Re: NEC-LIST:Modeling a conductive surface in NEC2

On Tue, 1 Oct 2002, Brandon Orchard wrote:

> In my previous e-mail I mentioned that you can model any surface using wire
> segments in NEC. By that I meant any metal surface of course.
>

As they say in South Africa: ``JaNee''. (Yes-No, a non-committal response
:-) We have used SuperNEC to model non-metallic surfaces very effectively:

  A R Clark and A P C Fourie and J Dresel ``Modelling lossy surfaces using a
  wire Method-of-Moments code'' IEE Proceedings---Microwaves, Antennas and
  Propagation, pp247--248, June 2000. Vol 147, No 3.

Subsequent work has proved the article's concluding assertion that this
method can be extended to dielectric surfaces, but has not been published.

>
>
> > Hi Jerry
> >
> > You can model any surface using wire segments that are about a 1/10 of a
> > wavelength in length.
> > I don't think that the diameter of the wires would be of such importance
> in
> > such a case as long as they satisfy the general rules for NEC with regards
> > to Segment length to radius ratios etc.
> > This same technique of modeling surfaces by a wire mesh can even be used
> to
> > model very complex structures such as aircraft, boats, cars etc.
> >
> > If you are looking for a easy to use, professional program to quickly
> model
> > this problem, why don't you try using SuperNEC. It has a full 3D graphical
> > interface so you won't have to write any text .nec files, and a really
> great
> > output interface for plotting 3D radiation patterns, smith chart plots etc
> > etc. It also has a number of predefined structure and antenna assemblies
> > such as wire mesh plates, cylinders, spheres, yagi's, helices, lpda's etc
> > etc. With the antenna assemblies, by changing your model frequency in
> > SuperNEC, the assemblies are automatically re-segmented so that your model
> > remains valid. Another great feature of this product is the structure
> > checker which checks your model to see that all the nec rules to produce
> > accurate results have been followed.
> > I am sure that these features of SuperNEC will considerably reduce the
> time
> > taken to model your specific problem.
> >
> > You can download SuperNEC from www.supernec.com and get a 30day evaluation
> > license file for running the software.
> >
> > Regards
> > Brandon Orchard
> > ----- Original Message -----
> > From: "Jerry Flanders" <jeflanders_at_comcast.net>
> > To: <nec-list_at_gweep.ca>
> > Sent: Friday, September 27, 2002 11:42 PM
> > Subject: NEC-LIST:Modeling a conductive surface in NEC2
> >
> >
> > > I am trying to evaluate a few ideas for a replacement antenna for my
> ham
> > > radio club's 2 Meter repeater.
> > >
> > > The new antenna will be mounted on one of our city's water tanks. I
> would
> > > like to take into account the effect of this nearby reflective surface
> in
> > > the model calculations.
> > >
> > > I understand NEC2 cannot model the surface, but that I can simulate the
> > > surface with a wire grid. I need help with some specifics on the grid
> > > structure.
> > >
> > > I have a copy of Arie's "Build" to actually generate a grid of the right
> > > size and shape, but I don't know what mesh size (size of the openings)
> is
> > > appropriate. If I use too large a mesh opening, I suspect it would not
> > > represent a plane surface accurately, but if I just use an
> > > arbitrarily fine mesh, I will cripple this 384MB computer. Is there a
> > rule
> > > of thumb for adequate mesh "pore" sizes in terms of wavelength?
> > >
> > > I could also use advice on what diameter "wire" to use for the grid, and
> > > what conductivity, if conductivity is important. I assume it is a steel
> > tank.
> > >
> > > Anybody?
> > >
> > > Jerry Flanders
> > >
> > > W4UK
> > > --
> > > The NEC-List mailing list <nec-list_at_gweep.ca>
> > > http://www.gweep.ca/mailman/listinfo.cgi/nec-list
> >
>

-- 
Alan Robert Clark, Pr Eng     Computational Electromagnetics
Associate Professor                 Wits University
School of Electrical and Information Engineering
Wits; Gauteng              ``Bugs are later known as features''
2050 South Africa                 Ps 111:10; Ps 37/150
Fax (+27 11)403-1929       clark_at_YingTongDiddleIPo.ee.wits.ac.za(Pref)
Tel (+27 11)717-7223(24hr)  <http://YingTongDiddleiPo.ee.wits.ac.za/>
     **Linux 2.4pl7(RH7.2)---the choice of a GNU generation.**
-- 
The NEC-List mailing list <nec-list_at_gweep.ca>
http://www.gweep.ca/mailman/listinfo.cgi/nec-list