Vertical path non-precision approach safety?

[PeanuckleCRJ]

A VNAV (managed descent/approach nav in airbus land) approach requires a single button push and will respect all altitude constraints on an approach. Far far safer and much more stable than all the twisting pushing and pulling needed for dive and drive. Garbage in/garbage out, of course, on the constraints. Always verify the programming and monitor it (be a pilot) to make sure the hamster hasn't fallen off the wheel.

I always go to north plan view with constraints pulled up and step through it all comparing with the chart.

And VNAV (or equivalent) is certified on all DL aircraft. One interesting caveat here is, on the buses at DL, we switch and fly a flight path vector FAF inbound on a LOC only approach. We still reference the path to help fine tune, but we utilize FPA mode as the vertical mode.

 

[derg]

I'm going to be a witch in church on this one and ask, once again, what was ever wrong with "dive and drive"? If you have legit, certified, copper-bottomed vertical guidance (ground based or otherwise), use it, obviously. But if you're flying a non-precision approach with no vertical guidance, why guesstimate and hope that the uncertified magic will get you where you want to be? That's an awful lot of button pushing and trust up against little things like "levelling off" and "moving the thrust levers", at least to my way of thinking. Non-precision approaches were designed with this in mind. Planes haven't gotten a whole lot bigger or harder to fly. Edumacate me.

Well, vectors to the final approach, in my plane, hit LNAV or LOC, hit VNAV and just remember to click off the automation.

It's really as simple as an ILS approach

 

[Boris Badenov]

I suppose that my concern is that, humans being what we are, we're liable to start thinking of an electronic "pseudo-glideslope" on a pure LNAV approach as being the same thing as an LNAV/VNAV approach. In my brief career flying as a "C/A" on a "crew" airplane, I can't count the number of times I saw the other guy pretty much throw away the TERPS once the "donut" appeared. If my very rusty "aviationese" is correct, that's called "inversion". Now, sure, you could have the TERPS open and be checking the cross radials, DME, etc etc, but how common is that, really? And if you aren't, you're basically just assuming that a huge litany of computer programmers never transposed a digit or "one zero".

Also, and as a pure aside, I've "ducked under" on "dive and drive" approaches more than once where I'm 100% sure I wouldn't have gotten in on a "pseudo-glideslope" approach. What's so hard about getting to the next step down, levelling off, and prepping for the next step down? Dudes were doing this stuff 50 years ago in planes that go just as fast as ours do now. Sometimes faster!

 

[PeanuckleCRJ]

On the CRJ the "snowflake" derived vertical path is to the runway pavement beginning, not to the touchdown zone. So if you fly it exactly centered up, you'll be below glide path.

Override the elevation and add 50 feet to it. Tada! A snowflake you can use!

 

[dasleben]

I suppose that my concern is that, humans being what we are, we're liable to start thinking of an electronic "pseudo-glideslope" on a pure LNAV approach as being the same thing as an LNAV/VNAV approach. In my brief career flying as a "C/A" on a "crew" airplane, I can't count the number of times I saw the other guy pretty much throw away the TERPS once the "donut" appeared. If my very rusty "aviationese" is correct, that's called "inversion". Now, sure, you could have the TERPS open and be checking the cross radials, DME, etc etc, but how common is that, really? And if you aren't, you're basically just assuming that a huge litany of computer programmers never transposed a digit or "one zero".

Also, and as a pure aside, I've "ducked under" on "dive and drive" approaches more than once where I'm 100% sure I wouldn't have gotten in on a "pseudo-glideslope" approach. What's so hard about getting to the next step down, levelling off, and prepping for the next step down? Dudes were doing this stuff 50 years ago in planes that go just as fast as ours do now. Sometimes faster!

We get it. You're a freight dog who won't give up the ghost. It's over.

Yes, these jets were certified to do LOC/VOR/NDB/SDF approaches in LNAV/VNAV 30 years ago. We don't even have a way to track a VOR course with the autopilot. You check what's in the FMC as part of your brief, and ensure that everyone is on the same page in terms of which mode the approach will be flown. It is much safer than a dive and drive, and it's not close.

 

[derg]

We get it. You're a freight dog who won't give up the ghost. It's over.

Yes, these jets were certified to do LOC/VOR/NDB/SDF approaches in LNAV/VNAV 30 years ago. We don't even have a way to track a VOR course with the autopilot. You check what's in the FMC as part of your brief, and ensure that everyone is on the same page in terms of which mode the approach will be flown. It is much safer than a dive and drive, and it's not close.

That was interesting in 767 school:

"Where's the VOR button?"

I guess you can raw data it and do the old 737-200 thing and either hand-fly it or heading select, but...

 

[Boris Badenov]

You check what's in the FMC as part of your brief, and ensure that everyone is on the same page in terms of which mode the approach will be flown. It is much safer than a dive and drive, and it's not close.

