[TowerTalk] Guying a self-supporting tower

[Jim Lux]

----- Original Message -----
From: "Doug Renwick" <ve5ra at sasktel.net>
To: <towertalk at contesting.com>
Sent: Saturday, April 09, 2005 1:49 PM
Subject: Re: [TowerTalk] Guying a self-supporting tower


> We are finally getting a technical discussion on this topic.
>
> K7LXC at aol.com wrote:
>
> >     The big problem with guying self-supporting towers is that the guys
have
> > the potential to increase the leg compression preload when properly
tensioned.
> > And as the leg capacity is the primary factor for tower strength, adding
> > additonal compressive force can lead to failure when a big windstorm
blows through
> > and the leg strength is exceeded.
> >
> Certainly leg compression will be increased with properly tensioned guys.
> I have believed that tower failure (guys remain intact) will be caused by
> torsion or twisting action on the tower, not by compression failure.
Someone
> with actual failure experience should comment on this belief.



> >     Delhi towers are sort of a special case. Yes, they're the classic BX
> > design and include straight sections along with the tapered ones and,
yes, the
> > manufacturer does say "Guy wires must be used for larger loads or if
straight
> > sections are added".
>
> I do not believe that Delhi towers are a special case...they have cross
arms
> and three legs and can be tapered or straight like so many other designs.
> They are not heavy weight sections, but they will take a lot of
compressive
> forces on the legs.

Someone who has the engineering analysis for a tower can take a look, but in
my recollection (it was about 8-9 years ago that I last checked this sort of
thing), you need to look at the loads in three kinds of members:

The longitudinal forces in the legs.
The forces in the diagonal cross braces
The forces in the horizontal members between the legs.


The usual truss design analysis (and antennas are  a form of this) treats
the thing as a series of smaller short fat units (i.e. the "repeat length"
along the truss... typically a few feet).  The loads and deflections on the
members are all considered as straight linear compression/tension loads,
and, I think, with the short column assumption, within the "repeat cell".
You calculate all the parameters for that "chunk of truss", and then analyze
the complete structure as if it were a bunch of those chunks stuck together.

So, the real question is what are the relative loads in the various members
for compression, bending, and torsion. Of particular interest would be if
any of the loads tend to result in a deflection in a different direction.
Think of a coil spring... when you squeeze it down, it tends to twist.  Most
trusses alternate the "hand" of the diagonal braces for this reason, so that
the torsional effects cancel out.


It may well be that while the vertical tubes/legs are the same size in all
the towers (subject to the manufacturer's preferences), but that the cross
and diagonal braces are a different diameter/section modulus.

The result would be that a given segment of the tower may have different
stiffness, even though the failure load is the same.  That change in
stiffness could have a very different effect in a freestanding and guyed
configurations.

Consider this as an example.  You have a piece of 1" PVC pipe as an
unsupported vertical mast.  It's pretty flexible, but will actually take a
fair amount of side load before failure (flexing a bunch in the process).
If it fails, it will probably fail at the base (because that's where the
loads are highest).  This is the standard canteliever beam thing.


 Now guy it.  You've increased the total down force on the pipe, and
constrained the position of the top of the column, changing it from a
fixed/free configuration to a fixed/pin  (or maybe a pin/pin) configuration.
Now the failure might be more likely in the center (the buckling failure).
You might want a much stiffer tower in this case, to reduce the amount of
bend from the axial load. When you put a side load on the tower, you do two
things.. one is that you increase the total down force  because of the
increase in the guy tension (unlike in the free standing, where the total
down force remains the same), the other is that the wind load tends to bend
the middle of the tower away from the wind (the top can't move, because it's
constrained by the guys).  Both of these increase the chance of buckling.

So, it seems that for a guyed mast, you really, really care about how stiff
the mast is, because that determines how much deflection you'll get for a
given axial load, and that ties directly to what the failure load is going
to be..  For the free standing, you don't really care (other than for
dynamics reasons and aesthetics), as long as it's strong enough.

>
> >     No other self-supporting tower or manufacturer makes that kind of a
> > statement or suggests guying the tower.
>
> Can you provide the self supporting tower manufacturers that state their
towers
> are not to be guyed?  And can you provide supporting references?  And if
there
> are some that do say 'do not guy', I would be interested to find out why
based
> on engineering reasons.

What a manufacturer will put on paper tends to be driven by concerns other
than straight engineering analysis.


>
> > They're designed to be self-supporting and guying them goes against
their
>  > engineering and design.
>
> Explain please how guying these towers 'goes against' their design and
> engineering.  It's obvious that they are designed and engineered to be
> self-supporting.  How does guying degrade their design?

By changing a canteliever beam designed for straight bending into an axially
loaded column, which has very different structural properties.

Whether it degrades or not probably depends on the details.


Jim, w6rmk