How to create a circular orbit and an angry bird space like orbit ?

Hi there,

Regarding your first problem, it seems that you have forgotten to apply the velocity perpendicular to the distance vector. Rather it seems like you are applying it in the direction of the attracting body. You do this by making the normalized vector perpendicular: n(x, y) -> n(y, -x) or n(-y, x) depending on wether you want a clockwise or ccw orbit.

Regarding the second problem, angry birds space has very little to do with actual physics, so there can't be a definitive answer. Can you describe the desired effect in a little more detail? (Haven't played the game much).

Cheers,

Mike

Well if youve got it attracting to the planet you just need to go around it now. Simplest way would be to just apply a force to move perpendicular to the vector from the object to the planet.

vec2 dirToPlant = normalize(planetPos - objectPos);

side = vec2(-dirToPlant.y, dirToPlant.x);

object.AddForce(side * someForce); // <-- recalc side and do this each frame while inside the attracting region

For the 'orbital decay' due to atmosphere (2nd problem), you can just apply some drag if the object is in the atmosphere.

Calculate the resistance force and apply it in the direction opposite to the velocity vector (as in slow down using that force)

IIRC

drag = objectCrossSectionalArea * objectSpeed^2 * airDensity * someConstant

where airDensity is a function of distance from planet center (or surface, doesnt really matter), see wiki article (seems complicated, so maybe you can just try a few simple functions like constant, linear, exponential etc. and see what works best).

You might want to not apply atmospheric effects above a certain altitude (define some cut-off point), so anything above that altitude wont be slowed down due to atmospheric drag.

A simple model would be constant air density + cut-off at some height (like in that angry birds pic, you could draw similar 'atmosphere circle' with constant air density inside and no air outside), which might be how angry birds does it.

Thank's for every one on their answers, actually i've forgot to apply a vector prependicular to dist vector, But when i did that, i got another problem :

The orbiting body is going to far while moving in a circular way, but i don't know why , here's my code again :

dist = new Vector2 (Planet.renderer.bounds.center.x - renderer.bounds.center.x, Planet.renderer.bounds.center.y - renderer.bounds.center.y).normalized;
r = dist.magnitude;
Vector2 t = new Vector2 (-dist.x,dist.y);
		
float f = Mathf.Sqrt ((5.00f * (mPlanet + mBody)) / r);
rigidbody2D.velocity = t * f ;

Dist is the normalized distance vector, t is the prependicular vector to the dist one, f is the velocity which is calculated from wikipedia : http://en.wikipedia.org/wiki/Circular_orbit#Velocity

Please tell me where i'am wrong ?

thank's

r always has the value of one, because you normalized dist :-) Hence the equation applies the velocity needed for a circular orbit one unit away from mass center.

Maybe you should reform your celestial bodies movement.

Quote

V = sqrt(G * ((m1 + m2 ) / r))

This formula is the classic real world one, V is the velocity of the game object .

There exists no formula for speed upon absolute values in general (speed is a property), this is a deviation of newton formula for force of gravitation between two objects at a moment in space, which is F=G*((m1*m2)/(r*r)), and speed at a time equals V=V_o+(F/m)*t where V_o is current speed vector (tendence) of attracted object, F is gravity force it is attracted by, m is the object's mass and t is time step that F takes effect - until you compute new F vector again pointing to new direction since it sticks to mass center while object has moved (the same also applies on the other object since gravitational force applies on both of the objects, but you can omit the second object if its mass is extremly larger than of the first object)

If there are also other attractors acting on the object, you simply only add also those other forces to the orginal speed of object, and after time step recompute them (vector variables are italic):

V=V_o+(F₁/m)*t+(F₂/m)*t.....+(F_i/m)*t

I would reform your code like this:

dist = new Vector2 (Planet.renderer.bounds.center.x - renderer.bounds.center.x, Planet.renderer.bounds.center.y - renderer.bounds.center.y)/*.normalized*/;

float rdist=dist.magnitude;

//r = dist.magnitude; beware, you have normalized previous subtraction, this would be just equal to 1.0, not real distance

dist=dist.normalized;

//Vector2 t = new Vector2 (-dist.x,dist.y);

Vector2 t = new Vector2 (dist.x,dist.y);

float forcesize=consG*((mPlanet * mBody) / (rdist*rdist));

