[aspect-devel] Discrepancy in L2 Error of Velocity between deal.II and ParaView
heister at clemson.edu
Tue Oct 25 06:04:49 PDT 2016
> Why is there this discrepancy in error between the dealii function, VectorTools::integrate_difference and the discrete L2 norm computed through Paraview? Is this at all related to the geometry of the domain?
Visualization output is
1. using a linear representation of the quadratic FE space for the velocity
2. not representing curved faces using higher order mappings (manifold
descriptions) but linear interpolated "straight" edges.
Finally, I doubt paraview is using quadratures as accurate as we are
using. Of course you can not expect the results to be the same.
More information about the Aspect-devel