Tilted Forum Project Discussion Community - View Single Post

kel · 07-27-2003, 06:44 PM

Would take a completely different approach, it would simulate the displacement and flow of currents of air over each individual surface. You can treat air as a vector and sorta ray trace it, or you can create a partical system with actual air particles with mass and velocity and implement a concept of high and low pressure. Equations of orientation and position don't even enter into the picture because the particle system "knows" by it's very nature how to translate the helicopter. The lockheed approach requires a supercomputer :-P. I just want to create a believable simulation, I don't care about the simulation being accurate, just believable. My question was how to phrase the problem as a simple mathematical system. It's very doable with the right background. And it's mindless to implement in code once someonce formalizes it into a set of functions.

At every point in time t I need to calculate the position and orientation of the helicopter based on a few dependent variables, namely, rotor speed, rudder speed (the rudder is actually a second rotor, the rudder pedals slow down and speed it it up), and rotor angle. Those are the only dependent variables I want to start out with. This is mostly some complex linear algebra which I am studying right now for this purpose, and some multi-variable calculus. Undergrad level ;-)

The equations that calculate the position and orientation at time t are constant for every set of dependent variables. So everytime the pilot modifies rudder, rotor, or collective the constants in the equations are recalculated. I just want the equations and the method to calculate the constants. Simulating various effects and nuances of h/c is pretty easy by simply tweaking various cases once you have a system like this.

The problem is that all the existing engineering texts just tell the engineer how to predict performance based on certain phsyical features and not how to actually flat out calculate the whole shebang. That's left to the programmer with the supercomputer :-P

07-27-2003, 06:44 PM	#8 (permalink)
kel WARNING: FLAMMABLE Location: Ask Acetylene	A realistic simulation like the one lockheed would use Would take a completely different approach, it would simulate the displacement and flow of currents of air over each individual surface. You can treat air as a vector and sorta ray trace it, or you can create a partical system with actual air particles with mass and velocity and implement a concept of high and low pressure. Equations of orientation and position don't even enter into the picture because the particle system "knows" by it's very nature how to translate the helicopter. The lockheed approach requires a supercomputer :-P. I just want to create a believable simulation, I don't care about the simulation being accurate, just believable. My question was how to phrase the problem as a simple mathematical system. It's very doable with the right background. And it's mindless to implement in code once someonce formalizes it into a set of functions. At every point in time t I need to calculate the position and orientation of the helicopter based on a few dependent variables, namely, rotor speed, rudder speed (the rudder is actually a second rotor, the rudder pedals slow down and speed it it up), and rotor angle. Those are the only dependent variables I want to start out with. This is mostly some complex linear algebra which I am studying right now for this purpose, and some multi-variable calculus. Undergrad level ;-) The equations that calculate the position and orientation at time t are constant for every set of dependent variables. So everytime the pilot modifies rudder, rotor, or collective the constants in the equations are recalculated. I just want the equations and the method to calculate the constants. Simulating various effects and nuances of h/c is pretty easy by simply tweaking various cases once you have a system like this. The problem is that all the existing engineering texts just tell the engineer how to predict performance based on certain phsyical features and not how to actually flat out calculate the whole shebang. That's left to the programmer with the supercomputer :-P __________________ "It better be funny"