[TowerTalk] Octahedron Towers

[Jim Lux]

At 03:53 PM 11/22/2004 -0800, Jim & Velma Woods wrote:
>I am new to the tower-talk group but have interest in building an unusual 
>tower for a small HF beam or quad.  It is based on a design from the 80's 
>where 1-inch EMT conduit is used to build sections of the tower in a 
>series of octahedrons.  Height would be about 44 ft.  Plans are sketchy 
>and engineering analysis appears to be almost non-existent.  The towers 
>were used to support wind generators.  Most wind generators have quite a 
>bit of wind loading compared with ham antennas.
>
>Any initial comments?
>
>73,
>
>Jim Woods, W7PUP
>woods at grantspass.com
>Grants Pass, OR

I've fooled with using EMT as a building material quite a bit over the 
years. (see, for instance, http://www.luxfamily.com/jimlux/robot/emtsphere.htm)

In general, EMT is terrible as a structural material.  It cracks on the 
welded seam in particular.

However, one could build something out of EMT and then proof test it to 
several times the expected loads.  That is, bolt the base of the tower to 
something sturdy so that it's hanging out like a canteliever beam a few 
feet off the ground.  Start hanging weights on the structure to replicate 
the design wind loads.  If you get to 2 or 3 times the load and the 
structure survives without permanent deformation you're in pretty good 
shape. It will deflect under load, but you want to make sure it comes back 
to where it was before you loaded it.  Use a tape measure! Don't estimate 
the deflection by eye.

Make sure you do this in several orientations.

Some other things to think about:
Fatigue failure.  Fortunately, you'll probably have the structure 
overdesigned enough (because of the crummy material properties for EMT) 
that you'll never have enough load on it so that fatigue is a 
problem.  However, if you haven't analyzed it, you don't know.

Buckling. In a structure like this where the materials are sort of iffy, 
you can't really be sure how the stresses are distributed.  You might be 
perilously close to buckling some member.  Proof testing helps, but, still, 
if you didn't happen to load it just the right way.

I'd be interested in seeing more about this approach, for a variety of 
reasons.

Maybe someone should get bold and build a tensigrity tower.

Jim Lux
W6RMK