Non-Newtonian Flow


Your product looks good from what I can see. My current problem involves non-newtonian flow, Bingham, pseudoplastic and some other variations. I haven't seem anything on the site that mentions whether or not it handles non-Newtonian flow. If so, I'd like to give it a test drive.


[Edit: based on an email exchange]

Viscosity Model

I'm not an expert on non-newtonian flow and we don't support the Bingham model. However, we do support the 'Bird Carreau' and 'Cross Power Law' non-newtonian viscosity models, if that's any help. To change the viscosity model:

  1. Drag and drop a substance from the Physics Tool Palette onto your flow volume
  2. In the Properties Panel select the Volume tab (far left), if it isn't already selected
  3. Set the Property Substance->Properties->Transport->Viscosity Model = Cross Power Law | Bird Carreau

User Defined Viscosity Models?

Those viscosity models are helpful, but I'm interested in some other ones as well. If you support those viscosity models is it possible for me to define my own? The "usual" Chemical Engineering technique is to define a shear rate dependant viscosity, which is what I imagine was done for the Carreau and Power law models. So is it possible for me to do the same: grab whatever parameter is used for the magnitude of the shear rate tensor, and use it to write a relationship for the viscosity and then let whatever non linear engine grind away on it, much like it does for the other non-Newtonian fluids?

The test case I would like to try is pretty straightforward: a cylindrical tank with a hole in the bottom, emptying under gravity. I'd be looking for a steady-state solution using some open boundary conditions on the top surface and outlet of the tank, so idealizing it as if I am constantly replenishing the tank in a way that the surface remains level.

As a minor issue, I see Caedium lets me look at stream lines. In the model tank problem, once in steady-state, I'm interested in the residence time distribution in the tank, so integrating along each stream line to find how much time a particular part of the fluid spent in the tank. Then, going further (not necessary for the moment), the rheology parameters are actually time dependant, so I would want to model them with how long a particular fluid element has been in the tank. I can imagine fudging this by defining a fictitious chemical concentration that starts at a constant value entering the tank with a simple linear decay rate, the using the concentration at any point as a surrogate for time. Anyway, would Caedium have either the native facilities to look at time along streamlines or allow me add a continuity equation to simulate chemical reactions along the way?

Use OpenFOAM Directly

It sounds like you have a good knowledge of CFD. I don't think we'll have these capabilities any time soon in Caedium. However, Caedium uses OpenFOAM as its RANS Solver, so you might have more fun using OpenFOAM directly. If you need custom code development services then I can pass your contact details on to OpenFOAM development experts if you like?

Might Give Caedium a Spin

Thanks. Not sure I want to go the developer route at the moment. If I have some time, I may give Caedium a spin anyway.