VNAV

[stratone]

@tup61 @Stearmandriver Let's reduce this to VNAV and it's use in RL by pilots.

This thread is not about what can go wrong at all. After about 40 years in aviation I do have a rough idea about ...

Best regards, Detlef

 

[jcomm]

18 minutes ago, stratone said:

@tup61 @Stearmandriver Let's reduce this to VNAV and it's use in RL by pilots.

This thread is not about what can go wrong at all. After about 40 years in aviation I do have a rough idea about ...

Best regards, Detlef

 

Quite on the contrary, IMO @tup61 point, just as the precious observations by @jon b are spot on...

 

[gb09f]

Not R/L, but I hardly ever use it in managed mode (Airbus). 

It's all OP DES and V/S mode for me.

It's a handy reference to check if your profile calculations are correct. 

[Sky_Pilot071]

7 hours ago, Stearmandriver said:

😬 THAT'S not going to end well lol.

In reality, we have an RNP approach to 26 that snakes through the Gastineau channel.  It gets impressively close to terrain.... And it's still way safer than a straight in would be!

I just wanted to see how the G3000 would handle it.  After I cleared the high stuff the airport came into view so it was a perfect setup and landing coming from Calgary.

Also had some interesting buffeting due to an xwind.

sp

[Bobsk8]

I use V Nav in the PMDG 737 all the time, on departure to cruise altitude, and on descent  to approach. It works great all the time, and when I am a bit high, the FMC always displays drag required. 

[Stearmandriver]

14 hours ago, stratone said:

Let's reduce this to VNAV and it's use in RL by pilots.

Isn't that what we're talking about? I said earlier that none of the workload management value to VNAV matters in the sim.  In real life however... If I ever flew with an FO who somehow instinctively distrusted VNAV and wouldn't use it, well... That would be a discussion point.  ;)

 

[g-liner]

On 3/5/2023 at 6:01 PM, jon b said:

Well, the 787 VNAV works better when it has quite a few descent constraints on the STAR then it can build itself a refined descent profile and it seems to behave better.

Normally though it presents you with its precise calculations for top of descent , where it’s going to slow down, even suggesting where the flaps need deploying, it even presents all this in a vertical plan view on the VSD.

It then sits there all pleased and smug with its work , however within 10 seconds of starting its descent it panics and shouts “ speedbrake, I can’t do it I need speedbrake’ not literally of course but an FMC message comes up with words to that effect.

It’s as though the VNAV program was written for another aircraft and doesn’t account for the 787’s slippery wing.Most people I fly with on the 787 seem to throw VNAV away by 20,000ft and switch to FLCH SPD because it’s making a mess of things, on the jumbo we used to use it all the way to glide slope capture on some occasions.

The 787 VNAV is also different in that it remains  on ,and prioritises path. For example in the 747 should you wish to expedite your descent for whatever reason while in VNAV PATH you can open the speed window and wind up the speed. The FMAs will change to HOLD (thrust) and VNAV SPD and the aircraft will pitch for speed vastly increasing its descent rate. At the bottom of this high speed descent the aircraft will level off and if you time it right , normally around 500ft above, you can wind the selected speed right back again and the thrust stays at idle and the aircraft will use its massive inertia to coast for 10 to 15 miles while it bleeds off speed. I used to get cleared high speed and level off at 5000ft downwind at 330kts and coast all the way downwind and decelerate to 180kts  flaps 5 for the base turn, all at idle or there abouts, very fuel efficient. 

Try opening the speed window and increasing it  in a VNAV PATH to increase descent on the supposedly super fuel  efficient 787 and  the thing stubbornly remains in PATH and  will add power to achieve the speed.  Not very efficient, I’m sure there’s a logic to this but I don’t know what it is, and no one I’ve spoken to really does either, it may just be the 787 is different , rather than “worse” than the 747, but the amount of people who just throw VNAV away and switch to FLCH SPD which a lower level of automation tells me I’m not the only one who’s a bit dubious about it.

That’s odd, I usually use VNAV from TOD until glideslope (or equivalent) capture and think it works really well. Are you uploading descent winds and temperatures etc? That’s usually when I see it not being able to cope well I.E. if there’s not enough data for it to work on or if the horizontal route/distance isn’t being kept up to date with intercept legs etc 

[G550flyer]

On 3/5/2023 at 12:01 PM, jon b said:

Well, the 787 VNAV works better when it has quite a few descent constraints on the STAR then it can build itself a refined descent profile and it seems to behave better.

