Home › Forum › SOFA › Using SOFA › [SOLVED] Stiffness change when refining mesh
- This topic has 3 replies, 2 voices, and was last updated 8 years, 8 months ago by
Hugo.
-
AuthorPosts
-
24 March 2017 at 16:42 #8824
Cristóbal Rodero
BlockedHi everyone,
I’m trying to build a cube with FEM with a variable number of (hexahedral or tetrahedral) elements. I found the component SparseGridTopology and I thought it could provide me what I want.
The problem I encountered was using the example SparseGridTopology.scn, because it seems that when the mesh is refined, there is a loss of stiffness.
Am I right? If that is the case, why and how could I possibly keep the same stiffness?
I’m pretty new with SOFA, so maybe there is an easier way to do what I want (I’m open to suggestions).
Thanks in advance,
Cristóbal24 March 2017 at 17:06 #8825Hugo
KeymasterDear Cristóbal,
Welcome to the SOFA community!
The SparseGridTopology is an option to create a sparse grid (obviously). If all you want is (love) create a basic cube, you can also have a look at RegularGridTopology (see examples/Components/topology/RegularGridTopology.scn) or the CubeTopology (see examples/Components/topology/CubeTopology.scn).As you noticed in the example SparseGridTopology.scn, the coarser the mesh, the stiffer the dragon. But this has nothing to do with the mesh or the mechanical constitutive law: the scene uses a UniformMass. This mass defines a mass per element. The more elements you have in the mesh, the higher the mass.
Is that clear?
You should instead use the MeshMatrixMass or the DiagonalMass (for lumped mass).Best
Hugo.
28 March 2017 at 15:53 #8839Cristóbal Rodero
BlockedDear Hugo,
yes, all clear! After some tests I decided to keep using UniformMass, but specifying the TotalMass instead of the Mass.
Even though, all the information has been useful.
Thank you!
Cristóbal
28 March 2017 at 18:51 #8842Hugo
KeymasterHi Cristóbal,
Alright, I set the topic as resolved then.
But you need to know that this uniform mass is not “physically” correct, since each node has the same mass, regardless from the topology. It corresponds to a mass matrix M diagonal with one single constant value on the diagonal components.
Do not hesitate to create a new topic for any further question.
Best wishes,Hugo
-
AuthorPosts
- You must be logged in to reply to this topic.
