CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   FLUENT (http://www.cfd-online.com/Forums/fluent/)
-   -   Non-Newtonian flow!!!!! (http://www.cfd-online.com/Forums/fluent/96566-non-newtonian-flow.html)

adambarfi January 26, 2012 02:13

Non-Newtonian flow!!!!!
 
Hi,

Any one know how can I Analyze a visco-elastic fluid in Fluent, I can just solve for a power law fluid.

Amir January 26, 2012 02:51

Quote:

Originally Posted by adambarfi (Post 341220)
Hi,

Any one know how can I Analyze a visco-elastic fluid in Fluent, I can just solve for a power law fluid.

Hi,

The FLUENT is not a proper tool for this purpose. You have 2 options; you can use "poly flow" software available in ansys pack which I don't recommend but the better option is using the "OpenFOAM" software that the viscoelastic solver has been added since 2009.

Bests,

adambarfi January 26, 2012 02:56

Quote:

Originally Posted by Amir (Post 341226)
Hi,

The FLUENT is not a proper tool for this purpose. You have 2 options; you can use "poly flow" software available in ansys pack which I don't recommend but the better option is using the "OpenFOAM" software that the viscoelastic solver has been added since 2009.

Bests,


Thank you Amir

arjun January 27, 2012 22:47

Quote:

Originally Posted by Amir (Post 341226)
Hi,

The FLUENT is not a proper tool for this purpose. You have 2 options; you can use "poly flow" software available in ansys pack which I don't recommend but the better option is using the "OpenFOAM" software that the viscoelastic solver has been added since 2009.

Bests,


very strange thing to say. Why you do not recommend Polyflow???

As far as I understand (based on my experience with polyflow), polyflow is most robust software when it comes to non Newtonian viscoelastic flows. It can run circles around openfoam as far as non newtonian flows are concerned.


The only issue that one could face with polyflow is that it takes lots of memory because some (not all) of the solvers in it are direct solver types for example it has multifrontal solver. So if your case is small enough that polyflow can run on it then there is NO match to it.

shk12345 January 28, 2012 00:11

Quote:

Originally Posted by adambarfi (Post 341220)
Hi,

Any one know how can I Analyze a visco-elastic fluid in Fluent, I can just solve for a power law fluid.

there must be several papers by prof Raj Chabra
I think you must see them for this purpose.


Regards
Shk

Amir January 28, 2012 03:06

Quote:

Originally Posted by arjun (Post 341600)
very strange thing to say. Why you do not recommend Polyflow???

As far as I understand (based on my experience with polyflow), polyflow is most robust software when it comes to non Newtonian viscoelastic flows. It can run circles around openfoam as far as non newtonian flows are concerned.


The only issue that one could face with polyflow is that it takes lots of memory because some (not all) of the solvers in it are direct solver types for example it has multifrontal solver. So if your case is small enough that polyflow can run on it then there is NO match to it.

Dear Arjun,

First of all, I think that polyflow uses finite element method in its procedure which is not the best for fluid calculations (correct me if it's not right). secondly, I had many studies regarding viscoelastic algorithms and numeric features and stabilizers specially in transient viscoelastic flow fields which need necessary changes in its solver code or libraries and so on where the openFOAM as an open source code helped me a lot. Thirdly, if I remember correctly, there are few restrictions in polyflow solver such as combination of viscoelastic and dynamic mesh technique which was exactly what I wanted but in openFOAM, there isn't any restriction. lastly, it may be seen so strange but I found openFOAM a user friendly software in comparison with the polyflow which depend on different users.

Bests,

arjun January 28, 2012 08:33

Quote:

Originally Posted by Amir (Post 341615)
Dear Arjun,

First of all, I think that polyflow uses finite element method in its procedure which is not the best for fluid calculations (correct me if it's not right). secondly, I had many studies regarding viscoelastic algorithms and numeric features and stabilizers specially in transient viscoelastic flow fields which need necessary changes in its solver code or libraries and so on where the openFOAM as an open source code helped me a lot. Thirdly, if I remember correctly, there are few restrictions in polyflow solver such as combination of viscoelastic and dynamic mesh technique which was exactly what I wanted but in openFOAM, there isn't any restriction. lastly, it may be seen so strange but I found openFOAM a user friendly software in comparison with the polyflow which depend on different users.

