A glideslope / localizer actually has some similarities in the way it works to the old low frequency "AN" 4-lobe radio ranges that were the primary means of aircraft navigation in the 1930's through 1950's, before VOR.

I don't remember which sim platform it was, but I definitely remember that the old-fashioned "AN" radio ranges were modeled. It's been a very long time ago and my memory isn't what it used to be.

I think you reasoning is incorrect. Provided that the surface described by the glideslope is a cone (and, as you confirmed, the glidepath is linear, so in the space it should describe an arc of cone), the described effect will happen.

The glideslope is not a cone. Rather, it is defined by the intersection of TWO cone-shaped beams, and further defined NOT by the intensity of the radio frequency energy in each beam, but by the amount of modulation applied to the RF carrier waves being emitted by the antennas.

 

A plane surface, derived from the intersection of two cones, does not itself curve.

 

This system for deriving the glideslope (and localizer), from the intersection of two discrete beams was chosen when ILS systems were first designed, specifically to insure course linearity at all times.

 

Now, there IS an effect, that José "The Uninstaller" mentioned in which the angular relationship between the glide path center and the aircraft's position in space can vary at extremely close distances from the glideslope antenna array, (if the aircraft if far off the centerline), but that effect occurs precisely BECAUSE the glideslope is "flat".

 

Practically speaking though, to be of any consequence it would require that the aircraft be so far offset from the center of the localizer, while very close to the runway threshold (i.e. a full-scale localizer defection at minimums) that the pilot should have declared a missed approach long before the effect coukd become an issue.

 

Here in the U.S., the "flatness" (i.e. linearity) of both the localizer and glideslope is something that the FAA tests rigorously with highly instrumented "Flight Check" aircraft before any new ILS can be certified for use, and after any repair or modification of the transmitters or antenna systems of an existing ILS. This is especially true for ILS systems certified for CAT II or CAT III

I don't remember which sim platform it was, but I definitely remember that the old-fashioned "AN" radio ranges were modeled. It's been a very long time ago and my memory isn't what it used to be.

I believe there were still a few A/N ranges operational in the US as late as the early to mid 1960's.

 

Now that you mention it, I believe there WAS a scenery / gauge combo (perhaps for FS9?) that would give the ability to set up an in-sim A/N range.

 

I also saw an article once in QST within the past couple of years, where a fellow Ham had built a scale-model A/N range which operated in the 2 meter band. There might even be a YouTube video on that one...

Now, there IS an effect, that José "The Uninstaller" mentioned in which the angular relationship between the glide path center and the aircraft's position in space can vary at extremely close distances from the glideslope antenna array, (if the aircraft if far off the centerline), but that effect occurs precisely BECAUSE the glideslope is "flat".

 

That is exactly the effect we were talking about.

 

The glideslope antenna is _NOT_ on the centerline of the runway, but it is displaced to one of the sides of the runway. So, if the aircraft is following the localizer (and hence moving directly towards the localizer antenna), it is at the same time moving _ASKEW_ from the glideslope antenna. This is the reason why the glideslope appear to be an hyperbole and not linear. It is because the aircraft is actually not moving towards the actual linear glidepath, but is going askew of it.

 

Some references:

 

https://etd.ohiolink.edu/rws_etd/document/get/ohiou1174609165/inline :

 

page 18: "The mast must be offset from runway centerline so that it is not an obstruction to aircraft. The shape of the ddm contour intersecting the vertical plane containing the runway is then an hyperbola."

 

http://moscom-e-learn.blogspot.it/2011/04/gilde-path-or-glide-slope-gp-gs-signal.html :

 

"The plane DDM 0 radiated by the glide path antenna is hyperbolic and does not touch the ground, as the dotted line shows. [...] this produces an offset of the glide path antenna mast with respect to the runway threshold of the distance D. This offset is 286 to 344 m depending upon the glide path angle selected. Due to this the optimal vertical glide path is not a straight line in azimuth direction of the centre line of the extended runway, but is a hyperbola."

 

http://www.dtic.mil/dtic/tr/fulltext/u2/a034492.pdf

 

page 24-25: "The actual ideal surface of zero CDI is a hyperboloid of two sheets whose axis of rotation is parallel to the z-axis and passes through the antenna. For comparison purposes it is convenient to have the aircraft travel along this surface and see how the real CDI deviated from 0. If the glide path is used for linear flight in the near field (between threshold and antenna) large CDI's will occur because of the hyperbolic shape. The program will allow the user to gererate a hyperbola which is the intersection of the 0 CDI surface and the plane containing the runway centerline parallel to the z-axis."

