gravity in a particle system

Programming Questions

nootropic..

in Programming Questions • 2 years ago

i guys i've tried to implement a simple particle system

i've got a Circle class with location, position and acceleration PVector

and this method:

void gravity() {
for(int i=0; i<particles.length; i++) {
Circle body = particles[i];
PVector attractor = new PVector(mouseX, mouseY); //center of attraction
PVector between = PVector.sub( attractor, body.location); //distance vector
float distBetween = between.mag(); //distance value
if(distBetween>=60) { // ###
distBetween=60;
}
between.normalize(); // why?? because i have subtracted?
float force = (50 * body.radius ) / (distBetween*distBetween); // newton gravity formula
between.mult(force); // why not normilizing again?
body.velocity.add(between);
}
}

i don't understand why to normalize,

and i'm gaining - dont know why - a "spring" linke movement:

particles get mouseX and mouseY coordination but oltrepass it and come back in a circular way,

and sometimes (there are ranodm initial velocity) orbit around it

.. but i'm adding gravity vector on velocity, not accellaration (as in spring formula!)

i thought that if gravity is too strong is bad, and so i put the if condition (###):

no more absurde velocity but if i left particles goes on for minutes the never achive mouseX,mouseY..

Replies(2)

torfuspol..

Re: gravity in a particle system

2 years ago

The first thing you need to understand is that you want to calculate a vector which points into the direction of gravity, and whose length is equal to the magnitude of the gravitational force. Now if you normalize a vector, you get a vector pointing in the same direction, but with length 1. This is important, because you don't know how far the particle is from the mouse pointer, and you're mainly interested in the direction. Since the length of the vector is given by the magnitude of gravity, you then multiply the vector by this magnitude. Multiplying a vector with a number gives you a vector pointing into the same direction, but with a length that has been multiplied by that number (if the number is positive).

The reason why you are getting very high velocities is that the gravitational force gets very strong if the particle is very close to the mouse pointer: Then, distBetween is very small, and since force is inversely proportional to its square, force gets huge. So instead of using an upper limit on distBetween like you're doing it now, use a lower limit:

if (distBetween <= 3) {
distBetween = 3;
}

nootropic..

Re: gravity in a particle system

2 years ago

ok thanks, very clear.

ok, lower limit is better but i have one more problem now:

if i put the mouse exactly on the trajectory, on the path of particles..

the trick broke down, velocoty get bad values like:

location: 1198655.0 368343.06

velocity: 787.2558 251.44264

also if i put something like

location.x = constrain(location.x, 0, width);
location.y = constrain(location.y, 0, height);

in render method.. something strange happend:

i follow the particle (i'm testing with only 1 particle) with me mouse,

velocity get high, and the particle stay in the angle of stage..

[EDIT: resolv ed, i put un upper bound limit to float force.. there is a cleaner ways?]

the other thigs, just for curiosity.

now i'm simulating gravitational force with a gravity center.

like planets over the sun..

that's ok

but in the reality i think that every body act some force to every others, like in galaxy

(well, mayebe not to all but to near ones)..

how can i simulate that? with a for loop the apply the formula to all others?

put some algorithm for neighours?

Top Reply