How would you interpret this VNAV situation?

[tkyler]

So this is my daily "I've been debugging/coding for 8 hours, I need a break, I'll type some brainless stuff in the forums".   I thought I'd share what I'm working on....and why its relevant, and why we can't just 'release' it without this stuff done.   So here's the situation....you enter a 'AT' or 'BELOW' altitude restriction at a waypoint not part of a departure or arrival procedure....and that waypoint happens to be in the enroute section of the route (Case 'A')....BUT?.....is it really in the enroute section?   'at' or 'below' restrictions generally propogate back towards their origin and affect T/C and T/D points.....but this is a case of asking, "what is the reference to apply it to"?  Case B shows the VNAV if you assume the restriction to be part of the climb.   Case C shows the restriction if you assume it to be part of the descent.  Case D shows it as not part of either and finally, Case E shows it as totally clipping the whole route such that you never reach your cruise altitude.

So how would you interpret this?   I think that the more you think about it, the more you would say, "well it depends....IF X, then I'd do this...if Y, then I'd do this...if Z, then this....and I could demonstrate why in each case, there are more potential combinations of waypoints and instances that would create an entirely NEW set of rules....and the code to handle every possibility gets REAL long.

So....why is this relevant?  Well according to big-daddy Jan Vogel....its common enough to enter altitude restrictions at waypoints while you are flying...especially at the behest of ATC.  And since we all want to fly on VATSIM and likely to get such a directive....the FMS really needs to handle it...or at least reject it if its crazy enough. 

You need to know, there isn't necessarily a right answer here...Jan has confirmed that these situations are very fringe cases....almost "pilot error" for entering it this way as procedures and flight regimes do reside in a well dedined box..and getting out of that box gets funky....BUT as long as folks have fingers to pry open the box, they'll find a way to explore....so we have to have something at the end of the tunnel.....back to debugging and coding.  

-tkyler

 

Edited February 10, 2016 by tkyler

[SwissairMD11]

Used in PMDG 737 3 times VNAV, and I won't press the button again as it's useless to me. However I do respect the work you put into it, my compliments. Have a nice day "ya'll"  

[SkyCoyote]

Allow entries for case B, and C and prompt the pilot with "That's a Rookie mistake" for anything else. Then you can roll this aircraft out of the factory before it is looked upon as the 1903 Wright Flyer. 

[bigbasspic]

I guess it comes down to the question: How would the real thing handle this? Do it the same :-)

Sorry Glad for being pragmatical, I am a Bass Player... ,-)

Bassy regards
Benjamin

Edited February 11, 2016 by bigbasspic
strikethrough!

[Litjan]

5 hours ago, bigbasspic said:

I guess it comes down to the question: How would the real thing handle this? Do it the same :-)

Sorry Glad for being pragmatical, I am a Bass Player... ,-)

Bassy regards
Benjamin

 

Hehe, that is our problem - we don´t know! I have never done that in 10 years of flying it - as there would be no reason to have a restriction at an enroute waypoint. But I know some of you guys will do so, so we need to cover it.

 

Jan

[bigbasspic]

2 hours ago, Litjan said:

 

Hehe, that is our problem - we don´t know! I have never done that in 10 years of flying it - as there would be no reason to have a restriction at an enroute waypoint. But I know some of you guys will do so, so we need to cover it.

 

Jan

Seems to be a step into no-man's-land, I would opt for the "That's a Rookie move: Expect the unexpected!" Just kidding...

In case of "B", "C" and "E" I would expect the plane to act as laid out, case "A" feels wrong to me, in case of "D" I would not expect automatic climb after the "AT" or "Below" point – as this has to be cleared anyway, or was defined in the flightplan. So worst case would be flying on the "AtOrBelow" level till the T/D, then joining the descend.
But as said, I am a Bass Player, not a pilot  

Bassy regards
Benjamin

[jimbim]

Isn't there anyone around that has access to a real 733 and can just test it for us ? :-)

Just kidding... But what about that LEVEL-D Simulator you already used Jan?

