Board index FlightGear Development Aircraft Flight dynamics model

Fundamental FDM design problems.

Good sims require good FDMs (the "thing" that makes an aircraft behave like an aircraft).

Fundamental FDM design problems.

Postby helijah » Sun Oct 06, 2019 7:09 pm

Split off from the topic Next-generation scenery generating?

legoboyvdlp wrote in Sun Oct 06, 2019 4:32 pm:Ok - how about this - you add procedural lights, a windshield glass effect, interior shadows, and a proper 3D cockpit to your 737-100 and I'll bring you a JSBSIM FDM - incidentally I have enough documentation to do so if I was interested in it - but lack of time might be a problem :roll:


The answer is simple:

With more than 300 models to maintain the work is done slowly, but the 737-100 I have at home is already much more advanced than the one on the depots. Simply there are hundreds of instruments that I have to create in 3D before adding them. I did it for the Mirage 2000 and I do it for others too. A little bit here, a little bit there. Not to mention the new models that are in progress and will arrive soon.

Then an FDM JSBSIM? It just shows that you don't know what to talk about. If you had studied my work at least a little bit before writing here you would know what I think of JSBSIm and you would know that ALL my models are YASim.
Offering me an FDM JSBSim is the certainty that I will do nothing for you and that you are talking without taking the time to check first.
Some planes (and other) for FlightGear
http://helijah.free.fr
and
http://embaranger.free.fr
User avatar
helijah
 
Posts: 1049
Joined: Wed Dec 27, 2006 12:35 pm
Location: Chartres (France)
Callsign: helijah
IRC name: helijah
Version: GIT
OS: GNU/Linux

Re: Next-generation scenery generating?

Postby Thorsten » Sun Oct 06, 2019 7:12 pm

And when someone asks me for improvements, I make them without any problem


Really?

Unless you happen to be... not interested:

Then an FDM JSBSIM? It just shows that you don't know what to talk about. If you had studied my work at least a little bit before writing here you would know what I think of JSBSIm and you would know that ALL my models are YASim.

So quite obviously when someone asks you about a JSBSim FDM, suddenly you do not 'make without any problem'. When someone asks you to merge ALS support in, you do not act 'without any problem'.
Thorsten
 
Posts: 11441
Joined: Mon Nov 02, 2009 8:33 am

Fundamental FDM design problems.

Postby helijah » Mon Oct 07, 2019 1:44 pm

The discussion is restarted:) But God, it takes effort and abuse to get there:)

There are many little things to do. Another rather logical example. Very logical even.

The interactions of aircraft with the ground, water, sand, dust trails, water jets etc.....
This should be integrated into FlighhtGear's 3D engine and not in aircraft.This would bring homogeneity to the different aircraft and vehicles and for many users it would make the aircraft folders much lighter. Making downloads lighter and faster at the same time. But now, I recognize it, I speak first and foremost for my parish lol. And anyway, it's only possible with YASim, JSBSIm makes him bounce planes on the water ridiculously. Although this could also benefit JSBSim in this regard.

It's an idea like any other. But the homogenization of the models, a more efficient rendering of the effects since directly in the 3D engine of FG (with GLSL why not as some like this programming :) ).
Some planes (and other) for FlightGear
http://helijah.free.fr
and
http://embaranger.free.fr
User avatar
helijah
 
Posts: 1049
Joined: Wed Dec 27, 2006 12:35 pm
Location: Chartres (France)
Callsign: helijah
IRC name: helijah
Version: GIT
OS: GNU/Linux

Re: Next-generation scenery generating?

Postby Thorsten » Mon Oct 07, 2019 2:05 pm

This should be integrated into FlighhtGear's 3D engine and not in aircraft.


Well, the way rendering works you'll just have to put up with it. It's neither reasonable to let FG guess whether a surface is internal or external glass nor is it feasible to let the renderer compute the airstream across an external surface just to get a droplet impact pattern.

