Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Koty]

I guess I'd make living as a proffesional necromancer... ^^;

Anyways, finally had a look into the class-files (decompiled without an error, luckily...) and FM...

Issues found:
It uses jet code to compute "thrust" (lift caused by rotor), this should be redone (includes touching the engine itself) into own set of things, specially for helicopters. That, because a lot of things cannot be solved by current methods. Now this comes just from basic observations, so no quantification (yet), but I should be able to find some sources that actually explain it in numbers.

Let's start with static Mi-24.

Helicopters have two controls (not touching the "joystick" yet) - throttle and "colective". In current model, you have only throttle, no colective. Colective acts as propeller pitch setting. You select RPM by throttle - and fly using propeller pitch.

The more pitch you have, the better you climb and the more lift (thrust) your rotor provides. 0 pitch means flat rotor, 100% pitch means max AoA of blades. This has to be done in the Engine.


Next - moving Mi-24.

In the FM, the wings and stabilizators have to be set, it helps the heli hold the direction at higher speeds. Also, when going sideways, it means you have to input "rudder".

The rotor itself has to be providing lift as well, when you pitch up, it should work as a wing - now this could be solved by the stabilizator surfaces (pic below)


How it (roughly) works, here:
(I am not realy sure where I was going with this, lost track multiple times, I hope it makes some sort of sense.)

[Vampire_pilot]

I have no idea what state of development the Mi-24 is in so I just wanted to ask about AI capable FM.

As far as I know there is no AI capable version yet? I am using it from BAT and that version is not flyable by AI, I presume this is the latest version.

[Koty]

Ahem.

AI version will be unnecessary, once this is finished. As it, as the original plan was, will pack own AI.

That however needs finished flight model.

That said, I have picked up the work.
What I will need to do now is completely turn the two engines into jets. (Currently, they work to lift the helicopter, I want to do the lift in code exclusively.)

Now working on drag values, just polishing out values now.

[Vampire_pilot]

oh, sounds interesting!

[Koty]

mkay, thrust from the two engines transfered completely to exhaust thrust

building rotor FM started

currently present:
-lift of main rotor
-drag momentum of main rotor rotation
-tail rotor steering momentum
-tail rotor drag momentum (it pitches nose, effect is minimal, but hey, why not include it)

and of course, the code is now extremely messy, as I kept the old one in form of notes and the new one is not really cleanly written, which does not help... 

[Koty]

Currently present all important forces, now polishing values and my math errors.

After this is polished, next will be aerodynamic moments caused by different flow relative speed over blades as the heli is moving.

In final FM, I want to also include effects such as Corriolis moment etc.

[Koty]

Turns out, it is very unstable, once you initiate a rotation, there is no momentum to stop the rotation. IRL there is a momentum caused by center of mass(gravity) being offset to the rotor center (now it acts as if it was identical point), both vertically and (in Mi-24's case at least) horizontally.

Tried adding this momentum, but the thing started doing all kinds of weird things. I'll concentrate on this for now.

EDIT: Math.sin() (literaly just sinus function) expects radians... I used degrees... I feel dumb. And tired.

[Vampire_pilot]

reads like a mad scientist's diary:P

great initiative

[BT~wasted]

I am very glad that work on my project is going on! I had a ready cockpit for Mi-24 and code for missile guidance like IRL written. I need to find it and hand it to someone who wil lcontinue working on it.

[Koty]

Great to hear Wasted!

However, I have to let you know that I currently struggle with the mathematical side of the project... cannot calculate rotational speed (rate of rotation) for pitch and yawing.

Well... I can get the absolute value of it, cannot determine the direction Which gives some really funny behavior.


EDIT: Now I got the direction right... but something else is off... this will be a long day.

EDIT2: Just why do I have to manually reset all the forces at the end of update(); when it should get overwritten since the code repeats over and over?

[Koty]

You can guess where I made a mistake 
Hint, look at the blue part...

[BT~wasted]

So how it looks now in game? I remember that one of the biggest problems were to simulate separate throttle and collective. I assume that you are still using throttle as your main control?