman, Marshall Space Flight Center and the Manned Spaceflight Center, would make their own cases to the committee and seek approval for their own preferred option.
 
Very interesting. I admittedly wasn't expecting the S-1C to be used as the basis of it (why is it always the F-1 that has to be used in space TLs?) but will be interesting to see.

Also perhaps utilize the threadmarks for this TL so someone could breeze through in reading the TL without the comments?
 
Last edited:
Very interesting. I admittedly wasn't expecting the S-1C to be used as the basis of it (why is it always the F-1B that has to be used in space TLs?) but will be interesting to see.

Also perhaps utilize the threadmarks for this TL so someone could breeze through in reading the TL without the comments?
We have threadmarks on Althistory? I didn't realize that was a thing here. I use it on other forums but didn't think it was included here.
 
Because the F-1B in this case would need to be restartable and reusable, compared to the single use F-1 and F-1A

I had meant more the F-1 in general, and added the B by accident there (which I removed). Just a little of a pet peeve of mine how it is always the F-1 that survives. But it detracts from this TL (which is just going to be very fascinating to read through) and I would be happy to continue this kind of conversation with you over PM.
 
Subscribed

This very interesting idea Booster and Orbiter as Phased program
James Fletscher proposed analog idea that Orbiter use J-2S engines and later R&D the High pressure SSME.

This is Saturn Shuttle from TL 2001: A Space-Time Odyssey
it use a modifed Boeing F-1, the Cargo version is similar to INT-22 study from this TL, use engine are the F-1A and J-2S (10 reuse)
But the depicted orbiter use integral tanks not the External tank with orbiter described in this TL

15071265380_e85c5b33c6_o.jpg
 
Just to clarify: What is the exact point of departure here?
The exact PoD is the inclusion of the INT-22 in the 1965 studies (the 1962 study is actually historical), though the first point where the timeline diverges from OTL in a way anyone but a major space nerd would notice is actually still yet to come.
 
Back when I was a BIS member, I watched with horror and bewilderment as the Shuttle program become a slo-mo train-wreck...

Didn't learn until much later that the *huge* cargo bay was specifically sized to ferry TLA spy satellites with their long, long optics or vast antennae.

And, of course, when the Shuttle project ran late, the TLA stuff went back to trad launchers, costing NASA that business...
 
Of course subscribed!

My questions in ETS, posted just as this thread went up, are pretty well answered--it will not be a TL where a parallel burning SSME stack with separate return of the SSMEs makes an appearance, or anyway if it does it will be by a very different path. My notion also diverged ideally from established STS tech in wishing to drastically revise the boosters--ideally replace them with much smaller strap-on liquid boosters, so that various sizes of launch could be accomplished by simply adding on boosters around various sizes of tank, with various numbers of SSMEs clustered on the bottom. This would be how economy would be achieved, right-sizing the system for various loads and accomplishing the rare super-large launches with lots of clustered boosters and engines. But it would only make sense to do this if the STS were first developed in OTL form--someone visionary enough to conceive of the sort of launch system I was aiming at from the beginning, at the time of the "Shuttle Decision" or before, would probably not design key elements of STS.

In such a TL--or say, this one--for instance, there is really little reason to design any SSMEs at all. Five or six J-2 type engines do the job of 3 SSMEs quite well--except for the hitch that the J series engines would not be much good at sea level, due mainly to the gas generator pumping system. In principle gas generators can deliver power to turbo pumps at sea level just fine, but it seems the designers of the original J-2 took advantage of the knowledge that the engine would be lit only at high altitude to assume near-vacuum for the turbine exhaust, and thus could go with lower pressure gas generation which presumably is much easier to make work in a lightweight yet reliable fashion. But trying to run it at sea level means an anemic and probably unreliable GG, unable to deliver the power to pump the main chamber fully and thus further impeding the performance of a hydrogen engine which I gather is already inherently handicapped at sea level. I am unsure whether it is true that the higher the vacuum ISP of a rocket, the more severely in proportion its performance suffers at sea level, aside from other issues such as nozzle design, or if in fact the hit they take in ISP is higher in seconds just in proportion to the higher theoretical ISP. But in any case the original J-2 would be useless at sea level.

It may be that J-2S already addressed this issue robustly, and without increases in engine weight or complexity, could handle sea level burns--my impression is, it could handle them better than the original engine but still rather poorly compared to a ker-lox engine designed for sea level burning.

Thus, the part of the Shuttle Decision that resolved that the high-altitude upper stage sustainer engines should also burn at sea level and provide significant lift, I gather mainly because they wanted to be certain before lighting the SRBs that the main engines would all perform well, is what mandated the development of some engine that in some way or other would resemble the SSME. That is, apparently the extremely high 200 atmosphere chamber pressure of the SSME is mainly to enable gas to flow well against sea level pressure; this in turn required extremely sophisticated and complex pumping systems and ultimately in two stages of staged combustion and turbo pumping the propellant to two pressure levels.

