Averaging of driving force

[Nils]

Hi there,

I was looking into the parameters on averaging of the driving force across the interface in the 'Phase interaction section', (avg max smooth) and wondered:

i.) what they mean exactly
ii.) how to set them properly?
iii.) Is it useful to use different setting for different simulation (dendrites, eutectics, solid-state)?
iv.) And do they affect the (efficient) anisotropy of the interfacial properties?

Thank you very much for any help and guidance provided.
Cheers
Nils

[Bernd]

Dear Nils,

you are raising a difficult question about these "dG options" which are all "dirty" parameters in terms of a scientifically pure phase-field model. I already tried to explain part of your query in this thread, maybe you have already read it.

The question which remains open is how to use them properly. For the averaging length avg, my personal default is 0.5. Averaging can sometimes cause trouble in solid-solid interfaces because of the slow diffusion. Then I turn it off by setting it to 0.

The maximum allowed driving force max is cutting off the driving force to a maximum absolut value in a smooth way. I use it always, for 3 reasons:
1.) While using TQ coupling, there is always a risk of getting local numerical instabilities which can result in "astronomically" high values of the driving force. These, in turn, produce astronomical increments of the phase-field variable, and such additional trouble. This is a good reason to set a (less than astronomical) limit to the driving force!
2.) The DP_MICRESS output of the driving force (which I always use to check the numerical stability of a simulation) gets nicer (without the need to rescale), if the maximum driving force is in the same order of magnitude as the majority of the values.
3.) Sometimes, (in the initial transient or after nucleation), the intermediately forming microstructures cannot be properly resolved at a given scale. Limiting the driving force gives a chance to reduce numerical trouble without having to change the phase-field mobility.

"smooth" is a "direction noise" on averaging which I believe can be helpful for reducing the effect of grid anisotropy. There is not much experience with this optional parameter, so please find out yourself whether this is true (and tell us later!).

Bernd

[Nils]

I am interested in the smooth parameter. I understand that this is an angle, how is it defined? Grad phi +- angle? Is then the whole vector, along which the driving force is averaged taken along this, or is the direction of the averaging re-evaluated every dx?

From my simulations is appears that the smooth parameter has a big effect up to 90, after that no further changes are seen. Is there hardcoded maximum value?

lots of questions (and I have more...)
Nils

[Bernd]

It is like you say: the local phase-field gradient is randomly rotated with a maximum amplitude of the given angle (in degrees), in 3d along three axes. This procedure is done for each interface cell independently, i.e. for each cell a complete path through the interface is calculated.

There is no limit on 90 degrees, but 90 already means (in worst case) averaging along the interface instead of across, what can you do more ?

Don't work all night!

Bernd