Boundary Conditions on Pressured Vessel 1

[ecFem]

Hello,

I hope you are doing well.

I am analyzing a tank, and the idea is to simulate a burst test. Below I leave an image of the tank.

The thing is that I would like to study it as if it was floating in the space, i.e., analyze only the pressure and neglect the effect of the necessary boundary conditions.

In this case, is there a set of BCs or an strategy that minimizes the BC effetct?
Moreover, any other tip or advice you'd like to share for this kind of analysis would be appreciated.

Thank you in advance !

Cordially,
EC

 

[FEA way]

The 3-2-1 method could help but it would be much better to utilize symmetry (assuming that the geometry and loads are symmetric).

 

[rb1957]

"is there a set of BCs or an strategy that minimizes the BC effect?" ... "any" scheme that removes the 6 rigid body motions should have "zero" impact to your model (as these will react zero energy).

the rigid body BCs set up a datum plane in the model, that the model deflects about, so there can be convenient planes (and inconvenient ones too).

Constraining a single point in all 6 dof is probably not a good idea, as the bending stiffness to the shell elements is not great (but there should be little to excite the model). The 3-2-1 method (constraining the translational dofs at 3 nodes) is a typically good approach.

One thing from my experience of modelling pressure vessels ... make sure your geometry construction behaves as you expect. I'd model a cylinder and see that it inflates as expected, then add the end caps.

I'm not sure why your model shows a very narrow shell (at the transition to the cylinder) ?

"Hoffen wir mal, dass alles gut geht !"
General Paulus, Nov 1942, outside Stalingrad after the launch of Operation Uranus.

 

[rb1957]

Also, your endcap geometry doesn't look like it'll mesh well (ie with rotational symmetry) ??

"Hoffen wir mal, dass alles gut geht !"
General Paulus, Nov 1942, outside Stalingrad after the launch of Operation Uranus.

 

[ecFem]

Thank you very much for your responses.
I continued working with a symmetrical model, which is way better from all prespective points.
I improved the geometry and the mesh is all mapped.

Thanks !