
Copyright © 20002004 Kurt Andress, K7NV All Rights Reserved
Updated April 29, 2001
A study of what happens when we make changes to a tower configuration
Preface
The following discussion is aimed at observing
tower behavior subsequent to changes in configuration, and has selected the
EIA/TIA RS222F spec as the standard for the comparisons. The information
presented is just part of the work pursued here to understand how guyed towers
work. I found the results interesting, and contrary to some amateur folklore,
making it a shame not to share with others.
The discussion is not intended to address
whether or not the EIA spec is a going to tell how much you can really get
away with at your own private installation. The EIA/TIA spec exists on its
own merit. Many have installed towers that have loads exceeding the EIA
spec, and they have survived. There are those who have not been so fortunate!
The main problem is that very few know what the real conditions were at
the many sites that were overloaded! This study is intended to establish
how a guyed tower system reacts to a variety of configuration changes.
Methodology
All of the towers were modeled with finite element analysis (FEA) software. The FEA approach to analyzing mechanical structures is very much like the MiniNEC and NEC codes. The models are made up of discreet segments called "Finite Elements." The program looks at each individual segment in the model with the loads applied and computes how it behaves under its loads and how it interacts with the others under their loads.
The FEA model is constructed in a modeling module with the elements containing the appropriate properties to represent the type of tower and guys used. The FEA models used in this exercise are the simplest models that one can make. They are called "stick models." They are simply a collection of line elements from one point to another. Each element has the properties of the section (tower or guy) that it represents.
Tower geometry, guy sizes, feedlines, antenna projected areas, and basic wind speed were entered into two of 5 linked spreadsheets to produce the input loads to be applied to the towers in the FEA models. All loads applied to the towers were calculated in accordance with ANSI/EIA RS222F.
The tower element properties were calculated from the information in the Rohn drawings with the cross sectional area adjusted to cause the correct tower weight to be developed in the analysis via the application of 1G acceleration to the model. This requred the extraction of results in the form of momernt, shear, and axial loads, instead of viewing the FEA stress results, because the tower cross sectional area was not correct for stress determination. This proved less work than the alternative of calculating and inputing all of the tower section weights into the model.
The guy properties used the formulas and data
offered by the Macwhyte Wire Rope Company, Kenosha, WI. to find a
solid diameter for the cables that would accurately provide stretch behavior
with an elastic modulus of 29 Msi. One should view the information available
at the Guy Cable Link on this site, to understand how the guy properties
were derived.
Then the calculated loads are applied to the
model in a load module. The analysis is run and the results are obtained in
a post processing module.
The post processor provides bending moments
in two axes, shear in two axes, axial loads, and deflections at the ends of
each element in the model.
This data is entered into another spreadsheet
to calculate the combined tower stresses and safety factors along with the
guy safety factors.
The FEA software is a linear finite element
code called GBEAM. It is quite nice for software that resides at the very
affordable end of the spectrum. The shareware version is quite adequate for
most simple problems.
Those interested in obtaining more information
can go to http://www.grapesoftware.mb.ca
There is another website that has an excellent
listing of public domain analytical software.
Those who are interested in this can go to
http://www.engr.usask.ca/%7Emacphed/finite/fe_resources/node74.html
Linear FEA code is ok for analyzing structures
that do not experience large deflections. One needs to use nonlinear software
when the deflections are large.
Most of the tower deflections in this study
are close to the width of the tower face, so the analyses are expected to
be acceptable. Questionable models are noted.
Comments to aid readers that are not familiar
with this subject:
The following analytical results contain
the following values.
Guy loads are simply in Lbs of load, compared
to the guy cable breaking strength to determine the safety factor.
Combined tower stress is the sum of stress
due to bending, stress due to shear, and stress due to axial compression.
Bending stress is caused by guy stretch that allows the tower to lean
over, Shear stress is caused by the horizontal windloads on the tower and
antennas, and the axial compression is caused by the guy cables resisting
the wind loads and their angles from the tower or ground. As each guy resists
the wind loads it produces vertical loads in the tower based on its angle.
The tower displacement values are simply how
far each point on the tower moves horizontally under the wind loading.
The EIA/TIA basic wind speed is NOT a peak reading on a wind gauge (anemometer). It is the average wind speed for one mile of wind passing over the structure. This is expected to be lower than the peak readings on anemometers installed at the site.
The base of the towers and guy anchors can
be modeled with different constraints at each of the ground connections.
Constraints in Fx Fy Fz are restraints that prevent movement along the X Y
& Z axes. Constraints Mx My Mz are restraints that prevent rotation about
the X Y & Z axes. The guy anchors in all models used Fx Fy Fz constraints.
Tower base constraints are given for each model.
The following values were used for determining
the safety factors:
Allowable combined tower stress = 50,000 psi,
allowing for a 1/3 increase due to combining of the fundamental stress
development modes, per the Rohn drawing, et. al.
Guy breaking strength for 3/16 EHS = 4000
Lbs
Guy breaking strength for 1/4" EHS = 6700
Lbs.
The Guyed Tower Study
The Baseline Model:
I selected a tower configuration from the
"Rohn Ham Tower Catalogue", May 1998. The tower designs in it are intended
to conform to the ANSI/EIA RS222E specification.
The baseline model is the 100' configuration
for a 90 Mph basic wind speed. Refer to Rohn drawing C870478 R1, P/N 45G90D100.
The tower configuration is the same as the
one offered in my new, 1999 commercial catalogue, obtained from Champion Radio.
This is a 100' tower with 3/16 EHS guys at
31' & 61', and 1/4" EHS guys at 91'.
The listed antenna projected area at the top
of tower is 13.5 SqFt + 6.7 SqFt (see note 4) = 20.2 SqFt.
The area listed for the side arm mounts in
note 4 is the effective area. When this is corrected for round member projected
area (as in note 2) the additional area for antennas, when not using the
side mounts, becomes 6.7 SqFt.
The tower also has 3 each 1/2" Dia and 3 each
7/8" Dia feedlines running from the base to top of the tower.
The guy loads include pretensioning of 400
Lb for the 3/16" cable and 600 Lbs for the 1/4" cable.
The Rohn drawing does not state the weight
of the lumped antenna load at the top of the tower, so I selected a weight
of 250 Lbs for use in these models.
Tower Model Showing Applied Loads and
Constraints
Results: Model 45G100B Load Case #1  Wind @ 30 Degrees to an upwind guy set.
Tower base constraints  Fx Fy Fz Mx My Mz
representing tower buried in a concrete footing.
Combined stresses taken at the bottom end of each span.















