Learning Forum

[Battery]

 

Hello,

For my thesis I'm designing a batterypack for an application. Currently I'm trying to determine the U and Y coefficients for the NTGK battery model. Because I'm not able to do any experiments yet I used a graph delivered by the battery supplier (Voltage in functon of time in minutes). You can see the graph below. I extracted the values from the graph and wrote them in three .txt files (1C;2C and 3C). Completely below the text files are shown.

With the parameter estimation tool I've readed the .txt files into Ansys. The tool calculated the Y and U parameters as shown below. I also added the model options (50Ah 3,2V pouch cell). As you can see the parameter bo isn't correct because it's shown in a red box. This parameter should be a positive number.

The experimental and fitted results are shown below. As you can see these results doesn't match the first graph at all and I can't figure it out what I'm doing wrong.

Text file 1C rate (time in seconds | voltage in V):

Crate 1.0

0001e+00  3.400e+00 

0068e+00  3.252e+00 

0194e+00  3.252e+00 

0332e+00  3.249e+00 

0466e+00  3.247e+00 

0597e+00  3.239e+00 

0862e+00  3.231e+00 

0994e+00  3.225e+00 

1260e+00  3.217e+00 

1392e+00  3.213e+00 

1657e+00  3.206e+00 

1922e+00  3.198e+00 

2320e+00  3.188e+00 

2585e+00  3.172e+00 

3248e+00  3.119e+00 

3703e+00  2.889e+00 

3832e+00  2.576e+00 

3878e+00  2.368e+00 

3949e+00  2.146e+00 

3989e+00  2.003e+00 

 

Text file 2C rate:

Crate 2.0

0001e+00  3.400e+00 

0060e+00  3.187e+00 

0200e+00  3.178e+00 

0333e+00  3.172e+00 

0465e+00  3.165e+00 

0598e+00  3.156e+00 

0730e+00  3.151e+00 

0863e+00  3.141e+00 

0995e+00  3.131e+00 

1261e+00  3.108e+00 

1515e+00  3.058e+00 

1656e+00  2.976e+00 

1744e+00  2.838e+00 

1776e+00  2.750e+00 

1802e+00  2.599e+00 

1825e+00  2.504e+00 

1850e+00  2.383e+00 

1892e+00  2.195e+00 

1901e+00  2.101e+00 

1913e+00  2.012e+00 

 

Text file 3C rate:

Crate 3.0

0001e+00  3.400e+00 

0001e+00  3.134e+00 

0134e+00  3.118e+00 

0267e+00  3.109e+00 

0400e+00  3.098e+00 

0532e+00  3.090e+00 

0665e+00  3.073e+00 

0798e+00  3.055e+00 

0919e+00  3.027e+00 

0997e+00  2.990e+00 

1041e+00  2.955e+00 

1075e+00  2.912e+00 

1087e+00  2.869e+00 

1113e+00  2.823e+00 

1155e+00  2.623e+00 

1201e+00  2.472e+00 

1198e+00  2.388e+00 

1226e+00  2.219e+00 

1226e+00  2.180e+00 

1250e+00  2.009e+00 

Extra information:

This is how the console looks like and what I've entered as input parameters in Ansys:

And here are the DOD in function of the U and Y parameters:

 

 

[Battery]

Bump

[Rob]

What happens if you run the Fluent models on? Ie is it a rogue point and the curves then fit?  Just wondering if it's a function of the polynomial fitting rather than an error. 

Note, as far as I understand batteries they are things that go into toys (not necessarily small) so I'll push this to the wider team: that will take time. 

To add, thanks for helping with the other thread.  

[Peymangh]

HI.when i want to write the name of txt file it says that "could not find file please specifi the file again"

do you know what the problem is ?

[Rob]

What file?

[Battery]

I'm not sure if I understand your question correctly but if you want to find the .txt files in Ansys through the parameter-estimation-tool this is the path:

[Battery]

What happens if you run the Fluent models on? Ie is it a rogue point and the curves then fit?  Just wondering if it's a function of the polynomial fitting rather than an error. 

Note, as far as I understand batteries they are things that go into toys (not necessarily small) so I'll push this to the wider team: that will take time. 

To add, thanks for helping with the other thread.

@Rwoolhou, thanks I hope I can find an answer soon because I can't continue with my thesis without the thermal models. I don't exactly know what happens when you run the parameter-estimation-tool and I don't know yet how the fitting results should look like. Documentation about this is very limited. Thanks for the reply

[Indraneel]

@Battery

Where do you keep those .txt files for input? Fluent Console says 

file name for curve 1 [] Crate 1.0

Could not find file "Crate". Please specify the file again!

 

Could not find file "1.0". Please specify the file again!

