cracking analysis OF a 3-points bended sandwich beam in the negative stressed area of the beam (above the neutral axis)

Hi,all

I'm looking for ideas-ways to model the above mentioned thema and solutions to overcome some obstacles. Here are some observations of mine:

1) Above the neutral axis of a bended beam the stress field is negative.Furthermore the traditional type of K factors estimation in Ansys is insufficient.The reason as i believe is that we haven't relative despacements of the crack nodes along  y(normal) axis.In contradiction we have shear desplacements.So i retrieve Ki nearly zero and Kii some noticable results.(using contact elements)

2)calculations using J integral seems to agree with the above results.

3)calculations using the V11 of ansys of J integral estimation (without using singular elements around the crack tip ) and using some included algorythms also agree.

4)I have observed that there is a penetration between the crack edges and i used contact elements,along the crack but i am'not sure  about the results.(Eventhough in some cases I take Ki=0 AS I THINK It should.)

Everyone's opinion is welcome

Best Regards

Ilias