[TowerTalk] MODEL FOR TOWER

[jimlux]

On 4/26/19 10:45 AM, Gedas wrote:
> One thing I have not seen mentioned (maybe I missed it) is the fact that 
> a whole lot of time & effort is going to go into generating a super 
> accurate model for a complex tower and antenna system but then, what of 
> the grounding of the tower itself at the base? Grounded or un-grounded 
> or sorta grounded are all going to give different results.

That's easy to test in NEC. Just put a resistor between tower and ground.

I actually don't think it will have a huge effect.  the tower is 
basically a bunch of vertical wires.  They'll change the tuning of the 
Yagi a bit, and there will be some current in the area of the Yagi.  But 
60 feet away, there won't be much current.

I'd worry more about the horizontal and sloping cross members.  They're 
more "aligned" with the segments in the antenna.

However, NEC is a good way to test this. You put the wires in, and see 
if there's current. If there's current, that's something to think about.

Even in NEC2, which isn't great about connections between antennas and 
ground, you could do a quick check.  If you have NEC4, then it does 
great at modeling wires penetrating the ground - that was one of the big 
improvements from NEC2.

To be honest, I don't know why NEC4 is still export controlled, other 
than LLNL has no incentive to re-examine it. It's probably not a huge 
money maker (and they're non-profit anyway), but as long as someone has 
to vet the recipient, they're going to charge for it.  And it probably 
helps some statistics - Gerry Burke's organization at LLNL probably gets 
some kudos for "number of licenses issued" (that's the way it works at 
JPL) and that helps justify their funding from DoE.


> 
> There are going to be deviations from real world results if one models 
> with the limitations of objects touching the ground with NEC-2 and even 
> with NEC-4.

NEC4 is pretty good at modeling things penetrating or beneath the soil 
surface. The key is that at the point of contact, you need to have a 
segment boundary. The recommended (in the manual) way to do this is to 
have different wires for the above ground and below ground parts of the 
structure. That guarantees that there's a node at the surface.

What NEC4 doesn't do well is model a wire laying on and just touching 
the surface.  When I model, what I do is model the wire at 10cm, 5cm, 1 
cm, 1mm, -1mm, -1cm, -5cm, and -10cm and see what changes. (assuming the 
wire is <2mm in diameter).

To be honest, people have spent their lives studying the problem of a 
wire just touching, or partly immersed, in a dielectric boundary.  J.R. 
Wait at NIST has lots of papers on this.

The big uncertainty is the soil dielectric properties.


> 
> Gedas, W8BYA
> 
> Gallery at http://w8bya.com
> Light travels faster than sound....
> This is why some people appear bright until you hear them speak.
> 
> On 4/25/2019 7:40 PM, jimlux wrote:
>> On 4/24/19 5:03 PM, Steve Maki wrote:
>>> On 04/24/19 8:40 AM, jimlux wrote:
>>>
>>>>> I have always just guess when converting a tower to a wire 
>>>>> diameter. The original poster seemed concerned that his tower would 
>>>>> effect his horizontally polarized HF beams, I have not seen that 
>>>>> occur.
>>>
>>>> I think the question would be about the SSV/BX style tower which is 
>>>> larger at the bottom than the top. Rohn 25 or 45 are "small" 
>>>> compared to a wavelength in the horizontal direction, so they can be 
>>>> modeled as a "fat wire" - just like a cage dipole element, for 
>>>> instance.
>>>>
>>>> The tower in question is 7.5 ft at the bottom and 2 ft at the top 80 
>>>> ft high.
>>>> The OP was asking about a 20m Yagi to be mounted at 60 ft, where 
>>>> you'd effectively have big square loops that are about 3 1/2 ft on a 
>>>> side (14 ft total perimeter)  near the antenna, as well as diagonal 
>>>> struts of some length.
>>>>
>>>> The wavelength is 60-70 ft, so those squares are about 1/4 
>>>> wavelength in perimeter.  If they were 1/10th wavelength, I'd say 
>>>> "model it as a big wire", but that's big enough that there might be 
>>>> some interaction, especially since they will be effectively "inside" 
>>>> the Yagi.
>>>
>>> In the scenario where you have a large enough tower that a nearby 
>>> horizontal antenna is impacted by the tower's horizontal members - is 
>>> there a fundamental difference between a lattice tower compared to a 
>>> cylinder of like diameter?
>>>
>>> I've assumed no, but now you have me wondering.
>>>
>>> -Steve K8LX
>>>
>>
>>
>> Here's what the latest NEC documents say:
>> "The wire radius a relative to {lambda} is limited by the 
>> approximations used in the kernel of the electric field integral 
>> equation. NEC uses the thin-wire approximation, neglecting transverse 
>> currents on wires and assuming that the axially directed current is 
>> uniformly distributed around the segment surface. The acceptability of 
>> these approximations depends on both the value of a/{lambda}  and the 
>> tendency of the excitation to produce circumferential current or 
>> current variation. Unless 2*pi*a/{lambda}  is much less than 1, the 
>> validity of these approximations should be considered."
>>
>>
>> So NEC does not model transverse currents in a conductor - so while 
>> you can model a tower as a wire of comparable diameter to the tower, 
>> the model will only work for (mostly) fields that are vertically 
>> oriented.
>>
>>
>> A further hiccup in modeling a lattice tower might be the "short 
>> segments forming loops" problem.
>>
>> NEC2 doesn't deal well with very short segments. NEC4 deals with them 
>> just fine.
>>
>> However, for loops where the circumference is <0.002 wavelength, the 
>> results may not be valid.  IN practical terms.. if you've got a 
>> triangular tower with face width 1 foot (perimeter 3 feet), if the 
>> wavelength is >1500 feet, you might have a problem.  Topband and Cheap 
>> TV antenna lattice *might* get into trouble.
>>
>> Modeling 4" reinforcing mesh or a dense rebar lattice might also run 
>> into troubles.
>>
>>
>> _______________________________________________
>>
>>
>>
>> _______________________________________________
>> TowerTalk mailing list
>> TowerTalk at contesting.com
>> http://lists.contesting.com/mailman/listinfo/towertalk
> _______________________________________________
> 
> 
> 
> _______________________________________________
> TowerTalk mailing list
> TowerTalk at contesting.com
> http://lists.contesting.com/mailman/listinfo/towertalk