Hole internal torsion by applying rotation to one edge of hole

[halston]

I'm trying to model some sort of internal torsion by applying rotational displacement to one edge of a hole in the middle of a disk-shaped specimen. After getting the result, I see that only displaced nodes have non-zero displacement(equal to the applied displacement) values and all the other nodes even nodes which are on the elements located on the hole surface have absoloutely zero displacement values which is not actually the case. the displacement should penetrate radially in the specimen up to some point depending on the magnitude of applied displacement.
Can anyone tell me why this is happening?
I've attached the .cae file!

 

[OptiEng]

Hi halston,

Interesting problem you are looking at... I noticed that your torsion is applied in the U2 direction, could this be causing the issue you are having?

 

[halston]

Hi thanks for your response.
I'm applying torsion in U2 direction in a cylindrical coordinate system which represents Theta and that's the direction that torsion should be applied! Do you think it should be in another direction?

 

[OptiEng]

I see, sorry I missed the fact you were using a cylindircal co system. I wonder if you need to use Adaptive meshing (you can find more about this in the abaqus manual). Could you consider modelling a rigid rotating shaft through the hole in contact with the hole and model the interaction properties to those in reality? I hope this helps.

 

[halston]

I've already done that. But the problem is that the kind of torsion I'm dealing with is not uniform along the axis of the hole. I want to find the torsional stiffness of the specimen with this specific type of torsion which has a gradient along the axis of the hole and I don't know the correct distribution of displacement in that direction. All that I know is the displacement on one edge and I want Abaqus to tell me how the distribution of displacement is along the thickness of the specimen on the hole surface. What is obvious is that by using a rigid body inside the hole I'm forcing all nodes to have same displacements which is far away from my real problem. It seems that the displacement doesn't propagate radially to other elements. I think there is a discontinuity between elements in this problem!

 

[OptiEng]

Sorry if you have already tried this, but have you lowered the U2 value to see if you obtain a more realistic distribution?