Interesting story, last night I went down the shop
to do some work and found a small puddle on the floor.
A solder joint where the copper hot water pipe was joined
to a 3/8 NPT adapter into an old bronze valve had let
loose and was leaking water onto the floor.

I took the opportunity to show my daughter how to
do a sweat solder repair (per harold's and my
discussion of a few days ago) and we had the
water back on in an hour or so.

After that we took the leaking fitting and hacksawed
it apart so that we could find the problem, and it
was pretty apparent that the plumber who did it
didn't clean it or flux it nearly enough so there
were major voids inside the joint that were full of
oxide. It lasted about 50 years, but I guess
nothing lasts forever.

The interesting part though was that today I took the
parts into work to give them the close hairy eyeball
under a good microscope, and while looking at them,
I realized that whoever made up the threaded fitting
between the adapter and the valve, used some unusual
compound in the threads - in fact, he had used nothing
more than cotton sewing thread. It was all bunched
up at the root of the male copper threads, but there
were several layers of thread that I could unwind and
inspect.

There wasn't any dope or tape or matrix material of
any kind mixed in with the thread, all of the fibers
were clear and freely visible.

I wonder if this was a common practice 'way back
when,' using sewing thread to seal pipe fittings?

It seems to have worked just fine, as the only leak
in evidence was at the copper-to-copper joint, all
of the threads were leak-tight.

The bronze valve btw had a) no washer left in it,
and b) the screw to hold the washer was mostly gone
as well.

Ah the joys of old houses!

Jim

==================================================
please reply to:
JRR(zero) at yktvmv (dot) vnet (dot) ibm (dot) com
==================================================