CFD Online Discussion Forums

CFD Online Discussion Forums (
-   CFX (
-   -   Strange multicomponent source behaviour (

Zitron July 12, 2007 02:38

Strange multicomponent source behaviour

I am doing a multicomponent simulation of a room, in which air is slowly being converted into CO2 at a boundary wall. To simulate this, I made a wall with a mass fraction wall source that introduces CO2 at 1.1e-5 kg/s.

I have built a transient simulation at room temperature with buoyancy, which worked just fine converging to 0.0005, except:

1. For some reason CO2 comes in at a very low temperature, a couple of the cells next to the wall can be 60K below the specified wall temperature of 310K.

2. It seem like more CO2 is being added than I specified. Using the

volumeInt(CO2 Ideal Gas.Density * CO2 Ideal Gas.Mass Fraction)@Air

function, after 1 second, the total CO2 mass in the domain is ~1.5e-5 kg (should be 1.1e-5), after 600 seconds, it becomes 0.0099 kg instead of the predicted 0.0066 kg.

I've just got into multicomponent modeling, does any one know what might be going on?



opaque July 12, 2007 09:43

Re: Strange multicomponent source behaviour
Dear Zitron,

How did you setup the boundary source? Is it a Continuity equation Mass Source? or just a Mass Fraction equation source.. They are different animals and you must understand what they do. At what temperature is the CO2 coming in?

Is it laminar or turbulent flow?

Hope it helps


Zitron July 12, 2007 12:19

Re: Strange multicomponent source behaviour
Hi Opaque,

Thanks for your reply. The source on the wall is a mass fraction source, so I think there is no actual fluid introduced into the domain (no continuity source). There are also no inlets or outlets anywhere to add or remove fluid. The setup is similar to the 2D buoyant cavity tutorial, except with multicomponent fluid.

I don't know if there is a way to set the CO2 temperature, since on the mass fraction source options there is only a input for total source or flux of source. I can add an energy source, which will raise the temperature, but I can't understand why I would need to do that.

Since the flow speeds in the domain are below 0.6m/s, I chose not to use any turbulence model. I think that is a good assumption, correct? I also used 5-8 layers of inflated mesh on all walls.

By the way, I think I solved my second problem:

"volumeInt(CO2 Ideal Gas.Density * CO2 Ideal Gas.Mass Fraction)@Air" is the wrong way to do it, I should use:

"volumeInt(Density * CO2 Ideal Gas.Mass Fraction)@Air" instead. Now It gives me 0.00659 kg of CO2 after 600 sec, which is pretty close to the 0.0066 kg theoretical.

Thanks for your help,


opaque July 12, 2007 14:01

Re: Strange multicomponent source behaviour
Dear Zitron,

If that is the case, then you need to account for the energy balance as well. CO2 does not show at wall for free. The control volume at the wall has a finite amount of energy before the conversion. That is, summation of Mass Fraction_i * energy _i = Energy_Value. When you change the composition distribution (mass fractions), the solver will determine a which temperature the new composition will return the same Energy_Value.. CO2 has a non-trivial enthalpy of formation; therefore, the temperature must be very low such that the energy_value constrain is satisfied.

Hopefully I explained correctly (and clear).. You need an energy source/sink.

Hope this helps,


Zitron July 12, 2007 15:32

Re: Strange multicomponent source behaviour
Ah! OK now it makes sense! So I need to add a energy source equal to the enthalpy of formation of CO2 x amount of CO2 produced, or perhaps I can simply set the enthalpy of formation of CO2 to zero, since I'm not interested in the chemistry.

Thanks a lot, that was really helpful!


All times are GMT -4. The time now is 08:29.