I generally try to model piles and pile group problems using volume piles rather than using embedded beams. However, sometimes the need for using embedded beam elements is required.
As these elements don't have a true interface, what is the most efficient way to validate the normal stresses on the embedded beam element so that I can generate soil springs (kN/m3) which I can feed the structural engineers. (being a function of normal stress and displacement).
I should state this is only truly valid for Plaxis 3D, as there is no 'mean' displacements used to account for pile spacing in the put of plane direction (Sluis, 2012).
In PLAXIS 2D, we indeed model the averaged/smeared effect of the soil body + pile row in a plane strain model.
For PLAXIS 3D, when looking at the normal stresses, I should point you to thiis publication: T.P.T. Dao, Validation of PLAXIS Embedded Piles For Lateral Loading (2011): https://www.plaxis.com/support/publications/validation-of-plaxis-embedded-piles-for-lateral-loading/ in which a detailed study of the file in lateral behaviour is given. Here we see that the lateral behaviour and the resulting skin friction (T2 or T3) can be predicted when the mesh is dense enough and that we do not have too much differential displacements between soil and embedded beam. When you have a concentrated large shear force at the top of the embedded beam, this might be the case.
Many thanks Micha - I was not aware of this research and will read with interest.