Mooney-Rivilin Constants Calculation

[Byoung-Chul Lee]

Hi, everyone!

I'm trying to design a EPDM rubber pad with ANSYS to reduce vibration or noise. And I'm going to do planar compression test to get a stress-strain curve.
Then If I have a experimental data, how can I calculate Mooney-Rivilin Constants?
I know how I can calcualte constants with ANSYS. But I want to make Mooney-Rivilin Constants calculation program with MS Excel or Matlab.

My question is when I have an experiment data, how can I calculate Mooney-Rivilin Constants?

Thank You.

[skacel]

First you have to derive a proper stress-strain relation and then find the constants associated with minimal distance between your experimental data and this function. This leads to optimization problem.
Instead of ANSYS, you can also try a freeware fitting utility HYPERFIT that can help you to fit the material constants for various hyperelastic models:

www.hyperfit.wz.cz

[nandish.rv]

Mooney-Rivilin Constants Calculation
to find the mooney rivlin constants curve fitting method can be used.in that it can be selected for veriable parameters, means 2-parameter model or 3,5,9 parameter models. with this it can observe that the which parameter model is suitable for the model and we can directly get the model constants.to get this we require experimental data.

[napoleonm]

Hello skacel

Have you got the hyperfit exe that run properly,, I tried to download it and dosen't works,
could you share it with me please?..

Thanks

[skacel]

Hello Napoleonm,

could you send me your PC configuration? I tried to run the HYPERFIT application on several PCs and it works. In one case however the application did not work (AMD Athlon 64 + WIN XP 32-bit), but I still do not know why. Your configuration may help me to solve that problem. Is there any system-generated error after the instalation?

Regards,
Pavel Skacel

[napoleonm]

Thanks Skacel.
now is working, it seen because I change the file name.. and run it, it did not reconized it.

now it is working..

call my attetion the 8 moths trial..

[nepolean.r]

Hi Pavel,
I dowloaded from www.hyperfit.wz.cz

But this error msg comes when I try to install it. ( filename change is not doing the trick)
"The setup files are corrupted. please obtain a new copy of the program"

following is my PC config.
Windows XP-SP3 32 bit
Intel(R) Core(TM)2 Extreme

I am working with Hyperelastic models for various simulations. Your application seems promising and I would love to try it for my thesis..

Can you kindly send me setup files to rajasenn@mail.uc.edu
(or) nepolean.r@gmail.com

thanks a lot

[skacel]

Hi Nepolean,

it is nice that (as you wrote me):
"It is working now in my original PC... I didnt do any changes but somehow it is working.."

I hope that HYPERFIT software will be helpful for you.

[nfarbod]

Hello Pavel,

I am using the Hyperfit v 1.x at to retrieve the Mooney Rivlin
constants for my Uniaxial, biaxial and planar tests on rubber. I can easily extract the
constants for each individual test, but how can I get one set of constants for all three
tests? (as in a set of C10 & C01 that takes into account all three tests). I need this to
represent my material under global loading conditions. Your reply would be very much
appreciated.

Regards,

Farbod Nassiri

[skacel]

Hello Farbod,

sorry for late reply:

you can add more data-sets into one fit of constants to take into
account more tests.

1/ Just push the "Add new data-set" button in
the left top corner of application form and load/add your data-file. Repeat this for any additional
data-set (number of data-sets is not limited). This will add other data into application for further
fitting. Do not forget to change the "Protocol" listbox for individual
data-set.
2/ Then you can arrange the sub-windows (containing the data-sets)
through menu (Window - Cascade)
3/ When you fit this arrangement now the resulting parameters take into
account all the data-sets.
4/ You can simply "activate" or "deactivate" any data set to try to fit the parameters from only a sub-set of data without the need of deleting and re-loading the data from files.

Pavel Skacel

[JMB]

Hello,

Regarding Hyperfit, I tried to run it using Wine in Ubuntu Linux, but get the following error:
fixme:actctx:parse_depend_manifests Could not find dependent assembly L"Microsoft.VC90.CRT"
err:module:import_dll Library MSVCR90.dll (which is needed by L"C:\\Program Files\\Hyperfit\\Hyperfit.exe") not found
err:module:LdrInitializeThunk Main exe initialization for L"C:\\Program Files\\Hyperfit\\Hyperfit.exe" failed, status c0000135

Questions:
1. How does one fix the problem?
2. Can one run this under wine in a linux PC?
3. If not, is a linux version available? If not, when will one be available?

Regards,
JMB

PS: Suggestion: It appears that the D/L process is very error prone. There are frequent dropouts and truncations. Publishing an md5sum (or similar) verification method on the website will enable a user to verify the integrity of the downloaded software.

[skacel]

Hello,

unfortunately I do not have Wine/Linux instalation here to fix the problem. Did you try to run HYPERFIT using WINDOWS? (this is the primary platform for which Hypefit is compiled).

ad3.: Unfortunately Linux version of HYPERFIT utility is currently not available. However, HYPERFIT is written in Python, therefore should be a way how to provide Linux version. I do not have any experience with Linux platform, but if you are ready to help me with that, please contact me directly at skacy@email.cz and we can try to do something with that.

Pavel Skacel

[Lowell]

when I select the 'Add new data-set' button at the top-left, a Windows explorer window opens. I select the text file with my data (column 1: strain, column 2: stress) and nothing happens. The experimental data column in Hyperfit generates the correct number of cells but they're all blank. Below is the name of my text file and the data inside of it. Please advise.