Guy Loads












Tower Displacements along the upwind
guyset axis @ 10x Scale
Tower Displacements










Results: Model 45G100C Load Case #2  Wind @ 60 Degrees to guy set (Directly between two sets of guys).
Combined stresses taken at the bottom end of each span.















Guy Loads












Tower Displacements along either guyset
@ 10x Scale
Tower Displacements










Comments:
This analysis shows that all safety factors
are > 1.0, so the tower is considered safe. This indicates that the methodology
and analytical results more or less agree with that represented by the Rohn
recommendations.
The tower safety factors are highest with
the wind closer to a set of guys, the guys work the hardest and hence, have
their lowest safety factors.
The lowest tower safety factors occur when
the wind is directly between two sets of guys. Here the guy safety factors
are highest because they share the loads with another guy set. The reason
the tower stresses increase is because of guy geometry. The guys anchors are
their farthest from the direction of the wind. Eventhough, the loads get
shared equally by two sets of guys, the poor angles result in higher tower
deflections that result in higher bending stress. Additionally, the fact
that two adjacent guys share the loads, results in an increased in tower
compression.
A very interesting observation in this model
is that 71% of the tower stress at the base is caused by bending,
27% due to compression , and 2% due to shear.
The same general distribution was observed
in all other models with fixed bases.
We can say that tower bending is the limiting
factor in this design.
For an optimized design, we would expect to
see similar safety factors for all tower sections and all guys. The guy
and tower safety factors might not be the same, due to other considerations.
But, for each unique tower feature we would expect all margins to be fairly
closely grouped.
This is clearly not an optimized design, simply
one that will work.
Configuration Variant #1:
Let's take the same 20.2 SqFt of projected
antenna area and spread it across different vertical locations on the tower.
I divided the antenna projected area and weight into 3 equal parts and placed
them at 100', 61' and 31' leaving all other tower properties and geometries
the same.
The feedline runs were changed to have 2 each
of the original 6 lines run up to each of the 3 antenna locations.
Results: Model 45G100F Load Case #1  Wind @ 30 Degrees to an upwind guy set.
Tower base constraints  Fx Fy Fz Mx My Mz representing tower buried in a concrete footing.
Combined stress taken at the bottom end of each span.















