During descent, the autothrottle is not capable of maintaining speed (usually 250 knots). Why? Picture is enclosed.

 

During descent, the autothrottle is not capable of maintaining speed (usually 250 knots). Why? Picture is enclosed.

 

Always pay attention to your FMA. It says "ARM" which means "I'm here if anything goes terribly wrong, but I'm deferring to you for now."

 

In order to be sure you follow the descent path, ensure that you've used the descent forecast function, otherwise, you will be above/below path/speed. In your case, you're on path, but under speed, bump the throttles up a little to get yourself back on speed. Next time, if you run your DESC FORECAST, it will be better about your path, and therefore, speed.

Always pay attention to your FMA. It says "ARM" which means "I'm here if anything goes terribly wrong, but I'm deferring to you for now."

 

In order to be sure you follow the descent path, ensure that you've used the descent forecast function, otherwise, you will be above/below path/speed. In your case, you're on path, but under speed, bump the throttles up a little to get yourself back on speed. Next time, if you run your DESC FORECAST, it will be better about your path, and therefore, speed.

 

Hi Kyle, thanks for responding so quickly.

 

Yes, in fact, I always use the descent forecast option to make my flights as realistic as possible. Maybe I'm doing something wrong, of course. 

 

When you say "bump the throttles up a bit", does that mean disengage auto throttle and manually increase the speed, then arm it once it reaches lets say 250 knots? My flights has no issues following the calculated descent paths, especially after I excessively started to use the descent forecast option.  

 

I noticed that in this flight, I had a bit of a headwind. Could that influence speed in such a way maybe?

Yes, in fact, I always use the descent forecast option to make my flights as realistic as possible. Maybe I'm doing something wrong, of course. 

 

If you're getting off path (vertical or longitudinal - see below), then:

• You're using old data
• You're doing it improperly; or
• The data just isn't matching up with the sim (this also happens in real life - the wind numbers are not actual observations in the real world either - they're approximations)

 

When you say "bump the throttles up a bit", does that mean disengage auto throttle and manually increase the speed, then arm it once it reaches lets say 250 knots? My flights has no issues following the calculated descent paths, especially after I excessively started to use the descent forecast option.  

 

As I mentioned earlier:

It says "ARM" which means "I'm here if anything goes terribly wrong, but I'm deferring to you for now."

 

In both HOLD and ARM modes, you can set the throttle manually while the AT is still on (ensure you have the appropriate, realistic setting of Override AT: IN HOLD/ARM MODE ONLY set in the PMDG SETUP > SIMULATION pages).

 

If you're getting below speed, then you're having issues with descent paths. Granted, it's not getting above/below path, but it's getting below speed in that screenshot. See below for why this is an issue.

 

 

I noticed that in this flight, I had a bit of a headwind. Could that influence speed in such a way maybe?

 

This is exactly the issue, and this is why the DESC FORECAST page exists. If head/tailwind didn't matter, then there'd be no point in the page, though there is a slight change due to pressure as well.

 

Your decent is a 3-dimensional path, which a lot of people forget about. The point of a descent is to move about the vertical axis, sure; while on a lateral path, of course; but what most people forget about is the longitudinal path. If you're descending at 280 knots today, that means you're going to use the same distance to descend at 280 knots tomorrow, right? Wrong, unless the weather is exactly the same.

 

Ignoring the change in TAS due to atmospheric conditions (so, assuming IAS=TAS, just for simplicity of the example), consider the following scenario:

You're cruising at 250 knots, at FL320. Your destination airport is at sea level (or 0' MSL). Your preferred descent rate is 1200'/min (VNAV doesn't force a set rate in an IDLE path descent, it simply pitches for speed, but bear with me here - it makes the math simpler).

 

Using a bit of maths, you get:

Total time in descent - 26:40 (ALT / RATE = 32000 / 1200 = 26 2/3 = 26:40)

NM/MIN: 4.17 (SPEED(H) / MIN IN HOUR = 250 / 60 = 4.1666R > 4.17)

Distance Covered: 111nm (TIME * SPEED(M) = 26.6666R * 4.1666R = 111.1111R > 111)

Vertical FT/NM: 288 (ALT / DIST = 32000 / 111.1111R = 288)

 

So, assuming zero wind, you are going to descend 288' for each nautical mile.

