Slowing Down the 787

  • I don’t know if this is a bug, or if the settings or physics are off, or if I’m doing something wrong. I’m guessing it’s me. I want to love the Dreamliner and I am SO grateful that we have this beautiful plane. But I just did three separate go arounds because I could not slow the plane down. Flaps won’t deploy and the speed brake won’t deploy until I’m under 200 knots (pulling all the way down on the throttle). So as I approach the airport on a standard descent I cannot get my speed below 250 and nothing will deploy so my speed stays the same or even starts slowly increasing as I descend. I couldn’t even get my gear to deploy. Any advice from you pros out there? Thank you!

  • Jet-Pack (IPACS) I set up the same route and started my descent and turned on autopilot. Speed brakes immediately deployed. The plane did not attempt to descend with the navigation marks while it slowed. Instead it followed the flight path and entered a moderate descent of 600 fpm which allowed the airspeed to bleed down to 250. Autopilot set the airspeed to 200. The plane followed the flight path to LAX but was about 7000 feet above the nav markers. At 6 nm from the runway and 9000 feet up it finally deployed 10 degrees of flaps at 203 knots. It then followed the flight path seen below, not adding any flaps the entire way until right base when it added full flaps. I will try the approach again and see if I can get speed brakes to stay deployed once I start the descent on autopilot and switch it off to stick with the nav markers.

    *Update… next approach speedbreaks worked without ever going to autopilot. So all I can think of was there was an issue coming out of autopilot that wouldn’t allow me to deploy them. I’ll let you know if it repeats. That bird still doesn’t like to slow down and come out of the air! I love it! It’s a beautiful plane.

    *Yes I have those assists off.

  • In pprune a lot of 787 pilots run the descent with pilot intervention, a lot do not like the PTH mode VNAV descent, often the real plane doesn’t work things out sensibly.

    I tried to set my speed to M0.84/300 kt for a descent but it overspeeds if the throttles idle, selected descent rate is over 1,700 fpm, I have to add 45% to my normal 737 (3 miles per thousand feet) Top Of Descent distance out, I was expecting about 8%.

    I monitor the speed but actually have the autothrottle set significantly lower to avoid unwanted power inputs in the descent.

    I read that for a 787-8 the typical rates would be M0.84 at 2,600 fpm down the FL240 and 300 kt at 2,800 fpm down to 10,000’ and 240 kt and 1,500 fpm down to the approach.

    Is the power off descent drag a bit low? Is the idle thrust a bit high?

    I tried a 144 kt APP autoland on flaps 25 in the 787, which in the sim’s CDU thrust limit page has a flaps 25 VREF of 142. At 22 feet on the radar altimeter it cut the throttle and did a longish flare slowing 12 kt to 132 kt and gave a tail strike.

    The indicated surface wind was 2 kt at 40 degrees from the right with no gusts.

    It was dropping a right wing slightly just as it touched.

    It looks like that VREF is good, it seems to stall with flaps 25, gear down and power off at about 109 kt (142/1.3).

    Version 20.22.11.01

  • The 787 glide ratio is now much higher than previously. It's now somewhere in the mid twenties which I think is still pretty realistic. The vertical speed values vs. indicated airspeed are not constant which makes it difficult to compare. We need TAS and true rate of descent plus true airspeed trend. It would be easier to know the distance flown over height lost or the actual level altitude deceleration rate (x knots per NM).

    For the A320 I was able to reverse engineer the glide ratio used in the default descent planning tables. Was about 1:19.

    The 787 should be much more slippery though, the glide ratio has to be quite high for efficient cruise.

  • A ‘suck it and see’ method (trial and error) will give a working solution. For a descent from say FL330 down to 3,000 feet that descent profile should give the idle descent rates without using the airbrake.

    (‘for a 787-8 the typical rates would be M0.84 [changing to 300kt] at 2,600 fpm down the FL240 and 300 kt at 2,800 fpm down to 10,000’ and 240 kt and 1,500 fpm down to the approach’).

    Using a basic 3 miles per thousand feet ratio to give a TOD distance it should place the plane at the 11,000 feet point to allow a 500 feet per minute deceleration to 240 kt before reaching 10,000 feet. The current 787 seems to need under ten miles for this phase.

    At 240 kt the descent can continue to 3,000 feet at 240 knots to the approach location. The 787 should then be at about ten miles from the final location at 240 kt and at 3,000 feet for that approach.

    Suck it and see alteration to the idle thrust descent drag (not exactly the same as gliding) can give a satisfactory result without having to generate a horribly detailed continuously varying descent angle exact flight-plan, with increasing indicated airspeed in the (brief) initial Mach 0.84 hold descent phase followed by reducing true airspeed in the 300 kt hold indicated airspeed phase and with changing flight idle residual thrust (though a flight planning app could probably do it instantly).

    The 3 miles per thousand feet value can be adjusted for anticipated winds.

    This is a zero wind 99 mile descent from FL330 towards the HAM VOR intended to be ten miles out on reaching 3,000 feet. It shows how much speed brake is currently required. The right seat view lets the captain’s hand act as a position reference for the speed brake lever. The throttles were on manual and fully closed.

    300 kt indicated.

    240 kt indicated.

    Edited once, last by Overloaded (January 25, 2022 at 6:51 AM).

  • Jet-Pack (IPACS) I set up the same route and started my descent and turned on autopilot. Speed brakes immediately deployed. The plane did not attempt to descend with the navigation marks while it slowed. Instead it followed the flight path and entered a moderate descent of 600 fpm which allowed the airspeed to bleed down to 250. Autopilot set the airspeed to 200.

