Discussion Closed This discussion was created more than 6 months ago and has been closed. To start a new discussion with a link back to this one, click here.
Incorrect Implementation of Model for Packed-Bed Reactor [Model 238]
Posted 3 ott 2014, 02:27 GMT-4 Chemical Reaction Engineering 2 Replies
Please login with a confirmed email address before reporting spam
Hi guys,
I refer to this sample model.
www.comsol.com/model/packed-bed-reactor-238
I was trying to figure out how Ergun's equation was implemented into the model, and I did some reverse-engineering from this file. However, I found that the implementation was mathematically-incorrect, and this is even more so if the velocity drop across the packed-bed was significant.
I've attached my proof. Hope that someone who's also invested in Packed-Bed reactors can comment.
I refer to this sample model.
www.comsol.com/model/packed-bed-reactor-238
I was trying to figure out how Ergun's equation was implemented into the model, and I did some reverse-engineering from this file. However, I found that the implementation was mathematically-incorrect, and this is even more so if the velocity drop across the packed-bed was significant.
I've attached my proof. Hope that someone who's also invested in Packed-Bed reactors can comment.
Attachments:
2 Replies Last Post 4 feb 2015, 12:02 GMT-5
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Hi,
I found the same problem while I was trying to figure out how to implement the Ergun equation for pressure drop.
The underlying reason for this problem - their reason for using this beta*(F/F0)*(Pfeed/P) approach, instead of implementing the Ergun equation in a more "normal" way, is because the Ergun equation requires the evaluation of several physical properties - density and viscosity - and as far as I can tell, you cannot arbitrarily evaluate these properties, either for a material or for a mixture, without adding additional physics. (For example, you can add a Darcy Flow physics node, and density and viscosity are available through it - but then it is calculating a pressure drop, defeating the purpose of implementing the Ergun equation.)
The way the Ergun equation is implemented currently is a kludge, because it requires the initial values of viscosity and density to be defined and then assumes proportionalities for relating those physical properties to their value in the feed.
For situations where you're using materials, a better alternative would be to allow you to get physical properties of materials (like mat1.rho), which doesn't work.
If you're dealing with more complicated mixtures defined by Chemkin files via the Reaction Engineering module, a better approach would be to allow the user to compute viscosity and density from the reaction engineering module (since it would have all the information it needs to do so).
I can't figure out a workaround for this, and I have large density/temperature/pressure/velocity changes. So if anyone has insights on how they worked around this issue, please share!
I found the same problem while I was trying to figure out how to implement the Ergun equation for pressure drop.
The underlying reason for this problem - their reason for using this beta*(F/F0)*(Pfeed/P) approach, instead of implementing the Ergun equation in a more "normal" way, is because the Ergun equation requires the evaluation of several physical properties - density and viscosity - and as far as I can tell, you cannot arbitrarily evaluate these properties, either for a material or for a mixture, without adding additional physics. (For example, you can add a Darcy Flow physics node, and density and viscosity are available through it - but then it is calculating a pressure drop, defeating the purpose of implementing the Ergun equation.)
The way the Ergun equation is implemented currently is a kludge, because it requires the initial values of viscosity and density to be defined and then assumes proportionalities for relating those physical properties to their value in the feed.
For situations where you're using materials, a better alternative would be to allow you to get physical properties of materials (like mat1.rho), which doesn't work.
If you're dealing with more complicated mixtures defined by Chemkin files via the Reaction Engineering module, a better approach would be to allow the user to compute viscosity and density from the reaction engineering module (since it would have all the information it needs to do so).
I can't figure out a workaround for this, and I have large density/temperature/pressure/velocity changes. So if anyone has insights on how they worked around this issue, please share!
Please login with a confirmed email address before reporting spam
Posted:
10 years ago
Hi Wu,
we checked our derivation once more and it seems to be correct. Make sure you have the latest version (5.0) of the model and documentation.
Looking at your derivation, you add the feed rate, u_feed, into the differential equation, when it should be u. That essentially means that you only solve the equation at the inlet.
If you have further questions or comments, I suggest you send a message to support@comsol.com.
we checked our derivation once more and it seems to be correct. Make sure you have the latest version (5.0) of the model and documentation.
Looking at your derivation, you add the feed rate, u_feed, into the differential equation, when it should be u. That essentially means that you only solve the equation at the inlet.
If you have further questions or comments, I suggest you send a message to support@comsol.com.
Hi guys,
I refer to this sample model.
www.comsol.com/model/packed-bed-reactor-238
I was trying to figure out how Ergun's equation was implemented into the model, and I did some reverse-engineering from this file. However, I found that the implementation was mathematically-incorrect, and this is even more so if the velocity drop across the packed-bed was significant.
I've attached my proof. Hope that someone who's also invested in Packed-Bed reactors can comment.