Index Home About Blog
From: (Henry Spencer)
Subject: Re: Why does the shuttle roll?
Date: Wed, 29 Nov 1995 16:59:25 GMT

In article <497q52$> (Brian S. Thorn) writes:
>>Second of all, SSTO advocates who advocate powered flight over
>>populated areas generally assume an extensive flight test program
>>first.  I haven't seen any detailed plans, but if operations are truly
>>aircraft like, hundreds of test flights are not out of the question.
> Not out of the question, perhaps, but does anyone truly believe that
> Rockwell, Lock-Mart, or McDonnell will fly their SSTOs hundreds of times
> before committing it to operational service? ...

It's the responsibility of the regulatory authorities to insist that this
is done right.  Rockwell, LM, and McDD *do* fly their aircraft hundreds
of times before committing them to operational service.  This is not a
radical and unprecedented way of doing things for them; they just have to
be told that aircraft standards, not missile standards, apply.

Note that they do even have some financial incentive to do this.  With a
reusable vehicle, reliability becomes an important issue, because you
lose more in a failure -- you lose a vehicle you didn't expect to have
to replace.  This encourages an incremental test program (assuming that
your vehicle permits it) to find problems before they become fatal.

> So the question remains, if the SSTO flies a test program of about
> 100 flights, just how will it differ from Shuttle, which by then
> should have 100-125 flights under its belt?

The shuttle test program lasted four flights and made no great attempt to
explore the flight envelope thoroughly, as any aircraft test program must,
and as one would hope an SSTO test program would.  As a particular
example, operation in sub-zero temperatures is a routine test case for
high-performance aircraft.  As another example, the major abort options
of the shuttle have never been flown, while demonstrating them is an
important part of an aircraft's test program.  A hundred repetitions of
the same near-nominal flight plan doth not a test program make.
Look, look, see Windows 95.  Buy, lemmings, buy!   |       Henry Spencer
Pay no attention to that cliff ahead...            |

From: (Henry Spencer)
Subject: Re: Low Tech Space plane
Date: Thu, 14 Dec 1995 14:06:30 GMT

In article <> (Alan Anderson) writes:
>>Actually, there is very little in a traditional SSTO that is for use
>>only in the atmosphere.  The gains from high-altitude launch come in
>>improved performance, not elimination of equipment, by and large.
>I can think of only one thing that can be "left out" because of a
>high-altitude launch:  short-bell rocket nozzles.  Since they would
>probably be replaced with *longer* bells, this doesn't quite count...

Actually, this is precisely why I waffled with "by and large":  the nozzle
hardware might get a bit simpler, by elimination of complications like
telescoping nozzles.  (Of course, if you're sensible :-) and just use an
aerospike, this isn't an issue...)
Look, look, see Windows 95.  Buy, lemmings, buy!   |       Henry Spencer
Pay no attention to that cliff ahead...            |

From: Henry Spencer <>
Subject: Re: SSTO alternatives (was Re: MSFC engine)
Date: Fri, 29 Dec 1995 03:20:43 GMT

In article <4bpjcv$> (Doug Hendrix) writes:
>>Many of the people who have investigated [BDB] say that it does have
>>the potential for that order of magnitude...
>If we can PROVE an order of magnitude cost reduction right now, we'd be smart 
>to put SSTO on back burner and go do BDB. I don't think we can prove it.

The only way you *prove* it is to build it.  Nothing less will convince
the die-hard naysayers, who think it is a law of nature that launches cost
$5000/lb and always will.  Nobody's yet put up the money for that.

But even if you are convinced that BDB will deliver the order of magnitude,
it is *not* smart to put SSTO on the back burner.  SSTO offers potential 
for more than just one measly order of magnitude.

>>>I don't know much about reoxing but it sounds 
>>>preliminary and fraught with problems that'll drive costs sky high.
>>Why?  The USAF does flight refuelling every day...
>I read about a jillion articles on this a couple of months ago and don't wish 
>to reopen the subject. 

In other words, you aren't willing to defend your position.  I take this
as a concession of defeat. :-)

>>>Near SSTO stuff will be light weight payload delivery.
>>Why?  The JATO-assist schemes take the full vehicle into orbit.  The
>>kick-stage schemes don't, but there's no reason why the payload has
>>to be small.
>I assumed near SSTO is a weight limited vehicle (or else you'd go on to 

It's a mass-fraction limited vehicle.  Mass fraction is a ratio, not an
absolute quantity.  It's quite possible to have very large near-SSTOs.

And you're still ignoring the fact that some near-SSTO schemes supply
the little extra kick at the beginning, not the end, so the vehicle drops
some strap-on boosters and then ascends into orbit.
Look, look, see Windows 95.  Buy, lemmings, buy!   |       Henry Spencer
Pay no attention to that cliff ahead...            |

From: Henry Spencer <>
Subject: Re: tourism (was Re: After Station? (Outreach II))
Date: Wed, 3 Jan 1996 15:11:07 GMT

In article <4cbjjr$> (Undamped) writes:
>>Safe is not a big problem if you design and build your spacecraft as
>>high-performance airliners rather than manned missiles.
>Aren't you seriously understating the difficulty of doing this?  At
>take-off, an airliner is typically 50% dry mass (structure, engines,
>controls, etc..), 38% fuel, and 12% "pay"load.  In comparison,
>hypothetical RLVs are usually quoted as 90% fuel, 9% dry mass,
>and 1% payload.  The airliner essentially must lift its own empty weight,
>in comparison, the rocket must lift 11 times its own weight...

This is quite true.  On the other hand, the thrust/weight ratio of the
very best aircraft engines is under 10:1, while rocket engines reached
125:1 thirty years ago.  So yes, the task (expressed this way) is 11
times harder, but the technology is 12+ times better, and has been for
quite a while.

(This argument does disregard a variety of complications.  For example,
the 125:1 engines are not the ones currently proposed for RLVs, because
other aspects of their performance are less impressive... but on the
other hand, the 10:1 engines are not the ones used in airliners either.)

