Yaw dynamics

[744student]

Hi!

First of all, thank you for the great 744 sim! Its accuracy and attention to detail have not gone unnoticed!

I don't know if this has been discussed previously, but I find that during crosswind landing de-crabbing, the aircraft seems to start drifting downwind much quicker than in real life (or FFS for that matter), making it quite hard to do what I normally do in real life (de-crab during flare, wings level until touch down) under certain cross wind/runway conditions. Has this ever been brought up or is it just me?

On another topic, the ground steering radius seems a little too wide, making a 180 turn on a 60m runway impossible. This is despite following the FCTM techniques and using full tiller input... Is there a way I can adjust the parameters?

Thanks in advance!

P.S. I'm rated on 737NG/MAX,744,757/767 and 777 so I'd like to think I know what feels natural during Xwind landing in a Boeing...

[ASCTU744]

Hi!

First of all, thank you for the great 744 sim! Its accuracy and attention to detail have not gone unnoticed!

I don't know if this has been discussed previously, but I find that during crosswind landing de-crabbing, the aircraft seems to start drifting downwind much quicker than in real life (or FFS for that matter), making it quite hard to do what I normally do in real life (de-crab during flare, wings level until touch down) under certain cross wind/runway conditions. Has this ever been brought up or is it just me?

On another topic, the ground steering radius seems a little too wide, making a 180 turn on a 60m runway impossible. This is despite following the FCTM techniques and using full tiller input... Is there a way I can adjust the parameters?

Thanks in advance!

P.S. I'm rated on 737NG/MAX,744,757/767 and 777 so I'd like to think I know what feels natural during Xwind landing in a Boeing...

Maybe its your wind settings, are you using aerowinx on it's own or coupled to another sim for the visuals?

[Hardy Heinlin]

Hi,

welcome aboard!

No, this has never been brought up so far. It's always important to describe your expectations in numbers: "I notice PSX drifts x meters in y seconds; on the real 744 I notice xx meters in yy seconds." – This way we can analyse the problem better. The more details the report includes, the better we can discuss it. Also important is wind data, weight, USB flight control calibration, what scenery system is used – PSX scenery or MSFS, X-Plane etc. External scenery interfaces include a motion smoothing algorithm which doesn't necessarily always agree with the actual dynamics within PSX; so for discussions about the flight model in PSX, the reference should be the simplified scenery in PSX and the PSX flight instruments. This also affects taxi turns. In PSX, when your tiller is fully deflected and there's no icy taxiway and your groundspeed is low enough, you should be able to get the correct and realistic turn radius. When you are too fast you will hear a noise and experience nose gear scrubbing; that's probably not the case in your situation.

To make it short: Are you using any external scenery system with MSFS or X-Plane etc.?


Regards,

|-|ardy

[United__744]

+1 for noticing the turn radius is very large. Not sure when this changed, but I'm trying to turn around at the RW28 end of SCIP, using the turn-around pad, and the 747 ends up off the other side of the runway.

It's like the body gear is not steering.

[Hardy Heinlin]

Nothing has changed since 2008.

I just retested it at 5 knots and my turn radius was 25 meters. Is that too large?

Do you use X-Plane scenery with an interface add-on? Maybe the interface add-on dampens the turn rate in the X-Plane scenery view.

Tip: Never judge PSX behaviour by looking at add-on behaviour. Always refer to the instruments in PSX and the tools on the Instructor.

[macroflight]

I really wanted to say "yes, the turn radius feels a bit too large", because that's how I often feel when using PSX. But I don't really know how the real aircraft is supposed to behave. It is very large and airports are very small, after all...

However, after having taught Claude to taxi a 747 and measure things... I can't prove Hardy wrong here. As far as I can tell the low-speed turn radius seems realistic.

The only real discrepancy found was that the effect of disable body steering seems to be larger in PSX than what Boeing says it should be. But unless 744student and United__744 are taxiing around with body gear CBs pulled or HYD PRESS DEM 1 showing on the ND that shouldn't be a factor.

AI slop report, Python code and data

[Hardy Heinlin]

What you guys still haven't told is whether you refer to an external scenery view or to the original PSX behaviour. On the track plotter in PSX you can see the turn radius agrees with that of the real 744 (circa 25 meters) and a turn is possible within a 50 meter wide runway.

External scenery links usually apply a smoothing algorithm to smooth out stuttering effects in the data stream of the position and heading. This way the external view gets an additional inertia and lags behind the actual PSX position and heading.

You could also watch the actual track on an external moving map with an interface add-on. Those typically don't use a smoothing system.

[macroflight]

My tests yesterday were done by measuring PSX's turn rate by looking at Qs121 position data. Those tests showed a 25m turn radius at low speed (unless body gear steering was intentionally disabled).

I got the same 25m radius both with plain PSX and with MSFS connected. So having MSFS connected does not appear to affect PSX' own movement in the turn.

But note that this is the turn radius of the Qs121 reference point. I don't know where that is located, but it's probably somewhere on the aircraft centreline and so its turn radius would be different than the turn radius of e.g the right outer gear.

I did some more tests today, visually measuring the minimum runway width needed to make a 180 degree turn to be 61m (stopped with outer gear no runway edge marking aligned with runway heading, then full tiller (verified Qh426 reached 999), keeping speed around 5kts (GS on ND) using light braking on inner wheels and a little assymetric thrust).

Even if there would be a significant delay in the PSX-MSFS link, that should not affect the stationary aircraft position before and after the 180 degree turn, right?

The 61m value was measured using MSFS' 3D "Ruler" tool. To verify the tool, I also measured the distance between the runway edge lines to be 45m, which matches the published test runway width (ESSA rw 26).

