I already asked this question in another topic, but i guess no one answered, because the topic is already solved.
How do I calculate the stiffness for an elastic boundary condition?
Usually one would calculate K=E*A/L, but I think this does not apply for a surface boundary condition, does it?
I presume that you know the stiffness (=A*E/L) that you want to simulate. Right? Let's call that stiffness Keqv (for equivalent stiffness).
In the model, the equivalent stiffness is being represented by N elastic boundary conditions ("1D Springs" in the ribbon menu) working in parallel. So you find the formula for calculating the equivalent stiffness for parallel springs, and enter that into the stiffness field. (It's either Keqv = sum of k, or 1/Keqv = sum of (1 / k). I never remember which is for parallel springs and which is for springs in series.)
Since you need to know the number of nodes, it would be better to select the surfaces, then select the nodes on those surfaces (right-click > Select Subentities > Vertices). The title bar of the elastic boundary conditions will indicate how many nodes are selected.
John Holtz, P.E.
Senior User Experience Designer, Simulation
Current version of Mechanical & Multiphysics: 2013 SP1 (2013.01.00.0012 28-Jun-2012)
Okay, your explanation makes sense. If springs are working parallel it is the sum of all stiffnesses K=SUM(k).
So I also tried to apply a stiffness to each node (like you said: selecting the subentities and dividing the stiffness K by the amount of nodes n: k=K/n) and I got another, better looking result.
However, don't I get a higher "stiffness density" in local refinement areas, now?