744 Forum

Apron => Hangar 7 => Topic started by: Hardy Heinlin on Sun, 11 Feb 2018 14:23

Title: CDU standby navigation
Post by: Hardy Heinlin on Sun, 11 Feb 2018 14:23
Good evening,

as we all know, each CDU is hardwired with other specific devices; when the FMCs fail, the data from other devices is no longer shared within the network. So in a standby navigation scenario:

CDU L gets data from IRU L, ILS L, VOR L, ADF L.
CDU R gets data from IRU R, ILS R, VOR R, ADF R.
CDU C gets data from IRU C, ILS C.

Non-conditional route waypoints are copied from FMCs to CDUs whenever a modification is executed, and only if the respective FMC is operative, of course. A failed FMC cannot transfer any data.

As far as I recall, the links between FMCs and CDUs for standby memory actions are hardwired as well:

CDU L gets data from FMC L.
CDU R gets data from FMC R.
CDU C gets data from FMC L or R, whichever is the master FMC.

This means, when FMC R fails, CDU R gets no data from FMC R anymore. It gets no data from FMC L either, because CDU R is hardwired with FMC R only. Data synchronization across FMCs is only possible as long as both FMCs are operative. As there is no wire directly from FMC L to CDU R, CDU R cannot get data directly from FMC L. CDU R can only get FMC L data via FMC R. So if FMC R fails: CDU R gets no data.

Correct or wrong?


As we all know, cheers!

|-|ardy
Title: Re: CDU standby navigation
Post by: Hardy Heinlin on Sun, 11 Feb 2018 14:58
Test demo:

Load the situation file Basic 004 - Cleared for takeoff.situ

Adjust your layout so that you see CDU L and ND R (you can return to your old layout by pressing the "0" key on the numpad.)

Set the range of ND R to 80.

Set F/O's NAV SOURCE selector to CDU R (check that "CDU R" is displayed in green at the bottom of ND R).

On CDU L, put NOXAD over OLITO and execute.

Notice that the CDU-R-route on ND R draws a direct-to to NOXAD. (FMC L has sent this route modification to FMC R, and FMC R has sent this to CDU R and ND R.)

Nice.

Now go to Instructor > Situation > Malfunctions > Nav, and click "Activate" in the line "FMC failure - R".

Wait until the message RESYNCING OTHER FMC disappears (it tries to resync, but it will fail).

Clear all scratchpad messages.

On CDU L, put SINDA over NOXAD and execute.

Notice that the CDU-R-route on ND R remains unchanged. This is because CDU R gets no data from FMC R anymore. And it cannot get data directly from FMC L either as there is no direct wire from FMC L to CDU R.
Title: Re: CDU standby navigation
Post by: mark744 on Sun, 11 Feb 2018 16:23
Hardy

I think that if one FMC is operable, it would synchronise waypoints to CDU L R and C
so that they will be available for Standby Navigation if the other FMC fails too.

As we know, if both FMCs are failed, Standby Navigation is by using just CDUs &  requires lat/long waypoints to be entered into all 3 CDUs along with three ILS frequencies. but only if both fail

But I am only making assumptions about the synchronising.

BTW
The QRH for ONE failed FMC, makes no reference to route mods being different in each CDU, surely it would.

( the QRH for two failed FMCs goes into lots of detail  )

Title: Re: CDU standby navigation
Post by: Hardy Heinlin on Sun, 11 Feb 2018 22:21
I just checked an engineering diagram and it looks like either FMC is directly connected to all three CDUs, while on the same diagram each IRU is explicitly connected to just one CDU. So you are probably right, Mark.

If I only could remember where I got that other information from. I'm still not 100% sure as these diagrams are simplified. But I tend to believe now that all three CDUs, indeed, can get data from one FMC directly, probably from the master FMC.


Regards,

|-|ardy
Title: Re: CDU standby navigation
Post by: John H Watson on Sun, 11 Feb 2018 23:52
From my training notes:

"FMC Inputs - 1"

"The Left FMC receives data from:
Left and Right CDU." (Book 14 p194...) Centre CDU not mentioned (I assume no C CDU data is sent to the FMC)


"FMC Inputs -2"
"The Left FMC receives data from FMC Master Relay
Captain's and F/O's instrument source switches"
"FMC Master Relay" (A list of Master switched data is given on another page).

