Pegasus/SciSat Launch Cost

Kim Keller wrote:
"Henry Spencer" wrote in message
...

Uh, no, that's ridiculous. Launchers operate for only a few minutes
between leaving the atmosphere and finishing their job; the chances of a
radiation upset during that time are minimal.


This is not true. Most launchers that access GTO use multiple burns of the
second stage. Some missions last for six hours before spacecraft sep. During
that time, their trajectory takes them through the lower edges of the Van
Allen Belt. Since the "brains" for the whole launcher are carried in the
second stage, they'd better be SEU-resistant.



SEU-resistant is a *system* requirement, not a individual parts
requirement. You could achieve the exact same results of a rad-hardened
chip by a slightly different avionics package design. Calculate out
the radiation protection requirement for the chips, then seal them
in a box that meets those requirements.


Non-rad-hard computers have

operated in LEO for years without any difficulties, on the amateur-radio
satellites. The requirements in this area are ridiculously inflated; the
launcher companies inherited them from the missile business, and stick
with them because they're traditional, not because they're needed to
complete the mission.


Again, the difference is the altitude of the operating environment. LEO
altitude does greatly reduce the incidence of SEUs, but it does not
eliminate them, particularly around the South Atlantic Anomaly.


Except that much of that effort is no longer necessary. It's an artifact
of trying to build high-reliability hardware with 1950s electronic parts.
Modern parts are so much better -- often *superior* to the mil-spec ones
when you look at the actual numbers -- that the old design practices are
now in the "nobody ever got fired for doing it this way" category.


That's an opinion, not a fact. Given the cost of qualification testing, one
must wonder why a launcher constructor would willingly spend that much money
to get their vehicle ready for flight, particularly if they're paying the
tab, not a subsidizing government agency. The reason is the testing
philosophy works and booster reliability increases as a result.



Out of curiousity, what would the differences be in testing a Mil-spec
vs non-Mil-Spec systems? Seems to me that the test program is the same
for the vehicle no matter it's individual parts. The test hardware is
the same either way. The tests are the same. The costs are the same.

Henry Spencer wrote:

In article ,
John Schilling wrote:
Believe me, some of us have thought about it. :-) Really hard to get the
Canadian government to fund it, however. Partly there's an ingrained
inferiority complex, a belief that Canada couldn't *possibly* do something
that daring all by itself,

And this from the country that built the Arrow?
Oh, how the mighty have fallen.

The Arrow fell partly for the same reason, alas -- it was inconceivable
for Canada to succeed in something so ambitious, so the program *must*
somehow be headed for disaster, and the sooner it was terminated, the
better. Being unspoken and implicit made this belief all the stronger,
since there was no way to openly challenge it.

The people who built the Arrow built a "Jetliner" for airline use several years
earlier. It was advanced enough to get orders in the US years before Boeing got
into jets, but the Canadian government shut it down because they wanted the
company to focus on fighters.

John Halpenny

"Charles Buckley" wrote in message
...
SEU-resistant is a *system* requirement, not a individual parts
requirement. You could achieve the exact same results of a rad-hardened
chip by a slightly different avionics package design. Calculate out
the radiation protection requirement for the chips, then seal them
in a box that meets those requirements.

Okay, works for me.

Out of curiousity, what would the differences be in testing a Mil-spec
vs non-Mil-Spec systems? Seems to me that the test program is the same
for the vehicle no matter it's individual parts. The test hardware is
the same either way. The tests are the same. The costs are the same.

The differences would depend on the manufacturer or his customer: what
levels do they want their parts tested to? The tests would be basically the
same in set-up, but the key question is, do they envelope the components'
expected operational environment plus a satisfactory safety factor? That
would be called out by the customer's spec.

-Kim-