Hi,

Of course you can only answer a complex problem from a distance in a very general way. I myself do not know MIDAS but Plaxis. First of all, they must realize that they have a two-material system (rock, concrete). As a result, they have a jump in stiffness between rock and concrete. I suspect that you assumed linear elasticity for concrete and elasto-plastic behavior for (surrounding) rock. It seems to me that you should not use Mohr-Coulomb here, because with stress rearrangement a convergence to the Mohr-Coulombian flow area in MIDAS can not be achieved. You would need to be able to easily control this by applying linear-elastic behavior to both materials. If you insist on MIDAS, use a different flow function and try it with dilatancy angle = phi.

If you get results, please let me know.

Regards

M.L.

Hi Michael:

I'm sorry for answer you so late, but I found the problem. As you said, the problem remains in the interaction with two different models of materials. So I used Mohr-Coulomb in sand (not rock) and the structure of concrete.

I don't remember the results for cylindrical model, but in a square shaft with tiebacks, the results is FoS ~ 1.8. So this approach solved my problem.

Thanks Michael!

Hi Charbel Chapana,

some programs can also cause convergence problems with large jumps in stiffness. The causes can lie not only in the material models but also in geometric nonlinearities.

With computer calculations one should lay very much care into the model formation.

M.L.