Non-constant orientation of orthotropic material

I'm trying to model a curved shaft/handle, made by laminating layers of wood. The layers are aligned but the direction of the grain follows the curve of the shaft. It has non uniform cross section so I believe I need to model it as a solid rather than shell? At the moment I'm modeling it as massif with the material directions constant. Is it possible to specify a changing orientation for the material?

Any help would be much appreciated.

Thanks
Joe

I've not managed to arrive at a 'proper' solution, so if anyone has any ideas of a better way to model this then I'm still interested.

What I have done is break the shaft into 8 parts and model it as an assembly. Each part covers a curve of 5 degrees so the maximum deviation of the modeled orientation to the actual grain orientation is 2.5 degrees. I don't want to do this repeatedly for each variation of the shaft, but the results are helpful

- the stress distribution between the two models is more or less identical (just one or two concentrations at joins between sections)

- the deformation is much reduced in the model with the multiple material orientations.

As such I plan on continuing to use a single orientation (correct at point of max moment) to check stress levels unless a better method presents itself.