Outerra forum

Please login or register.

Login with username, password and session length
Advanced search  

News:

Outerra Tech Demo download. Help with graphics driver issues

Author Topic: Hang gliding simulator  (Read 15116 times)

DeltaSim

  • Jr. Member
  • *
  • Posts: 48
Hang gliding simulator
« on: July 27, 2010, 10:15:49 am »

Hi,

I just finished my first setup of a hang gliding simulator which I might publish for training purposes (there is a real control bar and you hook into the simulator with your flying harness). I currently use FlightGear for it and based the flight dynamics on JSBSim, which is where Outerra comes in: Obviously FlightGear's graphics are far from what you need to have fun flying your hang glider while Outerra's fractal algorithms and beatiful forests are *exactly* what is needed for a good training simulator. I was delighted when I heard that you use JSBSim!

Now to the questions:
Is the xml-file that defines the flight dynamics in JSBSim basically unchanged?
Will there be anything special needed to assign joystick axes to variables in JSBSim (I use my own force sensors which appear as joystick axes)?
Weather: Do you plan to have wind, thermals and turbulence simulated? That's what a simulator would really need to use it in training.

Other than that - any news on when your first demo appears?

Great work, keep going!

- Martin
Logged
"The only way of discovering the limits of the possible is to venture a little way past them into the impossible." - Arthur C. Clarke

angrypig

  • Sr. Member
  • ****
  • Posts: 454
Hang gliding simulator
« Reply #1 on: July 28, 2010, 04:38:11 am »

Quote from: DeltaSim
Is the xml-file that defines the flight dynamics in JSBSim basically unchanged?

Yes these definition structures are unchanged.
The only changes we made are in FGLGear to improve collisions and dampening. These changes are based on Bullet library implementation.

Quote from: DeltaSim
Will there be anything special needed to assign joystick axes to variables in JSBSim (I use my own force sensors which appear as joystick axes)?

No nothing special. But currently we are propagating out only a few variables, like axes,trims and throtlle. All variable bindings are configurable through a text JSON file. We are using low level HID functions to work with external controllers so you will be able to use any USB device to control Outerra vehicles.

Quote from: DeltaSim
Weather: Do you plan to have wind, thermals and turbulence simulated? That's what a simulator would really need to use it in training.

All stuff like this we want to have implemented as plugins/modules and it will be possible after we will finish Outerra core. We don't know much about this domain. We would be interested what requirements are there on the engine itself  from such a perspective plugin/module, with regards to the thermals and turbulence modeling using terrain data.

Quote from: DeltaSim
Other than that - any news on when your first demo appears?

Cameni was writing something about this here.
Logged

DeltaSim

  • Jr. Member
  • *
  • Posts: 48
Hang gliding simulator
« Reply #2 on: July 28, 2010, 09:03:44 am »

Quote from: angrypig
We would be interested what requirements are there on the engine itself  from such a perspective plugin/module, with regards to the thermals and turbulence modeling using terrain data.

I am not at all an expert in computer games or simulation, but here is what I know:

FlightGear has a fairly simple model which - as far as I understand - consists of horizontal wind which can be defined in strength and direction for different altitudes. It also simulates the planetary boundary layer, i.e. decreasing wind and a change in direction as you come near the ground. On top of that you can choose a turbulence factor which probably just adds a random component to the wind. Thermals are modelled as fixed vertical cylinders with rising or falling air. You can have a cumulus cloud on top of each thermal if you wish. All this is already quite OK to practice your flying skills.

Since Outerra is obviously a completely different league to FlightGear in terrain modelling, it might be interesting to look into more realistic wind modelling as well, like solving some form of fluid flow equations. From simple to complex, the buzzwords would be

1 Potential Flow: Static over time, no eddies, no turbulence, follows terrain, no thermals, very simple equation to be solved once (Laplace)
2 Stationary Navier-Stokes: Static over time, eddies, no turbulence, no thermals, quite complex but doable equation to be solved once
3 Instationary Navier-Stokes: Changes with time, eddies, turbulence, no thermals, complex equation to be solved every few seconds
4 Instationary Navier-Stokes with heat transfer: Everything, incl. real thermals - not quite the holy grail of engineering maths but almost...

