Regular Article

Estimating Al₂O₃–CO₂ nanofluid viscosity: a molecular dynamics approach^★

¹ Mechanical Engineering, IIT Gandhinagar, GJ 382355, India
2 Chemistry, IIT Gandhinagar, Palaj, GJ 382355, India

^* e-mail: atul.bhargav@iitgn.ac.in

Received: 3 July 2018
Received in final form: 3 December 2018
Accepted: 10 December 2018
Published online: 14 January 2019

Abstract

High-viscosity CO₂ is of interest to the oil and gas industry in enhanced oil recovery and well-fracturing applications. Dispersing nanoparticles in CO₂ is one way of achieving increased viscosity. However, parametric studies on viscosity estimation of CO₂ nanofluids is not found in the open literature. A comparison of various interatomic potentials for their accuracy in predicting viscosity is also missing. In this work, we studied Al₂O₃ nanoparticles in CO₂ base fluid. We screened the inter-molecular interaction potential models available for CO₂–CO₂ interactions and found that the TraPPE-flexible model (with MORSE potential) to be most suitable for conditions used in this work. We estimated the CO₂–Al₂O₃ interaction potential using quantum mechanical simulations. Using this combination for CO₂–CO₂ and CO₂–Al₂O₃ interactions, we explored the effects of temperature and nanoparticle size on viscosity using molecular dynamics simulations (MD). We predicted that the viscosity would increase with increase in temperature and particle size. We also calculated the base fluid self-diffusion coefficient to investigate the effect of Brownian motion and its contribution to changes in viscosity. We found that it decreases with increase in particle size and temperature, thereby indicating that Brownian motion does not contribute to the increased viscosity. Further, the nanolayer formed at the Al₂O₃–CO₂ interface is studied through density distributions around the nanoparticle; the thickness of this nanolayer is found to increase with nanoparticle diameter. Finally, we examined the structures of CO₂ fluid in presence of nanoparticles at different thermodynamic states through radial distribution functions. The current work sheds light on the viscosity enhancement by the addition of nanoparticles; it is hoped that such studies will lead to tools that help tailor fluid properties to specific requirements.