I tried this path to keep the file Crate 1.0. It does not work. Where am I going wrong?

D:IndraneelBTMS model30Ah pouch_filesdp0FFFFluent

[Battery]

@Battery

Where do you keep those .txt files for input? Fluent Console says 

file name for curve 1 [] Crate 1.0

Could not find file "Crate". Please specify the file again!

Could not find file "1.0". Please specify the file again!

I tried this path to keep the file Crate 1.0. It does not work. Where am I going wrong?

D:IndraneelBTMS model30Ah pouch_filesdp0FFFFluent

 

Did you define the file? It needs to be something like "crate-1.txt"

[Pratik]

@battery

Hello, Sir. I am looking for model parameters for the NTGK model in Ansys Fluent for a single battery.

But when I mention the text file name in the command window of Ansys, it gives an error as “Could not find file ‘1C.txt’. please specify the file again!”
Do you know what the problem is?

Where to keep those .txt files for input? is there any specific directory/location for saving files?

[Indraneel]

@battery

Thank you. The thing that went unnoticed with me and probably everyone who asked similar doubt here is... mentioning .txt after the filename.

Thanks.

[Rob]

Using spaces in file names is usually a bad idea. Fluent was originally written for UNIX which doesn't allow spaces: as I used Fluent before it was available on Windows I never use spaces. 

[Battery]

@Indraneel, what about your results? Do you get correct parameters through the parameter-estimation-tool?

[Battery]

Bump

[Battery]

Bump

[Rob]

We've kicked this, and your other post towards one of the Academic managers. Short answer is I don't know, and it's getting too detailed for us (ANSYS staff) to be able to help via the Community so I can't push to a colleague. 

Please be patient: you've found a hole in the support system and we can't fix it at a technical level. 

[Rob]

I've spoken to a colleague.

 

The problem is likely caused by the input data: you need to make sure that

Y coefficient > 0  

So tidy the raw data at DOD =~0 so that it is smooth as this could be the cause of the issue.      It's covered in training, but missing in documentation: we've requested an enhancement to DOC. 

 

 

[Battery]

I've spoken to a colleague.

The problem is likely caused by the input data: you need to make sure that

Y coefficient > 0  

So tidy the raw data at DOD =~0 so that it is smooth as this could be the cause of the issue.      It's covered in training, but missing in documentation: we've requested an enhancement to DOC. 

Thanks for speaking to your colleague. I know the Y coefficient should be greater than zero, it's show in a red box. I tried different things/changes with the input file but the parameter is always negative. Is it possible that you can provide me with an example of an input file of your colleague? This might help me alot to find the issue.

[Rob]

The process involves tidying up the data: it's likely that the curve will need checking/amending. More generally, most polynomial related issues are due to sharp changes in data curves so I'd focus my attention near x=0 on the graph. 

[Battery]

The process involves tidying up the data: it's likely that the curve will need checking/amending. More generally, most polynomial related issues are due to sharp changes in data curves so I'd focus my attention near x=0 on the graph. 

I'm sorry but I've changed the data several times by adding more data to the .txt files but I can't get it to work. It would be really helpfull if I could see a working .txt file.

[Rob]

Change one of the curves to avoid the spike at the y-axis (x=0) and see what that does. 

[Battery]

Change one of the curves to avoid the spike at the y-axis (x=0) and see what that does. 

@rwoolhou after a long time I finally managed to solve the issue. The spikes at the y-axis indeed caused these false results. But there is still one problem as you can see in the graph below which shows the experimental data and the fitting results. All graphs (0,5C; 1C; 2C; 3C) should end at 2V as described in the input .txt files. The second graph shows how it should look like. I'm not sure if this influences the Y and U parameters but this should be solved. 

 

Below you can see the parameters and thermal models at 1C and 5C. They seem to be correct.

 

[Rob]

Please check:

1) the Min. Stop Voltage is set to below 2V
2) what is printed in the Text User Interface window when the solver stops?
3) if more time steps are needed
4) if smaller time steps are needed per time step to ensure convergence. 

[Battery]

Please check:

1) the Min. Stop Voltage is set to below 2V
2) what is printed in the Text User Interface window when the solver stops?
3) if more time steps are needed
4) if smaller time steps are needed per time step to ensure convergence. 

@rwoolhou. The graph of the fitting and experimental results in my previous post is a graph that is outputed as an .xy file when you run the parameter-estimation-tool. 

Documentation:
To help you assess and debug the fitting process, after running the fitting tool, ANSYS Fluent automatically creates a folder called fittingresult in your working directory. In this folder, Fluent stores the following automatically generated files:

 

• ntgk-curve#-t#-crate#.dat: A fitting result text file with three columns of data (t, experimental data, and fitting data) for each Crate curve. You can use this file in a third-party software for further processing.

