Dye diffusion in a stationary fluid

kunalrb5 · August 21, 2014, 07:48

I am very new to CFD. I am working on project in which I need to simulate the diffusion of dye in a Nanofluid. So how should I start off this problem?
Note: the fluid is stationary in a box and a pulse input of a dye is given at the center of the box.

mprinkey · August 21, 2014, 10:17

If the "squirt" of dye has no significant momentum, you call solve this problem using any diffusion/heat equation solver. That includes all finite element and finite volume CFD codes. For a quick (and free) way to get started, download OpenFOAM and run through the tutorials and focus on ScalarTransportFoam. That solves the unsteady diffusion equation. To make the fluid stationary, set the mass flux field (phi) and velocity field (U) to zero.

kunalrb5 · August 22, 2014, 06:39

Thanks for your reply. Yes the squirt can be assumed insignificant but what about the nanofluid? Can we simulate the nanofluid (basically a base fluid containing suspended nanoparticles)?
I want to see the enhancement of mass transfer in nanofluid as compared to normal base fluid. So ultimately I will be needing the diffusion coefficient for the case of dye diffusion in nanofluid as well as base fluid so as to do the comparison.

mprinkey · August 22, 2014, 09:47

The diffusivity is an input to any CFD code. So, the base fluid and nanoparticle suspension will have different diffusivities that need to be measured. Perhaps there is theory that can produce modifications to the base diffusivity based on the nanoparticle properties. Honestly, the CFD simulation you are describing is very simple. If the simulation is 3d, you will see spherically symmetric transient result that can probably be solved with analytical techniques, or, at worst, a 100 lines of transient 1D code--I would not call that CFD.

If you are looking for meso- or microscale simulations of the dye interaction with the base fluid and nanoparticles, you may well need to look at molecular dynamics codes instead.

kunalrb5 · August 25, 2014, 03:23

Hi Michael,
Actually I don't at all want to look at any interaction between dye and fluid/ particles, rather I would assume it to be negligible. My sole purpose is to find out the enhancement in mass transfer due to the presence of Nanoparticles.

If I input the diffusivity value by using any theoretical correlation then it already assumes the enhancement phenomena. Therefore the CFD results will obviously show an enhancement in mass transfer.

What I want to study is the effect of Brownian motion of nanoparticles on the mass transport phenomena in any base fluid and according to literature, we expect an enhancement. I don't want to presume the enhancement but want to observe it as a result of the Brownian motion of the nanoparticles. So is there any way by which I could track the enhancement? as I thought, effective diffusivity is the only parameter which would tell us the relevant information.

I hope now you have a clear vision over my problem statement.

Thanks again for helping

mprinkey · August 25, 2014, 13:36

Most CFD codes start with equilibrium pressure, density, and velocity as the solved variables. There are no distribution functions, no statistical behavior at all (apart from bulk scale turbulence, etc.) If you want to model Brownian motion, you will need to look at either molecular dynamics (direct Lagrangian simulation of fluid, dye, and nanoparticles) or Boltzmann/Fokker–Planck formulations. There are techniques of CFD that can inform the solution of those systems, but you can't start with Ansys FLUENT or OpenFOAM standard solvers and get the results you are looking for.

For molecular dynamics, you might look at LAMMPS (http://lammps.sandia.gov/). I can't recommend any Eulerian non-equilibrium solvers.

August 21, 2014, 07:48	Dye diffusion in a stationary fluid	#1
kunalrb5 New Member Kunal Bhoyar Join Date: Aug 2014 Posts: 3 Rep Power: 12	I am very new to CFD. I am working on project in which I need to simulate the diffusion of dye in a Nanofluid. So how should I start off this problem? Note: the fluid is stationary in a box and a pulse input of a dye is given at the center of the box.

