Errors in lubricate/poly ?


Sorry for another email addressing the same issue. Maybe I am beating a dead horse here, but another way to highlight the same error in omega calculation in both the lubricate and lubricate/poly pair styles would be to do a two particle shearing example as described earlier and in the attached input script (with one particle at the bottom edge and the other in contact) and increase the simulation box size. Even if one assumes that there should be a force between the particles, it should surely not depend on the box size. However, there is a box size dependency with these lubricate pair styles.

I also want to make it clear that the problem with omega is not the only problem with the current implementation of lubricate pair styles. Another interesting situation is when particles overlap when shearing (it can happen in simulations of dense suspensions paired with granular forces), where one has to be careful about how the lubrication force is calculated. There’s also an issue with the non-symmetricity of the stress tensor due to lubrication force as currently calculated in LAMMPS. Specifically for lubricate/poly, there are other issues that I pointed out earlier.

I can discuss all these issues in more detail if the LAMMPS strategy is to modify the current implementation.

One clarification about one of my earlier comments. I quote from my earlier email.

“4) a.The squeeze term in lubricate/poly is taken from the force given in Eq. 9. 33 of Ref. (1). According to the resistance matrix formulation, the first term in Eq. 9. 33 should be multiplied by a prefactor of “2/(1+\beta)” , and the second term by “\beta” (apologies for the mistake in my previous email). One simple way to see that is that the magnitude of the leading order terms given in Sec. 11.2.2 should be twice of Eq. 9. 33, which is not the case in the textbook.”

The first term between Eq. 9. 33 and Sec. 11.2.2 of Kim and Karrila is consistent, since they scaled the gap distance by particle radius in the first case, however the second term is still inconsistent between the two sections of the book. I think that the Chapter 11 is the correct version in the sense that the terms are consistent with Eq. (3.19a, b) of Jefferey and Onishi, J. Fluid Mech. (1984), wol. 139, pp. 261-290, which is the original research article.

I’ve also cc’ed the LAMMPS mailing list this time, apologies for not doing in my last emails.

Ranga (1.28 KB)