>There is simply no current way you can design a spacecraft like a
>high-performance airliner and have it come anywhere close to acheiving

You're reading me too literally.  I was addressing design philosophy, not
detailed design approach.  The safety problems of our current spacecraft
designs are problems of philosophy, notably the "failures are intolerable,
we will spend whatever it takes to make sure they never happen" approach,
which not only is very expensive but doesn't work.  The corresponding
airliner design philosophy is "failures are inevitable, so the aircraft
and payload must survive all likely failures and most unlikely ones".
Look, look, see Windows 95.  Buy, lemmings, buy!   |       Henry Spencer
Pay no attention to that cliff ahead...            |

From: Henry Spencer <>
Subject: Re: Space Shuttle SSTO?
Date: Mon, 29 Jan 1996 07:07:33 GMT

In article <> Ruediger Klaehn <> writes:
>Is there any chance of converting a Space Shuttle into a SSTO or perhaps
>TSTO (first stage Boeing 747 or Antonov 124) with horizontal starting
>and landing? ...


Doing SSTO with horizontal takeoff and landing is virtually impossible
without some kind of cheating.  The landing gear -- well, more precisely,
the takeoff gear -- is too heavy.  Boeing spent years, and a lot of money,
studying the possibility, and concluded that there was just no way to do
it without some sort of fudging like sled launch.  (They *did* conclude
that if you were willing to use sled launch, it could be done.)  Moreover,
starting with the shuttle orbiter wouldn't be much better than starting
from scratch, and might even be worse -- the current orbiter design is
totally unsuited to this application. 

As for air launch, the shuttle orbiter is much too heavy for air launch by
any existing aircraft (bearing in mind that it needs a large external tank
accompanying it, to supply fuel).  This approach is workable, but only for
considerably smaller vehicles. 

>With LH2 / LO2 Fuel you get an Exhaust velocity of about
>4000 m/s and therefore a weight ratio of 7.3890 to 1 for 8000 m/s
>orbital speed...

Actually, good LH2/LOX engines get 4400 m/s or thereabouts, although you
have to specify whether you are talking about sea-level or vacuum
performance.  (A rocket SSTO does most of its accelerating in vacuum, but
the time spent within the atmosphere is not negligible.)  However, you need
something like 9100 m/s total to reach orbit, because you lose some to air
drag and gravity losses.  Wings help with gravity losses but add drag and
structural mass.

>One could use a nuclear engine. These have an exhaust velocity of about
>8500 m/s with LH2 and therefore would require a weight ratio of 2.553.

Unfortunately, there are a few problems here.  There is some release of
fission products in the exhaust, which is considered unacceptable for
operation within the atmosphere nowadays.  The tested nuclear-engine
designs have relatively low thrusts, and clustering them is tricky (the
reactors interact via neutrons).  And to cap it off, the mass you've
gotten rid of is LOX, which is cheap, compact, and easy to deal with.
You actually need *more* LH2, which is horrendously bulky and makes it
relatively difficult to achieve good mass ratios.

>...Equipping a shuttle with nuclear engines would probably
>be cheaper than developing a completely new SSTO based on LH2/LO2,
>because most of the research has already been done in the 60s (Rover).

The research needed for LH2/LOX SSTOs has already been done.  In 1982,
Boeing quoted the USAF a firm fixed price of $1.4 billion to develop a
sled-launched HTHL SSTO.  Boeing does not quote fixed prices on advanced
development projects very often; they were awful damn confident that they
could do it. 

I believe the SEI people were estimating more than that -- admittedly
in somewhat later dollars -- just to get operational nuclear engines,
never mind vehicles using them.  It's harder than it looks; for example,
it is no longer considered acceptable to just vent the exhaust from test
engines to the atmosphere, and a scrubber system capable of handling the
exhaust from a large rocket engine is a major project in itself.
The Earth is our mother.                           |       Henry Spencer
Our nine months are up...                          |

From: Henry Spencer <>
Subject: Re: A New US Space Development Policy
Date: Fri, 16 Feb 1996 16:22:14 GMT

In article <Pine.SOL.3.91.960215134818.3369B-100000@seds> Mike Dicenso <mdicenso@SEDS.LPL.Arizona.EDU> writes:
>> Reliability rates of well over 99% are not fantasy. They are *required* in 
>> the airline industry.
>Maybe, but where do we meet the diminishing returns for chemical rockets? 
>Reliability over 99% is achivable, but 9.9999999...% In the extreme 
>enviroment that an RLV faces, is this realistic in the short term?

What "extreme environment"?  The only thing that is even particularly
unusual about the RLV operating environment is reentry, and the durability
of the heatshield -- how much inspection and refurbishing it needs between
uses -- is the one thing that I would class as a major technological worry
for RLV builders today.  In all other areas -- yes, including engines --
extremely high reliability over a long life is well within the state of
the art today, if you only *try*.  Yes, it does require some sacrifices,
in areas that have traditionally been sacred cows for launcher designers,
but it's not impossible.

The key thing to remember is that airliners do not achieve their extremely
high reliability by having ultra-reliable components.  They do it by
having backup plans, including intact abort after all but the most severe
and unlikely failures.  An airliner engine failure doesn't even make the
news unless it's of a very unusual sort, because the airliner lands
safely, nobody dies, and no large amount of money is lost.  The reason why
failures of liquid-fuel rocket engines usually total the vehicles is not
that the engine failures themselves are any more catastrophic, but that
the vehicles have not been designed to survive failures. 

On the rare occasions when a rocket has been designed to cope with
failures, it has worked.  The sixth Saturn I lost an engine in mid-flight
and went on to complete its mission, as did Apollo 13's Saturn V.  DC-X
had its whole side bashed in by a launch explosion, landed safely, and was
repaired to fly again.

We know how to do it.  We just haven't been trying.

>> Do you have any idea how many bases the US built during World War II using 
>> supplies carried in exclusively by DC-3s?  A lot of them had buildings that 
>> were larger than the DC-3 itself...
>Would that everything be done that way, why need heavylift aircraft like 
>the 747, C-5, and the An-124?

