Troubleshooting

Issue: checkerboarding (grid scale random oscillations) in equilibrium runs (usually open dipole grids)

Possible appearance:

Possible fixes:

1. Decrease resolution (increase grid viscosity and may wipe out artifacts)
2. Increase artificial viscosity (xicon in phys_consts.f90). This could impact stability so needs to be done with dare
3. Recommended: switch to first-order solver (for eq only; I've never see this to be necessary for disturbance runs).

In multifluid.f90 replace this:

! param=backEuler3D(param,A,B,C,D,E,dt,x%dx1,x%dx1i) !1st order method, likely deprecated but needs to be kept here for debug purposes, perhaps? param=TRBDF23D(param,A,B,C,D,E,dt,x)

with this:

  param=backEuler3D(param,A,B,C,D,E,dt,x%dx1,x%dx1i)    !1st order method, likely deprecated but needs to be kept here for debug purposes, perhaps?

! param=TRBDF23D(param,A,B,C,D,E,dt,x)

Issue: Discontinuity in velocity fields with the use of open grids in equalibrium runs.

Possible appearance:

Possible fix:

1. Change the size of the grid. For example slightly increase or decrease thetatd in grid creating script (e.g. makegrid_tilteddipole_3D.m)

if(thetatd<pi/2) %northern hemisphere thetamax=thetax2min+pi/25; % initial - 25 else thetamax=thetax2max-pi/25; end

Issue: Occasionally MUMPS will throw an error because it underestimated the amount of memory needed for a solve.

Possible fixes:

If this happens a workaround is to uncomment (or add) this line of code to the potential solver being used for your simulations: mumps_par%ICNTL(14)=50 If the problem persists try changing the number to 100.