Regarding outputs from the FMC:

"In the normal configuration, the left and center CDU receive data from the L-FMC." The right CDU normally receives data from the Right FMC. Source selection is determined by the Captain's and F/O's ISSS position." (ref book 14 page 216).

I think the FMC Master Switch only affects those systems listed on page 208: 

FCC's, ILS receivers, ADF receivers, DME interrogators, VOR receivers, AFCS mode control panel and Engine EECs.


Title: Re: CDU standby navigation
Post by: Jeroen Hoppenbrouwers on Mon, 12 Feb 2018 00:35
I had the pleasure of sticking my finger in FMC data busses a while ago, and although block diagrams may suggest "connections", the actual wiring typically is split in different busses for different purposes.

I am pretty sure that the connection from, say, FMC-Left to the three MCDUs only carries DISPLAY DATA.
There most likely is a DIFFERENT bus from FMC-Left to MCDU-Left that carries the flight plan data.

We need wiring diagrams, not block diagrams, to be sure.


Hoppie
Title: Re: CDU standby navigation
Post by: John H Watson on Mon, 12 Feb 2018 01:23
Hoppie, Wiring Diagrams generally only show wiring... not the data going on them. Wiring Schematics on the other hand, do sometimes show data.
Title: Re: CDU standby navigation
Post by: Hardy Heinlin on Mon, 12 Feb 2018 01:27
The normal stuff is clear to me. I'm just wondering whether those typical descriptions also refer to the data transfer into the CDU route memory. Maybe they just refer to CDU screen data and keypress signals.

We have:

• CDU keypress signals going to the FMCs --- NAV source selection dependent
• FMC text going to the CDU screens --------- NAV source selection dependent
• FMC route updates going to CDU memory -- Hardwired or NAV source selection dependent?


In case of a single FMC failure, with FMC L operative, I'm seeing two possibilities:

1. All CDU route memories get data from the operative FMC, independent of the NAV source selections.

Or:

2. CDU C memory gets data from the operative FMC; CDU L memory gets it if captain's NAV source is set to FMC L or CDU L; CDU R memory gets it if F/O's NAV source is set to FMC L or CDU R. (If the source selection is relevant, at least the selection of CDU C or of the inoperative FMC should disconnect the link.)


In other words: The NAV source selections are irrelevant and the memories of all CDUs get data in any case, -- or the NAV source selections are relevant and CDU L and R memories only get data when the related sources are selected.


|-|ardy
Title: Re: CDU standby navigation
Post by: Jeroen Hoppenbrouwers on Mon, 12 Feb 2018 09:49
Quote from: John H Watson on Mon, 12 Feb 2018 01:23
Hoppie, Wiring Diagrams generally only show wiring... not the data going on them. Wiring Schematics on the other hand, do sometimes show data.
Correct -- I intended to say that if the wiring diagrams show multiple parallel 429 lines, i.e. more than one bus going from the same FMC to the same MCDU, it could reveal that some of these busses go to one, two, or three MCDUs.

What data goes over that then is phase two of the research project...


Hoppie
Title: Re: CDU standby navigation
Post by: Hardy Heinlin on Mon, 12 Feb 2018 14:13
Phase 2 has started:

From observations on the real deck (some years ago) we know this: When FMC R fails and F/O's NAV source is set to CDU R, CDU R is in standby navigation mode and its IRS LEGS page shows its CDU route memory (with all legs it previously got from the FMCs, except for conditional waypoints).

Now if these IRS LEGS are still being updated by FMC L, these updates should be visible on the screen of CDU R. I'm having difficulties to imagine that you can see any updates from the FMC on this screen when that CDU is in standby navigation mode.


