Personal tools
You are here: Home Forum Extension of the platform Suggestions of improvements

Suggestions of improvements

Up to Extension of the platform
  • Suggestions of improvements

    Posted by Joël Cugnoni at September 09. 2005


    As I am progressing in my discovery of the Salomé Platform, i have noticed some points that could be improved (from version 2.2) in order make the user's life a bit easier:

    A MESH hypothesis to define a seed of the element size at a vertex could be a very positive addition, especially if another 1D hypothesis is added to interpolate the element size on an edge from the vertex seeds. In this way, it would be very easy to define a mesh refinement even in 3D with a large number of volumes (this option is the basics of GMSH). This could be done for example by implementing a full set of hypotheses based on GMSH: Vertex Seeds, 1D, 2D triangles and 3D tetrahedrons.

    Also a meshing tool to convert from linear to quadratic finite elements could be a simple way to add a first support of the quadratic finite elements (especially for 3D tetra!!) because these quadratic finite elements are of primary importance for mechanical stress analysis. The support of quadratic elements should be also added to Postpro in order to visualize nodal data as a surface map.

    Regarding Postpro, the visualization of tensorial data is not working properly, only 3 components can be selected and the displayed results are "random".

    Another improvement could be to allow to define a Local Hypothesis on several geometric entities in a single step in order to quickly create a refinement in complex geometries with a lot of edges for example.

    Finally, the entity selection could be simplified by first adding a toolbar to define an entity selection filter (by type: vertex, edge, face, volume, compound, node, element) and also by enabling the selection of sub-objects of an entity (for example select a vertex from a 3D compount).

    I think that with these additions, the ease of use of Salome could be compared with commercial software like for example Abaqus CAE that is in my opinion not so far from Salomé's approach.

    Just for information, I am currently developping a LiveDVD Linux distribution that contains Salomé & Code Aster and I am also writing several tutorials on the use of these two great sofwares in the scope of finite element analysis of mechanical structures.

    Thanks a lot for your tremendous work.

    Best regards.

Powered by Ploneboard
Document Actions