Nor does it make sense to let the renderer guess what amount of rotor-wash or propwash you currently have. It's not what the renderer is for - the renderer creates graphics on the screen, physics questions are solved by the FDM.

Generally, things which 'just work' in rendering are pretty costly for the framerate - which is why we won't do that, especially if the aircraft author can usually set the correct parameters in five minutes.

And, well, we do have effects available which can be utilized to get rid of particles for xXX-trails - you just have to use them (which seems to be the theme here).

integration generally is very simple:

* assign the effect to a surface / volume
* include the master effect from FGData
* set the parameters in your own effect file


and for many users it would make the aircraft folders much lighter


Get real - aircraft-side effect definitions are a few ten kb ascii files, you can duplicate them ad nauseam and not have any negative impact on size.

And anyway, it's only possible with YASim, JSBSIm makes him bounce planes on the water ridiculously.


Seems you have no clue at all of JSBSim, because it not only is used to simulate sailing ships, but also it can get water planes with full buoyancy calculations for the swimmers.

***

Anyway, you claimed that you react to user requests 'without any problems' - you got a clear request - so... get coding please!
Thorsten
 
Posts: 11441
Joined: Mon Nov 02, 2009 8:33 am

Re: Next-generation scenery generating?

Postby legoboyvdlp » Mon Oct 07, 2019 3:24 pm

helijah wrote in Sun Oct 06, 2019 7:09 pm:
Then an FDM JSBSIM? It just shows that you don't know what to talk about. If you had studied my work at least a little bit before writing here you would know what I think of JSBSIm and you would know that ALL my models are YASim.
Offering me an FDM JSBSim is the certainty that I will do nothing for you and that you are talking without taking the time to check first.


You were the one that claims to speak for those who want realistic VFR flight - so how on earth could that be accomplished with YASIM? All your models are YASIM - obviously - but some people actually care that an aircraft actually flies properly - not a fantasy aircraft!

And anyway, it's only possible with YASim, JSBSIm makes him bounce planes on the water ridiculously. Although this could also benefit JSBSim in this regard.


You ever tried the 172? :)
User avatar
legoboyvdlp
 
Posts: 7388
Joined: Sat Jul 26, 2014 1:28 am
Callsign: YV-LEGO
Version: next
OS: Windows 10 HP

Re: Next-generation scenery generating?

Postby helijah » Mon Oct 07, 2019 7:03 pm

legoboyvdlp wrote in Mon Oct 07, 2019 3:24 pm:You ever tried the 172? :)


Tiring of that kind of childish reaction. Unlike you ( it seems) I test ALL the aircraft in FlightGear. ALL without exception. The C172 is an exception. Not bad at all indeed It is able to land on its back, turn around and take off again...rather funny...but as far as realism is concerned............. That being said, compare my DR400-120 or 140 with the PAF DR400 (which in passing uses 40% of my work without reporting it (but that's another subject)
And to criticize as you do my models is also to make fun of some real pilots who have spent time making the flight models more realistic (In particular one instructor Belgian pilot on 777 and one Alphajet pilot )......, . And I don't congratulate you on that. Indeed I provide a minimum for my aircrafts. And over the past 2 or 3 years, if you had taken the time to test my latest ones, you would have noticed that the flight models have improved significantly.

I have always pointed out that JSBSim is a fantastic job and a fantastic project. Even if some misrepresent what I'm saying. I used it for almost 2 years before I realized that it didn't bring anything to the general public. But the general public, the basic users you don't care what it seems.

My planes are not perfect. Yes. I take it and try to improve by listening to the people who write to me. FlightrGear could be better for basic users. But you and others always find a good reason to do nothing but continue your personal little work.

On this I would not continue this discussion whith people totally hermetic and locked in their certainties do not accept any discussion.
Some planes (and other) for FlightGear
http://helijah.free.fr
and
http://embaranger.free.fr
User avatar
helijah
 
Posts: 1049
Joined: Wed Dec 27, 2006 12:35 pm
Location: Chartres (France)
Callsign: helijah
IRC name: helijah
Version: GIT
OS: GNU/Linux

Re: Next-generation scenery generating?

