CFD Online Logo CFD Online URL
www.cfd-online.com
[Sponsors]
Home > Forums > FLUENT

*Q* H2 passes through membrane PEMFC in FLUENT

Register Blogs Members List Search Today's Posts Mark Forums Read

Reply
 
LinkBack Thread Tools Display Modes
Old   August 17, 2017, 18:10
Question *Q* H2 passes through membrane PEMFC in FLUENT
  #1
New Member
 
Ryan Lee
Join Date: Jul 2017
Posts: 10
Rep Power: 2
dui0828 is on a distinguished road
Hello,

Thank you in advance for viewing my question.

I am working on "PEM Fuel Cell Module" in Fluent.

There are two flow channels, at the anode and cathode, and membrane blocks the flow of the gas into each side.(Hydrogen from Anode side and Air or Oxygen from the Cathode side)
Hydrogen molecule is not supposed to pass through membrane, but only hydrogen ions.

When I run the simulation of it and check the mass fraction of Hydrogen, I could check it passed through the membrane and exist at the another side. (Fig.1)
Fig_1_hydrogen penetration.jpg
I am not sure why it happened and how to deal with this.
I am suspecting 2 things below.
  • Is hydrogen ion is also included the mass fraction of H2?
  • When I check the Iso-surface at a certain point, the edge of flow channel end is not connected with the wall end. (In other words, the corner is not connected completely.) Fig.2

    Is this just problem of the expression?
    In other words, it is actually connected but the image is expressed like not connected?
    not connected edge.jpg

Is there anyone who is familiar with this issue?
It will be very pleased to get any response from you guys.

Thank you again.
dui0828 is offline   Reply With Quote

Old   August 21, 2017, 00:03
Default
  #2
Senior Member
 
A CFD free user's Avatar
 
A-A Azarafza
Join Date: Jan 2013
Posts: 219
Rep Power: 7
A CFD free user is on a distinguished road
Hi Friend,
For the first question,you can set the mass fraction for hydrogen zero for the membrane layer. However, doing this, makes it difficult to get the converged solution. You can do that by going to cell zone "membrane" and from that go to hydrogen mass fraction and set it to "0". Then you can check and see that hydrogen is not permeated through the membrane any more. Probably something should be done with the PEM customized UDF, however, as I have not worked on fuel cell for several months, I can not recall the details now.

Try it and let me know the outcome.

Cheers,
__________________
Regard yours
A CFD free user is offline   Reply With Quote

Old   August 21, 2017, 01:50
Default
  #3
New Member
 
Ryan Lee
Join Date: Jul 2017
Posts: 10
Rep Power: 2
dui0828 is on a distinguished road
A-A Azarafza,
First of all, I really thank you for your response.

To go further with your advice, I want to make sure about cell zone condition.

At the cell zone condition, all parts except current collectors(anode and cathode), were set as fluid.
That is because when I set them as solid, there were errors about segmentation or improper list.

So when I click the specific part, for example, membrane,
material name is 'fc-micture'.

Only when I change the type as solid, I could see the material name of electrolyte-default.

Is this normal?
Though I set them as a fluid and materials name is just fc-mixture, it will automatically adapt the property of the electrolyte(default), since I set the part as an electrolyte at the Model?
dui0828 is offline   Reply With Quote

Old   August 21, 2017, 01:52
Question
  #4
New Member
 
Ryan Lee
Join Date: Jul 2017
Posts: 10
Rep Power: 2
dui0828 is on a distinguished road
Quote:
Originally Posted by A CFD free user View Post
Hi Friend,
For the first question,you can set the mass fraction for hydrogen zero for the membrane layer. However, doing this, makes it difficult to get the converged solution. You can do that by going to cell zone "membrane" and from that go to hydrogen mass fraction and set it to "0". Then you can check and see that hydrogen is not permeated through the membrane any more. Probably something should be done with the PEM customized UDF, however, as I have not worked on fuel cell for several months, I can not recall the details now.

Try it and let me know the outcome.

Cheers,
A-A Azarafza,
First of all, I really thank you for your response.

To go further with your advice, I want to make sure about cell zone condition.

At the cell zone condition, all parts except current collectors(anode and cathode), were set as fluid.
That is because when I set them as solid, there were errors about segmentation or improper list.

So when I click the specific part, for example, electrolyte,
material name is 'fc-micture'.

Only when I change the type as solid, I could see the material name of electrolyte-default.

Is this normal?
Though I set them as a fluid and materials name is just fc-mixture, it will automatically adapt the property of the electrolyte(default), since I set the part as an electrolyte at the Model?
cell_zone.jpg
dui0828 is offline   Reply With Quote

Old   August 21, 2017, 03:51
Default
  #5
Fy_
New Member
 
YZH
Join Date: Apr 2017
Posts: 8
Rep Power: 2
Fy_ is on a distinguished road
Firstly, if you use the Fuel cell module, the Fuel Cell and Electrolysis Model allows for water and the ionic current to pass through the electrolyte/membrane.

