Hi, I would like to determine the bearing capacity of mat foundation in PLAXIS 3D. I have tried two methods, one where I applied a surface displacement along the z axis (free along the x and y axis), the other where I applied a prescribed surface load along the z axis followed by a strength reduction phase (I would obtain the bearing capacity by multiplying surface load by MSF obtained). The results obtained are different.
Can i check what is the difference between both methods mathematically? Which method would be the correct method to determine the bearing capacity of a mat foundation? When applying the prescribed load, should I add a dummy element to model the mat foundation?
Hi Alvin,
When a prescribed displacement (displacement control) is applied, it is considered for a rigid footing, or the surface to be applied displacement has the same displacement. Therefore, when you use the load (load control), please make sure that the surface has also the same displacement, so, use very stiff footing for example. If not, it is a flexible footing, resulting in different load-settlement response. So, basically both approaches are fine but should be consistent when comparison.
When you apply a plate (or volume) for the load control, it should be very stiff to make sure it is a rigid footing.
Can I check that for both displacement control and load control, given the same boundary conditions and soil type, the load-settlement response will be different but the ultimate bearing capacity generated will be the same?
If you can make sure the same behavior or mechanism failure, the ultimate bearing capacity should be the same
Dear Hung, can you please elaborate on how the ultimate load should be extracted in a load control test?
1- For my Master's thesis, I'm modelling a plate with a huge point load at it's top (10,000 kN for example) to ensure shear failure of soil, then the ultimate load is extracted from the Mstage vs Uz (Mstage multiplied by 10,000 kN). No safety phase (Msf) is used to get the load!
2- Secondly, It's seems not possible to mitigate the load-displacement curve of the General Shear Failure for high dense sand similar to the one proposed by Terzaghi, as from my model the load becomes perfectly horizontal when it reaches the ultimate capacity.
3- Finally, I'm having a hard time locating the ultimate load for the cases of loose soil where Local and Punshing Shear Failures take place. Visually, the point where the curve becomes more erratic is not clear so I'm currently using the second derivative as an indication. Do you have any recommendation?
Thanks a lot in advance!
Mohamad
Dear Mohamad,
1) That is the correct way to do it. I'm puzzled about your last remark though ... because it's not possible to get a failure load out of a Safety analysis so why would one try.2) The plotted curve suggests softening behaviour after reaching the peak strength. To model that in PLAXIS you would have to use a constitutive model that includes softening behaviour. If softening is not included in the constitutive model, it cannot show in the results. Unfortunately softening is not so easy to calculate in finite elements, so therefore most of the models we have do not include softening and the ones that do are not the easiest models to work with, like the Hypoplastic model and the OC-Clay model. Hence, it's probably not worth trying just to simulate a literature graph.
3) Local and punching shear are often quite mesh dependent, hence when and how they occur depends on the element size around the footing. So it is indeed not easy to determine - there are some graphical methods described in literature (also used to determine preconsolidation pressure from oedometer tests, for instance) that you could look into, but it will always have an amount of uncertainty.
With kind regards,
Dennis Waterman