Guy Loads












Tower Displacements along the upwind
guyset axis @ 10x Scale
Tower displacements:










Results: Model 45G100G Load Case #2  Wind @ 60 Degrees to guy set (Directly between two sets of guys).
Combined stress taken at the bottom end of each span.















Guy Loads












Tower Displacements along either guyset
@ 10x Scale
Tower Displacements










Comments:
This change was expected to increase the safey
factors from the baseline model. The safety factors increased everywhere
except at the tower base, where they decreased.
After reviewing the components of the combined
tower stresses it became apparent that moving the antenna point loads to
lower levels resulted in more guy stretch in the lower 2 sets of 3/16" cables,
resulting in higher bending stresses at the tower base.
This would indicate that the original Rohn
tower configuration was specific to the original load case, I.E. all the load
at the top.
It also confirms the Rohn position, stated
by K7LXC, in a Towertalk post, that the rated antenna area for the tower applies
to all methods of area distribution along its length.
Looking at the tower deflections, it appears
that the problem is that the new lower antenna loads have created higher
elongation in the lower guy sets, which result in increased deflection in
the lower sections.
Configuration Variant #2:
Since the tower stresses seem to be driven
by the stretch in the guy cables caused by placing antennas lower on the
tower, let's see what happens if we increase the lower cable sizes to 1/4"
EHS. All guy cables in these models are 1/4" EHS. Antenna loads are the same
as the previous example.
Results: Model 45G100F2 Load Case #1  Wind @ 30 Degrees to an upwind guy set.
Tower base constraints  Fx Fy Fz Mx My Mz representing tower buried in a concrete footing.
Combined stress taken at the bottom end of each span.















Guy Loads












Tower Displacements along the upwind
guyset axis @ 10x Scale
Tower Displacements










Results: Model 45G100G2 Load Case #2  Wind @ 60 Degrees to guy set (Directly between two sets of guys).
Combined stress taken at the bottom end of each span.















Guy Loads












Tower Displacements along either guyset
@ 10x Scale
Tower Displacements










Comments:
This configuration results in an increase
in tower safety factors, which now allows us to increase antenna areas to
get back to the original baseline safety factors.
Notice that, in this configuration,
the tower sections are much more aligned on a straight plane. The base stress
has been reduced by 22%.
Since the antenna loads were moved down the
tower, we had to increase the guy sizes to reduce the guy stretch to keep
the tower bending under control.
Again, nearly 70% of the combined stress in
the tower base is caused by bending, suggesting that guy cable stretch is
still the predominant factor!
The safety factors for ultimate guy loads
are just fine. But, the cable stretch is still allowing the tower to lean
over and produce the high base tower stresses that limit the entire installation.
This comparison should shed some light on
the continuing controvery over whether it is "okay" to put larger guys on
a tower. If the tower is similar to our example (limited by guy elongation
and tower bending stress), larger guys will always make the situation better.
The increase in tower compression caused by the larger guys and their weight
and pretension was insignificant compared to their ability to reduce the
tower deflection.
Increasing guy size only become a problem
with the tallest configurations that have relatively low antenna loads,
and the compression from the guys, reacting the extended tower sections,
becomes the predominant stress development factor.
Configuration Variant #3:
In the tower load calculations I noticed that
the number and size of the feedlines running up the tower seemed to significantly
affect the tower loads. Let's see what an extreme case can tell us about
this.
I took the baseline Rohn model with 3  1/2"
& 3  7/8" feedlines running to the top of the tower and replaced them
with just 1  1/2" feedline running to tower top to feed just one big antenna
(like an 80 meter beam) . Still 20.2 SqFt and 250 Lbs up there.
Results: Model 45G100B1 Load Case #1  Wind @ 30 Degrees to an upwind guy set.
Tower base constraints  Fx Fy Fz Mx My Mz representing tower buried in a concrete footing.
Combined stress taken at the bottom end of each span.