Postby Thorsten » Mon Oct 07, 2019 7:44 pm

I have always pointed out that JSBSim is a fantastic job and a fantastic project.


Like you did here?

JSBSIm makes him bounce planes on the water ridiculously


Or here?

If you had studied my work at least a little bit before writing here you would know what I think of JSBSIm and you would know that ALL my models are YASim.


Do you actually realize that you contradict yourself in nearly every post?

My planes are not perfect.


Amen.

And to criticize as you do my models is also to make fun of some real pilots who have spent time making the flight models more realistic


Well, unlike you, we're probably well aware of the limitations of YaSim...
Thorsten
 
Posts: 11441
Joined: Mon Nov 02, 2009 8:33 am

Re: Next-generation scenery generating?

Postby legoboyvdlp » Mon Oct 07, 2019 8:14 pm

Not bad at all indeed It is able to land on its back, turn around and take off again...rather funny...but as far as realism is concerned

That's totally and absolutely wrong - it won't.

I used it for almost 2 years before I realized that it didn't bring anything to the general public.

I started to type out an extensive argument as to why this is not correct - but I really don't have the energy right now...
User avatar
legoboyvdlp
 
Posts: 7388
Joined: Sat Jul 26, 2014 1:28 am
Callsign: YV-LEGO
Version: next
OS: Windows 10 HP

Re: Next-generation scenery generating?

Postby swampthing » Mon Oct 07, 2019 8:38 pm

JSBsim will have ground reactions as good as the author makes them. If you don't add contact points and do it right of course it will be bouncy or zing off to the side forever, thats part of setting up your contact points properly. People who use JBSsim know this.

I will grant Helija this, he has listened when I told him about problems I found in some aircraft and worked to make them fly better, so this is not a lie. None of us were born knowing how to do any of the work thats done now. We woke up and wanted a drink of milk.

I personally prefer JSBsim. If you crash a YAsim plane its a yard dart. YASim does NOT seem to be as flexible working with the terrain from all of my experiences but its also a strange FDM thats hard to wrap your head around when you are use to finding real data to build your FDM. Helija you have to get past hating JSBsim, for some reson I still have not understood. You have probably more models than anyone but they don't get improved because people don't want to offend you, Why? If they are going to help improve 3D they will also want a good FDM YASim has the terrible death pitch and look how much time people spend trying to correct YASim problems with nasal. Your plane would be converted to JSBsim, we all know you don't like that.

For the majoraty of YASim planes I've flown it just feels very un-natural. If you fly a plane that has bad ground reactions that can be fixed by doing them correctly with JSBsim. Its not the fault of JSBsim. I don't know of any YASim planes (even some I fly often) that have some outstanding reactions in the water. I'm not here to bash YASim even though I just gave some bad reviews but it is what it is. I can say it no other way. Either way you are going to have to spend time on an FDM to make it work properly, you can have good aerodata and ground reactions with JSBsim (takes time) or fight with nasal hacks with YAsim (takes time).
www.opredflag.com
I have sworn upon the altar of God, eternal hostility against every form of tyranny over the mind of man. -Thomas Jefferson-
swampthing
 
Posts: 560
Joined: Wed Oct 28, 2015 4:10 am
Location: Missouri
Callsign: swamp
Version: 2018.2
OS: multiple

Re: Next-generation scenery generating?

