Outerra forum
User mods, screenshots & videos => Aircraft => Flight Model Development => Topic started by: Uriah on March 15, 2015, 05:48:31 am
-
I am working on an Aircraft Development Toolkit for Outerra that will help people build accurate flight models, taking measurements directly from the 3d model and real world data. Some features, like force vector visualization for flight testing, will use a bone (e.g. BONE_CG or BONE_AC) and different colored 3D objects, such as Vector Arrows or Center Marks, that will be a attached to these bones. In the JS script I will set the bones to the position/rotation of the corresponding center or force vector, and use an autopilot to hold level flight at a specific altitude for standardized flight testing. There will be a variable to hide/show the visualization objects, as well as this.log_inf() these properties and output a CSV file. As part of the Aircraft Development Toolkit I'll include a users guide, 3ds Max file including the visualization objects, bones and helpers, the script code to animate each visualization feature, and FDM examples of a number of different flight modeling methods that can be used for different purposes/types of aircraft .
The JS script will have three versions, airplane, helicopter and rocket, which use different visualization techniques. For wings, I will be able to visualize multiple wing-span sections, with individual and collective lift force vectors with the origin at each Aerodynamic Center. Using the same method, I can model a helicopter rotor, modeling each blade separately as a rotating wing plane, with each blade having multiple blade-span sections, to compute lift, drag and stall forces for each blade tip section, mid section and root section. This helicopter FDM will account for prop pitch angle of attack and true airspeed of the rotor airfoil, modeling retreating blade stall, engine compression stall, true auto-rotation of the main rotor, and rotor blade flex using properties for rotor rpm, coefficient of flex and position along the blade-span, as well as other features. In the future, I will develop an automatic FDM and script generator that will use a configuration file to build a working flight model ready for detailed optimization and flight testing.
I am thinking that this would either be integrated with the package a developer will release in a mod, so everyone can turn this feature on in the script, or it could be used just for development. For that matter, you could export the FBX package without the actual aircraft mesh for dev, and use instead a flat plane plan-view orthographic representation of the aircraft, modeling each surface as a separate plane, including wing sections, control surfaces, flaps, etc. This would allow for clean flight testing.
I will eventually develop a GUI plugin for OT that displays critical flight data in a nice looking panel using html layout and js.
I'm thinking it would also be good to display the position of the datum relative to AC/CG.
I would also like to make the vectors increase in length relative to magnitude of the vector norm (0-1), and flip inverse direction if negative. For example the lift surface vectors for a wing tip stall would flip to negative magnitude, visualizing stall propagation.
I was thinking it might be possible to make a boundary layer separation (BLS) visualization, like they use on real aircraft during flight testing by attaching many strings to the upper surface of the wing. During normal flight the string strips will lay flat along the wing, but during stall when the boundary layers separates they will flap and wave around. I am already looking into modeling BLS in the FDM.
(http://upload.wikimedia.org/wikipedia/commons/8/8c/Wing_with_minitufts.jpg)
And of course I will add all control surfaces and other forces/centers/vectors for different types of aircraft as necessary for different types of aircraft; airplanes, helicopters/VTOL, spacecraft, rockets, etc.
I understand there are a number of other efforts underway towards something like this, (AH-DG?), but I don't know what stage these projects are at and when they would be ready.
Look forward to feedback, suggestions for improvements, and additional features!
Simon, are there any features you need specifically?
Regards,
Uriah
Red = Thrust vector (inverted normal)
Blue = Aerodynamic Center of Lift and Lift Vector
Violet = Lift surface vectors (could be added to control surfaces as well)
Black = Center of Gravity and Gravity Vector
Yellow = Velocity Vector
Green = Longitudinal axis
Orange = Drag Vector
F-22, low AoA, landing approach glide slope
(http://i.imgur.com/5URu3i5.jpg)
(http://i.imgur.com/vfEwSq7.jpg)
F-22, high AoA, high pitch angle, high thrust vector angle
(http://i.imgur.com/RUSc5dS.jpg)
-
(http://i.imgur.com/VE7MLZr.jpg)
(http://i.imgur.com/3QiCoqq.jpg)
(http://i.imgur.com/2oOZIz5.jpg)
-
I understand there are a number of other efforts underway towards something like this, (AH-DG?), but I don't know what stage these projects are at and when they would be ready.
Hello Uriah,
mine in english knowledge are limited, but still what would like you to know?
Gruß,
Stefan
-
Wasn't there kind of an auto aircraft toolkit from the JSBSIM designer? I think he never finished it, or something?
-
Hi Stefan,
Well currently there is not many good tools for building flight models using more advanced methods, which are possible using jsb.
As HiFlyer mentions, there was supposed to be a toolkit, but it never made it very far that I know of.
I know that SilentEagle and yourself have mentioned building a better dev tool for jsbsim. Any progress on anything that would be available in the near future? We are in desperate need of better tools..
This toolkit I am building can be used directly in OT, which is a plus side, but I am interested in external tools for generating the FDM from a far more advanced configuration than Aeromatic.
I am working with DATCOM currently, but it is limited.
What are the best current methods for creating high fidelity model in jsbsim in your opinion? I mean, if you want to get extremely accurate results.
