Jump to content

High Fidelity Flight Dynamics and Tech/Academic Initial Demonstration


CptSmiley

Recommended Posts

  • Replies 1.1k
  • Created
  • Last Reply

Top Posters In This Topic

Not to knock the OPs efforts, but AFAIK you will have a hard time building a LUT based FM of the same quality for anything else but the Viper, simply because there is nothing of sorts like the NASA report in the public domain for other fighters. It requires a real motherload of data. :)

 

Very nice project though, anything that expands the documentation so people can get into the API more easily is direly needed at this point.

 

Hornet. And I have the data. :D

Link to comment
Share on other sites

Excellent work. Just yesterday I started working on an F-16 flight model for my sim using the same sources you mentioned.

 

I'm not sure whether you are aware but there is an open source implementation of that NASA paper at Caltech:

http://www.cds.caltech.edu/~murray/projects/afosr95-vehicles/models/f16/

 

I used the Java implementation, since I want to use multi-threading to allow full flight models to be used for each aircraft.

 

The NASA model is ok in flight but the implementation has quite severe deficiencies such as:

* Not-a-Number produced at zero airspeed.

* no support for negative airspeed (tailslide!)

* no model for fuel consumption or transfer between tanks.

* no model of mass distribution (important once you add weapons, or stall).

* missing the relationship between pilot command and control surface deflection, as in the flight control system (fortunately there is other material out there for this)

* no model of ground handling

* etc.

 

I don't mean to denigrate your work in any way - it is an awesome achievement to get it into DCS, and so quickly too. All I mean is that the NASA model should be seen as a starting point, but you probably already know this :)

 

Keep up the great work.

 

You make some great points!

 

No simulator is of 100% fidelity, for situations like NaN incidents in an algorithm, those are easily overcome. For mass model distribution for stuff like weapons, that is also easily overcome as using nominal equations for moment of inertia calculations knowing weights and associated positions of stores, even analytical aerodynamic force/moment calculations are reasonable assumptions for decent fidelity.

 

Change in mass distribution due to fuel moving around and consumed gets complicated...do you model sloshing, do you just model it as a point-mass for inertia, all of these you take into concern. Now, is it reasonable to simplify the modeling in this area with little reduction in fidelity...probably, it's all a trade-off.

 

For the flight controls logic, this is available and I am currently working on this. I have currently all the actuator logic working with the transfer functions delays/lags as described in the NASA reports. I also have the leading edge flap modeled per the flight control logic. I just got started working on the other control surface mixing today.

 

The ground reaction model is the hardest of them all. For some reason ED decided that the only force they want to control when you use the external flight model is the weight force. It no longer uses the gear impact information to do it's own internal ground reaction model...that means you have to code it up yourself. I'm not sure how to tackle that yet, but something that needs to be overcome. I wish temporarily I could "enable" the SFM ground reaction modeling for now.

 

All good points and I hope to make this project a "living" project where others can improve with their own analysis and assumptions

"Witness mere F-14s taking off from adjacent flight decks, gracefully canting left and right, afterburners flaming, and there’s something that sweeps you away—or at least it does me. And no amount of knowledge of the potential abuses of carrier task forces can affect the depth of that feeling. It simply speaks to another part of me. It doesn’t want recriminations or politics. It just wants to fly.”

― Carl Sagan

Link to comment
Share on other sites

Yeah just started a couple weeks ago with this after I realized it was possible, I was pumped.

 

I still have a ways to go but I fully intend on providing the code that backs this up :)

No that's not possible it was stated a hundred times in the AFM/SFM discussions that such work takes a year.;)
Link to comment
Share on other sites

For the ground reactions a library like Bullet/JBullet might be a starting point:

http://bulletphysics.org/wordpress/

http://jbullet.advel.cz/

 

Also, there's JSBSim if you want an existing general flight model (not as good as the F-16 model, but will work with other aircraft):

http://jsbsim.sourceforge.net/

 

There is a proposed standard for describing flight models using XML:

http://daveml.org/

Of course, you then need to evaluate the model but at least you can exchange the data (there is an example XML file for the F-16 using the well-known NASA data).

 

This is all great stuff.

 

@aaron886: I'd be very interested to find more flight data for the Hornet. I have the NATOPS manuals with performance charts but would be very interested if you have any other data (eg. inertia matrix coefficients). I did have some FORTRAN code around somewhere but I'm not sure if it was as good as the Viper model.


Edited by Moa
Link to comment
Share on other sites

Hornet. And I have the data. :D

 

And I can back that up. The amount of data this man has on the Hornet is mind-boggling.

i7 7700K | 32GB RAM | GTX 1080Ti | Rift CV1 | TM Warthog | Win 10

 

"There will always be people with a false sense of entitlement.

