PSHELL ? is a propert card ... i think you have CQUAD.
"common nodes" ? if the nodes are common to each quad, then this represents a rigid connection and your rods should be doing nothing. if you have two nodes (one on each element) joined by the rod, then you'll have some load in the rod.

CRODs were probably not the best element to pick for the rivets, they only carry tension load, and you want something to carry shear, yes? CELAS zero length springs work well. if you've modelled the two skins at their mid-plane, then use an RBE3 to create co-incident nodes (for the zero length element). or BEAM elements.

Hello RB1957,

Thank you fo the reply. I think I have not communicated properly. Pls ref the picture. It shows the fastener arrangement. I hope this will be clearly show you the arrangement. I appriciate your feed back.

• http://files.engineering.com/getfile.aspx?folder=b7fe9982-2380-46fe-9213-b0

i have no idea how that model would work ... you say you're getting close to zero in the rod ... not very surprising ... rods react only axial load.

what you identify as "PSHELL" i call "CQUAD" since that is the element. you're trying to connect the two panels, right?, with a shear splice.

if the two elements share common nodes, this models a rigid connection. you can use GPF to find out the force being transferred between the panels.

using CELAS zero length single dof springs would allow you to create a finite stiffness between the elements; mind you the load will be pretty much the same (load in from the LH element is reacted by the RH element).

if you want the distribution between the fasteners shown, you'll need a finer mesh.