"jim rozen" wrote in message
...
In article , Harold & Susan Vordos says...

What can happen is when you feed the compound on the wrong flank of the
thread, the pressure of the cut can overcome the friction of the carriage
and move it away from the screw, and it often does. The net result is
what is termed a drunken thread.

I've always heard that term used for threads that were
die-cut. And that it can be caused by a die cocking
on the part and still threading along.

I would think that threading with the compound set
along the axis, but with the compound screw not
loaded up against the force of the cut, would cause
a thread that simply had the wrong pitch when
it was done - by the amount of backlash in the
compound screw, basically.

Jim

In this scenario, you'd realize the compound had shifted because the tool
wouldn't pick up the thread on a succeeding pass. The problem I speak of
tends to go unnoticed because the pitch wanders back and forth, usually at
the same spot in the thread, although that wouldn't necessarily be true.
The problem can come on the first pass, which then steers the carriage so
the drunken condition repeats at the same place each time and you don't
notice it happening. The amount of movement can be very minor, but that's
all it takes to screw up the thread.

I experienced just such a thread on a part for Thiokol many years ago while
in the employ of my last employer, just before starting my own shop. I was
running a Cincinnati tray top, a machine for which I have little regard,
cutting an acme thread. The carriage was very loose, and the hand wheel
was not balanced. As the handle of the hand wheel came over the top, the
weight of the handle bumped the carriage ahead slightly, until the lead
screw caught up with it. Needless to say, it did so repeatedly until the
thread was finished. Had I been running the machine for some time, I'd have
been aware of the machine's propensity to cut threads poorly, but it was my
first day on that particular machine. One of the rare returned items I
experienced while working there. Took a little searching to understand why
the thread was not good.

Harold