Bolted Connections 1

[stessedout]

I have six panels to make a box. These six panels are bolted together. The panels are built with Solid elements (Hexa and Penta). I am trying to figure out the best modeling method of the bolted connection from one panel to adjacent panel. The mesh density of the panels is a minumum of 3 through the thickeness. At the bolt locations i have removed the elements to simulate the bolt hole. I then place a center node at the center of each opening through the thickness of the hole. I use cbar elements to simulate the bolt shaft (minor diameter). I then spider (wagon wheel) to the outside edge of the hole using rbe3's. Is there a better or more acurate way of modeling the connection of the panles? Is it better to use rbe2's, springs? Also keep in mind that i have to look at isothermal conditions. Some elements just jack the stresses up unrealistically.

I need to determine the loads that are running through the joint so i can determine the stress resulting in the bolt with it preload which would be a hand calculation.

 

[mert1]

I use RBE2 elements for simulating bolts. And constrain all 6 dofs. RBE3 actually distributes the forces to the connection points by inversly propotional to the RBE3 element length. So I would not prefer using RBE3s. Depending on the stiffness of the materials, only rotational dofs can be constrained.

 

[SWComposites]

If all you want from the model is the loads in the fasteners, do the following
1) convert to a plate model (solids are overkill, more complicated to deal with, and not necessaryily more accurate
2) do not bother modelling the hole (overly complicated and you probably won't get the correct bearing deformation anyway).
3) model each plate with separate nodes; locate a node at each fastener location in each plate, these can be at the same location or offset by the plate thicknesses.
4) connect each pair of fastener location nodes with spring elements connecting the in-plane degrees of freedon and bar elements (if the nodes are not coincident) and springs (if the nodes are coincident) for the out-of-plane degree of freedom.
5) use fastener flexibility formulas to calculate the spring/bar stiffnesses; note that you have to include the plate bearing flexibility in the in-plane spring stiffness
6) I recommend making a small model of a one fastener joint loaded axially; use this model to calibrate/test your fastener flexibility values (spring stiffnesses). If you don't go with the above approach you should make a small model using your solid element/bar approach for the joint and use it to verify your joint idealization.
7) rigid elements in a model at fastener locations are almost always too stiff and will not give correct results unless used very carefully and with a full understanding of their behavior and effects.