I am glad it worked out for you.
1 Like
I’m wondering if this DCOUP3D is working properly for bending moments. I get the expected values when applying forces, but unexpected results when applying moments.
I’m using a simple test case with a cantilever beam. In PrePoMax I defined a node set (called Nodes_face) and a reference point (called RP) where one moment around the y-axis is applied. You can see an screenshot below for reference:
Then I define the connection between Nodes_face and the reference point. To do so, I add the next CalculiX keywords:
*Element, Type=DCOUP3D, Elset=RBE3
13804, 22915
Where 13804 is the number of a new element and 22915 is the node number of the reference point for rotations.
Then I define the distributing coupling as:
*Distributing Coupling, Elset=RBE3
Nodes_face, 1
Where I provide the node set (Nodes_face) and a weight of 1 for all the nodes.
I can see that the moment is applied in the right node and in the right direction:
However, when I run the calculation, I can observe that the cantilever beam rotates around the z-axis instead of the y-axis:
Am I doing something wrong? Here you can find the file that I’m using in case you want to run it at your side:
Testing_RB3_for_moments_PrePoMax_v202.pmx (4.0 MB)
Thanks for the support!
looking the input definition was set to force in Y(2) direction, not a moment. DCOUP3D element can not be used for moment type, required Coupling type Distributing instead. Below, an example used from another threads discussion.
With the right type of coupling for this task:
Testing_RB3_for_moments_PrePoMax_v202_mod.pmx (1.6 MB)
Abaqus also has both versions of coupling constraints and *COUPLING followed by *DISTRIBUTING is the recommended approach. This one is implemented in GUI.
1 Like
Thanks for the support! I confirm that the *COUPLING *DISTRIBUTING works as expected for forces and moments 
I’ll only use the DCOUP3D approach to link a lumped mass (*MASS) to a given set of nodes as I’m not aware of other ways to do this without rigidifying the system.