Sign in to follow this  

Integrating angular velocity ( Corioles term )

This topic is 4359 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

The Newton-Euler equations in a rigid body simualtion have the following form: M * dv/dt = f_ext I * dw/dt = t_ext - [w]* I * w The term [w] * I * w is sometimes called Corioles term in the literature. IIRC this term is responsible for making a simple semi-implicit Euler instable - especially for long and thin objects and larger time-steps. 1) How can I compute an estimation of the max timestep I can take given the local intertia tensor of an object? 2) Can I invert this and compute a max ratio of the min element to the max element of the inertia tensor in order to detect critical objects? Are there any rules of thumb? 3) What would be a good rescue operation after accidently making a critical timestep for such an object? 4) Are there any other factors I need t take into account and that I overlooked here when dealing with long and thin objects in a velocity constraint LCP simulator? These critical objects (long and thin) are normaly not part of an articulation, so I am looking for some kind of heuristic to deal with them. I added some kind of clamping, where I define the minimum of an element of inertia tensor to be at least some amount (e.g. 5%) of the maximum element. You can't get save here without making the simulation unrealistic. Another solution would be to clamp the applied torque before integration, but which upper bound should I chose. I am bound to a semi-implicit Euler because of the velocity constraint LCP formulation I use, so I can't use Verlet, RK4 or implicit Euler which might help here... Any help is greatly appreciated! Regards, -Dirk

Share this post


Link to post
Share on other sites

This topic is 4359 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.

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

Sign in to follow this