Board index FlightGear Development Aircraft

Announcing NEW Boeing 777 'Seattle' - Pre-Release

Questions and discussion about creating aircraft. Flight dynamics, 3d models, cockpits, systems, animation, textures.

Re: Announcing NEW Boeing 777 'Seattle' - Pre-Release

Postby Clive2670 » Mon Apr 22, 2019 7:43 pm

Well done I find no issues with this version (so far) flown it for around 25 hours and it seems flawless, also my frame rate has increased thanks!
Thanks Clive aka: G-BLS01
Can be seen flying the Boeing 777-200LR
Toshiba laptop,
Intel i5 4210U,
16Gb RAM,
AMD R7-M260 graphics 2Gb RAM
User avatar
Clive2670
 
Posts: 593
Joined: Mon Feb 08, 2016 7:11 pm
Location: Anywhere in the World at some point on Flightgear!
Callsign: G-BLS01
Version: 2018.3.1
OS: Windows 10 64bit

Re: Announcing NEW Boeing 777 'Seattle' - Pre-Release

Postby Octal450 » Mon Apr 22, 2019 10:00 pm

Hi biloute974,
I didn't change anything FDM. That's a YAsim fault. Probably won't ever be fixed.

Hi Clive,
I am glad to hear that! I will get it into FGAddon soon.

Kind Regards,
Josh
Waste of time. Goodbye forever.
Octal450
 
Posts: 4398
Joined: Tue Oct 06, 2015 12:51 pm

Re: Announcing NEW Boeing 777 'Seattle' - Pre-Release

Postby tdammers » Wed Apr 24, 2019 8:09 am

it0uchpods wrote in Thu Apr 18, 2019 4:08 pm:It appears that the -200s have weird trim behavior, needing much to much nose down trim at all times. YAsim is weird, so I will set a trim bias for the 200s only, as the 300s trim is less weird. Maybe something to do with the hstab being more aft on the -300, and how YAsim calculates pitch moments.


Rebrand it as a 737-MAX. Problem solved. (Too soon?)
tdammers
 
Posts: 227
Joined: Wed Dec 13, 2017 10:35 am
Callsign: NL256
IRC name: nl256

Re: Announcing NEW Boeing 777 'Seattle' - Pre-Release

Postby Isaak » Wed Apr 24, 2019 9:45 am

:D
Isaak
 
Posts: 467
Joined: Sat Jun 04, 2011 2:52 pm
Location: Leuven, Belgium
Callsign: OO-ISA
Version: 2019.1.1
OS: Windows 10

Re: Announcing NEW Boeing 777 'Seattle' - Pre-Release

Postby tdammers » Wed Apr 24, 2019 11:28 am

OK, so I gave the 777-300 a spin (EHAM - EDDF); lots of great improvements. A few issues I encountered that may or may not be a fault on my part:

- The route manager seems to jump back instead of forward sometimes (but not consistently) when reaching a waypoint - I ended up flying the leg from ARNEM to COL three times (out and back and out again), and it also wanted to turn me back towards the first fix on the SID when I was about to reach the fourth.
- VNAV engages and appears to pick up my cruise altitude from the route manager, but does not initiate a descent in time to meet the altitudes set for the arrival route (I had it set to FL130 by ROLIS, but had to initiate a manual descent). Not sure if this is a bug, a missing feature, or a PEBKAC.
- Thrust reversers were impossible to disengage after landing - tried my joystick binding, Delete key, clicking the reversers in the 3D cockpit, nothing. Might be a nasal error somewhere causing the reverser logic to stop running; I've seen a similar problem when working on the CRJ700-family (there, a Nasal crash caused the FADEC code to stop running, and it became impossible to adjust thrust after landing). Didn't spot anything suspicious in the log output though.

Autoland works beautifully though.
tdammers
 
Posts: 227
Joined: Wed Dec 13, 2017 10:35 am
Callsign: NL256
IRC name: nl256

Re: Announcing NEW Boeing 777 'Seattle' - Pre-Release

Postby Isaak » Wed Apr 24, 2019 1:39 pm

Hi tdammers,

- I've seen the route manager jump back in the past, but I've never been able to reproduce it. It didn't change in the last updates.
- VNAV DES is not fully implemented yet. It is only able to calculate the TOD for a descent to the landing rwy. It has until now never been able to calculate cruise descents or match STAR altitude restrictions.
- the thrust reverser issue is new for me. I use the reverser positions of my Saitek throttle quadrant to toggle the reversers and used the delete key quite often last month but never had issues with them. Did your console say anything?

One issue I had on my flight last night was the cruise speed in the Route Manager getting reset to 0, causing the craft to throttle back in cruise and diving into a stall. I had to disable autothrottle and operate the throttle manually until I was descending to overcome it, as the speed got reset everytime I was cruising level again. It could be that I caused it myself by tampering with the property tree though.

