[1.0.4] LEGS page and VNAV stuck, LNAV continues

[daemotron]

Sorry for the title of this post, I didn't find anything more suitable to describe this experience in just a few words. So here's what happend on my last flight from LEMG to LEPA, using this route:

LEMG 13 ROLA1H ROLAS UN851 IZA IZA1P LEPA 24L (IAP: ILS Z 24L, MJV transition)

Navigation data used: NDP AIRAC 1605

I started descending some 25NM before the calculated T/D using V/S mode (VNAV mode was not used during whole flight, but the FMS was completely set up so it could have been used, even including ISA and wind). Lateral navigation was done using LNAV mode (engaged shortly after TO and maintained for the whole flight until final).

Passing waypoint IZANB (first fix of the STAR, comes after the T/D), the LEGS page got "stuck" on that waypoint - i. e. it continued displaying it as the active waypoint. Also the VNAV path indication showed its deviation based on that waypoint, as if the aircraft had never reached it. At the same time however (remeber I was flying with LNAV) the flight director continued tracking perfectly the magenta line to X232 and SABAS without any issue (despite two HDG changes on these segments). Usually when missing a waypoint, LNAV will make you go round in circles until you alter the routing or properly intercept the missed WP. So to me it seems the VNAV part of the FMS and the CDU "lost" the route progress, while the LNAV part of the FMS correctly followed the route - the HSI consequently however did not highlight the waypoint LNAV was flying me to.

Maybe this screenshot helps illustrating the situation:

 

Next thing I tried is recovering VNAV and the legs page. Directly after having passed SABAS, I activated MJV (next waypoint after SABAS) by coyping it to LSK1 and EXEC. The legs page was then updated accordingly, however the VNAV deviation was still stuck at IZANB, and the HSI was still not highlighting MJV as active waypoint (however LNAV was not much impressed by this, it tracked the WP anyway):

 

After the aircraft passed MJV and started tracking MJV11 (next waypoint after MJV), without further intervention from my side, the situation recovered:

• the LEGS page updated automatically to MJV11
• the VNAV path deviation shrunk to something plausible (MJV11 has an at or above restriction on it, so the FMS estimation was not visible at that moment - therefore just "plausible", but correct with high probability)
• HSI shows MJV11 as active waypoint (magenta highlighting)

 

Attached the log files (Gizmo will show you the mandatory soft crash for me having forced a debug output on this, but since it seems like a VNAV/LNAV story, I thought it could be useful...)

 

 

 

Edited April 9 by daemotron

[Litjan]

Thanks for the report! We have seen some cases where the waypoint won´t sequence, usually due to a wide turn and the plane not passing close enough - the LNAV is strictly following the magenta, but the wpt sequencing is dependent on crossing the wpt close enough...

THis sounds different, as a direct-to did not seem to help (it usually does)...

I will add this to the list and we will see if we can recreate and fix it!

Thanks, Jan

 

[daemotron]

On 5/28/2016 at 3:53 AM, Litjan said:

Thanks for the report! We have seen some cases where the waypoint won´t sequence, usually due to a wide turn and the plane not passing close enough - the LNAV is strictly following the magenta, but the wpt sequencing is dependent on crossing the wpt close enough...

That's interesting, never had seen that  before. When I had WP misses, usually I was also off the magenta line, and the FD started maneuvering wildly to recover (usually "goin' round in circles"). In that case however it would have been hard to miss the wp; the last turn before (at IZA) should be just a 20° RT, and I didn't have any wind-driven offset.

Quote

THis sounds different, as a direct-to did not seem to help (it usually does)...

Oh it did, but not before next WP was sequenced.

Maybe this also has to do with how this STAR is implemented in the nav data. The original STAR as published by ENAIRE (former AENA) looks like this (chart available here):

 

So this chart suggests to me the STAR is "fly HDG 88°, then intercept MJV R232". However in the nav data (at least NDP AIRAC 1506) this STAR is modelled differently; instead of intercepting MJV, it uses artificial fixes (IZANB and X232) which are both defined as "Normal" type WPs with long/lat. Digging into the nav data provided, I found that both, IZA and IZANB do have the same geographical coordinates:

IBIZA                   IZA  NDB  38.915472   1.470417394.00N

(extract from wpNavAID.txt)

<Star Name="IZA1P.24L" Runways="24L">
  <Star_Waypoint ID="0">
    <Name>IZANB</Name>
    <Type>Normal</Type>
    <Latitude>38.915472</Latitude>
    <Longitude>1.470417</Longitude>
    <Speed>0</Speed>
    <Altitude>0</Altitude>
    <AltitudeCons>0</AltitudeCons>
    <AltitudeRestriction>at</AltitudeRestriction>
    <Flytype>Fly-by</Flytype>
    <BankLimit>25</BankLimit>
    <Sp_Turn>Auto</Sp_Turn>
  </Star_Waypoint>
  [...]

(extract from LEPA.xml).

I think I didn't notice when checking the route and just had the RTE DISC fixed by moving IZANB directly after IZA, which is pretty much unflyable (unless the FMC is able to spot that two subsequent WPs are the same and just sequences them both).

Quote

I will add this to the list and we will see if we can recreate and fix it!

Thanks a lot Jan!

Edited April 9 by daemotron

[Sweet19blue]

Hello,

I did my first complete flight yesterday and add that same problem. It screwed up my whole flight as I failed to identify the problem early.

I departed runway 23 at LSGG on MEDAM4B departure. This departure starts with a left turn to intercept a radial almost 90° from runway heading but behind the Aircraft. I guess the wide turn required should have bypassed a waypoint which prevented the LNAV waypoint sequencing in FMC (even if the Aircraft itself still folowed the LNAV path). The stuck waypoint was the INTC waypoint.

I will try the same flight again today and see what happen.

Best regards,

Pierre

 

 

[Litjan]

1 hour ago, Sweet19blue said:

Hello,

I did my first complete flight yesterday and add that same problem. It screwed up my whole flight as I failed to identify the problem early.

I departed runway 23 at LSGG on MEDAM4B departure. This departure starts with a left turn to intercept a radial almost 90° from runway heading but behind the Aircraft. I guess the wide turn required should have bypassed a waypoint which prevented the LNAV waypoint sequencing in FMC (even if the Aircraft itself still folowed the LNAV path). The stuck waypoint was the INTC waypoint.

I will try the same flight again today and see what happen.

Best regards,

Pierre

 

 

Thanks, Pierre - let me know what happens. I think the lateral distance is set to 2.5NM - so whenever you pass by a waypoint further than that, it will not sequence. We will tweak that in the upcoming weeks, for now when it happens just fly direct to the next waypoint (insert into LSK 1L, execute)

Jan