Stuck? 
support@mecway.com
support@mecway.com
Mecway
FEA
Nodal Stress Calculation - Average over Adjacent Elements of the Same Material Only
Hi, I'm using version 23. Two issues got my attention: Calculation of the nodal stress using average over all adjacent elements and uniqueness of the element stress.
Currently, Mecway V23 calculates the nodal stress by taking the average over all adjacent elements. It is not difficult to show that this approach can lead to significant error if the elements have very different material properties. This will become apparent if we include the free space into the mesh and set the elastic modulus to approach zero. Then the stress calculated at the interface nodes would become 50% too low. By default, at the interface nodes, we should take average only over the elements of the same material properties and only report the higher stress.
Turning to the 2nd issue, I found that the stress for the same element can sometime have different values when Mecway displays the nodal stress, which is calculated from the average over adjacent elements. As far as I understand, the solution for displacements is unique and the element stress calculated from the displacement (defined at the nodal points) must also be unique. Is this correct?
Currently, Mecway V23 calculates the nodal stress by taking the average over all adjacent elements. It is not difficult to show that this approach can lead to significant error if the elements have very different material properties. This will become apparent if we include the free space into the mesh and set the elastic modulus to approach zero. Then the stress calculated at the interface nodes would become 50% too low. By default, at the interface nodes, we should take average only over the elements of the same material properties and only report the higher stress.
Turning to the 2nd issue, I found that the stress for the same element can sometime have different values when Mecway displays the nodal stress, which is calculated from the average over adjacent elements. As far as I understand, the solution for displacements is unique and the element stress calculated from the displacement (defined at the nodal points) must also be unique. Is this correct?
Howdy, Stranger!
It looks like you're new here. If you want to get involved, click one of these buttons!
Comments
I'm not sure what you mean in your second point about the different values on the same element. Element stress should be unique (single valued) at each of the element's nodes.