Modeling soil in ANSYS workbench v14.5: Problems using Drucker Prager via APDL commands

[tomdeturck]

For my master thesis I have to construct a model of a tower and it's underlying soil using ANSYS v14.5.
For now, i have made a foundation slab embedded into a layer of soil. I have no trouble converging to a solution by defining the soil using only it's density, elasticity and poisson factor.

However as soon as try to include Drucker Prager (defined by cohesion, internal friction and dilatancy) by using APDL commands onto the solid "soil" geometry volume, it doesn't converge anymore.

http://imgur.com/Qq6VO6X

http://imgur.com/wyM37ol

Anybody have experience modelling soil in ansys?

 

[omidomidi]

when you include drucker prager in your model, it takes in to account cracks generated in the model, try incrementing load slower and defining a displacement convergence of 5 % ,also playing with material parameter cohesion and changing it may help to converge better.

http://ansys24.ir

Solving engineering problems with ANSYS

 

[rickfischer51]

Are you loading with force and/or gravity loads? If so, that's a problem. Simple Drucker-Prager (not the extended D-P) assumes no hardening (i.e. perfectly plastic response), so if your load produces equivalent stress significantly higher than the yield stress, you'll get a plastic instability and no convergence. Ramp the loads slowly, save the intermediate results, and look at your last converged substep. Also, you could try loading with a deflection, but that may not make physical sense.

Rick Fischer
Principal Engineer
Argonne National Laboratory