Calculating Bearing Life without Fe-Safe: possible?

[Fabi0]

Hi all

I'm dealing with the problem in object: usually we use standard method (ISO) to estimate L (number of revolution before collapse), but now we are looking for a better solution.

I'm totally out of concepts about fatigue analisys in abaqus, does anybody can provide something? The boss idea was: could we try to simulate life of a working bearing, so that the model itself collapse when the crack grows up? Is it possible or it is science fiction?

I can't imagine how make it...

THX all

Fabi0

 

[Fabi0]

Nothing at all?

Fabi0

 

[corus]

You can model a crack using crack tip elements to calculate the stress intensity factor with J integrals and so on. From that you can calculaute the critical crack length to cause failure. You can also model crack growth but for that case you need to know the crack growth path beforehand. You should find lots of examples in the manuals, if you have them. For a bearing though, I'd have thought you'd be interested in crack initiation and that would be a simple comparison between calculated Mises stress and the fatigue property of the material, factored for surface finish etc.?

Tara

http://tinyurl.com/4ydjg7m