If it is good enough for real pilots, it should be good enough for us as well :-p

 

best regards,

Sebastian

 

[tkyler]

As said, there really isn't a right answer because the proximity of the waypoint  to either the T/D or T/C will make it "feel" different depending on it location.  It is a rookie move for the most part but we have seen how differing versions of FMS software might refine behavior after some sort of rule or heuristic seems to stand out...and that is what we have chosen to do.  This is fringe case...and we have applied a rule that reduces that fringe case to further fringe, thereby isolating it in about 99% of the cases to be "logical".  If one were to set out to really test every enroute waypoint by putting altitudes on every one, the results might get a little funky, but would, most likely, end up like CASE C

-tkyler

[X-Plane Junkies]

Of course I may be an idiot but to me it seems like it should behave differently depending on when you enter the restriction.  For example, if you are below the "At Or Below" restriction is entered the aircraft will climb to that restriction and stay there until it passes (Example B).  On the other hand if the restriction is entered when you are already above the restriction the aircraft will descend at the appropriate time to meet that restriction (Example C).  I am also back to coding flight sim stuff... just not for IXEG.

Rizzo

[tkyler]

Well lets say the overall route was 1500 miles...and the last enroute waypoint was 150 miles from the arrival runway...just before the T/D with enough time to get down. ...and you put a restriction of 11000B on that point (before takeoff, while on the ground, so you're clearly under it at the time of calculation).   It doesn't make much sense to fly 11000B for @ 1300 miles, quite inefficient, especially if you had a cruise alt at 35k or so.  It IS logical in such a case to think that kind of restriction should be applied to the descent, it would certainly be the cheapest and we know the airlines like that ....if said enroute waypoint was closer to the climb regime, then it might make sense to apply it to the climb.   From an optimization point of view, you'd apply it to the regime that is the cheapest...i.e. keeps you at altitude the longest.  If there ever was a "most right" answer, it'd probably be 'B' in the given example.

-tkyler

Edited February 11, 2016 by tkyler

[Litjan]

1 hour ago, X-Plane Junkies said:

Of course I may be an idiot but to me it seems like it should behave differently depending on when you enter the restriction.  For example, if you are below the "At Or Below" restriction is entered the aircraft will climb to that restriction and stay there until it passes (Example B).  On the other hand if the restriction is entered when you are already above the restriction the aircraft will descend at the appropriate time to meet that restriction (Example C).  I am also back to coding flight sim stuff... just not for IXEG.

Rizzo

Yes, but what if you entered "At or Below 6000" really close to the end of your 2000NM route (last waypoint before the STAR starts)? While you are still sitting at the departure airport?  Would you want the plane to fly at 6000 feet for 4 hours (its not very fast at 6000´) for the whole route?

 

Edit: Tom beat me to it!

Edited February 11, 2016 by Litjan

[tkyler]

I WIN.....(in my most evil Plankton voice)

[X-Plane Junkies]

7 minutes ago, Litjan said:

Yes, but what if you entered "At or Below 6000" really close to the end of your 2000NM route (last waypoint before the STAR starts)? While you are still sitting at the departure airport?  Would you want the plane to fly at 6000 feet for 4 hours (its not very fast at 6000´) for the whole route?

 

Edit: Tom beat me to it!

Yes... I guess I was taking the scenario of ATC adding a restriction that is not published...  I was assuming (which isn't a good thing to do in software development) that a Clearance/Tower/Departure controller wouldn't assign a restriction on the "arrival" portion of the flight.

 

8 minutes ago, tkyler said:

I WIN.....(in my most evil Plankton voice)

We will see if you win after release...  I would assume that if the controller/pilot enters a restriction that the plane cannot perform it will display a remark like "Go home pilot, you're drunk!"  Call it an easter egg for idiots.

[Litjan]

2 minutes ago, X-Plane Junkies said:

Yes... I guess I was taking the scenario of ATC adding a restriction that is not published...  I was assuming (which isn't a good thing to do in software development) that a Clearance/Tower/Departure controller wouldn't assign a restriction on the "arrival" portion of the flight.

 

We will see if you win after release...  I would assume that if the controller/pilot enters a restriction that the plane cannot perform it will display a remark like "Go home pilot, you're drunk!"  Call it an easter egg for idiots.

I think Tom meant that he won over me - speedy answer.

You are right - usually ATC wouldn´t add a restriction that far out of their jurisdiction. But what if the pilot thought "oh, I know these guys, they always want me to be at FL240 at KERAX already (this is a real-world situation!), I will just put that in right now, so the FMS gives me some warning on when to descend AND an accurate fuel prediction for EDDF...

[X-Plane Junkies]

5 minutes ago, Litjan said:

I think Tom meant that he won over me - speedy answer.