But it seems clear from what we are told thus far that the ATL Boeing proposal assumes that the high-altitude engines are air-lit, in near vacuum. In this case there is no need to take such radical measures to enable firm thrust at sea level, instead the already developed, tested and improved J series would suffice pretty well. This in turn takes off much of the pressure to actually recover those engines--although developing, along with a reusable F engine, also a reusable J engine would be a possible choice to make.

Alternatively the SSME does deliver moderately higher vacuum ISP than the J-2S, although nothing ever seems to catch up to the moving target of the ISP of the ever-improving RL-10--which however suffers from a lower T/W ratio and having a really small thrust indeed. Along with, or instead of, making a reusable J engine is the option of making it more efficient with a raised ISP. And indeed the J-2S is notably improved over the J-2 in ISP.

It seems to me that if the designers step back from the mantra of "everything is reusable" and focus on "step 1, let's use reusability to cheapen the most expensive part of the rocket," they must immediately focus on the booster, and let slide the upper stages that are both harder to recover and much cheaper to make, being much lower in mass.

The opening posts do make it clear that in fact in these early days, NASA is indeed still obsessed with making everything reusable and all at once, and thinking anything less is a job half done. It is only the grim reality of OMB being immovable on the matter of the billion dollar a year budget cap that can force them to back away at all, and the external propellant tank concept as in OTL is seriously considered only because it does lower the development cost a whole lot.

With the mention of the Flax committee though I wonder if NASA will do a Hail Mary Pass and strive to give Grumman something while they accept that step one is to do the booster, and seek to package a smaller Flax mini-Orbiter atop the Mark 1 Boeing Booster. An Orbiter meeting OTL standards without SSMEs would need to have some 5 or 6 J-2S or better engines installed. But a smaller one would use fewer, perhaps all the way down to just one. The fewer J engines the Orbiter needs, the less important it is to bring any of them back to Earth. A design then with the J engines on the bottom of a tank, essentially a traditional disposable upper stage, would be viable; instead of stacking the Shuttle sidesaddle it could go on the tip, where safety emergency escape mechanisms are much simpler to design and incorporate, and where it would be immune to any debris coming off a hydrogen upper stage.

Sidesaddle is still possible of course. And so is placing the engines on the Orbiter.

A Flax mandated small Orbiter would not be viable as a cargo carrier, to be justified at all it would have to be focused on roles where crews are required by the nature of the task. However the same stack that can orbit a small Orbiter can also orbit substantial payload, in the 20-30 ton range at least, if it does not have to be accompanied by an Orbiter, and a stack of doubled capacity can launch that cargo and an Orbiter to service it at the same time, though the Orbiter still can't recover much as down mass.

An Orbiter freed of the obligation to bring back to Earth the engines that put it into orbit can be much lighter for a given mission capability.

All right that's enough jumping the gun about mini-Orbiters. It seems likely NASA will try to grasp at any straw to keep both Boeing and Grumman happy, but push will come to shove, and the agency will choose the booster over an Orbiter handicapped by the lack of a funded sufficient booster. The author highlights how by "Shuttle" NASA and the US industry mean specifically the Orbiter part, not the boosters, and to postpone the Orbiter is in public perception to abandon "Shuttle" completely. With that kind of thinking, there may never be an Orbiter of any similarity either to STS Orbiter or the smaller fly-back proposals suggested like HL-20, Hermes or so on. Instead, strapped for development funds but equipped with a flexible fly-back booster, NASA may either spend a decade in the wilderness with no manned launch capability as OTL but with no goal of having one in particular in sight, or finding small sums of extra money later, use it for minimal development of a capsule based system.

Martin having developed Gemini, and also supplying the Air Force with Titan missiles and derived boosters, development of a very flexible Boeing Booster will hurt their missile-launching division pretty badly--so compensating in funding log-rolling by awarding them the manned orbital vehicle competition and essentially mandating a low-cost, low-risk Big Gemini reiteration might be in the cards, to smooth the way for general acceptance of the Boeing booster.

The idea of "fuel ballasting" is new to me and in a perverse way, pretty ingenious. As I said above my own idea for how to provide for a very wide range of payload goals with one economical system was to design fairly small strap-on boosters sized for the smallest loads and add on more and more of them as payload goes up.
 
your quite right, Shevek 23
there were three reasons the SSME was build
One: reusability for 100 ignitions compare to 10 for J-2S
Second: needed Higher thrust as J-2S can provide
Third: the J-2 issue to ignite under sea level air pressure.

North American Aviation (Rockwell) proposed next reusable Two stage wings Saturn V.
the use of S-II stage as launcher Saturn INT-17 to INT-19
but it issue were J-2 on sea level so they proposed either J-2 modified to a lower expansion ratio engine.
or use Solid motors booster from Minuteman or Titan with altitude ignition of J-2
But NASA drop the proposal 17 to 19 because of Poor performance and cost-effectiveness...

