VNAV Profile and AT Mode

[3-2-1-Now 126]

Hi,

 

I'm not sure if this is a VNAV problem or really an auto-throttle problem, but several times now VNAV has bust the vertical profile whilst on descent.

 

The required descent angle to maintain profile has been shallower than the aircraft has been flying, putting it below path. When it has done this, the thrust has been in HOLD mode. The AT system does not appear to want to engage to maintain sufficient thrust once it enters this mode.

 

If I manually add thrust (mode still HOLD) then the aircraft will level off and capture the path, but I'm basically in manual thrust at this point, which is not what I would expect.

 

VNAV SPD INTV makes no difference - it just alters how fast VNAV flies, but it has no material affect on the AT mode and whether it flies the required path.

 

Does anyone have any insight into what it is doing and why I'm essentially having to manage thrust manually?

 

I've also had it bust the vertical profile during RNAV approach segment, ending up too high. It's like VNAV doesn't really care about the path at all. In this case thrust was idle.

 

Best regards,

Robin.

[FLEX1978 373]

If in HOLD just add thrust yourself, also ensure your descent winds and predictions are entered and accurate

[3-2-1-Now 126]

I understand that, but what I want to know is why won't the AT system manage thrust on its own? i.e. VNAV pitches for path, and AT takes up the slack for the speed?

 

It seems once it enters HOLD mode you're on your own with the thrust...

 

Best regards,

Robin.

[FLEX1978 373]

Well that is what HOLD does.  A/T will kick in & revert to speed if required. 90% of the time I will leave it idle in HOLD, sometimes I will assist the descent path with manual thrust in HOLD if winds are not as predicted and my descent is off target.  

 

The real machine is not 100% accurate, hence why two guys are up front with many tools at their disposal

[scandinavian13 5,192]

 

 

The real machine is not 100% accurate, hence why two guys are up front with many tools at their disposal

 

What??? There's a reason people are still up front?  I thought they were just supposd to watch the dog...

[KriVa 307]

HOLD is exactly what you describe... A/T is armed, but not actively controlling speed.

[Driverab330 74]

HOLD basically the pilot has to control the speed.l by moving the thrust levers. Because we don't normally have a thrust levers for flt sim which is back driven by the A/T system so it is quite hard to understand why it's there.

 

VNAV calculated its descent profile based on uplink or pilot input descent wind before TOD. Which of course won't be 100% accurate. It will only re calculate the path again during descent if you perform a DIR to another way point or change the speed/ALT restriction. So after reaching TOD the A/T mode would change from IDLE to HOLD and the airplane fly down in VNAV PATH tracking the FM calculated profile. You can imagine the PATH is an imaginary Glide slope calculated by the FM. So imagine on final in G/S there's wind shift the IAS changes and the airplane is descending on a constant geometric path, we need to maintain speed by moving the thrust lever if we are using manual thrust. Same applies for VNAV PaTH with HOLD mode on descent before approach phase is activated after lowering flap 1 or after the IAF waypoint on the legs page of the FM sequenced, in approach phase the A/T will be in speed mode and vertical remains on VNAV PATH.

 

On the real airplane People use 280kts on descend especially coming out from a jet stream with 100kts tailwind or more to protect against over speeding the airplane. I normally allow -5kts from target speed up to 320kts before I intervene with either thrust or speed brake because most of the time the speed change is transient.

 

For approach to avoid VNAV gave you the wrong profile, make sure on your legs page the speed on the way point is logical. Because if you perform DIR to a waypoint, VNAV modify the waypoint speed so that in extreme case it puts 240kts at 15nm from Touch down which will be way too fast. So real world pilots modifies the speed on the legs page so that is something sensible of course procedural restriction has to be respected at all time. If the IAF is within 20nm give it a speed of 210kts depends on ALT restriction of the approach. If you have a waypoint 15nm from TD with 4000' alt constraint give it a speed of less than 200kts but if you see 240kts you know the path is wrong.

 

VNAV is useful as long as you know what alt and speed it is trying to achieve further down the track.

 

Hope it helps. Happy landings.

 

---- a bit more back ground info on VNAV descent after digging into the books --------

 

It divides into two phases "above speed transition altitude" and below "speed transition altitude" which by default is 240/10000ft on the VNAV DES page. -- speed transition altitude is an altitude where we have to slow down from high speed to below 250kts unless high speed is granted by ATC in some part of the world, it's not the same as the QNH transition altitude where we change our altimeter from local QNH to 1013/29.92 --

 

Above speed transition alt which is 10000ft by default. If in VNAV PATH and the IAS goes 15kts below Target speed, the A/T will wake up in SPD mode and increase the IAS back to target speed.

 

below transition altitude 10000ft by default in VNAV path before approach phase, the AT will only wake up from HOLD to SPD when the speed reaches Vmms (too of the amber hockey stick).

 

Since I am not very good at remembering stuff. Feel free to correct me if I am wrong. : )

[3-2-1-Now 126]

Hi,

 

Thanks for the feedback so far, but I think I need to re-state the question.

 

Why is VNAV busting the vertical profile? It will only fly the profile if I manually add power. It holds the speed like FL CH. It is NOT flying a path.

