2D axisymmetric Hydrostatic load not working properly

Bug Reports
1

Here are demo files to show the bug
2D_axisymmetric_bug.zip (2.9 MB)

The 3D version of the load seems to work just fine. However, the 2D axis symmetric version does not seem to work as desired. It seems to be 10 times less pressure and not evenly distributed. I’m not an FEA expert by any means, but I think it is related to the way the pressure is integrated over the nodes. I saw a post about a 2D application where the stress was 10 times more than expected. Its possible that when that was fixed, it affected this one.

here is the uniform load in 3D. I applied 10MPa and 10MPa is shown.

3D_uniform
3D_uniform1018×487 153 KB

here is the uniform load in 2D. I applied 10MPa and 10MPa is shown.

2D_uniform
2D_uniform792×487 120 KB

Here is the hydrostatic load. I applied 10MPa at two end points and direction from end point to end point:

2D_axisymmetric
2D_axisymmetric825×475 143 KB

the little fold is the center of the axis where one would expect there to be a more even distribution. And notice how it shows around 1MPa and not 10MPa as applied.

2

Right, it seems to be a bug. It happens in v1.3.5.1 as well. Another way to verify this is to check the reaction forces.

1 Like
3

You are probably right. I did not consider applying it to the axisymmetric case. In the case of a 2D or a shell model, the area of the surface is computed from the thickness of the model to correctly compute the nodal loads (internally the hydrostatic pressure is converted to cload). I think that the thickness of the model is not computed correctly in the case of axisymmetry.

Probably the surface traction load could also be problematic. I am wondering what a surface traction load in such a case should actually do? Apply a load that equals a constant pressure?

4

That’s interesting about the surface traction. It would seem that only axial and radial forces would make any sense.