[TowerTalk] Tower Failure Analysis

[Grant Saviers]

A very cool finite element analysis!  A lot of work to set that analysis 
up for the LM470.

If I see the overload points in your FE analysis correctly, the 
overloads are in/near the overlapped sections of the middle sections.  
The top and bottom sections appear ok at failure loading.  So it is 
interesting to note some of the differences between a circa 1970/60's 
designed tower and a current model UST HDX589.
I have a Tri-Ex LM354 which I could measure and observe a few 
differences, the goal being to understand how designs have changed, not 
pro or con either manufacturer.

1. UST uses 50ksi steel in the legs vs the 42ksi in the TX.  The TX 
appears to fail in leg tension/compression so higher yield steel allows 
20% more stress.
2. The UST overlaps are "double Z" braced for the external and internal 
section vs single Z bracing in the TX, which should better distribute 
the stress and stiffen the joint.
3. The LM354 bottom section is 1.5" od tube with a 0.125" wall vs the 
UST #7 (3rd from the top) of 1.66"od by 0.19" wall.  It appears that the 
UST has always a heavier wall tube, although a section for section 
comparison is not exact.  The face center line is 18" on the 354 bottom 
and 19.94" on the UST #7, so they are approximately the same.
4. The top and bottom of section cross plates are two stamped 0.125" 
thick x 1.25" bars on the upper sections of my 354 vs the larger flat 
plates on your 470.  The UST uses thicker section channel.
5. The runners on the 354 are flattened and offset segments of the leg 
tubing as shown in your model, which precluded me measuring these tube 
wall thicknesses.  I'd estimate the wall to be 0.10" on the top two 
sections, your model rules on this one.  UST minimum wall is .154" for 
sections #3,4,&5.  (5 is the 589 top section).  On the UST the runners 
are welded on U stampings that ride inside the lower section, I think a 
stronger connection between sections.

I doubt that Tri-Ex had FEA available when they designed the LM series.  
I recall having a project in the 1970's that burned through several 
$100k of CDC6600 supercomputer time for a FEA of a structure less 
complex.  Now, PCs can do this easily and faster.  Towers have gotten 
stronger and more reliable as a result of better analysis tools, 
materials, and tougher standards.  If the tougher standards are a true 
reflection of what towers will experience as wind forces, then antenna 
ratings should be considered as optimistic for older tower designs.  
However, site specific conditions dominate and the standards & 
regulations track worst case for a larger geography.

Where will they fail when overloaded?  Probably in the middle for LM's 
and near the base for HDX's.

Grant KZ1W


On 2/9/2013 12:40 AM, KM5VI wrote:
> In response to the thread about crank-up tower failures I put some selected
> output views from a finite element stress analysis of my LM-470 loaded to
> theoretical yield failure at:  http://www.flukey.cc/km5vi/finite%20Element/
>
>   
>
> The analysis indicates the theoretical points of failure where the peak
> stress exceeds the material yield (for this tower 42,000 psi).   As someone
> stated, these points certainly are not at the lower tower section.
>
>   
>
> KM5VI
>
> _______________________________________________
>
>
>
> _______________________________________________
> TowerTalk mailing list
> TowerTalk at contesting.com
> http://lists.contesting.com/mailman/listinfo/towertalk
>