PPRuNe Forums

Go Back   PPRuNe Forums > Flight Deck Forums > Tech Log
Forgotten your Username/Password?

Tech Log The very best in practical technical discussion on the web


Reply
 
Thread Tools
Old 26th Dec 2012, 04:56   #1 (permalink)
 
Join Date: Jan 2006
Location: australasia
Posts: 388
Descent away from the magenta line (Boeing)

Can any of the more technically astute, give a figure as to the point at which a VNAV PATH becomes invalid, when descending NOT on the magenta line.

More specifically, at what displacement is the FMC unable to calculate an accurate profile reference

Personally, I get out of VNAV as soon as I leave the magenta, but observe quite often, very experienced pilots hanging on to VNAV at considerable displacements. The manuals I have read are mute on this issue.

Only interested in the Boeing case BTW.

Thanks

Maui

Last edited by maui; 26th Dec 2012 at 05:02.
maui is offline   Reply
Old 26th Dec 2012, 08:29   #2 (permalink)
 
Join Date: Aug 2000
Location: UK
Posts: 1,916
I used to wonder this in 747, 757 and 737. When you deviate, it doesn't seem to start going off very rapidly and seems to maintain a steady profile. I suspected it continued using a 'direct to' dtg rather than the equivalent magenta dtg, though nobody seemed to have any idea. I did find the 737NG VNAV descent used to make some sudden peculiar and unpredictable speed reductions for no reason that I could fathom, even when high on profile.
Notso Fantastic is offline   Reply
Old 26th Dec 2012, 08:50   #3 (permalink)
 
Join Date: Mar 2009
Location: R220/A590
Posts: 107
Quote:
Originally Posted by Notso Fantastic View Post
I did find the 737NG VNAV descent used to make some sudden peculiar and unpredictable speed reductions for no reason that I could fathom, even when high on profile.
Could it be because you entered a non waypoint speed / altitude constraint at Speed Restriction on VNAV DES? In order to be on desired speed at altitude indicated, the VNAC performs a deceleration.

Quote:
Originally Posted by maui View Post
Personally, I get out of VNAV as soon as I leave the magenta, but observe quite often, very experienced pilots hanging on to VNAV at considerable displacements. The manuals I have read are mute on this issue.
There's some non-precision approaches with guidance for a continuing ROD until the MDA. One must apply a controlled flight path to stay above or at the profile altitudes indicated and that requires the use of VNAV Path.
Claybird is offline   Reply
Old 26th Dec 2012, 09:04   #4 (permalink)
 
Join Date: Aug 2000
Location: UK
Posts: 1,916
I didn't often enter those unless absolutely needed. It just used to happen for no apparent reason that I could fathom. The NG VNAV descent profiles were very unreliable and frequently had to be ignored. The 747-400 VNAV used to be almost perfect, 15 years later the 737NG VNAV profile could be anything up to 8000' out and always needed to be checked by mental rule of thumb all the way down. Never to be trusted without checking.
Notso Fantastic is offline   Reply
Old 26th Dec 2012, 09:12   #5 (permalink)
 
Join Date: Jan 2006
Location: australasia
Posts: 388
Clay bird.

Thanks for that, but it's not what I am after. This is about "enroute" descents, not "approach" descents.

Maui
maui is offline   Reply
Old 26th Dec 2012, 09:13   #6 (permalink)
 
Join Date: Mar 2009
Location: R220/A590
Posts: 107
To NF: Another reason could be due to changes in the LEGS page. Because the FMC recalculates a new VNAV path for the constraints. When that happens, you lose the VNAV Path Deviation Scale until recalculation and hence have to then re-acquire the VNAV path.

Very interesting, though. Any additional insight or explanation by other members would be useful.

--

Edit: Maui, apologies, Jumped the gun and didn't read your post carefully.

Last edited by Claybird; 26th Dec 2012 at 09:28.
Claybird is offline   Reply
Old 26th Dec 2012, 09:19   #7 (permalink)
 
Join Date: Aug 2000
Location: UK
Posts: 1,916
<<Thanks for that, but it's not what I am after. This is about "enroute" descents, not "approach" descents.>>

What are we talking about here? Your question didn't specify what sort of descent. Is it any wonder I assumed you meant 'descent to approach'? Why are enroute descents any different? You specify where you want to be at what speed/altitude and let the computer chomp it out.
Notso Fantastic is offline   Reply
Old 26th Dec 2012, 09:26   #8 (permalink)
 
Join Date: Jan 2006
Location: australasia
Posts: 388
Claybird.

We are in different universes, perhaps I was not clear in my request.