I've created a detailed screen shot that more easily explains my problem. You can find it in the attachments along with a copy of the text file below.

R04_edgestress_Test03.txt
0 0
0.016320475 0.000574235
0.025222552 0.001577914
0.037091988 0.002531952
0.059347181 0.003301407
0.080118694 0.003983397
0.086053412 0.004562064
0.095667656 0.005135314
0.112759644 0.005609181
0.118694362 0.006108952
0.142433234 0.006414094
0.148367953 0.00690372
0.167655786 0.00716385
0.172106825 0.007598318
0.176557864 0.00789952
0.200296736 0.008257457
0.206231454 0.008561023
0.225519288 0.008871583
0.232937685 0.009120779
0.238872404 0.009437741
0.259643917 0.009622225
0.28041543 0.009917025
0.271513353 0.010097569
0.292284866 0.010339674
0.305637982 0.01065112
0.318991098 0.010860031
0.341246291 0.011023239
0.345697329 0.011211663
0.351632047 0.011477308
0.367952522 0.011696562
0.368941642 0.011839973
0.385756677 0.012022191
0.395647873 0.012288624
0.419881306 0.0124802
0.423343225 0.012924912
0.43768546 0.012939096
0.446092977 0.013189769
0.462908012 0.013077089
0.471810089 0.013259209
0.484668645 0.013440541
0.49851632 0.013674668
0.505934718 0.014019209
0.517309594 0.014038908
0.534124629 0.014045212
0.545004946 0.01403024
0.556874382 0.014573449
0.575667656 0.014377145
0.583580613 0.014978762
0.588526212 0.015228647
0.599406528 0.015391856
0.621661721 0.015093806
0.628090999 0.015643318
0.642927794 0.016193716
0.648862512 0.015775796
0.670623145 0.015600767
0.681503462 0.015608647
0.682492582 0.016043903
0.705242334 0.015842084
0.715133531 0.015987071
0.726013848 0.016274878
0.744807122 0.016524074
0.741839763 0.016528014
0.764094955 0.016581596
0.786350148 0.01694426
0.792284866 0.017040393
0.807121662 0.01708905
0.827893175 0.017158392
0.837784372 0.017309781
0.84371909 0.01737075
0.859050445 0.01743231
0.884272997 0.017461465

[skacel]

Hello Lowell,

you used wrong separator in your input file (see documentation of HYPERFIT). Columns have to be separated by comma: ",". This was chosen to be directly compatible with ANSYS FE package. I attached the corrected data file that works well when fitted to Neo-Hooke material model (see the attached figure generated by HYPERFIT).

Pavel Skacel

[Lowell]

Thanks for the data-set solution. I must have missed the instructions regarding the comma.

Can you explain why the lines that are fit to the data using Hyperfit and ANSYS look almost identical and yet the estimated values are slightly different?

See the attachment where I use the Ogden model in both programs.

[skacel]

I cannot read the numbers (estimated parameter values) from the figure you attached (higher resolution would be beneficial). However the described situation is typical for phenomenological models such as Ogden type. Moreover, Ogden model is highly nonlinear and therefore non-linear regression method (iterative) are used for parameter estimation and probably the method used in ANSYS is different from that used currently in HYPERFIT software. Therefore different local minima can be found by different numerical method (in general) and these minima can represent similarly "good" fit (or also very different fit), mainly depending on starting point that is crutial in any non-linear regression.
The second thing you should check (when you are comparing ANSYS and HYPERFIT) is the minimized objective function (the type of residuum). ANSYS and also HYPERFIT allow you to use "sum of absolut delta sigma" or "sum of relative delta sigma". You should check if you are using the same option in both programs.

[nicolasponce]

hello i need to simulated a tire, and y want to simulate rubber as real as posible. so i find that you can aply a mooney rivlin but for thi one y cant fine the constant y need. also i fine that y can aplay some damper spring nonlinear models but y realy dont know how to aplay then and the other opcion y found a explicit material call rubber2 but y dont know how good it will be for the simulation. I know that simulating i tire is realy dificult so I just want to strat with a litltle model were i can see how it works. can some one please help me by guiding me what to do, like what steps i need to follow.
thaks i am willing your answer will help me

[kmgraju]

Dear friends
I am trying to model mooney revilin model for artery wall in ANSYS .I plan to use 3 parameter model.material constants C01, C10, C11 and incompressibility parameter D1 have been taken from literature. My question is: should I provide any other properties like young's modulus poisson' ratio, density etc to the model?

Thank you for your kind reply


Regards

Govind

[skacel]

Dear Govind,

C01, C10, C11 and D1 constants are enough for specification of the material model you choosed. You don't have to specify any other material parameters (except for density if you plan to calculate with inertia forces -> dynamics).
Nevertheless, there is more suitable model for artery wall in ANSYS. Anisotropic exponential model (according to Holzapfel et al., 2000) is available in ANSYS 14.0 (and later).

Regards,
Pavel Skacel

[anton]

hello,
I would like to know how to implement the load deformation curve in Ansys without having to enter a point at a time, but all the points directly.
thanks a lot.

[skacel]

Dear anton,

if you mean Curve Fitting sub-module, there is possibility to import the whole stress-strain curve via a text file containing all the data points. This is the standard way I use.