Hoop Stress for 'inner tube-type' design 2

[Nella95]

I need help setting up COSMOS for a inflated tube. It's simple enough to use the equation: ?=PD/2t . But I wanna use COSMOS, and it doesn't want to work right.

P= 80psi
D=10.22 (though I'm unsure this is the nominal diameter)
t=thickness of tube


(Or is it this simple?)

 

[daveboy]

When you say it doesn't want to work what do you mean - is it because it's under restrained, it won't mesh or some other error?

Is this like a pipe inner tube / liner (i.e. linear) or like a tyre inner tube (i.e. circular)?

Can you advise on what you want to achieve as there maybe alternatives ways to doing this?

Any particular reason you want to use COSMOS over classical methods for what appears to be a simple problem.

 

[Nella95]

Oh, I was just kinda wanting the practice in COSMOS. The calculation was simple.

-It's like a tire inner tube - you're from England, aren't you?

 

[daveboy]

I'm from Scotland!

I guess your modelling the inner tube alone with a pressure on the inside and no outer wall support from the tire etc and if you apply a constraint to the model it doesn't work correctly and if you don't apply any constraint then you get messages something like it is under constrained and static analysis failure?

If this is the case then if you apply weak springs then you can get it to work. You need to take a bit of care with the results as you are constraining the model and verify it but I tried doing this and the results are within a reasonable degree of accuracy.

If this is not the case if you let me know how you are constraining the model.

 

[Nella95]

Weak spring, eh? OK - done.

 

[corus]

No need for springs. This is an axisymmetric problem. For a toroidal shape also use symmetry about the plane that the tube lies in. Alternatively simply assume that you have a pipe and the problem is purely axisymmetrical and a model of the pipe though the wall thickness will do with an internal pressure and symmetry restraints in the axial direction. I wouldn't be sure of the analytical solution though if you're using properties of rubber.

corus