Vector2 theforce=new Vector2 (t.x*forcesize,t.y*forcesize);

//float f = Mathf.Sqrt ((5.00f * (mPlanet + mBody)) / r);
//rigidbody2D.velocity = t * f ;

rigidbody2D.velocity = rigidbody2D.velocity + theforce*(1.0/mBody)*fixedtimestep ;

establish original rigidbody2D.velocity vector, choose timestep, and estbalish both(all) attracted objects masses

If you will wish for circular orbit, then establish your distance D, masses m1 and m2, and the velocity of circulary orbitting object based upon those values would be

V*t=(unit 2d vector orthogonal to distance)* cos(arcsin( (D-G*m1*m2*1/(r*r)*t) /D ) )*D

pick t as a second likely,

I've tried what everyone said to get a circular orbit, but the gameobject is still going too far to the left while the planet is in the right, I don't know why ?

deltaweb, I'm not certain what you mean, could you make a more detailed description? does the gameobject go into orbit even if it's not a circular one? Can the central planet move by applying force to it, or is it fixed in place?

Here's a quick attempt to help you define the Earth's initial position and initial velocity vector if it were to have a circular orbit path, where the Sun is fixed in position and has no velocity. Try using these parameters first so you'll know if it works as it should, before you start monkeying around with the value of G.

G = 6.67384e11 // Newton's gravitational constant
M = 1.9891e30 // Mass of Sun in kilograms
r = 149597887500 // Earth's semi-major axis -- its average orbit distance in metres
v = sqrt(GM/r) = 29789 // Earth's average orbit speed in metres per second -- NOTE: THIS DIFFERS FROM THE EQUATION THAT YOU'RE USING IN YOUR INITIAL POST.

Sun's fixed position = {0,0,0}
Sun's fixed velocity vector = {0,0,0}
Earth's initial position = {0, 149597887500, 0}
Earth's initial velocity vector = {29789, 0, 0} // orthogonal to the position vector

It's that simple.

If you want to, you can rotate the Earth's initial position and velocity (make sure you rotate both using the same angle) along the z axis before the simulation begins: http://en.wikipedia.org/wiki/Rotation_matrix If you need help with this, let us know. Hint: you can just use the 2D rotation since the orbit path will lie along the xy plane http://en.wikipedia.org/wiki/Rotation_matrix#In_two_dimensions

If you're still stuck, then you're probably doing something wrong when it comes to how you're using a time step (assuming you're calculating acceleration correctly). Check out gaffer's Fix Your Timestep if you haven't already: http://gafferongames.com/game-physics/fix-your-timestep/ and also see the integration basics tutorital http://gafferongames.com/game-physics/integration-basics/ If you're still stuck after reading those two gaffer articles then upload your code into a zip file and post it here.

- Shawn

P.S. For more information, the following link discusses elliptical orbits:

http://www.gamedev.net/topic/447624-calculating-an-initial-velocity-for-desired-planetary-orbit/#entry3961958

So we're clear, the eccentricity parameter is set to 0 for circular orbits.

Oh, thank's for everyone who answered, i've found a way to do that by playing with rotation and gameobjects, For anyone who would like to know how i did it :

1. Create an empty gameobject and set it postions to the planet pos
2. make your orbited gameobject a child of this empty object .
3. Rotate the empty game object with a Z angle ( 2D way ) .

This can only be done using unity.

Once again thank's for everyone, i've learned a lot .

Oh, thank's for everyone who answered, i've found a way to do that by playing with rotation and gameobjects, For anyone who would like to know how i did it :

1. Create an empty gameobject and set it postions to the planet pos
2. make your orbited gameobject a child of this empty object .
3. Rotate the empty game object with a Z angle ( 2D way ) .

This can only be done using unity.

Once again thank's for everyone, i've learned a lot .

NOt realy, you cane establish excenrtig behaviour upon those values (super acceleratred black hole horizont objects/ or, objects that escape observable universe in microseconds). Simulations of newton (no time) law are interesting. Though, this law seems to truly express tendence of attractions if no time is considered towards functional universe

If you are after silly circle orbiting stuff, you can just use sine and cosine relations towards distance.