You can want it, you can ask for it, but you don't automatically deserve it. "

Link to comment
Share on other sites

Just as I have discovered. I've been stuck for months waiting for ground handling support and fuel mass distribution explanation. I know how it should work, but my aircraft weight doesn't seem to reflect the fuel.

 

Yep, I know modeling the ground reaction is going to be an interesting task...I'm concentrating right now on getting the flight model looking good but I hope to develop even a low fidelity ground reaction model just so I can land and take-off.

"Witness mere F-14s taking off from adjacent flight decks, gracefully canting left and right, afterburners flaming, and there’s something that sweeps you away—or at least it does me. And no amount of knowledge of the potential abuses of carrier task forces can affect the depth of that feeling. It simply speaks to another part of me. It doesn’t want recriminations or politics. It just wants to fly.”

― Carl Sagan

Link to comment
Share on other sites

No that's not possible it was stated a hundred times in the AFM/SFM discussions that such work takes a year.;)

He had a massive head-start on that year thanks to NASA doing all the work on the analysis of the aerodynamics. :)

 

Getting said aerodynamics data (lift coefficients, drag profiles) is the meat and potatoes of flight dynamics modelling.

 

Aerodynamic equations are easy to find, and many will give good results. They can be added in a few days (let's be generous and say 2 weeks). It takes yet more work if you want to expand on what you have (i.e. model certain behaviors not accounted for in simpler equations, or unique to your aircraft).

 

After that is the hard task of tuning the flight model so it is believable to fly. That takes a lot of trial and error, and time.

 

I had a very basic flight model running in a matter of a couple of hours - but it wasn't flyable. All it did was fly straight and level.

 

Best regards,

Tango.

Link to comment
Share on other sites

Thinking of what you wrote just now, Tango: "model certain behaviors not accounted for in simpler equations, or unique to your aircraft" => How can be modeled behavior such as hypermanoeuvrability and tricks of the SU-35? (for exemple the double 'steady loop' or 'tail slide' it can perform)

That can be a nightmare to code in an AFM, isn't it?

DCS Wish: Turbulences affecting surrounding aircraft...

[sIGPIC] [/sIGPIC]

Gigabyte GA-Z170-HD3P - Intel Core i5 6600K - 16Gb RAM DDR4-2133 - Gigabyte GeForce GTX 1080 G1 Gaming - 8 Go - 2 x SSD Crucial MX300 - 750 Go RAID0 - Screens: HP OMEN 32'' 2560x1440 + Oculus Rift CV1 - Win 10 - 64bits - TM WARTHOG #889 - Saitek Pro Rudder.

Link to comment
Share on other sites

Not so much a nightmare as requiring a lot of analysis to determine how/why it does what it does, so it can be integrated into the modelling. When it comes to creating computer simulations, there is not one set of equations/data for all situations. You will probably need to switch equations/data on the fly in order to create the desired effect(s). This effectively requires the production of multiple flight models and data sets.

 

Take backwards flight for example. As the NASA models demonstrate, equations that are good for forward flight are useless for backwards flight. It is not simply a case of reversing the flow over the wings. In that situation an entirely new analysis needs to be conducted, requiring new data and equations, to handle reverse flows.

 

Best regards,

Tango.


Edited by Tango
Link to comment
Share on other sites

Yep, I know modeling the ground reaction is going to be an interesting task...I'm concentrating right now on getting the flight model looking good but I hope to develop even a low fidelity ground reaction model just so I can land and take-off.

 

Well from what I know, the basic ground handling should be there, but I don't know how to make it work. Don't think they planned on us making our own.

Link to comment
Share on other sites

Well from what I know, the basic ground handling should be there, but I don't know how to make it work. Don't think they planned on us making our own.

 

Do you know if any 3rd party dev utilizing the AFM has got the DCS internal ground reaction modeling working? I would be very intested to know how this works.

 

Tango, you're exactly right. That is single reason why I chose the F-16 as it was an ideal proof of concept to help learn how to make an AFM and I hope it will be a useful tool for others to learn from as well.

 

There is certainly a ton of tuning required that I call "fudge factors" to make the aircraft match performance charts, or fly how it should it areas outside of the provided data areas. With the right amount of tuning and know-how this can be overcome, but it takes a massive amount of trial and error. It also doesn't help that you can't debug into your code while the sim is running, I think due to DCS's DRM.

 

Like I have said before, this is just a stepping stone for other projects and never intended to share it, but at the same time I saw how much trouble others were having and I hoped we can all help each other in this process...

"Witness mere F-14s taking off from adjacent flight decks, gracefully canting left and right, afterburners flaming, and there’s something that sweeps you away—or at least it does me. And no amount of knowledge of the potential abuses of carrier task forces can affect the depth of that feeling. It simply speaks to another part of me. It doesn’t want recriminations or politics. It just wants to fly.”

