On 3/11/21 11:57 PM, Máximo EA1DDO_HK1H wrote:
Hi Jim,
As far as I know, for Cubical Quads, MiniNEC (Mmana) gets more
accurate results than NEC2/4.
I haven't tested myself but it's what I remember to read somewhere else.
Do you know how NEC5 performs on same Cubical Quads?
Thanks
73, Maximo - EA1DDO
No idea off hand, since I don't use MiniNEC or MMANA. A lot of the
issues reported with NEC are for NEC2, which has problems where wires
join at acute angles, and with numerical precision with very short
segments. NEC4 doesn't have those problems. I would *think* that NEC4
would do fine on a cubical quad - it's wires, it's simple, there's no
"rapid changes in diameter" or "wires of radically different cross size
connected at an angle" or "wires connecting to surfaces".
Both are method of moment codes, so the basic solution approach is the
same. The differences are in how they represent the current along each
segment. NEC4 is significantly better when dealing with transitions -
For all the codes, they make an assumption of what the current
distribution along a segment is - flat, sloped, typically, some sort of
A + B*sin() + C*cos(). What NEC4 did is change this a bit to
A+B*sin()+C*(1-cos()) to improve the numerical performance for very
short segments. With NEC2, if you took a dipole and made it
successively smaller and smaller segments, at some point, the solution
blew up - too many equations with terms that are either 1.000000001 or
0.00000001 and even with double precision it didn't work.
The other significant difference from NEC2 to NEC4 is a difference in
where the "current filament" is considered to exist. In both versions,
the current is assumed to be entirely along the axis of the segment
(there's no "around the segment" current flow). In NEC2, the current is
assumed to be at center of the wire, but in NEC4, it's on the surface of
the wire. For long straight uniform wires, this makes no difference,
but on a step or at a corner, or where wires are close together, it does
make a difference.
https://physics.princeton.edu/~mcdonald/examples/NEC_Manuals/NEC4TheoryMan.pdf
_______________________________________________
_______________________________________________
TowerTalk mailing list
TowerTalk@contesting.com
http://lists.contesting.com/mailman/listinfo/towertalk
|