[Top] [All Lists]

[TowerTalk] earth anchors & screw-in anchors

To: <>
Subject: [TowerTalk] earth anchors & screw-in anchors
From: (Jim Lux)
Date: Wed Feb 26 16:23:24 2003
At 03:32 PM 2/26/2003 -0500, Pete Smith wrote:
>At 11:00 AM 2/26/03 -0800, Jim wrote:
>>But, when you use multiple anchor points, you run the risk of not sharing 
>>the load among the guys as intended. The relative tension in the guys 
>>will change with their relative length, which changes with 
>>temperature.  If you have separate attachment points, you'd need to 
>>design each guy for more (conceivably, all) load, or design some sort of 
>>load equalization into the system (i.e. maybe the tower flexes a bit).
>I'm NOT an engineer, but if Jim's point is that the dynamic loads on a 
>tower are equally shared among the guys at all levels when they are 
>attached to a single equalizer plate and anchor, then I'm 
>unconvinced.  K7NV has done some finite element analysis on guyed Rohn 45 
>that seems to demonstrate that deflection of the tower, under load, is 
>very unevenly distributed among the different levels of the 
>structure.  Seems to me that this must reflect different tension on the 
>guys at different levels.  I wish Kurt were around to comment, but I 
>suspect he's off in NZ tending to composite spars at the America's Cup (at 
>least I hope so, for your pleasure, Kurt...).

I'm not surprised the load and deflection is uneven. This rapid gets so 
complex that you're probably better off designing with a large margin and 
being done with it.

>Jim's point about different guy materials rang a bell with me.  I know 
>that my Phillystran top guys stretch about twice as much as 
>EHS.  Fiberglass is somewhere in between, and the first 21 feet of my 
>middle guy set are that material.  Only the bottom guy set is all 
>EHS.  Does the Rohn equalizer plate really have the ability to manage all 
>this variation and apply it neatly to the single guy anchor?

Probably not..or, maybe it's good enough..given sufficient design margin in 
all the components. Or, maybe the plate's really just a convenient 
multipoint joiner designed to solve an installation problem for people who 
want to install just one anchor.

>  Or might it be better, in my situation, to anchor the top guys 
> separately, set their static tension per spec, and rely on the huge 
> margin of the entire system to ensure that if the wind ever blows hard 
> enough to approach the breaking point, I'm going to be worrying about a 
> lot of things other than my tower.

The huge margin approach... simple, effective, works well, unless you have 
some other constraint that forces you away from it.

>>The Rohn approach is a good one, sort of following the "put all your eggs 
>>in one basket, and then watch that basket very carefully" 
>>philosophy.  Design the one anchor with a huge margin and then hang 
>>everything off of it.  Of course, if you "cheat" and put an unreliable or 
>>underspecified anchor in, then you've defeated the purpose.
>This may be what they are doing, but it still worries me.  Sure, you can 
>inspect the anchor rod down to the point where it enters the concrete 
>block anchor, but if you've parged it with an asphalt sealant, it can be 
>hard to see what may be going on under the coating.  Even then, there's 
>also the possibility of failure through electrolysis inside the block.  It 
>may be a choice among non-optimum solutions, but I think there's a 
>legitimate case to be made for the scheme I proposed.

Or, what about two anchors, attached to an above ground equalizer attached 
to all guys.. Each anchor would need to be able to take the full load. 
Then, your suspect point (below ground anchor) is redundant.

In any case, as has been pointed out many times, erecting a guyed tower 
with multiple tiers of guys is non-trivial.  For some (many) cases, all the 
issues have been worked out in advance (because lots and lots of people 
have done it over the years) so you can just "cookbook" it. But, it still 
needs an appreciation for the limitations of the original analysis, so you 
can know if you are getting "close to the edge".

Jim Lux, P.E., W6RMK

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