NLgeom ''On'' effect on the calculated reaction force

[biofemer]

thread799-158806

Hi

In a large, non-linear deformation of a spherical capsule, when I use NLgeom, at the end of simulation, I get a linear Force-Displacement graph. This is strange, because I know that in this case, force-disp relationship should be closer to that of a second order curve not a line.
Do you have any suggestions why this happens?

Best,
Biofemer

 

[rstupplebeen]

Biofemer,

Likely we will need more details to help more but here are a couple of my thoughts.

You mention NLgeom 'at the end of the simulation' Do you not have it on the whole time?
Do you have non-linear material properties such as hyper-elasticity?

I hope this helps.



Rob Stupplebeen

http://www.optimaldevice.com

https://sites.google.com/site/robertkstupplebeen/

 

[biofemer]

Sorry for the confusion caused by the structure of my message. The NLgeom is "on" during the simulation. But, The results that I get at the end are not consistent with my expectations.

The case that I am simulating is the deformation of an axisymmetric shell between two rigid plates. The shell is elastic and meshed with CAX4RH elements. Therefore, the deformation is not linear. Then I need to use NLGeom for reasonable results in terms of deformation. The visualization of deformation is fine with NLgeom "on" but not acceptable when it is off.

For more details I can also attach the .inp files if required.

cheers,
Biofemer

 

[rstupplebeen]

If you post your INP I'll try to run it.

Rob Stupplebeen

http://www.optimaldevice.com

https://sites.google.com/site/robertkstupplebeen/

 

[biofemer]

Plaese find attached the *.INP file of the mentioned case.

 

[biofemer]

If you have already run the simulation, you realised that the RF2-U2 relationship is linear when nlgeom is on and non-linear when nlgeom is off. However, this should be the other way around. Like the examples of skewplate in linear and non-linear analysis, or axisymmetric mount with nlgeom on and off.

 

[rstupplebeen]

I ran the simulation with NLGEOM on and off. Both are showing nonlinearities with respect to reaction force.

I have always used NLGEOM whenever I'm modeling contact. I hope this helps.
Rob

Rob Stupplebeen

http://www.optimaldevice.com

https://sites.google.com/site/robertkstupplebeen/

 

[biofemer]

Hi Rob

Thank you!
As you see, in your Force-time graph, the "linear" curve, which you probably obtained by toggling nlgeom "off", is more non-linear ( force magnitude-wise uplifted compared to the linear relationship that you can get by extrapolating the common linear part of the two graphs which occurs in very small deformations) than what you got by using nlgeom "on".
Besides, regarding the physics of the problem [1] the blue curve can not be correct.





Reference:

[1] " Determination of the elastic properties of single microcapsules using
micromanipulation and finite element modeling " by Ruben Mercade´-Prieto, in Chemical Engineering Science 66 (2011) 2042–2049.

 

[rstupplebeen]

Biofemer,

Unfortunately I do not have access to the reference you provided. In the reference and in your case is the sphere filled with a contained fluid or gas? If so a fluid cavity will be needed.

With the very large deformations as shown in the image below any simulation without NLGEOM is suspect. Personally I would turn it on and work on the fluid cavity and providing a hyperelastic material model for the rubber.

I hope this helps even though I am not directly answering your questions.

Rob

Rob Stupplebeen

http://www.optimaldevice.com

https://sites.google.com/site/robertkstupplebeen/