Fuel cell modeling

[fth987]

Hello,

What should the mesh strategy be when performing a CFD study for a fuel cell? For example, can a non-conformal mesh approach be used for all the geometries inside a fuel cell? To what extent can the fuel cell geometry be simplified?

[Aleksey Gerasimov]

Hello Fth987,

The non-conformal interfaces can be used, the only restriction is not to place them in the catalytic layers. Typically, three main options exist:

1) NCI between the gas diffusion layers and flow channels & bipolar plates.

2) In the midst of GDL(s)

3) In the middle of membrane.

We have a workshop available on Ansys Learning Hub for the placement of non-conformal interfaces in PEMFC, which uses the first method. Regardng the choice of the actual meshing strategy: This is can be design-specific. It is difficult to give a specific advice without knowing how the design looks like and what you are trying to simulate exactly. If you have access to the ALH, then you are able to find some relevant tutorials and best practices.

Hope this helps,

Aleksey.

 

[fth987]

Thank you very much for your response, I am very pleased. Can we say the following for the second and third options?

 

When the GDL consists of two different structures, namely macroporous and microporous layers, an NCI can be defined between these layers to model the physical properties of each region separately

 

When the membrane is divided into two distinct regions, such as the upper and lower halves, a non-conformal interface (NCI) can be introduced between these regions to independently model the physical properties of each section

[Aleksey Gerasimov]

Hello Fth987,

You can place the non-conformal interface (NCI) between the gas-diffusion layer (GDL) and microporous layer (MPL). This should not create a problem if the mesh scales are consistent on both sides. The only aspect to think about is that the properties of both these regions are different in terms of porosity, permeability & wetability. This is something that affect the transport of species and water management in the cell. An introduction of the NCI, can also introduce numerical issues that might become apparent on this interface too. In other words, if we place the NCI there and then observe noticeable differences in species concentrations, temperature etc, then it becomes difficult to assess whether this is due to materials or whether this is a numerical artifact.  For this reason, my personal choice is to place the NCI right in the middle of the regions so that all the properties are the same on both sides of the NCI. This way can show me if further action is needed to reduce NCI-related issues or not. However, if the NCI have facets of similar length scales on both sides, then the numerical artifacts should be small.

Regarding the membrane: typically membranes are made of the same uniform material and there is no problem to slice them in the middle with the NCI.  There are some designs where a reinforcement material is added, but this is typically done for structural integrity purposes, and we do not accout for their presence explicitly. Another advantage of the mid-membrane split is that, fewer equations are being solved there, in comparison to MPL or GDL or GDL with channel/BPP interface regions.

Hope this helps,

Aleksey.