Circling Approach Minimums - Descent Rate Constraint

[Keith Smith]

There is absolutely no guarantee of being able to execute a straight in descent from the MAP at the MDA on non-precision approaches.

That's the reason I prefer the "dive and drive" method!

I'm a dive and driver, too. There is new technology with WAAS GPS receivers that will calculate an advisory glideslope to give you a constant gradient descent, rather than diving between stepdowns, but I have to say, dive and drive isn't as difficult or dangerous (in my mind) as it sounds. You're just alternating between two configurations, descending and leveling out. It's more work, yes, but not a crazy amount of work.

The other issue I'm potentially not wild about with the advisory glideslope is that it may be a non-standard gradient (ie, shallow), not associated with a typical 3 deg GS that you see on precision approaches. The power settings that you normally use on a precision approach may not work well with the advisory glideslope model. I'm completely theorizing here, btw, I've never flown with a WAAS receiver or an advisory glideslope, have only read about them. Would be great to discuss with those that have used them!

My 'dive' in drive & dive is not overly aggressive, btw. I know that the plane could descend way steeper than the approach actually requires to hit the stepdowns, so I don't dive particularly hard. Power idle, or close to it, and let the plane fall at the currently trimmed speed, but nothing more aggressive than that. It does the job.

The only part that requires serious attention is the level out at MDA. It's critical to add power, otherwise, you're either going to get slow, or low, neither of which is good for your long term health.

[Peter Grey]

I guess I'll be the dissenting opinion.

Additionally, for no extra charge I'll be the person who can cover this area as well:

I'm completely theorizing here, btw, I've never flown with a WAAS receiver or an advisory glideslope, have only read about them. Would be great to discuss with those that have used them!

I am opposed to the dive and drive method of handling non precision approaches, here is why:

Let's assume the goal of the dive and drive is to get to MDA as soon as possible to allow the maximum time to possible see the runway at MDA. We will also assume the weather is at (or near) minimums (otherwise the MDA portion quickly becomes irrelevant). I'll use the VOR 25L approach into LAS for my example (because it's a very commonly used approach in the real world).

We will run through this approach a couple of times, the first time we are going to say the weather is exactly at visibility minimums 3/4.

So your magic airplane instantly drops down to 2720' at 8.1 DME from LAS and you see, well nothing that matters. You continue forward and 1 mile from the runway (2.7 DME) you see the approach lights meaning you can descend down. Now let's look at some gradients here. You are 7000' from the aiming points and 9500' from the end of the touchdown zone. This means you will need to descend at 558-441 ft/nm to make the touchdown zone.

Is this doable, yes technically it is in most GA aircraft, is it a standard descent gradient for most people in a GA airplane, no. These numbers also ignore the need to flare (trust me you'll want to do that). Would doing so comply with 91.175? (maybe, honestly most likely yes). This means you'll be descending at the runway at a rate you are not used to (in poor wx no lest), this can be bad.

Ok so you very quickly determine that no, this isn't going to work, so what can you do? Head down anyway and land past the touchdown zone (at LAS, sure you'll most likely be ok, other places, no)? Missed approach (Never can go wrong with a missed)? or Circle?

So let's talk the circle. The circling MDA is 3080 (or 360' above you). Anyone know how many feet clearance you get at MDA? It's 300', so not only are you below circling minimums, you are BELOW OBSTACLES for the circle. So we can agree that attempting to salvage the approach with the circle is bad.

So what's the summary here, there is a point on the approach where continuing on has no purpose except to get you in trouble (either by excessive descents, or bad attempted to salvage a circle out of the approach).

Ok, opposite scenario the ceiling is bad but the visibility is unrestricted. So there is a 651' ceiling (well 652'). So our magic airplane gets down to 651' at 8.1 DME and woo there is a runway. Ok so let's start down, we descend down nice and slow (hey so in case the clouds are down low we don't need to worry about them). Get to 300' low, and are no longer in protected airspace, ok no big deal we can see everything. So no real risk here, well unless there is a ground fog layer, or everything closes in nice and early, and once again we are below MDA with no protection and no sight. That is bad, unlikely? yes, but possible. So there is no real benefit to getting to MDA early,

So what's the answer if we know that being at MDA too early can be bad, and too late can also be bad, we should define an "ideal" point to be at MDA. This point is called a VDP (visual descent point). On newer charts you'll see it as a V on the profile view. Assuming we want a 3 degree glideslope (the standard) this point is calculated as:

(HAT of MDA)/300 = NM from the runway the VDP is.

In this case you get around 2 NM (you can round). What does this mean?

