[Aicharem]

Hi All,

I am trying to create an OpenFOAM model for an electrohydrodynamic thruster. I am using the work performed by AA Martins [1] as a base for the model, it involves solving Navier Stoke's equation, the electric field equations and the charge transport equations in three separate modules.

My questions is this: Is it best to modify the existing simpleFoam solver by adding in the electrostaticFoam module, and then manually insert the charge transport equation? I only need the steady state solution.

I have seen the tutorial for adding temperature to the icoFoam solver and wonder if the electric field strength could be added in a similar fashion.

I'm really struggling with this so any advice would be really appreciated!

Thanks
A

[1] Modeling of an EHD corona flow in nitrogen gas using an asymmetric capacitor for propulsion: http://arxiv.org/ftp/arxiv/papers/1008/1008.0552.pdf

[chriss85]

You can add additional fields and equations similar to what is shown in that tutorial to a solver. You will have to figure out if you need to couple these additional equations inside a (pseudo-) time step to archieve convergence and stability. Relaxation might also be needed, depending on the modelled system.

[Aicharem]

Thanks for the pointers Chris, I have been working on doing this for a couple of weeks now, but I have hit a wall.

I have combined the electrostaticFoam with simpleFoam with a few modifications. The solver compiles successfully but when the case runs I get a "floating point exception (core dumped)" error.

I have attached the solver I am trying to run, I'll also post the case. It is a modified version of the airFoil2D tutorial case. If anyone has some time to look at it and see if I've made any obvious errors it's be a real help.

One of the things that I am unsure of is the modified UEqn.H, where I have added in a new term for the electrostatic force. The other thing that I haven't figured out is whether or not I need to include the d/dt terms since the solver is a steady state one.

Thanks
A

[Aicharem]

Here is the case I am trying to run.

The properties folder has the files that belong in the constant folder. I couldn't upload the entire constant folder as the polymesh folder within it is too big, however this is unchanged from the original case.

A

[chriss85]

Floating point exception is often caused by division by zero. Make sure none of your physical properties are completely zero. For example when I want to simulate a resistive material (laplacian(sigma, phi)=0), sigma must not be completely zero.

[Aicharem]

Thanks, I have checked over my physical properties and transport properties but I don't think these are the problem. None of them are zero, I have also tried moving them all into transportProperties (previously the electrostatic properties were defined in physicalProperties and rho and nu were defined in transport properties. This seems to have had no effect, positive or negative).

If you could provide any further ideas or guidance it would be hugely appreciated. I have to submit my project progress next week and so far I have extremely little to show for it....

Thanks,
A

[chriss85]

You could try to use different equation solvers, sometimes this helps as well.
Oh and if you have a steady state solver you can just set the ddt schemes to steady (or steadyState?) in fvSolution.