create coincident nodes and use coupling to release the moments. It will be appropriate to use released beam local coordinate system.

Since ansys does support scripting, it shouldn't be difficult to script to do it automatically for multiple nodes.

Yeah.. Abhimanyu Kumar is right.

In case you want to go into the non-linear range too, use a hinge element like COMBIN39 to connect the 'co-incident' nodes, although, if i remember correctly, COMBIN elements need a small gap between the nodes, and not perfectly co-incident..

Abhimanyu , Rahul and Claudio thank you very much for your answers!

Claudio Thanks a lot for such a detail explanation, it is detailed enough so that even a complete beginner like me can understand it. Indeed so far i have barely had any contact with the APDL commands and I see I will need to get familiar with them to increase the benefit of using ANSYS.

Claudio I believe your example is a good solution, as you said there is the problem of rotation around its own axis but it can be solved by using the commands around a group of elements with similar orientation therefore I could do the procedure for a big group of elements at the time and release the appropriate rotational degrees of freedom, it is indeed some work but is still a lot better than doing it one by one in workbench, So i think this is indeed the best solution and still a big improvement from my originally proposed method.

Thanks a lot!

For the case where the element would freely rotate about its axis freely, I presume, one could release torsional DoF on one end alone, while flexural DoF's can be released at both. Still one has to look carefully and visualize whether the beam can freely move.. like for example, release Rz at one and and Uy at the other, the second end will simply have a 'free fall'.

Claudio Pedrazzi

ENDRELEASE,,-1,ROTX,ROTY,ROTZ

But then I don't know how to indicate in the code the couples of nodes i want to connect with this command. Can you help me? Thanks a lot, and sorry for this one-year-late answer.

I feel a simple way to do this is by not having the links connected by common nodes rather having seperate end nodes for different members. But the nodes should be coincident or nearly coincident ( with in the tolerance). The nodes corresponding to all moment free joints can then be selected and couple coincident nodes command can be used. It must be noted that only the displacements must be coupled thus allowing free rotation if the nodes at the joints.

I think this will be a convenient way to achieve moment free joints. I think this is what Abhimanyu Kumar had also suggested. Hope this helps.

I feel a simple way to do this is by not having the links connected by common nodes rather having seperate end nodes for different members. But the nodes should be coincident or nearly coincident ( with in the tolerance). The nodes corresponding to all moment free joints can then be selected and couple coincident nodes command can be used. It must be noted that only the displacements must be coupled thus allowing free rotation if the nodes at the joints.

I think this will be a convenient way to achieve moment free joints. I think this is what Abhimanyu Kumar had also suggested. Hope this helps.

Lucia Zerruso, hello. I was actually having a similar problem trying to model a truss structure with beam elements (purple lines) and truss elements (blue lines). I intended for the transverse trusses (horizontal ones) to be jointed to the battens with free z-rotations. I had several unsucessful attempts using ENDRELEASE and each time I ended up with automatically-created additional nodes, which I still can't understand. If someone can give me some insights into ENDRELEASE, I would greatly appreciate and I think Lucia Zerruso would so as well.

As to the problem I had, I decided to go with Mir Aamir Abbas suggestions and it seems to be working (check the pic). I used coincident nodes and I coupled all DOF's except z-rotation.