Simulation of decarburization process, Need your help
 

Hey, experts
I want to simulate the decarburization process in molten steel. In my model, the very low concentration of carbon and oxygen are dissolved in the molten steel. The reactor is full of molten steel, the specified flow rate of argon (bubble) will be injected into the molten steel. Above the bath surface of the molten steel, the pressure is very low. So in this condition, the dissolved carbon and oxygen will transfer to the argon bubble surface, and generate the monoxide gas which will be flow out of molten steel. I want to simulate this process.
My question is how to define the dissolved carbon/oxygen and carbon monoxide? multicomponent or additional variables?
Maybe the multicomponent is best choice. I can define the molten steel and the dissolved carbon/oxygen as the variable component mixture, and treat molten steel as constraint part. But, how to define the generated carbon monoxide?
If I define it as variable component mixture with molten steel, like carbon/oxygen, because the molten steel is liquid state, the carbon monoxide will be also treated as liquid state. But in this way, the generated carbon monoxide can not flow out from the degassing bountary condition.
If I define it as gas state, there will be have more one phase in my system. So I need consider of the interphase force between molten steel and carbon monoxide? The flow field is not perfect.
So what can I do? I really need some good guys to give me some advices. Thank you very much.

This sounds like you need to model some chemistry as well, and you need multicomponent mixtures for that.

If a degassing boundary does not work then use a source term to get rid of the CO.

Hey, Glenn
Thanks for your reply. Can I ask you some other questions?
In actual process, the dissolved carbon and oxygen will transfer to the argon bubble surface and react to form carbon monoxide. The carbon monoxide gas will flow out from the molten steel. So, if I treat steel, the dissolved carbon/oxygen and carbon monoxide as a multicomponent mixture, I have known that the degassing boundary does not work, because the state of carbon monoxide is liquid. If I will use this way, I also have some questions.
1) How to control the carbon-oxygen reaction only be happened in argon bubble surface? This means only the part which the specified multicomponent contact to argon bubble can have a reaction.
2) Actually, This is my first time to some simulation about mass transfer and chemical reaction. So I am unfamiliar with the CFX function of these two parts. My question is what is different between the transport equation in component models and fluid equation source in Subdomain source? If I want to use the mass transfer process to control the chemical reaction, how can I do this?

Regards
Bohong