[TowerTalk] Wind Ratings

[Grant Saviers]

Some more for the discussion:

1. Complex structures are pretty accurately analyzed with Finite Element 
Analysis (FEA).  The error to real behavior can be made very small if 
you have enough compute power and build a good model.  Way back it took 
a $10M CDC 7600 (I rented one, $1k/hr) to model a part smaller than your 
hand, now a i7 PC 8Gb SSD is at least several hundred times that power 
and there are several free derivatives of Nastran FEA for them (NASA 
paid for it).  I think upgrades to tower designs and ratings, besides 
changes in the standards, are the result of more available analytical 
tools.  While the detailed CAD drawing of a tower section construction 
(braces, welds, joiners, etc.) is not public, the section properties as 
a single structural element is "out there" for PE analysis of towers.  
So it is practical for hams to FEA towers such as Kurt K7NV did with his 
classic work.  http://k7nv.com/notebook/topics/feastudy.html

2. The complexity of the wind field is largely ignored in the standards 
except for the exposure classes.  Then again, Computational Fluid 
Dynamics has progressed to doing complex flow modeling.  Again free on 
the web and it works fine on a big/fast PC.  So a specific tower site 
could be modeled.   Consider that NASA infrequently uses their numerous 
wind tunnels (the big one at Ames needs 104 megawatts to run!).  
Computers to it faster and cheaper. I have to wonder if the link re wind 
tunnel testing of cell antennas refers to the difference from "rules of 
thumb" vs tunnel results or CFD modeling vs tunnel?

3.  Dave Leeson W6NL in "Mechanical Design of Yagi Antennas" recommends 
down boom guys.  He states many booms fail with updrafts, so add that to 
our list of turbulence induced problems.  I suspect his QTH on a ridge 
in the Santa Cruz Mts. is particularly prone to that.

4.  At least for the 3 towers of PE analysis I've had done, the "load 
margin" of various loads to the manufacturers engineering spec is 
calculated.  Obviously higher than 100% isn't going to get approved and 
how much margin is desirable is a judgement call.  I had this discussion 
with an FAA inspector re United Airlines using a 727 at 99% of rated 
maximum range.  He said, "under 100% I can't do anything, but bad s**t 
will ultimately happen".   There is some safety margin that the 
manufacturer used to set their spec based on material, aging, and 
process variations.  Some of those can be imputed from their specs.

I agree that the administrative state has gone crazy with regs, delays, 
obstruction, and pinheads abusing their power.  "No" is the safest 
answer for a bureaucrat.

Grant KZ1W

"more towers are killed by trees than trees by towers"  (2 towers killed 
by trees in the last 3 years near my QTH, falling tree hit guy)

On 6/11/2017 12:34 PM, Wilson Lamb wrote:
>
> I don't have expertise in this area, but I know people who do.
> ALL such ratings are based upon a long string of assumptions, some 
> pretty loosey goosey.
> The drag on a nice clean cylinder, in straight line wind, with a 
> surface finish of known characteristics, is fairly well known and agreed.
> When you group it with other members, make it much shorter than the 
> original specimen, add some bracing at various angles and add 'em up, 
> you are fairly far out on a limb.
> What's the wind profile? Where are the trees, buildings, hills?
> How does the environment change with azimuth?   Hurricane winds can 
> come from ANY direction and, if the eye goes over your location, you 
> can see up to 180 deg of change.
> Is there small scale turbulence involved, to modify the profile?
> What about downdrafts?  Could they make wind that's not horizontal at 
> the tower?
> Is the drag of your antenna a function of the relative wind direction?
> What were the "safety factors" used for CYA?  There's no point in 
> comparing an analysis with a SF of two to one with a SF of 1.5, unless 
> you know the SFs
> If the engineer published the SF, lots of people would just get 
> comfortable with the unfudged numbers, pre SF, and uprate their plans 
> accordingly.
>
> It would be neat to strain gage the legs of a free standing tower and 
> measure the forces actually seen at the base, while measuring the wind 
> profile with an anemometer that can go up and down to get a profile.
> Of course you'd likely have to wait years for wind strong enough to 
> make the data relevant to the design max spec's.
>
> Precision analysis/design is no better than the basic assumptions.  
> Adding decimal places doesn't make it better.
>
> OK, this was a rant, because it pains me to see local government 
> people who know nothing, or less, about towers accepting design 
> criteria put out by committees of people who are paid to write codes 
> and imposing them in circumstances where it doesn't matter what 
> happens.  If MY tower can't fall on my house or anyone else's 
> property, it should not be subject to regulation.  Am I sensitive?  
> Yes.  A friend on 52 acres was just told that a 40X40 foot shop was 
> too much impermeable area and would disturb the watershed!  What the 
> disturbance might be was not disclosed, but I'm pretty sure the water 
> is going to go into the river no matter what!  Said water will be a 
> lot cleaner if his dirty stuff is inside, not out in the rain.
>
> Does anyone have records of failed ham towers, pictures?  The fact 
> that something survives says very little about the analysis by which 
> it was approved, except that the analysis plus SF produced a design 
> that survived the load experience.  There is no bound on how much 
> overdesign there may have been.
> How about tower sections tested in a wind tunnel?  Full size, of course.
> We had a massive failure here a few years ago, of a 1500 ft TV tower, 
> brought down by ICE.  Not too much ice, but a dynamic overload caused 
> by rapid ice shedding from guywires.  At least that's the theory 
> published.
> www.oldradio.com/archives/warstories/WRAL.htm
> I feel better, target up.
> Wilson
> W4BOH
>
>
>
>
>
>
> _______________________________________________
>
>
>
> _______________________________________________
> TowerTalk mailing list
> TowerTalk at contesting.com
> http://lists.contesting.com/mailman/listinfo/towertalk
>