Modelling necking with hyperelasticity

[Fredrik]

Hi!

Is it possible to simulate necking with only a hyperelastic model?
Isn't the criteria for necking only a softening material, or even
a minor hardening?

If I let Abaqus do a cuve-fitting, it results in a softening curve for
i.e. Mooney-Rivlin. However, a negative slope makes the material
unstable above that strain level with no convergence as a result.

Is there a way to get necking using only hypelasticity for strains about
5-10 % in unaxial tension in a test specimen.

/Fredrik

[Jorgen]

Yes, it is possible to simulate necking using a hyperelastic model. Although I would not think that it is the best approach.

You are right that in order to get necking you need a "softening" material, and that you can achieve that using certain specific values of the hyperelastic material parameters. You are also right that you might get convergence problems due to the strain localization.

I would use a viscoplasticity model instead of hyperelasticity in order to simulate necking, I might also consider using an explicit FE simulation to improve the convergence.

- Jorgen

[hhspiny]

necking probably no neccessarily requires softening material. if the hardening if very minor, necking could also happens. if you look at true stress-strain curve of typical steel, although there is necking, the true stress actually is not decreasing. only the hardening speed is less than necking speed.

for model this in polymer, I agree hyperelasticity is not the correct model to use, as it might never give you the correct ratio between necking and 'hardening', and hence produce the necking as you observed. i.e., the necking will not be localized in small volume but tend to spread out in your model