The answer is "very rarely".  Almost all air cargo goes in standard-size
containers that will fit in most modern airliners, including quite small
ones.  The 747 was built to carry lots of small cargos (passengers), not
to carry single big items.  The C-5 and An-124, as well as the Shorts
Belfast, are specialist military aircraft.  There is a small commercial
market for oversize cargo transportation, so small that no major airline
has ever bothered to get involved and no civilian aircraft production run
has ever been done for it.  Last I heard, half a dozen ex-military
Belfasts were serving most of it, supplemented by an occasional An-124
charter.  Compare that to the enormous amounts of standard-size air cargo
moved every day. 

>Getting the task done, using an HLV,  
>speeds the construction up a bit. That ain't so bad either. Instead of 
>700 flights in 14+ years, why not reduce it to 70 in 7-10 years?  Let the 
>747/C-5 type boosters do their job and leave the rest to the 737, and 
>DC-9 type RLVs?

None of the existing launchers bear any resemblance to a 747 or even a
C-5.  700 flights of an RLV will be much cheaper than throwing away 70 
Space will not be opened by always                 |       Henry Spencer
leaving it to another generation.   --Bill Gaubatz |

From: Henry Spencer <>
Subject: Re: HTHL vs.VTVL ( Why a DC-X w/o wings??)
Date: Sun, 9 Jun 1996 23:21:30 GMT

In article <4p432d$> (Marcus Lindroos INF) writes:
>...Lockheed's design will require a linear aerospike engine (forgot
>to mention this although it is a VTHL too), the Delta Clipper will need
>eight new, ultra reliable engines...

Why "ultra reliable"?  Eight is enough for engine-out tolerance.  As for
the linear aerospike, that's a new nozzle type rather than new engine
hardware; while there are lingering uncertainties about it, most of those
should be resolved within the next few years.

>...a VTVL would need highly economical AND
>reliable engines with advanced restart/deep throttling capabilities...

Gosh, yes, that's hard new technology, like the RL10, which is only 35
years old now and can do all those things.

*All* SSTOs need economical and reliable engines.  The engine redundancy
at VTVL landing is so extreme that the need to get a couple of them 
restarted really doesn't do much to reliability requirements.

The ability to do a restart at all takes some attention, particularly for
hydrogen engines, which have chilldown requirements.  However, it's not
something that's deep and mysterious and fundamentally hard, it just takes
attention during design.  Deep throttling is a bit more of a challenge,
especially for the hotter engine cycles, but it too has been done.

>Perhaps X-33 will demonstrate that VTVL is economical and more reliable
>than VTHL. But the database of succcessful unpowered glide landings is
>far, far bigger than that of rocket-powered vertical landers, so I'd say
>VTVL has more to prove here.

Indeed so... but as has been pointed out in a number of other connections,
you have to separate things that are hard from things that are merely
unfamiliar.  The database of glide landings of high-performance aircraft
is *not* all that large, and it includes some crashes.  The database of
powered vertical landings is small only if you arbitrarily exclude vehicles
with air-breathing engines from it.

>Uh, nose-first VTVLs have to transfer from vertical to horizontal and 
>finally flip-flip back to vertical again shortly before landing...

Nose-first VTVLs have to do only one transition, the flip before landing.
(Although that one transition is part of why nose-first is a silly way to 
build a VTVL.)  Going from vertical to horizontal is part of the ascent 
for any SSTO.
If we feared danger, mankind would never           |       Henry Spencer
go to space.                  --Ellison S. Onizuka |

From: Henry Spencer <>
Subject: Re: SSTO vs nSTO
Date: Mon, 10 Jun 1996 03:28:39 GMT

In article <> writes:
>Current engineering opinion seems to be against you on ELV-SSTO vs.
>ELV-nSTO.  It has been clear for a number of decades that with Saturn
>level engineering one could make a throw away SSTO.  For decades
>however, no designers of new launch systems have chosen this route...

Now, count the number of all-new launch systems which have been designed
from scratch (without being constrained to use existing stages or solid
motors or etc.) in those decades.  There aren't very many.

The shuttle was constrained not to be a throwaway (and there was one serious
proposal to build it as an SSTO, by the way).

Ariane was constrained to use storable propellants in its lower stages,
and was generally constrained to be a low-risk design.

Zenit was explicitly required to maximize the payload available within a
given launcher size -- the Russians have said this, in print, although
they haven't said *why* -- and that inevitably gives you a multistage

Energia's constraints are poorly understood, but large payloads on a
vehicle that was already pushing practical size limits probably had
something to do with it.

Neither Zenit nor Energia had any cost constraints.

>Of course, this is merely the engineering judgement of the
>"establishment" - it could be wrong...

No, it's primarily the political judgement of the establishment's bosses.
The engineers don't set these constraints.
If we feared danger, mankind would never           |       Henry Spencer
go to space.                  --Ellison S. Onizuka |

From: Henry Spencer <>
Subject: Re: HTHL vs.VTVL ( Why a DC-X w/o wings??)
Date: Mon, 10 Jun 1996 02:57:23 GMT

In article <4ommge$> (Marcus Lindroos INF) writes:
>Not *now*, but maybe they _should_ be just like aircraft... I.e. take off and
>land horizontally. No stupid launch pads. Full compatibility with existing
>airport infrastructure. 

As others have pointed out, concepts like Skylon are *not* compatible with
existing airport infrastructure, because they need strengthened and
lengthened runways, different fuel supplies, probably a major relaxation
of noise restrictions, and certainly major changes to air-traffic control
procedures and flight regulations.

Operating a well-designed VTVL out of an airport should be just about as
easy, perhaps easier.  Some special facilities, but no need to upgrade
the runways.  Less of a noise problem due to vertical ascent and descent
(the noise issue for HTHL vehicles occurs over the neighboring areas, not 
over the airport itself).  Similar problems with fuel and procedures.

>However, unlike a VTHL the HTHL uses its wings during ascent as well...
>The result is a far more pleasant airplane like 'lifting trajectory'
>which would be preferable to human passengers on board...

I don't know, I think I prefer a nice smooth elevator ride over bumping
around through the clouds horizontally.  Besides, any difference here is
for the first few minutes only -- either way, the bulk of the acceleration
is done at extreme altitude in vacuum.

>The Delta Clipper,
>which would enter the atmosphere nose-first and then 'flip-flop' shortly
>before landing is about as unsuitable as it could be for hypersonic
>passenger transport of humans...

