VNAV Path construction and Flap Speed
 

Hello Technical Types,

I am trying to sort out a question regarding the above. Particularly the 737 with 10.8a.

When constructing the planned VNAV path is the FMC expecting you to reduce to minimum maneouver speed for the current config. before selecting flap ie. within about 10kts of UP speed before selecting F1 and so on; or is it expecting you to select flap just under flap limiting speeds, which would seem like using flaps as drag to me, but there you go.

I've consulted all the usual sources and cannot find a reference either way, the implication in Vol.2 and Billy Bluffer seems to hint to the former. However many of my colleagues are prepared to stake their mortgage on the latter "...its why you always end up hign using VNAV!"

Any thoughts or better yet references you may have are gratefully recieved.

"Always end up high"? I don't have that problem.

I bet that's the same person who exits VNAV because LVL CHG will magically give them a better descent rate than VNAV SPD at the same IAS.

I'll do some digging, but I bet your sources are as good as mine.

VNAV expects a continuous deceleration to the proposed flap setting when coming back to clean configuration.
A seasoned aviator will roughly know how long it takes for flaps to extend, thus ensure you select the next flap setting required for phase of flight with that time in hand, e.g. approximately clean speed +20kts works reasonably well to get flap 1 and immediately afterwards 5 out in time for approach segment, depending on angle and weight flap 10 may be required to maintain speed on the subsequent final descent angle coded.
Think of it that thrust levers stay idle as long as possible until throttling up to maintain speed on intermediate approach segment before gear goes down and further deceleration takes place...

Unfortunately I have yet to find references to the question you're asking...

From the 747 FCTM 4.21:Since the flap speed for FLAPS 1 is 280 KIAS, and the descent profile has a speed reduction to 240 at 10,000', extending flaps at or near flap maneuvering speed should be assumed.

Further (4.22):...and on the next page:Going forward to 5.7:HOWEVER, the FCTM doesn't seem to address a constant descent approach. So, in order to be able to decelerate while descending, our FCOM used to say to start extending flaps at 17 miles instead of 12 (it is now silent on the subject...).

"Angle"? A CDFA is a Constant Descent Final Approach and a CDA is a Constant Descent Approach. I have not heard of a Constant Descent Angle Approach and it would be illogical if there was one, as you would, as you say, have to add drag at various points to stay on the constant angle to reduce speed, thereby wasting fuel.

VNAV is a path both vertically and with a speed consideration. The most common mistake is for guys to chase the magenta path indication using VS and then arrive at the decel point way above the VNAV path calculated speed. Re engaging VNAV at this point will cause the aircraft to fly level until with the FMC speed parameters, thus giving the impression that VNAV leaves you high. The ND clearly shows the green decel dots, the point at which the vertical rate will erode along with the speed to the speed schedule shown in the legs pages. Obviously if one fails to extend the lift devices, or take account of tail winds or anti ice then the issue is basic airmanship and lack of understanding of the systems. Equally, the messages " unable next altitude" or " bypass" should be an indication that somehow you have managed to overcook the approach, a quick check of the Legs Data will then show that according to your advanced planning you have asked the machine to arrive in two places at the same time , or loose 2000 feet and 40 Knots in 1 mile for example. The Systems manual navigation section has comprehensive descriptions of the process: for instance:
Normally, the target speed is economy speed above the airspeed restriction altitude and 240 knots below that altitude, until deceleration for approach. VNAV will not permit descent below the airspeed restriction altitude until the airspeed is at or below the restricted value plus ten knots. The start and end of the airport speed restriction deceleration segment is shown on the map as a green open circles with no labels.
The descent path assumes deceleration to reach the final approach fix (FAF), or the glideslope intercept point at VREF 40+20 knots.
Target speeds are changed by entries on the DESCENT page. Entries made on the LEGS page are “at or below” and may limit the target speed. Wind and thrust assumptions are changed on the DES FORECASTS page.
Deceleration points show on the map as green open circles with the label DECEL. Deceleration points show prior to:
• airspeed constrained waypoints
• holding patterns
• approach flap extension.
If more than one deceleration segment exists in the flight plan, only the next deceleration point shows. Deceleration points can also show prior to cruise holding patterns or other speed reductions.

Bloggs,

CDFA = as you say
CDA = continious descent arrival

Hm?

Off topic,

Pans ops 8168

CDFA is the constant descent final approach and there is no level off segment.

CDA is constant descent angle type of approach and there is a level off segment!

Ah, bureaucracy at it's best...