Bests,


Here are few things:


" I think that polyflow uses finite element method in its procedure which is not the best for fluid calculations"

Yes it is finite element, but one can not say that it is not suited for fluid flow simulations. Fluid flow is easier to write in finite volume, but it does not make other approaches unsuitable. It is pretty accurate and very robust solver.


"I had many studies regarding viscoelastic algorithms and numeric features and stabilizers specially in transient viscoelastic flow fields which need necessary changes in its solver code or libraries and so on where the openFOAM as an open source code helped me a lot."

Not everyone likes to add or modify the code to his and her needs. Some of the people just like to use the code. You can not 'NOT RECOMMEND' based on it. If other person said he wants to modify the code then it might make sense.

I also find it very easy to modify MY solvers codes but it does not mean that other codes are not useful for others.


"Thirdly, if I remember correctly, there are few restrictions in polyflow solver such as combination of viscoelastic and dynamic mesh technique which was exactly what I wanted but in openFOAM, there isn't any restriction."

I do not know where you looked but I think there are no such restrictions. Dynamic meshes is there in it for long time.

I can think of many issues with openFOAM with regard to viscoelastic or non newtonian flows, take for example how do you apply wall Slip based on navier law in openFOAM, slip is one of the very important aspect of non newtonian flows. I did not see mesh superposition type methods in openFOAM?

Amir January 28, 2012 10:30

Hi,
Quote:

Originally Posted by arjun (Post 341636)
Yes it is finite element, but one can not say that it is not suited for fluid flow simulations. Fluid flow is easier to write in finite volume, but it does not make other approaches unsuitable. It is pretty accurate and very robust solver.

I don't have enough knowledge about finite element procedure but I guess that the more accurate solvers need more accurate shape functions which is both time and memory consuming right? If it's correct do you think it's a robust code?

Quote:

Originally Posted by arjun (Post 341636)
Not everyone likes to add or modify the code to his and her needs. Some of the people just like to use the code. You can not 'NOT RECOMMEND' based on it. If other person said he wants to modify the code then it might make sense.
I also find it very easy to modify MY solvers codes but it does not mean that other codes are not useful for others.

Ok, calm down :D . The recommendations are based on the comparisons between softwares. As you said "Some of the people just like to use the code" Ok; they can also use openFOAM to just run simple cases but as you know the viscoelastic field is an enigmatic field specially in the transient one and usually need modifications. Consequently, I recommend a FREE software with capability of modifications in all of its aspects instead of polyflow. ;)

Quote:

Originally Posted by arjun (Post 341636)
I do not know where you looked but I think there are no such restrictions. Dynamic meshes is there in it for long time.

I apologize for this. According to my goal, polyflow has a limitation in modelling of viscoelastic flow fields in domains with a moving parts according to the polyflow help 12.1; I don't know maybe it's solved in later versions (21.1.2 manual)
Quote:

Originally Posted by arjun (Post 341636)
I can think of many issues with openFOAM with regard to viscoelastic or non newtonian flows, take for example how do you apply wall Slip based on navier law in openFOAM, slip is one of the very important aspect of non newtonian flows. I did not see mesh superposition type methods in openFOAM?

As I know, you can change the BCs according to your goal by introducing new library or even use simpler ways such as groovyBC ...
Look, there isn't even viscoelastic with dynamic mesh solver but you can simply merge many solvers; about the mesh superposition the procedure is the same but as I said it has a restriction in polyflow as least in ver 12.1.
Now this is the question: is there the capability of merging solvers in polyflow? can you introduce new forms of BCs? and so on .....

Bests,

adambarfi April 14, 2012 05:59

Hi guys

thanks, I decide to work with OpenFoam!

but I learned that how we can analyze a power law fluid with Fluent,

again thank you.


All times are GMT -4. The time now is 02:22.