This Boeing flyback booster is expensive to build
it not simply put a cockpit and wing with engines on Saturn I-C stage and glue a coating on it
this RS-IC need titan and steel alloy to replace some of the aluminum-magnesium alloy structural parts to fit the Wings and take the forces the Wings endure during take off or landing.
Next to that are other issue like F-1 engines covers made in time from Asbestos
 
...
This Boeing flyback booster is expensive to build
it not simply put a cockpit and wing with engines on Saturn I-C stage and glue a coating on it
this RS-IC need titan and steel alloy to replace some of the aluminum-magnesium alloy structural parts to fit the Wings and take the forces the Wings endure during take off or landing.
We'd surely expect the BB to be a major effort. "simply put a cockpit..."--Simply? I am always amazed that they seriously intended to put a human flight crew on a booster rocket! When the model was horizontal takeoff, as with the usual late 60s assumptions of 2 stages both flyback proper Shuttles, putting a flight deck on the booster did not seem odd, but when they switched over toward vertical liftoff, then one wonders why someone did not suggest that automatic flight control combined with remote piloting to bring the thing to the ground?

From some things I've read recently, one factor in keeping crew on the Booster was the astronauts as a lobby--they were gung-ho to pilot spacecraft and insisted all flyback boosters be flown by crew on board.

Do I overestimate the capabilities of automation and remote control in these days and underestimate the importance of human being in the loop to the routine task of recovering from a vertical launch, aerobraking and turning to return to launch site? Then of course the landing is always a critical maneuver--but this is where remote flying by radio link ought to be most feasible too; the range is minimal, the direction the approaching craft is in is known.

By the 1980s, civil aircraft were already using auto-landing systems.

As to the economics of the whole thing--no, the idea is not to make the Booster cheap! The idea is to make it reasonably priced enough and maintainable that a given airframe launches many many rockets, and then part of the economy is spreading the admittedly large up-front cost over those many flights.

I believe there was some hope of lowering routine maintenance costs, but it would not surprise me if these are actually large for the fly-back booster.

I suppose the first generation Booster will require flight crew, in the sense that it will be designed to have them; there would be no mass savings to speak of by leaving them off. My concern is to avoid a loss of crew incident.

I hope that later iterations omit the crew cabin completely!
Next to that are other issue like F-1 engines covers made in time from Asbestos

Kevlar is substituted nowadays for asbestos. I suppose that by the time public concern about asbestos begins to grow (I remember it getting on to the national evening news around 1977 or so) substitutes would be developed.
 
Kevlar?
Oh that will be hell of Job, to get that burn stuff of F-1A Engine After each flight
Cover made from Mineral Fiber would be better
 
Of course subscribed!

Good to hear, and thanks to everyone for the positive reception so far!

My questions in ETS, posted just as this thread went up, are pretty well answered--it will not be a TL where a parallel burning SSME stack with separate return of the SSMEs makes an appearance, or anyway if it does it will be by a very different path. My notion also diverged ideally from established STS tech in wishing to drastically revise the boosters--ideally replace them with much smaller strap-on liquid boosters, so that various sizes of launch could be accomplished by simply adding on boosters around various sizes of tank, with various numbers of SSMEs clustered on the bottom. This would be how economy would be achieved, right-sizing the system for various loads and accomplishing the rare super-large launches with lots of clustered boosters and engines. But it would only make sense to do this if the STS were first developed in OTL form--someone visionary enough to conceive of the sort of launch system I was aiming at from the beginning, at the time of the "Shuttle Decision" or before, would probably not design key elements of STS.

In such a TL--or say, this one--for instance, there is really little reason to design any SSMEs at all. Five or six J-2 type engines do the job of 3 SSMEs quite well--except for the hitch that the J series engines would not be much good at sea level, due mainly to the gas generator pumping system. In principle gas generators can deliver power to turbo pumps at sea level just fine, but it seems the designers of the original J-2 took advantage of the knowledge that the engine would be lit only at high altitude to assume near-vacuum for the turbine exhaust, and thus could go with lower pressure gas generation which presumably is much easier to make work in a lightweight yet reliable fashion. But trying to run it at sea level means an anemic and probably unreliable GG, unable to deliver the power to pump the main chamber fully and thus further impeding the performance of a hydrogen engine which I gather is already inherently handicapped at sea level. I am unsure whether it is true that the higher the vacuum ISP of a rocket, the more severely in proportion its performance suffers at sea level, aside from other issues such as nozzle design, or if in fact the hit they take in ISP is higher in seconds just in proportion to the higher theoretical ISP. But in any case the original J-2 would be useless at sea level.

It may be that J-2S already addressed this issue robustly, and without increases in engine weight or complexity, could handle sea level burns--my impression is, it could handle them better than the original engine but still rather poorly compared to a ker-lox engine designed for sea level burning.

Yes, these are a lot of the issues we looked at when researching this TL. SSME is a decent first-stage engine, but for some reason proved difficult to air-start IOTL. J-2, conversely, is a decent upper-stage engine but next-to-useless at ground level. The various NASA attempts to account for that in the INT series (partially-fueled stages, among them) actually were part of the inspiration for "Propellant Ballasting." The difference of the two engines ties their development intimately into the chosen architecture, as you note--so it's a decision we'll be giving some attention in future posts...

