David,
It would only
make sense to test against the high pressure face of the viewport. It
didn't look to me like they used the sub for the test. That would have
required inverting the viewport and filling the sub with water completely.
It looked like they tested in a barrel device of some kind and inverted the
viewport on that so the low-pressure side was exposed to the shop. It's
unclear if their mounting technique in this test case contributed to the
failure, but it certainly could have and precipitated an early failure.
I'll check the Stachiw numbers later on this weekend, or someone else who has
the book could do the same. I'm estimating the diameter at 20-24
inches.
Jon
|