Implémentation of capillary force during mechanical contact

Hi,

What is your exact problem with ViscElCapMat ? (and what is your sign convention for your delta_n ?)
Is it that the case of strict contact is excluded at [*] ?

An easy solution for that would be to add a "=" after the "<" in that source code line, and recompile.

Other than that, note that you have CapillaryPhys and related classes (see examples/capillaryLaplaceYoung/capillaryBridge.py [**] for an example) that will also compute for you and apply capillary forces, contact case included.

[*] https://gitlab.com/yade-dev/trunk/-/blob/master/pkg/dem/ViscoelasticCapillarPM.cpp#L138
[**] https://gitlab.com/yade-dev/trunk/-/blob/master/examples/capillaryLaplaceYoung/capillaryBridge.py for current YADE version

> the Capillary force in ViscElCapMat is only defined for delta_n > 0

I don't think it is the case, else it would be a mistake in the code.
I think automatic _initialization_ of the bridge in that class is only when there is contact.
But after a bridge is created it will (should) persist even if delta_n < 0.
B

Actually, I'm realizing my above link [*] is exactly about this case: when particles go apart after they were in geometrical contact (hence showing a liquid bridge). So there probably is nothing to change there, and my answer was not really helpful, sorry about that..

...Thank you very much for your replies..

For ViscElCapMat (*), the point is that A Gladky and al in his work (Comparison of different capillary bridges model for application in the discrete element method: https://arxiv.org/pdf/1403.7926.pdf) precises "We are neglecting effects of capillary forces during mechanical contact because we are not aware of a proved model for that case.... and the capillary forces are in order of magnitude much smaller than the mechanical ones".

For Capillaryphys (**), Luc Scholtès in his thesis (p.62) explains that the capillary force during penetration delta_n >0 is kept constant equal to its maximal value Fcapmax (value of capillary force at contact delta_n = 0) (and supposing small penetration distances compared to radius)....

So my point is that a value for the capillary force may be calculated in case of penetrating particles for both formulations but I have to see if the specific geometry of penetrating particles is correctly taken into account by the implemented models, I would like a more reliable calculation of the capillary force in this case....
???

All the best

Vincent

Thanks Jérôme Duriez, that solved my question.

> So my point is that a value for the capillary force may be calculated in case of penetrating particles

I fully agree. From a theoretical standpoint there is no need for any empirical force model, it is enough to integrate Young-Laplace and that integration doesn't depend on the fact that there is or isn't a contact.

In Law2_ScGeom_CapillaryPhys_Capillarity1 the force is a smooth function of distance, regardless of the sign of of delta_n.
If delta_n>0 it keeps changing (unlike in Scholtès), by neglecting the deflection near the contact region, which is ok for larger bridges.

I realize only now that this algorithm has been left behind with the obscure name "Law2_ScGeom_CapillaryPhys_Capillarity1".
It should be improved...
Cheers
Bruno