Descent FMA on Beta 13

[emerydc8]

Sounds good to me. We will probably never see speed reversion in the sim anyway.

[Hardy Heinlin]

I just tested the MCP window opener.

It looks strange. Non-intuitive. I don't like it. I'll remove it.

Instead, I'll implement a reminder that is set when the path is captured. When set, it allows VNAV SPD to go to VMO-16. The reminder is cleared at CRZ or when pushing EXEC.

[emerydc8]

I haven't been able to get PSX to get anywhere close to VMO-11 while descending in VNAV PTH. Are you inserting an artifical wind to push it high on the path when you test it? I think you mentioned last week that when you removed the FMC+15 restriction that it was able to maintain path within ECON + 20ish.

[Hardy Heinlin]

I set max thrust. Then shortly before VMO-16 I set idle.

[emerydc8]

I guess that would do it. Thanks for working on this Hardy. I don't know when you sleep.

[Hardy Heinlin]

Beta 17 is uploaded.


And I download myself now into the pillow. Good night ...

[emerydc8]

Hi Hardy,

I had a chance to test-drive Beta 17 into SFO on the Big Sir Arrival. I was impressed. First, I flew it without entering the 12,000' at SKUNK to see how it would do without throwing it any curves. It was very nice the whole way down. The Econ command descent speed was 284 knots. It seemed to get pushed a bit low and it slowed at one point to just under 269 knots and, predictably, the power came in and the FMA correctly indicated SPD||VNAV PTH. Back to the path and then the power goes back to idle and back to HOLD||VNAV PTH. Right on.

The second time, I threw a curve at it right after the top of descent. I entered /12000 at SKUNK. The FMA correctly went to VNAV SPD and the new path was over 4000' below the aircraft. It descended right at 284 knots, just as I believe it should.

This is a good scenario if anyone wants to see how to get down without using the speed brakes and just using speed intervention. Load the scenario and as you are descending through FL345, insert the /12000 at SKUNK.

https://drive.google.com/file/d/0B-WRh0Hf7VdZMUk4OTAwTHR3VmM/view?usp=sharing

Note that it correctly drops out of VNAV PTH and into VNAV SPD since you are no longer within 150' of path. The speed stabilizes at the command speed of 284 knots. Now, speed intervene (open the speed window) and select 330 knots. You can watch it converge on the path from above as you descend. You will get the DRAG REQUIRED message but just ignore it and watch. When you intercept the path from above, close the speed window and observe it will go back to VNAV PTH. It will cross SKUNK at exactly 12,000 and 284 knots! I took it all the way down to OAK and selected flaps 1 as it slowed towards 210 knots -- Very nice! Once past the first speed constraint the VTK ERROR is virtually zero the rest of the way.

As for the speed reversion, I tried to add power as I descended to simulate a severe tailwind, but had some trouble keeping it in the descent while accelerating. The nose kept wanting to pitch up as I fed in the power and it decreased the descent rate to about 1500 FPM. I decided to stop it at 70% N1, which was enough to eventually push it up to 349 knots. At that point I was 3,000' high and still in VNAV PTH (maybe this is due to trying to create the unforecast tailwind with thrust which is not something you would ever do in the plane). At 349 knots it reverted to VNAV SPD just as you said it would.

I'll have some fun playing with this more over the weekend. Thanks again for making all the revisions. I think it's coming together very well. Maybe it can be tested over the weekend during the flights.

[Hardy Heinlin]

Hi Jon,

thanks. I just tried your situ.

This is a good scenario if anyone wants to see how to get down without using the speed brakes and just using speed intervention. Load the scenario and as you are descending through FL345, insert the /12000 at SKUNK.

https://drive.google.com/file/d/0B-WRh0Hf7VdZMUk4OTAwTHR3VmM/view?usp=sharing

Note that it correctly drops out of VNAV PTH and into VNAV SPD since you are no longer within 150' of path. The speed stabilizes at the command speed of 284 knots. Now, speed intervene (open the speed window) and select 330 knots. You can watch it converge on the path from above as you descend. You will get the DRAG REQUIRED message but just ignore it and watch.

To be precise, on my system, -- and that's intentional --, the DRAG REQUIRED message occured only after VNAV PTH was re-engaged, not when VNAV SPD was already engaged. I hope you were seeing the same on your system.

When you intercept the path from above, close the speed window and observe it will go back to VNAV PTH.

That's the point where I got DRAG REQUIRED (as expected). I would close the window a bit later to prevent a premature deceleration. It can't follow the path during deceleration because the path is designed for command speed related descent rates. During deceleration the aircraft is not at the command speed yet and thusly will fly a shallower path, causing the deviation to increase again for some seconds.

How do you control this mode transit on the real aircraft? I would slowly decrease the MCP SPD. If not that, I would open the MCP window when the aircraft is ca. 400 ft below the path.

As for the speed reversion, I tried to add power as I descended to simulate a severe tailwind, but had some trouble keeping it in the descent while accelerating.

I intentionally want it to climb, and expect a climb. It's just for creating a path deviation :-)

You can also create a tailwind using the tropopause wind on the Planet weather page, but the path error will not be that big because the FMC uses the currently sensed IRS wind data for the next 5000 vertical feet down the path and mixes this with forecast wind.


