Next, after a few questions, Dr, Walter MacNair discussed two
subjects, product testing and the external initiator program.
Dr. MacNair contrasted product testing in the manufacture of -
Product
Hosting
nuclear weapons vith the usual manufacturing: eltuation in which items
are produced for public use in large quentities.
In the latter case
large scale customer use supplies an overall statistical quality test
on the iten, a method not applicable to nuclear wespons.
The Sandia
Laboratory attempts to invent and develop substitutes. for customer
use testing; this effort accounts for about one third of the laboratory's total budget.
The tests include laboratory determinations of
the reactions of components to environmental conditions (impact,
vibration, acceleration, climatic exposure) ; wind tunnel+ experiments
assurance
e program ia carried‘out{in1the fashion of ‘industrial epot~
check inspections.
Finally, each completed stockpile item is sub-
jected to a continuing surveillance,
The surveillance program begins
with a complete non-destructive test when the item arrives in the |
stockpile.
It is tested subsequently at intervals of not less than
eighteen months.
The present stockpile items are tested every five
months, on the average.
In answer to questions, Dr. MacNair said
that components in the stockpile occasionally fail to meet specifica~
tions, but there is practically never a bomb that wouldn't work.