747 Honeywell FMC Problem

[atpcliff]

I talked to another pilot and an engineer about this problem: None of the 3 of us knew how to fix it...

This is on a Legacy Honeywell box (Not NG FMC) 747 pax converted freighter:

A pilot sent for winds, and loaded them. They were waiting for the Destination Time/Fuel predictions to be show, before executing the wind load.
The Time/Fuel predictions never appeared. It turns out that, past a point on the route of flight, about 3/4 of the way to destination, the Time/Fuel estimates were all blank for all subsequent waypoints.

There were no EICAS messages, or CMC Present Leg fault messages that applied to the FMC.

The pilots tried both FMCs in Left, and Right, and sending for and loading new winds in all configurations...it didn't fix the problem.

Very close to the end of the flight, the FMCs suddenly started working correctly, displaying both the time/fuel estimates for the Destinations and all waypoints.

The initial loaded Flight Plan was about 12 FMC pages in length.

[Intruder]

Did you try purging the wind load? Did you check how far downline the fuel calculations were completed on the LEGS and ROUTE DATA pages?

I would guess there was a dropped piece of the transmission that affected one waypoint downline. Once you passed that waypoint and it was cleared from the system, the calculations were completed.

If it happened on the ground before takeoff, I would purge the FMC cache by cycling the Active Nav Database on the IDENT page, then reload the FMS. Do NOT try this airborne, because it will dump ALL loaded data from the FMS.

[atpcliff]

We thought about trying to Purge the winds, but we didn't know how. I had seen a Purge prompt earlier, when I was in the middle of loading the Descent Winds, but didn't see it later when we were looking for it.

We were in cruise when it happened.

[MATMAX]

cycling the databases will "erase" everything you entered ...
did this happen after a new DTBS was loaded ?
do both FMCs are having the same DTBS ?
there are times where new DTBS have "bugs" ...
did you try to reset the FMCs CBs ?
when there are problems with fuel calculations , the fuel qty system might be creating the issue , therefore , i am also suggesting to reset the FQPU also ...
ask Maintenance/Engineering to re-load it again and then doublecheck ...

[MATMAX]

if still no fix after DTBS reloaded , replace both FMCs ...

[atpcliff]

MATMAX:
Thanx for the info. We don't check both FMCs for the database...only the one the pilot setting up the flight is using (whoever is flying sets up the boxes)...good thing to check in the future.
Thanx for all the suggestions.

[NSEU]

cycling the databases will "erase" everything you entered ...

Well, not everything. Only a long term power interruption will do that. As an engineer, I've played with FMC on the ground and have had to clear all my entries before the next flight. Cycling the databases only seems to affect the route entries, not performance entries. Also, you shouldn't be able to swap databases in flight anyway. My training notes say that, because of inbuilt safeguards, it's only possible on the ground.

do both FMCs are having the same DTBS ?

If the nav databases disagree, a Nav Data Crossload page should appear automatically.

[flite idol]

I'll take a WAG. Manually entered winds will propagate forwards for the purposes of time and fuel calculations. it sounds like a partial or corrupted uplink occurred and the computations ceased due to incomplete data. Maybe if it happens again you could check to see if there is missing wind data then go to the first waypoint with no wind data, manually enter it and see if it propagates forward and allows time and fuel calculations. Once the descent was commenced fuel and time probably reappeared because the FMC transitioned into DES phase.
The only time I recall fuel/time calculation going missing is when data was missing or a change was made and not executed.

[JammedStab]

I had a similar problem twice for the exact same STAR and runway and heard about others that had the same problem with this and other STAR's. A bunch of the last waypoints were missing along with VNAV predictions as well as the the arrival fuel prediction(either we did not check the arrival fuel prediction/time prior to departure or perhaps we did but changed the STAR enroute, I can't remember).

The first time, we just flew the arrival without VNAV. The second time, each of the waypoints without the proper predictions was individually deleted to see what the effect would be and then restored once there was seen to be no improvement. If I remember correctly, it was one of the last waypoints in the STAR that was causing the problem and deleting it resolved the problem.

Info on possible remedies that came from these experiences was to delete the problem waypoint, delete its restrictions, replace the restriction with a different restriction such as 100 feet higher, select another runway then re-select the original runway. The one that was personal experience was deleting the problem waypoint.

I think in the end, it is a database problem for the individual waypoint that perhaps needs to be mentioned to the manufacturer to be resolved.

Let us know if any of this helps.

[MATMAX]

"I think in the end, it is a database problem"
thank you for reading what i wrote before , an MOCC guy ... ;)

[atpcliff]

flite idol

I am sure this would have fixed it. Didn't think of this, as I have only loaded the winds manually when we could not download them for the whole route of flight.

[JammedStab]

MATMAX

I certainly would not go down the replace both FMC route that you wrote before.

[flite idol]

It's just a guess Cliff. I believe there is a scratch pad message for incomplete wind data similar to the "partial route uplink" message that is seen from time to time but it may be easy to miss and it may not be the cause in this instance. I do remember losing fuel and eta data on a flight and staring at the CDU like Pavlovs dog for quite some time. What caused it and how I remedied it eludes my aging brain at the moment. If I do happen to dig that data from the depths in a lucid moment I will update. Cheers.

[flite idol]

It's been a while but I did remember a scenario that caused loss of eta and fuel at destination on the Honeywell legacy FMC but could not remember the thread....hehe. Anyhoo if an arrival is executed and a constraint speed is higher than the ECON descent speed due to a low cost index then it appears that the propagated speeds (small font) prior to the constraint are replaced by dashes and ETA/fuel disappear. Manually entering a descent speed the same or higher than the constraint speed restores the ETA arrival fuel. Sorry to dig this corpse up.