launch

Thank you for your reply

>I am not so sure I agree with your analysis. General
speaking, Moore's law applies to doubling of the POWER
of a chip not the doubling of volume<

My analysis was based on an article on historical manufacturing cost data.
The article was a sidebar to put Moore's law in its historical context.
The relevant volume for Moore's law is the number of transistors, and yes
this does track the historical trend I cited. The number of transistors on
a chip has gone up sharply over time.

The unusual thing about Moore's law is the long time it has gone on, not
the trend itself.

>As for launch costs, they only go down so long as the
yearly volume increases, not the volume over 16+
years<

This may be true, but it would make rockets different from all other
products that have been studied.
I think this is unlikely. Any product could make a claim to uniqueness,
and yet a variety of very different products show very similar experience
curves with respect to cost. [cars, golf balls, limestone, phone calls,
paper bags, steel ingots, pins, cookies, airplanes, TVs, broiler chickens]

If, and note I do say if, the experience curve argument does apply to
rockets, what would be the consequences?

1. re: Al Globus' proposal.
The proposal may be a way around the experience curve, by paying for
development costs before the market justifies it. However, in the
historical examples technological advance is only one facet of cost
reduction, and I do think volume will be required to really push costs
down. I would also hesitate to favor RLVs over expendables - see below.

2. Expendable costs can still come down a lot. The experience curve says
the lack of cost reduction over the past couple of decades is to be
expected. Those people who point to the cost of materials, propellants,
and operations and say that costs could be a lot lower are probably right.
However, we need experience to get there. As an example, Intel could
probably have predicted the performance of the Pentium IV in 1990, but that
did not mean they could produce it in 1992. Experience is important.

3. As someone (Jack?) said earlier on this forum, for near term market
ideas we need to accept current launch costs. We would like ideas for
which the market would increase rapidly as costs are lowered, which would
produce the "virtuous cycle" typified by Moore's law. More volume = lower
cost = more volume = lower cost, etc., etc.

4. To produce cost reductions of 20-30 per cent per year, we need 4000
launches the first year, 8000 the second, and so on.
This is the key, and somewhat discouraging point. Maybe the market can be
segmented, allowing a start over at lower numbers.

5. Launches of any sort should be encouraged.