Guy Loads












Tower Displacements along the upwind
guyset axis @ 10x Scale
Tower Displacements










Results: Model 45G100C1 Load Case #2  Wind @ 60 Degrees to guy set (Directly between two sets of guys).
Combined stress taken at the bottom end of each span.















Guy Loads












Tower Displacements along either guyset
@ 10x Scale
Tower Displacements










Comments:
These models should be compared to 45G100B
& 45G100C.
This shows that the number and size of the
feedlines has a fairly significant affect on the tower loads.
The elimination of 5 feedlines resulted in
a reduction in tower stress at the base to 69% of the original configuration.
Configuration Variant #4:
Since, the limiting factor for the tower,
all along, had been base bending stress, I thought it would be interesting
to see what improvement would result by taking the tower base out of the
concrete and setting it on a free rotating base.
This configuration is the same as the original
"factory" model, except the tower has been taken out of the concrete filled
hole and mounted on a pier that has the equivalent of a ball socket
mount that allows the tower base to freely rotate.
Results: Model 45G100B2 Load Case #1  Wind @ 30 Degrees to an upwind guy set.
Tower base constraints  Fx Fy Fz representing tower base mounted on a free rotating connection..
Combined stress taken at the bottom end of each span.















Guy Loads












Tower Displacements along the upwind
guyset axis @ 10x Scale
Tower Displacements










Results: Model 45G100C2 Load Case #2  Wind @ 60 Degrees to guy set (Directly between two sets of guys).
Combined stress taken at the bottom end of each span.















Guy Loads












Tower Displacements along either guyset
@ 10x Scale
Tower Displacements










Comments:
This model should be compared to 45G100B &
45G100C.
This modification changes the entire problem!
The change eliminates the bending stress at
the base, which was responsible for approx 70% of the stress. The new limiting
factor for the design is the top tower section stress, where it is cantilevered
above the top guy. The top guy load is running a close second place.
The two lower sets of guys are working harder
so their safety factors went down a bit but are not a problem. Because the
guys are carrying higher loads, and no change was made in their elongation
behavior, the tower deflections are higher. The limiting load case
changes from load case #2 to case #1.
On a free rotating base the tower really doesn't
care how far it leans over, as long as it stays in column.
Overall, it is an improvement and the tower
would support additional antenna loads.
Here is what happens if we add another 10 SqFt of antenna to the top of this tower
Just for giggles!
Results: Model 45G100B4 Load Case #1  Wind @ 30 Degrees to an upwind guy set.
Tower base constraints  Fx Fy Fz representing tower base mounted on a free rotating connection.
Combined stress taken at the bottom end of each span.



























