Behaviour of VNAV vertical deviation indicator (on ND) when approaching T/D

[Hardy Heinlin]

Good morning,

as we all know, when approaching the predicted VNAV path from below (like approaching a G/S radio beam from below), the needle is inititally above us and then moves down towards the center of the scale. The descent starts when the needle is close to the center.

This behaviour can be seen during an early descent when approaching the predicted path from below at -1250 fpm, or after a path modification that puts the aircraft below the new path which will be recaptured at a shallower descent rate, or when on a level-off segment shortly before a VNAV guided final approach starts.

Now the question is whether this behaviour, on the real aircraft, can sometimes be seen at CRZ ALT as well, shortly before a normal idle descent starts.

Or will the needle be indicating the deviation from the CRZ ALT until the aircraft has reached the point where the actual descent starts?

I'm asking because I have received different pilot reports in the past years. Maybe the behaviour depends on certain system versions, or is differently modeled in the big sims (e.g. a difference between Atlas Air sims and BA sims?), or it depends on whether the command airspeed in DES is higher or lower than that of the CRZ, or the memory wasn't quite correct (maybe thinking of an early descent rather than of a normal one) ...?

I think we all agree that when an early descent has just been started, e.g. 30 nm before T/D, and we're only, say, 100 ft below CRZ ALT, the needle will not indicate -100 (with reference to the CRZ ALT) but the needle will immediately go up way beyond -400 ft (with reference to the predicted descent path, extended from the T/D backwards up to the current aircraft position).

In older PSX versions, in the early descent phase, the needle was indicating the CRZ ALT deviation (getting greater) until passing the original T/D lat/lon position after which the needle indicated the descent path deviation (getting smaller).

In newer PSX versions, when the early descent phase is started, the needle immediately indicates the deviation from the backwards extended descent path. This modification has not been criticized so far; so I guess everyone is happy with that.

In the latest PSX version, however, this is also applied when starting a normal descent (from high altitudes only, not from a missed approach cruise). Unfortunately, this doesn't agree with the latest observations in the BA sim :-)

So that's the question.

(In this thread I'd like to focus that needle behaviour only. Just needle cosmetics. I know the transition from CRZ to DES is a very complex subject, but let's discuss other related topics in other threads.)


Regards,

|-|ardy

[Britjet]

From my experience, no.
CRZ is CRZ, DES is DES. Nothing "clever" happens.
This applies similarly to any difference in target speeds between the two modes.
At TOD point the FMC pages change from CRZ to DES, the command speed changes to DES command speed (not before), and the thrust levers move accordingly if it is a speed decrease.
The VNAV pointer appears - at 0 vertical error, although it may stay at 0 for some time if there is a deceleration phase.
I've never seen it do otherwise.

HTH
Peter.
(I note we have quite a few current 747 pilots on the forum and would welcome any input - preferably with a video?)

[Hardy Heinlin]

There is no doubt that the DES speed is only commanded when in ACT DES. No problem with that.

I'm just focussing on the needle cosmetics.

I'm just asking whether the needle should -- in all airlines and versions -- stay at 0 while the aircraft is maintaining the CRZ ALT until the predicted descent path is intercepted.


Cheers,

|-|ardy

[emerydc8]

Hi Hardy,

Just did this in the 767 (Pegasus) from T/D (FL340). As soon as the VDI appears you are right on the path and it starts a descent. https://youtu.be/Epy0DoWxgsI

Jon

[Hardy Heinlin]

Thanks, Jon!

Nice to see that it agrees with the BA sim. Also nice to see that the DES mode here activates ca. 3 to 4 nm before the T/D. Was there any command speed change?


|-|ardy

[emerydc8]

I'll have to look at that again on the command speed change prior to descent. We just left it in ECON and, to be honest, we weren't really paying much attention to what speed it would descend at. Time to power sleep.

[Markus Vitzethum]

Not sure whether this video helps as it is not at T/D but rather at FL160 (level segment during descent).

But here the VDI appears again, too, but not centered.
https://www.youtube.com/watch?v=lm_ca9KvOCU&index=42&t=333s

Markus

p.s.
By the way, this is also one of the few videos I know of which clearly shows the curved segment in the LNAV route when executing a Direct-To at 7:16, a (closed) topic discussed some time ago.

[emerydc8]

I think we all agree that when an early descent has just been started, e.g. 30 nm before T/D, and we're only, say, 100 ft below CRZ ALT, the needle will not indicate -100 (with reference to the CRZ ALT) but the needle will immediately go up way beyond -400 ft (with reference to the predicted descent path, extended from the T/D backwards up to the current aircraft position).

Hi Hardy,

A few days ago we were given an early descent where ATC cleared us to a lower altitude (FL220) about 50 miles prior to the T/D. I used the DES NOW feature and it immediately put VNAV into the DES mode. The VDI counter apparently only goes up to 9999, even though we were 16,000 feet below the original path (1st video).

As we got within 9999', the counter next to the VDI started counting down until it captured the path (2nd video). By that time we were already cleared to descend further, so it just resumed the descent. Just thought you might like to see this.
Cheers. Jon

https://youtu.be/m__H7TiwkBo

https://youtu.be/N12cChWdmmU

[Hardy Heinlin]

Thanks, Jon.

I don't really know if the 747 ND can display 5 digits. Maybe the 767 ND shows 4 digits ("9999") because its ND screen is narrower and it needs more space?

What I do know is that on the 747 ND the indication is rounded to 50 ft when outside the 400 ft window. This seems different to the 767 ND here. The rounding may be lower in newer software versions.

Your second video nicely shows the disagreement between the FMC indication and the ND indication: The FMC sends the data to the EIU, and the EIU seems to dampen any fluctuations before it sends the data to the ND's symbol generator; i.e. the ND indication smoothly scrolls down to the FMC's target value. Or maybe they disagree because the indication on the CDU screen is updated at a lower rate.


Cheers,

|-|ardy

[emerydc8]

I'm thinking that there must be variations of the PIP, Legacy and NG software for each airplane. While you would think our 767-300ERW would be less sophisticated than the 744, it has the ability to set a speed segment on the LEGS page, a characteristic of the NG. Sometimes I am surprised at what the FMC is capable of.

On the other hand, it completely ignored all the speeds on the STAR going into DFW yesterday. We can't figure out why the target speed never ratcheted down to the crossing speed on the LEGS page. Then, going into ELP we had to shoot an RNAV X RW04 approach in actual weather. Weren't we surprised when I pressed VNAV, speed intervened and watched the A/T surge forward, completely ignoring the speed I set in the window (as if I had never intervened at all).

Those two A/T malfunctions might be good to program into PSX just to teach pilots not to trust the automation. Both times I had to disconnect the A/T to regain control, and there wasn't much time to think about it either.

[emerydc8]

Or maybe they disagree because the indication on the CDU screen is updated at a lower rate.

From the second video, it looks like this is the case. The CDU lags behind the VDI. Perhaps it just updates the display at certain intervals so you can read the digits.

As a follow-up to my post, coming into ORD this morning, they started us down early, so we used the DES NOW feature. It went into DES mode within a few seconds, but I noticed that it took about 30 seconds before it could come up with a VDI display and by that time we were already about 1000' below the path. Maybe they're using a Commodore 64 microprocessor in this one.