Simulation of cement ingress into sand

Hi all,

Apologises if this is not the correct software for this task.

The task I am trying to solve is that I have a volume of silica sand that has known properties of the particle sorting and sizes. I am then forcing a cement and water mix into the volume at a set pressure, again with a known particle sizes and viscosity.

I need to simulate where this cement reaches within the volume.

In field experiments we have observed that this is dominated by diffusive behaviour and dependent on the sand saturation the result is a sphere. However occasionally we see behaviour dominated by significant volumes of sand moving away from a ‘spur’ of high pressure cement. It is this behaviour I would like to simulate as we need to understand the causes of this.

Firstly is LAMMPS the correct package to use with this simulation? I’m not a materials scientist by trade, but understand some of these simulations methods from my work as geophysicist. Where would I start for a simulation like this?

Thank you for your time,

I think before looking for a specific software to do your simulations, you need to find a collaborator that has experience in that field and can advise you. It is not so much the software that is important, but how you use it, where and how you get the parameters and settings to properly represent your materials, and what are good strategies to obtain useful results.

Two questions you should ask yourself are: can the phenomenon of interest be modelled with Newtonian equations of motion for a reasonable number of particles, where the forces between particles can be written in terms of a short-ranged central force between particle pairs?

And can that phenomenon be observed within a reasonable number of integration time steps, bearing in mind that the larger the time step, the more inaccurate the integration of the equations of motion will be?

If the answer to both are yes, then there is a decent chance you may be able to model your phenomenon as a suitable LAMMPS simulation. My gut feeling is that you would be better off with some kind of finite element modelling. But, of course, my gut feeling is not a reliable source of academic information. (: