Expected behaviour of cylindrical beam under torsion

[TS77]

For a section of a piece of coursework we need to analyse how a beam is deformed when subjected to torsion. My expectation would be that it simply twists, but the deformation diagram obtained from simulating this in Ansys shows a deformation which seems to suggest the cross-sectional area of the beam increases at the end where the torsion is applied.

Is this the expected behaviour, am I misinterpreting this deformation diagram, or is the simulation just completely wrong altogether?

 

[GregLocock]

Wouldn't it be amazing if we knew what this was a plot of, and what the units are, and which end is fixed? guessing, left hand end is encastre, torque applied at RHS, angular deflection or torsional stress or strain, and yes that is exactly what you should have been able to predict from beam theory. If you couldn't work this out by hand then you might as well major in catering.


Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[TS77]

Apologies Greg, I don't think I was especially clear, it seems we've been taught with slightly different terminologies and conventions.

Your assumptions are correct, I should have included but them its been a very long day. It is a cylindrical beam, fixed at the LHS with torsion applied at the RHS. Units are metres.

I am aware that the expected behaviour of the beam under torsion will be an angular rotation which induces shear stress in the surfaces.

I suppose my question is better phrased as whether the plot itself is representative of the expected behaviour. While I'd expect the rotation, the diagram seems to imply an increase in cross-sectional area towards the point of application of torsion, i.e the increased radius on the diagram.
It's certainly possible that I'm misinterpreting the diagram, and instead the values refer to the displacement of a point on the surface, which would indeed be expected, but the change in size of the diagram introduced some uncertainty for me.

 

[GregLocock]

The post processor is moving the surface nodes tangentially. That will have the effect of making the cross sectional area bigger.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[BAretired]

A cylindrical beam subjected to torsion would feel shear stress varying linearly from zero at center to maximum at outer edge of cylinder. This would be typical throughout the length of beam, except for the disturbed regions at each end.

BA