On a related note, regarding VNAV path construction, aside from speed brakes, flaps or gear, the only other way to reduce airspeed during the deceleration phase is to reduce the descent rate. I think the original poster's question -- whether you are expected to add flaps near the maximum flap extend speed to decelerate and stay on the path, or reduce descent rate until close to minimum flap maneuver speed -- can be asked in a different way:

How does the FMC deal with the deceleration "knees," which must be calculated in advance and inserted into the predicted overall idle descent path? I noticed the vertical deviation indicator stays centered while the vertical speed decreases to about 500 fpm during the deceleration segment. Obviously, the path itself is not a constant gradient if you can change vertical speed from 2400 fpm to 500 fpm while decelerating (aproaching 10,000' or some other speed restriction point) and continue to stay on the path. I suppose you can keep the 2400 fpm descent rate going and decelerate if you add flaps just as you get below each max flap placard speed, but this is usually not what is preferred.

There must be some target speed that the FMC is anticipating for these deceleration segments. Is it the minimum flap maneuver speeds, maximum flap placard speed, or some speed in between?

Indeed. But that's how they had to separate the 3 types of NPA approaches back then to establish design criteria for each procedure. But it's there written.

CDA today means for me constant descent arrival, like LHR. Which doesn't really have to be constant in terms of angle, but rather a constant descent achieved.

I suppose you can keep the 2400 fpm descent rate going and decelerate if you add flaps just as you get below each max flap placard speed, but this is usually not what is preferred.

Indeed, frowned upon and I doubt very much you could achieve the performance you speculate on. There is not enough drag in the flaps to succeed in that.

Flat Cap: I can only tell you what I've observed. First some basics, which Avenger has touched upon. VNAN PTH is an energy profile not a strict descent profile like a glide slope. If you are descending towards a WPT with an ALT/SPD constraint the ALT has priority at the sacrifice of speed. When descending to 250/100 then 250kts has priority.
The forecast winds entered on DESC page help with the calculation of TOD but are then not updated during descent, automatically. If the real world & winds are different the a/c will try to stay on the calculated path profile and the speed will be sacrificed. You can update the FMC to actual wind by reselecting DGT TO the active WPT and this actual wind is now used as if it was at TOD from np]ow to landing.
When decelerating approaching FL100 it generally goes 500fpm if on an accurate path and accurate descent speed. If the speed is higher than FMC speed it could even level off at FL100. This can cause you to go high. It could then dive for the path for the next WPT constraint and accelerate above 250kts in doing so. You'll have seen 2 green rings on the map showing the calculated start & end points for decel to 250kts and then continued descent.
The next green ring you'll see is DECEL. This is the calculated point where F1 is expected. Assuming this whole profile is flown in VNAV PTH with A/T as you approach this DECEL marker you'll see the ASI speed bug start to reduce towards UP. The a/c goes 500fpm to decelerate. If its got it correct you should hit DECEL at UP - UP +10kts. Taking F1 s now your decision. The FMC has calculated the reduction in energy from this point in space to achieve the next constraint which maybe IAF or OM. You'll see the SPD/ALT in LEGS. There are no more DECEL rings, but you'll see the ASI speed bug reducing and you are supposed to match the flap setting as the speed bug reduces below flap man speed.
It is a real gotcha if you forget the flaps and end up slow, but there is a speed buffer between Flap Man bug and absolute Min Man speed; that is why it is very common, IMHExp, for many operators to either not use VNAVPTH below FL100, or t go SPD INV. You then control speed and not FMC. A variety of reasons: it improves your direct control of the a/c; the ATC route may not be the same as FMC route and so the energy profile will be different.

But it ain't calculated on using flaps at max speed.

Ahhhh......VNAV.....where art meets science. It's all about energy management old boy.

Indeed. And sometimes it threw up some real gremlins, which highlighted serious lack of appreciation of energy, and therefore training. I didn't know the answer at the time, but that was not important: I watched PF RHS sitting idly by in VNAV-PTH, on the pipper, on speed. I suggested they look at PROG 1 and distance to TDZ and then the altimeter and consider if all was well. "But we are in VNAV PTH." so what can be wrong. I pointed out the discrepancy and we decided what was the best course of action. That in itself was a small adventure. On the return sector we analysed the previous 'curiosity' and came up with a reason hidden in the FMC. The PF was astounded how this could happen. It all looked so good. But why were they not applying their own 'on-board computer'? Another lesson ticked off that should have been done in LT and continually since.

Thanks for all the input so far, much appreciated.

I had never heard the "VNAV leaves you high because we take flap at min maneuver rather than flap limit" until I arrived at my latest company.

Anecdotally my on the line observation of what VNAV does seem to agree with most peoples. Such that the VNAV path constructs its "energy profile" (I like that, going to steal it), that if you fly a standard Boeing decent and approach profile as in the FCTM you generally end up in the same place as the VNAV. (Subject to correct FMC programming and the forecast info you've enter being similar to the actual met you encounter.)

