# Free body under thermal load. Unexpected Stresses

Hi,

I’m finding VM stresses at the end of a free expansion of this beam. Which could be the reason?

Abaqus results from this .inp:

Ohh!.

I agree with the free stress state, that’s what I would expect but displacements are weird isn’t it?

I would expect 0.4 and -0.4, symmetric expansion around the center and an overall 0.8 elongation of the beam.

Without Nlgeom, the values are different:

With symmetry BCs to keep it in place:

linear:

It’s like with unconstrained pressure vessel subjected to uniform pressure load - minor disturbances can make it respond non-uniformly.

Thank you very much Calc-em. That last picture is very interesting. Abaqus LINEAR result for displacements it is the same I’m obtaining but I am supposedly using the nonlinear solver. I have looked at the inp that Prepomax generates and I have seen that Nlgeom is correctly there.

¿?¿

CalculiX result for the case with BCs and Nlgeom (should have the same results as in the second image in my previous post):

Maybe Nlgeom is not working correctly in this procedure.

I agree with your. NLGEOM is not active for the thermal part of the expansion.

Might be worth reporting on GitHub then. Another important bug to notify Guido.

I wonder if you need to make an arbitrary choice of strain measure for large strains. Since CCX and Abaqus disagree on the length change, maybe they just have different assumptions about the meaning of CTE?

CCX hints at a difference from other codes which I don’t quite understand:

“This establishes an exponential expansion relationship. This is fine, as long
as the thermal strain was also measured according to Equation(18), i.e. with
reference to the actual length. The latter approach, used by a lot of Finite
Element codes requires linear expansion coefficients for linear calculations (using
linear strain) and exponential ones for nonlinear geometric calculations (using
logarithmic strain). The approach in CalculiX avoids this.”

It doesn’t explain the non-zero stress though. But perhaps that doesn’t really matter if there’s no real material that can have 40% thermal expansion and still remain a solid?

I think you didn’t achieve the steady state.

Hi JuanP74,

I’m imposing an increasing temperature to the body as a whole. Just in case I have look at final Heat flux and it’s negligible.

You don’t have flux BC on your model so all heat is taken as internal energy by the beam. It is thermally insulated. Try using steady state parameter

Hello JuanP74,

I have try the Steady Option and it doesn’t help. (Convergence is failing).
I have also increase some orders of magnitude the Thermal conductivity to help spreading any small temperature gradient and the result is the same.

Temperature is uniform in all the body after each iteration. From my point of view there shouldn’t be any Stress for this particular geometry. Something different would be in pressence of Gaussian curvature but this is not the case for a straight bar.