Nanofluid modeling

ezlidor_shb · December 1, 2013, 12:10

Hi
Can I model nanofluid applying eulerian-lagrangian two-phase approach using ANSYS CFX 14.5? is ANSYS CFX 14.5 capable of nanometer sized particles simlulation?
Any help would be appreciated.

ghorrocks · December 1, 2013, 16:17

It depends on what physics you want the nanoparticles to show. Also it is unlikely that a multiphase model will be appropriate as multiphase assumes mixing at the micro scale level, not the nano scale. A multi-component flow might be more appropriate.

But this all depends on exactly what you want the nanoparticles to do. What physics do the nanoparticles do in this simulation?

ezlidor_shb · December 2, 2013, 00:29

i want to investigate the effect of the nanoparticles on the continuous fluid in curved pipes. in CFX-Pre processing (domain definition) can 10e-9 [m] be specified as the particle diameter for particle transport solid?
thnx

ghorrocks · December 2, 2013, 04:22

What effect on the continuous fluid?

Nanoparticles generally have no slip relative to the continuous phase as they are so small. Multiphase models are generally based around the idea of inter-phase slip - that is why multiphase models are usually unsuitable for nanoparticles.

Yes, you can put 1e-9 as a particle diameter. But as I said, this will just mean you get no interphase slip and so the multiphase model is unlikely to be suitable.

ezlidor_shb · December 2, 2013, 05:42

thank you so much,
which model do you suggest for nanofluid modeling?

ghorrocks · December 2, 2013, 05:59

Let me ask this more clearly - I cannot suggest what model would be suitable until you tell us what is the important physics you are trying to model. There are many possible effects of nanoparticles and some can be modelled in CFX and some cannot.

You have to tell us what physics you are trying to model.

ezlidor_shb · December 2, 2013, 13:41

I want to investigate the convective heat transfer coefficient and pressure drop enhancement compared with base fluid

ghorrocks · December 2, 2013, 16:13

How does the nanofluid affect the heat transfer coefficient and pressure drop? Does simply change the fluid thermal conductivity and/or viscosity or does it do something more complex? If so, what?

ezlidor_shb · December 3, 2013, 00:42

it changes fluid properties simply

ghorrocks · December 3, 2013, 00:54

And how does it change fluid properties.........

ezlidor_shb · December 4, 2013, 01:07

it increases thermal conductivity, convective heat transfer coefficient, density, viscosity and decreases specific heat,
and i don't exactly know how

ghorrocks · December 4, 2013, 06:28

If you don't know how it affects those properties then how can you model it?

You need to understand that physics of what you want to model before you touch CFX. So you need to do some research until you can write a mathematical function which defines how those properties vary as a function of whatever inputs they have.

salehorafi · December 23, 2013, 10:28

Hi
I'm trying to use the nanofluid to enhance the heat transfer rate, but not sure if the two phase mixture model is suitable for modelling, the properties will be modeled using UDF, so the thermal properties will be modeled using the known thermal conductivity models, my question is: is it appropriate to use the two phase mixture model for nanofluid?

ghorrocks · December 23, 2013, 21:20

This thread has many posts about the suitability of multiphase models in nanoparticle flows. But the general answer is probably not, but until you give specific details of what you are modelling I cannot be sure.

salehorafi · December 24, 2013, 06:33

Thanks for the reply, I'm modelling 2D flow natural convection in a square enclosure utilizing nanofluid in the presence of gravity, the hot and cold walls are 310 and 285 K respectively, the upper and lower walls are adiabatic, the nanofluid is Al2O3 -water based fluid , I'm trying to model the effective thermal conductivity and viscosity and density using mathematical models proposed by recent researchers , the properties will be modeled using UDF as a source code written in C, my question which model should I use? I've heard that some people used the two phase mixture model for modelling the nanofluid, this is an example:
http://www.sciencedirect.com/science...5193313001759#
and another clear example of using two phase mixture model:
http://www.sciencedirect.com/science...0072913001920#
I hope I could explain my case so waiting for explanation and how to use the UDF to model the properties? is the UDF used for both primary and secondary phase or only for the primary phase (water in this case), I've tried both but the contour for temperature looked that no temperature change along the body.
Regards

ghorrocks · December 25, 2013, 04:24

There are many approaches to model this.

The simplest is to model the nanoparticle concentration as an additional variable with convection and diffusion. Then you can make the fluid properties a function of the concentration. But this approach cannot model slip and more complex interations.

An alternate approach is to model this as a multicomponent fluid (not multiphase). This allows more complex fluid property interactions but will not allow slip between the nanorparticles and the fluid.

