Intersting read about the Glide Slope and how it works IRL...

[cagarini]

Post at AVSIM's Hangar Chat forum


BTW, is this effect also modeled in PSX ?

[John H Watson]

I think you'll find that the real G/S is calibrated to show the most accuracy over the threshold, which may involve a slight adjustment of the angle of the beam overall.

Photos of glideslope towers appear to show 3 radiating elements at various hights on the tower. At what height above the ground would the beam begin?

[cagarini]

From one of the tests run with FSX, mentioned on the above AVSIM link:


"What would have a much greater effect is the height of the glideslope beacon. I guess the cone's apex is on the middle antenna, which is about 30 feet above the ground, so the entire hyperbola is that much higher again. That would put the GS/LOC flightpath at 104 feet over the threshold, quite a difference from the original sixty-five."

[Hardy Heinlin]

Sure, this geometric effect is modeled in PSX as it's simply a side effect of the ground station being off the runway. On the Instructor, on the Position map, you can see the G/S stations, they are shown as crosses (LOC stations as circles, DMEs as rectangles); and on the Profile page, when tracking the runway centerline, you can see the magenta G/S plot gets more and more curved when getting more and more abeam the G/S station.

[cagarini]

Thx Hardy.

[G-CIVA]

Of course the premise of their whole discussion over there in FSX Land is that EVERY G/S antenna inside FSX/P3D is located correctly in relation to every runway & vice versa - in the same way that every visual approach path indicator should be - which they are not - even in many a payware scenery.

Users of Gary's suite of software marrying up PSX with FSX or p3D will immediately know the real answer & therefore the huge assumption that they are all making about testing in 'stock FSX'.  How do they know that the G/S beam in their example is in the correct position in the first place?  A huge assumption.

Remember even the 'guts' of something as new as p3Dv3 is still FSX - when I say this I mean much of the data like airport layouts including navaids & runway assignments etc.

Interesting piece all the same.

[Hardy Heinlin]

However, Jim has made a good point ...

I guess it's a mix of both models. If the beam is a disc rather than a cone, the G/S angle would be 0° when abeam the station. That is, it can't be a constant 3° slope in that case either. It can't be 3° within +/- 89° of the front course, and then suddenly 0° at 89.9°. So, geometrically, in the disc model, when approaching the threshold, the G/S needle goes down, not up. The curve goes down in that case. -- Edit: The curve appears only when referring to the station, but that's irrelevant; the path to the threshold remains straight, and that is what we want. -- I think the actual cause of the needle going up lies in radio interference near the ground, not in the geometry. As the needle movement is very chaotic near the ground in PSX and real life (crossing two beams actually), there is no need to modify the station model.

[Hardy Heinlin]

I just added a cosine function so that the lateral G/S beam shape is not conic but plane. That is, the beam is now a straight line going into the asphalt rather than a parabolic curve going up over the touchdown zone.

I bet nobody will notice the difference during the flare. The autoland flare guidance ignores the G/S anyway. Abeam the antenna, over the touchdown zone, the pointer is now just a tiny fraction of a dot lower than in the previous versions, and everything's going very fast.


|-|ardy

[G-CIVA]

I shall try to look ... I am due my quarterly PSX AUTOLAND practice soon anyway 

[Hardy Heinlin]

Now available in PSX 10.0.9-beta12:

http://aerowinx.com/board/index.php?topic=3685.0

[United744]

FSX glide slope is CURVED ON PURPOSE in the vertical axis, as a hack to enable AI to land!

If you fly correctly in FSX, the GS will slowly rise as you approach the TDZ (this is the artificial curve). If you try and keep it centered, you will end up flying level.

FSX/P3D can NOT reliably used for any determination of actual behaviors for ANYTHING.