[Laexzzz]

Hey, i am modeling 3d reactor with chem. reaction, but I ran into some difficulties.

Concentration is changing very rapidly near the inlet, and then becomes constant.





Then I made a 2d version (very fine mesh).
270000 quadrilateral cells, zone 2, binary.
90 2D mass-flow-inlet faces, zone 3, binary.
90 2D outflow faces, zone 4, binary.
2220 2D wall faces, zone 5, binary.
538800 2D interior faces, zone 7, binary.
271201 nodes, binary.
271201 node flags, binary.

Result



(click to zoom in)



So, there is a small layer near inlet, where concentration is changing.
I tryed different boundary and different reaction - result is the same.
WhenI modeled the reaction in a tube, results were better

[Borsche]

From the tube reactor simulation it looks like your reaction equations are calculated alright. What kind of mixing do you have in your tank? Is it just the flow into the reactor that is causing the mixing?

Also, the reaction extent is greater in your tank than your tube. Did you specify a reversible reaction? If so you could have simply reached equilibrium.

Lastly, what is the length of your inlet to the tank? Is it similar to that of the tube?

[Laexzzz]

Originally Posted by Borsche

From the tube reactor simulation it looks like your reaction equations are calculated alright. What kind of mixing do you have in your tank? Is it just the flow into the reactor that is causing the mixing?

Also, the reaction extent is greater in your tank than your tube. Did you specify a reversible reaction? If so you could have simply reached equilibrium.

Lastly, what is the length of your inlet to the tank? Is it similar to that of the tube?

Thx for your response!
Mixing realized with MRF technique (50 rpm). Without mixing results also are not good.
Reaction is non-reversible.
Residence time in the inlet nozzle is much less than the residence time in the reactor. Residence time in the tube-reactor and in the tank-reactor are the same (average residence time = density*volume/mass flow rate).

[Borsche]

All I can think of is that the mixing is so great that your model predicts and ideal CSTR.

Is the model isothermal? It could be that the temperature in the tank is such that the reaction stops once it enters the tank.

It's strange that you have such high conversion in the inlet to your tank. If the CTSR and PFR are really of the same residence time than you would expect almost no conversion in the inlet to the tank. Do you have any backflow?

Check out the temperature and velocity profiles in your tank. That may give you a better understanding of your system. It doesn't seem to me that you've defined anything incorrectly at this point since your tube reactor shows good results.

[Laexzzz]

Originally Posted by Borsche

It's strange that you have such high conversion in the inlet to your tank. If the CTSR and PFR are really of the same residence time than you would expect almost no conversion in the inlet to the tank. Do you have any backflow?

No, threre is no backflow here.
You are right, conversion in the inlet is a main problem.
Also, as you can see at the picture below:



(click to zoom in)



I had same problem with 2d case without any mixing (high conversion near the inlet). I think, the main reason is a difference between geometry of the inlet and other space in reactor.

Velocity profile



Also, very interesting thing.
I check molar concentration of species.




Fluent show, that minimal value of molar concentration = 7.37. But its not true. Molar weight of specie = 104, density = 865. So, if specie mass fraction = 1, molar concentration should be = 1*865/104=8.31. Not 7.37.

[Laexzzz]

Any ideas?