"LVL CHG" button abnormal descend

[Andrey Novikov]

I was cruising at FL330 and .72M, have set altitude for descend (FL130) and pressed "level change" button. After that vertical speed quickly raised to -6000 fpm and airspeed became almost critical. So I had to manually level the aircraft and use V/S for further descend. I do not know how to reproduce it but I remember that this already happened to me once but these where my first days of flying B733 so I then thought that I was doing something wrong.

[Morten]

Seeing 4500-5000 fpm is normal, but 6000 seems a bit high.  Was the aircraft very light?

Also, if you are using XP11, lift and drag has changed significantly so wait for the IXEG XP 11 update which addresses these issues.

M

Edited June 20, 2017 by Morten

[Andrey Novikov]

I prefer to fly with 48-52 ZFW, do not know if this is light or not. Of course I'll wait.

[Litjan]

There could be a variety of contributing factors, especially descending in MACH speed after passing the normal MACH/SPD switchover altitude. MACH will stay constant, but IAS will rise rapidly while descending (into warmer air)...

Jan

 

[mfor]

Same thing here:

Selecting LVL CH at FL330 made the vertical speed drop to 6000+ fpm for about 2000ft before slowly recovering towards 2500 at FL290. Speed went up from .74 to .785+ (294)@~FL300 - still a bit below the red band though, so I didn't think much of it.

This was with a ZFW of 48.3 tons plus 5.1t fuel left and a 42 knots headwind on XP-10.

[Andrey Novikov]

@Litjan No, it was developing at lightning speed, in 2-3 seconds, just after the button press.

[Litjan]

Ok, try to capture a video if it, maybe - or a series of screenshots, so we can see whats going on...

Cheers, Jan

 

[mfor]

See attached a video (15MB) - if you need better resolution/sound, let me know.

lvlch.mp4

[Litjan]

Hi and thanks for the video - I think the problem is the turbulence - the PID algorithm may be fooled by the up and down of the airspeed. We did some adjusting for a similiar problem during MACH climb... but will look at this again.

In the meantime I would suggest to use FL CHG with MCP SPD instead of MACH - it might track a bit better.

Cheers and thanks for your help, Jan

 

[mfor]

I think I did a flight without turbulence and it was more or less the same, will try with MCP speed set to knots.

 

 

Edited June 22, 2017 by mfor

[mfor]

Did a few more tests without wind and no turbulences.

At first I thought they were inconclusive, because I got drops to 6000 with both settings and sometimes both would behave well (~4000-ish).

However during one test with a drop I noticed a slight stutter during the nose dive, which, I presume, was the result of either x-plane or an addon performing some loading action due to the view changing (my prime suspect is Skymaxx). Such a leap in the time continuum could easily throw off any PID controller.

So I disabled (almost) all plugins and relocated the testing area out to open see. So far I haven't seen any unruly drop in the vertical speed there. Haven't tried with turbulences yet and the number of test has been rather small - running out of fuel 

However I currently think that drop is due to fluctuations in the physics, either in form of the frame rate (stutter or slow downs) or maybe wind speeds due to turbulences - not sure yet about the latter.

 

 

Edit: It seems turbulences (set to 1) can cause this as well - although it seemed to be a little less extreme (would barely touch 6000)

 

 

PS: Didn't know you can ditch planes in X-Plane - good thing the operations manual includes a ditching section. Drag seems a little low though.

 

PPS: Hope the life vests are included with IXEG's model.

 

PPPS: Send boats... or nudes... or both... 

 

Edited June 22, 2017 by mfor

[juan0511]

Hi, this happened to me as well 3 times during the same flight (ZFW 44.48 tonnes, fuel 8.0 tonnes) until I switched to V/S mode for descent. The ATC was making me descend gradually:

• From cruise FL370 to FL350. Speed was Mach .80, there was a pretty strong headwind of around 90kt. No turbulence. As soon as I pressed LVL CHG the descent rate went to 6000ft/min and had to keep the speed in check with the use of speed brakes.
• I was then cleared to FL240, when I pressed LVL CHG this time, the same happened. But this time I believe it was a lot more than 6000ft/min, judging by the rate of descent seen in the altimeter. When the mach speed warning fired, I leveled off manually.
• I tried once more, I was at FL290 or so, but the behavior was exactly the same. At this point I began using V/S mode.

I've seen the descent rate go up to 4750 ft/min (on NG 737s though), but always keeping the airspeed in check, even if flying at high speeds before applying LVL CHG. I hope you can take a look into it, if there's a specific test I can do to help, please let me know.

Keep up the good work, this plane is a delight to fly! Definitely Xplane's best!

[Litjan]

I will take a look at high-speed Mach descents again to see if I can find out what went wrong there for you...

Thanks for the report!

Jan

 

[Morten]

Looking at some FDR data, and it is not uncommon to see 5000+ fpm for a short period of a few secs right after starting the descent at TOC in FLCH.  6000+ is a bit high though..

Edited October 13, 2017 by Morten

[juan0511]

Jan, Morten, thank you for your replies! I've been running some tests, flew up to FL370, down to 350, then up again to 370, down again to 270. Speed was mach .80.

• The first time the vertical speed went to 6000 ft/min and stayed there until it reached 350. Speed went up to mach .82.
• The second time was worse, this time I enabled the data output window (see screenshot). It was at 7600 ft/min for maybe 6-7 seconds then stayed in the 6000s until I told it to go to 370 again.

 

 

EDIT: I did three more tests, same altitudes, two times starting at .80 mach and the last time starting at .75 mach. This time I let the aircraft reach the selected altitude, all three times the result was the same. At first the descent rate goes to 7500+ for a few seconds, then 6000+ and finally it manages the descent properly, keeping the dialed in speed. When starting at .80, it will momentarily accelerate to .82, when starting at .75, the speed raises momentarily to .78 mach. Hope this helps.

Edited October 13, 2017 by juan0511

[Litjan]

Thanks for your tests.

I will see if we can tune the behaviour a bit - however an excursion of 0.1 Mach is not unusual when the autopilot flies the plane.

Also keep in mind that 0.8 is a VERY high speed for a 737-300 to be flown at. Usual cruising speed is .74, and we flew .78 if we were really in a hurry.

The FL CHG pitch mode is also limited by the g-load it will place on the passengers. At very high speeds, it can not "pull up wildly" to fight an overspeed situation, this would be very uncomfortable. In fact a lot of real pilots will refrain from using FL CHG mode in these regimes (high alt, high Mach) because of the exact reasons you describe. Very steep descents, low pitch.

Cheers,Jan