The problem of the coordinate system

[Pineapple-blowing-snow]

I’d like to ask a question: I’m reproducing a model from a paper where parameters 11, 22, and 33 correspond to the ([11-20]), ([-1100]), and ([0001]) directions of α-Zr, respectively. In the 3D case, x, y, and z also map to 11, 22, and 33, respectively. Now I want to convert the model from 3D to 2D and examine the plane formed by ([11-20]) and ([0001]). Can I directly model on the X-Z plane? Will MOOSE still default to the X-Y plane? The material parameters involve some complex rotations, and I don’t want to modify them. Thanks.

[GiudGiud]

Hello

2D is XY in moose, not a general two D plane.

I would model a 3D slab with reflective boundary conditions in XZ if rotations to XY are complicated

[Pineapple-blowing-snow]

Thank you very much for your helpful suggestion — that clarifies things a lot.

Just to confirm my understanding: would it be appropriate to model a 3D slab where the thickness in the y-direction is set to be very small (e.g., one element), so that it effectively represents an XZ cross-section?

Thanks again for your guidance!

[GiudGiud]

I think that would work and the computing cost would be commensurate (like 2x) with the cost of a 2D simulation