Yet another approach is multiphase. This only makes sense is if there is slip between the nanoparticles and fluid. And there are two approaches in multiphase:
Algebraic slip model - for materials with short slip time scales and where interface drag is the only important force this is the one to go with. A very cheap multiphase model.
Full eularian - this allows modelling of all multiphase interactions such as slip, gravity, wall effects and lots of other stuff. But this will be a complex model to set up and get running, especially if you are pushing the boundaries of what multiphase can do when you model nanorparticles.

The choice between these models is complex and I do not know enought detail about what you are doing to make a recommendation. But I do strongly recommend you try them all and get to know them so you can make an informed choice.

shenyan · April 15, 2014, 11:55

Quote:

Originally Posted by salehorafi

Hi
I'm trying to use the nanofluid to enhance the heat transfer rate, but not sure if the two phase mixture model is suitable for modelling, the properties will be modeled using UDF, so the thermal properties will be modeled using the known thermal conductivity models, my question is: is it appropriate to use the two phase mixture model for nanofluid?

hi
i want to model a nanofluid in two-phase eulerian model,so i have to define the nanoparticles as a fluid phase,but i don't know how do it,actually i don't know how define nanoparticle's viscosity.
can you help me please?

ghorrocks · April 15, 2014, 18:45

Have you read the posts on this thread? It says why nanoparticles should, in general, not be modelled using a multiphase model. My post #16 outlines the main possible approaches.

shenyan · April 16, 2014, 01:12

Quote:

Originally Posted by ghorrocks

Have you read the posts on this thread? It says why nanoparticles should, in general, not be modelled using a multiphase model. My post #16 outlines the main possible approaches.

yes,i did ,but didn't see my answer,i'm sure that must use eulerian model,but i don't know how define the viscosity of secondary phase(nanopartilces as a fluid phase)

ghorrocks · April 16, 2014, 06:06

Then I suspect you have defined the nanoparticles as a continuous phase and that does not sound correct.

Can you describe what you are modelling and why you are sure you need a two-phase eularian model for it? For most nanoparticle flow there is no slip of the nanoparticles relative to the continuous fluid, and this means a multiphase model is not appropriate.

December 1, 2013, 12:10	Nanofluid modeling	#1
ezlidor_shb New Member hesam Join Date: Dec 2013 Posts: 9 Rep Power: 12	Hi Can I model nanofluid applying eulerian-lagrangian two-phase approach using ANSYS CFX 14.5? is ANSYS CFX 14.5 capable of nanometer sized particles simlulation? Any help would be appreciated. aminamost likes this.

December 23, 2013, 10:28	thermal properties	#13
salehorafi New Member Join Date: May 2013 Posts: 4 Rep Power: 12	Hi I'm trying to use the nanofluid to enhance the heat transfer rate, but not sure if the two phase mixture model is suitable for modelling, the properties will be modeled using UDF, so the thermal properties will be modeled using the known thermal conductivity models, my question is: is it appropriate to use the two phase mixture model for nanofluid?

December 24, 2013, 06:33		#15
salehorafi New Member Join Date: May 2013 Posts: 4 Rep Power: 12	Thanks for the reply, I'm modelling 2D flow natural convection in a square enclosure utilizing nanofluid in the presence of gravity, the hot and cold walls are 310 and 285 K respectively, the upper and lower walls are adiabatic, the nanofluid is Al2O3 -water based fluid , I'm trying to model the effective thermal conductivity and viscosity and density using mathematical models proposed by recent researchers , the properties will be modeled using UDF as a source code written in C, my question which model should I use? I've heard that some people used the two phase mixture model for modelling the nanofluid, this is an example: http://www.sciencedirect.com/science...5193313001759# and another clear example of using two phase mixture model: http://www.sciencedirect.com/science...0072913001920# I hope I could explain my case so waiting for explanation and how to use the UDF to model the properties? is the UDF used for both primary and secondary phase or only for the primary phase (water in this case), I've tried both but the contour for temperature looked that no temperature change along the body. Regards aminamost likes this.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Nanofluid Modeling at Microchannel Heat-Sink	mech-engineer	FLUENT	6	June 12, 2014 03:32
Nanofluid Modeling at Microchannel Heat-Sink	mech-engineer	CFX	2	June 4, 2012 05:36
Nanofluid Modeling	safikhani_hamed	CFX	9	July 25, 2011 18:48
Nanofluid Modeling	safikhani_hamed	FLUENT	1	July 22, 2011 11:55
nanoFluid modeling in fluent	hamid	FLUENT	0	February 9, 2007 05:53

December 1, 2013, 16:17		#2
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,703 Rep Power: 143	It depends on what physics you want the nanoparticles to show. Also it is unlikely that a multiphase model will be appropriate as multiphase assumes mixing at the micro scale level, not the nano scale. A multi-component flow might be more appropriate. But this all depends on exactly what you want the nanoparticles to do. What physics do the nanoparticles do in this simulation?

