A integration scheme which is works good with less number of elements and fails when we increase the number of elements for the same geometric domain. What could be the reason behind that.
I thought this is maybe related with structured and unstructured mesh. In that case a new scheme may not be good accuracy with unstructured mesh. If it is so how to improve the accuracy of integration for an unstructured mesh.
I'm dealing with 2D structural analysis problem.
Hello,
I got this information (from https://www.pygimli.org/_tutorials_auto/1_basics/plot_6-mesh-quality-inspection.html) : " In the case of unstructured triangular meshes (or tetrahedral meshes in 3D), relatively large and small angles can lead to discretization errors. Large angles can cause interpolation errors, while small angles can lead to ill-conditioned stiffness matrices. " Do you think that your precision is related to small angles ; I mean that for structured 2D meshes, you control the minimum angle value, while you may have kind of degenerate triangles with the unstructured one?
Other point of view: form this paper, you may study which areas cause inaccuracies and try to mesh them in a more structured way.
https://www.numeca.com/readnews/article/619
Regards,
Gabriel
Dear Gabriel,
Thanks for your valuable suggestions.
I'm specifically looking for quad element not for triangular elements and also I don't have any constraints on minimum or maximum angle value for unstructured mesh.
Hi, good idea the use of quad instead of trianguler elements.
Gabriel
- Badges
- Science topic
- Similar topics
- Physical Sciences
- Materials Science
- Materials
- Inorganic Materials
- Ceramics
- Glass