• ntgk-curve#-t#-crate#.xy: An XY plot file that can be plotted in Fluent to examine the fitting quality.

• ntgk-generatepic.scm: A scheme file that can be run in Fluent to draw all the generated XY plots and automatically save them as image files.

• ntgk-t#-u-function.xy, ntgk-t#-y-function.xy, and ntgk-t#-intermediate.xy: XY plots of fitted U, Y and V~I curves.

The min. stop voltage doesn't influence the output of the parameter-estimation-tool and time-steps aren't used during the parameter-estimation-tool. Below you can find what is printed in the console:

[Battery]

@rwoolhou. To add to my previous post in the image below you can see the comparison between the generated .dat file through the parameter-estimation-tool and the input .txt file that I've written. In the image below you can see on the left the .dat file and on the right the input .txt file. The values are correct but the parameter-estimation-tool isn't using all the data in the .txt file for some reason, this is why the curves aren't correct. The .dat file stops at a time of 7187.08602483162 seconds.

 

[Battery]

I discovered that Ansys isn't taking these values into account:
- 0,5C: all values above 120 minutes or 7200 seconds
- 1C: all values above 60 minutes or 3600 seconds
- 2C: all values above 30 minutes or 1800 seconds
- 3C: all values above 20 minutes or 1200 seconds

So for example if I add this line to the 0,5C.txt file:

"7195.084562266433           3.1128621286722262"

It wil work.

But why is this because I need values up to 8250 seconds for 0,5C.

 

[spatel]

This could be due to the current limitation of 20 DOD levels when using Table option - I noticed you are using 30 DOD levels. Perhaps you try (if not already done) using Polynomial option:

[spatel]

[habtom]

Hi Battery,

I am new to MSMD battery model, Can you help me how to find the file using parameter estimation tool

 

Appreciate your help,

Thank you

[Battery]

This could be due to the current limitation of 20 DOD levels when using Table option - I noticed you are using 30 DOD levels. Perhaps you try (if not already done) using Polynomial option:

@Spatel, thanks I normally use the polynomial instead of table option but it doesn't change anything.

[Battery]

Hi Battery,

I am new to MSMD battery model, Can you help me how to find the file using parameter estimation tool

Appreciate your help,

Thank you

What file? If you're using .txt input files put them here. 

[KK2128]

hi battery,

my Y coefficient value is coming as negative 

my data curves has no spikes still the value is negative 

I would be really grateful if someone could help me out 

 

 

[Battery]

hi battery,

my Y coefficient value is coming as negative 

my data curves has no spikes still the value is negative 

I would be really grateful if someone could help me out

What's the difference between the first and the second curve? Is this experimental data for different C-rates? Problem is probably caused by the data in the .txt files but sadly enough it isn't described in the docs how detailed these should be. 

[Battery]

@rwoolhou. To add to my previous post in the image below you can see the comparison between the generated .dat file through the parameter-estimation-tool and the input .txt file that I've written. In the image below you can see on the left the .dat file and on the right the input .txt file. The values are correct but the parameter-estimation-tool isn't using all the data in the .txt file for some reason, this is why the curves aren't correct. The .dat file stops at a time of 7187.08602483162 seconds.

 

Still can't find why Ansys isn't using these values. Any ideas?

[Pranavnagaveykar]

Hello,

I am doing a thermal study of lithium ion batteries as well. Can you share me your contact details? I would like to connect and discuss about this. I have made a similar single cell model but my thermal values are not showing any difference. Would appreciate your help

Thank you

Pranav N

mail : pranav.nagaveykar@gmail,com

Whatsapp : +918605187137

[Battery]

Hello,

I am doing a thermal study of lithium ion batteries as well. Can you share me your contact details? I would like to connect and discuss about this. I have made a similar single cell model but my thermal values are not showing any difference. Would appreciate your help

Thank you

Pranav N

mail : pranav.nagaveykar@gmail,com

Whatsapp : +918605187137

 

I've sent u mail

[Pranavnagaveykar]

Dear Battery ~

I need your help to understand how to enter the data using the parameter estimation tool in ANSYS FLUENT MSMD model.

I saw that we can the command "define models battery-model parameter-estimation-tool" which asks us which model we want to use
1NTGK 2ECM  3xyz
if we select ECM
it asks us to enter following data :
number of temperature levels -
number of SOC-level curves per temperature level -

after entering that if asks us to give a text file input in following format
SOC     - 
Current -
Temperature -
time_1 voltage_1
time_2 voltage_2
time_3 voltage_3
and so on..

Then it asks the solver method
1 4P solver
2 6P sover
and then another selection option
1 J
2 LH

Once I load the file and select either of these option it gives a gaussj error and then it hangs for 2 minutes and my fluent crashes.