 

Now, consider the following:

You're cruising at 250 knots, at FL320. Your destination airport is at sea level (or 0' MSL). Your preferred descent rate is 1200'/min. You have a headwind of 30 knots.

 

Total time in descent - 26:40 (ALT / RATE = 32000 / 1200 = 26 2/3 = 26:40)

NM/MIN: 3.67 (SPEED(H) / MIN IN HOUR = 220 / 60 = 3.6666R > 3.67)

Distance Covered: 98nm (TIME * SPEED(M) = 26.6666R * 3.6666R = 97.7777R > 98)

Vertical FT/NM: 327 (ALT / DIST = 32000 / 97.7777R = 327.2727R > 327)

 

So, assuming zero wind, you are going to descend 327' for each nautical mile.

 

Remember that airspeed is your speed through a parcel of air. When that parcel of air is moving backward at 30 knots, your effective speed across the ground is 220. This change means that given the same distance traveled through the air, your track across the ground is a lot less, which is why you get more feet of descent for each nautical mile over the ground. A headwind will decrease the amount of distance covered during a descent (effectively moving your T/D later), while a tailwind will increase the amount of distance covered (effectively moving your T/D earlier).

 

Now, to throw a wrench into all of that, one thing many people forget is pressure.

 

Consider the following:

You're cruising at 250 knots, at FL320. Your destination airport is in America at sea level (or 0' MSL). Your preferred descent rate is 1200'/min. You have a headwind of 30 knots. The pressure at your destination is 30.92. Remember that in the FLs, you use 29.92, and local pressure (30.92) is used below FL180.

 

Altitude Error: 1000' (30.92" - 29.92" = 1.00" > 100' for every 0.10" > 1000')

Initial Descent: 14000' (32000 - 18000)

Second Segment: 19000' (19000 - 0)

 

To illustrate this mess in another way, while still at cruise (at standard pressure), change the altimeter setting to local pressure to see your actual altitude MSL. By rolling the setting up to 30.92, you increase the indicated altitude by 1000'. So, despite cruising at "FL320" using the standardized 29.92, you're actually at about 33,000' MSL. This is the precise reason we call them FLs and do not refer to them as specific altitudes. FLs are standardized pressure profiles more than actually resembling altitudes.

 

Using the math from earlier, you get 289 feet per nautical mile for the entire descent, which isn't significant over the 288 of the first scenario, but keep in mind that some STARS have early altitude restrictions. If you're not considering the pressure change, and you have a restriction to be at a fix at 17000', then you're probably going to be about 1000' off in the above scenario if you did not use the forecast function, as the FMC will assume a constant pressure profile if you don't provide an actual pressure. As soon as you hit STD to drop out of standard into local pressure, your altitude jumps 1000', as does your vertical profile deviation.

 

If all of this is confusing, I made a video a while ago with diagrams explaining the concepts, here:

 

If you had both a headwind, and pressure was particularly low below the transition level, then you'll be below path. That may require you to nudge the throttle levers (even with AT on) to keep your descent on path - vertically and longitudinally.

Wow nice informative thread / post. I'll have to remember to come back and digest it all when I have time.

Hi Nick,

 

I can reassure you, in reality 90% of descents using VNAV PTH will not descend the aircraft at your planned speed no matter what you put in the FCST page.

 

In VNAV Path the aircraft has absolutely zero regard for speed and will do anything just to remain on your planned vertical path using pitch control and idle thrust.

 

In a nut shell If your speed gets low increase thrust and if your speed gets high use speedbrake, it's honestly as simple as that!

Hi Nick,

 

I can reassure you, in reality 90% of descents using VNAV PTH will not descend the aircraft at your planned speed no matter what you put in the FCST page.

