Hello all,

I have problem regarding meshing with the shell elements. I have a 3D body which contains cylindrical part and I'm doing pressing simulation with it (this part is cca 20x stiffer than the other part that is pressed). I found out that pressing curves matches when I use 3D solid elements and shell elements for the stiffer part, of course the shell variant is much faster.

So I have a 3D geometry and I am meshing only the surfaces I need to be included in the simulation. It is important to have the radius that matches the measurement, so the radius changes (MIN NOM MAX variant).

My problem is, when I have a 3D cylinder with some radius, after meshing the radius is different. I found out when examining the mesh - see pictures attached. When I measure vertice vs node, I see the deviation that is not intended (meaning that X and Z distance in the measurement are non-zero). But when I use node vs node distance measurement, it has no deviation, but the radius is smaller than on the 3D geometry. 

So the mesh node does not lie on the vertice (which is weird, that doesn't happen normally to me).

Maybe it has something to do with shell elements? In the section assignment I use Top surface. Or mesh settings? Element type is S8R with 6 DOFs per node, mesh control is structured - quad-dominated.