Normally though it presents you with its precise calculations for top of descent , where it’s going to slow down, even suggesting where the flaps need deploying, it even presents all this in a vertical plan view on the VSD.

It then sits there all pleased and smug with its work , however within 10 seconds of starting its descent it panics and shouts “ speedbrake, I can’t do it I need speedbrake’ not literally of course but an FMC message comes up with words to that effect.

It’s as though the VNAV program was written for another aircraft and doesn’t account for the 787’s slippery wing.Most people I fly with on the 787 seem to throw VNAV away by 20,000ft and switch to FLCH SPD because it’s making a mess of things, on the jumbo we used to use it all the way to glide slope capture on some occasions.

The 787 VNAV is also different in that it remains  on ,and prioritises path. For example in the 747 should you wish to expedite your descent for whatever reason while in VNAV PATH you can open the speed window and wind up the speed. The FMAs will change to HOLD (thrust) and VNAV SPD and the aircraft will pitch for speed vastly increasing its descent rate. At the bottom of this high speed descent the aircraft will level off and if you time it right , normally around 500ft above, you can wind the selected speed right back again and the thrust stays at idle and the aircraft will use its massive inertia to coast for 10 to 15 miles while it bleeds off speed. I used to get cleared high speed and level off at 5000ft downwind at 330kts and coast all the way downwind and decelerate to 180kts  flaps 5 for the base turn, all at idle or there abouts, very fuel efficient. 

Try opening the speed window and increasing it  in a VNAV PATH to increase descent on the supposedly super fuel  efficient 787 and  the thing stubbornly remains in PATH and  will add power to achieve the speed.  Not very efficient, I’m sure there’s a logic to this but I don’t know what it is, and no one I’ve spoken to really does either, it may just be the 787 is different , rather than “worse” than the 747, but the amount of people who just throw VNAV away and switch to FLCH SPD which a lower level of automation tells me I’m not the only one who’s a bit dubious about it.

Here's what I have found with slippery aircraft like the 757, G5/G550. It seems as long as the STAR is not aggressive, VNAV works out great. When the descent gets aggressive, has speed changes and intermediate level offs in close proximity, it starts getting behind quickly. I have noticed two things that exacerbates the issue. One, a lot of FMS systems target slowdowns with a 500 feet per minute descent rate. Two, you get this cone of silence like period when the system re-calculates the descent. When slowing with that 500 feet per minute rate, the target speed it's trying to capture becomes hit or miss. You notice this a lot at the 10,000 slow down point. It starts, then realize that it's not going to make it, then chooses altitude over speed, drops the nose and speeds up at times. Unlike me that has 1200 feet per minute at 12,000 feet burned into my brain, it starts that 500 feet per minute slow down at or below 1000 above. At that point, its much too late to try to recover. So, you have to either use vert speed and use the 12 at 12 or 13 at 13 technique, or simply gamble and watch to see if it pulls it off. 12 at 12 seems to work great in every jet I've flown to include light and heavy jets. Now, for the re-calculation.

As you approach these intermediate level offs, the FMS will calculate a new TOD for each one. When it happens, it goes into this some what frozen state for a few seconds like a shark does when you grab it's nose. I've never tried that by the way. On the VSD, you will see nothing during this time and suddenly a indicator point for speed and TOD will appear on the path. When the intermediate level offs are close to each other, the calculations happen more often. It loses those few seconds of action and then realize it has missed the TOD/VPATH and adjusts to acquire it. As the calculations happen, it starts a snow ball effect. At some point, it gives up speed and targets altitude and dumps the nose to get down. Now, the G550 is a little bit smarter and it deals with this situation slightly better by going into to VFLCH. As it realizes it's struggling with the VPATH, throttles will move idle and VFLCH illuminates as it automatically enters FLCH mode. Now, combine this with the slowdown issue and it gets crazy. At some point, speed gets the middle finger as it chooses to guarantee the altitude restriction.

For aggressive STARs, I review the STAR prior to the descent/approach brief. When I notice that it's aggressive, I may slow down and put in an early descent request. I also use the vertical direct to altitude to get the jet to descend now when I need it to. I can also go in and adjust the VPATH angle if I think it may be helpful. I will also throw out the boards early and often. When I see these aggressive STARs, out comes the boards early on and I will ride them all the way in until configuration time. In fact, when you get the boards out early enough, you don't have to use as much vs realizing you are not going to make it and throwing them all the way out. 