I struggle to accept that Boeing would want the FMC to construct a profile that requires you to use flaps as drag on every approach, when they are at such pains to ensure you don't do that when you are working the MCP or manually flying. But I would like to be able to point to it on a page.

Not sure Boeing helped those new to FMC VNAV by labelling the numbers on the path as "Feet" in Vol. II. when as you say it actually combines both speed and height ,foot/knots, units of path, any other name except straight "feet" might have helped.

So the consensus so far seems to be that "[t]he a/c goes 500fpm to decelerate. If its got it correct you should hit DECEL at UP - UP +10kts."

Flat Cap,
If you do find something in writing that addresses this, I would be interested too. So much of Boeing's VNAV is a mystery to those who operate the airplanes.

that's why they call it "continious descent arrival, i.e. without recourse to level segment." There's nothing constant about that descent. [OFF]

So much of Boeing's VNAV is a mystery to those who operate the airplanes.
That's one good reason, in the early days, to let the automatics do their thing and watch. It's the best way to learn: far easier than reading manuals; and as you say there's not a lot there.

My most fav VNAV observation that tells me instantly that it is not understood goes like this. "I'm starting down early to get a head of the path." Ok says I and nobby promptly selects DES NOW or mashes that there altitude selector. We now start a shallow descent and THR HOLD presents itself. We continue on that 1200 ish fpm descent until re intercepting our original flight idle descent path that was 40 miles in front of us when we started down on this oddisy. The TLs go to IDLE and "What did we just accomplish" says I to a blank stare.

This may be airplane specific, and totally a matter of technique, but some guys on the B744 will bring the first constraint on an arrival down to the scratchpad and add a /-4 after it, then insert that waypoint over the first constraint waypoint and insert the same airspeed on the right as the respective constraint waypoint. This fools the FMC into planning on being where it should -- but often isn't -- about 4 miles prior. It seems to work well for them and it beats having to pull out the boards after it becomes apparent that the airplane is not going to make it down.

A few days ago I had a T.E.M class and the FOQA guys gave us some nice numbers to remember (data collected using a huge number of monitored flights) and help managing the energy during approaches that are going south...

1) If you're 1000ft high on the VNAV profile you'll need (using speed brakes):
- 6nm at 280kts
- 10nm at 250kts
- 12nm at VREF40 + 70

2) For every mile on a 3deg glide path you will lose, roughly, 10kts (SPEED DECREASING WITH GEAR DOWN)

3) For every dot high on the glide slope (<10nm from threshold), you will need 1 extra mile to get back to the profile (SPEED DECREASING WITH GEAR DOWN).

4) You'll need 3nm for every 1000ft high on the VNAV profile (GEAR DOWN / DECREASING SPEED)

5) You'll need one extra mile to dissipate the energy generated by 20kts of tailwind on a 3deg glide path (GEAR DOWN).

So... For example: Let's say you aim for 150kt at the FAF (5nm from RWY) and you're with 190kts, 1 dot high with 20kts of tailwind. When should you ask for the gear down ?

19 - 15 = 4 + 1 (high on the glide) + 1 ( 20kts of tailwind) = 6nm from the OM. According to what they observed on all those flights (900+ daily ops), you should ask for the gear around 11nm from the runway in this case.

It was a pretty nice class... Some great recordings of messed approaches !

On the other hand, I still prefer using V/B & V/S indications on the DES page. Much more straight forward with far less thinking !

Hope it helps :)

Wow. That's a lot to remember! It is interesting looking at the FOQA data in ground school for approaches that guys really botched; but I think it is good to keep in mind that it's easy to Monday-morning quarterback. And usually the FOQA guy, who is shaking his head in amazement at how badly it was botched, spends little time as the Man in the Arena, flying the line.

B737SFP: That would have been mind-exploding for most of the cadets I've had through TR's in the past 15 years. Ask them questions like that and the first thing that is brought out is their Japanese brain. Simple mental arithmetic skills seem to have disappeared.
What I found missing on the line was the appreciation of quick 'gross error' checks. This could be load sheets, flight profile, sector fuel required, etc. Perhaps, one is they were never taught, and two is they require mental sums.

I'll start down early if cruising above FL350 with 4nm/1000' before TOD. I find the NG tends to overspeed in the descent above FL350.

But that's 4nm early for FL360, not 40mn!

Decel segments are at -1.3° which can be about -500fpm.

I find the NG tends to overspeed in the descent above FL350.


Do you mean if TOD is >FL350, or for the 1000's when above FL350? But given that you are still >100nm from touchdown does a little overspeed matter; and that depends on your CI anyway.

I mean if the cruise is above FL350, as we initiate descent, the plane often pushes close to MMO. We use .78/280 in the descent usually. Even with automated descent winds, the plane will still get close to the clacker.

I'd prefer not to have to write up an overspeed 100nm, 50nm or 2500nm from touchdown.