We are on descent. LNAV VNAV path, on profile. ATC give us a vector and off we go away from the magenta line.
Leave the LEGS page until we have something to aim at.
In the meantime the auto pilot/FMC is following some data for descent. The derivation of that descent profile cannot be accurate as neither we, nor the FMC knows what is next to happen.
How far from the magenta line can we go before that data is invalid. At this stage we still have a vertical deviation indication and the autopilot has kept that zeroed.

It s my view that you might as well call the data invalid once you move off the magenta, because the unit has no way of predicting the next move. However, many pilots leave VNAV plugged in, using what would appear to be invalid data. This of course is verified when they update the LEGS page, and oops, all of a sudden we are 3000 feet low on profile.

The question is: How much lateral deviation will the FMC tolerate and still produce good data?
Does the FMC continue to calculate a profile based on a turn from PPS to the next waypoint, and if so out to what lateral displacement, before it throws up it's hands and says "can not compute"?

Thanks

Maui

Last edited by maui; 26th Dec 2012 at 10:00.
maui is offline   Reply
Old 26th Dec 2012, 09:34   #9 (permalink)
 
Join Date: Jan 2006
Location: australasia
Posts: 388
Not So Fantastic.

My appologies .
By approach descent, in that second post, I refer to descent inside the FAF, (something raised by Claybird not me). Not part of the original request.

The original question was about normal descents for arrival, (perhaps arrival would have been abetter description than enroute).

Hopefully the immediate past post by me, will clarify.

Maui.
maui is offline   Reply
Old 26th Dec 2012, 11:06   #10 (permalink)
 
Join Date: Mar 2009
Location: R220/A590
Posts: 107
Quote:
Originally Posted by maui View Post
We are on descent. LNAV VNAV path, on profile. ATC give us a vector and off we go away from the magenta line.
Now you're talking :- ) ATC radar vectors!

VNAV on arrival and radar vectors is not a great combination, when there's other auto flights modes more appropriate to the task.

Indeed, as you say, the FMC could mislead the pilot if he doesn't know where he is.

Quote:
Originally Posted by maui View Post
Does the FMC continue to calculate a profile based on a turn from PPS to the next waypoint?
Yes.

Quote:
Originally Posted by maui View Post
and if so out to what lateral displacement, before it throws up it's hands and says "can not compute"?
Good technical question indeed. So back to square one.
Claybird is offline   Reply
Old 26th Dec 2012, 12:11   #11 (permalink)
 
Join Date: Jan 2006
Location: australasia
Posts: 388
Thank goodness we now have the question sorted.

Now, does anyone have an answer?

Maui
maui is offline   Reply
Old 26th Dec 2012, 13:19   #12 (permalink)
 
Join Date: Mar 1999
Posts: 307
Maui I agree with your technique for this example. You are on a STAR passing FL160 with a constraint of FL100 at the next waypoint. ATC instructs "turn right heading 270 expedite descent to FL80 for traffic." Heading is selected and 8000 dialed into the MCP. If VNAV is used for the descent then the FMA will indicate VNAV SPD and then capture VNAV ALT at FL100 until passing abeam the constraint in the legs page and the resume the descent to FL8000. ATC will not be very happy with you and a TA/RA may be the result. I think VNAV looks at the distance along track from the abeam point to the waypoint/constraint to calculate the path but will honour constraints abeam to your postion thus the unintended level segment. I too see this lack of awareness from very experienced pilots and I always select FLCH or V/S when taken off the lateral track by ATC for descents.
flite idol is offline   Reply
Old 26th Dec 2012, 16:42   #13 (permalink)
 
Join Date: Mar 1999
Posts: 307
Should have clarified.....747-400
flite idol is offline   Reply
Old 26th Dec 2012, 22:08   #14 (permalink)
 
Join Date: Mar 2002
Location: Seat 0A
Posts: 5,073
Quote:
ATC instructs "turn right heading 270 expedite descent to FL80 for traffic." Heading is selected and 8000 dialed into the MCP.
Not being a 747 pilot, it sounds to me like you didn't expedite descent. Besides, you're now off the STAR so the constraints no longer apply.
Capn Bloggs is online now   Reply
Old 27th Dec 2012, 01:33   #15 (permalink)
 
Join Date: Jun 2001
Location: In da north country
Age: 53
Posts: 405
I"ve been on the -400 for over 4 years. I had the worst ground school in the history of aviation.

For me, I trust the FMS decent calculations when i am confident that we are flying whats programmed in the box. When I am off the programmed decent/arrival, I revert to FLCH and ALWAYS keep the 3:1 calculation going in my head. There are times i do not know what the VDI is looking at, so; back to the 3:1.