Thanks.
Best regards,
Uriah
-
Oh boy......
What's needed is the creation of the 'front end' of the flight model, which is nothing more than weight, derived from volume of airframe parts...And distances...
At present I have to do this the hard way... Which is what I'm soon to show for the cesna.
-
This is top notch, Uriah!!!
"taking measurements directly from the 3d model and real world data"
I have been modeling the tail rotor for the 64A.
(https://scontent-dfw.xx.fbcdn.net/hphotos-xpf1/v/t1.0-9/11067648_1444840852474561_8980740579933430880_n.jpg?oh=ff32693d7089fc854d87b6a3074fdf7f&oe=557C109A)
What should I take in consideration so this work with your toolkit. Just the length of the blades? The thick?
It is hard to model accurately without ortogonal references, and/or measures. If something is wrong in the 3d model, will be possible to override the values?
-
Hi there,
Sorry, I will have to clarify what I meant.
You will take measurements from the 3d model that are not available from real world data sources. However rotor aerodynamics need to be computed from airfoil and angle of attack lift coefficients. For instance, you may have the wing area and aerodynamic center from some arbitrary aircraft's technical data, but in order to break the wing into wing-span sections for the FDM you will need to calculate each area and aerodynamic center from the 3d model by taking coordinate positions and using a little geometry/trigonometry.
So for your model you can calculate the necessary rotor section areas and aerodynamic centers for each section. In the FDM these AC vectors for all the lift area sections will be summed to the Aerodynic Center for the aircraft. You be able to define the fuselage and control surface geometry similarly.
By visualizing the centers and vectors in real time, you will be able to detect flaws in the FDM more easily.
You don't need to change your model at all, and I will post all of the necessary files to configure the visualization toolkit for your aircraft.
Regards,
Uriah
-
...I know that SilentEagle and yourself have mentioned building a better dev tool for jsbsim. Any progress on anything that would be available in the near future? We are in desperate need of better tools...
Hello Uriah,
I do not know you whether I properly understands, but with team JSBSim I have not dealt.I cannot say about JSBSim at the moment a lot, because I have not dealt yet with it thoroughly.
Everything what I have one day said was that we would like to implement a ballistics module with pleasure in OT. In our small team one gives a lot and very well about ballistics understands and has developed a ballistic calculator. And because we build an exact and correct 3D-model (appendix original Fw190 work aircraft construction drawings) we would like to have of course a correct ballistics to the Fw190 arming.
This is everything what I in addition has to say. :)
Gruß,
Stefan
-
Hi Stefan,
Are you implementing your ballistics calculator in jsbsim or as a .dll plugin? That sounds awesome!
Regards,
Uriah
Here are some screens of what a helicopter visualization of forces and vectors.
I plan to add a display element for moments, like rotor rotation direction and magnitude, and pitch, roll and yaw angular direction and rate.
(http://i.imgur.com/KsDVZAy.jpg)
(http://i.imgur.com/SDWLSZ6.jpg)
(http://i.imgur.com/YDfqAoi.jpg)
(http://i.imgur.com/cAoZXVA.jpg)
-
I myself would have had it with pleasure as a module. At the moment our problem exists to find out how one synchronises it with airplane movement in the world surroundings.
To specify calculation frequency physical "cursor" of Outerra of coordinates of a purpose bullet, change of differential corners (in High-speed System of Coordinates) mismatches of the direction of the movement FO (fly object) with a construction axis. It is necessary for definition of a task for "depth" of ballistic calculation.
::)
-
Thanks, Uriah! I get it.
-
I myself would have had it with pleasure as a module. At the moment our problem exists to find out how one synchronises it with airplane movement in the world surroundings.
To specify calculation frequency physical "cursor" of Outerra of coordinates of a purpose bullet, change of differential corners (in High-speed System of Coordinates) mismatches of the direction of the movement FO (fly object) with a construction axis. It is necessary for definition of a task for "depth" of ballistic calculation.
::)
By "synchronizes it", what do you mean? Do you need to attach the ballistic projectile to the aircraft and "fire" the projectile from the aircraft/platform? If so, that is a general problem I've been trying to solve in OT.
Regards,
Uriah
-
By "synchronizes it", what do you mean? Do you need to attach the ballistic projectile to the aircraft and "fire" the projectile from the aircraft/platform? If so, that is a general problem I've been trying to solve in OT...
precisely.... + in addition must be exactly tuned the flight (or place) by airplane-A to airplane-B. That is to calculate the exact coordinates of both airplanes around synchronize time (in Hz). Otherwise one will shoot and meet nothing.
Thus says me to my ballistics engineer... :)
-
And recoil forces ... AH-DG ... dont forget those ...
-
And recoil forces ... AH-DG ... dont forget those ...
ja ja, but we want to this subject not further with the ballistics let to flood... ;)
-
Guys, I'm sure cameni has said the ballistics both firing and hit box detection is solved by them..It's just a waiting game for us till implemented.
Recoiled is a simple external force added to the airframe, see the bottom of the flight model front end I posted up yesterday.
And synchronisation is overrated.... If the shooter aims and hits the plane then it's hit, and the damaged plane is then passed ballistics data for it to calculate probability and damage..