|
[Sponsors] |
Solution diverges in packed bed granular simulation. |
|
LinkBack | Thread Tools | Search this Thread | Display Modes |
May 9, 2020, 23:51 |
Solution diverges in packed bed granular simulation.
|
#1 |
Member
Anshuman Sinha
Join Date: Oct 2018
Posts: 70
Rep Power: 8 |
I'm trying to model a packed bed gas absorption model. I'm using the Eulerian multiphase model with the species transport equation.
I have defined the materials as mixture materials, with material 1 = Air ( Oxygen + Nitrogen ) and material 2 = Carbon ( carbon_solid + product_solid ). Material 2 is the granular material and secondary phase. With the heterogeneous reaction being { Nitrogen + Carbon_solid = product_solid. } Problem: The solution runs well for normal fluidization bed, but as soon as I turn on the 'packed bed' option the solution diverges, especially the continuity equation diverges first. I'm using: Drag model = syamlal-obrien-para , void fraction = 0.4 and min fluidization velocity = 0.06 m/s Screen Shot 2020-05-10 at 8.17.00 AM.jpg Model: Bed zone = { volume fraction 0.6 Material 2 } and { 0.7 mass fraction species Carbon_solid. } Screen Shot 2020-05-10 at 8.17.10 AM.jpg |
|
May 10, 2020, 13:23 |
Packed Bed Model
|
#2 |
Senior Member
|
A requirement of using packed-bed model in Fluent is to fix the velocities of the cell zone representing packed-bed to 0. So, you need to go to Cell Zone Conditions, enable Fixed Values for the cell zone representing the packed-bed, and set all velocities to 0. This is in addition to selecting packed-bed under secondary phase properties.
__________________
Regards, Vinerm PM to be used if and only if you do not want something to be shared publicly. PM is considered to be of the least priority. |
|
May 10, 2020, 21:09 |
|
#3 | |
Member
Anshuman Sinha
Join Date: Oct 2018
Posts: 70
Rep Power: 8 |
Quote:
Also, I have tried modeling this particular system using the porous media approach. Where I defined 'wall surface reaction' in the porous zone. I have a question in that, 1. Whether the 'site species' of the wall surface reaction get consumed or not? If yes! Then the product of this surface reaction must diminish with time since the reactant is getting consumed up! But no such development is seen in my results! I'll post the model in detail if required. Thanks, Ansh |
||
May 11, 2020, 04:45 |
Packed Bed
|
#4 |
Senior Member
|
You only need to fix secondary phase velocity field to 0, not the primary phase velocity field. If the case is diverging, it could be because of many other reasons.
You need to specify the surface concentration of site species under Mechanisms.
__________________
Regards, Vinerm PM to be used if and only if you do not want something to be shared publicly. PM is considered to be of the least priority. |
|
May 11, 2020, 05:48 |
|
#5 | |
Member
Anshuman Sinha
Join Date: Oct 2018
Posts: 70
Rep Power: 8 |
Quote:
{ **The geometry of this problem remains the same as the original problem in the post, just instead of O2+N2 I'm using O2+H2 and the bed is now a porous zone instead of a granular eulerian zone and has sites of 'ga' (gallium) on it } mkmk.JPG in this simulation, 'ga' is the site species and 'gah3' is the solid species. As seen in the picture, it reacts with h2 to produce gah3. And the reaction is set in the porous zone only. { This is simply adsorption of h2 gas from the feed mixture, by preferential adsorption of h2 gas. The outlet mixture will have a lower h2 concentration } But after some time the 'ga' at the site should decrease in concentration as the time marches ahead. And since the 'ga' is getting lowered, the other reactant h2 now must increase at the outlet. { This phenomenon is not getting in place there } The image below shows the o2 (oxygen) component getting increased due to the preferential adsorption of h2 (hydrogen) but this o2 must decrease after a certain time when all the site species is consumed or gets covered with gah3 ( product of the surface reaction). But this doesn't happen! mkmkmk.JPG Please help me with this. Thanks |
||
May 11, 2020, 11:58 |
Consumption Rate
|
#6 |
Senior Member
|
Have you calculated how much time should it take for almost complete consumption of the site species?
__________________
Regards, Vinerm PM to be used if and only if you do not want something to be shared publicly. PM is considered to be of the least priority. |
|
May 11, 2020, 12:57 |
|
#7 | |
Member
Anshuman Sinha
Join Date: Oct 2018
Posts: 70
Rep Power: 8 |
Quote:
I can increase the reaction rate of the reaction and check whether the site species is getting consumed or not? |
||
May 11, 2020, 13:04 |
Consumption Rate
|
#8 |
Senior Member
|
Yes, artificially increased rate is a good idea.
__________________
Regards, Vinerm PM to be used if and only if you do not want something to be shared publicly. PM is considered to be of the least priority. |
|
May 11, 2020, 15:35 |
|
#9 |
Member
Anshuman Sinha
Join Date: Oct 2018
Posts: 70
Rep Power: 8 |
I tried to increase the reaction rate but no change was observed.
The solutions appear to reach a steady-state and even after increasing the time step size to 5 sec ( earlier I was using 0.001 sec) no divergence is observed and the solution remains the same. I also checked for the results of surface coverage of site species at different timestamps but it remains constant throughout to value 1.0 everywhere in the reactor zone. What may I do now? |
|
May 12, 2020, 08:22 |
Site species density
|
#10 |
Senior Member
|
I suppose the number you have is quite large for site density. What's the area of the surface where you have reaction?
__________________
Regards, Vinerm PM to be used if and only if you do not want something to be shared publicly. PM is considered to be of the least priority. |
|
May 12, 2020, 11:47 |
|
#11 | |
Member
Anshuman Sinha
Join Date: Oct 2018
Posts: 70
Rep Power: 8 |
Quote:
I have decreased the site density to 5e-10 kgmol/m2. which was earlier 0.005 , but still the solutions have come to standstill, I'm posting results after 400 seconds. Please also look at the residual plot. check.jpg residual plot resi.JPG |
||
Tags |
eulerian-granular, heterogeneous reaction |
Thread Tools | Search this Thread |
Display Modes | |
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
UDF for Automatic Solution Initialization for previous case data file | gartz89 | Fluent UDF and Scheme Programming | 6 | March 30, 2020 08:38 |
Steady-state solution as IC for transient simulation | pardoa | OpenFOAM Running, Solving & CFD | 3 | July 24, 2019 07:49 |
Solution diverges | giof | STAR-CD | 4 | November 5, 2015 13:04 |
porosity in packed bed with eulerian multiphase | Sree | FLUENT | 1 | April 15, 2015 07:17 |
Questions about gas-solid fluidized bed simulation | Sinferm | FLUENT | 1 | January 31, 2007 15:54 |