A(z,y)= (sin(x),cos(x)) = position...... what is a unit circle still a demanding analisis explenational group relation

Believe me deltaweb, it's a much more rewarding experience to actually learn the theory, implement it yourself and see it work than just hack something toghether that only looks real. I should have seen this coming, but I overlooked the angry birds warning. :-)

Okay, here's the full code that i'am using to implement the circular orbit from the physics side :

using UnityEngine;
using System.Collections;

public class Orbit : MonoBehaviour {
	
	Vector2 dist = new Vector2 ();
	Vector2 tdist;
	float r;
	float mBody;
	float mPlanet;
	public GameObject Planet;
	public float G;





	void Start () {

		mPlanet = Planet.rigidbody2D.mass;
		mBody = rigidbody2D.mass;

	}

	
	void FixedUpdate () {
		circularOrbit(); 

	}

	void applyGravity(){
		dist = new Vector2 (Planet.renderer.bounds.center.x - renderer.bounds.center.x, Planet.renderer.bounds.center.y - renderer.bounds.center.y);
		r = dist.magnitude;

		float force =  G * mPlanet * mBody / (r * r);
		rigidbody2D.AddForce (dist.normalized * force);


	}
	void circularOrbit(){
		// Get the distance between the two object centers .
		dist = new Vector2 (Planet.renderer.bounds.center.x - renderer.bounds.center.x, Planet.renderer.bounds.center.y - renderer.bounds.center.y);
		r = dist.magnitude;
		tdist = new Vector2 (dist.x, -dist.y).normalized; // 2D vector prependicular to the dist vector .
		float force = Mathf.Sqrt (G * ((mBody + mPlanet) / r)); // Calculate the velocity .
		rigidbody2D.velocity =  tdist * force;


	}
}

The important part is in the circular Orbit function, and here is a small gif of what's hapenning :

http://makeagif.com/i/-JxBG8

As you can see the monster is not orbited by the planet, here are some infos :

• Monster mass : 1 unit;
• Planet mass : 5 unit ;
• Gravity of the monster is off ;
• G is 1, but the probleme stay the same for all value that i've tested : 0.1 , 0.01 , 1 , 5 , 20....

All i want to do is to be able to make the monster get in a circular orbit, for the angry bird part, this is a whole another story, that i will work on once i get the circular orbit working .

Best of luck.

I've solved the issue : my perpendicular vector was (dist.x,-dist.y) where it should be (-dist.y, dist.x), more here : http://stackoverflow.com/questions/29704731/circular-orbit-formula-is-not-working-in-unity-3d/29704896#29704896

I'm not sure why your velocity code works.

Normally you would calculate the acceleration (where force F = G*M*m / r^2 = m*acceleration ---> acceleration = F / m = G*M / r^2) and then integrate that into the velocity.

ie:

velocity' = velocity + acceleration*(unit length vector pointing from gravitated object toward gravitating object)*dt

position' = position + velocity*dt

...

where dt is your time step size.

This was covered in the gaffer Euler integration tutorial... but I get the distinct feeling that you aren't interested. :)

Well i think that it's working, because the physics engine take care of all the other stuff, like the time step .

What I don't get is why this works: float force = Mathf.Sqrt (G * ((mBody + mPlanet) / r));

Have you read up on the subject of dimensional analysis?

http://en.wikipedia.org/wiki/Dimensional_analysis#The_fundamental_physical_constants

http://www.efm.leeds.ac.uk/CIVE/CIVE1400/Section5/dimensional_analysis.htm

For instance, let's analyze the equation for circular orbit velocity v = sqrt(GM/r)

G has dimensions of Length^3 Mass^-1 Time^-2

M has dimensions of Mass^1

r has dimensions of Length^1

GM has dimensions of Length^3 Mass^-1 Time^-2 Mass^1

GM/r has dimensions of Length^3 Mass^-1 Time^-2 Mass^1 Length^-1 which simplifies to Length^2 Time^-2

v = sqrt(GM/r) has dimensions of Length^1 Time^-1 (ie. metres per second, which stands to reason).

The problem that I have with your equation force = sqrt(G * ((mBody + mPlanet) / r)) is that it has dimensions of Length^2 Time^-2, which is not the dimensions of force (which is Mass Length Time^-2) -- in other words, the equation force = Mathf.Sqrt (G * ((mBody + mPlanet) / r)); makes no physical sense. I invite you to double-check my work, in case I made an error during the analysis of your equation. Plus, the practice wouldn't kill you. :)