[vtkusers] Streamlines vs. Lagrangian drifters

[Prabhu Ramachandran]

hi,

>>>>> "Berk" == Berk Geveci <berk.geveci at kitware.com> writes:

    Berk> 	1) A fundamental problem in generating particle paths
    Berk> is that vtk has no built-in time support. I am working on
    Berk> this.

<snip>

    Berk> vtk 3.2 (which I hope, will be out soon). I will, in time,
    Berk> implement a 5th order Runge-Kutta with 4th order step-size
    Berk> control.

Sounds very good.  So you mean with the 5th order adaptive RK, instead
of specifying IntegrationStepLength etc. one can specify a precision
and the time steps are taken adaptively?  That would be great.

    Berk> 	3) Although the underlying data object can't,
    Berk> vtkInitialValueProblemSolver can handle time. Therefore, it
    Berk> is a matter of adding that support to the data sets to make
    Berk> vtkStreamer capable of generating streaklines.

    Berk> 	I would appreciate feedback and suggestions on this.

Once you add streakline support path line support should be trivial.

What about the requirement for data.  Will you need a range of vector
data files that you will iterate over?

prabhu