 

In VNAV Path the aircraft has absolutely zero regard for speed and will do anything just to remain on your planned vertical path using pitch control and idle thrust.

 

In a nut shell If your speed gets low increase thrust and if your speed gets high use speedbrake, it's honestly as simple as that!

 

Huhm, noted. But I thought entering data into the forecast page helped the computer decide when to start slowing down/decelerating etc. ?

Huhm, noted. But I thought entering data into the forecast page helped the computer decide when to start slowing down/decelerating etc. ?

 

Yes completely! You should always populate the FCST page with the most up to date descent winds and destination QNH so you can get the most accurate DES Path from TOD to E/D (End of Descent). I've never used the TAI function.

 

I just wanted to reassure you that you were not doing anything wrong, in real life you can be nicely descending on your calculated path on speed and then ATC will throw a spanner in the works and tell you to increase speed or give you a significant shortcut etc so your calculated DES Path is thrown out the window!

 

A good tip is to frequently update your calculated DES Path during descent. On your snapshot you're inbound to GA28A. If you line selected LSK1L Twice (GA28A) and executed during your descent you'd get a new DES Path which would use the actual wind at your present altitude within the equation.

 

A good tip is to frequently update your calculated DES Path during descent. On your snapshot you're inbound to GA28A. If you line selected LSK1L Twice (GA28A) and executed during your descent you'd get a new DES Path which would use the actual wind at your present altitude within the equation.

 

 

 

That's a pretty neat trick. Always learning! Thx Sam. 

Another thing to note is that in the real world 737 in a VNAV descent the plane is planning a path at idle but if your speed does start to decrease as described here the autothrottles will come out of arm and increase thrust if you slow more than 15 knots. They will remain active until your back on speed then they will go back to arm.

I guess the question would be what do pilots typically do in real life... kick up the throttles manually or wait till the auto throttle corrects?

I guess the question would be what do pilots typically do in real life... kick up the throttles manually or wait till the auto throttle corrects?

Everyone does it differently, most people increase the thrust until the speed increases to the planned descent speed then reduce the thrust to idle. A few Captains I fly with, if using VNAV PTH, will just let the speed bleed off and let the A/T recapture the speed as Tom suggested. Before doing this we generally update the DES path as I previously suggested so a new DES Path will be generated.

I won't usually wait to correct the speed since we're supposed to be within 10 knots of assigned. I'll also let the plane get fast if there is no speed restriction and simply let increased parasite drag eat some energy. (That's assuming there's no restrictions downstream that would be affected.)

A good tip is to frequently update your calculated DES Path during descent. On your snapshot you're inbound to GA28A. If you line selected LSK1L Twice (GA28A) and executed during your descent you'd get a new DES Path which would use the actual wind at your present altitude within the equation.

Do you know if this technique works on the 777F?

I guess the question would be what do pilots typically do in real life... kick up the throttles manually or wait till the auto throttle corrects?

 

I'm not a real pilot, but my thoughts from a sim point of view...

 

I tend to try and look ahead and see what's coming up, because particularly during the descent, VNAV is pretty dumb. I might not always want that extra energy (assuming no speed restrictions etc). For example, we might be slightly low and/or slow at this point in the profile, but the next segment is particularly steep. No point stuffing the thrust levers forward now only to yank the speedbrake later -- it's just a waste of fuel. So in this instance I might speed intervene at the current speed for a few moments, or drop in to FLCH to keep the thrust levers at idle (and then later on I've got room to wind the speed up and get a better rate as Matt suggests).

 

On the other hand, if we really were low all the way, my inclination would generally to be to go for either V/S for a bit or just ease the thrust levers forward as suggested -- not sure about the 737, but the 747 is quite aggressive with its (automatic) application of thrust in a VNAV PTH descent and there are smoother ways of achieving the same result! Of course, this may not necessarily be the case in real life.

 

On the whole I'm quite reluctant to just "wait for the aeroplane to sort it out" -- if it's not doing what I want, especially if that involves getting slow, I'm inclined to take some form of positive action sooner rather than later. That's just me though.