VUSDFLD - Abaqus 6.10 vs. 6.19 1

[Ross11]

Hi,

I am trying to implement a Mohr-Coulomb model in Abaqus 2019 using the vusdfld user subroutine. The user subroutine was initially developed and tested in Abaqus 6.10.

The user subroutine computes two field variables (plastic shear strain and confining stress) that are used to update the friction and dilation angles of the soil. Plastic shear strain is computed using the strain components (PE11, PE22, PE33, PE12....), and confining stress is the average of S11, S22, and S33.

I have previous input files of a single-element triaxial compression test that were run in Abaqus 6.10 using the user subroutine. If I run these in Abaqus 2019 I get significantly different results than what was computed in 6.10.

Has anything changed across Abaqus versions?

I am thinking it may be a result of the plastic strain measure, however, the documentation seems to be consistent across versions.

Thanks,

Ross

 

[FEA way]

That's a really big difference in versions. I assume that you know that Mohr-Coulomb model is built into Abaqus and you want to implement its different version, right ?

 

[Ross11]

I am implementing a modified Mohr-Coulomb model to account for the dependence of friction and dilation angles on the accumulated plastic strain and mean stress. Important when modeling medium to dense sand. The built-in Mohr-Coulomb model assumes friction and dilation angles to be constant.