I'm inclined to agree -- base-first VTVL is considerably superior. 
Elevator ride up; elevator ride down. 

>Fuel economy? Does not appear to be an issue. SKYLON does require more
>expensive hydrogen fuel than the Delta Clipper, but the latter requires
>_huge_ quantities of (cheap-) oxygen which drives up the total cost.

Uh, do the numbers, Marcus.  Less hydrogen plus a lot of oxygen is less
expensive, not more.  It takes an awful lot of LOX to make up for needing
more LH2 -- the difference in cost per kilo is very large. 
If we feared danger, mankind would never           |       Henry Spencer
go to space.                  --Ellison S. Onizuka |

From: Henry Spencer <>
Subject: Re: SSTO vs nSTO
Date: Mon, 10 Jun 1996 03:14:11 GMT

In article <4ovumu$g74$> "Michael P. Walsh" <71544.2040@CompuServe.COM> writes:
>Whether an ELV-SSTO would be cheaper than an ELV-nSTO is a 
>debatable point when you measure cost by payload to orbit...

Ah, but remember the key point the BDB crowd make:  enlarging the vehicle
(not adding more pieces to an existing vehicle, but making the pieces of a
new design bigger) costs almost nothing.  If you don't like the size of
the payload, just make the launcher bigger.  Costs scale primarily with
parts count, thinness of margins, and nearness to the leading edge of
technology, but only very weakly with size.

The BDB crowd end up invoking this rule because low-tech construction
gives them heavy structure, while an ESSTO runs into it because it needs a
high mass ratio... but the rule applies regardless. 
If we feared danger, mankind would never           |       Henry Spencer
go to space.                  --Ellison S. Onizuka |

From: Henry Spencer <>
Subject: Re: aero-space weight margins.
Date: Tue, 11 Jun 1996 16:17:10 GMT

In article <> M R Atkinson <> writes:
>...We are talking at this stage about how good the weigth estimates
>are. If the 777 was 2% then for a SSTO with much less well known
>technology we will do well to be within 5%, wether it will be 5%
>under or 5% over is anyones guess.

A 5% mass margin -- that is, enough slack in the design so that a 5%
overshoot in dry mass causes *no* payload loss -- is well within the
real of possibility today.  The big problem with the SSTO designs of
the past was that they tended to have rather limited mass margins,
and this correctly made people nervous.  Modern structural materials
make a big difference.

>IIRC the C-17, F-22 and A340 have all suffered weight problems, as
>well as various mid-life upgrades, unmanned vehicles, etc...

Of course, then we have certain other examples, like the A-3 Skywarrior
and A-4 Skyhawk, of how firm management control by competent people can
*eliminate* weight growth (even on programs which "conventional wisdom"
says are so weight-constrained that their weight goals are impossible). 
Without blowing either the schedule or the budget, I might add. 

>...aircraft have the option of reducing the
>operational range, SSTO does not have that luxury, we have to make
>orbit so payload inevitably suffers.

Getting 95% of the way across the Atlantic doesn't work very well either.
New-airliner contracts typically make quite specific promises about range.
If we feared danger, mankind would never           |       Henry Spencer
go to space.                  --Ellison S. Onizuka |

From: Henry Spencer <>
Subject: Re: Illogical Conclusions (SSTO vs.TSTO)
Date: Tue, 11 Jun 1996 16:53:47 GMT

In article <4ouh8s$> (Marcus Lindroos INF) writes:
>If TSTO always is more expensive, then please explain why Interim HOTOL 
>costs less to develop than than a single-stage vehicle in the same payload
>range... (7-9t to LEO).

Because it's an SSTO with the unusual engines replaced by an unusual
launch pad. :-)  Don't play word games, Marcus:  when people say "TSTO",
they mean something that stages at hypersonic speed, not an SSTO which
cheats slightly by using air launch.  Both HOTOLs are SSTOs, and Interim
HOTOL uses less aggressive technology, so of course it's cheaper.

Note that development cost estimates for LART (essentially a German HOTOL)
were approximately half those for Saenger (Germany's favorite TSTO).

>: Why are you so hung up 19 metric tons?  Are you planning to launch 
>: elephants or something?
>If you'd care to actually study the X-33 proposals, you would notice that
>both Rockwell and Lockheed think 19-20t to LEO is essential...

No, they think it's the best way to attract US government funding.  There's
a difference.  Observe that McDD doesn't agree with them.

>Wales Larrison just said satellites keep getting heavier and heavier. If
>you have to spend $10 billion, you want as many customers as possible
>to keep the vehicle flying.

No, you want as many customers as possible *without* warping the design so
badly to accommodate the oddballs that it becomes ill-suited to the
majority.  There is only one customer who needs 19-20t payloads, and he
doesn't buy very many of them.  A rational RLV supplier aiming at
commercial operation will tell him to go fly on Titan IV, and will size
the RLV cargo bay to meet the bulk of demand rather than the unusual

It's true that the satellites are getting bigger, but they aren't that
big, not yet.  And we've already done a feasibility demonstration of mating
the payload and the upper stage in orbit, so there is no particular need
to fit both onto a single launch.  "The number of flights does not equal
the number of missions; it equals the number of missions times the number
of flights per mission." (Max Hunter)
If we feared danger, mankind would never           |       Henry Spencer
go to space.                  --Ellison S. Onizuka |

From: Henry Spencer <>
Subject: Re: SSTO vs nSTO
Date: Sat, 15 Jun 1996 16:49:05 GMT

In article <4pjs1c$> (Marcus Lindroos INF) writes:
>: The shuttle was constrained not to be a throwaway (and there was one serious
>: proposal to build it as an SSTO, by the way).
>You mean Lockheed's Starclipper...? Great idea-what killed it was probably
>that the DoD insisted on 65,000lb payloads rather than 25,000lb, so
>the vehicle grew by a factor of ten! 

No, I was thinking of Chrysler's SERV, actually, which met the full spec.
(In fact, it had twice the payload for ascent, although somewhat less for
return.)  Some aspects of SERV may have been a little optimistic, but it
merited more attention than it got.

>: Ariane was constrained to use storable propellants in its lower stages,
>: and was generally constrained to be a low-risk design.
>Yup, but not the Ariane-5 or H-II.