An issue I've found since one of the last commits is that the RM cruise speed is at 0 when loading the sim. This used to be at 320 by default. I don't know if this has had anything to do with my issues of last night.
Isaak
 
Posts: 467
Joined: Sat Jun 04, 2011 2:52 pm
Location: Leuven, Belgium
Callsign: OO-ISA
Version: 2019.1.1
OS: Windows 10

Re: Announcing NEW Boeing 777 'Seattle' - Pre-Release

Postby tdammers » Wed Apr 24, 2019 2:15 pm

Isaak wrote in Wed Apr 24, 2019 1:39 pm:- I've seen the route manager jump back in the past, but I've never been able to reproduce it. It didn't change in the last updates.


Haven't looked at the actual code too deeply yet, but my guess would be that this might be a bug in turn anticipation. IIRC I ran into a similar situation when implementing this for the CRJ.

Isaak wrote in Wed Apr 24, 2019 1:39 pm:- VNAV DES is not fully implemented yet. It is only able to calculate the TOD for a descent to the landing rwy. It has until now never been able to calculate cruise descents or match STAR altitude restrictions.


Fair enough.

Isaak wrote in Wed Apr 24, 2019 1:39 pm:- the thrust reverser issue is new for me. I use the reverser positions of my Saitek throttle quadrant to toggle the reversers and used the delete key quite often last month but never had issues with them. Did your console say anything?


I did a bit of digging, and I think I found the problem (and a fix):

Code: Select all
diff --git a/Nasal/engine.nas b/Nasal/engine.nas
index c575de8..85e6521 100644
--- a/Nasal/engine.nas
+++ b/Nasal/engine.nas
@@ -123,7 +123,7 @@ var Engine = {
                 }
                 else
                 {
-                    if(me.throttle.getValue() == 0)
+                    if(me.throttle.getValue() <= 0.0001)
                     {
                         var reverser_cmd = me.reverser_cmd.getValue();
                         if(reverser_cmd != me.reverser.getValue())


Meaning that the reverser setting can only be changed when the throttle is set to exactly 0. My joystick however seems to emit something like 1.0^-15 - very very close, but not *exactly* zero, and so the reverser refuses to disengage. Engaging it works, because when auto-landing, the A/T retards the throttle, setting it to perfect 0; but once the reverser is engaged, A/T disengages, and moving the throttle sets the value to something other than perfect 0.
tdammers
 
Posts: 227
Joined: Wed Dec 13, 2017 10:35 am
Callsign: NL256
IRC name: nl256

Re: Announcing NEW Boeing 777 'Seattle' - Pre-Release

Postby biloute974 » Wed Apr 24, 2019 7:17 pm

Isaak wrote in Wed Apr 24, 2019 1:39 pm:Hi tdammers,

- I've seen the route manager jump back in the past, but I've never been able to reproduce it. It didn't change in the last updates.


This often happens when there are too many points to the flight plan. It also happens at home after two or three flights without leaving FG.
Intel I7 7700 - 16Gb DDR4 - Nvidia GTX970 - FG 2017.4.0 from D&C
biloute974
 
Posts: 191
Joined: Mon Feb 23, 2015 8:49 am
Callsign: U974
Version: 2016.1.0
OS: Mint 17.2

Re: Announcing NEW Boeing 777 'Seattle' - Pre-Release

Postby tdammers » Wed Apr 24, 2019 7:30 pm

biloute974 wrote in Wed Apr 24, 2019 7:17 pm:
Isaak wrote in Wed Apr 24, 2019 1:39 pm:Hi tdammers,

- I've seen the route manager jump back in the past, but I've never been able to reproduce it. It didn't change in the last updates.


This often happens when there are too many points to the flight plan. It also happens at home after two or three flights without leaving FG.


Seems unlikely; I've done the same route many times in a different aircraft, without any issues.
tdammers
 
Posts: 227
Joined: Wed Dec 13, 2017 10:35 am
Callsign: NL256
IRC name: nl256

Re: Announcing NEW Boeing 777 'Seattle' - Pre-Release

Postby sidi762 » Thu Apr 25, 2019 9:10 am

Isaak wrote in Wed Apr 24, 2019 1:39 pm:An issue I've found since one of the last commits is that the RM cruise speed is at 0 when loading the sim. This used to be at 320 by default. I don't know if this has had anything to do with my issues of last night.


Hi Issak,

I'm guessing you are using FlightGear 2019.1. In 2019.1 the RM causes problems when the autothrottle is switching IAS to Mach. It's a known issue which prevents the updated RM goes into 2018.3, but somehow it goes into version 2019.1 without the issue fixed. Since 2019.1 is not officially(?) released, I assumed that it was done on purpose for testing or something.