I see these type of arrivals a lot going into the DC area.   

[jon b]

@Gstove yes all winds and temps are being updated regularly and also around 30 mins before TOD a new set comes through. I'm glad you're finding it ok to use, so it makes me think something isn't quite somewhere as what I'm seeing isn't   what I'd expect from a latest generation jet, i.e. it's a retrograde step from the 747.

I have a suspicion where the fault lies but I'd better not discuss it on an open forum 😉

@G550flyer Thanks ever so much for those tips Rick, I've never actually heard of the 12 at 12 before , I really like that one, a lot ! Every days a school day.

Yes this 787 is certainly a slippery character, on the 74 the speedbrake only ever came out on special occasions, normally the pilot or ATC had left something too late and it's use usually generated a smirk from the other pilots on the flightdeck. It was mentioned from day one on the 787 conversion that on this aircraft the speedbrake is needed and is not a sign of weakness anymore !

I'm not necessarily saying the VNAV behaviour I'm seeing on the 78 is wrong, it just different from the 74 which I'm very comfortable with. I found the 74's VNAV very flexible and you could open the speed window come out of path and into speed and back into path again as required to manage your energy and path, whereas as I mentioned earlier I find the 787 adding thrust in a descent very counter intuitive for an aircraft that is meant to be super efficient. I had a play around a few months ago and flew it like a 74 going high speed below profile and coasting on level off and it saved on flight planned fuel.

I suppose the danger on this thing is accelerating anywhere near the red bricks , it's so slippery it'll just keep going into an overspeed.

It's not just me that isn't liking VNAV , I recently did a line check and I played the game like you do, and kept it in VNAV as long as possible on the descent, it being the highest mode of automation and all that. On the debrief the checker said, personally I'd have thrown VNAV away much earlier ! ( you cant win)

[Sky_Pilot071]

I really don't care what VNAV does in RL.  Since this a MSFS thread all I can say is that the HJET VNAV does a really good job.

sp

[G550flyer]

4 hours ago, jon b said:

@G550flyer Thanks ever so much for those tips Rick, I've never actually heard of the 12 at 12 before , I really like that one, a lot ! Every days a school day.

I picked it up in the DC-1030 lol. If you set 1200vs at 12000, it would hit 250 right at 10000. If you were on the heavy side, you could set 1300 at 13000 and it would shack 250 at 10. Just so happen, it works well in the Gulfstreams too.

[g-liner]

14 hours ago, Sky_Pilot071 said:

I really don't care what VNAV does in RL.  Since this a MSFS thread all I can say is that the HJET VNAV does a really good job.

sp

The thread is asking if vnav is used in real life, thats why we’re discussing real life vnav. 

[g-liner]

10 hours ago, G550flyer said:

I picked it up in the DC-1030 lol. If you set 1200vs at 12000, it would hit 250 right at 10000. If you were on the heavy side, you could set 1300 at 13000 and it would shack 250 at 10. Just so happen, it works well in the Gulfstreams too.

We were always discouraged from using vs higher up with my lot on the B75/76 due to the lack of speed protection. I think someone had got into a low speed event using it in the old days so it was always frowned upon. We’d only really use it for doing a CDA or intercepting the glideslope or on an old school approach with vs for the vertical profile. 
 

Oddly now the U.K. CAA require an approach flown using VS or GPA  called a 2d approach in the sim for us where we have to control the vertical profile ourselves. Seems an unlikely scenario but I suppose if the approach was missing from the FMC database we might have to use VS on approach  irl? 
 

Now we just us IAN for vor or loc only approaches so it looks like an ils or VNAV if it’s an RNP (ar) type approach. 

[g-liner]

On 3/7/2023 at 10:32 PM, Stearmandriver said:

Isn't that what we're talking about? I said earlier that none of the workload management value to VNAV matters in the sim.  In real life however... If I ever flew with an FO who somehow instinctively distrusted VNAV and wouldn't use it, well... That would be a discussion point.  😉

 

Plus it’s going to make all the RNP ar approaches, where I expect you (like me on the 78) presumably use lnav/vnav kind of tricky? 
 

if you don’t trust vnav higher up using it down a valley in the alps (other mountains are available) is going to feel quite uncomfortable. 

