Nodal strain solution at a contact between different materials 1

[DavidCam]

Hi everyone!

I need help to interpretate FEM results. I would appreciate some responses.

I built a finite element model (using ANSYS) of a block (3D) composed by different horizontal layers, each one with different material properties (linear elastic). I used 3D solid elements. The boundary between materials is modeled so that they share nodes (rigid condition). Thus there is only have one node at a certain location belonging to the contact between different materials.

Now, I need to get the strain at a node that belongs to the contact between two materials to check if the upper material is good enough to resist. By default, ANSYS doesn't average nodal solutions across material discontinuities, so I get two different strain values depending on which material I select. Since I need to check the upper material, should I use the strain related to the upper material or should I use the average strain value between the two materials?

Thanks in advance!!

 

[Denial]

If the two materials are completely and continuously adhered to each other (which your description suggests is the case), then the strain at a node on the interface will be the same regardless of which material you are thinking about. It is the stress that will be different.

I have never used ANSYS, so what follows is a guess. The program is probably calculating the nodal strains by an extrapolation from the values at the Gauss points in the interior of the element. If so then small differences are to be expected, but large differences would indicate that your model is too coarse. If you need to ask the question whether to use the material-1 value, the material-2 value, or the average, then your differences are definitely large.

 

[DavidCam]

Thanks Denial for your reply.

I tried to make a finer mesh, but it didn't work. I got the same results. The horizontal strains are almost equal at the interface between different materials, but the vertical strain isn't (the load is also vertical), specially when the difference of the modulus of elasticity is big (ten times or more).

I'm not sure if it's relevant or not, but the model is simplified by double symmetry and I'm getting the strains at one of the planes of symmetry.