Steel beam-column connection torsion

[kabur88]

Hi I'm currently modelling a steel beam under torsion that is welded to a column. I am currently having problems modelling the contact surface as it says that in the Warnings page:

(576 elements are distorted. Either the isoparametric angles are out of the suggested limits or the triangular or tetrahedral quality measure is bad. The elements have been identified in element set WarnElemDistorted.)

 

[kabur88]

I don't understand what I am doing wrong here but I would really appreciate your help thank you

 

[ESPcomposites]

It would seem the message tells the story.

You have poorly shaped elements "distorted". Therefore, the results cannot be trusted. ABAQUS created a set for you to observe the specific elements it is referring to.

You will have to improve the quality of the mesh. Of course, then there is a question of why you are using contact in the first place, but that is another story.

Brian

http://www.espcomposites.com

 

[kabur88]

Thanks for your help. The reason why I was using contact was because I wanted to stop the endplate that the beam is connected to from warping into where the column is supposed to be located

 

[ESPcomposites]

Couldn't that be done with displacement boundary conditions or springs? Contact can lead to convergence problems and is time consuming.

Brian

http://www.espcomposites.com

 

[kabur88]

I see, well I just started using AbaqusCAE anyway so I'm still learning it as it now. Thanks for the info though, I'll try look it up

 

[kabur88]

I tried what you said about the displacement boundary conditions, however the endplate is now not warping at all. I wanted it to be able to warp towards the beam while not being able to do so towards where the column is supposed to be located.

 

[kabur88]

Does anyone know what I may be doing wrong and how to simulate what I want to do?

 

[ESPcomposites]

It would help if you posted some pictures.

Do you simply want to connect the parts (tie them together)? That can be done many ways. It seems you may be "shooting darts" if the previous results and warnings are stumping you.

You may want to start by asking questions on how to approach problems, rather than creating a potentially awkward model/scenario and trying to "debug" the issues. Being a novice user, you may have several issues that should be addressed.

Brian

http://www.espcomposites.com

 

[kabur88]

The endplate and beam are in union and the flanges at the square beam are to apply torsion. The endplate is fixed at the edges and can warp in any direction to the (Z)axis. What I really want to do now is to stop the endplate warping into the back but allowing it to warp in the direction of the beam. This is to simulate a column at the back that would stop the endplate from warping in the (-Z)axis but still allowing it to warp into the (+Z)axis.

Thanks

 

[ESPcomposites]

A couple options:

- Try the contact solution again, but improve the mesh such that you don't get warnings. I am guessing it is due to the thin endplate trying to be meshed with solids. You probably need more elements.

- If you can visualize where the contact will occur, just ground the appropriate surfaces/nodes and don't apply boundary conditions to the areas will separate.

General:

- You should consider plate/shell elements rather than solids. You will a lot of solid elements (and a few through the endplate thickness) to capture the bending in it.

- As a beginner, forget about the contact. You have bigger fish to fry, like understanding how to use the proper element, mesh density, etc. You may be eager to jump into difficult problems, but it will be counter productive in the long run. Break out your FEM book as well and start getting familiar with the behavior of elements.

Brian

http://www.espcomposites.com