― Carl Sagan

Link to comment
Share on other sites

How can be modeled behavior such as hypermanoeuvrability and tricks of the SU-35? (for exemple the double 'steady loop' or 'tail slide' it can perform)

That can be a nightmare to code in an AFM, isn't it?

 

If you have some programming skill then look at ED FM template, thrust vector control is generally possible. Thrust parameters consist of thrust vector length and orientation. Vector length depend on throttle position and altitude. Your task is create algorithm for calculating angles for thrust vector orientation depend on stick and rudders.

Generally it`s possible.

You can do it! :D

Link to comment
Share on other sites

Do you know if any 3rd party dev utilizing the AFM has got the DCS internal ground reaction modeling working? I would be very intested to know how this works.

 

Haven't seen any yet. I would assume you'll have to write your own reaction model and calculate a normal force for each tire when it impacts the ground. Then you can tweak elasticity and friction/grip.

Link to comment
Share on other sites

The biggest point is to be clear on the equations used, and their limitations. The rest of the source code should provide clarity on the order each function is called, and exactly what it is doing at each step. It might be worth creating a flowchart of the structure.

 

Best regards,

Tango.

Link to comment
Share on other sites

maybe ED can Donate their F-16 External Model so we can make something happen... it's been 2 years since those renders, the Max files are prolly degrading on some geeks flash drive..

Windows 10 Pro, Ryzen 2700X @ 4.6Ghz, 32GB DDR4-3200 GSkill (F4-3200C16D-16GTZR x2),

ASRock X470 Taichi Ultimate, XFX RX6800XT Merc 310 (RX-68XTALFD9)

3x ASUS VS248HP + Oculus HMD, Thrustmaster Warthog HOTAS + MFDs

Link to comment
Share on other sites

This does not mean it won't be available for you to fly with!

 

Could you please explain what you mean by that? Will you guys create a flyable F16, with all its systems?( radar, target pod, weapons etc)?

 

Would love a DCS F16 that can replace/ supplement to falcon 4.0BMS

 

Either way, keep up the good work and good luck

"When once you have tasted flight, you will forever walk the earth with your eyes

turned skyward for there you have been, and there you will always long to return"

Neck.png

132nd Public documents - Want to apply? - 132nd Youtube channel - 132nd Discord Channel

Link to comment
Share on other sites

Could you please explain what you mean by that? Will you guys create a flyable F16, with all its systems?( radar, target pod, weapons etc)?

 

Would love a DCS F16 that can replace/ supplement to falcon 4.0BMS

 

Either way, keep up the good work and good luck

 

Nope, my skill set nor free time do not allow for aircraft systems and weapons modeling.

 

My goal is to create a working advanced flight model example for other to work from.

 

Like Tango said, I am working to organize it in a way that make sense, document the flow, comment the units to be used. As it stand now it does not even have a cockpit...

 

This does not prevent others from taking the model and doing as they wish and you can certainly fly it as is when I publish it.

 

I hope this helps!

"Witness mere F-14s taking off from adjacent flight decks, gracefully canting left and right, afterburners flaming, and there’s something that sweeps you away—or at least it does me. And no amount of knowledge of the potential abuses of carrier task forces can affect the depth of that feeling. It simply speaks to another part of me. It doesn’t want recriminations or politics. It just wants to fly.”

― Carl Sagan

Link to comment
Share on other sites

i dont think NASA has test data for weapons either... lol

Windows 10 Pro, Ryzen 2700X @ 4.6Ghz, 32GB DDR4-3200 GSkill (F4-3200C16D-16GTZR x2),

ASRock X470 Taichi Ultimate, XFX RX6800XT Merc 310 (RX-68XTALFD9)

3x ASUS VS248HP + Oculus HMD, Thrustmaster Warthog HOTAS + MFDs

Link to comment
Share on other sites

/Thread cleaned of Rule 1.2 infringements.

Let's not let that happen again Gentlemen.

Ta

  • Like 1

Novice or Veteran looking for an alternative MP career?

Click me to commence your Journey of Pillage and Plunder!

[sIGPIC][/sIGPIC]

'....And when I get to Heaven, to St Peter I will tell....

One more Soldier reporting Sir, I've served my time in Hell......'

Link to comment
Share on other sites

Progress Update:

- Leading edge flap control law logic complete

- Actuator logic for all surfaces complete

 

To do this weekend:

- Higher level control laws

 

We'll be testing it internally and should have another video up soon of the improvements...

"Witness mere F-14s taking off from adjacent flight decks, gracefully canting left and right, afterburners flaming, and there’s something that sweeps you away—or at least it does me. And no amount of knowledge of the potential abuses of carrier task forces can affect the depth of that feeling. It simply speaks to another part of me. It doesn’t want recriminations or politics. It just wants to fly.”

― Carl Sagan

Link to comment
Share on other sites

  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...