Mr.Spyros,
Yes, in full integration the finite elements are over stiff, yes it is true in FEM ..the reason is very simple as you are dividing the geometry into number of smaller element ..then, stiffness in element level and global level will be formed and solution will be obtained by applying bc's..in this process you could see that the global stiffness matrix is higher than the exact..so with the use of that global stiffness with full integration the element behave as stiffer element..thus, if you go with reduced integration, the error will be compensated and the solution become more accurate..i suggest you to read any finite element text book..the one like the 'first course in finite element method' authored by Logan..
You are asking about "locking" effect in FEM. There are few types of "locking", one of them is explained here:
https://www.quora.com/What-is-shear-locking-in-FEA
Karthigeyan . S It is hard to understand what you mean. Locking could not be overcome with meshing with smaller FE.
to make things more intuitive, “locking” refers to the presence in the element of parasitic stresses/strains that absorb a fraction of elastic energy, so that the loads applied to deform the element result to be higher than the loads needed if such parasitic contributions were absent.
For example, a 3 node triangular element (CST) under pure bending also shows a shear strain, which however is not predicted by the theory.
Such parasitic contributions arise as a direct consequence of the approximations (i.e. shape functions) used in finite element method.
- Badges
- Science topic