Secondly, I noticed what you have said above. As you say, the membrane shoule be set as fluid in the cell zone condition and the material will be fc-mixture.

Do you make the ''Update Cell Zones" keep selected when you open the Eletrolyte tab of the Fuel Cell and Electrolysis Models dialog box to specify zones and properties of the electrolyte/membrane portion of the fuel cell?
Attached Images
File Type: png QQ??20170821155207.png (36.4 KB, 11 views)
Fy_ is offline   Reply With Quote

Old   August 21, 2017, 04:08
Default
  #6
Senior Member
 
A CFD free user's Avatar
 
A-A Azarafza
Join Date: Jan 2013
Posts: 219
Rep Power: 7
A CFD free user is on a distinguished road
Fy is quite right. Did you manage to figure out the problem?

Sent from my SM-G950F using CFD Online Forum mobile app
__________________
Regard yours
A CFD free user is offline   Reply With Quote

Old   August 21, 2017, 04:46
Default
  #7
New Member
 
Ryan Lee
Join Date: Jul 2017
Posts: 10
Rep Power: 2
dui0828 is on a distinguished road
Quote:
Originally Posted by Fy_ View Post
Firstly, if you use the Fuel cell module, the Fuel Cell and Electrolysis Model allows for water and the ionic current to pass through the electrolyte/membrane.

Secondly, I noticed what you have said above. As you say, the membrane shoule be set as fluid in the cell zone condition and the material will be fc-mixture.

Do you make the ''Update Cell Zones" keep selected when you open the Eletrolyte tab of the Fuel Cell and Electrolysis Models dialog box to specify zones and properties of the electrolyte/membrane portion of the fuel cell?
Thank you very much for the response.
And also about the figure you took.

I set as you mentioned.
(checked the box at the model, which is 'keep updated')
I could make it clear because of your confirmation.

Thank you again.

Would you check the residual is converged?
(though it is just 20 iterations)
residual_007.jpg
dui0828 is offline   Reply With Quote

Old   August 21, 2017, 04:56
Default
  #8
New Member
 
Ryan Lee
Join Date: Jul 2017
Posts: 10
Rep Power: 2
dui0828 is on a distinguished road
Quote:
Originally Posted by A CFD free user View Post
Fy is quite right. Did you manage to figure out the problem?

Sent from my SM-G950F using CFD Online Forum mobile app
no_h2.jpg

As you directed, I changed the mass fraction of H2 and O2 as a 0 constant at the electrolyte.
And then there was no penetration of H2 and O2 so far.

Thank you!

P.S. Is it right to check porous zone for the electrolyte?

I have an another question about the current density.

I only got
I_anode = 0.024808 (A/cm^2) ... I_cathode = 0.025970 (A/cm^2).

At the 0.65V fixed cathode voltage and
Mass flow rate (kg/s)
6.00E-07(anode) 5.00E-06 (cathode)

1mmx1mmx50mm 1 turn of the serpentine flow channel. (so 2 long channel.)

I think it is too small to compare other literature data.

Do you have any suspicion about it?
dui0828 is offline   Reply With Quote

Old   August 21, 2017, 05:15
Default
  #9
Fy_
New Member
 
YZH
Join Date: Apr 2017
Posts: 8
Rep Power: 2
Fy_ is on a distinguished road
Quote:
Originally Posted by dui0828 View Post
Thank you very much for the response.
And also about the figure you took.

I set as you mentioned.
(checked the box at the model, which is 'keep updated')
I could make it clear because of your confirmation.

Thank you again.

Would you check the residual is converged?
(though it is just 20 iterations)
Attachment 57888

Obviously it was convergent, but I don't know why it was so fast to reach convergence.

p.s. what field do you research about pemfc?
Fy_ is offline   Reply With Quote

Old   August 21, 2017, 07:27
Default
  #10
Senior Member
 
A CFD free user's Avatar
 
A-A Azarafza
Join Date: Jan 2013
Posts: 219
Rep Power: 7
A CFD free user is on a distinguished road
Quote:
Originally Posted by dui0828 View Post
Attachment 57889

As you directed, I changed the mass fraction of H2 and O2 as a 0 constant at the electrolyte.
And then there was no penetration of H2 and O2 so far.

Thank you!

P.S. Is it right to check porous zone for the electrolyte?

I have an another question about the current density.

I only got
I_anode = 0.024808 (A/cm^2) ... I_cathode = 0.025970 (A/cm^2).

At the 0.65V fixed cathode voltage and
Mass flow rate (kg/s)
6.00E-07(anode)5.00E-06 (cathode)

1mmx1mmx50mm 1 turn of the serpentine flow channel. (so 2 long channel.)

I think it is too small to compare other literature data.