Your suggestions are interesting, and would have been, IMO, one of the best ways forward for NASA in 2005. In 1970, however, much more is up in the air, as you note.

Thus, the part of the Shuttle Decision that resolved that the high-altitude upper stage sustainer engines should also burn at sea level and provide significant lift, I gather mainly because they wanted to be certain before lighting the SRBs that the main engines would all perform well, is what mandated the development of some engine that in some way or other would resemble the SSME. That is, apparently the extremely high 200 atmosphere chamber pressure of the SSME is mainly to enable gas to flow well against sea level pressure; this in turn required extremely sophisticated and complex pumping systems and ultimately in two stages of staged combustion and turbo pumping the propellant to two pressure levels.

But it seems clear from what we are told thus far that the ATL Boeing proposal assumes that the high-altitude engines are air-lit, in near vacuum. In this case there is no need to take such radical measures to enable firm thrust at sea level, instead the already developed, tested and improved J series would suffice pretty well. This in turn takes off much of the pressure to actually recover those engines--although developing, along with a reusable F engine, also a reusable J engine would be a possible choice to make.

Alternatively the SSME does deliver moderately higher vacuum ISP than the J-2S, although nothing ever seems to catch up to the moving target of the ISP of the ever-improving RL-10--which however suffers from a lower T/W ratio and having a really small thrust indeed. Along with, or instead of, making a reusable J engine is the option of making it more efficient with a raised ISP. And indeed the J-2S is notably improved over the J-2 in ISP.

Don't forget that the early-model RL-10 wasn't all that much better than J-2S in Isp. It only got to the extremes of Isp it has now in the 1990s.


A lot of these points will be addressed in coming posts, but you are generally on the right track. The pressure from OMB will have to be measured against the strong desire to retire the Apollo CSM and yet still have astronauts in orbit. As to how that comes out...stay tuned.

The idea of "fuel ballasting" is new to me and in a perverse way, pretty ingenious. As I said above my own idea for how to provide for a very wide range of payload goals with one economical system was to design fairly small strap-on boosters sized for the smallest loads and add on more and more of them as payload goes up.

"Propellant ballasting" is kind of an answer to the various multi-engine, different-propellant-loading schemes--after a certain point, standardizing on one vehicle design, even if you lift unburned fuel as ballast, is more economical than varying like that. Particularly if your first-stage vehicle is reusable. Kerosene, after all, is cheap compared to the manpower needed to reconfigure a vehicle, and extremely cheap compared to a vehicle at the bottom of the Atlantic...

Ballantine also helps reduce velocity at entry interface, enabling a much simpler TPS than typically needed for reentry.

I want to thank everyone again for the interest you've shown. We're planning on a regular Tuesday at 9:00 AM EST (1400 UTC) update schedule, so stay tuned for more of RS-IC Constitution.
 
Last edited:
But why ballast with kerosene instead of say, water? Obviously it would not be smart to just mix the water into the kerosene (won't mix without an emulsifier for one thing--"oil and water," non polar and polar molecules...) as this would just hurt engine performance and not even serve the purpose. But it would be possible to put a plastic bag in the kerosene tank and fill it with as much ballast mass as necessary, just water. This may still require making the fuel tank larger than ever necessary even for a max load launch, but less so that if one used kerosene, because water is denser. (There are kerosene rocket fuel formulations that are as dense or denser than water but I believe NASA has been using more normal density types, around 86 percent or so of water density, like typical diesel or jet fuel).

For that matter, because if I understand it the point of ballasting is to raise the total mass of the stack to maximum capacity of the booster, so that burn-out mass is the same as the maximum load it can launch, and therefore speed and altitude are also the same at staging as with such an upper stack load, the fuel tanks and oxygen tanks on the booster will always be loaded to maximum. Therefore if more mass is needed for ballast, one either has to make the fuel tank larger (or both the fuel and oxygen tanks--oxygen is less convenient to work with but denser, 1.15 times water density, and also despite its higher density the largest tank, so adding X cubic meters to it will cost less structural mass than doing so for the smaller kerosene tank--I also suspect that LOX costs about the same or possibly less by mass than refined rocket fuel would) or simply add an auxiliary ballast tank. The dry weight of the ballast tank will cost and require every launch to use more propellant than without it, but it seems a small price to pay for a simpler and safer ballast system.

In fact, it may not be necessary to install any tankage for ballast at all, if a suitable strap-on tank can be designed! Given that the Boeing Booster flies back as a glider (or possibly powered subsonic or even supersonic airplane, if I overlooked something about it having return turbine engines) and lands much as the Orbiter did, and therefore has a flat heavy TPS face (aka "the bottom of the aircraft") which still has hatches for landing gear, some clamps for a water tank or three could easily be added as well, perhaps cleverly using the landing gear hatches. Or they can be made of high-temperature metal perhaps. We know from OTL with the Orbiter's many accesses in the bottom high TPS (landing gear, oxygen and hydrogen inlets, brackets for the propellant tank, data connectors and I don't know what all else) that this is not problematic, especially because the Booster endures a much lighter dose of heat flux due to its much slower speed on burnout and staging, and therefore does not require the heroic measure of the ceramic tiles used on the OTL Orbiter--I'd think a layer of high-temperature stainless steel or one of the more exotic alloys would do--indeed, the post seems to indicate plain old aluminum might do. That surface would be far tougher against accidental damage of any kind--bird strikes for instance--and anyway there's nothing there to hurt it. Unless we put a ballast tank on it but that probably will not be any sort of threat!

