Understanding Elastic Collisions in 2D: Exploring Velocity and Rotation

[Isawyou0]

Hi all!
I'm developping a program, I wish if somebody helps me since I'm not good in physics, we know that:
v'1=v1-(j / m1)*n
v'1=v2+(j / m2)*n
1) when we use negative and positive sign? if we change them, there is error sometimes.
2)In 2d, do I have to calculate vx, and vy separately, what about n normal vector? should I do the same or use the magnitude?
And here we talk about relative velocity of j=-(1+e)*rV/1/m1+1/m2,calculate just magnitude or vx and vy, but doe'snt going to work in rotation:
w'=w+j/i*r*n;
j=j=-(1+e)*rV/(1/m1+1/m2)*n^2*r^2/I*n^2*r^2/I;there will be no wx and wy for vx and vy, I guess magnitude here?
3) Can somebody demonstrate to me how the heck do we passed from:
Vf = Vi+ 2m(vi-Vi)/(m+M);to:
vf=vi-(j / m1)*n;also for rotation;
4) how to add friction:(it's should be negative to the velocity, and f'=uf, with ucoeficient of friction), a dot product with perpendicular to n normalised + j ?
I'm lost here please help; I will appreciate :)

 

[mfb]

If you use variables, it would be useful to define and explain them.

 

[rowny]

Better explain the conditions.

 

[Isawyou0]

okay! it will be more complicated, just want to know how to clculate ellastic colision, and rotation

 

[sophiecentaur]

okay! it will be more complicated, just want to know how to clculate ellastic colision, and rotation

You will need to make the effort to explain what you want in better detail or no one can answer you.
Did you try a textbook?

 

[Isawyou0]

vx = vx + ((-1 * (1 + 1) * ((vx + teta *r) * nx+ (vy + teta * r) * ny)) / ((1 / m1+ 1 / m2) * ((ny* ny) + (nx* nx)))) * nx;
vy = vy + ((-1 * (1 + 1) * ((ix + teta * r) * nx+ vy + teta * r) * ny)) / ((1 / m1+ 1 / m2) * ((ny* ny) + (nx* nx)))) * ny;

nx &&ny: are normal of collison.
this code work wrong, why?

 

[Isawyou0]

it works finally! just want to know about rotation last n,
w'=w+j/i*r*n;
n magnitude?

 

[Isawyou0]

it's ok for rotation,
teta = teta + (-1 * (1 + 1) * ((ix + teta * (centx - ja) - ibx) * xx1 + (iy + teta * (centy - jb) - iby) * yy1)) / (1 / 1.0f + 1 / 2.0f + ((float)Math.Pow((centx - ja) * xx1 + (centy - jb) * yy1, 2) / In) + ((float)Math.Pow((centbx - ja) * xx1 + (centby - jb) * yy1, 2) / Ina));
tetab = tetab + (-1 * (1 + 1) * ((ibx + tetab * (centx - ja) - ix) * xx1 + (iby + tetab * (centy - jb) - iy) * yy1)) / (1 / 1.0f + 1 / 2.0f + ((float)Math.Pow((centx - ja) * xx1 + (centy - jb) * yy1, 2) / In) + ((float)Math.Pow((centbx - ja) * xx1 + (centby - jb) * yy1, 2) / Ina));

ja and jb point of collision coordinate;
xx1 and yy1 are conatct normal;
and finally In for inertia

just maybe an error here?

 

[mfb]

See above: without much more explanation what you are doing, how all those variables are defined, what you expect as result and what you get it is impossible to follow.

 

[Isawyou0]

Since that inertia is in kg.m^2, how to convert it to pixel(best way)? my rotation still going on the wrong sens

 

[mfb]

Like this, the thread is completely pointless. Feel free to open a new thread if you need help, but then explain what you are doing.