Modelling rubber interlayer between metallic plates under high strain rate

[Selçuk Başdemir]

Dear all,

I try to model a sandwich structure consisting two metal plates and a rubber interlayer. A high speed projectile (around 5000 m/s) is hitting the sandwich. I realized that any rubber model in Lsdyna does not support EOS. In this case, how can we model material response of rubber, especially volumetric response? As far as I know, these hyperelastic models simply use P=K*deltaV (pressure=Bulk modulus*volumetric strain) relation for volumetric response but I think this is not best option since strain rates are very high.

Also, is incompressible assumption valid under these strain ranges? İn real case, viscoelasticity takes role and initial shear modulus of the rubber increase enormously while bulk modulus is nearly constant. Therefore, i think, incompressible assumption is no longer valid.

I would be grateful for any help

Thanks

[James Kennedy]

Dear Selcuk,

 

See if this is of some help:

 

Best,

Jim Kennedy

 

For high strain rate rubber in LS-DYNA, common material models include `MAT_077` (Ogden)* and `MAT_181` (Simplified Rubber), which can be parameterized to capture rate-dependent behavior through techniques like stress relaxation data. For very high strain rates (>105is greater than 10 to the fifth power), `MAT_11` is an option, while `MAT_183`* is particularly useful when the material response under compression at varying strain rates needs to be precisely characterized using experimental stress-strain data. 

 

Common LS-DYNA material models for rubber

• `MAT_077` (Ogden)*:

This model uses a parameter-based approach and can incorporate rate-dependent behavior by using stress relaxation data (Prony series). 

• *MAT_181 (Simplified Rubber):

This model uses tabulated input data, making it a good alternative for representing rate-dependent behavior, especially when experimental stress-strain data is available. 

• *MAT_183:

This model is designed to fit any set of uniaxial experimental data exactly and can be used to simulate the elastomer's behavior under a wide range of strain rates, as discussed in this ScienceDirect article. 

• *MAT_11:

This material type is suitable for modeling materials at very high strain rates (>105is greater than 10 to the fifth power) and can handle pressure-dependent yield behavior. 

• `MAT_234` and `MAT_235`**:

These are specific models developed for body armor and loose fabric under high velocity conditions, as shown in this YouTube video. 

• LS-DYNA Tutorial Material Models

Mar 22, 2014 — and of course there are other thermal effect and so forth which uh which we talked about uh 18183 is simplified one uh...

Key considerations for high strain rate rubber

Experimental Data:

The accuracy of the simulation depends on having good experimental data, especially for rate-dependent models. Data from high strain rate tensile and compression tests are crucial for calibration. 

Calibration:

Material model parameters need to be calibrated using the experimental data. Tools like MCalibration can be used to fit data to constitutive models, including those available in LS-DYNA. 

Model Instability:

Some models may have instabilities at certain values for parameters like bulk modulus and shear modulus, which can lead to poor simulation results. Baseline values often need to be analyzed and adjusted. 

• Viscoelasticity:

Rubber is a viscoelastic material, meaning its behavior is dependent on the rate at which it is deformed. Capturing this non-linear, rate-dependent behavior is essential for accurate simulations of impacts and other high-speed events. 

Model Choice:

The best material model depends on the specific application and the type of data you have available. *MAT_181 and *MAT_183 can be good choices if you have experimental stress-strain curves, while *MAT_077 is useful if you have stress relaxation data. 

--
You received this message because you are subscribed to the Google Groups "LS-DYNA2" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ls-dyna2+u...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/ls-dyna2/d05d4aa8-5302-4f29-ac59-3f5b04f457e1n%40googlegroups.com.

[T Kim Parnell]

Thank you Jim Kennedy. 

These references address many questions on rubber and polymer materials. 

For 

Selçuk Basdemir

Are you performing any material testing?

If yes, what tests are you performing?

-------------------------------------
T. Kim Parnell, PhD, PE
Fellow, ASME
Life Senior Member, IEEE
Vice-President, Silicon Valley Engineering Council (SVEC.org) 

SVEC “Legends of Engineering”:
https://Luma.com/x4t0x4bu

Parnell Engineering & Consulting
E-mail: kim.p...@stanfordalumni.org
https://parnell-eng.com 
https://www.linkedin.com/in/ParnellPEC   <-- My LinkedIN Profile

To view this discussion visit https://groups.google.com/d/msgid/ls-dyna2/DM6PR13MB330715E558F09E610B46CCB691F5A%40DM6PR13MB3307.namprd13.prod.outlook.com.

[Selçuk Başdemir]

Thanks a lot Jim Kennedy. I wll go over material routines that you mentioned but I am still confused about modelling volumetric behaviour without an EOS at high strain rates. 

Dear Kim,

I have only uniaxial test results for the rubber. But if I decide to use MAT_077_OGDEN with viscoelasticity as Jim suggested, I need stress relaxation test for viscoelastic parameters. Also, Ogden model can capture S-shaped tension behaviour nicely if you calibrate your model with uniaxial and biaxial/equbiaxial test data. Using only uniaxial tension data for calibration of Ogden model parameters is not suggested generally. Therefore I need to perform biaxial/equbiaxial tests and simultaneuosly fitting. MCalibration is great but I have own codes for several rubber material models calibration.

20 Ekim 2025 Pazartesi tarihinde saat 20:43:42 UTC+3 itibarıyla parn...@gmail.com şunları yazdı:

[Oğuz Kağan GENÇ]

Hi Selçuk,

That is one of the fundamental assumptions in conventional hyper-viscoelastic modeling for rubber-like materials that material is already very resilient to volume change, i.e., incompressible or nearly incompressible, even under static loading conditions. Thus, it is assumed that the volumetric response does not change under different strain rates and that is why the volumetric response is only defined through bulk modulus in hyperelasticity, not in viscous part. The viscous part only deals with deviatoric response as a function of time, strain-rate, temperature etc. 

How do you measure the bulk modulus as a function of strain rates? What kind of tests you made? 

If the bulk modulus is very different under different strain rates, this is another story. You must check it by tests if you have a doubt about it. 

I guess it is not easy to carry out volumetric tests with that high strain rates. What I do is to test samples under uniaxial loading conditions with different strain rates and use 3D DIC measurements, calculating bulk modulus accordingly. 

Long story short, you do not need EOS regarding hyperelasticity due to the assumptions I just explained.  It is assumed constant. 

Kagan

To view this discussion visit https://groups.google.com/d/msgid/ls-dyna2/936a5280-51ef-44f9-b089-a35ac6c1fae6n%40googlegroups.com.