Can this kind of problem can be solved in Composite shell elements?

[pcomp]

I want to analyze a carbon fiber structure made from 350cure graphite/epoxy unitape. The geometry is a very simple T shape bracket. It is a simple square flange and a vertical rib on it.

Here is the lay-up. The carbon fiber lay-up is made, such that two L shape brackets are co-cured along the plane (half thickness) of rib. Additional pre-pregs were added in the top surface of the flange to increase the flange thickness.

Here is the loading condition. The top surface of the flange is bonded on a rigid fixture. A load is acting on the rib which is in normal direction to flange. The load is trying to separate the rib from flange. I feel that the failure will be simple bond separation in the flange layers. Can this kind of problem be solved using PCOMP card with 2D elements? How can I constrain (SPC) only top most fiber in the flange, so that other layer in the flange can have displacement? Is 3D brick element layers are only the option to study this kind of problem?

 

[SWComposites]

If I understand the configuration correctly, this is a T-section loaded in a pull-off manner. Modelling it with shell elements will not work as you cannot SPC the top layer, you can only SPC nodes. You need to either use a 3D solid model, or a 2D plane strain model in the cross-section plane. Also, using a stress/strain failure criteria to predict failure will not work. The failure mechanism is interlaminar delamination/debonding, and only fracture mechanics type approaches have been shown to work for predicting strength of this type of part. In the FE model you can use VCCT methods or cohesive type interlaminar elements to perform the fracture analysis.