Model of cantilever retaining wall in Plaxis

[plaxisbeginner]

Hi,

I tried modelling a cantilever sheet pile wall in Plaxis. The soil is mostly cohesive soil with water level located 3m below the ground surface. I modelled the cohesive material into two soil layers (first layer is 3m thick) with different soil properties. The first layer is modelled as drained mohr-coloumb and the 2nd layer is undrained mohr-coloumb.

My analysis results suggest that the wall goes into reverse curvature at the boundary of the two soil layers. This is also illustrated by the resultant shear and bending moment changing from negative to positive. I am wondering why. In order for the wall to have reversal, it must be restrained at that point typically achieved by having props or anchorage into the soil.

I created another model of the same cantilever sheet pile wall with different soil condition; the two soil layers now being drained. And the analysis results is what I expected. No reversal of curvature.

Could anyone please explain why there is this change on wall behavior?

Thanks very much for your replies.

 

[ntschwanz]

In addition to strength, the wall will respond to soil stiffness variation. When you switched to drained strength properties did you have a corresponding change in soil modulus (it's common to have drained modulus values lower than undrained). Without knowing your model it may be that the stiffness of your wall relative to that of the soil may be high enough in the drained model that the wall acts as a short pile. But with the undrained model it may act as a long pile with an inflection point.