uneven stress on pressurized barrel haz

Dorian_Busson
Dorian_Busson Altair Community Member
in Community Q&A

I encountered an issue with a pressurized barrel.


To model the welding between two barrels, I applied different material properties to simulate the Heat-Affected Zone (HAZ). The mesh of this component is linked to the other component (with the initial material properties) through nodal equivalence. The geometry has the same diameter for both the HAZ component and the raw material barrel.

The pressure is uniformly distributed and applied evenly across the entire structure. Therefore, I expected the results to show homogeneous stress distribution for both the base material of the barrel and the HAZ. For the barrels, the stress distribution is fairly even, as expected.

image.png

However, for the HAZ component, the stress distribution, as you can see, exhibits a sort of repetitive pattern with a variation of about 100 MPa.

This result doesn’t make sense, and there seems to be an error. I’m unable to determine where it could be coming from. Do you have any ideas?

image.png
Tagged:

Answers

  • Kranzeder
    Kranzeder
    Altair Employee

    Hi Dorian_Busson

    Sometimes small numerical meshing issues can cause such things.
    Please make sure that all of the nodes are in the 100% correct position. (radius / x-position)
    Please also check whether element normals all point in the same direction.
    Also check elemental coordinate systems pointing in the same direction.

    Hope this helps.
    Kind regards

    Juergen

  • Dorian_Busson
    Dorian_Busson Altair Community Member

    Hi Kranzeder,

    Thank you for your answer.

    In the meantime, I remeshed the shells using nodal equivalence instead of stitching the surfaces. Specifically, I meshed the different components separately, ensuring they had the same number of elements at the intersection before applying nodal equivalence.

    For some reason I can't identify, this fixed the issues and now the stress is consistent. It's not the first time I've encountered such problems with the stitch function in my office.

    If anyone knows why stitching surfaces might cause this, feel free to share your insights.

  • Kranzeder
    Kranzeder
    Altair Employee

    Hi Dorian_Busson

    Stitching alters the surface definition by moving one surface edge to the other.
    So you need to be careful which edge you move. E.g. moving the good edge to the bad edge definition instead of vice versa.

    Finally this could have caused position issues of the nodes created on this surface definition in your case.
    By luck nodal equivalence seemed to have it done in the correct direction. :o)

    Kind regards Juergen

Categories