    At 6 nm from the runway and 9000 feet up it finally deployed 10 degrees of flaps at 203 knots. It then followed the flight path seen below, not adding any flaps the entire way until right base when it added full flaps.

    Boeings on VNAV autopilot will reduce the speed for the pilots to apply the corresponding amount of flap. They are more hands-on than an Airbus.

  • Overloaded you assume a 3 miles per 1000ft glide ratio that is about 3 degrees or a glide ratio of 1:18 if I'm not mistaken. I think our current 787 is closer to 1:22 then it's 3.6 NM per 1000ft.

    If we use your airspeed of 300 KIAS assuming 420 KTAS and 2800 ft/min, that's 220m/s and 14.2m/s that's only 1:15.

    And if we use 240 KIAS assuming 290 KTAS and 1500ft/min that yields 1:20.

    So either the engines produce a significant airbrake at idle or the 787 is not nearly as slippery as I thought it was. Before jumping to conclusions I'll check if I can find some fuel planning manuals for the 787, they usually have a descent planning section.

  • Hi Jan, my point is how much, how very much airbrake is needed to hold the guidance ias to the corresponding paired fpm, it is very slippery. I must check the actual achieved descent rate and the mean ground speed to see how the measured numbers add up, they look good.

    Some decent info on-line can’t do any harm.

    No exam standard flight planning, life is too short.

    I should have really used about a 3.2 miles per thousand feet ratio for the 787’s 2,800 fpm descent, if I had no air brakes the 1,700 fpm descent would need about a 5.3 miles per thousand feet ratio. It only varies how far out the TOD is.

    The simplified (and taught) descent method gets the end result, it does not try to maintain any ideal instantaneous performance value (like 3.6 nm per thousand feet passing a particular altitude) on the way down. The start and the end points have to work.

    It gets steeper at lower levels as the TAS falls. At FL280 descending at 2,600 fpm it is 3.23 degrees, at FL220 descending at 2,800 fpm it is 3.81 degrees.

    Edited 3 times, last by Overloaded (January 25, 2022 at 10:02 AM).

  • Ok I found a chart in the manuals, the descent rate is affected by pressure altitude, airspeed or mach flown and of course weight as expected. That's why descent rate is not a good measure for idle glide performance. The higher up you are the greater the true airspeed and thus the greater the descent rate for the exact same glide ratio. Also even though the maximum glide ratio is almost the same you glide better at speeds higher than the speed for max L/D if the aircraft is heavier. That weight difference is significant, at FL200 at 290 KIAS you may have 2100ft/min if the aircraft is empty but only 1600 if the aircraft is heavy.

    From a 787-8 FCOM:

    For our current weight of 488000 lbs at 290 KIAS - roughly 334 KTAS, passing through 10000ft on the descent the descent rate should be 1500 ft/min with a pitch angle of around zero degrees. That's a glide ratio of 1:22.5 or 3.7 NM per 1000ft. I've increased the parasitic drag and our 787-10 now it does exactly that.

    For reference: At 20k at 290 KIAS (386 KTAS) the descent rate is stated as 1600 to 1700ft/min. That's 1:23 idle glide ratio or 3.78 NM per 1000ft.


    And while we're at it:

    777 FCOM, for 550000 lbs:

    at 270 KIAS (311 KTAS), 10k: -1400ft/min -> 1:22.5 idle glide ratio or around 3.7 NM per 1000ft

    at 270 KIAS (359 KTAS), 20k: -1600ft/min -> 1:22.7 idle glide ratio or around 3.7 NM per 1000ft as well

  • Version 20.22.15.01 787 no airbrake descent for Top Of Descent placement.

    I wanted to see just what the maximum descent without airbrakes looked like so I used FLCH to hold M0.84, 300 kt IAS and 250 kt IAS, the kink between 300 and 250 was from slowing down with minimal height loss just before reaching 10,000 feet.

    The Aero 787 remains very slippery, instead of a basic 90 miles to lose 30,000 feet it needs 130.

    Boeing 787 detailed design mission

    This is a 787-8 flight profile from the net, I cannot see what the background to the page is but the numbers make sense. It looks like the descent is made with Mach hold until the desired IAS creeps up then it is an IAS hold with a speed reduction at 10,000. The thrust is idle and the vertical descent rate is variable. It uses 134.62 nm to lose 39,850 feet.

    The current Aero 20.22.15.01 787-10 model needs a lot of distance to descend, it needs 176 nm to lose 40,000 feet. The current autopilot LNAV and VNAV overshoots so an early manual intervention using that chart’s TOD distance out to select FLCH or trying A LOT of airbrake with the autopilot LNAV/VNAV is a possible alternative. I’d like manual FLCH with enough airbrake to shorten the descent to something like 3.4 miles per thousand feet overall.

  • That's a great link, thanks!

    Maneuver segment 9 seems to be an idle descent at about 2.5 degrees with 277 KCAS and 138 metric tons. So that's only 60% of the weight that our 787-10 currently has. And they are flying at a CL of 0.34 which is quite low, hence 'only' the 18.84 L/D ratio, even though they calculated L/D to be 20.5 at CL 0.55 during maneuver segment 7. So I'd expect that our distance should be 22.5 / 18.84 = 1.19 times higher than in that maneuver.

    But more interesting is seeing the engine thrust which is actually negative there (idle about -3.1kN and max climb +169kN). With the previous setup where the engines didn't produce enough thrust at high altitude I think the idle thrust was actually negative at idle as well (fan windmilling mostly). But with the latest changes our simulated engine still creates noticeable idle thrust and I might have to adjust this. I'll probably have to adjust the nozzles a bit but without affecting the engine core speed that much.