Abaqus-Shell to Beam Element connection 1

[Error.n.n]

Hello,

How can I connect these two elements: the crossframe (truss elements) to the stiffener (shell elements). will merging instances work well in terms of results displayed?

Thanks
C

 

[FEA way]

Coupling constraint should work here.

 

[Error.n.n]

Hi,
Sorry for the long post in advance.

Yes I did try coupling constraint and I used the node at the joint of the cross frame and attached to the stiffener as the constraint control point and the surface of edge of the stiffener as the constraint region type.

However i get this error, "multiple usage of this node as a secondary node or combine the main surfaces together and use this surface instead.

Overconstraint checks: node 2 instance ub_girder-1-1 is used more than once as a secondary node in the *tie keyword. Remove multiple usage of this node as a secondary node or combine the main surfaces together and use this surface instead.

1 nodes are used more than once as a secondary node in *TIE keyword. These constraints are not removed either due to the presence of *CONSTRAINT CONTROLS,NO CHANGES keyword or because removing the constraints might affect the model. The nodes have been identified in node set ErrNodeOverconTieSecondary."

The error happens because I have tie constraints between the top girder surface and the deck bottom. I am using the top girder as my slave surface, hence it is has a secondary node that is connected to the stiffener.

I don't want to partition the stiffener to create a surface section at these joint locations and pick that section of the surface only because i am not sure how my results will be affected. (The stiffener is 928mm in height, what partition offset distance would I use? 100mm from the joint either side? Does it matter )

According to Abaqus I cannot choose the same secondary node for two constraints. Is there another way to do it or should I partition?

Another option,is to offset the stiffeners from the axis of the topflange. This way when i select the surface of the constraint, the node at the flange is not picked since it is not shared. In this model, I used the shell extrude function to model the stiffeners and the node at top edge of the stiffner is connected to the node axis of the ftop flange. However, when I offset the stiffener edge in another model( ie. not connecting the nodes of the top flange to the node at the top of the stiffener) , the stress results are affected. When the node of the stiffener and the flange is connected, the stress results i get at that location are significantly higher.

Please help me on how i should sketch the part of the stiffeners or how i can apply the constraints at that point. In the real bridge model, the cross frame is connected to the stiffener using bolts, I am trying to create an ideal model that reflects that situation without overcomplicating the model.


I forgot to mention that the elements are truss elements. I have another model where I used beam elements for the cross frame, merged the whole girder and the analysis run without any error about zero point pivot that I get when I merge the truss instance to the shell) no idea why that worked. Perhaps someone can explain it to me in the FEA language. Thanks

 

[FEA way]

I would try with some partition of a reasonable size (approximately corresponding to the area of the interface between those parts).

However, beam elements are usually the way to go and they are recommended in most cases instead of more simplified and traditional truss elements.

 

[Error.n.n]

Thank you for the insight, @FEA way. I have a follow-up question: The bolt connection of the cross frame is located towards the centre of the stiffener, but to simplify the model, I made the connection at the edge of the stiffener. Is that advisable?

If I did extend the cross frame connection node to the centre of the stiffener, would i have to create a contact constraint at those points, and is it a better way to model these sections?

 

[FEA way]

Turn on beam profile rendering and see if the geometries are properly aligned with respect to each other. Various simplifications are pretty much inevitable but it's best to reduce their number. Also, you can always create a copy of this model, modify the connections and see if the analysis works and if the results are significantly different from the previous ones.

 

[Error.n.n]

Thank you. I will do that.

 

[poping112a]

Another option,is to offset the stiffeners from the axis of the topflange. This way when i select the surface of the constraint, the node at the flange is not picked since it is not shared. In this model, I used the shell extrude function to model the stiffeners and the node at top edge of the stiffner is connected to the node axis of the ftop flange. However, when I offset the stiffener edge in another model( ie. not connecting the nodes of the top flange to the node at the top of the stiffener) , the stress results are affected. When the node of the stiffener and the flange is connected, the stress results i get at that location are significantly higher.