Too High Displacement!

I had created a relatively simple wooden structure to analyze using karamba. My Goal is to use 0 fixed connections. The problem is that the displacement value seems to be too high. I’ve been shifting all planes and changing all orientations in the model trying to figure our what the problem is, maybe its not related to the element planes, im not sure. It could be something fairly simple that my beginner experience is not aware of, I’m still learning Karamba and GH. Any help would be appreciated

ExampleFile.gh (76.0 KB)

Hi @mahradsa32, you have defined some elements as truss elements and therefore your stucture is kinematic. If you want to control your joints, use the BeamJoints components to define your joint conditions.

Thank you for you reply, I will try using that component and readjusting the definition.

I had another question if possible. Using the joint definition component is different from the support component in the sense that in the support component when the box is filled then that degree of freedom is locked, while in the joint definition when the box is filled it means that this degree of freedom is free or released ?

If that is the case if i am using the joint definition to identify the end connections (joints and D.O.F) then if there is one point that overlaps (like point C) in my example that is attached. Should I release it in both ends of the members (member BC and member DC ?) or releasing it one member is sufficient? releasing in both members means that the end connections are both hinged or does it mess up the calculation?

I am trying to reach a solution where the main connections have some degrees of freedom (hinges ) and the secondary supporting elements are fixed (Like BC’)

ExampleFile.gh (85.9 KB)

Hi @mahradsa32,

Yes you are correct. The nodes in the Beam Joint component releases the degree of freedom for the selected node.

When defining the joint, it releases the degree of freedom at that specific node for that specific member. In your current setup Column(BC) is hinged at node C, but Column(CD) and Beam(CrossBC) are still rigid. You can see that when you apply joint also on Column(CD), the structure becomes kinematic.

Thank you for the response and the clarification. I believe I know what I must do now as the reactions and numbers are making more sense. One last question ! :slight_smile:

There are two ways to go about joints in Karamba, With joint identifier (Joint Definition) or by toggling the bending state of a beam element to False, which makes it behave like a truss so all forces (only tension and compression applies ?). For more accurate results It makes sense to use the Joint Definition, But for faster results the Bending element works better ? I am trying to decide which approach makes more sense to use, I know given the design and the parameters it depends, but it seems like Joint Definition is flexible customization and more detailed, while the Bending element approach woks just for complex truss systems ?

Dear @mahradsa32,
toggling ‘Bending’ to false removes the bending stiffness degrees of freedom (dofs) and thus leaves three instead of six dofs. Therefore a model with truss (2 x 3 dofs) elements calculates faster than the equivalent model with beams (2 x 6 dofs).
Joint definitions add degrees of freedom to a beam. Therefore computation time slightly increases when using them as compared to the elements without joints.
Nodes which are connected to truss elements only, need to be supported in three non-planar directions in order to be stable.
– Clemens

1 Like