|
|
Can you provide a patch to correct this behaviour? |
|
|
|
No, as long as this isn't a straightforward implementation error (which I believe it isn't) |
|
|
|
It is not clear from your description that there is a problem. If a region of low shear is surrounded by a region of high shear there will be a blob of effectively solid material transported by the surrounding flow. The low shear regions would only be stationary if they are adjacent to walls and even in this case the current implementation would not guarantee no motion at all because the behaviour of the solid is approximated by a high viscosity which is sufficient for most purposes of driftFluxFoam. If you need the detailed behaviour of the solid regions to be correct, including residual stresses, you will need to create a completely different solver which includes the transitions between solid and fluid and solves for the stresses in the solid regions. |
|
|
|
I agree with what you described, but this didn't touch the problem.
To be more specific: increasing BinghamCoeff, BinghamExponent or muMax will increase the mixture viscosity and the yield stress up to a specific threshold. Above that threshold the increase of those coefficient has no effect on the mixture viscosity (we can keep increasing mixture viscosity by increasing the plastic viscosity) which seems like a limitation for the current formulation of the Bigham model. |
|
|
|
How would you rather have the Bigham model implemented? It is not clear what you want changed. If you can provide a patch it would help. |
|
|
|
In my opinion allowing increase of mixture viscosity by the mean of increasing the yield stress will yield a "more natural" behavior. |
|
|
|
Please provide a patch and a test-case which demonstrates the benefit of this. |
|
|
|
No feedback from the reporter |
|