I don't deny that the shape of the radiated antenna pattern is hyperbolic (curved) very close in, it has to be, since it begins at a what is essentially a point source at the antenna itself, and "spreads" outwards from there but the curvature is in the horizontal plane. The "skew" effect (caused by the antenna being offset from the runway) only occurs to a noticeable degree within the near field of the antenna - which is considerably less than the distance from the antenna to the threshold.

 

For an antenna with an effective aperture of 5 meters (typical for a glideslope array) at a operating frequency of approx 330 MHz, the radiated pattern transitions to the far field (where the radiated wavefront goes from being curved to being essentially flat) at approx 200 meters from the mast, or about 650 feet. On a Cat I, approach, a pilot is going to be fully visual and in the process of beginning to flare before he enters that region.

 

That is why the autopilot of an aircraft equipped to perform Cat III autolands smoothly transitions from using the ILS glideslope, to using the radar altimeter to control vertical path in the final moments of an approach, because the glideslope becomes unreliable as the aircraft gets very near to the GS antenna.

 

However, for all intents and purposes, the glideslope center, (where the two modulating waveforms are of equal amplitude) at distances beyond the antenna near field, can be considered to be angled flat sheet, whose height above the surface is essentially constant along a line running at right angles to the runway. To see any substantial glideslope deviation caused by horizontal offset from the runway centerline at points much beyond the runway threshold, would require that the localizer be completely off-scale, (which is not a stable or useable approach). At a mile out, the deviation would not be detectable unless one were approximately two thousand feet off centerline.

 

We may all be looking at the same point from different aspects... and probably boring most readers of this thread silly in the process.

Ok, and I just confirmed Aerowinx models it too.

 

Hardy actually posted this at their forum.

 

I now just have to try with DCS World :-)))), but I'm afraid I can't try with il-2 Battle of Stalingrad :-(

 

 

and probably boring most readers of this thread silly in the process.

 

By all means, keep discussing about this! I'm enjoying the conversation indeed!

I confess I learned something new for me from this thread, and I blame myself for never having thought about it and the reason for some effects, specially when flying VNAV approaches !

 

Any contribution will enrich our knowledge, and each one can read up to where we can understand, and try to find more about what we do not understand because it's too technical.

 

And, btw, great thread Jaime!

We may all be looking at the same point from different aspects...

 

Yeah I think we all agree. When following the localizer, the glidepath is actually a hyperbola. However, the non linearity is only evident when very close to the runway threshold. For the approach phase, it can be considered to all effects linear.

I confess I learned something new for me from this thread, and I blame myself for never having thought about it and the reason for some effects, specially when flying VNAV approaches

Yes! With the proliferation of WAAS-enabled GPS LPV approaches, you can have all the benefits of ILS precision, without the need to maintain ground-based equipment, with fussy signal strength and antenna pattern calibration requirements.

 

Of course, that's all relative, since the technical requirements of building, launching and maintaining the GPS satellite constellation are extreme indeed!

 

I have always thought that the government's decision to decommission the old LORAN system was a mistake. Though GPS is fantastic in the benefits it offers (I use it every time I play golf) I worry that a "thousand year" solar flare is going to erupt some day that might take the whole system out.

 

LORAN.... now THERE was a navigation system that was ALL about hyperbolas in the way that it calculated position!

 

 

I worry that a "thousand year" solar flare is going to erupt some day that might take the whole system out.

 

But think of it this way, Jim:

 

The day of a Fallout (good game by the way) and going back to being cave men, we will have to start from scratch all over.

 

Or is the average Joe capable of making a fire in the wild without a lighter? hahaha, no way!

 

just a fun offtopic comment

And that great MLS system, which was way better than the one which ended up being adopted .... A lot less prone to interferences and other problems, and even less expensive to explore...

 

 

that great MLS system

 

The MLS is the "Betamax" of landing systems, hahaha

 

One thing that intrigues me, and I would ask a real pilot to confirm: Is the GLIDESLOPE indicating BELOW glidepath when the aircraft is on the runway and ABEAM the glideslope antennas? Theoretically, it should.

It does not, the GS fluctuates and the GS flag is displayed just prior to passing abeam the GS antenna. Although I am not custom to watching instruments when landing, but paid attention today.