Pushing MCP ALT knob in VNAV ALT deletes constraint?

[emerydc8]

Thanks, Peter. That was my answer too. I think Hardy is going to change it so that it won't be operative in level flight.

Jon

[emerydc8]

Hi Peter,

Just to verify, in this example you did https://youtu.be/nUf8bYL7V1U?t=8m you are level at 4000 in VNAV PTH and you have a constraint at 4000 in front of you. You run the MCP up to 5000 and press it once. Will it take a second push to get rid of the 4000 on the LEGS page and start your climb to 5000, or does it delete the 4000 because you are already at the same altitude. I would think in this case the 4000 would be deleted and you will climb to 5000 with only one push.

Jon D.

[United744]

I figured you did and I'm glad you raised this. I'd still like to get Peter's thoughts on it. I found this in the Bulfer book:

Altitude Intervention does 5 things:
In climb, if LEGS page wpt altitude constraints exist between the airplane and the MCP altitude, each push deletes the next constraint, one at a time. A constraint equal to the MCP is not deleted. The same operation occurs in descent - one constraint deleted for each push.

When they say "climb," I don't think they are referring to climb mode but, again, input from Peter would be welcome.

I wouldn't call VNAV ALT "climb or descent", so I would expect the FIRST push to ONLY initiate a climb or descent and nothing else. Once the aircraft has entered VNAV ... something, only then would I expect subsequent pushes to delete altitude constraints, otherwise for a SINGLE push to initiate climb/descent AND delete contraints (how many? All between current and MCP ALT?) it would act like CLB/DES DIR instead.

This also fits with the description posted by emerydc8.

[United744]

Hi Peter,

Just to verify, in this example you did https://youtu.be/nUf8bYL7V1U?t=8m you are level at 4000 in VNAV PTH and you have a constraint at 4000 in front of you. You run the MCP up to 5000 and press it once. Will it take a second push to get rid of the 4000 on the LEGS page and start your climb to 5000, or does it delete the 4000 because you are already at the same altitude. I would think in this case the 4000 would be deleted and you will climb to 5000 with only one push.

Jon D.

I would expect it to delete the 4000 constraint and start climbing now. If it didn't delete the constraint and start climbing, or didn't do anything and waited to pass the waypoint before climbing, then pushing ALT would have no effect.

Imagine this situation:

Flying towards CLI:

CLI ---/4000 (hard)
DOVER ---/4000 (hard)
MAXIM ---/4000 (hard)
BAKER ---/12000 (hard)

Before CLI, I dial 6000 on the MCP and push once. I'd expect to see:

CLI ---/6000 (soft, plenty of performance to make MCP ALT before CLI)
DOVER ---/4000 (hard)
MAXIM ---/4000 (hard)
BAKER ---/12000 (hard)

The way I has this behavior explained to me was this: when you dial a new altitude and push the ALT knob, you are clearing the aircraft to climb/descend to that altitude NOW. For the FMC and MCP to be consistent with this, the current waypoint must also be updated when pushing the ALT knob (in this case, the constraint is deleted because it is updated via the MCP). The FMC reverts to a soft altitude and computes the crossing altitude.

[Britjet]

United744 - No.
You can put what you like in the ALTSEL. Pushing the intervene doesn't populate the legs line in this case.
Pushing the ALTSEL will delete one constraint at a time, so the aircraft will only go to 4000 (DOVER) in the example you describe. The legs line won't "populate" with 6000. It will merely delete the previous 4000 hard entry.
Peter.

[Kieber]

Peter I fully agree with you.

According to my notations, selecting an altitude above constraints (MCP) and pushing once, only the next constraint is deleted. Then every following push delete only the next constraint.

Walter

[emerydc8]

What I was getting at is in Peter's example, where he is at 4000, and he has a constraint of 4000 in front of him, and he selects 5000 in the ALT SEL. Because he is already at 4000, when he pushes ALT SEL, the constraint will be deleted and he will climb to 5000 with only one push. So this would be an exception to the constraint deletion feature only working when actively climbing or descending.

Compare this with where you are in the same position but you are level at 3000. If you put 5000 in the ALT SEL and pushed it, it would take two pushes to get to 5000 -- one to get it climbing and one to delete the 4000 constraint.

My point was that if you are level at the same altitude as a constraint in front of you and you select a higher altitude in the ALT SEL, constraint deletion will still occur (to delete any constraint in front of you at your present altitude) even though you are not presently climbing when you pushed it.

[Hardy Heinlin]

The condition "actively climbing/descending" is the condition I mean when I write "VNAV ALT not engaged".