None of that is different from the Bitchjet, systems-wise. Certainly there's more inertia involved, but it's not like I haven't sampled the wares of the "All Approaches Must Be On A Three Degree Glideslope From 30,000ft" mafia. I'm going to go out on a limb and guess that you've got some kind of manually tuned and dialed backup ILS/VOR/DME receiver (the "Green Monster" on the Bitchjet). They're all tools. Why wouldn't you have the cross radials tuned on the backup and the TERPS open? I submit to you that "briefing" the FMC and hoping it's right is a lot MORE dangerous than a well-executed "dive and drive" by a competent and awake crew. I AWAIT YOUR RESPONSE, SIR. PURSES AT DAWN, TWO PACES?

 

[typhoonpilot]

I don't have access to this full article. Anyone with an IFR subscription care to post it?

I bring this up because I've always been trained that following the FMC or GPS derived vertical path will keep you safe to 50 feet above the touchdown zone. Not long ago I was doing a day VMC visual backed up with the localizer into an airport I can't remember. I've gotten used to using the path indication as a substitute for an electronic glide slope in the rare occasion we don't have a full ILS. I tend to crosscheck the outside runway environment picture with the inside VNAV path and VSI descent rate, especially between 500 feet AGL until and I start looking solely outside to begin the flare. So, there I was on path, at about 200 AGL, and my F/O calls out "red over red, a little low". I looked out and sure enough, I was red over red on the VASI and I needed to make a correction. I looked back inside for a flash to confirm my VNAV path and it showed me dead on the proper path while the VASI showed me low. I was a surprised by this but just passed it off to me being an idiot who needs to get out more.

Then, we lost an A300 in BHM on a night non-precision, and it kind of got me doing some research. I clarified what our training guide says, which paraphrased is "use the VNAV path as a reference but outside visual contact with the runway must be maintained". I'm starting to wonder, though, if the VNAV path will keep you out of the trees on runways with high terrain on final approach. I'm thinking that the VASI or PAPI will be all that's keeping you safe and perhaps the path should be ignored, or certainly considered a secondary indication, once you are below the MDA.

Just wondering what others have experienced.

Mentioned once in the replies is the effect of temperature on vertical paths. Where I fly the temperature is often ISA + 20C ( +36F ) or more. With that kind of a temperature the effect on the FMS vertical path is quite significant, it amounts to about a 0.3 degree change to the FMS derived glideslope. With high temperature like that it increases the glideslope angle so one would show high on the PAPI, usually three whites and a red. If the temperature were colder than ISA then it becomes a bit more dangerous since the glideslope is now shallower than normal and one would appear low, three reds and a white.

So I would ask, in your example, what was the OAT?

Electronic glideslopes, i.e. an ILS, do not have this problem since they are fixed in space. When flying an electronic glideslope on a 3 degree slope with ISA deviations one notices the OM crossing height (on the aircraft altimeter) as being different than published. It reads lower when OAT is above ISA and it reads higher when OAT is below ISA. Hopefully everybody makes cold temperature altitude corrections to DA once the OAT gets below a certain threshold ( 0 degrees C or 32 F for us ).

Another problem of note for an FMS glidepath is that it may end before the runway. This happens when a non-precision approach is only coded to a missed approach point and not the runway. Stick the LOC 18 at BHM into your FMS and see where it is coded to, 50 feet above the runway or the missed approach point only.

Theoretically, on my aircraft and if on autopilot, the aircraft would level off at an MDA coded like this. In the simulator I have seen it tuck under and continue descending at a faster rate.


Typhoonpilot

 

[dasleben]

None of that is different from the Bitchjet, systems-wise. Certainly there's more inertia involved, but it's not like I haven't sampled the wares of the "All Approaches Must Be On A Three Degree Glideslope From 30,000ft" mafia. I'm going to go out on a limb and guess that you've got some kind of manually tuned and dialed backup ILS/VOR/DME receiver (the "Green Monster" on the Bitchjet). They're all tools. Why wouldn't you have the cross radials tuned on the backup and the TERPS open? I submit to you that "briefing" the FMC and hoping it's right is a lot MORE dangerous than a well-executed "dive and drive" by a competent and awake crew. I AWAIT YOUR RESPONSE, SIR. PURSES AT DAWN, TWO PACES?

I can't say that I've ever seen anyone monitor cross radials on one of these approaches, but it's not a terrible idea.

What are the "TERPS" you're referring to? We have our charts out, and crosscheck the chart with the FMC before commencing the approach. We also monitor raw data on the RMI.

Wait...you didn't think we just plug junk into the FMC and blindly fly it, right? You do realize that we're professionals?

 

[Boris Badenov]

What are the "TERPS" you're referring to?

Terminal Instrument Procedures. Commonly referred to as "plates", I think. I do not see the logic in relying on a computer to fly you down a glideslope that does not exist, is not certified, and exists only in the manufacturer's (or their sub-contractor's) database in lieu of looking at the chart which is certified, is regularly tested by statute, etc. The magic tells you stuff. Obviously, most of it is true, 99.99% of the time. But. The plate tells you things that you Know.

 

[BobDDuck]

Terminal Instrument Procedures. Commonly referred to as "plates", I think. I do not see the logic in relying on a computer to fly you down a glideslope that does not exist, is not certified, and exists only in the manufacturer's (or their sub-contractor's) database in lieu of looking at the chart which is certified, is regularly tested by statute, etc. The magic tells you stuff. Obviously, most of it is true, 99.99% of the time. But. The plate tells you things that you Know.

