[TowerTalk] Tower ACCIDENT, engineering

[Jim Lux]

At 01:51 PM 6/27/2007, Daron J. Wilson wrote:

>The entire tower structure design takes into account the guy tension,
>downward pressure, wind loading, bending, flexing, etc.  ANYTIME you change
>those parameters, you change the structure, and if you can do the
>calculations then you can demonstrate the 'over engineering'.  However,
>simply using bigger parts hardly qualifies.  Why not drill out the holes and
>use 1/2" bolts?  They're bigger, heavier duty, must make it stronger huh?

Indeed. There's a classic exercise in a structures class where you 
analyze adding a welded gusset to a "T" joint (where one member joins 
another member at right angles, in the middle)(think of a ladder 
rung), results in a structure that fails much sooner, because the 
stiffness of the gusseted joint increases the load on the member.



> > I over engineer my installs, within reason, based upon personal
> > conversations I had with Rohn back in the days before litigation became a
> > national sport. Ive never had a bit of static from any inspector or PE. A
> > well known fact among serious installers is that 25G is actually designed
> > at
> > 2.5 times the specs since they knew that hams would always push the specs
> > over the limit.
>
>Simply increasing the diameter of one of the components is hardly 'over
>engineering' unless you do the required engineering to the rest of the
>structure!  Using your theory, I bet I could 'over engineer' a tower with
>larger guy wires and turnbuckles until it buckled under its own weight!

Are you sure that this 2.5 times the specs isn't just a "factor of 
safety" to allow for variability in material properties and assembly 
workmanship?  That is, according to the calculations, if everything 
was exactly as expected and assembled perfectly, it would have a 
failure load of 2.5*x, where x is the design load.

One would use such a factor also when you make some simplifying 
assumptions in the analysis, with the extra margin there to cover the 
possibility that the assumption went "the wrong way".


It's sort of like why houses are wired with AWG14 wire, even though 
the fusing current for that gauge is >100 Amps. (there's also a 
voltage drop issue, too, I admit).