Learning Forum

[MS10]

I am working on a CFD simulation for CO2 capture using K2CO3 adsorbent in a 2D bubbling fluidized bed. Here are the details of my simulation:

- Using Eulerian CFD with the Gidaspow Drag model (Transient simulation).
- Modeling species transport with reaction.
- The reaction is: CO2 + H2O + K2CO3 --> 2KHCO3.
My questions are:
1. I modeled this as a heterogeneous reaction in multiphase modeling. Is this correct?
2. I didn't use a volumetric reaction. Is this okay?
3. I made a solid mixture of K2CO3 and KHCO3 to define KHCO3 as the product. Is this the right approach?
4. My simulation is diverging as soon as it starts. I've tried changing the model settings, but it didn't help.
5. I'm not sure how to correctly model this reaction.

Thank you 

Error: 

Divergence detected in AMG solver: vof-1

Divergence detected in AMG solver: pressure correction

Divergence detected in AMG solver: gas-phase-species-0

Divergence detected in AMG solver: gas-phase-species-1

Divergence detected in AMG solver: solid-phase-species-0

Divergence detected in AMG solver: vof-1

Divergence detected in AMG solver: pressure correction

Divergence detected in AMG solver: gas-phase-species-0

Divergence detected in AMG solver: gas-phase-species-1

Divergence detected in AMG solver: solid-phase-species-0

Divergence detected in AMG solver: vof-1

Divergence detected in AMG solver: pressure correction

Divergence detected in AMG solver: gas-phase-species-0

Divergence detected in AMG solver: gas-phase-species-1

Divergence detected in AMG solver: solid-phase-species-0

Divergence detected in AMG solver: vof-1

Error at host: floating point exception

 

Error: floating point exception

[Rob]

What you propose looks fairly sensible, but check your assumptions re reaction rates etc. You may also need to account for phase change in the material formation enthalpy. 

Assuming the mesh is good. Divergence typically means something is changing too quickly. So, check time step. If that fails turn models off until it works. Then turn things back on. That will pinpoint what's causing the problem to help figure out what is needed to fix it. 

[MS10]

Thank you for your response, Bob. The mesh appears to be fine, and I've verified the time step and under-relaxation factors, but nothing seems to work. When I turn off the model, it runs, but I still can't identify the issue. I've checked the model settings and the reaction modeling, but the problem persists. Another thing! I used the same settings to run the simulation on another computer, and it worked fine there. However, the simulation starts well sometimes, but the solid region becomes too fluidized and rises to the top of the reactor, which is unrealistic. What could be causing this issue with the simulation? How can it be corrected? What are the possible solutions to these issues? 

[Rob]

Looking at the chemistry, you're dropping two moles of gas and adding one mole of solid? What's controlling the reaction rate? That may explain the bed behaviour if material is switching phase too quickly.