Strain energy

[albert1981]

Hello everybody,

I have a deformated shell with a certain strain energy and a certain disposition of stress. Since is simmetric respect to both axis I model just a quarter of a plate putting th symmetry on the other part of it.
To simulate a crack I release some elements in the simmetry line in order to get them free. In this case the strain energy should be the same? if so why the free elements simulating the crack change the stress towards 0 while the other doesn't change at all?
I'm asking if there is some way to to fix the strain energy in order to have even the other stress redistribuited.

thanks,
Albert

 

[xerf]

If the crack propagates the new portions of the crack faces (resulted through node release) should be traction free if the material is elastic. If the material deforms plastically then there'll be a "plastic wake" and the stresses will not go to zero but rather to some residual values.

The strain energy is a history quantity and you can alter it only by modifying the state of stress or elastic strain in the body.

 

[albert1981]

xerf,

I have an elastic material so the traction free is correctly 0 but do you know why the other stresses are not influenced by imposing a crack? I think that the stress distribution will change not only on the crack introduced but also in the rest of the plate.
And how can I change the state of stress or elastic strain in a shell once I already have a state of stress on it?

thanks,
Albert