 

If it was flying as expected, then it would be on-path and the speed would drop to the stall if I did not add power, but this is not how it is behaving.

 

Best regards,

Robin.

[777simmer 117]

I understand that, but what I want to know is why won't the AT system manage thrust on its own? i.e. VNAV pitches for path, and AT takes up the slack for the speed?

 

It seems once it enters HOLD mode you're on your own with the thrust...

 

Best regards,

Robin.

The problem is two fold I think.

1) you are not in VNAV PATH but in VNAV SPD mode!

2) there seems to be a bug that prevent the AT mode changing from HOLD to IDLE once back on Vnav parh.

 

Do you remember the discussion we recently had where I said that the FMC calculated Vnav path is like a virtual line that remains there and does not change as long as you make no changes that cause a recalculation?

 

Well, you are below that imaginary caclulated line and thus Vnav jumps out of Vnav Path mode and into Vnav Spd mode!

Anytime you start a descend from cruise FL, before reaching the TOD point, you are in this situation.

This is actaully a normal situation.

We allways start a shallow descend slightly early (set lower MCP altitude and click the altitude selector before reaching TOD).

You end up a little below path this way initially (the path deviation indicator on the ND will increase untill you fly past the TOD point at which point the planned idle descend will catch up with you again.

with:

-AT in HOLD mode

-LNAV

-ptich mode is VNAV SPD.

-a shallow descend untill the path is reintercepted

 

 

Vnav Spd mode is similar to a descend in FLCH, but smarter (just not as smart as you, the pilot)!

Vnav Spd is smart in that it should adhere to all altitudes that are programmed into the FMC (FLCH does not). It should not bust them, if it did then that would be a bug!

 

And, ".......drum ruffles....., Vnav Spd is smart in that its goal is to get you back on path!

Read that again, because it is something many do not realize!

Vnav Spd is smart in that its goal is to get you back on path!

(but it is stupid at the same time, read on)

 

When you go "hey, why is the AT adding thrust? all I want is a descend" you are probably in Vnav Spd and below path (you can see that on the ND path deviation indicator) and all the thing is doing is trying to get you back on that vertical imaginary path.

 

Now here come the silly part, Vnav Spd is stupid at the same time, as it does not really know where (at what point in space) you want to reintercept your original virtual Vnav Path line! (but it should be back on path at/before the next programmed altitude constraint.)

Do you want to be back on path a.s.a.p? (a climb would be fastest if you are below path, but that is crazy ofcourse) or,

Do you want to intercept in 20nm or so? (maybe level flight would do in that case) or,

Do you want to intercept the path in 40nm? (which could maybe be achieved by a shallow descend untill intercept).

 

So since Vnav Spd does not know this, it just sets a little bit of thrust to maintain a shallow descend rate (I am sure it states the target descend value in the books somewhere, but it usually gives you something like 1000-1800ft/min) and then it waits untill the path is reintercepted.

And sometimes it never does (thrust too low, shallow descend not shallow enough).This is where you come in to adjust the thrust value!

In the mean time the AT is in HOLD mode so that the pilot can vary the point at wich the path is intercepted :-)

So thrust controls descend rate in this mode rather than speed.

Speed is controlled by pitch in Vnav Spd. More thrust at the same speed = less descend rate.

 

When it intercepts the path after this HOLD phase, then you should see the pitch mode change to VNAV PATH and the AT mode to IDLE (If you are using speed intervention then this will not work. So cancell speed intervention, if you were using that, once you are on path!).

 

****And I have seen the PMDG777 make the mistake of NOT going to IDLE here!*****

 

I did a few tests a few days ago and the AT stayed in HOLD mode which give that impression that you are in (unwanted) manual control of thrust!

This is a BUG (I have to do a few more test to be sure, but I was about to write PMDG this.....or should I wait untill after SP1, lol).

Once on Path the AT should pull back thrust to IDLE and continue the idle Vnav descend.

 

Unless...oh my this is getting complicated.....unless you are in a phase between two programmed altitude restrictions that requires only a shallow descend. Lets say and 800ft/min path from A to B. In that case you should see the AT change to SPD mode once Vnav Path is intercepted. (pitch is maintaining the required VS to stay on path and AT adjusts thrust for speed).

 

Pffff, does this help or make things more complicated :-)

If it was flying as expected, then it would be on-path and the speed would drop to the stall if I did not add power, but this is not how it is behaving.

 

Can you explain what you mean with that?

[3-2-1-Now 126]

Hi,

 

Great post! You nailed what I saw, namely that it was not going back to VNAV PTH on re-intercept, and HOLD was remaining instead of IDLE.

 

Can you explain what you mean with that?

 

If VNAV PTH is flying a path, and let's say it is descening at 500 ft/min then if it remains at idle the speed will bleed off until it stalls out, so I would expect it to be in some AT mode other than IDLE to ensure that it does not lose speed. Instead, VNAV PTH (and I need to check but I think it was actually VNAV SPD as you mention) was acting like FL CH in that it was just putting the nose down to hold speed and ignored the fact VDEV was increasing (whiuch for the mode would be correct, only it was in the wrong mode).