Well you should ideally see the runway 2 NM away to be able to do a normal descent to the runway. Does that mean go missed at 1.9NM away? No for most GA aircraft, but it means you need to start to be aware that you won't be doing a standard descent to the runway and be aware of the risks thereof. Note if you do go missed here be aware to NOT turn for the missed until you are past the MAP.

Ok so how do I prevent getting to MDA early. Well you can do some math again and determine a descent rate to be at MDA at the VDP. The formula is:

(altitude to descent from FAF to MDA)/distance between FAF and VDP * GS/60 = ROD

Assuming a 90 knot GS you get an answer of around 450 ft/min (once again rounding is 100% ok), which as a fun fact is a 3 degree glideslope (hmm, like they almost planned it that way {they do, unless it doesn't work at 3 degrees}).

If you combine the concept of a VDP and this constant descent rate, you'll notice in this case you in theory set a 450 ft/min descent at the FAF, make sure you see the runway at MDA, and then flare at whatever altitude you flare at with no change all the way down.

This leads us to the WAAS advisory glideslope in some modern GA aircraft:

I'm going to reference the G1000 advisory glideslope (LNAV+V) not to be be confused with LPV (which is for all intents and purposes a precision approach) Which simply put does all the math above, and provides an "advisory" glideslope to the runway, normally it's a 3 degree glidepath, but it doesn't have to be.

I'm going to leave it there for now (you could talk for hours on this one) and see where the discussion goes.

[Mark Hargrove]

Peter,

Your post was an very interesting read!

Given all the information in your post, it's got me wondering what the FAA's rationale is for the location of MAPs for non-precision approaches. Acknowledging (as you did in your post), that perhaps they're not actually dangerous, it sure seems like they're on the hairy edge of being so. Why isn't something like the VDP used instead of the MAP as specified now?

-M.

[Peter Grey]

Why isn't something like the VDP used instead of the MAP as specified now?

Note in the airline world (I fly a CRJ for a large regional airline) we basically do (we call it a DDA, but it's the same basic principle).

It's really simple actually.

1 is for circling approaches, if you are at circling MDA you want to get as close to the airport as possible on the approach to be able to do your circle.

2 Instrument approaches must be built for the lowest common denominator aircraft, which is an aircraft with 1 VOR and no DME. A VDP cannot be realistically calculated in this case so as a result you have several limitations on what you can do to build an approach. With no way to determine distance you need either a navaid for the MAP (which needs to be on the airport then {so you can use it for multiple runways}), or the use of a clock to determine distance (which is about as accurate as you can guess). Method 1 is preferable to method 2.

Obviously the second reason doesn't apply to a lot of approaches (anything with DME, RNAV, or GPS in the title), but the first applies to most (if not all approaches).

[twharrell]

Peter, your explanation was very clear and concise - a very interesting and educational read. This, to me, is one of the great values of Pilot Edge that can easily be overlooked. Thank you Keith and Peter.

Todd

[Keith Smith]

Speaking of lowest common denominator, also keep in mind that helicopters can shoot these approaches. Calculating a one-size-fits-all VDP is not really possible when you consider that helicopters can also shoot the approach.

The MAP is fixed, but the VDP is going to depend on the capabilities of the airplane/helicopter that's shooting the approach.

[Peter Grey]

Speaking of lowest common denominator, also keep in mind that helicopters can shoot these approaches.

True, and you can't account for the ability to hover at the MAP until your rotor wash clears out enough to see the runway and land.

<disclaimer>I know helicopters can't actually do that</disclaimer>

[twharrell]

Speaking of lowest common denominator, also keep in mind that helicopters can shoot these approaches.

True, and you can't account for the ability to hover at the MAP until your rotor wash clears out enough to see the runway and land.

<disclaimer>I know helicopters can't actually do that</disclaimer>

Actually, the downdraft from the rotor would push the ceiling down further, ruining it for everybody. ; b

Todd

[Peter Grey]

If you ignore the adiabatic heating from the descent it would, but otherwise in theory the cloud should dissipate from the pressure based compressional heating effect (after the pressure equals out from the draft itself).

Of course none of this really matters

[twharrell]

If you ignore the adiabatic heating from the descent it would, but otherwise in theory the cloud should dissipate from the pressure based compressional heating effect (after the pressure equals out from the draft itself).

Of course none of this really matters

True, but would that necessarily overcome the negative pressure (i.e. vacuum) created above the rotor, which would effectively pull the cloud layer downward. We might be kicked off this thread if we keep this up (or down, per our discussion).