Tower Displacements along the upwind
guyset axis @ 10x Scale
Tower Displacements










Results: Load Case #2  Wind @ 60 Degrees to guy set (Directly between two sets of guys).
This model was not analyzed as it became obvious
that this load case was no longer limiting the tower design..
Comments:
Putting the tower on a free base connection
has allowed us to increase the antenna loads by 49%.
I get kind of excited when I can find a single
configuration change that results in that much improvement!!
This tower would be helped by a larger set
of top guys to get it back in column.
Configuration Variant #5:
Phillystran aramid cable is a popular material
for guying towers to eliminate electrical interaction that can degrade antenna
performance.
If we were to select an aramid cable to replace
the 1/4" EHS in the top set of guys, based on breaking strength, we
would chose HPTG6700.
Let's put some HPTG6700 Phillystran on the
baseline tower (buried in the footing) and also use it to replace the 3/16"
EHS in the two lower sets, and see what happens.
Results: Model 45G100B5 Load Case #1  Wind @ 30 Degrees to an upwind guy set.
Tower base constraints  Fx Fy Fz Mx My Mz representing tower buried in a concrete footing.
Combined stress taken at the bottom end of each span.















Guy Loads












Tower Displacements along the upwind
guyset axis @ 10x Scale
Tower Displacements










Results: Model 45G100C3 Load Case #2  Wind @ 60 Degrees to guy set (Directly between two sets of guys).
Combined stress taken at the bottom end of each span.















Guy Loads












Tower Displacements along either guyset
@ 5x Scale
Tower Displacements










Comments:
This model should be compared to 45G100B &
45G100C
Yikes! This tower fell down (according to the EIA design spec)! When the safety factor goes below 1.00 a failure has occured.
Aramid (TM Kevlar) is very strong, but it
is not as stiff as steel. The aramid guys are strong enough to carry these
loads. Looking at the safety margins for breaking strength, we would expect
it to be okay!
They just elongate too much, causing the combined
stress at the tower base to exceed the allowable.
I'm not really sure the linear FEA analysis is providing real accurate values at this point, as the deflections have gotten out of hand.
Information to help you understand the various
guy cable elongations can be found at the "Guy Cable"
link on this site.
Let's free up its base and see what happens
Same model as the one above, except the base has been taken out of the hole and placed on a free rotating connection.
Results: Model 45G100B6 Load Case #1  Wind @ 30 Degrees to an upwind guy set.
Tower base constraints  Fx Fy Fz representing tower base mounted on a free rotating connection.
Combined stress taken at the bottom end of each span.















Guy Loads












Tower Displacements along the upwind
guyset axis @ 10x Scale
Tower Displacements










Results: Model 45G100C4 Load Case #2  Wind @ 60 Degrees to guy set (Directly between two sets of guys).
Combined stress taken at the bottom end of each span.















