Newsgroups: sci.space.shuttle
Path: utzoo!henry
From: henry@zoo.toronto.edu (Henry Spencer)
Subject: Re: launch rates
Message-ID: <1990Oct7.045635.6958@zoo.toronto.edu>
Organization: U of Toronto Zoology
References: <10195.26fde341@pbs.org> <1990Sep25.033816.16652@zoo.toronto.edu> <1990Oct1.191917.24542@cimage.com> <1990Oct2.031535.6556@zoo.toronto.edu> <1990Oct3.232201.5196@cimage.com>
Date: Sun, 7 Oct 90 04:56:35 GMT

In article <1990Oct3.232201.5196@cimage.com> gregc@dgsi.UUCP (Greg Cronau/10000) writes:
>Henry this is really getting ridiculous. You continue to put down my
>assertions based on a set of data that you contend is meaningless, and then
>turn around and make your own assertions based on that very same data.

If you really think that, then this discussion is pointless.  Let me make
one more try.

You do not find out whether Joe Doakes can run faster than Fred Bloggs by
holding a race between Joe, in sneakers and a track suit, and Fred, in an
arctic parka with a lead ball chained to one ankle.

You will not find out whether the Saturn V could sustain a higher launch
rate than the Shuttle by comparing the Shuttle, trying year after year
for the highest possible rate, to the Saturn V, deliberately held back
to a lower-than-possible launch rate to give more time for scientific
assessment of each mission's results.

If you want to compare maximum sustained launch rate, you look only at
the subsections of each launcher's record where they were trying for a
high launch rate with no extraneous restrictions.  That means you delete
the early development ramp-up, the aftermath of catastrophic problems,
and periods when launch rate was deliberately held back for reasons
unrelated to technical problems.

What are you left with?  For the Saturn V, a bit over a year, from about
Apollo 8 through Apollo 13.  If you want to compare launch rates, you
compare *that* against the period from about STS-5 to the present,
omitting the post-Challenger hiatus.  And you get the results I cited:
the Shuttle beats the Saturn V just once, and ends in disaster doing it.

If you look carefully at the numbers, actually, you conclude that even
this most-meaningful comparison doesn't mean much.  On the Saturn V end,
that's an awfully small set of data; clearly the Saturn V never really
had much chance to demonstrate what it could do.  On the Shuttle end,
cutting off the development period at STS-5 is a doubtful procedure,
even if that was the end of the development period according to NASA,
and you have to wonder whether things like this year's hydrogen leaks
should get special treatment.

Looking at the facts behind the numbers is harder, because you have to
make guesses, but we can try.  The Saturn V got about 5/year with no
great strain, and contingency plans to add an October launch in 1969
were far enough along -- with no serious problems seen -- that there
is at least weak evidence for a sustainable 6/year rate there.  On the
Shuttle side, many of the strains experienced in 1985 were preventable,
and would not recur today, so there is at least weak evidence for a
sustainable rate of 9/year.  The latter also fits with the NRC report
on shuttle launch frequency, which put the limit about there with a
surge capacity slightly higher.

>1.) I am not "Obsessed" with numbers. But the space launch question revolves
>    around numbers: maximum thrust, weight, tonnage to LEO, launch turn
>    around time, etc. If you can provide a qualitative vs. quantitative way
>    to evaluate boosters, please, I'd like to hear about it.

We need numbers, but we need to look at what the numbers signify, and not
just crunch through the calculations blindly without understanding the
meaning -- or lack thereof -- of the results.

>2.) You state that the "Saturn V achieved 5[launches]/year, briefly,"
>    It did that *once*. You asked me to talk to a statitician about my
>    conclusions, what do you think statiticians call a one-time unrepeatable
>    spike in data? I'll tell you: noise.

The same comment, of course, can be made about the shuttle's 9/year in
1985.  Deduct that "noise" and the shuttle average doesn't look so good.
As I discussed above, when one looks at the *reasons* for those peaks,
one concludes that they probably represent attainable performance much
more closely than the averages do.

>3.) You state that: "The evidence clearly shows..." No. Sorry. The evidence
>    shows no such thing. The design flaw in the SRB was aggravated by the
>    cold on the morning of the launch. It was an accident waiting to happen.
>    It would have happened if it had been preceded by 1 or 20 launches in the
>    previous year. I will grant, however, that the evidence does "suggest".

If you read the Rogers Commission report, you will find that they make
much stronger statements than this.  There were quite a number of things
that were accidents waiting to happen, many of them aggravated by the
higher launch rates.  And the major reason why the cold-weather launch
triggered that accident was the intense pressure to keep up the high rate.

>4.) If, however, you would like to use that argument, what about Apollo 13?
>    Memory fails me right now, and I don't have any of my ref books handy, but
>    didn't it occur shortly after that 5/year period you mentioned? It could
>    just as easily be blamed on the hurry up syndrome that you claim caused
>    challenger.

Please construct a plausible argument for this before proposing it, preferably
with references.  The Rogers Commission was quite explicit about the role of
the hurry-up syndrome in Challenger.  I know of no equivalent line of
reasoning for Apollo 13.  Deadline pressure does not seem to have been a
factor that I am aware of.

>... consider this: Saturn, because
>it is expendable, has an absolute lower limit to it's cost...

Unless you start looking at making the stages recoverable, which was indeed
on the agenda.  Recovering the first stage should not have been prohibitively
difficult -- a recoverable S-IC almost became the shuttle's booster -- and
I've seen proposals for recovering the third stage from Earth-orbit missions.

And that "absolute lower limit" is set by manufacturing considerations,
which change considerably if you commit to volume production as originally
planned for the Saturns.

Comparisons are difficult because of changes in the value of the dollar,
but I've seen studies which concluded that the cost per kilo of a Saturn V
launch to orbit -- at Apollo prices, no recoverable stages, no volume
production -- equalled or bettered current Shuttle prices.

>... The shuttle *can* be cheaper,
>it was promisied and designed that way...

I'd be interested to hear your rationale for this; there is no evidence
to support it that I'm aware of.  (There *is* evidence for cost reduction
due to volume production of expendables, by the way.)  Promised, yes;
designed, debatable; "can be cheaper", it's only been getting more
expensive as lessons have been learned and hidden subsidies removed.

I agree that *reusable launch systems* in general have the potential
of being cheaper than expendables.  It is not so very clear that actually
doing this is practical today, especially if you assume comparable funds
for development work on both approaches.  It is very clear that the
Shuttle is never going to do it.

>... Do you have any idea how hard it is to estimate what it's going
>to cost to build and maintian somthing *that's never been built before!!!*

The airliner manufacturers do it all the time, quite successfully.
I would recommend the OTA report "Reducing Launch Operations Costs"
as background reading on the subject.
-- 
Imagine life with OS/360 the standard  | Henry Spencer at U of Toronto Zoology
operating system.  Now think about X.  |  henry@zoo.toronto.edu   utzoo!henry