Learning Forum

[Carlos Pfeiff]

Hello, I am trying to simulate a slot die coating process with Fluent. I am using a 2-D domain, VOF model (Transient, Implicit). The boundaries are as can be seen in the first image of the mesh, both are pressure outlets and the fluid has a 10^-3 m/s velocity at the inlet. The mesh size is 10^-5 m and 10^-6 m near the substrate. Time step is 10^-5 s and number of iterations is 10.

I let the liquid bridge form between the die lips and the substrate, and then make the substrate move in the x direction with a velocity of 10^-2 m/s. A wet film starts to form, but the upstream meniscus is dragged downstream to the point where the liquid bridge breaks. If I lower the substrate velocity the liquid bridge overflows.

I believe this is a numerical issue, as I have been running tests on an actual slot die coater with the same parameters and this clearly does not occur; the liquid bridge forms and then is maintained throughout coating. Also, varying either flowrate (inlet velocity) or substrate speed should have an effect on the thickness of the wet film when the parameters are within the coating window (they are), but should not have such drastic changes on the liquid bridge.

Any ideas as to what may be happening are very much appreciated!

[Rob]

That looks to be a convergence problem, possibly not helped by the mesh resolution away from the substrate towards the nozzle. What did you set the wall contact angle as?

[Carlos Pfeiff]

The contact angle on the nozzle walls is 70 degrees, and on the substrate is 20 degrees. I coarsened the mesh away from the substrate as I was getting the floating point error, and I mostly need resolution to solve for the wet film boundary and profile. What I have not tried yet (I am still running another iteration of the model) is make the mesh finer just below the nozzle exit, between the lips and the subsrate. Would you suggest I follow this avenue, and if it does not work, what would be the next steps?

Thank you for your time, Rob.

[Rob]

You need to resolve the free surface spacially (mesh resolution) and with time (time step). Everyone focusses on the former and then won't look at the latter as "it takes too long". Settings for VOF are fairly simple, and problems tend to be mesh/time step related.