I'm not following you. The glideslope that does not exist really does not exist. It's simply the computer calculating a 3:1 (or whatever) that will get you to the minimums of the approach at the VDP while still respecting all the stepdown fixes published on the chart. Nothing changes other than that. You are still flying each of the stepdown fixes as you go (as referenced on the chart and loaded into the FMC) but the computer has figured out the descent rate for you.

 

[Boris Badenov]

You are still flying each of the stepdown fixes as you go (as referenced on the chart and loaded into the FMC) but the computer has figured out the descent rate for you.

Garbage in, garbage out. I guess I'm not understanding you. What conceivable reason would there be not to have the cross-radials tuned and identified? I dunno, maybe I'm a Nerd, but I tune every navaid I can on every approach, and identify them, too. I even turn the marker beacons on. What, are you really that freaking busy? I Trust, sure, but you'd better believe that I Verify. I'd rather skip a few steps on the 50 item checklist and make absolutely sure that I know where I am in three dimensional space than have the landing lights out at the right altitude so they can illuminate the mountain I'm about to fly in to. *shrug* Is the gear down? Are the flaps set? AM I ABOUT TO HIT ANYTHING? Checklist complete!

 

[dasleben]

Terminal Instrument Procedures. Commonly referred to as "plates", I think. I do not see the logic in relying on a computer to fly you down a glideslope that does not exist, is not certified, and exists only in the manufacturer's (or their sub-contractor's) database in lieu of looking at the chart which is certified, is regularly tested by statute, etc. The magic tells you stuff. Obviously, most of it is true, 99.99% of the time. But. The plate tells you things that you Know.

Next time you have those charts open, look at what it's really telling you (ignore the visual chart):

(Yes, the 3.14 degrees is built into the FMC, and in any case, the VNAV will not descend below the stepdown altitudes at CAPIT, WUGAL, RUSHY, or CEMUG until those fixes are passed).

 

[Boris Badenov]

I'm not seeing the McGuffin here. It seems like flying that approach in "dive and drive" mode would be just as safe as flying it in "sorta VNAV" mode, to me. Conceivably...dare I say it...safer?

 

[derg]

Garbage in, garbage out. I guess I'm not understanding you. What conceivable reason would there be not to have the cross-radials tuned and identified? I dunno, maybe I'm a Nerd, but I tune every navaid I can on every approach, and identify them, too. I even turn the marker beacons on. What, are you really that freaking busy? I Trust, sure, but you'd better believe that I Verify. I'd rather skip a few steps on the 50 item checklist and make absolutely sure that I know where I am in three dimensional space than have the landing lights out at the right altitude so they can illuminate the mountain I'm about to fly in to. *shrug* Is the gear down? Are the flaps set? AM I ABOUT TO HIT ANYTHING? Checklist complete!

Probably "AUTO".

The way the FMC works in my airplane is that it's looking at three GPS receivers, three IRUs and multiple DME-DME and I-DME's to triangulate where you are, usually with an ANP of 0.02 or less which is completely funky fresh for a NP approach.

Once you go into "manual" to tune another VOR, it "somewhat throws it off".

 

[dasleben]

I'm not seeing the McGuffin here. It seems like flying that approach in "dive and drive" mode would be just as safe as flying it in "sorta VNAV" mode, to me. Conceivably...dare I say it...safer?

The disconnect is that you're still assuming it's all "sorta VNAV." It's full VNAV, the jet and approach are certified for it, and it's appropriately charted.

 

[Boris Badenov]

Probably "AUTO".

The way the FMC works in my airplane is that it's looking at three GPS receivers, three IRUs and multiple DME-DME and I-DME's to triangulate where you are, usually with an ANP of 0.02 or less which is completely funky fresh for a NP approach.

Once you go into "manual" to tune another VOR, it "somewhat throws it off".

Well, I'll just say that I think that you're a trusting soul. There are worse things to be, for sure.

 

[Boris Badenov]

The disconnect is that you're still assuming it's all "sorta VNAV." It's full VNAV, the jet and approach are certified for it, and it's appropriately charted.

I was under the impression that we were talking about an LNAV approach with vertical guidance from the FMC (or FMS, or whatever silly acronym we're using this week for the computer), not an LNAV/VNAV approach.

 

[BobDDuck]

The disconnect is that you're still assuming it's all "sorta VNAV." It's full VNAV, the jet and approach are certified for it, and it's appropriately charted.

Yeah. In his defense, there are airplanes (the CRJ for one) that have "advisory" VNAV that allow for a sorta VNAV (flown in vertical speed mode).

But in an aircraft with VNAV or PROF or something like that... I don't see the issue.

 

[BobDDuck]

I was under the impression that we were talking about an LNAV approach with vertical guidance from the FMC (or FMS, or whatever silly acronym we're using this week for the computer), not an LNAV/VNAV approach.

Where do you think the VNAV portion of the guidance comes from on a LNAV/VNAV approach?