I tried it many times but same result.

Please help to resolve this issue and let me know how can I use this parameter estimation tool correctly.

Thank you

[Battery]

I’m not familiar with the ECM-model. You need to make sure that the .txt files are not opened when uploading them. I know if you have them still open in Notepad Fluent will crash.

[sabitbiruni]

Hello Battery,

Currently, I am doing research on thermal modeling of cylindrical li-ion battery for my final task as college student. I'm very interested to discuss about this topic. I have some issue on analyzing the this model with NTGK method. Can you give me your contact person? Would be grateful if you can help me.

 

Regards,

Sabit Biruni

 

mail : sabitakbarb@gmail,com

 

[Battery]

Hello Battery,

Currently, I am doing research on thermal modeling of cylindrical li-ion battery for my final task as college student. I'm very interested to discuss about this topic. I have some issue on analyzing the this model with NTGK method. Can you give me your contact person? Would be grateful if you can help me.

Regards,

Sabit Biruni

mail : sabitakbarb@gmail,com

Hello Sabit, what exactly is your issue with the NTGK-model?

[sabitbiruni]

I got an issue with the input parameter on parameter estimation tool. I stuck on define how to get DoD interval until got 1.00000 in last line after input the .txt. After many trial and error and after that i got a clue that i've wrong on define the time interval in .txt. But until now i still don't understand how to fix it. I want to ask you how to define time interval in your analysis so it's can make DoD from 0,- until 1.000000 like in your screenshot ?

And can you help me to give your last 0,5C.txt - 4C.txt that was close to the experiment data sir ?

Thank You

[alluch94]

Dear battery,

currently doing my project on thermal modeling of lithium ion battery.
can i know what is txt file for the estimating parameter and how can i get them?
this is my email
aliabbas861@gmail.com

[TSKhanh]

.

Dear battery,

Currently, I am doing research on thermal modeling of cylindrical li-ion battery for my final task as college student. Can you share me your contact details ? can i know what is txt file for the estimating parameter and how can i get them? I would like to connect and discuss about this. I have made a similar single cell model but my thermal values are not showing any difference. Would appreciate your help.

Thank you

Nguyen Duc Khanh

mail : khanh.nguyen181@hcmut.edu.vn

Whatsapp : +49355268928

.

[TSKhanh]

.

Dear battery,

Currently, I am doing research on thermal modeling of cylindrical li-ion battery for my final task as college student. Can you share me your contact details ? can i know what is txt file for the estimating parameter and how can i get them? I would like to connect and discuss about this. I have made a similar single cell model but my thermal values are not showing any difference. Would appreciate your help.

Thank you

Nguyen Duc Khanh

mail : khanh.nguyen181@hcmut.edu.vn

Whatsapp : +49355268928

.

[jfash]

.

Hello, I am doing a similar project and wanted to ask how you originally extracted the values from the graph below? Did you have data points for this already?

And also what was the electrochemistry of the cell you were using? The profile of the discharge curve doesn't look too dissimilar to an NMC battery discharge curve so I may want to use a similar polynomial function to yours if that is okay? (I would love to reference your work)

.

[salih kaya]

hi did you solve the problems? ı have the same problem, and  ansys doesnt helping :/

[Rob]

You emailed us on 3rd, and were pointed here within about 10minutes, it's then taken a week for you to post the above into a thread from 2019.  I'm assuming you've read https://ansyshelp.ansys.com/account/Secured?returnurl=/Views/Secured/corp/v231/en/flu_ug/flu_bat_MSMD_sec_use_battery_model.html%23flu_bat_MSMD_sec_NTGK_input  

[salih kaya]

link is not opening, ı have no custemer id

[Rob]

Click on Help, then paste the above into the browser that opens. 

[salih kaya]

when ı click on help, it wants custemer id

[Rob]

Clicking on Help in Fluent requests a customer ID? 

[abdur raheem adam]

hi, i am doing the same battery capacity for my project as well. i have the data for internal resistance vs SOC. can i do parameter estimate or Y or U table from my data? or should my data only be in time and voltage as mentioned above?

[Rob]

You'll need the data in whatever format Fluent requests: I don't know if what you have is sufficient. 

[ASHUTOSH SHARMA]

Hello battery,

Can you please help me with an similar issue as yours in the input files of PET.

email: ashutosh.sharma1@my.jcu.edu.au

Cheers!

[Pratik]

 

Hello, Sir. I am looking for model parameters for the NTGK model in Ansys Fluent.

But when I mention the text file name in the command window of Ansys, it gives an error as “Could not find file ‘1C.txt’. please specify the file again!”
Do you know what the problem is?

Where do you keep those .txt files for input?