Postby wkitty42 » Tue Oct 08, 2019 10:23 am

helijah wrote in Mon Oct 07, 2019 7:03 pm:
legoboyvdlp wrote in Mon Oct 07, 2019 3:24 pm:You ever tried the 172? :)


Tiring of that kind of childish reaction. Unlike you ( it seems) I test ALL the aircraft in FlightGear. ALL without exception. The C172 is an exception. Not bad at all indeed It is able to land on its back, turn around and take off again...rather funny...but as far as realism is concerned.............

i don't know which c172p you are testing but the one that is distributed with FG surely doesn't do that... not unless you specifically click on the repair button which is specifically included so you don't have to exit the sim and restart it just to go flying some more... perhaps you can/should provide a video of this situation you describe... if it is a real problem, the dev team for the craft can then fix it...
"You get more air close to the ground," said Angalo. "I read that in a book. You get lots of air low down, and not much when you go up."
"Why not?" said Gurder.
"Dunno. It's frightened of heights, I guess."
User avatar
wkitty42
 
Posts: 6105
Joined: Fri Feb 20, 2015 3:46 pm
Location: central NC, USA
Callsign: wk42
Version: git next
OS: Kubuntu 14.04.5

Re: Next-generation scenery generating?

Postby Thorsten » Tue Oct 08, 2019 11:22 am

@wkitty42:

You're taking this way too serious... it's actually more akin to a comedy, because the same person who does thorough testing

Unlike you ( it seems) I test ALL the aircraft in FlightGear. ALL without exception.


also happens to explain to us how that thorough testing is done in detail

Simply because I personally do not use FLightGear. I have been flying simulators for over 30 years starting with FS I. I only use it to test my models and offer a minimum usable.


So, we may safely deduce that

'ALL aircraft without exeption' in his language means 'MY aircraft only' and that 'test' means 'making sure it is usable and doesn't crash the sim' :D

In other words, he can't actually pilot and has never tried the 172p (since it's not his aircraft), he's just judging FDM quality from how it appears on the runway.
Thorsten
 
Posts: 11441
Joined: Mon Nov 02, 2009 8:33 am

Re: Fundamental FDM design problems.

Postby Octal450 » Tue Oct 08, 2019 4:54 pm

If you are developing 300 models at once - clearly you aren't focusing on the realism then as doing realism of 300 models at once is simply not possible.

So then use your YAsim because you don't need realism anyways.

Bit of a useless conversation guys...

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

Re: Fundamental FDM design problems.

Postby swampthing » Tue Oct 08, 2019 7:25 pm

@Octal450

Bit of a useless conversation guys...




No, No its not. You don't leave things that are not true as if they are truth. New people come to the forum and we just leave things that are not true?
www.opredflag.com
I have sworn upon the altar of God, eternal hostility against every form of tyranny over the mind of man. -Thomas Jefferson-
swampthing
 
Posts: 560
Joined: Wed Oct 28, 2015 4:10 am
Location: Missouri
Callsign: swamp
Version: 2018.2
OS: multiple

Re: Fundamental FDM design problems.

Postby abassign » Wed Oct 09, 2019 7:18 am

If you want a JSBSim airplane, complex and hard to pilot... with AI pilot system for look the panorama around you :D ... This is good start point:
http://wiki.flightgear.org/FIAT_G91R1B

But anyway I remember that making good airplanes is not easy, I see many FGFS fans who are working very well, but the best planes are few because they require a lot of work. I personally think that JSBSim is a very good FDM, I'd like it to be more flexible and complete to be able to replace NASAL more, as JSBSim is much more suitable for real-time applications.
abassign
 
Posts: 831
Joined: Mon Feb 27, 2012 5:09 pm
Location: Italy (living 5 Km from airport LIME)
Callsign: I-BASSY
Version: 2018.3
OS: Linux Mint 19. x

Re: Fundamental FDM design problems.

Postby Thorsten » Wed Oct 09, 2019 8:58 am

I'd like it to be more flexible and complete to be able to replace NASAL more, as JSBSim is much more suitable for real-time applications.


Why would you want your screwdriver heavier so that you can use it better to hammer in nails when you could... simply use a hammer for the job :?: :?: :?:
Thorsten
 
Posts: 11441
Joined: Mon Nov 02, 2009 8:33 am

Next

Return to Flight dynamics model

Who is online

Users browsing this forum: No registered users and 1 guest