Hi, this is probably common topic for FE design but I cannot find a good example on how it is solved, my problem is that I need a better connection between my beam and shell elements to avoid local maximum stresses. This leads to much higher utilization in the shell elements, ultimately governing the design which they should not.
I am able to extract the nodes (in green) which I would like to connect to the nodes of the beams respectively but I cannot figure out this beam-joint-agent component, or where to go from here. If there is another component or a better workaround, please let me know, regards
Hi Matthew, yes I have looked at those examples, but I cannot understand if this is possible in Karamba. I would like to make a rigid connecting between a branched list of nodes to the corresponding node on the line of the beam. I have all points as shown in the screenshots, regards Rick.
To clarify, this component does exactly what I am after but is meant for FEM-design, this allows me to connect the transversal stiffener in more than one point to the bottom cord beam.
Hello @Rick8,
you can avoid the stress singularity by placing a vertical beam (or two to have a connection point in the middle) at the end of the web of the transversal beam which is modelled with shell elements.
Placing a short spring between these short beams and the truss elements lets you control the type of connection. Zero length elements are possible but necessitate additional thoughts regarding node handling.
– Clemens
Hi Clemens, I did exactly that. I extracted each node from the mesh on the edge of the transversal stiffeners (shells) and made a line from each mesh node on the edge to the point on the beam end, and then made them into beam elements. This is an OK work-around but is not very pretty, as it also adds additional mass, especially when the mesh is fine and many beams are used to connect the shell to the main beam. I also tried the index to beam component but did not get it to work between the beam end points and the mesh nodes on the edges.
Hi @Rick8,
you can avoid the additional mass by using a dummy material with zero specific weight or springs at the edges.
Did you try to sort the nodes at the edges by their Z-coordinates to have one continuous set of beams and then connect their overall middle point to the truss?
– Clemens
I should of thought of changing the specific weight earlier! , if you mean my solution for the "ficticous beams? I tried it now, it gives the exactly the same results, with 0 specific weight, its prettier I assume as the dummy elements do not overlap this way
- Karamba3D 2.2.0
, if you mean my solution for the "ficticous beams? I tried it now, it gives the exactly the same results, with 0 specific weight, its prettier I assume as the dummy elements do not overlap this way 