Normally in rocketry we seek far and wide for heroically lightweight, yet strong and tough and temperature-tolerant exotic materials to make everything with. But in this case, the name of the game is to weigh down the booster, and it would be perfectly OK to make a water tank using heavy but reliable and cheap and easy to work with materials. To contain water we could easily get away with simple grades of aluminum, but if steel is cheaper despite being heavier, that's perfectly fine--we just deduct the excess weight of steel versus aluminum for a given volume from the total mass of water we wish to contain and make the tank smaller.

Being made of grades of metal (or if cost-effective at some date go over to composites or whatever--but cheap common steel grades would be pretty hard to beat for cost-effectiveness and easy management I'd think) that are chosen in part to be easy to work with, a facility, very possibly at the launch sites, could simply make cylindrical tanks with spherical caps (we don't care about air drag either, at least not if it doesn't create control problems, because air drag counts toward "ballasting" the load too) and walls cut out to whatever length we want for a precise ballasting, and just weld the thing--I suppose the seam would be where to install brackets to bolt it to the belly of the Booster. (Well, with some auxiliary bracing along side). There are no pumps, no pressure to maintain, no flows, monitoring the water is sort of optional--it's water, it won't boil mysteriously. We might worry about it freezing on a cold winter day, but some antifreeze is probably plenty in Florida or even Lompoc. These places get few snow days and the extremes aren't like in the north. It won't boil, even on a hot hot day.

As far as the Booster is concerned, once stage burnout is complete it can just drop the tank and reenter on an absolutely standard profile. We might want to wait until the combination is reentering and sufficient air drag exists to guarantee separation, and wait to drain out the water until they are well separated. But we could just leave the tank full and let it crash into the ocean whole and sink.

Or blow out the water, reseal it and declare the floating tank (if it survives impact) free salvage for anyone who wants it. To avoid charges of littering the Coast Guard will fish it up on a routine cutter patrol eventually if no one else claims it first, and NASA is budgeted to compensate that service for its time and trouble. Also NASA will purchase a retrieved tank for a nominal price. People can sell the tanks to space fans and museums, repurpose them, melt them down for scrap or whatever else they like--or sell it to NASA for a cheap fixed price. NASA might want to investigate the forensic fate of the tank, or melt it down for more tanks, or even use one for another launch.

At the other end of the spectrum, safety concerns might demand the tank be voided while still high up, and the material destroyed in mid-air to guarantee no hazardous debris ever strikes an innocent bystander in a boat. NASA never took that much trouble OTL for its STS tanks, which generally wound up shedding debris over some unpopular spot in the Indian Ocean, but to be fanatical I suppose the tanks can be made of an aluminum alloy with a layer of essentially a version of solid rocket propellant inside for a primer. Once the water is voided, redundant triggers ignite the interior material, which is less likely to be accidentally lit than real rocket stuff because its sole purpose is to heat up the tank alloy to the point it will burn in air, so the whole thing winds up so much dust. I don't foresee that actually being done but it is an option too.

With this option, the main Booster can be optimized for maximum payload launches without any built-in added mass for spare tankage volume requiring the basic load of propellants to be kicked up. Each mission simply gauges the deficit in total stack mass from maximum, orders a water tank of that mass to be attached to the belly of the Booster, and the only mass penalty there is versus a design with no ballasting options is the mass of the tank brackets. Even in the event of an abort after liftoff, the tank can simply be ejected at any point. A slightly more elaborate design would allow the tank to be partially drained while the Booster is burning--this might allow a boost to reach standard separation height and speed despite an engine going out, for instance, depending on exact circumstances.

I did wonder why not simply shut down or avoid usage of particular engines in the Booster, and only fuel it with partial propellant loads, to allow for reaching the same height and speed without ballasting and saving propellant too. As well as some wear and tear on some of the engines, but if left installed they would still suffer somewhat from thermal and acoustic battering by the other engines. Perhaps thermal heating is a problem--it is OK when all engines burn because each has kerosene circulating in the nozzle to regeneratively cool it, but an idle engine would not have its nozzle cooled? This might be addressed by an auxiliary pump to run coolant in idle engines though. We definitely need to leave the mass at the base of the rocket alone because the Booster, in aerobraking and flying back, must have its mass balanced over the aerodynamic center of mass, removing multi-ton F-1A engines would throw that off--unless we replaced the removed engines with ballast masses. Simply leaving the engines in place (or installing all of them in case as with STS the engines come out after every flight to be checked and refurbished) gives the option of emergency starting a designated idle engine, in case a designated burn engine fails for any reason.

