IAP: Accuracy outbound from LOM

[drauk]

This question is about an instrument approach (in the US if it makes any odds), where a NDB is located at the outer marker.

The procedure shows the NDB is tracked inbound, from a nearby VOR with a prescribed course from the VOR. Due to the angle of intercept to the final approach and the proximity to the runway, an outbound turn is required at the LOM, followed by a procedure turn back to the final approach course.

With no DME available from the VOR to the outer marker the way to know you have reached it will be the swing of the ADF. So as it starts to move, you make a turn to track outbound. Given that this won't be all that precise one would likely end up slightly off the exact reciprocal of the final approach course.

My question is, to what extent is it necessary to establish the exact outbound course, which would involve figuring out which side of it you're on (not hard) and manoeuvring accordingly? In a way it doesn't much matter, because the procedure turn (a somewhat imprecise manoeuvre itself) comes at a self-determined point (within 10 miles of the marker, as per the approach plate) and you then have the guidance of the approach for the exact inbound course.

If there is a difference, which I assume theoretically there would not be, I'm interested to hear views on both a real-world scenario and that of an IR flight test.

For anyone with US plates available, the particular case I'm wondering about is ILS 27L to South Bend, Indiana.

[FlyingForFun]

I'm trying to visualise this.... not sure if I've got it quite right in my head (if there's an on-line copy of the plate that would be good!), but if I understand you correctly, isn't this the way that all procedural approaches were done before the invention of DME?

I would imagine that there are rules somewhere about how much contingency must be built in to various procedures and how this affects the MDH and also the required distance from any obstacles. No idea where to find this stuff, though. Anyone got any ideas?

FFF
--------------

[Chilli Monster]

There's no reason why you shouldn't be exactly on the outbound course before the procedure turn (and no reason for the instructor / examiner to expect you not to be within 5 degrees)

Why?

It's an ILS :) (Which you will already have set up and idented - you just need to follow the back course to the procedural turn point).

So - actually flying the procedure, do it as follows.

ADF tuned to 341 (Misha) track inbound. Watch VOR/ILS (Nav 1, tuned to 109.3) and wait for the needles to become live. Start the turn BEFORE overhead the LOM, approx 2 dots on the CDI, so that you turn straight onto the outbound leg (fine tuning the turn as you go round). Determine Procedure turn point by flipping NAV2 over to 113.7 (GSH) and ensuring you're at 2400' level just past 347 Radial (NORNA). Execute procedure turn, flipping NAV2 back to 115.4 (GIJ) at the same time so that you're ready for the Missed Approach (if it happens).

[bookworm]

drauk

Approach procedures in the US are designed in accordance with FAA Order 8260.3, called TERPS available from here (http://www.dot.state.fl.us/aviation/text_version/fedpub.htm) but very big (28 Mb). Outside the US, the international standards aare laid down in PANS-OPS vol 2 published by ICAO.

The ILS27L for South Bend (http://www.myairplane.com/databases/approach/EastCentral/SBN_ir27L.pdf) doesn't look particularly unusual. You are expected to track outbound on the localizer. If I read TERPS correctly, the primary protected area on each side of the localizer is 6 nm. The protection extends along the length of the PT (10 nm).

[drauk]

Bookworm and Chilli Monster, thanks guys. I realise it is a fairly standard approach - it was just the particular one I happened to be looking at.

Flying the localizer backcourse makes perfect sense and I can now see that doing so would make it pretty easy - as the turn at the LOM is approaching intercepting the localizer is straightforward and thus would give an accurate outbound course.

What if there was no ILS? Would you still go to the trouble of intercepting a particular outbound radial from the LOM, or just make the turn and sort it out on the way inbound?

FFF, bookworm has kindly provided a link to the plate. It actually didn't work for me and I had to search for KSBN and then select ILS 27L. There is nothing special about this approach really - as bookworm says there are rules which govern the approach design but I was just curious about how to actually fly it in practical terms.

[Chilli Monster]

What if there was no ILS? Would you still go to the trouble of intercepting a particular outbound radial from the LOM, or just make the turn and sort it out on the way inbound? With no Glideslope available the lateral profile stays the same, you just have higher minima and have to calculate your own descent, so it doesn't change that much.

With no ILS at all (Localiser as well as Glideslope) the approach just isn't available in its present form. Time to revert to the NDB/GPS 27L. It's similar, but now you're looking at 10 degrees either side of track if you can manage it. Again the trick for accuracy is anticipating the beacon overhead and ever so slightly 'cutting the corner', instead of flying through it and then trying to play 'catch up' with the outbound track.

[MasterCaution]

In the UK, is it required to be within a certain bearing of the outbound track on the initial approach before starting the descent to the base turn platform altitude (in a similar manner to the half-scale deflection or +/-5 degrees for final approach as per AIP ENR 1.5 para 3.15.1)?

MC

[drauk]

ChilliMonster, your point about cutting the corner is well received and that is what got me thinking about the outbound track in the first place. I flew it on a simulator whereby I had the distance to MISHA (the LOM) courtesy of a GPS with the fix selected. But without the GPS how do you anticipate the turn? The plate shows no distance from the VOR. The sensitivity of the ADF would increase obviously when station passage is close, but surely that isn't accurate enough to gauge when to start the turn?

I guess in that case we're back to just nailing the outbound course by adjusting once we're outbound, within the greater tolerance you specify?

[FlyingForFun]

Ah - makes a lot more sense now I can see the plate! And can I suggest that we use the NDB/GPS plate, since there is one, rather than use the ILS plate and pretend that we're not using the localiser?

Yes, anticipating the turn would be nice, but I agree it would be difficult (and, ironically, the more accurately you fly in to the beacon, the harder it will be to anticipate). But the procedure designers will build enough contingency into the procedure to allow for this, and I would imagine that the document which bookmark has linked to will specify how much (if I'm really, really bored I might even look for it later!)

Interesting to note that, on the NDB/GPS procedure, the minimum altitude at the LOM is 2100', quite a bit higher than you would be at that point on the ILS.

FFF
--------------

[bookworm]

drauk

You're expected to fly the outbound on the localizer, not the ADF. That gives you a little lead time as the needle comes alive, though the intercept angle is large.

If it's an NDB approach with similar patterns, you do have to wait for station passage. However in the UK (i.e. under ICAO PANS-OPS) a turn of more than 30 degrees directly on to the outbound would not be permitted, and you'd have to go once round the hold. I don't think that's the case in the US.

MasterCaution

Good question. Where the level at the IAF exceeds the sector safe altitude, I've always felt justified descending to the latter, which usually gives you a minute or so to get to the point where 5 degrees is wider than your cockpit.

However, PANS-OPS requires that you are established (or +-5deg) before descending only for the inbound leg. I wouldn't want to be too far off though! I don't have PANS-OPS Vol 2 so I can't tell you how large the protected area is. Maybe one of the procedure experts could help. I'll ask on Tech Log.