Well my point was that if you start down early but leave the jet to do what it is programmed to do you will end up back on the idle path anyway. If you want down early and to "stay" ahead of the path then the thrust levers need to be closed when the servos release i.e. THR HOLD. This applies to the Boeing crates I have flown anyway that does not include the 73NG btw.

After you have programmed your legs page for a descent,connecting the arrival/approach without any discontinuities, you must have a bold speed, and altitude on your legs page, at a point, at the commencement of your initial approach segment,preferably in level flight.This has now instructed the FMC that you intend to arrive at that geographical position, at a specific speed and altitude. The next step is to program the variables into the FMC. Go to the descent page and select forecast. On the forecast page, enter your cruise altitude with the wind direction and speed (just before descent). Do the same for a medium altitude( during descent) and your destination altitude. e.g.If cruise was at FL350, enter the altitude,wind direction and speed.
Flight level. Wind SPD. Direction
FL350 260. 30
Fl160. 220 40
00300. 180 15
If you are unable to get winds at the desired altitudes, you can enter another altitude. It doesn't matter if it is an even or odd altitude. Neither does it matter if the altitude ends in 0 or any other numeral.
This allows the computer to calculate a descent path, considering the winds(Headwind or Tailwind). The final part is this.The computer calculates the descent path to arrive at the position mentioned earlier on, with the "hard" speed and altitude.The only thing is that it assumes that altitude is with a standard QNH(Altimeter setting) of 29.92 or 1013 mb.You now have to apply the variation from the standard setting, on the same descent forecast page, i.e. temperature variation from i.s.a. and the actual altimeter setting at the destination. You will see some green circles on your legs page.They indicate where the Aircraft has to change it's pitch/descent path, for a change in speed.The first two closely spaced dots represent when the Aircraft starts to slow down to 250kts, just before 10,000 ft.The second on represents when the Aircraft is slowed down to 250kts,the resumes the descent path.
The next dot represents when the Aircraft starts to slow down, in order to achieve the hard speed,and altitude programmed earlier. As the Aircraft is approaching the clean maneuvering speed + 10kts, or as shown in the FCTM/FCOM, the flaps are lowered to the initial setting. Once the initial flap setting is achieved, the speed bug automatically moves to the maneuvering speed for the initial flap setting.This process is continued for each stage of flaps, with speed or glide slope as the trigger for the next flap setting, as shown in the FCTM/FCOM.
If the FMC is programmed properly with the previous data(accurate wind direction/speed etc) it will give you the required performance. Remember "garbage in, garbage out"
​​​​​​If you do have any questions,feel free to ask.

Shinobi, thank you, very useful info.
Just a question.

The ND will show a DECEL point. Is this the point where the aircraft begins reducing speed to UP speed or is this the speed it wants FLAP 1?

90% of the time VNAV will work and can be trusted to bring you at the intercept at the right speed and altitude.
However there are times when it gets you there a bit too fast or with ATC shortcuts, a bit too high.
Of course people always forget to "update" the active waypoint from time to time as this also updates the wind info in the FMC and alas updates your path.
The same goes for updating the FORECAST page (especially the use of TAI).
But the FORECAST page is not of too great importance - it's a FORECAST after all.

On the ND(Navigation display), that open green circle (deceleration point), shows where the aircraft will start decelerating to the speed which is shown in bold on the legs to page.If you also refer to the descent page,you will also see the position referred to, with a display showing the "hard" speed and the flap setting required.
The more accurate the wind direction, speed forecasts, and ISA deviations input on the descent page, the more precise the FMS will maintain the desired flight paths, and speeds.
LIke I mentioned earlier, for the computer, it's "garbage in,garbage out".

Not sure what type you're on, but on the 767 when it comes to slowing down to clean speed you can use the speed restriction option on the descent page to slow it down at a particular altitude, or a hard speed on the legs page to slow it down at a particular point. Both options assume idle thrust and rely on the correct winds so you may find yourself needing to add power or speedbrake.

If you're really lazy and want to make VNAV do it all for you, build a geometric path: Pick a prior point and look at the predicted alt - if it says 14100 then type /13000. That will force a geometric path between that and the next point that is slightly shallower requiring more thrust than idle. The FMAs will change to SPD / PATH and the autothrottle will adjust the speed without the need for speedbrake. Just don't confuse it for approach logic!

I'm almost certain that the deviation indicator actually shows feet and only feet, in relation to the different segments your flying, whether they're flat or almost flat deceleration segments or regular ones.

What we accomplished maybe was to befriend the engines and not to go non stop from CRZ power to idle.

I think the blank stare is "Is this guy for real?"

Don't want to shock cool those cylinders.

Yeah, specially those big cylinders mounted under the T7 wing 🤣