All that fancy stuff is a tool, not a scientific given. Use it as much as you can, but trust it as much as you would a cop. Blame no one or no thing except yourself, if you screw up the decent. The 3:1 always works.

Merry Christmas, Enjoy the Egg Nogg!
Willit Run is offline   Reply
Old 27th Dec 2012, 02:33   #16 (permalink)
 
Join Date: Jan 2005
Location: Can't remember
Posts: 41
VNAV should still provide a reasonably accurate descent profile up to 20 nms to the left or right of the magenta line. This is based on the fact that at distances greater than 20 nms, the next waypoint will not be recognised as one passes abeam the position. In this case, the data will most definitely be invalid, including fuel, dist and time remaining.

Last edited by 777boeings; 27th Dec 2012 at 06:13.
777boeings is offline   Reply
Old 27th Dec 2012, 03:20   #17 (permalink)
 
Join Date: Apr 2009
Location: Earth
Posts: 470
Quote:
you're now off the STAR so the constraints no longer apply.
Without wishing to open a new can, that may or may not be correct, depending apon where one finds oneself.
falconeasydriver is offline   Reply
Old 5th Jan 2013, 22:42   #18 (permalink)
 
Join Date: Feb 2008
Location: Earth until ..........
Posts: 22
Flying B738.
Volume 2, Pg. 4.10.9
VNAV mode is terminated by any one of the following :

Crosstrack deviation exceeds twice the RNP value during PTH descent for
an active leg with a database vertical angle and LNAV not engaged.
( There are other reasons for termination of VNAV mode but this was the one
relating to the question ).
Seen it happen this morning.
B888 is offline   Reply
Old 8th Jan 2013, 13:44   #19 (permalink)
 
Join Date: Apr 2009
Location: San Clemente, CA
Posts: 1,913
flite idol:

Quote:
Maui I agree with your technique for this example. You are on a STAR passing FL160 with a constraint of FL100 at the next waypoint. ATC instructs "turn right heading 270 expedite descent to FL80 for traffic." Heading is selected and 8000 dialed into the MCP. If VNAV is used for the descent then the FMA will indicate VNAV SPD and then capture VNAV ALT at FL100 until passing abeam the constraint in the legs page and the resume the descent to FL8000. ATC will not be very happy with you and a TA/RA may be the result. I think VNAV looks at the distance along track from the abeam point to the waypoint/constraint to calculate the path but will honour constraints abeam to your postion thus the unintended level segment. I too see this lack of awareness from very experienced pilots and I always select FLCH or V/S when taken off the lateral track by ATC for descents.
Indeed "expedite descent" combined with an off-track heading assignment invalidates VNAV. As you say FLCH or V/S is making the airplane comply with the new ATC clearance. But, can some of the modern pilots think this through?
aterpster is offline   Reply
Old 8th Jan 2013, 15:36   #20 (permalink)
 
Join Date: Apr 2008
Location: nowhere
Posts: 846
Quote:
Originally Posted by 777boeings View Post
VNAV should still provide a reasonably accurate descent profile up to 20 nms to the left or right of the magenta line. This is based on the fact that at distances greater than 20 nms, the next waypoint will not be recognised as one passes abeam the position. In this case, the data will most definitely be invalid, including fuel, dist and time remaining.
Sounds familiar.....isn't that the max distance for a proper offset. Farther will not have WP recognition?

Quote:
Originally Posted by flite idol View Post
Maui I agree with your technique for this example. You are on a STAR passing FL160 with a constraint of FL100 at the next waypoint. ATC instructs "turn right heading 270 expedite descent to FL80 for traffic." Heading is selected and 8000 dialed into the MCP. If VNAV is used for the descent then the FMA will indicate VNAV SPD and then capture VNAV ALT at FL100 until passing abeam the constraint in the legs page and the resume the descent to FL8000. ATC will not be very happy with you and a TA/RA may be the result. I think VNAV looks at the distance along track from the abeam point to the waypoint/constraint to calculate the path but will honour constraints abeam to your postion thus the unintended level segment. I too see this lack of awareness from very experienced pilots and I always select FLCH or V/S when taken off the lateral track by ATC for descents.
If this is the case, would an appropriate push of the altitude button help?

Last edited by JammedStab; 8th Jan 2013 at 15:38.
JammedStab is offline   Reply
Reply
 
 
 


Thread Tools


Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off



All times are GMT. The time now is 09:06.


vBulletin® v3.8.7, Copyright ©2000-2014, vBulletin Solutions, Inc.
SEO by vBSEO 3.6.1
1996-2012 The Professional Pilots Rumour Network