CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   FLUENT (http://www.cfd-online.com/Forums/fluent/)
-   -   How to simulate electrostatic in dielectric (http://www.cfd-online.com/Forums/fluent/105647-how-simulate-electrostatic-dielectric.html)

danieldmgy August 5, 2012 10:24

How to simulate electrostatic in dielectric
 
Hi everyone, I am trying to make a electrostatic solver in fluent, but I am not quite sure how could I implement that.

Now here is my situation. I have dielectric material and air flow, which is a multiphase flow system in Fluent. If I add voltage to the boundary, the electrostatic field are suppose to be uniquely determined. Laplacian equation is applicable to both dielectric and air because there is no free electrons.

The problem is the interface that between the dielectric and the air needs to satisfied by Gaussian pillbox which are:
1. Tangent E in dielectric= Tangent E in air
2. Normal E in dielectric * permittivity =Normal E in air

How can I implement that in UDS or UDF?

Right now I see some paper mentioned that UDS(voltage as scalar) could be a solution. I checked FLUENT Theoretical Reference, that could potentially be erroneous because FLUENT arbitrarily impose the interface based on diffusion flux not on Gaussian pillbox.

If I use UDF, how could I quickly get started?

Thanks a lot for you opinion on my case. I really appreciate!

mohammadkm August 5, 2012 12:32

I don't have experience in modeling the dielectric flow. But I think you should use UDF_source and add the quantity as a source to the navier-stokes.

danieldmgy August 5, 2012 13:47

1 Attachment(s)
Quote:

Originally Posted by mohammadkm (Post 375464)
I don't have experience in modeling the dielectric flow. But I think you should use UDF_source and add the quantity as a source to the navier-stokes.

Hi mohammadkm, thx 4 ur reply!

I am not quite sure whether this UDS approach ( adding a source term, here, charge density Q, in combination with voltage as a scalar, e\nabla^{2}U=-Q) works in this case.

From Chapter 17.3.4.3 in FLUENT Theory Guide: "ANSYS FLUENT’s control-volume formulation requires that convection and diffusion fluxes through the
control volume faces be computed and balanced with source terms within the control volume itself", which means FLUENT will make e*\nabla ^{2} U_{dielectric}= \nabla^{2} U_{air}

However, the dielectric interface here, it seems the boundary condition in the image cannot be fulfilled:confused:

What do you think?

Ganapathy August 6, 2012 00:16

There are papers which deal with ElectroStatic Force on particles as implemented in Fluent.

You could use that approach.

You would need UDS as well as UDF.

Ganapathy August 6, 2012 00:21

http://www.electrostatics.org/images...4_Farnoosh.pdf

danieldmgy August 7, 2012 10:02

Quote:

Originally Posted by Ganapathy (Post 375511)

Thanks Ganapathy!

Frankly my understanding about the particles is that they do not require a dielectric-air interface because they are treated as charge dispersed in air flow. So I guess it is not applicable to my case.

Right now I have already get started to write my own electrostatic solver customized for this boundary condition, I think it should work even though it's the worst scenario


All times are GMT -4. The time now is 02:17.