When I tried to estimate the dry mass of the Booster (presumably greater by far than the Saturn V S-1C stage) and figure out how much mass should be stripped from the stack for one engine out, I found a drastic reduction in payload. Figuring the dry mass of the Booster to be say 300 tons (as opposed to 130 tons dry for the S-1C) meant to achieve the same overall orbit, we need to strip 170 tons from the upper stages for a nominal launch. I did not figure on upgrading to F-1A engines, but I suppose I should, so let me try this again. Also I'll be referring to the Skylab version of Saturn V for my baseline rather than the Apollo launchers since this should be more similar to the starting point in the ATL design process, meant for a Grumman Orbiter with external propellant tank but lifted by 5 J-2S engines.

OK, from this page (kudos to Norbert Brügge, of Germany--I use his data pages a whole lot and people should appreciate them) I can see that the Skylab launcher was all up 2857.15 metric tons. F-1A engines at 9189.6 kN thrust vacuum and 8003.8 kN sea level are consistently 18.25 percent more powerful than the original F-1 in both regimes and I presume in all between, so to account for the better engines we can raise the pad mass by 18 percent, to 3370 let us say. If we assume tankage and dry weights of both stages are scaled up in proportion, along with all other masses, payload rises to 120 tons--this is for a super 2 stage Saturn V. This may not be quite right since I don't believe the J-2S was 18 percent higher thrust than the J-2 (it was lighter though, and simpler and yet more capable).

Well, heck, Silverbird Calculator actually gives 163 tons to a 185 km circular orbit inclined at 29 degrees, from Cape Canaveral! The improved ISP of the J-2S accounts for a 43 ton bonus, I suppose.

All right. Now I have to make a wild guess as to the dry mass of the Boeing Booster. I presume its fuel tankage remains just the same, but any mass that I add to the dry weight clearly must come off the upper stack, which is payload and second stage. I guessed 300 tons and I will stick to that for now though it more than doubles the S-1C mass, even kicked up 18 percent. After all we are asking a lot of it. Indeed the engines might not be F-1A at all because we want them to be reusable. But it seems reasonable that their thrust will be a firmly held design point, and an "F-R" to suggest a name (or B-1, for "Booster mark 1) would probably be heavier to make it more durable, while for this exercise I assume the same thrust and ISP. It will have a winged planform--one can put a lot of kerosene fuel in reasonably thick wings to be sure. But it was also suggested Boeing do this job because of their expertise in supersonic big plane design, and that suggests it would aerobrake only down to moderate supersonic speed and glide/fly under thrust at supersonic speed back to the launch site. Well, that would probably be overambitious IMHO; one loses so much in trying to make an airplane cruise at supersonic speed versus high subsonic, which Boeing is also expert at though less uniquely so--and the rivals who also bid for the SST contract might argue that Boeing really wasn't superior to them in that sphere either--North American in particular would have a strong case for their work on the B-70 Valkyrie bomber, and Convair delivered an operational supersonic-dash bomber the B-58 Hustler--by the 70s these would be known as Rockwell and General Dynamics respectively of course. Lockheed also is a strong claimant, especially with the SR-71 design under their belt. So I think supersonic cruise is ill-advised even if we do want Boeing to get the contract--and indeed, Boeing's credentials for big subsonic aircraft are undeniable. I for one feel more confident with them doing the work! Lockheed and McDonnell-Douglas (replacing NA/Rockwell and Convair/GD because the former never made any big transport planes and the latter's did not do so well in the market although I suspect this may have been just bad luck, they seemed to be good planes) would be the strongest competitors, but only the former has experience with cutting edge supersonics. Even a subsonic cruising/gliding design will transition through a supersonic, indeed hypersonic, flight regime though I believe this will involve essentially a flat belly-flop on the TPS bottom of the plane--still experience with those regimes as they guide material choices is good even if the airframe is not designed for supersonic flight at all.

300 tons is bloody huge of course; it means a dry Booster masses almost as much as a fully loaded 747 of this era! That is odd, but after all we are going to be holding some 2500 tons of propellant so it is not crazy out of line. The proportionately scaled up S-1C+ would have dry mass of 155 tons so we need to shave off some 145 tons from the payload and second stage, which are 163 tons and 603 tons respectively. I see no reason not to do that proportionally--drop 31 tons from the payload and 114 from the second stage. The latter in turn is a case of cutting propellant and dry structure, but note that we probably won't reduce the engine count although we might, so really we should look at propellant and dry mass minus the mass of 5 engines that we cannot alter. Propellant was 560, and dry mass less the mass of 5 J-2S engines is 36, so in proportion we cut them by 107 and 7 tons respectively.

Thus the upper stage has a 36 ton dry mass all up and holds 453 tons of propellant--actually a bit smaller than the OTL standard S-2. But to use the standard would mean reaching booster burnout at a lower speed...but I say that like it was a bad thing! It's a good thing, it makes the reentry of the Booster easier, so let's save some money by just swapping in a bog-standard S-2 as designed for Skylab, with J-2S engines swapped in. Those engines, note, were specifically designed to fit exactly where old J-2 engines would fit, with their various fastenings and pipe fittings placed exactly where the J-2 required them. So swapping in the more advanced hydrogen engine is quite easy and cheap to do. The upshot would be 21 tons heavier than the optimized shrink-down I just did, but on the scale of a 3000 ton rocket I don't think that will overload the booster engines!

