Does anybody have data about either the Mooney-Rivlin constants (or the constants for any other hyper-elasticity constitutive law), or the stress-strain curves of Nitrilic rubber NBR-70 ? I found an example with 2-constants Mooney-Rivlin law having NUXY=0.4996, E=1.28 [MPa], c1=2.765 [MPa] and c2=4.596 [MPa] (parameters for a MOONEY table, not HYPER), which seems to me a bit strange because c2 is bigger than c1, leading mathematically to negative compressibility parameter and bulk modulus...
Eng-Tips is the largest engineering community on the Internet
Intelligent Work Forums for Engineering Professionals
modeling of rubber hyperelasticity 1
- Thread starter cloche
- Start date
-
1
- #2
- Thread starter
- #3
I already tried there: they have experimental true stress / true strain data for uniaxial test of NBR-70, but the list of values is extremely difficult to use as it contains a LOT of repeated lines, and is time-dependent, so that the interpolation function in ANSYS says "incoherent data". I tried to fix the list with Excel, but it's something like 6500 values to check... After 590, I gave up...
- Thread starter
- #5
Hi,
sure, that's I already tried without success, but perhaps we haven't seen the same file(s)... Or, maybe also the curve fitting algorythm has changed since ANSYS version 7.0 I'm using (will upgrade to 9.0 in the next few weeks)?
Can you write me the exact path from where you found the data in Polymerfem website?
Claudio
sure, that's I already tried without success, but perhaps we haven't seen the same file(s)... Or, maybe also the curve fitting algorythm has changed since ANSYS version 7.0 I'm using (will upgrade to 9.0 in the next few weeks)?
Can you write me the exact path from where you found the data in Polymerfem website?
Claudio
hi claudio,
I couldn't get convergence using that data for one of the rubber component analysis. Also the partiallly converged solution doesn't match with my experimental results.
In the web-site, right hand side material models(most accurate material models) -> natural rubber (55 pph CB) click ->below the graph -> click download experiemental data...
you can see bi-axial, planar tension and uni-axial data...
Have you upgraded to Ansys 9.0?
If you get any break through, please share it with the forum......
Regards,
Logesh.E
I couldn't get convergence using that data for one of the rubber component analysis. Also the partiallly converged solution doesn't match with my experimental results.
In the web-site, right hand side material models(most accurate material models) -> natural rubber (55 pph CB) click ->below the graph -> click download experiemental data...
you can see bi-axial, planar tension and uni-axial data...
Have you upgraded to Ansys 9.0?
If you get any break through, please share it with the forum......
Regards,
Logesh.E
- Thread starter
- #7
Hi elogesh,
I performed a first run of analyses with a model of a seal for butterfly valve. I used the M-R constants I wrote in my first post: Ansys seems to be clever enough to "sort" the 2 M-R constants so that C1 is always bigger than C2.
Accordance with experimental tests is acceptable but not really good. I'm now trying to improve the model.
I re-downloaded the uniaxial compression data for NBR-70 from polymerfem, and after some processing I gave it to the curve-fitter. Even with 9 constants M-R, the error norm is very high, but I tried anyway: now, the problem is to find out exactly the best value for the compressibility parameter "d": using the one corresponding to my first 2-consts M-R model (i.e. d=0.0246), the material gets ways too deformable; with 10E-4, it gets too stiff. Probably the best is to make a model of a real test specimen and adjust the data until all the results are in accordance with the experiments, then get back to the seal's model with the "correct" constants...
Another thing I'd like to do is make a sensitivity analysis to find out which of the parameters of the constitutive law (C1, C2, d,...) has the greatest influence on the model; unfortunately, for the moment I haven't enough time to do that. As regards our new seal, due to time issues we will make considerations/extrapolations based on my model + experimental data on existing seal...
Regards,
Claudio
I performed a first run of analyses with a model of a seal for butterfly valve. I used the M-R constants I wrote in my first post: Ansys seems to be clever enough to "sort" the 2 M-R constants so that C1 is always bigger than C2.
Accordance with experimental tests is acceptable but not really good. I'm now trying to improve the model.
I re-downloaded the uniaxial compression data for NBR-70 from polymerfem, and after some processing I gave it to the curve-fitter. Even with 9 constants M-R, the error norm is very high, but I tried anyway: now, the problem is to find out exactly the best value for the compressibility parameter "d": using the one corresponding to my first 2-consts M-R model (i.e. d=0.0246), the material gets ways too deformable; with 10E-4, it gets too stiff. Probably the best is to make a model of a real test specimen and adjust the data until all the results are in accordance with the experiments, then get back to the seal's model with the "correct" constants...
Another thing I'd like to do is make a sensitivity analysis to find out which of the parameters of the constitutive law (C1, C2, d,...) has the greatest influence on the model; unfortunately, for the moment I haven't enough time to do that. As regards our new seal, due to time issues we will make considerations/extrapolations based on my model + experimental data on existing seal...
Regards,
Claudio
- Status
Similar threads
- Locked
- Question
- Locked
- Question