Regards,

|-|ardy


P.S.: For a test I just changed my code so that the CDU R standby memory is still being updated by FMC L. Of course, standby route modifications in CDU R are not sent to FMC L. It only goes vice versa. So when I make a direct-to in this CDU route and execute, two seconds later this CDU route will be replaced by the FMC route. So the CDU route is shown, but I have no reasonable control over it. This makes no sense. I could allow this control only when the F/O's NAV source is set to CDU R. This way you could spend an hour making a fine new standby route in CDU R and be happy, and then lose all your work when you turn your NAV source selector away from CDU R for a second.
Title: Re: CDU standby navigation
Post by: mark744 on Mon, 12 Feb 2018 14:45
In answer to you P.S.

Having F/Os Nav Source to CDU R, with a working FMC makes no sense as the aircraft is not forced into standby navigation.

so loosing that route that is input while in CDU R isn't unreasonable when re  selecting FMC L, which  is a superior mode, with full Autopilot function and normal operation under the control of FMC L, with data being able to be input from either CDU,all as normal (except two FMCs cant confer calculations)

Title: Re: CDU standby navigation
Post by: Hardy Heinlin on Mon, 12 Feb 2018 14:56
Quote from: mark744 on Mon, 12 Feb 2018 14:45
there is no reason to set F/O's NAV source to CDU R if there is still an operating FMC

I agree. But if you do it anyway (accidentally or whatever), the system design should provide a clean solution for such a case, e.g. it could inhibit the access to the IRS LEGS page as long as the FMC is feeding the CDU's standby memory.

It does this for CDU C. You can only access the IRS LEGS page on CDU C in case of a dual FMC failure. For CDU L and R, however, the access is already enabled in case of a single FMC failure.

(Please note my "P.S." edit in my previous comment.)
Title: Re: CDU standby navigation
Post by: Hardy Heinlin on Mon, 12 Feb 2018 15:07
Quote from: mark744 on Mon, 12 Feb 2018 14:45
Come to think of it, I believe the Source Select Switch just directs the source of the date that goes to be displayed on the ND (and PFD if it's a FMC) and doesn't control where the CDU gets it's information.

The NAV source selection also determines whether this CDU is controlling FMC L or R, and whether this CDU screen is getting text from FMC L or R. It's not just for the ND and PFD.
Title: Re: CDU standby navigation
Post by: mark744 on Mon, 12 Feb 2018 15:38
I see.
However

QUOTE
It does this for CDU C. You can only access the IRS LEGS page on CDU C in case of a dual FMC failure. For CDU L and R, however, the access is already enabled in case of a single FMC failure.
QUOTE

For one FMC failure, the correct Source Select Switch selection is to the operable FMC not to CDU
Title: Re: CDU standby navigation
Post by: United744 on Mon, 12 Feb 2018 15:46
As I understand it, a *single* FMC failure does not cause reversion to standby navigation. The CDU on the failed side connects to the remaining FMC, and can update/receive data from it as normal.

Whether NAV source selection can result in CDU standby navigation, I don't know (I would guess if FMC R fails, and the ND R source is selected to FMC R, that it will cause the CDU to go to standby navigation mode because its on-side FMC has failed).

If you had a failure of FMC R, and selected the ND L to FMC R, I would expect CDU L to go to standby nav, again because the FMC R has failed.
Title: Re: CDU standby navigation
Post by: Hardy Heinlin on Mon, 12 Feb 2018 15:52
Quote from: mark744 on Mon, 12 Feb 2018 15:38
For one FMC failure, the correct Source Select Switch selection is to the operable FMC not to CDU

Yes. But why can the IRS LEGS page be accessed when the pilot incorrectly selects CDU R? Why does the system not protect itself against this incorrect selection? With the example of CDU C I just want to show that the system is able to inhibit the access if it just wants to.
Title: Re: CDU standby navigation
Post by: United744 on Mon, 12 Feb 2018 15:54
Quote from: Hardy Heinlin on Mon, 12 Feb 2018 15:52
Quote from: mark744 on Mon, 12 Feb 2018 15:38
For one FMC failure, the correct Source Select Switch selection is to the operable FMC not to CDU

Yes. But why can the IRS LEGS page be accessed when the pilot incorrectly selects CDU R? Why does the system not protect itself against this incorrect selection? With the example of CDU C I just want to show that the system is able to inhibit the access if it just wants to.

