CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   CFX (http://www.cfd-online.com/Forums/cfx/)
-   -   uniform mass fraction throughout domain (http://www.cfd-online.com/Forums/cfx/81661-uniform-mass-fraction-throughout-domain.html)

ranjith November 2, 2010 23:51

uniform mass fraction throughout domain
 
Hi,
I am simulating the mixing of two fluids with different viscosities. I am doing a transient analysis in which the domain is initially filled with fluid A and then gradually fluid B enters the domain and mixing takes place. The problem is that when I look at the results in post, the mass fraction is uniform through out the domain. There is no variation with respect to space. But there is variation with respect to time. I am expecting a variation with respect to both time and space. any ideas?

ranjith

AliTr November 3, 2010 02:29

assuming you have fluids with different densities, using isoVolume with density variable you should be able to track fluids' mixing.

ranjith November 4, 2010 01:16

Thanks ali for ur response.....but I am still not able to visualize the mixing. The properties are apparently calculated in bulk for the whole domain. I getting a uniform mass fraction and uniform density throughout the domain..

AliTr November 4, 2010 01:42

if your fluids get mixing (like brine and water) and model supposed to have a range of densities, I suggest you to follow the Tutorial about Reacting Flow in Mixing Tube, your case might be simpler but you have to follow the same method to create a multicomponent fluid which describes the mixture.

ranjith November 4, 2010 04:38

I did create a multicomponent fluid, in fact my fluid is a variable composition mixture of components A and B to form a mixture C. A has a mass fraction 1 initially in the domain and in the fluid(C) entering the domain B has a mass fraction 1. I then do a transient simulation with total time equal to the time taken by the fluid to traverse the length of the domain

ghorrocks November 4, 2010 06:25

If you set the initial mass fraction as 1 and inject stuff with a mass fraction of 1 into the domain then nothing will happen. As Ali suggests, have a look at the tutorial examples for how to set up multicomponent models. I think there is a CO2/air tutorial which will also be of relevance for you.

ghorrocks November 4, 2010 06:26

If the initial mass fraction is 1 then you have to inject stuff with a mass fraction of 0 (or anything other than 1) to get anything to happen.

ranjith November 5, 2010 23:55

Thanks Glen for ur response, but u have got me wrong. Initially the mass fraction is 1 for component A(0 for component B). The fluid entering the domain has mass fraction 1 for component B(0 for component A)....and I have taken total time equal to the time taken by the fluid to traverse through the domain.

ghorrocks November 6, 2010 07:17

If you have set the diffusivity too high it will rapidly diffuse over the entire domain. Are you sure the diffusivity is correct?

AliTr November 7, 2010 06:15

just for check, you should use the Multicomponent C material over your boundaries and then define the subcomponents fraction. you should not use the components A and B directly over the boundaries.

ranjith November 10, 2010 06:50

@glen -I have checked with different diffusivity values but I am getting the same results...
@ali- at inlet I have defined multicomponent fluid with mass fractions. my definition is correct

ghorrocks November 10, 2010 17:33

Your diffusivity could be out by orders of magnitude. Small changes will then have no effect. Try a diffusivity of 1e-20 and see if that makes a difference.

ranjith November 11, 2010 00:39

Thankyou Glenn!! you are absolutely right. I am getting the mixing pattern now... Thank you very much


All times are GMT -4. The time now is 06:04.