Piled-Raft Modelling in PLAXIS 3D

Question

Hi, there. I want to ask a question regarding piled-raft modelling in PLAXIS 3D. 
 
 Currently, I am modelling the raft as a volume and the piles as embedded beam. As in the theory stated by Poulos (1994), the load-settlement curve of the piled-raft foundation generally should look like this. 
 But currently in my model, the model has collapsed even though it is still far-far away even from the ultimate load of the raft foundation. I found out after that the reason for the collapse is because the capacity of the embedded piles have all been mobilized. This causes me a real problem because it means that I can't be able to model the Curve A-B as specified in the picture I attached. It's as if the model behave like a slab on pile rather than a piled-raft. Can someone please provide me an alternative or solution regarding this problem? Appreciate it very much

Hung Ho Manh · Answer

Hello Edward, With an enquiry related to a specific project, we recommend you to submit a SR using the link here: https://apps.bentley.com/srmanager/ProductSupport

Hung Ho Manh · Answer

If you do not want us to check the project, so, you need to understand why it fails first. General idea to check it is shown by our colleague here: https://communities.bentley.com/products/geotech-analysis/f/forum/243632/plaxis-3d-tunnel-modelling adjusting possible things after that.

Hung Ho Manh · Answer

communities.bentley.com/.../plaxis-3d-tunnel-modelling

Hung Ho Manh · Answer

No way to run the model if the model fully collapses due to bearing capacity because no convergence can be achieved between driving force and resistance. This is also a real physical behavior.

Dennis Waterman · Answer

Dear Edward, 
 In principle if all the embedded beams have fully mobilized their capacity but the raft still has capacity, the calculation should not stop because there is no failure yet. So considering that it does stop in your case, it either means that the raft has no more capacity either, or it's a numerical issue. For instance that the calculation needs to do a few unloading steps to redistribute stresses before it can continue to load the raft, and it considers the successive unloading steps as a sign of failure. This can be overcome to increase the Number of unloading steps in the Numerical control parameters of the Phase settings. But there can be other numerical reasons as well, which would then indeed by model-specific as my colleague already suggested. In which case you would indeed have to send us a Service Request. 
 With kind regards, 
 Dennis Waterman