Hi everyone,
I am trying to model a steel helical pile with a mass on top in a soil block that I will perform dynamic analysis on (by applying a time history on the base). For now I am trying to simplify the model and just model the helical pile inside the soil and apply a static load on the piles head.
Until now I have successfully run the soil circular block and applied the initial predefined field and performed the second step (geostatic). In the results I see the correct stresses and almost 0 displacement (something ^ -19).
Now I am trying to include the pile into the mix, the issue is I dont want the pile to settle down under its own weight when placed into the soil. What I am currently running as I post this is the same as above but with adding the pile and fixing it (no interface between it and the soil yet). I am planning on adding the interface in the next step.
Is there an easier way to do this or has someone done something similar? If so please give me some advice :).
Keep in mind in the big picture is that I am going to use this model to mimic a full-scale seismic shaking table test I have performed on steel helical piles with masses on top.
Thank you in advance!
Hi Mustafa
The proper way to go is to have your pile as different part and then assemble them into the whole model (body of soil) with the frictional interaction element. This way you can model the soil-structure interaction properly. However, an easy way to have your model running just for adjustment and rough results is to define the whole pile as beam element inside your soil body. This way, the pile is fixed with soil body (100% friction) and it will settle with the soil.
I do not know if there are other easy ways to do this simulation.
Best regards,
Hi Mustafa,
I usually put pile and surrounding soil (as parts) in the assembly, along with interface and appropriate boundaries in-place. If the predefined geostatic gradient and gravity load are correct, then the max displacement after geostatic step would be around 1.E-7 m. I don't see a problem with compiling it all in one first step unless your soil block is sloped or has very low modulus.
Just curious, which boundary are you going to use for the dynamic analysis?
Best regards,
I read somewhere people using infinite boundaries. I will try using that once I perfect my static model and move onto dynamic
You mean apply the shaking on the boundaries? Dont I need some sort of absorbing boundaries to eliminate reflection of waves back into the system?
Hi Thajesh,
Thanks alot for your suggestions. Can you provide me some of the published articles you are referring to?
Thanks in advance!
Come to think of it, why should I link the boundaries? The shaking table container is a laminar box which allows for differential movement between the layers (not rigid boundary moving together).
So for example if my laminar shear box had 5 plates, I would split my soil into 5 layers with the same thickness of the 5 plates then apply the boundary condition you have mentioned?
Another question, how do I know that my model is large enough and no waves are reflected back into the system? Is it by making the boundary as rollers and checking the displacements? If so, then the constraint mentioned above is useless because when my model is large enough my boundaries wont move?
Sorry for the amount of questions :D
I am asking that because rhey have not mentioned anything regarding why they didnt use infinite elements. I understand infinite elements may absorb all waves and not mimic the test exactly. But is the method you suggested going to reflect waves back or?
https://www.youtube.com/shared?ci=dRUR0xZmKuc
Herw is the test im trying to model, should I model the soil as the same size as the laminar hox? Notice how rigid the laminar box is.
Initially, simulate the geostatic condition in the entire soil domain (which you have done). Now, in the next step, remove the soil mass or remove the elements and introduce the pile elements and establish contact. (in this step, either use gravity load or body force of the pile). Run and check the settlement. If is in millimeters, that can be ignored and the remaining simulation steps can be assigned.
Generally, for hollow foundations, the displacement under self-weight is very less. In case, you need zero displacement, then you would have to remove the body force/gravity in the contact establishment step. If you fail to get convergence, apply a small displacement. Once you get the convergence, make a new step and remove the load in the current step. And proceed for the remaining part of the analysis.
Thanks
- Badges
- Science topic