Do you have any suspicion about it?
Is it the Fluent PEMFC tutorial? If yes, obviously, something must have gone wrong with your model definition.

Sent from my SM-G950F using CFD Online Forum mobile app
__________________
Regard yours
A CFD free user is offline   Reply With Quote

Old   August 21, 2017, 09:35
Default
  #11
New Member
 
Ryan Lee
Join Date: Jul 2017
Posts: 10
Rep Power: 2
dui0828 is on a distinguished road
Quote:
Originally Posted by Fy_ View Post
Obviously it was convergent, but I don't know why it was so fast to reach convergence.

p.s. what field do you research about pemfc?
This is just parametric study for serpentine and parallel type flow channel.

When I ran it 90 iterations, the error below appeared and the graph roared.

Error: Divergence detected in AMG solver: species-1
Error Object: #f
dui0828 is offline   Reply With Quote

Old   August 21, 2017, 09:37
Default
  #12
New Member
 
Ryan Lee
Join Date: Jul 2017
Posts: 10
Rep Power: 2
dui0828 is on a distinguished road
Quote:
Originally Posted by A CFD free user View Post
Is it the Fluent PEMFC tutorial? If yes, obviously, something must have gone wrong with your model definition.

Sent from my SM-G950F using CFD Online Forum mobile app
What is the meaning of the tutorial?

If it is single straight channel, no. it is serpentine channel.
dui0828 is offline   Reply With Quote

Old   August 21, 2017, 09:55
Default
  #13
Senior Member
 
A CFD free user's Avatar
 
A-A Azarafza
Join Date: Jan 2013
Posts: 219
Rep Power: 7
A CFD free user is on a distinguished road
Quote:
Originally Posted by dui0828 View Post
This is just parametric study for serpentine and parallel type flow channel.

When I ran it 90 iterations, the error below appeared and the graph roared.

Error: Divergence detected in AMG solver: species-1
Error Object: #f
There is something seriously wrong with your model based on my previous experiences. Check the geometry and BCs carefully.

Sent from my SM-G950F using CFD Online Forum mobile app
__________________
Regard yours
A CFD free user is offline   Reply With Quote

Old   August 23, 2017, 16:29
Default
  #14
New Member
 
Ryan Lee
Join Date: Jul 2017
Posts: 10
Rep Power: 2
dui0828 is on a distinguished road
Quote:
Originally Posted by A CFD free user View Post
There is something seriously wrong with your model based on my previous experiences. Check the geometry and BCs carefully.

Sent from my SM-G950F using CFD Online Forum mobile app
Would you check the picture I uploaded with this article below? (3 pics)
BC_1.jpg
BC_2.jpg
BC_3.jpg

I only correct the boundary condition of below.
Inlet(anode) – mass flow inlet
Inlet(cathode) – mass flow inlet
Outlet(anode) – pressure outlet
Outlet(cathode) - pressure outlet
Terminal(anode) – wall
Terminal(cathode) – wall

As Solution Methods, I used “Simple” or ”coupled”

Did I miss something important for the boundary conditions?


I am very pleased for your sincere answers.
dui0828 is offline   Reply With Quote

Old   September 3, 2017, 12:44
Default
  #15
New Member
 
Join Date: Jan 2017
Posts: 11
Rep Power: 2
Elvis143BRA is on a distinguished road
Did you figure out what was wrong? Currently facing the same problem.
Elvis143BRA is offline   Reply With Quote

Old   September 4, 2017, 12:18
Default
  #16
New Member
 
Ryan Lee
Join Date: Jul 2017
Posts: 10
Rep Power: 2
dui0828 is on a distinguished road
Quote:
Originally Posted by Elvis143BRA View Post
Did you figure out what was wrong? Currently facing the same problem.
As A-A Azarafza advised, when you change the H2 and O2 mass fraction to constant 0 at the fixed value at the cell zone condition, Hydrogen doesn't penetrate onto cathode side (but little into electrolyte.)

And when I changed the mesh, the comparable current was got. But still have some convergence problems.
dui0828 is offline   Reply With Quote

Reply

Tags
fluent, pemfc

Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are On
Pingbacks are On
Refbacks are On


Similar Threads
Thread Thread Starter Forum Replies Last Post
Problems for simulating PEMFC using fluent module slan89 FLUENT 4 March 16, 2017 06:53
How to simulate PEMFC by CFD (Fluent)? dmfo FLUENT 5 February 9, 2017 07:53
Lift and Drag pattern change wit FLUENT 16 and 13 PISO for same mesh n solver setting arunraj FLUENT 0 June 2, 2016 22:58
Error in reading Fluent 13 case file in fluent 12 apurv FLUENT 2 July 12, 2013 07:46
how to define membrane layer boundary in gambit and fluent RICHA FLUENT 2 March 15, 2012 10:37


All times are GMT -4. The time now is 06:38.