Kangaroo

3008 members

Description

THIS FORUM IS NO LONGER ACTIVE. PLEASE POST ANY NEW QUESTIONS OR DISCUSSION ON:

https://discourse.mcneel.com/c/grasshopper/kangaroo

The discussions here are preserved for reference, but new questions posted here are likely to go unanswered.

Kangaroo is a Live Physics engine for interactive simulation, optimization and form-finding directly within Grasshopper.

Comments Discussions Members

Damping/ drag

by Nick Cole

Dec 19, 2010

Daniel

 

Could you give a brief explanation of how the damping coefficient and drag coefficient are implemented in Kangaroo? I've found that increasing the damping or drag obviously lets the model settle down quicker but can significantly alter the final result in some cases (position, extension, symmetry). For very simple models, where i know what the answer is meant to be, reducing the damping to as small as possible gets closer to the correct result but never quite gets there. Related to this i've found that the forces in/ forces out never seem to quite sum up either. Also if you set both the damping and drag to 0 it still settles down. Is there an inbuilt minimum damping?

 

Sorry for all the questions.

Cheers

Nick 

  • up

    Daniel Piker

    Drag applies a force to every particle opposite to its velocity and proportional to its speed multiplied by the value of drag.

     

    Spring damping is similar except it only acts in the direction of the spring.

     

    There is no inbuilt minimum damping, and things should keep oscillating indefinitely when there is no drag or damping.

    There were some errors in Kangaroo 0.02 with the way forces were added up which have been fixed in the new version. This may be what is causing these errors.

    If you could post one of your files I'll take a look and see if this is the case.

    Dec 19, 2010

    • up

      Nick Cole

      Daniel

       

      I am on 0.02 so that may be the problem. See the attached file. You'll find that as you vary the damping the final values change. In the second example it is particularly noticable. The node settles in the centre when you set the lengths to 5 but with any situation that leaves forces in the springs the  node is always off centre. 

      Dec 20, 2010

      • up

        Nick Cole

        Daniel

        What was the verdict on this one? Is it something that is corrected in the latest version of Kangaroo or am i doing something wrong in the model?

        Dec 24, 2010