Dear all,

I am trying to build a simulation aiming to describe the behavior of a dense colloidal suspensions made of silica particles.

I defined a DLVO potential (pair_style yukawa/colloid+colloid) for what concerns the conservative inter-particle interactions and I chose a non-inertial Fast Lubrication Dynamics (nFLD) scheme (pair_style brownian and lubricateU) in order to describe the hydrodynamics interactions. In such configuration, since the net force on the particles is zero (we assume no inertia), as reported in LAMMPS manual (pair_style lubricateU), we solve the following linear system in order to get velocities and angular velocities (vector U) of the particles (see Bolintineanu, Dan S., et al. , 2014 for further details):

-R_{FU}(U-U^{infty})= -R_{FE}(E^{infty})-F^{rest},

where (in my case), F^{rest} accounts for forces and torques relative to the DLVO potential and the Brownian dynamics.

What I would like to do now, is to add also forces and torques deriving from effective particle-particle contact.

Thus, I thought to define also the pair_style granular in order to have Hertz contact forces +adhesion+rolling resistance etc. But here it comes my problem:

First of all, I saw that LAMMPS automatically generates an "overlap error" (defined in pair_style colloid) if the center-to-center distance b/w two particles (r)

becomes smaller of the contact distance (d =2r_p, where r_p is the particle radius. We have particle of the same type here).

But r<d is the precise condition to have contact and no contact forces can be taken into account if r<d is not allowed !

Second problem is relative to the damping terms of the contact forces.

If for example we consider the normal damping force, i.e. F_{n,damp}=-C_damp U_{n,rel}, we have a force that is function of particles velocity

since U_{n,rel} is the relative component of the normal velocity b/w the particles in contact. It follows that we need to modify the coefficient matrix

(R_FU written in the balance of forces at the beginning) in order to account for this term. This procedure needs then to be repeated for all the damping

terms included in the contact forces (i.e. the damping term of the tangential contact force, of the rolling resistance torque etc.)

Is this done in LAMMPS? Can I simply define the pair_style granular and the coefficient matrix of the hydrodynamics interactions is automatically

corrected ?

Thank you all in advance,

all the best

Gianluca