Silverbird gives 138.5 tons to LEO, which is 6.5 better than I guessed above. To be sure SB Calculator might be failing to take into account added air drag during boost due to the Booster stage's wings. So let's take it with a grain of salt. If Grumman had the budget and would follow my advice, their Orbiter design would be essentially the Skylab S-2 with a 120-135 ton spaceplane on top, more similar to Buran than the STS Orbiter because it would have no main engines installed, only orbital maneuvering system and possibly some turbojet flyback engines. Otherwise it is in the same ballpark as OTL Orbiter, but we must dispose of 5 J-2S engines with every launch. Also of course it might have been designed to resist the Space Winnebago syndrome of OTL Orbiter, with more of its total mass being actual cargo and less being infrastructure for keeping 8 human crew happy and functional in a mini-space station for weeks--such capabilities might be an option achieved by a module loaded into the bigger cargo bay. Then again, I don't know the cargo bay can take more mass without messing up the center of mass distribution on reentry. Just bear also in mind that again I can't guess how much the wings on the Booster will impede launch performance.

Obviously if the dry weight of the Booster is substantially below 300 tons, we can improve the size of the payload more, while if it is heavier the payload obviously suffers. I don't see any really strong arguments for either variation, barring careful detail work. Cutting it below 150 tons is absurd in the light of the fact that the simple expendable version is already that big, raising it to 500 or 600 tons seems pretty alarmist. I'll defer to the experts but for now, stick to 300 tons.

Now then--what happens if we simply let one of the Booster's 5 F-R (equivalent to F-1A in thrust if not dry weight) engines take a little vacation in the next launch, and launch firing just 4 of them instead? That means we are consistently deprived of 20 percent of all thrust during boost. If we wish for the boost phase to stick close to the standard launch profile, we must omit to load 20 percent of the Booster's propellant, since that will govern burn time on the remaining 4 engines to match, and we must delete 20 percent of overall stack weight.

The 5-engine launch version is 3448.5 tons on the pad, so now we must delete 689.7 of those to bring the boost within the capability of the 4 active engines. Note air drag remains just the same so actually we probably need to shave more weight but I don't know how to estimate how much so let that ride. 1/5 of the standard full propellant load is 500 tons, so we need to knock off another 190 from the upper stack, which masses 648.5 tons on the main version. We bring down both payload and second stage by some 30 percent, so the latter is 360 tons all up. This time we scale back the engines too. 70 percent of 5 is 3.5 which is awkward because we must choose either to go down to 3 engines or just down to 4. Well, let's look at both. We have the same non-engine dry mass either way, 20 tons, and 332 tons of propellant, but either 4.2 or 5.6 tons of engines. With 4 J-2S Silverbird suggests a payload of 99 tons, and with 3 upper stage engines, 96.2--I guess it comes down to the question of whether the 5.8 extra tons are worth expending an extra J-2S engine or not.

Now tonight I am getting much less dismal results than my rather slapdash methods yesterday implied, so I will forge on with the question of simply not using all the F-R engines. Let's look at 3 engines in use only. We clearly need to lower the masses in the same places, that is Booster propellant loads and upper stage all up once again. Again we eliminate another 500 tons of booster propellant and must remove 190 from a 357.6 ton upper stage and 99 tons of payload. 149, call it 150, must come from the second stage, which reduces it to 207.6. If the dry mass is 14.6 tons then we have 193 tons of propellant and just 2 J-2S engines left. SBC suggests a payload of 56.7 tons!

Can we keep going? Let's try it. Down to 2 Booster engines, again lower the Booster propellant by 500 and shave off 190 tons from the upper stack. Note that now the upper stage is almost wiped out, since we must delete 147 tons from it and it only massed 207.6 all up before! We are clearly going to go so low we might want to get rid of the J-2S engine completely and go with arrays of RL-10s, but that's drastic since the program must now include a third engine type--well they probably are anyway for deep space applications. And we must accept a lower thrust/weight ratio; RL-10 is not really designed for putting stuff into low Earth orbit after all. Let's see if we can keep a single J-2S first. We have 54.5 tons of propellant, and a dry mass of 4.5 tons for 59 all up. SBC says we can put up almost 17 tons!

I think that clearly we cannot go down to just one F-R engine because the second stage mass would become negative. We've reached the limit of going down by simply neglecting to use booster engines, and still maintain the same ascent profile for the booster. Yet here we are, with a large 17 ton payload, at the very bottom of a range that outperforms OTL Saturn V at the top. The baseline booster is fantastic for a TL where huge amounts of mass must be launched annually, say an ASB TL where there is a race to get control of alien technology or some such. Or just one where politicians are space nuts and quite happy to fund space exploration on a scale say 10 times as great as OTL. But it is clearly not "rightsized" for a TL operating on a budget similar to OTL's!

