Nigel Pearson writes:
Question about using safety wired bolts in high-stress applications.
People go to the trouble of polishing connecting rods to eliminate
sharp edges which may cause stress cracks.
Wouldn't drilling holes in hardened or forged bold heads do exactly
the opposite? (i.e. make the bolt more likely to fail)
Surface flaws in stressed regions of components lead to failure because the
flaw causes
locally increased stresses. When these stresses are cyclic (as in connecting
rods) fatigue
failures become more likely. A fatigue crack initiates at the flaw (or at
thread root in bolt)
and grows under cyclic loading to become large enough that the remaining
material
cannot carry the applied loads. Connecting rods are sometimes shot peened to
condition
the surface (by providing residual compression stresses which help reduce the
fatigue
problem), or polishing the surface (to remove flaws such as scratches,
machining marks,
or micro-cracks).
There is a significant difference holes in bolt heads and flaws in connecting
rods. Since
the heads of bolts are not significantly stressed in service, holes drilled
in the head, or
other flaws in the head, do not cause this type of problems.
Incidentally, fatigue of bolts is significantly (or totally) alleviated by
proper preloading
(torquing) the bolts, which is an idea getting much attention in the last 2-3
mailings (the first
ones I've read, as I'm new to this group). This concept, of eliminating cyclic
(fatigue)
stresses from bolts by proper preloading, is always hard to communicate to the
engineering students I teach. Basically, the superimposed static preload, even
though
causing much higher stresses than the cyclic stresses which cause fatigue, is
much less
severe for the bolt. A loose head bolt (or stud) can fail in fatigue while a
properly
preloaded (torqued) bolt will not fail, if designed properly. I always use a
torque wrench
when the bolt is critical.
-------------------------------------
Name: Ray W. James
E-mail: Ray James <rwj4123@sigma.tamu.edu>
Date: 05/27/94
Time: 10:00:51
-------------------------------------
|