Re: 1teraflops cell processor possible?

Bill Todd <[EMAIL PROTECTED]> wrote:
>"George William Herbert" <[EMAIL PROTECTED]> wrote:
>> [Shuttle budget cutbacks]
>> The wide variety (literally hundreds) of concept designs
>> before and after the rescope, going back into the early 1960s
>> in fact, are well documented in Jenkins' book (and other
>> places).
>
>That's all well and good, but does nothing to address the criticisms that
>were offered.  To do so, you'd have to demonstrate that the areas that
>were changed materially affected the viability of the result, rather than
>simply scaled back its scope.

The Jenkins book and related Shuttle program technical
histories demonstrate that the scalebacks materially affected
the reliability, maintainability, reusability, and margins
of the resulting vehicle.

>> I am an aerospace engineer and have designed re-entry
>> vehicles (not that have flown, but done concept design
>> work) and I have *no* idea what you are talking about
>> regarding the lifting body shape not being 'safe'.
>> There are tradeoffs from ablative thermal protection
>> systems (heavier than tiles and metallic shingles and
>> the like, but can withstand higher peak loads) that
>> make them work better with capsules.  But there is no
>> law of nature that reusable thermal protection is unsafe
>> or that lifting re-entry / lifting bodies are somehow
>> inherently unsafe.
>
>To say that there's no known such law of nature avoids addressing the
>reality that with today's technology (let alone the 30-year-old technology
>used in the Shuttle) we seem to know how to create considerably safer
>ablative shielding than reusable thermal protection.  Until that situation
>changes, the conclusion stands.

You're saying this like there aren't any failure modes for
ablative heatshields, Bill, and that's not even vaguely true...

They're large, thin composite structures engineered to vaporize
in a controlled manner not be as strong as possible.  There have
been cracks in them in processing, due to damage (Apollo 1) and
there were very real fears in Apollo 13 that the heat shield was
busted by the tank explosion.  They are not easily testable;
you can test reusable TPS incrementally, but ablators are one
use only items.

Again, I say these things as an ablator fan and a capsule fan;
I like using them, but doing so responsibly requires knowing
the tradeoffs, and the limitations of the technology not just
its advantages.  Application of hubris to aerospace technology
often leads to fatalities, even with generally well known
and well liked and robust technologies.

>  I say this as a capsule bigot and
>> someone who pushed very hard for OSP to consider capsules.
>> Capsules are cheaper.
>
>Exactly.  So the primary argument for the shuttle - that it would
>dramatically *lower* the costs of placing and maintaining objects in orbit -
>evaporates.  The Russian concept of 'big dumb boosters' may not have pushed
>the frontiers of several areas of technology that the Shuttle pioneered, but
>as a means to reach and work in space cost-effectively it seems to have won
>hands-down.

A reusable vehicle, to cost less, has to have lower per-flight costs,
AND has to fly often enough to amortize its greater development
costs over the flights it flies.

Shuttle was backed into the corner of the trade space where it
wasn't going to be reusable enough and robust enough to be
able to fly often enough to do that.  By having its development
budget axed, and then not doing a proper top level systems
capabilities requirements reassessment to see if the amount of
money available really was enough to do RLV or not.

>>  Safety can be done right or wrong
>> with either capsules or lifting bodies.
>
>While that is certainly true in a relative sense, when compared on an
>absolute basis capsules seem considerably safer given current technologies.

There have been four fatal flight accidents in manned spaceflight
history (ignoring for the moment ground training and such);
Soyuz 1 and 11, Challenger and Columbia.

One was launch phase, two were re-entry, one was a re-entry failure
due to a launch phase incident.

Soyuz 1 was a parachutes failure.  Those nice, reliable parachutes
that everyone keeps saying never fail?  Oh, and we also lost one
of them on an Apollo, though the redundant units were enough to
prevent any serious mishap that time.

Soyuz 11 was a systems failure (stuck valve depressurized the
craft during re-entry) and was unrelated to the vehicle's
shape and re-entry mode. 

Challenger was the ascent loss.  Shuttle wasn't designed to
survive certain failure modes during ascent, due to budget
cutbacks forcing them to a design space where they couldn't
solve some of the technical problems reasonably.  No reasonable
thrust termination and escape rocket systems were likely
possible once the large SRBs were baselined.  Russia has
had two launch accidents with Soyuz which were not fatal.
A better done winged vehicle wouldn't necessarily be
dangerous like Shuttle is, though.