I read Ariane 5 as another low-risk design, actually.  That argument can
even be made for H-II, although somewhat more weakly.  National megaprojects
which are crucial to the future of entire space programs tend to have a
strong low-risk bias.

>: Neither Zenit nor Energia had any cost constraints.
>The Ariane-5 aims for a 50% cost reduction compared with existing Western

No, it aimed for a 10% reduction compared to Ariane 4, per flight, which
was about all that was realistic given a conservative low-risk approach.

>I would like to have a cost breakdown for expendable 1.5STOs such as
>Ariane-5 and Energia. I'd think the core plus LOX/LH2 engines cost
>far more than the boosters.

In the case of Energia I'd be quite surprised.  Russian LOX/LH2 engines
are not substantially more complex than Russian LOX/kerosene engines
(oh, a little, but not a lot), and there are four strap-ons.  For
Ariane 5, it's harder to call, but as I've recently observed in another
discussion, gimballed-nozzle solids (especially segmented ones) are not
as simple as people think.
If we feared danger, mankind would never           |       Henry Spencer
go to space.                  --Ellison S. Onizuka |

From: Henry Spencer <>
Subject: Re: weight growth (was Re: HTHL vs.VTVL)
Date: Sat, 15 Jun 1996 17:08:28 GMT

In article <4pit45$j44$> "Michael P. Walsh" <71544.2040@CompuServe.COM> writes:
>>A 5% weight growth in a modern VTVL SSTO would be within the
>>available weight margin, and hence would have no effect 
>>whatsoever on payload.
>...I can believe that 
>a successful design should have sufficient margin to accomodate a 
>5% weight growth but to say it will requires a leap of the 

It must; it's a practical necessity.  Part of the reason why SSTO is now
being taken seriously is a steady stream of studies showing that realistic
mass margins are now feasible.

>...I believe the comment about the air breathers have 
>less sensitivity to weight growth is somewhat correct, but they 
>start out with the awful penalty of having to carry inlets and 
>associated air-breather plumbing to orbit and back.  On the other 
>hand, the rocket SSTO is very sensitive to weight gains, as it 
>all comes out of payload.

Actually, it all comes out of payload either way.  The airbreathers'
advantage is that they have -- on paper -- somewhat superior engine
performance, which means they can haul somewhat more dry mass into
orbit.  It's easier to allocate comfortable margins that way.
If we feared danger, mankind would never           |       Henry Spencer
go to space.                  --Ellison S. Onizuka |

From: Henry Spencer <>
Subject: Re: Use of STS facilities after STS program
Date: Sat, 13 Jul 1996 04:43:11 GMT

In article <4s4i9p$> (Jud Ready) writes:
>>Why do you assume that it will use the VAB and Complex 39 at all?  An
>>operational SSTO will not be a government-owned vehicle.
>But does that prohibit the corporate entities from using the
>existing government facilities.  The infrastructure is there.  It makes
>little sense to not use it...

The infrastructure unfortunately comes complete with lots and lots of people,
who cost a bundle.  And why *bother*?  The whole point of the VAB is to do
vehicle stacking out of the weather.  SSTOs don't need to be stacked.  They
need hangars for maintenance, but you can build a lot of hangars for what
KSC spends on operating the VAB complex.

It is marginally possible that an operational SSTO might operate from the
Cape.  Not from KSC, mind you, but from the Cape.  Sensible people
wouldn't do that, because even the Cape has lots of expensive bureaucrats
whose professional specialty is getting in the way -- I kid you not, that
is the only way the safety people can show that they are doing something
to earn their pay -- but government contractors aren't always sensible. 

> ... Especially, when you consider that the money
>spent on developing an SSTO is gonna be *huge*...

While lots of people have made exorbitant estimates, do note that Lockheed,
which just won the X-33 competition, said of SSTO development costs:  
"people say five billion dollars, but we could do it for a lot less".

>Why spend more money buying
>land, building launch pads, and the associated plumbing?

Because land, pads, and plumbing are cheap, and using KSC isn't.  The
entire ground-support setup for DC-X cost a few million dollars.
If we feared danger, mankind would never           |       Henry Spencer
go to space.                  --Ellison S. Onizuka |

From: Henry Spencer <>
Subject: Re: New shuttle X-33 vs DCX delta clipper
Date: Sat, 13 Jul 1996 04:48:23 GMT

In article <> (Thomas J. Frieling) writes:
>Ok, let me get this straight. This NASA initiative was foisted on Clinton and 
>Gore against their will...

Damn nearly.  It has not been easy to keep reusable-rocket development
work alive under the Clinton administration, and Henry Vanderbilt and
others have sweated blood accomplishing it.  Even cheap little DC-X nearly
died, repeatedly, because of foot-dragging in the administration.
If we feared danger, mankind would never           |       Henry Spencer
go to space.                  --Ellison S. Onizuka |

From: Henry Spencer <>
Subject: Re: Use of STS facilities after STS program
Date: Mon, 15 Jul 1996 17:18:21 GMT

In article <Pine.SOL.3.93.960712132518.28975C-100000@seds> Mike Dicenso <mdicenso@SEDS.LPL.Arizona.EDU> writes:
>It makes sense to use the VAB, at least as a hanger of sorts for
>protecting vehicles that cannot be made ready for flight soon enough to
>escape severe weather, as is currently the case with hurricane Bertha. Or
>to do non-routine, high level maintaince work.

The VAB is a very costly hangar, especially for relatively low vehicles
which can use ordinary aircraft hangars.  You don't need the VAB unless
you have something that's really tall.  Even the VTHL SSTO concepts
generally are erected on the pad, not before rollout.

Incidentally, in a sensible launch organization, you do not even roll the
vehicle out of its hangar until it's ready to launch and the weather looks
good.  Parking it on the pad for weeks and weeks is ridiculous.  In fact,
the Complex 39 mobile-launcher concept was specifically meant to *avoid*
this, to permit preparations to be done under cover, away from weather,
until the very last minute.  The one time NASA actually did things that
way -- Skylab -- it worked quite well. 
...the truly fundamental discoveries seldom        |       Henry Spencer
occur where we have decided to look.  --B. Forman  |

From: Henry Spencer <>
Subject: Re: Venture Star specs
Date: Mon, 15 Jul 1996 19:49:57 GMT

In article <necrolar.837198170@merle> (Rene
Carlos) writes:

>c)Possibly.  I'm rather keen on the idea of ETs.  Why
>drop engines when it's the weight of the tankage you
>want to lose?  After all, a tank is far less complex 
>than an engine or an entire stage...

