[Novel]

Hello,
I'm trying to simulate a Taylor instability by using the electric potential formulation of the MHD equations (similar to http://www.tfd.chalmers.se/~hani/kur...rt_Tassone.pdf ).

The domain is a cuboid (96x96x120 mm) where the patches inlet,outlet are the sides with the size 96x96 and all other sider are defined as wall. All boundaries are defined as closed domain.
U -> noSlip
p -> zeroGradient
Epot -> zerogradient at the wall, and fixedGradient 0.5... at the inlet at fixedValue 0 at the outlet. This corresponds to a current of around 18kA.
B -> fixedValue where the value is defined with the analytic solution.

The lorentz force is defined as:

lorentz = (J0 ^  B);

the force is added in the momentum equation the following way:

tmp<fvVectorMatrix> tUEqn
(
    fvm::ddt(U) + fvm::div(phi, U)
  + MRF.DDt(U)
  + turbulence->divDevReff(U)
  - (1.0/rho) * (lorentz)
 ==
    fvOptions(U)
);

The simulation runs and the velocity field looks just fine but close to the boundary the velocity increases rapidly and than falls down to zero due to noSlip. I'm wondering if I should add the force term not in the momentum equation or if I need to use other boundary conditions but which one?

In the link bellow I uploaded some pictures where I measured the velocity, pressure and lorentz force along the y-axis.

http://imgur.com/a/uRXP8

Maybe one of you had a similar problem and has some ideas what I could try.

regards,
Roman