[safikhani_hamed]

Dear my friends,
I am going to modeling nanofluid heat transfer in a tube, is there any tutorial or document please ?

Sincerely

[ghorrocks]

None that I know of. The concepts are the same, it is just a matter of choosing appropriate physics. Obviously it will be low Re and therefore laminar, but otherwise choose options appropriate for the model.

But be aware that CFX uses the Navier Stokes equations, and this assumes the fluid can be treated as a continuum. If this assumption is not valid (possibly the case at the nano scale) then CFX may not be suitable. Depends on exactly what you are modelling.

[safikhani_hamed]

Dear friends,
Thanks for your guidance, but shoud I use a multiphase flow.

[ghorrocks]

No problem - I do MEMS simulations at the moment for inkjet printers so am quite familiar with the issues of micro scale modelling.

But if you say you are doing nano scale modelling then the multiphase assumption (I assume you mean free surface modelling, not eularian or particle tracking) may not be appropriate, as on a nano scale free surfaces cannot be assumed thin and surface age time scales become important, and the chemistry of what is happening on the surface governs the behaviour. For instance if you are talking about heat transfer in a carbon nanotube then I doubt CFX can help you and you will need to look at direct modelling of the molecules themselves.

Can you describe what you are doing a little more?

[safikhani_hamed]

Dear friends
In fact I want to investigate the effect of nano particle in heat transfer enhancement in a circular tube. the base fluid is water and the nanoparticles are cu.

[ghorrocks]

What is the circular tube? How big is it? What is driving the flow? What is the thermal conditions?

[safikhani_hamed]

Dear friend
The diameter of tube is 6 mm and the driven force is pressure and the length of tube is a meter.

[ghorrocks]

I assume the idea is to use copper nanoparticles to enhance heat transfer in the tube. Is this correct?

If the nanoparticles simply increase the thermal conductivity then simply put a representative thermal conducitivty in. You may need to modify other parameters such as viscosity, density.

Or do the nanoparticles cause some more complex heat transfer effect?

[safikhani_hamed]

Dear friend,
Thanks for your guidance, yes you are correct. Do you have any document or tutorial ? moreover should we modify the material of both gas and particles ? should we define two material in CFX and a discrete solid type material ?

[ghorrocks]

I think you will find copper nanoparticles will not require a particle model. They are so fine that the length scales associated with slip is insignificant, so a particle tracking model is not useful. A eularian approach could be useful, but I suspect (without knowning anything of the detail of your model ) that nanoparticles are too fine for the eularian assumption to be useful either.

For instance is Brownian motion important? Neither the particle tracking or eularian models account for Brownian motion by default.

That is why I suggest you consider just modelling this as a single phase simulation with the fluid properties modified to account for the nanoparticles. If that is all the nanoparticles do (ie modify the fluid properties) then this is by far the simplest and easiest approach.