I would suggest because the selected FMC has failed, it reverts to standby nav (thus, selected side IRS nav data).

If it didn't revert to standby nav, that would be even stranger, as you would have selected the failed FMC, yet still be using the operable FMC.
Title: Re: CDU standby navigation
Post by: mark744 on Mon, 12 Feb 2018 16:15
also, there is little risk in being allowed to access the IRS Legs page it cant be followed by LNAVs or autopilot.

so its only a map for one pilot side, then if the NS Select is set to the operable FMC, the CDU gets updated with the FMC waypoints that were or can be used by the LNAV/ autopilot as normal
Title: Re: CDU standby navigation
Post by: Hardy Heinlin on Mon, 12 Feb 2018 16:34
What I want to avoid is this: Say, the pilot incorrectly selects CDU R on the NAV and modifies the CDU R route for 10 minutes, and then notices the incorrect NAV selection and selects FMC R. -- 10 minutes of work in vain. If the access to the IRS LEGS had been inhibited from the start, the pilot wouldn't have wasted 10 minutes in useless work. The deleted route mods are useless anway. But the wasted time could have been avoided by not inviting the pilot to work on the IRS LEGS page.

In short: If I let CDU R get data from FMC L, I'd like to let the data flow to CDU R directly regardless of the NAV source selection. Any mods on the IRS LEGS will then be overwritten by FMC L after a few seconds already; so the pilot can instantly see that the work is useless.
Title: Re: CDU standby navigation
Post by: mark744 on Mon, 12 Feb 2018 17:29
I completely see your logic, but I don't think the aircraft should override the selection the pilot has made.
This is a Boeing, after all, not an Airbus  ;)

I dont think Boeing philosophy would prevent the pilot from doing something incorrect with no danger to the aircraft, just wasted time. Crew should follow the QRH and Boeing would expect this to be done.
(not Boeing philosophy for this age of aircraft anyway)

They dont prevent Alternate flaps if there is no fault, or alternate gear without a fault etc etc
Title: Re: CDU standby navigation
Post by: Hardy Heinlin on Mon, 12 Feb 2018 17:47
Quote from: Hardy Heinlin on Mon, 12 Feb 2018 01:27
In case of a single FMC failure, with FMC L operative, I'm seeing two possibilities:

1. All CDU route memories get data from the operative FMC, independent of the NAV source selections.

Or:

2. CDU C memory gets data from the operative FMC; CDU L memory gets it if captain's NAV source is set to FMC L or CDU L; CDU R memory gets it if F/O's NAV source is set to FMC L or CDU R. (If the source selection is relevant, at least the selection of CDU C or of the inoperative FMC should disconnect the link.)

Just got a good hint from a trustworthy expert. I now take possibility #2 -- with a slight variation:

CDU C memory gets data from captain's NAV source selected FMC
CDU L memory gets data from captain's NAV source selected FMC
CDU R memory gets data from F/O's NAV source selected FMC

Title: Re: CDU standby navigation
Post by: mark744 on Mon, 12 Feb 2018 18:35
Sounds like a good plan

What if either pilot has the unserviceable FMC selected?    Map fail on ND? and Timeout/Menu on CDU?
until a valid input is selected? ie the good FMC or CDU s?

What affect should the FMC Master Switch have in this situation?
The QRH for a single FMC fail calls for the Master FMC Switch to be set to the operable FMC, so it must have an affect. What if it is set incorrectly?
(Is it selection of FMC used for FCC and  EECs ?)
Title: Re: CDU standby navigation
Post by: Hardy Heinlin on Mon, 12 Feb 2018 21:28
All these what-if questions are answered by PSX. Just try it in PSX and you will see :-)

I will only change that nuance discussed in the original post. The rest is correct, I'm sure.
Title: Re: CDU standby navigation
Post by: Hardy Heinlin on Tue, 13 Feb 2018 01:34
Quote from: Hardy Heinlin on Mon, 12 Feb 2018 17:47
CDU C memory gets data from captain's NAV source selected FMC
CDU L memory gets data from captain's NAV source selected FMC
CDU R memory gets data from F/O's NAV source selected FMC

These modifications are now available in PSX 10.24:

http://aerowinx.com/board/index.php?topic=4191.0


|-|ardy
Title: Re: CDU standby navigation
Post by: mark744 on Tue, 13 Feb 2018 12:38
Thanks Hardy for all the mods on this

A few comments from an old Boeing manual:

Boeing Operating manual
quote A
"The CDUs do not have a performance or navigation database. The CDUs continuously copy the active route from the FMC."