Best regards,
Sidi
sidi762
 
Posts: 130
Joined: Sat Jun 18, 2016 8:15 am
Location: Beijing, China
Callsign: DAG0762
Version: 2018.3.2
OS: MacOS Mojave

Re: Announcing NEW Boeing 777 'Seattle' - Pre-Release

Postby Isaak » Thu Apr 25, 2019 9:36 am

Thanks for the pointer, Sidi! I use 2019.1 indeed. I'll let James know that I 've come across this issue!
Isaak
 
Posts: 467
Joined: Sat Jun 04, 2011 2:52 pm
Location: Leuven, Belgium
Callsign: OO-ISA
Version: 2019.1.1
OS: Windows 10

Re: Announcing NEW Boeing 777 'Seattle' - Pre-Release

Postby biloute974 » Thu Apr 25, 2019 9:58 am

This is the observation I make.
Yesterday again, leaving from LMFL to EGKK, everything is going well. Then I empty the RM and I do EGKK at LFMN, everything is fine. I decide to do LFMN to LEMD and there, the plane from the departure wants to go directly to the last point. Require to validate each point in the RM by hand. However after a while, removing the previous points and those very close, the plane follows the road again. I will try to reproduce it and film it. I specify that I do not restart the simulator between each flight.

Edit: it's possible to do anything for the speed limitation during TO with flaps ? Or it's a Yasim restriction?
Intel I7 7700 - 16Gb DDR4 - Nvidia GTX970 - FG 2017.4.0 from D&C
biloute974
 
Posts: 191
Joined: Mon Feb 23, 2015 8:49 am
Callsign: U974
Version: 2016.1.0
OS: Mint 17.2

Re: Announcing NEW Boeing 777 'Seattle' - Pre-Release

Postby tdammers » Thu Apr 25, 2019 10:29 am

biloute974 wrote in Thu Apr 25, 2019 9:58 am:Edit: it's possible to do anything for the speed limitation during TO with flaps ? Or it's a Yasim restriction?


It's not *impossible* to fix, it's just that the way Yasim works, it would involve some serious kludges.

JSBsim is mainly built around multi-dimensional lookup tables, and interpolation between table cells. For example, to model a flap, you would have a 3D table, with inputs airspeed, AoA, and flap setting, and each cell would give lift and drag attachment points and vectors, and then JSB would pick the closest table cells for the current situation and interpolate between them to get the correct forces. This is a very flexible model, and doesn't rely on any particular aerodynamics simulation model - all it cares about is a bunch of sample values, and the assumption that interpolating between them is going to be close to correct.

By contrast, in Yasim, you specify flaps by parameters - position, angle, width, height, etc., and then Yasim applies some aerodynamic modelling to calculate the lift and drag forces from that. The problem with that is that the aerodynamic model is, by necessity, an oversimplification, and not necessarily correct across all flight regimens. However, changing any of these parameters will affect *all* parts of the flight envelope: fixing flap behavior in the takeoff situation will also affect flap behavior during approach, whereas with JSBsim, you could precisely tweak parameters for a specific situation. E.g., if the flaps produce too much drag at low speed and high AoA (takeoff), you might want to change their drag coefficient, but that will also cause them to produce less drag at higher speeds and lower AoA (approach), and the aircraft will be unrealistically difficult to slow down. In JSBsim, you would just adjust the table entries governing low speed / high AoA, and leave the high speed / low AoA ones untouched.

To fix this in Yasim would require a lot of experimental tweaking, and probably even a bit of "faking it" - you could for example define flaps with unrealistic parameters that happen to do what the flaps in the real aircraft do, or you could add a second set of (invisible) flaps and use property rules or Nasal to secretly extend or retract them based on the current situation. (The CRJ700-family uses a similar trick to model the aircraft's wing shape - the Yasim FDM has two pairs of wings, one modeling the inboard portion, and another the outboard part).
tdammers
 
Posts: 227
Joined: Wed Dec 13, 2017 10:35 am
Callsign: NL256
IRC name: nl256

Re: Announcing NEW Boeing 777 'Seattle' - Pre-Release

Postby Octal450 » Thu Apr 25, 2019 2:04 pm

Then maybe the A/P needs to be fixed. My IT-AUTOFLIGHT System works fine on 2019.1, as do any aircarft based off it (A32X, etc)

Kind Regards,
Josh
Waste of time. Goodbye forever.
Octal450
 
Posts: 4398
Joined: Tue Oct 06, 2015 12:51 pm

Re: Announcing NEW Boeing 777 'Seattle' - Pre-Release

Postby biloute974 » Thu Apr 25, 2019 5:10 pm

tdammers,

thanks for this precise answer.
Intel I7 7700 - 16Gb DDR4 - Nvidia GTX970 - FG 2017.4.0 from D&C
biloute974
 
Posts: 191
Joined: Mon Feb 23, 2015 8:49 am
Callsign: U974
Version: 2016.1.0
OS: Mint 17.2

PreviousNext

Return to Aircraft

Who is online

Users browsing this forum: No registered users and 2 guests