CFD Online Discussion Forums

CFD Online Discussion Forums (
-   FLUENT (
-   -   Fluent-Mixing Problem (

Sundar February 5, 2002 14:59

Fluent-Mixing Problem
Hi guys, I am trying to model a mixing problem. I have fresh feed and recycle and they mix in a pipe with bends. The mixture is fed to reactor and some of the species in the mixture is consumed. I know the composition of species in outlet and at feed and recycle and I have rightly fed those values. I am able to observe the change in the individual species concn, but globally I couldn't observe a change in the mass. What I mean is I am monitoring inlet surface and outlet surface of reactor and species do vary as i expect to be, but the total mass which should be less due to comsumption remains the same. All your responses and insights are most welcome. Any insight in modeling such kind of mixing problems is also most welcome. Thanks Sundar

Erwin February 5, 2002 17:36

Re: Fluent-Mixing Problem
I'm sure everybody would like to know why the Law of Conservation of Mass doesn't apply to your system...

Maybe you need to check molar flows instead of mass flows.

Sundar February 5, 2002 18:09

Re: Fluent-Mixing Problem
I checked molar and mass flows. Everything looks fine. Also as such at the reactors I have a sink term and the boundary conds are velocity inlet and pressure outlet. The problem I am having is with the material properties. I would really appreciate any good thoughts in mixing analysis. Sundar

Greg Perkins February 5, 2002 19:17

Re: Fluent-Mixing Problem
I don't understand your say there should be less mass due to consumption doesn't occur. Where does this mass go (actually where do you think it should go, and why)?

If you've added source terms (maybe negative) to the equations you need to add a mass source to the fluid phase that equals the the sum of the species mass sources. If you don't do this, Fluent won't recognise a change in total mass conservation, due to your source terms.


Sundar February 5, 2002 21:20

Explaining Fluent-Mixing Problem
hi greg, thanks for ur response, let me explain the problem again. Fresh feed and recycle mix well and enter a reactor, In the reactor consumption takes place which i have modeled by putting a mass sink term in the (reactor)fluid zone in fluent for the respective species. I am solving the problem by specifying the composition at (velocity inlet condn)fresh feed inlet and recycle inlet and pressure outlet. I have put surfaces at reactor inlet and outlet to monitor the species in fluent. The species concn changes as i expect according to my specifed conds, but when i am referring at flux reports, the mass in flow at the reactor is equivalent to the mass outflow, which is not right. for eg. fresh feed = 135 g/s. recycle = 105 g/s. mixture at reactor inlet= 240g/s. at reactor outlet=215g/s. But the flux report in fluent is indicating fresh feed =127g/s, recycle= 97 g/s. and at outlet = 224 g/s. I am modeling for energy, turbulence (rke realizable) and multi species. well something is wrong in the way fluent calculating the material properties, I am modeling fluid density by ideal gas incommpresible flow. If any one has any insights in this problem i would really appreciate your responses.From the analysis stand point, I am intersted in finding whether the feed are recycle are mixed well before they enter the reactor. thanks sundar

Greg Perkins February 5, 2002 21:27

Re: Explaining Fluent-Mixing Problem
For this problem you will need a mass source for the fluid continuity equation which is equal to the sum of the species mass sources. In your case this will be a negative value.

This source is added in the fluid zone b/c panel under mass source. I assume you've added all the species sources in this panel too.

Have you ensured you have this. If not, this would explain your problem.


Sundar February 5, 2002 21:35

Re: Explaining Fluent-Mixing Problem
greg , thaks again for ur response. well i have a good question on this. I have 5 species. H2,CO2, H2O, N2 and Ar. Well the way fluent model muti species problem is we have to specify the mass fraction of 4 species and fluent calculate the 5 th one to balance it into 1. Also when specifying the mass term (-ve source in fluent), the help menu in fluent informed me that to specify the negative source terms for the 4 species in the respective species column in the mass source panel and to specify the negative source term for the 5th species in the main panel. Is the approach right. Have u solved multi species problem before. I am jsu following as per the indications in the panel. Anyway I had the question in my mind should i be mentioning the total mass. ur response is most welcome. thanks SUNDAR

Greg Perkins February 6, 2002 00:18

Re: Explaining Fluent-Mixing Problem
This is not correct. If your species source term for the ith species is Si, and you have N species, you should put the source terms Si, i=1..N-1 in the species panel as you describe. Now the total mass source is S = sum(Si,i=1..N) should be added in the mass source box in this panel (top box).

This does two things....first the total mass source is added to the continuity eqn to determine the total cell mass conservation. If this is non-zero you have cell mass addition/subtraction. The volume integral of this term, will be the global mass addition/consumption for your problem. Second, the Nth source term, Sn, is simply Sn = S - sum(Si,i=1..N-1).

So it looks like you put the source term Sn in the total mass source box rather than the sum, S. If Sn was zero as I suspect, this will explain the fact you have no global mass addition/consumption, but you do have changes in the species mass fractions.


Sundar February 6, 2002 15:35

Thanks Greg
hi Greg, Thanks a lot, that really worked. Well one thing Fluent should do is get rid of these ambigous statements. I am a regular Fluent user for nearly 2 years and finding it difficult in trusting the help menu. All the times it makes sense to use common sense rather than following the instruction given. However this forum is cool becos of the level of interaction. Again Greg, thanks for your time and suggestions. Well I am a full time CFD Engineer and would really like to keep in touch with ppl who are into CFD, so please keep in touch. My email id is Thanks Sundar

Greg Perkins February 6, 2002 18:29

Re: Thanks Greg
Your welcome,

...if you think the documentation for the program is lacking, try getting details out of the udf manual......anyway, fortunately, after about 2 years I've worked out pretty much what their code does, but it took ages to work it out and on occaisons reverse engineer some of the assumptions in the code.

Regards Greg

All times are GMT -4. The time now is 18:41.