It's also much lighter, so ditching it doesn't gain you nearly as much.
(The logical extreme of this is the Atlas, which stages by dropping
engines only, since its pressure-stabilized tankage weighs almost
nothing.)  The benefit of drop tanks is mostly that they minimize the size
of the part that has to be recovered.  Unfortunately, you pay a price for
this, because a vehicle which is mostly empty tankage has a much easier
time of things on reentry and landing than a vehicle which is mostly
dense, heavy engines and equipment. 
..the truly fundamental discoveries seldom        |       Henry Spencer
occur where we have decided to look.  --B. Forman  |

Date: 11 Apr 91 18:11:15 GMT
From: ssc-vax!bcsaic!hsvaic!  (Dani Eder)

In article <> (Donald Heskett) writes:

>IMHO it's a very poor idea to complicate your launch system that much
>for a mere 300mph.

According to Dr. Dana Andrews of Boeing, who is the study manager of one
of the four industry SSTO studies underway for the SDIO, the addition of
1000 feet/second (680 miles per hour) to a horizontal take-off SSTO rocket
that uses oxygen/hydrogen propulsion will triple the payload.

This is because SSTO vehicles using chamical propulsion are today just
this side of feasibility, and so are very sensitive to how much velocity
they have to provide.  So in this case, a relatively small ground accelerator
can have a large effect on payload capacity.

Date: 11 Oct 91 03:01:34 GMT
From: ssc-vax!bcsaic!hsvaic!  (Dani Eder)
Subject: Re: DCX

burger@oldcolo.UUCP (Keith Hamburger) writes:

>It is not true that margins on SSTO vehicles are that tight.  Gary Hudson
>wrote out a thought experiment that shows that vertical-take off SSTO's are
>not that difficult.  

>I don't know the number offhand but he shows that using an Saturn-IVB stage
>(third stage of the S-V vehicle) with a Space Shuttle Main Engine (SSME) if
>you add the thrust structure necessary for the conversion and delete no other
>structure you can place a couple thousand pounds in LEO with a single stage.

I beg to differ. When structure is designed for multiple uses it has to
be heavier than when it is designed for one use.  The reason for this
is that cracks that start at flaws in the metal (and all metal has
flaws unless it is an unformed single crystal) grow at a highly
non-linear rate dependant on the stress in the material.  To delay
stress-induced failure (when the crack gets big enough that the
structure no longer supports the design load), you have to lower
the stress level, which means the structure gets heavier for the
same load.

In aluminum the stress has to be lowered, or the weight increased,
by about 10% for each factor of ten in service life in the range
of 1 to 1000 stress cycles.  So an SSTO with a life of 100 flights
would have to be designed 20% heavier than a 1 use structure (like
the Saturn third stage).  Note that the S-IVB was actually designed
to something like 5 pressure cycles (1 pressure test at the
factory, a few loadings of fuel with aborted countdown, plus
actual use in flight), and commercial aircraft are typically 
designed with a structural fatigue life 4x the design service
life (which is why few fail structurally in service).

So, if we make an SSTO out of an S-IVB, we would have to beef it
up by 20% in structure to get it to last 100 uses, which would
make it too heavy to get to orbit.

Data: (Source: Boeing Saturn V System Familiarization Training
Manual, October 1966)

S-IVB Gross weight 251,900 lb
Empty Weight 21,900 lb

Data: (Source: Shuttle Flight Operations Manual, Vol 8A: Main
Propulsion Systems)

SSME specific impulse = 453 sec.

Data: Velocity to Orbit = 9144 m/s (typical for existing rockets)
(includes losses from atmosphere and vertical flight)


Mass Ratio to Orbit = exp(deltaV/(g x Isp))
                    = exp(9144m/s / (9.80665m/s^2 x 453sec))
                    = exp(2.06)
                    = 7.83
Therefore propellant used = 1-(1/7.83) = 0.872 
of takeoff weight.
Propellant = 251,900 - 21,900 = 230,000 lb
Therefore takeoff weight = 230,000/0.872 = 263,660 lb.
Since stage gross weight = 251,900, and the difference between
a J-2 and an SSME is 6,990 - 3480 lb (from Rockwell engine data
sheets), and allowing 500 lb for thrust structure beefup (the
SSME is twice the thrust of the J-2, Im using 1/6 of the orbiter
thrust structure weight), and 1000 for larger ancillary systems
(gimbal actuators, etc, based on 2000 lb/SSME) we have a total stage 
weight of 256,910 lb, leaving 6750 lb for payload, guidance system, and
payload fairing.  Let us assume that the guidance system is 
only 500 lb, which covers the accelerometers, computer, batteries,
etc.  And Payload fairings typically run 20% of payload weight,
although they only count 1/5 of their weight in payload loss, since
you don't carry them all the way to orbit.

 Thus we have a net useful payload of 6000 lb.
This much I agree with the previous poster.

Now, If we beef up the structure 20% for a 100 flight life,
we get a 4380 lb increase, which only leaves us a 1620 lb
payload.  Of course, if you want to get a stage back, you
need a recovery system (heat shield, parachute, landing legs) 
75% of the thing you are returning, which in this case is
26280 lb, or 19,710 lb, which shoots your payload to hell and

To get to zero payload, you need to get a 25% reduction over
Saturn/Shuttle technology weights.

Dani Eder

Dani Eder/Boeing/Advanced Civil Space/(205)464-2697(w)/461-7801(h)/#905, 1075
Dockside Dr.,Huntsville,AL35824/Member: Space Studies Institute
Physical Location: 34deg 37' N 86deg 43' W +100m alt.