Before turn:


After turn:


As far as I know the minimum runway width required to make a 180 degree turn is 46m for the 744, so the 61m result is way too large.

[Hardy Heinlin]

So having MSFS connected does not appear to affect PSX' own movement in the turn.

I didn't mean to say that PSX is affected by the add-on, but that the external scenery view in X-Plane or MSFS may be affected by the add-on's smoothing system. E.g. when the heading in PSX turns by 5° in 100 ms, the external view may turn by 5° in 120 ms because it has an additional visual inertia. That does not affect the dynamics in PSX.

You are right that the radius reference in PSX is the aircraft center and not the nose gear. The radius at the center might actually be zero and the turn would then be made by the nose gear alone, and the radius would then be the distance between the aircraft center and the nose gear, which is about 27 meters. A certain fraction of these 27 meters will be added to the main radius of 25 meters. So, that makes sense; I probably need to decrease the main radius by that fraction.

[Hardy Heinlin]

I just checked my update history. Actually, there was a modification:
______________________________________

Version 10.42 - 16 August 2018

42.03. Nose gear steering turn radius modified; it was too small at very low speeds and too big at high speeds.



Version 10.43 - 3 September 2018

43.07. Nose gear steering turn radius further fine-tuned (source: BA sim).

[Hardy Heinlin]

PSX update 10.190 is uploaded. The taxi turn radius referring to the aircraft center is now reduced to 14 meters on dry surface at slow speed, symmetric thrust and brakes, with body gear steering operating. This allows a U-turn within a 46-meter-wide runway.

https://aerowinx.com/board/index.php/topic,4191.0.html


|-|ardy

[macroflight]

As usual - amazing support from Hardy. Thanks!!

I have verified that with 10.190, I can do a 180 degree turn on a 46 m wide runway.

Same test as with .189, but now we almost stay within the 45m wide runway edge lines on ESSA/26 instead of going into the grass.

[macroflight]

I find that during crosswind landing de-crabbing, the aircraft seems to start drifting downwind much quicker than in real life (or FFS for that matter),

Let's not forget this part of the original post. It is an interesting observation.

I have no idea how the real aircraft or the FFS behaves, but I can try to start the discussion by offering some possible causes other than PSX behaving incorrectly:

1: You put in more rudder than you think (e.g due to rudder calibration or pedal resistance), so the nose swings towards the runway heading faster than you expect, so the aircraft drifts more before touchdown

2: Your visuals are not aligned correctly (screen size + zoom level + eye to screen distance not matching), so you perceive the aircraft as yawing slower than it actually does, so you put in more rudder to compensate, which leads to a faster drift downwind

[macroflight]

I did a very basic test to rule something out: aircraft on runway centerline on a normal ILS. I then freeze motion at 300 ft RA and set a 30 knot crosswind in PSX, switched the autopilot to heading hold and V/S and then released the motion while monitoring the aircraft drift via Qs121 data.

I did not expect this to be a problem, and it is not. PSX drifts downwind at 30 knots with a 30 knot crosswind. I also tested calm, 5, 10, 20 and 40 knots direct crosswind and a 45 degree 30 knot crosswind, all with the expected results.

[Hardy Heinlin]

I guess what he means by "quicker" is not too high a crosswind but too quick a track change; he probably expects a higher inertia, a longer transition time from non-sidedrift to full sidedrift.

[jawomba]

I found two related videos:

B747-400 turn of 180 degrees on runway turn node

Minimum Turning Radius of Aircraft | Derivation | Airport Enginneering

cheers Jan

[United__744]

Here is a Boeing guide on airport planning. It specifies for -400/ER/F types that:

* Nose gear +/- 70 degrees, and;
* Body gear steering +/- 13 degrees
* Low-speed taxi
* Minimum width for 180 degree turn with BGS operative is 152 ft or 50.5 m.

PDF page 102 onwards.

https://www.boeing.com/content/dam/boeing/v2/airports/acaps/747-400_Rev_F.pdf

[jawomba]

X = 14 m is in accordance to D6-58326-1 section 4.2.1

=> required (min.) runway width for a 180 deg. turn - within runway edge markings: B (3) Nose Gear (2) + C (3) Wing Gear (2) in a perfect world, right?
(refering to D6-58326-1 section 4.3.2)

Cheers Jan

[ahaka]

After trying the latest update, it seems that gusty crosswind landings feel more stable/easier. Was the inertia increased in this update as the aircraft seems to resist the wind blowing from the side when de-crabbing? I also thought I noticed that you need to hold the rudder a bit more in the de-crab when previously you needed to neutralize it or even put a bit of opposite rudder after touchdown.

Until now a gusty strong crosswind with turbulence was quite difficult to handle sometimes and often resulted go-arounds but now the plane settles quite nicely every time.

Is the new behaviour more like the real thing? I thought the previous model was already regarded as spot on by those flying the real thing.

[Hardy Heinlin]

I haven't changed the aerodynamics but the intensity of the turbulence in the weather model based on this new discusssion.

I just checked your forum posts, Antti; your last post was from 2020 :-) Welcome back. There was a special modification in the past years that partially generates an "automatic" decrab on touchdown due to asymmetric tire contact which can also be observed on the real 744 in many videos.

-------------------------

Version 10.100 - 27 February 2020

100.01. Aerodynamics & ground motion model: A crabbed touchdown will now cause a self-decrabbing yaw moment (most effective on dry runways). As the main wheels are behind the center of mass (behind the yaw axis) and flying sideways along the inertial track, the ground contact will swing the tail toward the track, so that the main gear and the center of mass will be nearly in-line with the track (source: Video analyses & Britjet's theory).