Hi, I’m trying to model a wing with a strut on it, but I can’t figure out a way to model the wing-strut joint without modelling a physical joint. The joint should allow rotation along 1 axis, and move as the wing deflects up or down. I’ve tried to use *EQUATION with 3 translational dof fixed and 2 rotational dof fixed to create a joint but the joint just acts stiff as the wing deflects. What is the best way to model this type of joint in Calculix?
How did you decide the joint acts stiff? It’s not really clear to me from a deformed view.
The geometry of the hinge looks mostly OK - both pairs of nodes differ only in x which is the free axis.
However, you shouldn’t usually couple nodes with simple equality like that when they aren’t at the same location because it doesn’t transfer moments in a physically realistic way. For example, a displacement in z on one node will cause an equal displacement in z on the other, but without the moment it should require because of the z offset between them.
I just remembered you can’t use rotational DOFs in constraint equations in CCX. It probably doesn’t need them anyway but maybe they’re somehow causing a problem.
It’s also risky to use constraint equations on shells because they might conflict with the MPCs of the shells which usually causes an error but maybe just goes wrong sometimes. As JuanP74 suggested, elastic connections are a good idea because they’re more reliable in CCX.
Thanks for both of your replies. I’m still pretty new to FEA, could you elaborate on rigid bodies connecting coincident nodes joined by 3d springs? Which part of the model should be a rigid body?
you should create 2 rigid bodies, 1st is a fitting in the wing, 2nd a fitting in the strut, they all end in a node which should be coincident but different for each rigid body. The elastic springs are attached to these 2 nodes
*RIGID BODY,NSET=rigid1,REF NODE=100,ROT NODE=101
*RIGID BODY,NSET=rigid2,REF NODE=200,ROT NODE=201
pink are rigd bodies, yellow are nodes (REF NODEs and ROT NODEs), cyan are springs
hope is clear.
You could do something similar using elastic bodies instead of rigid as well, rigid bodies are easy to create in prepomax for example
this case of shell models can be problematic, i’m not sure rigid body with reference nodes is possible to use in connection, i.e two node spring element. Distributing coupling can be used as connector mixed with another element (spring, truss or beam) but only allowed in model with solid element due to surface based constraint approach. Maybe it can be achieved by adding more detail in models of connector parts.
*edited
i checked again, using reference node to connect with two node spring is allowed. Probably i forget in CalculiX versions since previously did not succeed, great news since nonlinearity seems to be possible also.