From: (Henry Spencer)
Subject: Re: New Venturestar Layout
Date: Mon, 3 Apr 2000 15:13:43 GMT

In article <>,
Jonathan A Goff  <> wrote:
>Take say HTP/JP-4 combo.  Bulk density is ~1.35kg/L.  Isp is
>averaged out to roughly 325sec (using a nice E-D nozzle for
>altitude compensation), and a T/W of roughly 150-200 for the

That's a pretty high T/W, although perhaps achievable.  Also, if we're
talking about the same "E-D nozzle" concept, note that it's been tested
(in wind tunnels) and doesn't seem to work very well.  See NASA TN D-4462,
"Performance of annular plug and expansion-deflection nozzles including
external flow effects at transonic Mach numbers", Robert A Wasko, 1968.

>What would the likely Delta-V required be?  I was
>guessing roughly 8500m/s, but that is a total SWAG...

Well, the shuttle is 8930m/s, so with a little gain from dense-fuel
effect, 8700m/s is a good first guess.

>> While challenging, this isn't an obvious disaster.  The Titan II first
>> stage was operational with such a ratio in 1963.
>Two points about the above.  I don't like balloon tanks for
>MCD vehicles...

Titan II did not use balloon tanks.  In fact, that was a structure which
supported not only itself but a heavy upper stage.

>That said, my second point is that you could likely get
>fairly similar mass fractions with simple fiberglass/epoxy
>fillament wound tanks if you have them both be spherical...

Note that with filament winding, there is no great advantage in making
them spherical.  Spherical tanks have an advantage with metal, because
cylindrical tanks have more longitudinal strength than they need, but with
filament winding you can just use more hoop windings to give the necessary
hoop strength without adding unnecessary longitudinal strength.

>> The one big disadvantage of it is one shared by all expendables:  the
>> inability to repeatedly fly the same hardware, to establish reliability
>> of, and shake the bugs out of, each flight article.
>However, by eliminating most of the systems, you
>also eliminate most of the error potentials.

It still doesn't get you to the same level of reliability, and confidence
in reliability, as being able to test-fly the vehicle before trusting it
with a paying cargo.  The alternative for reaching the same level of
reliability is automated manufacturing which builds hardware very
consistently, plus a *lot* of rockets expended on test flights.  A full
and proper test program, the sort that's used for aircraft, will take
several dozen launches at the very least (with heavy optimism on how well
you can substitute careful data analysis for more test flights, and how
much you can gain from a much more constrained operational flight plan).

Remember that design flaws don't always show up until you hit just the
right conditions.  For example, combustion instability can afflict even
very simple designs -- in fact, simple designs may be more vulnerable,
since they tend toward single large engines rather than clusters of
smaller ones.

>Of course, I don't blame you for your opinion, as
>it is pretty much industry standard to design stuff
>so complexed that it is doomed to have to be tweaked
>into submission.

Debugging is a constant in all engineering.  Rocketry is the only major
field of endeavor where we have tried to pretend that it's unnecessary,
that we can get it right the first time.  There have been repeated
demonstrations that this optimistic view is simply wrong, yet people still
cling to it.
"Be careful not to step                 |  Henry Spencer
in the Microsoft."  -- John Denker      |      (aka

From: (Henry Spencer)
Subject: Re: New Venturestar Layout
Date: Mon, 3 Apr 2000 22:14:39 GMT

In article <>,
Jonathan A Goff  <> wrote:
>> ...See NASA TN D-4462, "Performance
>> of annular plug and expansion-deflection nozzles including
>> external flow effects at transonic Mach numbers", Robert A Wasko, 1968.
>...Was it that it didn't work as well
>as expected, or that it just plain doesn't work?  I probably
>won't have access to that report you cite for quite some time.

Let's see, I don't have a copy of it myself, but I do have the notes I
made from reading it...  Pulling out just the E-D parts...  (Caveat, I was
much more interested in the plug-nozzle stuff, so I didn't pay careful
attention on the E-D side.)

  "E-D performed poorly, little evidence of altitude compensation...
  Base pressure was consistently below ambient...  Adding secondary flow
  [base bleed] didn't help it much."

The conclusion was that the E-D nozzle really doesn't altitude-compensate
well at all, at least not in the specific form they tested.  The base
cavity isn't effectively connected to the outside air, and aspiration
pulls its pressure down, inducing extra drag.

>> Well, the shuttle is 8930m/s, so with a little gain from dense-fuel
>> effect, 8700m/s is a good first guess.
>Really?  That's all?

You might be able to do a bit better, but this number is fairly easy to
defend... and substantially lower than the requirements one sometimes
sees imposed.

>You are right, it was the Atlas that used the balloons.
>My mistake there.  However, I did hear that the cost
>for fabbing Titan II tanks was hideously expensive.

I wouldn't be at all surprised.  That's an impressively light structure,
considering when it was built.

>> Note that with filament winding, there is no great advantage in making
>> them spherical...
> it easier to fab a fillament wound spherical
>tank that didn't have varying hoops?

I don't think so, at least not very much.  Winding direction is something
you get to specify essentially for free with filament winding.

>> It still doesn't get you to the same level of reliability, and confidence
>> in reliability, as being able to test-fly the vehicle before trusting it
>> with a paying cargo.
>I'm sorry, but I just don't subscribe to this philosophy.
>A design if good should work every time.

True in principle, but there is a big gap between theory and practice
here. In particular, you don't necessarily know that the design is 100%
good, or that you are controlling every manufacturing or materials issue
that matters.  More than anything else, testing defends against
*surprises*, things you did not think of.

>Look at Porsche.  For a long time, every time a car would come
>off the line, they'd have 40 engineers who'd test it,
>tweak it, and then eventually get it to work...  Finally they started
>having cars come off the line that weren't defective.

Betcha they still test them, though.  When they hand the customer the key,
he isn't starting the engine for the very first time.  They may expect
the thing to work when they test it -- and treat it as an urgent problem
in the manufacturing process if it doesn't -- but that is VERY DIFFERENT
from not testing it at all.  Surprises happen.

>...If you have to test a ship too much
>before it flies, some of the parts are actually already
>getting to near the end of their life-cycles...

Note that I'm not saying "before it flies".  I'm talking about *flight*
testing.  Once the design is shaken down, I agree that there is little
benefit in extensive ground testing.  The way to be sure that it will
fly successfully is to fly it successfully.

