Today, with DEMs available to generate the "tiles" for a full 3d model would be
straightforward. That was one of the challenges when Breakall did his work.
There are commercial (and non-commercial) codes out there that can do this kind
of thing (RadioMobile for VHF and up, although I think it does only radial
rays).
Returning to R.W. Hamming's comment "The purpose of computing is insight, not
numbers", particularly for amateur radio, I'm not sure an ever increasing model
fidelity is useful. If you are a SW broadcaster or setting up ALE links on HF,
there are commercial codes out there that do this kind of thing.
Having developed a lot of models over the years for both work and play, there's
value, to me, in developing a code like this - because I'm interested in how it
works, and the subtleties and second and third order effects, but when
developing for me, I don't have to work about making it work for others.
What might be useful is not necessarily making a generalized tool, but to do
some example scenarios - the vertical on top of a hill or next to a cliff. And
with some parametric variation in runs. That might give enough information to
allow a conceptual model or explanation to be developed, and that's probably
more useful.
I'm reminded of folks who build SPICE models of circuits in incredible detail,
find that their amplifier oscillates, when you could have seen that coming with
the original block diagram and seeing that reverse gain was >1. OTOH, folks
who have circuits that "sometimes work" - that's where the model is nice,
because it's a heck of a lot easier, and more informative, to vary component
values and temperatures in a model than to solder 100 different components and
run them in a temperature chamber.
On Tue, 14 Jan 2025 12:53:24 -0800, Brian Beezley <k6sti@att.net> wrote:
"A good way to think about this kind of 'antenna on the side of a hill'
is to rotate the pattern for flat ground to the angle of the hill."
That's how I think of it, too, Jim. It's a good first approximation, but
there's more to it. In the simplest case, the far field includes a
direct ray, one from sloping ground, one from flat ground after the
slope quits, and diffraction from the bottom of the slope. There might
also be a secondary bounce from sloping to flat. Although sloping ground
can certainly concentrate power at low angles, you never get power right
at the horizon due to the inevitable phase cancellations. W6NL has an
old photo me of explaining this very geometry to W7EL by drawing rays in
the dirt with a stick at his hilltop QTH!
My TA program will calculate the resulting pattern for a vertical
antenna at the edge of a slope, including all possible reflections and
diffractions. But I don't have a working copy.
One thing to note about these ray-tracing programs is that they all do a
one-dimensional analysis along a given bearing. For mountainous terrain,
this is often inadequate. Power can easily reflect or diffract from
terrain to the right or left and wind up on the bearing of interest. The
magnitude of these off-bearing paths can be quite large. This really
restricts terrain for which results are accurate.
I've been thinking of resurrecting TA for Windows. If I do, it will only
be for a full 2D analysis. That should be possible in a reasonable
amount of time with modern computers, especially with multiple
processors gnawing away at all of the potential paths. But the main
problem is getting reliable data to check the calculations. What I used
for TA was the original WA3FET IEEE article. I had to extract data by
hand from the curves. But as I recall, the helicopter that took the data
flew directly over a hill with the RF source so it is not a good test of
2D terrain. I would not write a 2D program without having reliable data
to check it with. There are just too many ways to make ray-tracing errors.
Brian
_______________________________________________
_______________________________________________
TowerTalk mailing list
TowerTalk@contesting.com
http://lists.contesting.com/mailman/listinfo/towertalk
_______________________________________________
_______________________________________________
TowerTalk mailing list
TowerTalk@contesting.com
http://lists.contesting.com/mailman/listinfo/towertalk
|