 

I'll try and fly an apporoach this evening and see if I can confirm this.

 

I understand what you are saying regarding starting DES before TOD point. I do it this way as well using the descent circles for guidance (I can't remember its name right now - had my head in another aircraft's manuals all day). Last night when I flew I left it until the TOD point to see how it would cope, and well, it didn't go as well as I thought it would due to the above problem.

 

I'll try and do a flight this evening, and will get copious screenshots which I'll post here with my findings, but what you posted is what I thought as well - that it should change back again when it re-establishes the path, but for some reason it isn't.

 

Best regards,

Robin.

[777simmer 117]

VNAV PTH (and I need to check but I think it was actually VNAV SPD as you mention) was acting like FL CH in that it was just putting the nose down to hold speed and ignored the fact VDEV was increasing.

 

Yes that is what Vnav Spd does, just like FLCH, it does not follow a path, and with the AT anywhere between idle and full thrust (wherever it was left after HOLD engages) one can only hope that Vnav Path is re-intercepted (this is normal and requires manual adjustment of the throttles sometimes).

 

 

I still dont get the stall part, though, but I think that is not the important part of the discussion here so we can leave that for now I guess.

 

But as additional info, Vnav Path has a path it will try to follow with a target speed (on the dscend page).

If actual speed deviated too much from target speed, then at some point two possible things can happen.

 

1) You get the "drag required message" calling for speed brakes.

(example: Thrust is already at idle but you have stronger tailwinds than anticipated. The higher groundspeed during descend requires a higher descend rate (VS) to stay on the precalculated path. So the AP puts the nose down more (to stay on path) and thus speed goes up and up and up unless you extend speed brakes or throw the gear or so)

 

2) the AT wakes up from IDLE and starts adding thrust.

(example: stronger than planned headwinds have reduced ground speed and thus VS required to follow your preplanned path is less. The AP increases pitch to reduce VS and with the AT in IDLE, you loose airspeed).

 

However once a certain deviation below target speed (FMC descend page) is reached the AT wakes up and starts adding thrust to get back to speed (while Vnav Path is still maintained by pitch).

So it would never stall. Because it does not stay in IDLE.

 

Looking fwd to your screenshots and results.

 

Vnav can be a challange but it is cool isn't it ;-)

[3-2-1-Now 126]

Hi,

 

Flew the LAM4A arrival into Heathrow last night, and I think I see what is going on.

 

It is initially in VNAV PTH, then it goes to VNAV SPD. For some reason on this arrival it wants to decelerate to 240 kts at around FL200 which is far too early, so I went for SPV INTV at 300 kts. This is where things get interesting. It intercepts the vertical profile, but as long as you're in SPD INTV it will remain in VNAV SPD mode even though it is blowing through the vertical profile. The instant you switch from SPD INTV to VNAV managed speed it will select VNAV PTH and AT SPD mode and capture the vertical profile.

 

Now for the interesting part. If you press SPD INTV again, it goes OUT of VNAV PTH and back to VNAV SPD! It did this most of the way down the arrival, until at some point it decided it would stay in VNAV PTH and allow SPD INTV as well.

 

For some reason the VNAV mode cares if you're in SPD INTV which I don't think it should - it is merely a way of overriding the desired speed.

 

Best regards,

Robin.

[FLEX1978 373]

Robin, VNAV path is the calculated profile using the FMC speed profile.  If you intervene that whole calculation goes out of the window. 

 

The logic you are describing seems correct.  I am about to fly from EDDF to EGNX,  will do a few tests myself, will also send your post to a 777 captain.  

[scandinavian13 5,192]

 

 

I am about to fly from EDDF to EGNX, will do a few tests myself, will also send your post to a 777 captain.

 

PM me what you find out about the scenery issue.

 

 

 

Also, to all:

The earlier commentary wasn't made to downplay the topic here.  It was just generic commentary in an effort to be funny.

[3-2-1-Now 126]

Robin, VNAV path is the calculated profile using the FMC speed profile.

 

VNAV PTH doesn't care for speed. It's like an ILS glideslope - it is a specific angle regardless of whether you fly it at 100 kts or 1000 kts. VNAV PTH will change the vertical rate to suit.

 

That isn't really the issue though - the problem is the modes don't seem correct, and VNAV SPD doesn't like to revert to VNAV PTH.

 

Try the LAM3A arrival into Heathrow, RNAV RW27R. Why does it want to be at 240 kts at FL200?

 

Fly East of LAMBN by about 15 nm and turn around. Be at FL300. Use SPD INTV to keep the speed up (320 in the descent is fine) and tell me if it ever goes into VNAV PTH despite FLYING THROUGH the path in VNAV SPD during the initial part of the descent. To me it looks totally screwed up.

 

If you take it out of SPD INTV it will want to slow to 240 kts, but it will then also go to VNAV PTH. I find at this point if I try and SPD INTV it will revert to VNAV SPD! End result is having manual thrust to fly the path it should be flying in VNAV PTH mode. Extremely annoying.

 

Best regards,

Robin.