You are right - usually ATC wouldn´t add a restriction that far out of their jurisdiction. But what if the pilot thought "oh, I know these guys, they always want me to be at FL240 at KERAX already (this is a real-world situation!), I will just put that in right now, so the FMS gives me some warning on when to descend AND an accurate fuel prediction for EDDF...

I get that situation for sure...  I cannot confirm nor deny that I have ever assumed an instruction from ATC and been wrong.  I can also not confirm nor deny that I have flown flights online while drinking the fanciest of extra anejo tequilas.  

On a slightly ore serious note, maybe say if the "At Or Below" restriction is within X nautical miles from the T/C then hold the climb at the restriction.  If it is beyond X then climb to the set cruise altitude and later descend for the restriction.  X can be calculated the difference between the set cruise altitude and restriction and the time/distance it takes to meet the restriction... Plus a little buffer.  

But I do stand behind my idea of having the CDU display "Go home pilot, you're drunk!" because it is probably true in my case.

[tkyler]

we're not saying it doesn't.  I already let the cat out of the bag with my "rookie move" easter egg message caught on video, DOH   There'll be more

-tkyler

[torgeir]

Interesting to get some backstage (guitar player here) insight. Keep on with the jokes. Not rolling on the floor yet, but I do chuckle.

[KirMi]

Hello Guys,

maybe I can bring in some more light to this topic? I found this in "Bill Buffer/FMC User's Guide B737".

It covers the "Smith FMC" version U0 - U10, used in B737s (300-800).

Michael

 

 

[Litjan]

50 minutes ago, KirMi said:

Hello Guys,

maybe I can bring in some more light to this topic? I found this in "Bill Buffer/FMC User's Guide B737".

It covers the "Smith FMC" version U0 - U10, used in B737s (300-800).

Michael

 

 

Oh, NOW you are telling us...

We actually came up with a solution quite similiar - we will allow entry at all enroute waypoints, but based on total distance along the route we will assume they are "climb" or "descent" restrictions. There is the chance that you ment a restriction to be a "climb" one, and we interpret it as a "descent" one (based on proximity to T/D), but then there is only so much mind-reading that the FMS is able to do. Some pilot sanity is required ;-)

Jan

[KirMi]

Sorry for very BAD timing!

I had to dig thru very dusty boxes, but now I know were the book is ;-)

Michael

p.s.: We're all praying for sane pilots!

 

[BaBene]

Maybe Jan was sarcastic and I didn't get it, but I would be quite surprised if you guys at IXEG don't have a whole bookcase full of books concerning the FMC, Systems, etc. of the 737. I mean, don't get me wrong, but the smartest and most obvious thing to do if you're facing problems like this, is to look it up. Either from trustworthy (if this actually do exist ) sources on the internet or from a book. Especially when the book is, judging by a quick internet search, quite affordable. Time is money.

 

Cheers,

BaBene

Edited February 12, 2016 by BaBene

[LPNils]

They probably have some manuals. My guess. Do you guys?  b737.org.uk also had some manuals for sale some time back. Haven't checked in a while. So expensive. Plus, I have quite a lot of digital manuals. 

But even the books sometimes do not tell the whole story. Or they sometimes are ambiguous as to what the actual situation will be. 

[aljaz41]

Which version of the FMC are you implementing? I'm curious becuase the picture from the manual Michael posted is valid for versions from U1 to U6 and searching through google I also found an online shop where these manuals can be purchased from, which I actually might do.

[tkyler]

6 hours ago, BaBene said:

but the smartest and most obvious thing to do if you're facing problems like this, is to look it up

...If we were out to simulate a specific FMC software version, then i'd agree....but software is updatable, versions change and we can always take the position of, "well we have the latest version that is 'one better' that the ones real airlines have"    "....this one goes to 11"...so we did not target a specific software version.  I'd guess that they all have about 99% similar functionality though.  I'd go with the latest version (assume that its the most modern in theory) and then you can just hound us if we don't have the feature.

but the FMS page above does bear out that there are "what ifs" and "it depends" when it comes to this situation and FMSs have taken different positions on handling it over differing versions......so we took that tact of providing our solution that we feel is "most expected" for the given situation given Jan's input.  The fact that our solution mimics U7.0 functionality is a form of confirmation that we're all thinking the same thing.   

-tkyler

Edited February 12, 2016 by tkyler