Today I got a rather tricky question from a manufacturing expert. They are looking for a tool to help them to assess what happens when you dip a rather complex component into a liquid.

They want to be able to assess how the liquid fills all holes and cavities in the component. How the component should be turned and twisted to ensure filling everywhere, and then as the component is lifted up, how the liquid drains from all holes and cavities in the component.

Not exactly the normal kind of fluid-dynamics problem for a turbomachinery aero guy like me. Anyone have any ideas? Are there tools out there to do this kind of assessment? The geometry of these components are usually quite complex.

Typical applications are dip-painting and treatment of casted components with various etching liquids etc.

Jonas,

It'd be worth a call to C. W. (Tony) Hirt, the founder of Flow Science Inc (Santa Fe, New Mexico, US) and originator of Flow3d.

In addition to developing the original VOF method, Hirt has had a long interest in the effects of surface tension, filling and emptying simulations, etc.

More info at www.flow3d.com.

BTW, I'm not connected to the company, derive no income from them (although my company once paid license fees to use the code). I just think it's a good CFD code for many purposes - this is perhaps one of those purposes.

To my knowledge, this is quite a typical application for free surface solvers - you need a surface capturing method and, depending on the properties of paint, some additional bells and whistles, like contact angle handling or non-Newtonian viscosity laws.

People who already solve this kind of problem using CFD are the automotive companies - painting of a car body is done in the same way as you describe.

As for the choice of solvers, almost any surface capturing method will be OK because the scales are such that surface tension does not play significant effects.

Hrv

Yea, I guess I could use an advanced CFD code with surface tracking etc. Feels a bit like using a chainsaw to pick flowers with though.

I was hoping there was a simpler more specialized tool to do this kind of assessment. The manufacturing guys aren't really interested in the exact flows. They just want to make sure that the liquid will fill all hole in the component as it is dipped and that the liquid will drain out nicely as the component is lifted. And if they identify problems they'd like to be able to experiment with various turns and twist to of the component to get the liquid in and out.

The tool doesn't necessarily have to be a flow solver. It could be something that just assesses the topology of component in a smart way. I dunno. Thanks for the feedback anyway. Perhaps a good opportunity to try FOAM Could FOAM do this kind of simulation btw?

Good suggestion. They sure have a lot of experience from free-surface simulations. Is Tony Hirt still working btw? I thought he was retired.

The problems with coating are things like air bubbles and velocity fields on the surface - therefore, if you decide to only simulate the surface you won't pick up these effects. Simulations of this type I know about are done by modelling a tank of liquid and then dipping the object (maybe modelled as a shell, depending on the shape) into in.

There may be some simpler tools for coating extrusions and dip-coating smooth profiles, but you will probably need to postulate that no air pockets are being created. In this case, you can afford only to solve for a liquid film on the surface.

Foam can easily do this kind of thing; there may be problems with mesh generation, depending on the motion of the object and its shape, but a combination of mesh motion and sliding interfaces (say, for twisting the object in the dip) should do the job.

Hrv

I agree that meshing is probably the main problem. Perhaps one could use one of those cartesian mesh teqniques that some people use to simulate engine blocks etc. I'm thinking of tools like harpoon or hexpress. Would a coarse cartesian mesh without any boundary fitted cells produce anything useful? Or would a coarse auto-generated tet mesh produce anything useful?

The objects I'm thinking of are quite complex in shape - think of a car wheel and add 5 times as many cavities, spokes etc.

I was hoping to be able to avoid meshing this thing manually.

Gee Jonas, I don't know if Tony's retired.

There's a picture of Tony on the website, chatting up a client and enjoying a libation. Maybe that's what retired folks do : ) .

Thought I'd add some more background info... The application I'm interested in is not coating or painting, it is etching of casted components. Hence, small air-bubbles are not important, neither are how the liquid surface looks or behaves in detail. What is important is that you do not get pockets of air inside the holes and cavities of the object where the liquid does not reach, and, as you lift the object, that the liquid drains out of all cavities.