🐛[BUG]: MHD_PDE equation "B_grad_u" uses v * Bx.diff(y) instead of By * u.diff(y) #2021

Version

PhysicsNemo 2.1.1

On which installation method(s) does this occur?

No response

Describe the issue

examples/cfd/mhd_pino/losses/mhd_pde.py L58:
    self.equations["B_grad_u"] = Bx * u.diff(x) + v * Bx.diff(y)

(B·∇)u should be Bx * u.diff(x) + By * u.diff(y). The next line (B_grad_v, L59) follows
the correct pattern, as does the hand-written version in losses/loss_mhd.py L539.

Impact: none for the shipped examples as far as I can tell. loss_mhd.py recomputes the
term correctly, and the *_physicsnemo.py losses only evaluate computations 12–24
(B_grad_u is index 2). Anyone reusing MHD_PDE.equations["B_grad_u"] gets a wrong term.

Suggested fix:

•    ```
  

self.equations["B_grad_u"] = Bx * u.diff(x) + v * Bx.diff(y)

•    self.equations["B_grad_u"] = Bx * u.diff(x) + By * u.diff(y)
  

Found by checking MHD_PDE against an exact decaying Alfvén wave; all other terms
(u/v/Bx/By/A right-hand sides, div constraints) match it.

### Minimum reproducible example

```shell
cd examples/cfd/mhd_pino/losses
python -c "from mhd_pde import MHD_PDE; print(MHD_PDE().equations['B_grad_u'])"
# prints Bx*u__x + v*Bx__y  — expected Bx*u__x + By*u__y

Relevant log output

Environment details