View Full Version : Simple Tension Validation
I need a clarification. I have a simple tension data of foam rubber. I used a mooney-rivlin model to curve fit the stress strain curve and obtained three coefficients for c10, c01 & c11. Now I tried simulating the actual simple tension test in FEA and used the coefficients obtained previously. I was under the impression that the stress-strain data from this analysis will be matching up with stress-strain data obtained from the actual experiment. But somehow the curve from the analysis seems to be softer than the experimental curve. Is there some reason for this? Iím aware of the fact that stress-strain curve from one mode of deformation will not be enough to represent the material but as a basic logic shouldnít the two curves correlate pretty well?
(1) First, how did you find the Mooney-Rivling constants? Are you sure you did it right?
(2) In general, you cannot expect the predictions from virtually any material model to be exactly matching the experimental data that was used to calibrate the model. In your case, the Mooney-Rivlin model is a very simple non-linear 3-parameter model that estimates the response of certain rubbers. It will not produce highly accurate predictions in many cases.
(3) Since you are working with a foam you need to carefully consider the volumetric response. The Mooney-Rivlin model is likely not a great choice is this case.
It may have to do with how the M-R constants were fit. Commonly, the constants are fit to the "pure" deformation modes, i.e. pure shear and pure dilatation. Try running a pure shear simulation (one element) and see if the fit is better. If it isn't, there is probably an error in the constants.
Keep in mind that simple tension is really a combination of shear and dilatation, so for a real material (especially something complicated like foam) you could have a perfect fit for shear and a perfect fit for dilatation but a crummy fit for tension. You may need to consider the deformation mode that is important in your particular application (maybe it's tension) and base your fit on that mode.
Thanks Jorgen and Grant for your response and suggestions.
Yes Iím aware of the fact that MR model does not represent the foam materials in all modes of deformation and also simple tension test will not be enough to understand the actual foam behavior. Well the reason why I was doing it was because I was trying to compare the stress-strain curves generated from simple tension tests (the easiest of all experimental tests) on two foam specimen of different specific gravities. To understand them better I was trying to create material model for each specimen and compare them on a compression simulation of a hollow foam tube. So to do that, I was trying to characterize the foam specimen using MR model available in MSC Marc.
I agree with both of you, I will probably have to do a simple shear and simple compression along with simple tension to characterize the foam specimen better. MSC Marc does have a foam curve fit but I will have to input a fictive poisonís ratio which I guess I can obtain from the lateral and axial strain.
Powered by vBulletin® Version 4.2.0 Copyright © 2013 vBulletin Solutions, Inc. All rights reserved.