Hi everybody,

for learning purpose, i try to re-implement the known HGO-Model in Abaqus using UANISOHYPER_INV. This is the first time i use the invariant subroutine for anisotropic hyperelastic materials but I have some experience with UANISOHYPER_STRAIN.

What I did:

- I made a little test-model of 8 cubes, each one consisting of only one Element (C3D8).

- 4 of them are associated with the original HGO material and 4 with my homemade material, so i can compare the resulting material behavior optically as a first evaluation.

- The 4 cubes of one particular material have one preferred fiber orientation each (represented through vectors (1,0,0), (0,1,0), (0,0,1), (1,1,1) ).

- the BCs hold the cubes in position while the constrained nodes can shift in the corresponding plane.

- the 4 top (max z) nodes of each cube are moved half the side lentgh in z-direction

--->   SCREENSHOT   <---

When i run the model, no errors occur, but the results look pretty strange:

- the 4 cubes with my material simply shrink (same amount in all directions!!), but generate no stress in response, while the cubes with original HGO behave like expected. These are the right 4 cubes in the screenshot, the left cubes are are associated to the original HGO material.

- while debugging with VS2005, my routine seems to calculate everything as it should. I check that by modifying the values for Ainv() and compare the results with corresponding values out of Mathematica

- strangely the input values for Ainv() always stay the same with the exception of shrinking Ainv(3) = volume ratio J. (values of reference configuration or rather zero deviatoric deformation)

Anyone here with an idea, what i could have missed?? 

Thanks in advance for every answer!

Christof