Check vertical alignment of a tower

If you know the different diameters at the top and bottom you could calculate the design offset versus the actual offset.

Ie if the tower is 3.0m round at the bottom and 2.0m at the top, the top should be 1.0 further horizontally away than the bottom.

Maybe measure the top and bottom on all four sides? The top four points should be centered within the bottom four points.

We had a small job where we had to ensure some tapered mobile phone towers got installed exactly vertical. It would have been easy except the towers were cast concrete in two parts and because the join between the upper and lower section was so rough, the towers were actually slightly kinked! We did the best we could and the client ended up happy thankfully.

Hi Roveryan

Is this a 'lattice tower', square in plan with four legs tapering up to a 'shoulder'?
Usually above the shoulder there's a bolted on pyramid with a lightning rod, and as you suggest, near but not on the centre.

If you can't climb the tower and place a centre mark up at shoulder level you will have to measure the horizontal component of the taper preferably on all four legs.
I used to do this regularly on towers, stacks and chimneys.

If the taper is the same for all four legs then the tower is vertical. The method I use is to lay a levelling staff horizontal at base level, set up a 5" theodolite well back from the tower so that sighting up to the top is comfortable. Sight top and bottom and rotate down to the staff and record the readings, averaging face left and face right is best. You could do two set ups at 90 degrees to each other and aligned with the tower faces or, four set ups at 45 degrees to the tower faces, ie at right angles to the tower diagonals. All depends how many obstructions there are to seeing the structure. Make sure you get the exact same points every time, it's easy to confuse the main structure with brackets, plates etc, especially against a bright sky.

A typical UK lattice tower, it tapers then the upper part is vertical. This photo was taken about three years ago, there's even more stuff on it now!

Artie,
Yes that is the type of tower I need to measure. There are so many attached components that you don't know where the center actually is. I will try to understand your method that you suggested. Thanks guys.

This might explain slightly better than my rough description!

Reflectorless measurement would be another option, which would pick up any torsion or twisting of the tower due to wind loading, ie construct a 3D model, that would require x4 accurate stations on the ground opposite each face of the tower.

Hi,
These towers are supposed to be symmetric.
On each of the 4 sides you take readings on the left side and right side of the tower(at different elevations if you want) and average them to calculate the centerline of each side.
In most cases taking readings on 2 sides will do I guess.

chr.

> How do you check for the verticality of a tower? There is no vertical line in the structure since the components are angled up to the top.....
I had a similar job about a year ago. In my case it was a water tower with 6 round legs. We established control around the tower, then shot the outside of each leg at points around the circumfrence with reflectorless at bottom, middle, and top. I then used the best fit curve routine in Carlson to establish center points to compare.

You could do similar thing with you tower. Shoot the outside corner of each leg various heights. The points where the horizontals connect would be convenient. The center point is the average of the 4 northings and the 4 eastings at each level. Maybe.

Shoot the 4 mains at the top, shoot the 4 mains at the bottom at given elevation, draw diagonals in cad, analyze the intersections at top and bottom. Done

Depends on your field software, but you can shoot four corners at the bottom as the same point and average all four to give you the center (centroid). (I'd also shoot them separately to check later). Then do the same thing at the top if possible (i.e. prismless) The horizontal inverse between those two averaged solutions will give you the horizontal distance from the center at the bottom to the center at the top.

Just out of curiosity, what are they going to do if it isnt?

While I might be tempted to go prismless and average shots for top and bottom to establish the vertical alignment, it might be just as easy and perhaps more accurate to set three intervisible control points in three quadrants around the tower, assuming at least the four corners of the tower, top and bottom are visible from any two stations, and simply turn angles to the center of each corner from two stations and intersect the angles turned in CAD to establish the center of each corner of the tower, top and bottom. Then compare the offsets to determine which way it's leaning. Then after determining that it is leaning, get out my handy man jack and set'r right!

>averaging face left and face right is best

Averaging both faces (Direct & Reverse, older terminology) is absolutely required because this measurement is very sensitive to any error in the instrument's vertical motion alignment, and that sensitivity doesn't show up so much in the work you do more often.

> Just out of curiosity, what are they going to do if it isnt?

> Just out of curiosity, what are they going to do if it isnt?
Resurvey when the wind is blowing the other way.

😉

Seriously, though, does anybody have data on how much towers move? Say with wind, or with sun heating one side on a summer morning? The tower in Artie's picture obviously is going to be better than towers that have a large height-to-width ratio. Guyed towers will move more than tapered self-supporting ones. But I'd like to see some example data.

I can remember being at a construction meeting in the 70s when we began staking metal towers for SWEPCO and talking with an engineer as he explained what the towers were designed to take from outside forces.

SWEPCO would be my first call or the people I had a contract with.

0.02

There are a lot of good ideas on how to measure this. I like seeing how people solve the same problem differently. I guess which method you choose depends on the situation. The way I described is the quick and dirty method to get a solution as quick as you can take 8 shots and inverse. Some others may be more accurate, but it depends on how much time you have and what tolerance you're looking for.

I used to monitor a gas processing plant with several cooling towers of heights over 100 feet. To do this we set up three control points outside the plant and turned angle sets to the perceived outer edges of each tower from each control point. The data was processed using StarNet. The resulting error ellipses were almost perfect circles and allowed us to determine the circumference top and bottom and the slope and angle from the centerpoint of the top to the centerpoint of the bottom. These towers had up to 3 feet of lean and one had an increase in lean of 1 foot in the last year we monitored the plant.
Control points outside the plant had the advantage that we did not have to wear respirators which was required in the site.

Im going with your method ..

Set up opposite one face, and measure the angle between the bottom legs, than turn one half of the measured angle and mark it at the bottom of the tower. This will be the center at the bottom. Measure the angle at the top of the tower, than turn one half of this angle and mark it at the bottom of the tower next to your other mark, this will be the center at the top. The differnce between the marks will be the amount out of plumb. Do the same for the othe face.