[ARC5] Fw: Screw failure

[Michael Hanz]

Heh, heh...I've learned never to say either never or always.  Heat
cycling may indeed be a factor, Jay, but I guess I've been lucky, or the
east coast air has something in it.  I have only had perhaps half a
dozen heads break off when removing the #3-48 screws, out of literally
hundreds.  Each of these were corroded by the "aluminum to nickel plated
brass battery" that gets activated in the usual damp basement haunts of
these sets, and required a pretty good torque to get moving.  Which they
did catastrophically...  :-)  However - I suspected this was more of a
symptom than an explanation...so I took a look at both NOS and well used
screws under the microscope, and what I noticed is that the radius of
the transition to the threaded portion under the head looks like it's
smaller than I would consider prudent by using the stress tables in
Machinery's Handbook, though that's a value judgment to be sure.  It is
certainly smaller than that of the stainless steel binding head screws I
sometimes replace them with.  In other words, the automatic screw
machines that spit these out by the thousands apparently used cutting
tools with a sharp point on the side that forms the head to body
transition.  That kind of right angle transition inevitably leads to
stress cracks at the boundary, exacerbated by the tension of the screw
after tightening, in-service vibration, heat cycling, and yes,
subsequent corrosion among other factors.  Some metals are more
sensitive to this than others.  Brass isn't as bad as aluminum for
example, but like any metal it does have that stress cracking
potential.  I don't know of any other rational explanation for the
consistent failure location, other than the plain over-torquing you
mentioned.

73,
Mike

> -----Original Message-----
> From: COWARD,JAY (A-SanJose,ex1) <jay_coward@agilent.com>
> To: 'arc5@mailman.qth.com' <arc5@mailman.qth.com>
> Sent: Fri Mar 08 20:54:53 2002
>  It seems that most of us have run across the screw heads popping off with
> minimal torque applied.I think that expansion coefficints may play a part in
> this failure mode.