PLAXIS 2D - global error tolerance selection

Hi,

My deep excavation model is not converged after large swell occurs at the excavation bottom due to high pwp in the soil beneath the excavation. 

When choosing the default value 0.01 as tolerated error, the deflection is still pretty small when Mstage is 0.94. Then Mstage would not converge to 1, but deflection continue increasing. 

After set tolerated error 0.015, the calculation is completed. And the swell at the excavation bottom is small. It seems there is  a small area or isolated stress points beneath excavation level with negative vertical stress.

Not very sure whether using a slight higher tolerated error is acceptable or not.

Besides it seems the another way to get the calculation completed is optimizing the mesh below the excavation level. However, I dont have the clue how to optimize the mesh properly. What I found if choosing a smaller mesh, Mstage could increase to 0.98; but it will stop at 0.95 or less if I use a further smaller mesh. 

Regards

Zhilong

  • Dear Zhilong Liu,

    Changing the tolerated error means your analysis may be inaccurate, which are not in equilibrium, therefore it is not recommended.

    The fact that the calculation stops can indicate a failure which you need to investigate. In order to find out what the mechanism for failure is, and so possibly determining the cause, these Output plots should be inspected:

    • deformed mesh
    • total displacements (shaded plot)
    • incremental displacement (shaded plot and/or arrows)
    • total deviatoric strain (shaded plot)
    • incremental shear strains (shaded plot)
    • Plastic Points plot
    • when applicable: excess pore pressure plot

    A combination of these plots will usually give insight into what is happening.

    Another aspect is the constitutive model being used. The more advanced constitutive model used, e.g. HS or HSsmall the more accurate results you will get.

    Concerning the meshing, indeed it seems that the elements are relatively large for this analysis. Refining should be done in the area of the excavation (including the walls) and tunnel to increase the accuracy of the results, thus of the calculation.
    Note that there is no straightforward guideline on how to refine the mesh as it highly depends on the model. I would say if your excavation pit is 20m wide, having elements 0.5m big might not be able to capture the behaviour of the soil.

    Finally, make sure that the pore pressure definition (if present) is done correctly as this can also cause issues in the stability of the excavation bottom.

    If you still have problems, please submit a service request.

    Then, one of our support engineers can help you in detail:
    https://apps.bentley.com/srmanager/ProductSupport