......At the moment in PSX, with both FMCs working and just F/OsNav Source Sel to CDU R, CDU R is not  'continuously" updating from any FMC


quote B
"If both FMCs fail, the CDUs retain flight plan waypoints except for conditional waypoints"

......so  they only retain the waypoints they have, at the point when BOTH fail (until then, they continue to copy from the FMC)

quote C
"Failure of the primary FMC causes both CDUs to display the MENU page and the scratchpad message TIMEOUT-RESELECT"

......I assume this is because the L FMC is the master and both CDUs are trying to display info from that FMS when it fails
If the Master FMC switch were then to be moved to R FMC, then I assume you get the behavior in quote D below

...this is not reflected in PSX at the moment



quote D

"Failure of the secondary FMC displays the respective CDU MENU page and the scratchpad message TIME-OUT RESELECT. Rotating the respective Navigation Source Selector to the operable FMC restores the CDU display and NDs to normal"

.......this is happening just like this in PSX


Thoughts anyone?


Title: Re: CDU standby navigation
Post by: Hardy Heinlin on Tue, 13 Feb 2018 15:50
QuoteBoeing Operating manual
quote A
"The CDUs do not have a performance or navigation database. The CDUs continuously copy the active route from the FMC."

......At the moment in PSX, with both FMCs working and just F/OsNav Source Sel to CDU R, CDU R is not  'continuously" updating from any FMC

In PSX the CDUs do not have a performance or navigation database either. So PSX is correct here.
Yes, in PSX the CDUs continuously copy the active route from the FMC when the related NAV source selectors link the related CDU with an operative FMC. This info in bold font is new, and modelled in PSX 10.24 now. This info is the most trustworthy info I have (less vague than other infos); it's being teached in an airline based on a large european island, if you see who I mean :-)

The Boeing Operating manual just says "continuously copy", but it doesn't tell you when. It's just a general, simplified statement like "eagles fly continuously", which is correct. But when do they fly continuously?


Quotequote B
"If both FMCs fail, the CDUs retain flight plan waypoints except for conditional waypoints"

......so  they only retain the waypoints they have, at the point when BOTH fail (until then, they continue to copy from the FMC)

Yes, they copy from the operative FMC when the NAV source selectors are set as instructed by the QRH (see bold font above).


Quotequote C
"Failure of the primary FMC causes both CDUs to display the MENU page and the scratchpad message TIMEOUT-RESELECT"

In PSX, the CDU which is linked with the remaining operative FMC will not return to the MENU page as this CDU is already linked with the operative FMC (e.g. CDU R with FMC R), regardless of the FMC master switch position. This behaviour in PSX is based on observations on a real 744 parked on the ground. I recall it was differently modelled in an earlier alpha version based on manuals, and then remodified after this real-deck observation. The FMC master switch position is also irrelevant to the EICAS thrust limit indications. It's relevant to the autothrottle (and LNAV/VNAV get a yellow line). I can remodify this detail again. But not before there is 100% evidence that this is a general, non-airline-specific feature.


Cheers,

|-|ardy
Title: Re: CDU standby navigation
Post by: mark744 on Tue, 13 Feb 2018 16:50
Hardy,

Thanks for taking the time to explain.

Maybe my manual was never updated, airline specific, or just plain wrong.
I'm sure your inside info has PSX absolutely correct now.

Regards,
Mark.
Title: Re: CDU standby navigation
Post by: Hardy Heinlin on Tue, 13 Feb 2018 17:09
Don't give up :-) I'm not saying your manual is wrong. I'm just saying that those words "continuously copied" don't tell the "when". Those words per se are not wrong, they're just incomplete, they just tell one part of the truth.