Cheers,

|-|ardy

[emerydc8]

I think that's where I got the DRAG REQUIRED message too -- right when closing the speed window.

I experimented and tried letting it get 100' below the path before closing the speed window, so it had more room to decelerate back to 284 knots. I think it resulted in the actual airspeed settling in five or ten knots lower than when I closed the window right on the path.

In the real plane, I do just as you suggested and just slowly reduce the speed with the MCP knob to make the transition smoother. Some guys will pull the speed brakes out in addition to speed intervention, then close the window when right on the path and slowly take the speedbrakes out as it slows back to the command speed.

Actually, I'm not a fan of using speed brakes in anything other than V/S. You have to be careful in VNAV SPD or FLCH mode because the nose will pitch over to attempt to maintain the MCP speed when you pull the speed brakes out. If you pulled them fast enough, everything not strapped down would probably be in the air, so it's kind of a balancing act. In V/S, you don't have to worry about that.

[Hardy Heinlin]

On the real aircraft, with HOLD | | VNAV PTH, and 10 knots or so below command speed due to being below the path, and not much acceleration in sight, do you add some thrust manually or do you wait until SPD engages?

[Hardy Heinlin]

I just rediscovered a public Boeing document that confirms the Lufthansa quotes re maximum target speeds:

Vertical Navigation (VNAV) limits the
maximum target speed as follows:

747-400: 349 knots (VMO/MMO minus
16 knots) or a pilot-entered speed greater
than 354 (VMO/MMO minus 11 knots).

Source:

http://www.boeing.com/commercial/aeromagazine/articles/qtr_2_07/AERO_Q207_article5.pdf


Cheers,

|-|ardy

[Kieber]

Hi,

Sorry answering late....but the family.....

Although you are now headed in the right direction, I try to put it straight and send my message. May be it's to late now and just an overseeable roundup for our hobby-pilots.

VNAV PATH descent in connection with speed reversion or VNAV SPD descent are two very different animals.

1.   When descending in VNAV PATH the system behavior is exactly as discussed above and represented in the OM, involving above or below the first speed constraint. After speed reversion VNAV PATH changes to VNAV SPD, VNAV resets the target speed to 349 knots (VMO/MMO minus 16 knots). This high speed remains constant unless you like to change the speed with speed intervention by MCP or CDU-VNAV-page and the first speed constraint is not anymore relevant.  (I think when back on the path and closing the MCP speed window VNAV SPD changes again to VNAV PATH).

2.   When descending in VNAV SPD caused by different modifications like starting descent after cruising beyond the ToD, changing/entering altitude/speed constraints and much else, the system behavior is not like described in the OM and it doesn't matter whether the airplane is above or below the first speed constraint. The speed is always the speed from the FMC calculated or pilot entered. Speed changes are now possible with speed intervention by MCP or CDU-VNAV-page.

HTH

Cheers,

Walter

[Britjet]

Walters point number two is what I observed in the big sim last night.
FMC Descent speed was 290 kts if I remember, and I deliberately didn't reset the ALTSEL at TOD until the aircraft was 7000ft above profile, then I selected the first constraint altitude and pressed the ALTSEL.
The aircraft went into a descent but the speed stayed at 290 kts and I didn't get a DRAG REQUIRED message.
It never tried to regain the path.

Peter

[Hardy Heinlin]

Beta 17 behaves like that.


|-|ardy



P.S.: Just an operational question:

On the real aircraft, with HOLD | | VNAV PTH, and 10 knots or so below command speed due to being below the path, and not much acceleration in sight, do you add some thrust manually or do you wait until SPD engages?

[Kieber]

Thank's Peter...... confirming my very simple explanation.

Walter

[Kieber]

Super Hardy.........great job!

Walter

Reg. your speed question. In the early days of the 747-400 (1990 - 1997) we mostly used FLCH or VNAV with speed intervention.

[jtsjc1]

Hardy as in Peter's example would you use the Offpath Des function? Is this function used by pilots?

[emerydc8]

2.   When descending in VNAV SPD caused by different modifications like starting descent after cruising beyond the ToD, changing/entering altitude/speed constraints and much else, the system behavior is not like described in the OM and it doesn't matter whether the airplane is above or below the first speed constraint. The speed is always the speed from the FMC calculated or pilot entered. Speed changes are now possible with speed intervention by MCP or CDU-VNAV-page.

Hi Walter,

Thanks for confirming my understanding. When you say FMC calculated, are you referring to what is actually on VNAV 3/3 at 2L (the command speed that appears in magenta on the speed tape)? I think Hardy has it right in Beta 17. Thanks.

Jon D.

[Hardy Heinlin]

Hardy as in Peter's example would you use the Offpath Des function? Is this function used by pilots?

The word "offpath" in that context refers to lateral offpath, not vertical offpath. The energy management circles are not needed when you are on the route. For more details I suggest to start a separate thread :-)


Cheers,

|-|ardy

[jtsjc1]

Thanks Hardy I still got ALOT to learn about this airplane.