ABAQUS torsion question

[Ashu28]

Hello all,
I have a really simple question (most of you might know the solution from top of your head as well). I am trying to get started in ABAQUS and am in process of modeling torsion in a hollow cylinder. I constrain all DOFs on the one face and use a Reference point kinematic coupling on the other face. Then I give a BC at STEP 1 (after initial step) of rotation UR3 (in axial direction) = 1.5 (about 90 deg) to the reference point. This should theoritically proide torsion to the cylinder but what i see is that the cylinder kind of bulges. I am not sure why is this happening and how can I fix it. I am simply tyring to simulate the load (RM) required to twist it but really doing something wrong.

I am attaching some pitures showing coupling, boundary conditions and result alongwith. Can someone please tell me what is the issue with this simple problem.

Kinematic Coupling:

http://files.engineering.com/getfil...a941-4320-a486-fc076029d09e&file=Coupling.jpg

Boundary Condition and Load:

http://files.engineering.com/getfil...-bc7d-4295-af41-7fd2e0a0fb57&file=BC_Load.png

Result:

http://files.engineering.com/getfile.aspx?folder=5b333728-9a31-4a58-9ec4-fe5ea415abdc&file=Bulge.jpg

Regards,

 

[corus]

Has it bulged or just necked due to the torsion? Plot the undeformed mesh on top of the deformed mesh to see. Incidentally, the pictures you provided were tiny and pretty much useless.

The manager

 

[Ashu28]

Hi,
Thank you for the reply. The job is not necked. It is conical. The constrained face (with DOFs 0) is constrained throughout but the other face just grows in diameter, both ID and OD and the whole part kind of gradually grows in diameter along the whole length. I am putting in another picture, but i think this one is coming out small too....I donno why. I am attaching my small CAE file as well (ABAQUS ver 6.11).

http://files.engineering.com/getfil...e14-4cfa-a1e0-e628b9ce286c&file=DefnUndef.png

CAE file:

http://files.engineering.com/getfil...ef16-4b9d-a206-db0ca282bf33&file=Torsion1.cae

Can you please look into and advise as to whats goin on?

Regards,

 

[friedmind]

I think the results are OK end the apparent grow in diameter is due to the scale factor of the deformed shape and/or the perspective.
You should measure the displacement at one point of the outer diameter of the unconstrained face and check if the result is consistent.
Hope it helps.

 

[corus]

It's nothing to do with the perspective or the scale factor as the tube simply forms a conical shape under the applied rotation. I'm not sure why this occurs and you might be able to tell why by the stress components but as you've used tetrahedral elements, which are basically useless, then you can't see these easily. Change to a hexahedral mesh and run it again and ask your tutor.

The manager