Dear Ahmed Faris

As it investigated in many recent papers, for nanofluids generally, using two-phase simulation yields more precise results (Because effect of interface is not neglected). Also, type of multiphase model ( formulation) plays an important role in results. For example, Eulerian-Eulerian model, Dispersion Model, Mixture model. Comparing all mentioned models vs. single phase model (heat transfer of nanofluids in microchannel), demonstrated results with higher accuracy, which were more compatible with experimental data.

Here, I recommend you to study following articles which would be effective for better understanding of nanofluids heat transfer mechanism' :

- Fard, M.H., Esfahany, M.N., Talaie, M.R., 2010. Numerical study of convective heat transfer of nanofluids in a circular tube two-phase model versus single-phase

model. Int. Commun. Heat Mass Transfer 37, 91–97.

- Lotfi, R., Saboohi, Y., Rashidi, A.M., 2010. Numerical study of forced convective heat

transfer of nanofluids: comparison of different approaches. Int. Commun. Heat

Mass Transfer 37, 74–78.

My guess is that two fluid approximation shall be better, particularly as you shall get a better resolution of the effects at the interface

Is it liquid-liquid interface OR liquid-solid interface?

Dear Ahmed Faris

As it investigated in many recent papers, for nanofluids generally, using two-phase simulation yields more precise results (Because effect of interface is not neglected). Also, type of multiphase model ( formulation) plays an important role in results. For example, Eulerian-Eulerian model, Dispersion Model, Mixture model. Comparing all mentioned models vs. single phase model (heat transfer of nanofluids in microchannel), demonstrated results with higher accuracy, which were more compatible with experimental data.

Here, I recommend you to study following articles which would be effective for better understanding of nanofluids heat transfer mechanism' :

- Fard, M.H., Esfahany, M.N., Talaie, M.R., 2010. Numerical study of convective heat transfer of nanofluids in a circular tube two-phase model versus single-phase

model. Int. Commun. Heat Mass Transfer 37, 91–97.

- Lotfi, R., Saboohi, Y., Rashidi, A.M., 2010. Numerical study of forced convective heat

transfer of nanofluids: comparison of different approaches. Int. Commun. Heat

Mass Transfer 37, 74–78.

I guess the size of the nano particles will be very important in the multiphase model; since it correlates directly with the solid surface area and the interface interaction with the fluid.

I should agree with Ahmad Ismail,

Due to my masters' thesis, I have buckled a lot with two-phase liquid-liquid interaction considering mass transfer mechanisms. Although it seems precise at first glance, but you should use many simplifying issues in order to extract results and during this stage you might face Some baffler conditions in the multi-dimensional physics of problem and do not know how to take care of(such as homogeneousity and the scheme of slip velocities between phases .)

So I would suggest you to make sure of the necessity of taking two-phase modeling into account or not. A good reference could be CFX user manual and CFX theory guide in each of which, the procedure and limitations of single/multi phase is explained sufficiently.

I should also mention that CFX theory guide would be a reasonable reference if you are building up your own coding or using any other softwares.

In a flow, especially at high velocity, what will be the effect of the brownian motion? Decreasing?

I´d say that multiphase flow simulations in standard CFD code are meaningless for simulations of nanofluids: in a stabilized suspension, the behavior of the colloidal particles rely on detailed features of interactions between particles (in principle, both hydrodynamic as well as non-hydrodynamic, say DLVO type) . These interactions are essential to describe the microstructure (i.e, state of aggregation of nanoparticles) of the nanofluid, which in turn is responsible for the observed 'anomalous' effective properties (viscosity, thermal conductivity) of the nanofluid.

For example, a Peclet number based on local flow conditions (say, from energy dissipation rate) will determine the importance of shear driven versus brownian driven coagulation rate, the non-hydrodynamic potentials will determine how and/or if particles aggregate. High shear could also break aggregates depending on bonding strength, etc. None of these details can be described with standard multiphase CFD simulations.

Therefore, I think that only single phase simulations which incorporate the effect of nanoparticles by means of effective properties make sense. In turn, effective properties need to be determined by experiments and/or theoretical/specific numerical models, and possibly could have a dependence on local flow conditions.

How about the effect of the collisions among the nanoparticles and also with the solid walls? Are they significant?

Collisions between nanoparticles are important as they lead to aggregation; with walls, as long as clusters remain of moderate size (so that we can properly keep calling the suspension a nanofluid), I don´t think they are relevant (other than some deposit can be formed and the implications that could have).

Hi Mauricio,

Try using a single phase which will be the solvent (a continuous phase). Represent the nano-particles as dispersed solid particles with zero slip velocity. Use full coupling between dispersed nano-particles and continuous phase. Full coupling will allow the nano-particles to influence the liquid flow.

We also find that the nanofluids have self-lubrication property. How to include this in the simulation?

There is also viscosity change in nanofluids.

Should we treat nanofluids as Newtonian fluids or non-Newtonian fluids?

Some researchers observe that nanofluids are not Newtonian fluids. So, which non-Newtonian model should we use?

We are looking at power law fluid model and self-lubrication.

We are also looking at the best mixing-law for nanofluids.

For nanofluids (with solid particles), will the solid-liquid interactions influence the simulation results? How to handle this?

When we run the CFD simulation using "transient" conditions, will the single phase produce faster solutions?

At what percentage/concentration of the nano particles will the effect of collision become significant?

I will recommend the paper by Ekene Onyiriuka et al who analyzed four different models for modeling nanofluidsArticle Evaluation of single-phase, discrete, mixture and combined m...