I would suggest going down to designing around 3/5 what I had above on top, using 3 F-R engines for the baseline maximum mission.

In that case, we'd start out with 80.8 tons payload with all three F-R engines firing. Eliminating one of the three booster engines as always costs us 500 tons booster propellant and 190 off the upper stack, for 42.5 tons to orbit. Can we go down to just one F-R? I would say, no we can't, we'd only have 6 tons left for the whole upper stack! To go lower than 42 tons then, we'd have to do something like ballasting.

If I cut the upper stage mass in half, that is 77 tons--but with only 2 engines burning I have reduced the fuel load in the booster by over 150 tons, so there is plenty of room there to put 77 tons more kerosene that will not be used in flight as fuel, thus holding the whole stack at the standard burnout mass (for 2 engines). This allows some 22 tons to reach orbit. Since we have nearly enough margin in the booster fuel tank alone, I suppose it is fair to say here we have a booster that can put any size payload below that into that some orbit, because we have enough room there to just about eliminate all upper stack mass!
--------
So in this case, it would not be necessary to attach an extra ballast water tank, because the volume does exist in the fuel tank which is only 2/3 full versus its fully fueled state. But we would want to use kerosene, not water. We don't have to make any modifications at all if we do that, no water bag that might break, and dumping the excess 77 tons or more of ballast kerosene is easier than devising a means of dumping water from a bag within the fuel tank.

We could do the same with the 5 engine, 300 ton dry super-booster, for payloads less than 17 tons.

But in either case--note how tremendously wasteful we are being compared to say the Saturn Multibody system developed in ETS. Practice since the Shuttle Decision suggests that we only rarely want payloads much larger than 20 tons, which is the minimum we can get with the big 300 ton 5 engine job (running on only 2 of them) or with 2 of 3 engines in the 180 ton dry smaller version, if we ballast with nearly 80 tons of extra kerosene. The latter is somewhat more efficient overall for the 22 ton load, being 180 tons dry, loaded with 1077 tons of propellant, putting up a 22 ton payload (normal sized OTL) using a 77 ton upper stack, 1357 tons all up. To put up 17 tons the big booster is 1376 tons on the stack, with none being ballast. With either version we can go lower on payload but the total stack mass will remain the same all the way down to zero payload. Both of these mass essentially the same--and nearly twice the mass of an ETS Multibody single core launch. Most of the extra mass is propellant and excess first stage dry mass.

I am not sure which version to favor, seeing that the big 5 engine installed version is not more wasteful at low payload masses than the apparently more svelte 3 engine version. With either we mainly buy the ability to occasionally launch really gigantic payloads--up to 80 tons with one, up to nearly 140 with the other--at the cost of normal launches involving tremendously more mass than they have to.

With the 3 engine version, it occurs to me that something like the ETS Multibody core unit could be developed, essentially 500 tons of propellant in say 60 tons of stage, and attached to the belly of the Booster by the same means I proposed to attach water ballast tanks. The Booster might also be designed to take two balanced some distance out on the wing, or conceivably in the "elbow" where the wing meets the upper body of the main Booster. This would allow up to three to be attached, doubling the capability of the standard Booster--at the cost that the three LRBs and their F-R engines would not be recoverable. But how often do we want to put up 160 tons or more at one shot? It may be more sensible than designing the 5 engine version with its ability to put up nearly 140 tons without tinkering with attached boosters. It turns out not to waste fuel versus the smaller one on this scale, and reaches high performance with no kludging around with external boosters.

But that 300 ton dry weight really is daunting! A 180 ton system with fewer upper range options (unless we develop the F-1A based LRBs to attach to it), it makes me wonder what we'd get if we downgraded further and had a two-engine, even a one-engine version? But if we usually use the 3 engine version with one shut down for most business, we have a backup engine to substitute in for either. We can't do that with one engine systems, or even two; no redundancy.

We can't go much lower using F-1A engines. But switching over to say H-1 (upgraded for better ISP perhaps) is a huge jump down too, resulting in maximum payloads not much over 10 tons, which is too low. And we would want the big payloads occasionally. The question is I guess, do we really need to double the pad mass for routine 20 ton launches in order to have a reusable first stage with the ability to every now and then put up 7 times the standard payload?

At any rate 180 tons is a less insane mass for Boeing to design to fly back; at 300 tons dry they are asking a Boeing 747 to be a glider!
 
A particularly revolutionary innovation in this study was the concept of “propellant ballasting.” By carrying more propellant than strictly necessary for lower-end payloads, and burning it off in a second post-staging burn of the first stage, reentry velocity could be reduced considerably for smaller payloads (like those needed to service a space station), extending stage life.

I think he's talking about using it as fuel, so Kerosene and oxidizer is required, not literal ballast. They use it to slow the 1st stage down after separation to reduce re-entry speeds. At least as I understand it from the chapter posted. True it's 'ballast' on launch because it's basically payload until separation, then it gets used.
 
Top