[G550flyer]

1 hour ago, Gstove said:

We were always discouraged from using vs higher up with my lot on the B75/76 due to the lack of speed protection. I think someone had got into a low speed event using it in the old days so it was always frowned upon. We’d only really use it for doing a CDA or intercepting the glideslope or on an old school approach with vs for the vertical profile. 

That's unfortunate because V/S is an essential part of the auto pilot and is very useful in various situations. It's also a workload reliever when in busy airspace when you need to have eyes and ears outside instead of working the box with edits and vertical direct to altitudes. The big thing about V/S is it's not a set and forget item. In fact, no item is. In regard to V/S, you just need to know that you can command a V/S that will get you slow or fast. I believe it all comes back to the basic instrument scan. One of my pet peeves is that people will get above 10000, stop aviating and go heads down with paper works and EFBs. I remember this one time I was giving an examination as a check airman. We get past the mach change over and I noticed that the climb rate is dropping off. I sat and let this go on to see if they would catch it. They were both heads down, yacking and working the oceanic charts. I knew this would be a problem early on because we departed late and they perf'd the box at 320/.84 to make up some time. We were heavy and I knew it would kill the climb rate after mach change over. Once the rate dropped below 700 feet per minute, I asked, "who is flying the jet?". The left seater looked up responding "I am". They immediately looked puzzled at the climb rate and I could see the hamster wheels turning. I then dropped the QRH on the center console and asked if we could climb to target altitude at this speed and weight. I guess with my nudge, they were able to find out that .84 was too much and perf'd it back to .80. I don't think V/S is the issue to be frowned on, it's the lack of basic piloting skills. Here are some situations where I find V/S helpful.

 

The first one is when departing at light weight and capturing an altitude shortly after departure. The DC10 was a beast at light weights with the nose at 25 degrees pitch and still accelerating past V2+10. I have clocked the G550 at 6500 feet per minute after lift off at V2+20. Needless to say, that low altitude comes very quickly and the jet will pitch aggressively to capture the altitude. To slow things down and prevent hi yucka pitching, you just enable V/S and use 2000 up. Of course this all depends on your min climb gradients if you have one.

 

Second, passenger comfort using FLCH. When you are light or have powerful engines, the engines spool and then the nose pops up sharply and then adjusts back. I like to set V/S first and dial it up until the engines get near climb power and then engage FLCH. Makes a nice transition and without the boss in back spilling Cognac or wondering why we are pulling Gs. 

 

Third, temperature inversions. Usually you will see these coming based on the cloud layer. As you enter, the ASI will bump in speed and the jet will aggressively pitch based on how bad the temp change is. I immediately smash V/S, especially when I can anticipate it. This locks in the current climb rate and then I re-enter VNAV/FLCH once things settle down. Passengers won't even notice when you are good at it.

 

Fourth, FLCH after mach change over. The jets I've flown tend to get pitchy when climbing in mach mode FLCH. I find that in the G550 its a little better when the jet is really heavy. It's similar to temperature inversions where the mach will bump and the jet will pitch up or down to adjust to speed. It starts off light and then goes into this sea saw action. As it sea saws, the speed deviations get larger and larger. If you don't intervene, you end up on the back side of the curve and it takes for ever to get back on speed without stopping the climb. As soon as I see this start, I smash the V/S button, let it settle down and then re-engage FLCH.

 

Fifth, when slowing. To me, 12 at 12 and other slowing techniques works out better than the FMS calculations. The FMS plans that perfect slow down with 500 feet per minute. In average, I find it has you above 250 for a little bit. The G550 reports your indicated airspeed to ATC and in certain areas, they will call you out. So you are stuck there minimizing your descent as best you can to slow. If you are not careful when slowing, you may not realize initially that dialing back the speed in FLCH may cause a pitch up making things worse when you are trying to slow and descend at the same time. Manually controlling the slow down with V/S just works out better.

 

Last, is the scenario you mentioned. If you are doing a non FMS based approach, you could be in this scenario. Though it's not often, I have seen certain approaches removed from the database. For example, Honeywell will remove approaches from the DB if there's an error or if the FMS directs a right turn when it should have been a left, etc. I have also been to places where the airport did not exist in the database. Maybe you are old school and just want to practice and fly that LOC only approach in V/S. You calculate your timing and descent rate on approach anyway, now you just have to dial in your estimate and adjust as needed.