The Influence of Bubbles on Foamed Cement Viscosity Using an Extended Stokesian Dynamics Approach

Abstract

We want to study the influence of bubbles on the viscosity of suspensions with a computational approach that also accounts for the arrangement of the bubbles due to shearing flow. This requires a large number of bubbles to properly simulate and requires a large amount of computational resources. Here we develop a set of equations to define the viscosity ratio from the simulation results to show the influence of the bubbles on the viscosity as a function of the volume fraction. One application of this work has been used to study a specific type of cement that has bubbles injected into the slurry while it is still fluid. The bubbles are added to reduce the density but they also improve the properties of the cement with the increase in viscosity. We show that the computed results match the few experimental results that have been reported.

A Review of Rheological Modeling of Cement Slurry in Oil Well Applications

“…Murphy et al [295] simulated the shear flow of two Bingham plastic cement slurries containing Portland cement and fly ash particles by applying fast lubrication dynamics-discrete element model with LAMMPS. Rosenbaum et al [296] studied the influence of bubbles on foamed cement viscosity using an Extended Stokesian Dynamics Approach with LAMMPS.…”

Recent Advances in Mechanics of Non-Newtonian Fluids

“…Rosenbaum, et al [26] study the influence of bubbles on the shear viscosity of foamed cement; the cement slurry is modeled as a fluid suspension and in their numerical simulations, they use an extended version of Stokesian Dynamics along with Fast Lubrication Dynamics. Bubbles are introduced/injected into the cement slurry to reduce the density, increase the viscosity, and improve the properties of cement.…”

Displacement Flow of Compressible Non-Newtonian Fluids in Annular Geometries for Well Cementing Applications

“…In this work the function described in Eq. 3-22(Rosenbaum, 2019) was used in the simulations.𝑓(𝜑) = 1 + 1958𝜑 + 15810𝜑²(3-22)The choice of the models proposed byDucloué et al (2014) and used by Hanachi (2018) and Rosenbaum et al (2019) in the studies conducted in this dissertation comes from a relatively straightforward way to adjust the viscosity model of energized fluids with variation of the gas fraction. However, any other correlation 𝜂(𝛾, 𝜑(𝑝, 𝑇)) can be implemented if sufficient data is available.…”