I have a list of questions regarding a fracture analysis of a bonded, concentric cylindrical specimen, with a predefined crack.

Please see attached file and below screenshots for context. The crack exists at the boundary between the first (innermost) and second layer of the concentric cylinder. One end of the cylinder is fixed, while internal pressure, torque and a tensile axial load are applied.  

  • Are cylindrical coordinates prohibited in fracture mechanics? Defining the crack front using a cylindrical coordinate system would make sense, because it will most likely propagate tangentially around the bonded interface ... but ANSYS will not allow me to do so...
  • Should crack fronts be applied to both edges of the internal crack, with two opposing CS's? I have seen examples defining the crack CS on the face of the crack instead of the edge(s), but I believe that would dictate the crack to move in only one direction.
  • Is there a method for evaluating the resulting penetration in an analysis? My results seem to indicate significant penetration but I cannot tell is this is an artifact of scaling, or if penetration truly is occurring...
  • I want to eliminate or minimize penetration between the 3 concentric layers... What is the best method for doing this? I have tried placing limits on the allowable penetration in the contacts, but I still see penetration in my results. 
  • Lastly, I think this model favors quarter symmetry. Would it be preferable to set the crack as a pre-meshed edge crack in this case? (as opposed to an internal crack?) I am hoping to try this after I get the full model running...

 

Combined Loading:

Pressure Only:

Tension Only:

Torque Only: