• Advertisement
Sign in to follow this  

problem with rotational dynamics

This topic is 1804 days old which is more than the 365 day threshold we allow for new replies. Please post a new topic.

If you intended to correct an error in the post then please contact us.

Recommended Posts

Hi everyone. I'm doing my class project and i'm having troubles with Euler integration (witch seems to be the basic)...
The main loop:
 
 
        torque = position.cross(force);
        
        momentum += dt * force;
        //velocity = momentum * (1.0f / mass);
 
        position += velocity * dt;
 
        ang_momentum += torque * dt;
        m_inertia_inverse = MakeMatrixFromQuaternion(orientation).transpose() * g_inertia * MakeMatrixFromQuaternion(orientation);
        ang_vel = multvm(ang_momentum, m_inertia_inverse); //multiplies a vector by a matrix
 
        orientation += (Quaternion(0, ang_vel.x, ang_vel.y, ang_vel.z) * orientation) * (0.5f * dt);
        orientation.normalize();

 

"g_inertia" is the following matrix3

 

Ix = Iy = Iz = 0.4f * mass * Mathematics::sqrt(radius);
 
        inertia.m11 = Ix;            inertia.m12 = 0;        inertia.m13 = 0;
        inertia.m21 = 0;            inertia.m22 = Iy;        inertia.m23 = 0;

        inertia.m31 = 0;            inertia.m32 = 0;        inertia.m33 = Iz; 

 

and i'm rotating like this:

 

 

orientation.angleAxis(angle, axis); //quaternion to angle axis
    
        glTranslatef(position.x, position.y, position.z);
        glRotatef( angle / Mathematics::pi * 180.0f, axis.x, axis.y, axis.z );

 

Is the code right? i've spend 5 days trying to solve it but i think is a little problem and i can't see

Seems like i'm having trouble with the numbers... they are large, especially with the angular velocity...

Share this post


Link to post
Share on other sites
Advertisement
Sign in to follow this  

  • Advertisement