Guy Loads












Tower Displacements along either guyset
@ 5x Scale
Tower Displacements










Comments:
Well, freeing up the base connection again
improved the situation. It got the safety factors back up above 1.0, so the
thing should still be standing.
This is still not a very happy tower, but
it is happier than it was with the fixed base.
The top guys set is way too small for this
material. This is obvious from looking at the tower displacements and the
deflection plots in the FEA program.
My guess is that using HPTG 8000 or 11200
on the upper guys would help get the tower back in column.
Hopefully, this example will make it clear
that, when replacing EHS steel guys with nonconducting synthetic materials,
we should primarily consider elongation.
Strength is important., but it is not always
what keeps our towers standing!
General Discussion
Redistributing the antenna loads along the tower:
Splitting up the rated area to several locations
on the tower can be done. It can yield an ability to increase the total
antenna area if the guy sizes are adjusted according to the load redistribution.
If the guy sizes are not adjusted (for the catalogue designs), the total
rated antenna area remains essentially the same..
As with so many other things, there is really
only one best solution for any given set of conditions.
Tower Bases:
Guyed tower bases buried in the footing suffer from a system limiting feature that is caused by guy stretch. The elongation of the cable allows the tower to lean over enough to create large base bending stresses. Freeing up the tower base connection to rotate, eliminates this problem. Making this change increases the guy loads, but does not cause them to exceed acceptable safety margins.
For many years I've seen hundreds of commercial
installations with tapered bases. Until now, I always wondered why the commercial
installations had tapered bases and the amateur ones did not.
After looking at my new Rohn commercial catalogue,
the first thing I noticed is that all sections for guyed towers from 45G
and up are offered with tapered bases. One of the rigid tube towers "J" offers
a roof mounting base that actually has a ball and socket connection. This
is exactly what was modeled in Configuration Variant #4.
This indicates that Rohn and their commercial users understand the problem, and agree that the base of the tower should not only be free to rotate about the vertical axis to reduce torsional load development, but also be free to lean over just as far as the guys will allow. The small tapered bases sit on bearing plates on a pier pin. It is obvious from the drawings that these connections will allow enough rotation, about the horizontal axes, to prevent development of significant bending loads.
The problem with the common pier pin bases for the 25G & 45G is that they are full size sections at the tower/pier interface. The wide footprint on the pier may allow the tower to rotate about the vertical axis, but resists the ability to lean over without developing significant bending loads. In the load case #2 scenario (wind between guys), if the base was to be able to rotate far enough to eliminate the bending loads, it would be sitting on only one leg. It is unlikely that this can occur with the tower compression pushing the base against the pier. If it actually did occur, the poor downwind leg would be carrying all the tower compression.
Although not commonly used in amateur service, Rohn has a tapered base section for 45G. Ask for P/N 45TG. There are also two other sections, 45TGIA & 45TGIA47, which are made to mount on a base insulator. These are used frequently in broadcast installations. I cannot find an equivalent base section for 25G.
There has been significant discussion on amateur
sites about the pros and cons of fixed vs pier pin tower bases. Most of these
discussions are concerned with the full sized sections at the tower/pier
interface.
During my review of the discussions, I have
found that the amateur community seems to be equally divided over which
type is best. Much of the discussion centers around how stable the tower
feels when climbing it, the fixed base being preferred.
In spite of our varying personal preferences, on the issue, we should consider the fact that most commercial users want maximum efficiency from their tower installation, and the people who pay for the towers don't have to climb them. Most of these installations use tapered pier pin bases.
The importance of guy elongation on the system
is not an original idea on my part. The EIA 222F specification states in
paragraph 3.1.2 "For guyed structures, the displacement of the mast (this
means tower to us) at each guy level shall be considered when computing
stresses." This acknowledges that lateral tower displacements
do occur and that they contribute to stress development. I have simply presented
information about where they come from, and what can be done to mitigate
the problem.
Aramid Guy cables:
These are very strong and lighter than the
steel cables. They exhibit higher elongation than comparable strength steel
guys.
Since, some guyed towers can be sensitive
to cable elongation, replacement of steel guys with aramid cannot be only
done based on strength alone.
Guy elongation must be considered.
SEE the Guy Cable Link on this site for more information.
Final
Comments
The preceding information is not the final
word on the subject. It is simply a presentable piece of the work I have
done, subject to review.
There are likely several other issues to be
explored. The exercise is entirely incomplete! And likely to remain so.
Comments from others working in this area are welcome.
The preceding information is a just a study of tower behavior, using a widely recognized standard for approaching the problem.
It may have little to do with what actually happens at your site on any given day, or what you have observed you can get away with when you are not experiencing a 50 year wind event.
I believe the relative behavior presented to be accurate. And, would expect it to be repeatable under any other standard.
The anecdotal experiences of many may not agree with what I have presented, but most of those do not come with enough information to make any meaningful evaluations. They can, however, be an entertaining source for years of conjecture and discussion!
I have evaluated a couple of real life amateur
installations, equiped with anemometers, that would defy "the general notion
of what might be acceptable", according to the book designs, and found that
the same methodology presented above showed them to be as sound as the baseline
models in this study, if not better. They were not the "canned" tower configurations,
they had more and/or larger guys, and had far more antennea than the catalogue
designs permit.
Empirically, we know they work, because they
are still standing, and we know what wind speeds they experienced. Imagine
that!
I set out to satisfy my own curiosity and desire to understand what makes these things tick and develop tools to process the problem. I found what I was looking for so I can now find the best solution for my pile of treasures.
I hope it helps you with yours!
May the forces be with you!
73, Kurt, K7NV.