December 2, 2013, 00:29		#3
ezlidor_shb New Member hesam Join Date: Dec 2013 Posts: 9 Rep Power: 12	i want to investigate the effect of the nanoparticles on the continuous fluid in curved pipes. in CFX-Pre processing (domain definition) can 10e-9 [m] be specified as the particle diameter for particle transport solid? thnx

December 2, 2013, 04:22		#4
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,703 Rep Power: 143	What effect on the continuous fluid? Nanoparticles generally have no slip relative to the continuous phase as they are so small. Multiphase models are generally based around the idea of inter-phase slip - that is why multiphase models are usually unsuitable for nanoparticles. Yes, you can put 1e-9 as a particle diameter. But as I said, this will just mean you get no interphase slip and so the multiphase model is unlikely to be suitable.

December 2, 2013, 05:42		#5
ezlidor_shb New Member hesam Join Date: Dec 2013 Posts: 9 Rep Power: 12	thank you so much, which model do you suggest for nanofluid modeling?

December 2, 2013, 05:59		#6
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,703 Rep Power: 143	Let me ask this more clearly - I cannot suggest what model would be suitable until you tell us what is the important physics you are trying to model. There are many possible effects of nanoparticles and some can be modelled in CFX and some cannot. You have to tell us what physics you are trying to model.

December 2, 2013, 13:41		#7
ezlidor_shb New Member hesam Join Date: Dec 2013 Posts: 9 Rep Power: 12	I want to investigate the convective heat transfer coefficient and pressure drop enhancement compared with base fluid

December 2, 2013, 16:13		#8
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,703 Rep Power: 143	How does the nanofluid affect the heat transfer coefficient and pressure drop? Does simply change the fluid thermal conductivity and/or viscosity or does it do something more complex? If so, what?

December 3, 2013, 00:42		#9
ezlidor_shb New Member hesam Join Date: Dec 2013 Posts: 9 Rep Power: 12	it changes fluid properties simply

December 3, 2013, 00:54		#10
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,703 Rep Power: 143	And how does it change fluid properties.........

December 4, 2013, 01:07		#11
ezlidor_shb New Member hesam Join Date: Dec 2013 Posts: 9 Rep Power: 12	it increases thermal conductivity, convective heat transfer coefficient, density, viscosity and decreases specific heat, and i don't exactly know how

December 4, 2013, 06:28		#12
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,703 Rep Power: 143	If you don't know how it affects those properties then how can you model it? You need to understand that physics of what you want to model before you touch CFX. So you need to do some research until you can write a mathematical function which defines how those properties vary as a function of whatever inputs they have.

December 23, 2013, 21:20		#14
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,703 Rep Power: 143	This thread has many posts about the suitability of multiphase models in nanoparticle flows. But the general answer is probably not, but until you give specific details of what you are modelling I cannot be sure.

December 25, 2013, 04:24		#16
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,703 Rep Power: 143	There are many approaches to model this. The simplest is to model the nanoparticle concentration as an additional variable with convection and diffusion. Then you can make the fluid properties a function of the concentration. But this approach cannot model slip and more complex interations. An alternate approach is to model this as a multicomponent fluid (not multiphase). This allows more complex fluid property interactions but will not allow slip between the nanorparticles and the fluid. Yet another approach is multiphase. This only makes sense is if there is slip between the nanoparticles and fluid. And there are two approaches in multiphase: Algebraic slip model - for materials with short slip time scales and where interface drag is the only important force this is the one to go with. A very cheap multiphase model. Full eularian - this allows modelling of all multiphase interactions such as slip, gravity, wall effects and lots of other stuff. But this will be a complex model to set up and get running, especially if you are pushing the boundaries of what multiphase can do when you model nanorparticles. The choice between these models is complex and I do not know enought detail about what you are doing to make a recommendation. But I do strongly recommend you try them all and get to know them so you can make an informed choice.

April 15, 2014, 18:45		#18
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,703 Rep Power: 143	Have you read the posts on this thread? It says why nanoparticles should, in general, not be modelled using a multiphase model. My post #16 outlines the main possible approaches.

April 16, 2014, 06:06		#20
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,703 Rep Power: 143	Then I suspect you have defined the nanoparticles as a continuous phase and that does not sound correct. Can you describe what you are modelling and why you are sure you need a two-phase eularian model for it? For most nanoparticle flow there is no slip of the nanoparticles relative to the continuous fluid, and this means a multiphase model is not appropriate.