Regarding that return to the MENU page: I can't gurantee that this is the final truth for all aircraft for all times. It's just what I have learned so far. It may disagree with other observations. My philosophy is the method of falsification (https://en.wikipedia.org/wiki/Falsifiability); we can never know the final truth of all things at all times, so we can never claim the absolute truth. We can only keep the latest theory until someone discovers something different. Every step is just another theory (a better one, but never the final one).


Regards,

|-|ardy
Title: Re: CDU standby navigation
Post by: Jeroen Hoppenbrouwers on Tue, 13 Feb 2018 19:55
Background: you get TIMEOUT-RESELECT three seconds after the MCDU display feed of the currently selected subsystem stops sending keepalive data (words 172 and 377 basically). It happens also for ACARS, SATCOM, etc. Possibly this gives another clue on what exactly is connected through which switches/relays.


Hoppie
Title: Re: CDU standby navigation
Post by: Phil Bunch on Tue, 13 Feb 2018 22:16
From Hardy:

Regarding that return to the MENU page: I can't gurantee that this is the final truth for all aircraft for all times. It's just what I have learned so far. It may disagree with other observations. My philosophy is the method of falsification; we can never know the final truth of all things at all times, so we can never claim the absolute truth. We can only keep the latest theory until someone discovers something different. Every step is just another theory (a better one, but never the final one).

-----------
I've come to believe that physics also works this way.

Karl Popper was one of the best known advocates of this philosophy.  Science develops testable theories, and progress occurs when experimental evidence shows that the theory is wrong.  A theory is not shown to be correct, but can only be falsified. 

From wikipedia, an excerpt:

https://en.wikipedia.org/wiki/Karl_Popper

Popper is known for his rejection of the classical inductivist views on the scientific method, in favour of empirical falsification: A theory in the empirical sciences can never be proven, but it can be falsified, meaning that it can and should be scrutinized by decisive experiments. Popper is also known for his opposition to the classical justificationist account of knowledge, which he replaced with critical rationalism, namely "the first non-justificational philosophy of criticism in the history of philosophy."
-------------------

It's interesting that PSX is being improved in much the same way as modern physics. 

These comments are meant to be complimentary to PSX and its continuing development.

(FYI - I'm a <retired> physical scientist.)
Title: Re: CDU standby navigation
Post by: John H Watson on Tue, 13 Feb 2018 22:39
QuoteMaybe my manual was never updated, airline specific, or just plain wrong.

QuoteRegarding that return to the MENU page: I can't gurantee that this is the final truth for all aircraft for all times. It's just what I have learned so far. It may disagree with other observations.

Not sure if we are talking about the same thing, but my FMC fail ground observations seemed to vary from aircraft to aircraft (within the same fleet). At one point, I had a theory that the results varied with IRU alignment (but I can't explain why it would affect the results). Also, aircraft system behaviour does differ on the ground to behaviour in the air for certain systems, so anything is possible.

At one point we were pulling CBs for the FMC prior to every flight to flush out the data in the ATC (?) logs. Too much data was locking up the FMCs. However, this usually involved pulling both FMC CBs at the same time. 
Title: Re: CDU standby navigation
Post by: Jeroen Hoppenbrouwers on Wed, 14 Feb 2018 00:53
Quote from: Phil Bunch on Tue, 13 Feb 2018 22:16
(FYI - I'm a <retired> physical scientist.)

We are all engineers -- retired or not   :-)


Hoppie

When people speak about "mad scientists", they usually really mean "mad engineers". Real scientists do not build working machines of greatness -- they publish what-if games. Many so-called scientists are actually engineers, mad or not.






:-D
Title: Re: CDU standby navigation
Post by: torrence on Wed, 14 Feb 2018 02:48
In my world (space/planetary science) we get a real chuckle out of the label 'rocket scientist'.  There is nothing such as a rocket scientist - just brilliant rocket engineers who make it possible for us scientists to get some interesting data to mess around with  :).  I bow down before their altar. 

Cheers,
Torrence
Title: Re: CDU standby navigation
Post by: Will on Wed, 14 Feb 2018 16:24
"This task should be easy for you. It's not rocket surgery after all."