>Why not do it right the first time, instead of doing it wrong several

Because if you base your plan around being sure that you will do it right
the first time, you'll have customer payloads aboard when you discover
that you didn't.  And you will.

>...Not that I don't have respect for engine designers,
>just that I feel we need a more firm grasp on what
>causes the problems, and a set of computer analysis
>tools and other tools that allow us to not repeat
>the same mistakes every time...

Good luck.  People have been *trying* to do that for combustion stability
for forty years now.  Instability is an extremely complex process which
simply is not well understood, and it's not for lack of effort.  There are
some general rules of thumb about how to avoid it... but some of them have
awkward exceptions, and more of those might lurk somewhere.  (Everyone
thought small engines were pretty dependably stable, until the shuttle RCS
thrusters had stability problems.)

>If the engine isn't combustion stable, then no ammount of incremental
>tests will make it stable.

Correct.  But with the present state of the art, despite all the effort
invested in this, the only way to find out whether the engine is stable is
to test it.  And if you find it's unstable, there are rules of thumb about
how best to fix it, but the only way to be sure is to test the fixes.  If
you skimp on the testing, you get what happened to the second Ariane 1.

>> Debugging is a constant in all engineering...
>Actually, I see the exact opposite view in most industries,
>and they are finally coming to the realization that it is
>possible to produce stuff that works without extensive
>tweaking up front.

You're confusing two separate issues.  Industry is working quite hard to
get the design right the first time, and this is a good idea.  Relying on
"extensive tweaking" is a bad plan.  But that is very different from
assuming that the thing is correct by decree, without ever testing it;
that's the mistake the Ariane 5 management made.  The purpose of testing
is not to tweak a bad design until it works, but to catch surprises...
which affect even good designs.
"Be careful not to step                 |  Henry Spencer
in the Microsoft."  -- John Denker      |      (aka

From: (Henry Spencer)
Subject: Re: Is Roton Dead?
Date: Sat, 6 Jan 2001 06:33:31 GMT

In article <>,
Michael J Wise  <> wrote:
>> Roton was to be an SSTO.  No boosters, no upper stages, no pieces falling
>> off along the way.  It takes off, flies to orbit, comes back down...
>Sorry, I thought building an SSTO was a bigger issue. Has anyone actually
>built a (non-Solid-Fuel) SSTO yet?

Roton was definitely considered rather daring.  Nobody has actually
*flown* an SSTO (with any sort of fuel), but there are some existing --
indeed, museum-piece -- expendable rocket stages that have the necessary
performance.  The big uncertainty about building an SSTO is whether you
can make it reusable at a reasonable price.

>But back to the Roton:
>What fuel/oxydizer was it going to use?


>How was it going to solve the weight problem?

Good design and modern materials.  Expendable stages light enough to have
definitely solved the "weight problem" were flying in 1962 (the Titan II
first stage).  Economical reusability is the only remaining trick.

>Is there an engine (other than the AeroSpike) that would allow you to
>build a non-nuclear SSTO vehicle?

Aerospikes are popular for SSTO designs, because the extra performance
edge is certainly useful, but it's not at all necessary.  Any good
lightweight engine will do for non-hydrogen designs; hydrogen systems have
to sweat harder because hydrogen hardware is so heavy, but it is doable.
(If you put six SSMEs under a shuttle ET, it could reach orbit all by
itself... with a fairly heavy load of cargo.)
When failure is not an option, success  |  Henry Spencer
can get expensive.   -- Peter Stibrany  |      (aka

From: (Henry Spencer)
Subject: Re: Is Roton Dead?
Date: Tue, 9 Jan 2001 21:11:51 GMT

In article <93eaqs$6a4$>,
David Kinny <> wrote:
>> fact, the central problem with using
>>the Titan II first stage as an SSTO is that it has *too much* thrust to
>>fly an efficient trajectory.
>How exactly does too much thrust prevent flying an efficient trajectory?
>Difficulties in flipping over to horizontal?  Or something else?

Basically, in the time it takes to climb clear of the atmosphere, it picks
up too much vertical velocity.  This thing was an ICBM, designed to move
out fast... and flying as an SSTO, it hasn't got a hulking great second
stage on top to slow it down.  (In fact, a secondary problem of having too
much thrust is the bone-crushing acceleration toward the end, when the
tanks are almost empty.)  An SSTO launcher wants to take things a bit
slower, so that it can tip over to horizontal gradually, as it leaves the
atmosphere, and still have most of its fuel left for horizontal acceleration.

You can't just throttle back the engine, first because it wasn't
throttlable :-), and second because you need to keep it operating
efficiently, which throttling usually sacrifices to at least some extent.
However, *reducing* the performance of an engine is usually not a
difficult engineering problem!
When failure is not an option, success  |  Henry Spencer
can get expensive.   -- Peter Stibrany  |      (aka

From: (Henry Spencer)
Subject: Re: SSTO and economy class syndrome
Date: Tue, 9 Jan 2001 21:19:54 GMT

In article <1emyoho.12gsk6d16jb8mgN@[]>,
Steve Schaper <> wrote:
>Here's a marketing opportunity for SSTO:
>Eliminating economy-class syndrome...

Potentially quite an interesting market... but it probably has to wait for
at least the second generation of SSTOs.

Ballistic passenger service needs spaceships with proven reliability,
proven to the point where airworthiness authorities will clear them for
passenger service *and* for routine operations near major cities.  They
also have to be able to fly with 99.9% on-time reliability in 99% of all

Finally, people have to be sufficiently confident of all of this --
confidence that will probably come only from seeing the things operate for
a while -- to be willing to put up a multi-billion-dollar investment for
enough spaceships, and enough ground support facilities, to establish at
least the beginnings of a route network and a regular service.  You can't
do this with one or two ships, and you can't do it without solving a
variety of problems with things like air-traffic control, and so there's
going to be a big up-front effort and investment before a single ticket is
bought.  The technical unknowns have to be brought down to nearly zero
before people will be willing to tackle that.
When failure is not an option, success  |  Henry Spencer
can get expensive.   -- Peter Stibrany  |      (aka

Index Home About Blog