August 22, 2014, 06:39	Nanofluid and Diffusion coefficient	#3
kunalrb5 New Member Kunal Bhoyar Join Date: Aug 2014 Posts: 3 Rep Power: 12	Thanks for your reply. Yes the squirt can be assumed insignificant but what about the nanofluid? Can we simulate the nanofluid (basically a base fluid containing suspended nanoparticles)? I want to see the enhancement of mass transfer in nanofluid as compared to normal base fluid. So ultimately I will be needing the diffusion coefficient for the case of dye diffusion in nanofluid as well as base fluid so as to do the comparison.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Can CFX do CHT simulations with a solid domain rotating in a stationary fluid domain?	acro	CFX	15	September 23, 2016 12:16
Multiphase simulation of bubble rising	Niru	CFX	5	November 25, 2014 14:57
Overflow Error in Multiphase Modelling with Two Continuous Fluids	ashtonJ	CFX	6	August 11, 2014 15:32
elastic cylinder in stationary fluid	doctsh	ANSYS Meshing & Geometry	0	January 9, 2010 04:47
Fluid Dynamics Eng - PAX Mixer - San Rafael, CA	Gary Jong	FLUENT	0	February 25, 2008 21:46

August 21, 2014, 10:17		#2
mprinkey Senior Member Michael Prinkey Join Date: Mar 2009 Location: Pittsburgh PA Posts: 363 Rep Power: 25	If the "squirt" of dye has no significant momentum, you call solve this problem using any diffusion/heat equation solver. That includes all finite element and finite volume CFD codes. For a quick (and free) way to get started, download OpenFOAM and run through the tutorials and focus on ScalarTransportFoam. That solves the unsteady diffusion equation. To make the fluid stationary, set the mass flux field (phi) and velocity field (U) to zero.

August 22, 2014, 09:47		#4
mprinkey Senior Member Michael Prinkey Join Date: Mar 2009 Location: Pittsburgh PA Posts: 363 Rep Power: 25	The diffusivity is an input to any CFD code. So, the base fluid and nanoparticle suspension will have different diffusivities that need to be measured. Perhaps there is theory that can produce modifications to the base diffusivity based on the nanoparticle properties. Honestly, the CFD simulation you are describing is very simple. If the simulation is 3d, you will see spherically symmetric transient result that can probably be solved with analytical techniques, or, at worst, a 100 lines of transient 1D code--I would not call that CFD. If you are looking for meso- or microscale simulations of the dye interaction with the base fluid and nanoparticles, you may well need to look at molecular dynamics codes instead.

August 25, 2014, 03:23		#5
kunalrb5 New Member Kunal Bhoyar Join Date: Aug 2014 Posts: 3 Rep Power: 12	Hi Michael, Actually I don't at all want to look at any interaction between dye and fluid/ particles, rather I would assume it to be negligible. My sole purpose is to find out the enhancement in mass transfer due to the presence of Nanoparticles. If I input the diffusivity value by using any theoretical correlation then it already assumes the enhancement phenomena. Therefore the CFD results will obviously show an enhancement in mass transfer. What I want to study is the effect of Brownian motion of nanoparticles on the mass transport phenomena in any base fluid and according to literature, we expect an enhancement. I don't want to presume the enhancement but want to observe it as a result of the Brownian motion of the nanoparticles. So is there any way by which I could track the enhancement? as I thought, effective diffusivity is the only parameter which would tell us the relevant information. I hope now you have a clear vision over my problem statement. Thanks again for helping

August 25, 2014, 13:36		#6
mprinkey Senior Member Michael Prinkey Join Date: Mar 2009 Location: Pittsburgh PA Posts: 363 Rep Power: 25	Most CFD codes start with equilibrium pressure, density, and velocity as the solved variables. There are no distribution functions, no statistical behavior at all (apart from bulk scale turbulence, etc.) If you want to model Brownian motion, you will need to look at either molecular dynamics (direct Lagrangian simulation of fluid, dye, and nanoparticles) or Boltzmann/Fokker–Planck formulations. There are techniques of CFD that can inform the solution of those systems, but you can't start with Ansys FLUENT or OpenFOAM standard solvers and get the results you are looking for. For molecular dynamics, you might look at LAMMPS (http://lammps.sandia.gov/). I can't recommend any Eulerian non-equilibrium solvers.