#1 Gives wind that you can use to soar along a ridge, which is already great. But it will not model the wake behind mountains, which is a bore. Implementation should be easy, turbulence and thermals can be added as in FlightGear.
#2 Will also model the wake behind a mountain which is particularly interesting because these wakes make flying in the mountains tricky. Again, turbulence and thermals can be added artificially. My personal favourite in terms of complexity and what you could expect.
I am not sure whether current PCs would be able to cope with #3 next to all the other jobs in the simulation. And for #4...

There is an open source fluid dynamics library called OpenFoam which you might want to check out.

By the way: You could combine #2 and #4 to give you a static flow field with thermals and only add turbulence - that would be quite cool. You would need to define ground temperature according to Sun illumination, ground colour, thermal capacity etc. to make something trigger the thermals.
Logged
"The only way of discovering the limits of the possible is to venture a little way past them into the impossible." - Arthur C. Clarke

Jon S. Berndt

  • Jr. Member
  • *
  • Posts: 24
    • http://www.jsbsim.org
Hang gliding simulator
« Reply #3 on: August 07, 2010, 08:12:27 pm »

Quote from: angrypig
Quote from: DeltaSim
Is the xml-file that defines the flight dynamics in JSBSim basically unchanged?

Yes these definition structures are unchanged.
The only changes we made are in FGLGear to improve collisions and dampening. These changes are based on Bullet library implementation.


That's interesting. I'd like to see that code. One of our developers has recently updated our landing gear model considerably using Erin Catto's paper. Maybe that is the same approach?

Jon
Logged
Jon S. Berndt
Development Coordinator
JSBSim Project
JSBSim.org

cameni

  • Brano Kemen
  • Outerra Administrator
  • Hero Member
  • *****
  • Posts: 6721
  • No sense of urgency.
    • outerra.com
Hang gliding simulator
« Reply #4 on: August 08, 2010, 03:35:26 am »

Quote from: jonsberndt
That's interesting. I'd like to see that code. One of our developers has recently updated our landing gear model considerably using Erin Catto's paper. Maybe that is the same approach?
It's Angrypig's adaptation of Bullet's raycast vehicle code, suitable to handle the bumpy surfaces in Outerra.
But it's nice to hear the landing gear model got updated, we can try it to see if it works here.
Logged

Ron Jensen

  • Newbie
  • Posts: 3
    • http://www.jentronics.com/fgfs/
Hang gliding simulator
« Reply #5 on: October 19, 2010, 09:19:20 am »

Quote from: DeltaSim
Quote from: angrypig
We would be interested what requirements are there on the engine itself  from such a perspective plugin/module, with regards to the thermals and turbulence modeling using terrain data.

I am not at all an expert in computer games or simulation, but here is what I know:

FlightGear has a fairly simple model which - as far as I understand - consists of horizontal wind which can be defined in strength and direction for different altitudes. It also simulates the planetary boundary layer, i.e. decreasing wind and a change in direction as you come near the ground. On top of that you can choose a turbulence factor which probably just adds a random component to the wind. Thermals are modelled as fixed vertical cylinders with rising or falling air. You can have a cumulus cloud on top of each thermal if you wish. All this is already quite OK to practice your flying skills.

FlightGear has had a "ridge lift" function for about a year now.  
http://wiki.flightgear.org/index.php/FlightGear_Newsletter_July_2009
http://www.flightgear.org/forums/viewtopic.php?f=6&t=3377
Logged

Jon S. Berndt

  • Jr. Member
  • *
  • Posts: 24
    • http://www.jsbsim.org
Hang gliding simulator
« Reply #6 on: October 19, 2010, 08:56:16 pm »

Quote from: jentron
FlightGear has had a "ridge lift" function for about a year now.  
http://wiki.flightgear.org/index.php/FlightGear_Newsletter_July_2009
http://www.flightgear.org/forums/viewtopic.php?f=6&t=3377

Hey, Jentron!

One of the JSBSim users from Germany recently added a very nice MIL-STD turbulence model. I'll see if I can get info on how to rurn that on. I think it would add a lot of realism to the flying experience.

Jon
Logged
Jon S. Berndt
Development Coordinator
JSBSim Project
JSBSim.org