Sunday, May 14, 2017

Correcting Tower Lean

I've straightened towers that have developed a lean. It's usually a routine repair job. However like any structural work on a tower it can be hazardous if you become careless. If you are at all uncertain of how to go about or if your knowledge of towers is weak I suggest hiring an expert.
In an earlier article I mentioned that my Trylon tower is not vertical. In this one I'll talk about correcting lean in general and how I corrected the lean in my current tower.

Effects of wind, ice and time

Over time self-supporting towers can develop a lean, sometimes quite pronounced. Most often in my experience the lean is away from the direction of prevailing winds. For most hams this means a lean towards the east since in mid-latitudes the prevailing wind is from the west. Major storms are the exception, which can push the tower in almost any direction.

If the tower was perfectly vertical when installed lean can develop for a variety of reasons:
  • Poor quality or missing fasteners: Unrated or improper fastener selection is almost always trouble waiting to happen.
  • Inadequate torque on splice bolts or other fasteners: Do you know how many inch-pounds of torque splice bolts require? Tightening large bolts while strapped onto a tower is difficult but must be properly done. Large size grade 5 hardware requires a lot of torque.
  • Excess antenna load: Overloading a tower might not bring it down but it will place it under severe stress that can bend structural components, and shift or shatter fasteners. I've seen many leaning towers that lean directly away from the prevailing wind direction.
  • Inadequate foundation: Some hams fail to construct a proper base for reasons of cost or inconvenience, or are simply negligent. A compromised base must be dealt with immediately; that is, the tower must be taken down before it chooses the time on its own.
Towers can be curved or lean when first built due to construction problems:
  • Bent or improperly aligned sections: Towers like the Trylon that rely on bolts to connect all the structural components must be properly aligned during construction, whether by the factory, the dealer or you. Do it wrong and the tower will lean. Bent components have a similar effect. Even when properly built rivets and bolts can wear or loosen and welds can crack.
  • Poorly seated section splices: If the upper section is not sitting on all the splice bolts it will not align with the lower section. This can occur since the holes are typically larger than the bolts.
  • Crane lifts: Many hams splice many tower sections on the ground and lift the assembly by crane. While faster and less dangerous than using a gin pole an improper lift can severely stress the tower. Splice bolts slip or, worse, structural members bend or break.
Problems can be delayed or entirely avoided with regular tower maintenance. Check all fasteners and structural components at least once each year. In areas with extreme temperate swings between summer and winter consider doing it twice yearly. After a severe wind or ice storm sight along each tower leg and face to look for changes. Use a long level if you don't trust your eyes.

On a guyed tower regularly measure the pre-load tension in all guys. Guying hardware may be failing or the anchors may have shifted due to faulty design or construction, or soil movement due to floods and seismic events. Guy tension drops and the tower wobbles or leans. There is a substantial risk of structure failure.

Considering a repair

If you discover or suspect the tower is leaning immediately perform a thorough inspection. Identify the locations where the tower deviates from a straight line. Most often it'll be a section splice. That's the type of repair I'll cover in this article. Other damage such as broken rivets, bolts and structural components or a shifted or broken foundation are more serious and must be dealt with before the lean can be addressed. It may be unwise to climb a tower with damage of this type.

Let's proceed under the assumption that the lean is due to nothing more serious than splice slippage. Before we begin it is recommended that all splice bolts be inspected for proper torque and structural components for cracks and bends. If a splice has slipped it is possible that there is less visible trouble lurking elsewhere due to the same stress event(s).

My tower

The lean in my tower is mainly due to the way I put it up. The gin pole I built had a few design flaws, one of which caused the pulley to occasionally jam against the section as it was being slipped into the one below. The low width-to-length ratio of the lower sections made it difficult to get them to sit vertically well enough to seat itself on the bolts. The ⅝" bolts for the bottom splices were difficult to torque since I have only one 15/16" wrench and it isn't nearly long enough.

