Notes about the Pointwise grids :
   B1-HLCRM_UnstrTet_PW
   B2-HLCRM_UnstrPrismTet_PW
   B3-HLCRM_UnstrHexPrismPyrTet_PW

The various grids are very closely related, as described below.


Tet Grids
  - Triangular surface mesh is created on HL-CRM.
  - Each vertex on the surface mesh is deformed normal to the surface based on wall spacing.
  - Once each of the surface mesh vertices have been deformed; you have one layer of tetrahedra 
    which is essentially equivalent to a stack of three right-angle tets on each surface triangle. 
  - The triangle on top of this layer can then be deformed in the same manner and this process 
    continues until the final right-angle tets in the stack of layers have hit an isotropic state, 
    they have collided with a grid boundary or another front of the grid, or quality criteria have 
    been violated.
  - The remainder of the volume is filled in with isotropic tets.


Prism-Tet Grids
  - Triangular surface mesh is created on HL-CRM.
  - Each vertex on the surface mesh is deformed normal to the surface based on wall spacing.
  - Once each of the surface mesh vertices have been deformed; you have one layer of tetrahedra 
    which is essentially equivalent to a stack of three right-angle tets on each surface triangle. 
  - This stack of three right-angle tets are evaluated to determine if the prism that could be 
    created from them would contain a negative component volume. In other words, Pointwise cannot 
    predict how flow solver might decompose a prism so each of the 12 different ways a prism 
    could be decomposed into 3 tets are reviewed and evaluated.  If any of those decompositions 
    result in one of the component tet volumes being negative, the three tets are not combined 
    into a prism.
  - The triangle on top of this layer (whether it's the top of a right-angle tet or a prism) can 
    then be deformed in the same manner to create a new layer.  This layer is evaluated to see 
    if an acceptable and valid prism would be created by the same process used in the previous 
    step.
  - The stack of layers continue to grow until it hits an isotropic state, they have collided 
    with a grid boundary or another front of the grid, or quality criteria have been violated.
  - The remainder of the volume is filled in with isotropic tets.

Comparing the two techniques, it's not just a brute force combination of cells. There is the 
possibility that there could be a small pocket of tets in a few spots amongst the prisms if 
creating prisms would create negative volume cells.  That being said, there does not appear 
to be very much of this in most grids, and they should be almost identical.  Nonetheless, there 
is still the possibility of a small chunk of tets left uncombined.


Hex-Prism-Pyr-Tet Grids

These are formed in a similar manner, except at each layer we then evaluate if every pair of 
prisms could make an acceptable quality unstructured hex.  

If not, you could have a pocket of prisms sandwiched in the layers but this does not happen 
often.  Once a complete stack of hexes has been created, the surface mesh underneath the stack 
has its two triangles combined into one unstructured quad.  So the surface and volume mesh 
essentially retain the same vertex locations.