[Top] [All Lists]

[TowerTalk] Tower Falling Radius (long)

To: <>
Subject: [TowerTalk] Tower Falling Radius (long)
From: (Steve Sawyers n0yvy)
Date: Fri, 04 Apr 1997 22:26:08 -0600
What we are sort of talking around is a Failure Mode and Effects
Analysis. (This is also called the systematic application of Murphy's
Law.) This is where you look at every part and do a "what if" analysis.
Things like what if this breaks, or this is left out, or this particular
form of overload is applied. Generally you have to look for the "weakest
link in the chain" to see what breaks first, after it breaks, what
happens next and so on. Different "what ifs" give different effects. 

This gets especially hard under complex loading situations like long
flexible members in compression, supported laterally with less than
rigid supports. (Fancy definition for what a tall tower looks like to a
structual engineer).

I agree that tall towers (600 feet or greater) generally fail due to
initial buckling of one or more legs. Primary causes are someone working
on the tower and damaging or improperly removing a lateral, vertical ice
load, or unbalanced ice load which moves the effective center of the
compressive force away from the center of the tower. Granted, sometimes
high wind will complicate the ice loading and sometimes a straight wind
will overload a tower or tornadic wind will twist a tower and start the
collapse. But in most of the cases I have seen or read about, the
lateral wind was less than half the design wind. Something else was the
major contributor to the initial failure that brought the tower down.

As a general case for all towers, I cannot say that they will all fail
in a localized pile unless they are designed and built to do so. The
question is what is the Failure Mode. 

If the Failure Mode is compressive collapse of the tower sections, then
the falling radius is small. If the Failure Mode is a guy wire failure,
then two things can happen. On taller towers, the tower can start to
move lateraly until it either stabilizes, or some localized tower member
and/or one or more legs reach their limit stress which initiates the
vertical collapse. If the tower is shorter and/or more rigid, there is a
possiblity that a guyed tower will fall straight out.

Tall towers (600 feet an above) have multiple guy levels and are really
rather flexible. The towers sections are usually handling their maximum
compressive load. 

I agree with Rod W5HVV when he said:
"Isn't it entirely feasible (and even probable) that the tower would
then essentially collapse straight or nearly straight down?  Why would
the tower, given that form of failure parameter, "lean over" and crash
some distance away when there were all those upper level guys to keep it
from doing so until after the momentum was essentially straight down?"

There is so little force to start the tower moving laterally as compared
to the huge inertia of the tower and even the guy wires, that it is not
probable to have tall towers fall straight out.

On shorter and relatively stiffer towers, say Rohn 25 at 70 feet, the
situation is different. It is loaded at about 25% of the maximum
compressive section capabilities. (See Note 1 below). In this case if
you had a failure of the guys from one side, the tower would have enough
inherent strength to maintain its shape until it hit the ground. In this
case, straight out. I have seen it happen where a tree took out a set
guys and the tower fell over straight as a string. In this case the
tower was stronger than the guys for the applied load. (Yes, I know that
you shouldn't have to design for trees falling on your guylines - or
should you? Most towers that I know of will take more wind or ice load
than most trees that I know of.)

Now that we have outlined the primary Failure Modes, what are the

That depends on what it falls on, and how much damage it does when it

In the case of the failed Rohn 25, it fell across a couple of back yards
onto a standard 4 foot high chain link fence. It did not do that much
damage as the tower buckled over the fence. In this case the weight per
unit length of the objects coming down was less than a 4" diameter tree

On the other hand,if this had been a section of Rohn 65 and it had
landed on a house, then it would have done lots of damage.

In my case, I am planning 70 foot of Rohn 25 next to my daughter's bed
room, my preferred failure mode on this tower is to fall straight out
rather than collapse straight down into her bedroom. The guys will get
designed to meet the wind and ice load for the installation, but the
tower and its appertances will by sized to assure that the weak link is
the guying, and not the column buckling of tower sections.

For the 300 lbs. per section 114 foot tower that is going in 60 feet
away, I want that one to fail straight down, no matter what the loading
conditions. Thus I am going with double 5/16" top guys and 3/8" second
and third level guys and huge concrete guy anchors. The column buckling
will be the weak link on that one.

For the 150 lbs per section 80 foot tower that is 100 feet from the
house, I will go with a pretty much standard approach where I will
balance the guy strength and the tower column strength to get the
maximum wind load up.

IMHO a Failure Modes and Effects Analysis is very neccessary. It lets me
sleep a lot better.

Sorry this was so long. I hope it was educational and gives you some
other insights into tower installations.

de n0yvy steve
Steven H. Sawyers PE
ARRL Volunteer Consulting Engineer

Note 1:
The tower base reaction from the Rohn literature for a 70 tower is only
3010 lbs in a 90 mph wind zone with no ice. In this case the bottom
tower section is not loaded anywhere near it's 12,000 lb maximum
compressive stress (this theoritical load is only applicable if the load
is exactly centered on a section and there are no external bending
moments). I support  of the assertsion that 3010 is a small tower
section load, I woul dlike to point out that Rohn does show a 190 foot
tower having a load on the bottom section of 9870 lbs in the same wind
conditions with 31 foot guy spacing. Dwg Ref C870488 R1 from my latest
Rohn Consumer Products Catalog.

FAQ on WWW:     
Administrative requests:

<Prev in Thread] Current Thread [Next in Thread>