The winter was cold and I did not always take time to correct errors when they occurred. Although I did notice most of the problems as they occurred each on its own seemed minor. Deviations from the vertical are amplified as you go up, and that caused a noticable lean. Well, at least I noticed it. None of the hams who've visited noticed it without me drawing it to their attention.

The photography skills to capture the lean is beyond my skill. These pictures underplay the extent of the problem. I added a closeup of the southwest leg since that is the leg with poorly seated splices. The southeast leg is seen on the right.

The worst splice is between the #10 and #9 sections. The splice between the #12 and #11 sections is more subtle but contributes a lot to the lean since it is the base section. Although there are a few smaller deviations higher up the tower they are not of immediate interest. It is the two mentioned splices that I intended to correct.

The geometry of improper section splicing

The small amount of play in the splice bolt holes can have a surprising impact on the straightness of a tower. It can be approximately modelled by a rectangle representing the face of the upper tower section.

In the domain of small angles we can simplify calculations using the approximation:
x = sin x = tan x, where x is the angle in radians
The upshot is that the ratio between the bolt slippage and section width (W) is equal to the ratio between the offset at the top of the section and the section height (H). The approximation works well whether you measure W at the top or bottom of a tapered tower section.

For example, with bolts sitting 1/16" high in the splice holes, W = 24" and H = 96" the top of the section is offset by ¼". The lean angle α = 0.15°. If there are 6 sections above the improperly done splice the offset will be 1.5" at the top of the tower.

In my case the lean is worse since there are two sets of poorly seated bolts and one has an error worse than that of the above example. Standing on the top of the tower the lean is very noticable.

Lean correction procedure

I waited for a day with little wind, no rain and no one in the vicinity except me. This took a while because of the horrid weather we're having and the regular presence of workers doing house renovations.

I tied a rope between the southwest leg ⅔ up the tower and a suitable anchor to the southwest. The impromptu anchor is a post supported a balcony on the house. The strength of the rope and anchor is not critical since the required tension is only in the tens of pounds. The balcony is in no danger. We want just enough force to encourage the tower to sit back onto the splice bolts when they are loosened.

The 2 ton winch is certainly overkill. I used it because it made it easy to finely adjust tension on the rope. You can set the tension by hand if you prefer, in which case I suggest using a temporary rope cleat to allow a similar and rapid method of adjusting tension. The ladder you see in the picture belongs to the renovators, not me. To the left you can see a few elements of the Hy-Gain TH7 I am assembling.

To begin we put some tension on the rope. I prefer rope over steel cable because it is more forgiving of excess zeal. Stop when there appears to be just enough tension to draw the tower back when the splice bolts are loosened.

For the next step it is necessary to loosen some of the splice bolts. The picture shows the #11-#12 splice on my tower. At the centre is the southwest leg. All 4 bolts on that leg are loosened, but only enough for the lock washers to relax. Do the same for the nearest 2 bolts on the adjacent legs. This allows the tower to pivot on the back legs without stressing the steel. The 4 bolts on the opposite face are not touched.
Note: The bolts you loosen are different on towers with tubular legs (e.g. Rohn) or towers with bolts on one surface of the legs (e.g. DMX). Even so the basic procedure is the same. It is fair to say that adjusting the Trylon is the more complicated of the three because of the leg shape. The big guyed tower I am putting up this year, the LR20, has the same splice bolt and leg pattern as the Trylon. The procedure is similar on the LR20 but with modifications due to its being guyed.
Don't be surprised if you see no movement of the tower when the bolts come loose. The amount of downward motion is slight and may occur in small steps as each bolt is loosened. With the selected bolts loosened I returned to the winch. I discovered that the rope was quite slack, so obviously the tower shifted in the desired direction. Looking up the tower from the bottom the shift was visible. I put tension back on the rope in small steps until I was satisfied that the bolts were fully seated. On the tower I could tell by the fact that the bolts were being pushed down. I used a level to confirm I had the result I wanted.