So we are all on the same page :-)

The only difference being me talking in programming language. The words "actively climbing/descending" are undefined. So I need to program a definition. I define it like this:


Old version:

If MCP knob pushed and ... then delete constraint.


New version:

If MCP knob pushed and VNAV ALT disengaged and ... then delete constraint.


Quite trivial modification. Inserted in two seconds.


Cheers,

|-|ardy

[emerydc8]

Hardy,

What if you are in VNAV PTH and level? This would mean that you are at the same altitude as what's on the LEGS page. If you select a higher altitude and push, the existing constraint for your present altitude should be deleted (as in my example above).

[Hardy Heinlin]

OK, then I'll do this:


New version:

If MCP knob pushed
and ( VNAV ALT disengaged XOR MCP=constraint )
and ( VNAV ALT disengaged OR abs( aircraftAltitude - activeConstraint )<50 ft )
and ...
then delete activeConstraint.

[emerydc8]

That sounds good. I'm trying to imagine a situation, other than CRZ where constraint deletion is disabled, where you would be in VNAV ALT and at the same altitude as a constraint in front of you on the LEGS page. I guess it doesn't matter as long as MCP=constraint. In Peter's example, he was in VNAV PTH at 4000 -- not VNAV ALT.

[Hardy Heinlin]

I edited my post above.


You edited your post as well :-)

In Peter's example, he was in VNAV PTH at 4000 -- not VNAV ALT.

That's right. So VNAV ALT will not engage anyway. If I recall correctly, in some engineering manuals (and in PSX) the logic is so that VNAV ALT will only engage at MCP capture if the MCP differs from the FMC constraint by more than 150 ft. Maybe it was 100 ft. But certainly not less than 50 ft. Anyway, if it's less, the winner will always be VNAV PTH.

[emerydc8]

Sounds good, Hardy.

[Hardy Heinlin]

Now implemented in PSX 10.0.9-beta11:

http://aerowinx.com/board/index.php?topic=3685.0

[emerydc8]

Thanks!

[emerydc8]

Hi Hardy,

When you get a chance, can you take a look at this situ and see if you are able to delete these at-or-above altitude constraints on the PORTE 5 departure from SFO? I can't seem to get them to delete using the altitude intervention, but if I change them to hard altitudes, they will delete.

https://drive.google.com/file/d/0B-WRh0Hf7VdZb1JXNjB1R2w1a1k/view?usp=sharing

Thanks,
Jon D.

[Hardy Heinlin]

That's not a bug, and not new, but it possibly disagrees with the real FMC. -- Here's the whole story:

PSX is intentionally programmed so that it can distinguish between level-off constraints and non-level-off constraints.

An "A"-constraint will never cause a level-off on the climb path, and a "B"-constraint will never cause a level-off on the descent path. Such constraints do not prevent a direct climb, respectively a direct descent. Therefore they've been excluded from the MCP-one-push-one-deletion process since PS version 1.0.

In one of the PSX updates I saw it was necessary to allow the deletion even if the constraint doesn't prevent a direct path. Why? Because the altitude constraint may also include a speed constraint, and that is something that must be erasable.

However, in the current PSX version, the deletion of such constraints is possible only if they are on the active waypoint. I need to modify this.

In your situ, in the current PSX version, you can delete that "A"-constraint at PORTE only when PORTE is the active waypoint.


|-|ardy

[emerydc8]

Okay. Thanks, Hardy. I don't think it's as important on the list as the other items that are slated to be modified, so I can live with it. Thanks again.

Jon

[United744]

United744 - No.
You can put what you like in the ALTSEL. Pushing the intervene doesn't populate the legs line in this case.
Pushing the ALTSEL will delete one constraint at a time, so the aircraft will only go to 4000 (DOVER) in the example you describe. The legs line won't "populate" with 6000. It will merely delete the previous 4000 hard entry.
Peter.

That's what I wrote? CLI is at L1, and upon pressing ALT SEL, it deletes the constraint at CLI only? I'm confused now.

In my example, the aircraft is flying TO CLI, initially at 4000 ft.

[Hardy Heinlin]

United744, you wrote:

"CLI ---/6000 (soft, plenty of performance to make MCP ALT before CLI)"

Maybe Peter thought that you thought that this "6000" in big font is set big ("populated"?) by the MCP push.

I guess by "soft" you mean small font. However, this value in small font can be 4650 or 2160 or whatever, depending on the FMC's dynamic path prediction.

The word "hard" is typically used in context of "A", "B", "AB", and exact (hard) constraints. E.g. 9000A is in big font, but it's not a hard constraint; it's an "A" constraint ("at or above").