Columbia was... complicated.  A lot of details specific
to the engineering of the Shuttle all collided in negative
ways at once.

>> The comments regarding pilots and flying things are off
>> base as well.
>
>Really?  I could imagine that the rationale owed as much to the potential
>for developing related military technology (where the benefits of an ability
>to operate in both atmosphere and vacuum are obvious - the same is also true
>of single-stage-to-orbit vehicles) as to a simple 'fly-boy' mentality, but
>that's about it:  there is no obvious rationale whatsoever for the
>atmospheric capabilities of the Shuttle (or SSTO) for NASA use, and several
>good cost and safety arguments against it.

The reason for the wings configuration is crossrange on re-entry.
High hypersonic lift to drag.  NASA wanted lower hypersonic L/D
and lower crossrange; the DOD wanted higher crossrange so they could
loft certain payloads into polar orbit and return to land in one
orbit if they had to, and certain related mission issues.

Fundamentally, again: if you're going to reuse the vehicle
it has to land softly and to land softly it has to be
able to land on specific points and at specific attitudes.
Reusing a capsule which lands blunt end down on random
soil may not be possible, even if you want to.
Salt water is very hard on aerospace equipment.
To reuse you need to either make very rugged equipment,
or land it softly with wings or with rocket motors.

>  To reuse a large vehicle it has to be
>> flown to a relatively low velocity pinpoint landing.
>
>Reuse may be over-rated (at least it certainly never operated to reduce
>Shuttle costs below that of its non-reusable competition), or the definition
>biased.  The only part of the system that usually needs to return to Earth
>is the crew compartment, and reuse of that (minus its ablative shield)
>should be relatively easy - especially if the landing uses water to cushion
>the impact on the structure.

Water immersion generally voids the warranty on most aerospace
alloys.  Jets that skid off the runway into the ocean, even if they
just get lightly soaked, are junked.  Same thing with capsules.
Stainless steel may not be so subject to it, and Titanium isn't so 
bad, but most capsule structures are aluminum.

>Given the difficulty of putting mass into even low Earth orbit, putting any
>more up there than is necessary seems silly.  The Shuttle structure not only
>contains a large cargo bay (which usually returns to Earth empty) but also
>its heavy and powerful main engines (and it must be strong enough to survive
>their thrust) - neither of which is of much use once LEO has been attained.

The engines still being along is a function of the rocket being a
stage-and-a-half vehicle, not it being reusable.  Any single stage
or stage-and-a-half rocket has the same problem, going back to
the early Atlas rockets.

The cargo bay is, in fact, not a major weight component of
the overall shuttle structure.  

The whole airframe, needed to return all the parts safely
to the ground, *is* a major weight component... an expendable
vehicle with the same basic technologies would save at least
fifty tons over the shuttle's weight.

>As long as propulsion systems and fuel dominate cargo capacity in launch
>vehicles (exactly the reverse of the situation with general-purpose
>aircraft), using purpose-built vehicles built up from standard (and, where
>feasible, reusable) components rather than attempting to create
>general-purpose, reusable vehicles at coarser grain seems to make much more
>sense.

Fuel is cheap.  If you ever build a rocket where the fuel bill
is a noticable fraction of the launch cost, you will receive
serious congratulations from the space launch community.

Using nitrogen tetroxide is cheating, however.

Propulsion systems are not cheap, however.

We could go on ad infinitum about big dumb boosters;
I happen to agree that they're a really good technology
option at this place and time, but if the volume of
launches increased by a factor of ten, as it might
if costs dropped down to low hundreds of dollars per
pound launched, then reusable vehicles will likely
be a significantly more cost effective option from
that point onwards.

I say that as someone with man-years of so far unreimbursed
engineering time into some BDB booster designs and tech
analysis, plus two so far failed rounds of commercializing
and capitalizing to get the full scale development program
going to build one.  BDBs are great, until they succeed
wildly, and then RLVs will win in the long run.
I hope to help make that happen and make some money
in the decade or so that it takes for that to happen,
if I am lucky.


-george william herbert
[EMAIL PROTECTED]