I tightened two of the southwest leg splice bolts, one on either side of the leg, and left the others as is for the rest of the procedure. I then repeated the rope tension and bolt loosening procedure for the #9-#10 spice further up the tower. The rope slackened less this time when the tower sat back towards the southwest so I added more tension to the rope. With the bolts fully seated there was some residual lean at that splice.

Perhaps the #9 section is improperly aligned or something else is going on. Adjusting a misaligned section is not a task that can be safely done on an erected tower. I therefore chose the next most desirable approach which was to slacken the rope a small amount and loosen the remaining 4 splice bolts. At this point all 12 splice bolts are loose, but not so loose as to allow the tower to wobble freely and fret at the bolt threads (the bolts are harder than the tower steel). It is at times like this that you appreciate doing this job on a windless day.

Despite that warning this procedure is safer than you might expect since the tower is trapped by the bolts and can move only a small amount, an amount that is in general not a danger. The danger that does exist increases the further down the tower this is done and less the further up, due to the amplification of the lean upward from the splice being repaired.

With all the splice bolts loose I added tension to the rope in small steps, visually inspecting the tower after each increase. When the tower was about as close to vertical as I could reasonably expect I went back up the tower and tightened all the splice bolts, first at the #9-#10 splice and then the remaining 2 bolts at the #11-#12 splice.

The bolts opposite the southwest leg on the #9-#10 splice are no longer sitting on the bolts. When the bolts were tightened it was the pressure between the two legs that held the sections in position. This is not ideal but acceptable under the conditions I found. There is the possibility that due to ordinary load forces that over time the section will slump and sit back on those bolts. That will reintroduce a small amount of lean.

In the end I had a tower that had ~95% of the lean corrected. The couple of small deviations on the upper sections is not a concern since the absolute amount of lean is very small. I declared success and put my tools away.


I remain unhappy with a few remaining issues with the tower. The ⅝" bolts on the #11-#12 splice and the splice of the #12 to the base stubs need more torque. Using a long level there is a small error in the base section's vertical orientation that was not there when built. It is amplified over the height of the tower. I will buy the tools I need to deal with those large bolts.

There are a few small areas of rust where the cold galvanizing paint isn't up to the job. Those will need touching up. Alignment issues remain with a couple of upper sections, including the #7-#8 splice. The splice problems can be dealt with in the same manner I described in this article. There are also a few diagonals that have resisted repair and ought to be replaced.

As I said, these are minor concerns.

Does it matter?

A few years ago I went on a road trip back home to VE4. An emergency road closure rerouted us and thousands of others to the more northerly highway 11. It's actually a shorter route but more isolated and less scenic than the 17. One stretch in particular, 200 km long, was devoid of towns or service stations. That's what we get for living in a large country with a modest-size population.

The route is not entirely wild. In addition to numerous indigenous communities and wilderness recreation areas off to one side or the other there are regularly spaced maintenance depots, for road work and other services. They are interconnected by radio links. This requires antennas up high to reach its neighbours over the rough terrain.

With little other than forests, hills and lakes for this long stretch those depots drew my eye. Each had a Trylon tower, perhaps 70' to 80', with a collinear VHF antenna. What surprised me is that pretty much every one of those towers was not vertical. They deviated from vertical by a few degrees, which is a lot. It was readily apparent to my eyes and I expect by anyone who takes the trouble to look.

This is an example of incompetence. If not for the light load those towers would not survive long. Ham towers are more susceptible since their owners often push the load to the limit, and beyond. Leaning towers are also not much fun to climb.

So it can matter. Keep it vertical and eliminate the concern. If your antenna load is light in comparison to the tower capacity don't fret about it too much. But I'm the sort of person who walks into a room and straightens the pictures hanging on the walls.

No comments:

Post a Comment

All comments are moderated, and should appear within one day of submission.