[Geoqus] Plasticity Convergence problem

[bsaad at po-box.mcgill.ca]

Dear Hiva

By the way for Sand you should not have physical cohesion. You might 
get some apparent cohesion due to the suction (small in case of sand) 
that is in turn generated while the dense sand is being sheared.
-By increasing E, you increase the strain energy and the same time 
decrease the plastic dissipation energy, therefore excessive element 
deformation may occur. That is why sometimes involving artifical 
damping helps to achieve convergence in elasticicity.
-When you have some points with low confining stress (e.g. surface 
points) then the poor cohesion at these points will lead to convergence 
problem.
-Regarding the dilation angle and friction angle: it depends on your 
flow rule. I suggest that that you try to write down the relationship 
among the Dilation Angle , The Friction Angle, and The Mobilized Shear 
Strength, respecting the signs adopted by ABAQUS .E.g, according to 
Taylor's assumption (work is entirely dissipated in friction) the 
mobilized shear strength should increase by increasing the friction and 
decrease by  increasing the dilation. Indeed that what I noticed when I 
used ABAQUS Drucker-Prager model.
Please note the above are only my views according to my understanding, 
so feel free to express/defend your points of view regarding these 
interesting observations.

Bassam






Quoting hiva at engr.mun.ca:

>
> Hi everyone,
>
> I am working on 3D model pipe-soil interaction. The soil is dense sand and
> I am modeling the hardening-softening behaviour of the sand by making a
> friction angle, dilation angle and cohesion of the soil as a function of
> magnitude of plastic strain. (The constitutive model is Mohr-Coulomb)
>
> Based on available tests I obtained the hardening rule of this dense sand
> but unfortunately for this set of parameters the convergence is not
> achievable.
>
> This is the error:
> SUBSPACE MOHR-COULOMB PLASTICITY DID NOT CONVERGE FOR ELEMENT SOIL-1

>
> By doing  sensitivity analyses, I understand that the convergence had a
> direct relation with cohesion and dilation angle and has a reverse
> relation with soil young modulus and friction angle.
>
> On the other hand if I want to get convergence I need to do one or the
> combination of these: 1-increase the soil cohesion 2-increase the soil
> dilation angle 3-decrease the soil friction angle 4- decrease the soil
> young modulus.
>
> Does anyone have explanation for this? Why there is such a relation
> between my model convergence and mentioned parameters?
>
> I appreciate any help.
>
> Thanks,
> Hiva
>
>
> --
> Hiva Mahdavi
> PhD Candidate
> C-CORE
> Captain Robert A. Bartlett Building
> Morrissey Road
> St. John's, NL
> Canada A1B 3X5
> Tel: 1-709-737-4378 (office)
> Email: hiva at engr.mun.ca
>
> _______________________________________________
> Geoqus mailing list
> Geoqus at lists.ruhr-uni-bochum.de
> http://lists.ruhr-uni-bochum.de/mailman/listinfo/geoqus
>