Hi,
I have a 3D model simulating TBM tunnel construction through a clay layer (modelled using undrained A, HSS). At the end of construction, I have excess pore water pressures (pwp) in the model, as expected. I ran a consolidation analysis after the last phase to dissipate the excess pwp to zero.
While checking the results, I noticed that there is a change in the total ground vertical stress (eg. about 10% in my model for the node at tunnel crown) between the final phase of the undrained analysis and the following consolidation analysis phase, despite having no additional loading/external stress change. Is there any explanation for this behaviour? Everything else is working as expected (Pexcess=0, Pwater = Psteady+Pexcess, s=s'+Pactive).
I was curious and ran a simple 2D model with this approach, and observed the same phenomenon. Any inputs on this is appreciated!
Thanks in advance,
Mihir Pillai
Hi Mihir,
Interesting question, so I will follow the answer from Plaxis. But just thinking out loud, don't you over-simplify the problem by trying to follow the soil mechanics principles in a complex model? I felt so lazy to try, so maybe you might try this on a simple model without any structural elements and complexity, just one layer of soil consolidating and see the total pressures. If total pressures are constant in that case, we might say that there is an interaction with tunnel due to settlements during consolidation.
Thank you Berk for the reply. Really appreciate you taking the time to think about this.
I had the same idea as well, so had run a simple 2D model with just undrained material (and no tunnel), added a surcharge in the next phase (Phase 1) and let excess pwp dissipate to 0 (Phase 2). The total stress remains the same between Phase 1 and Phase 2, as expected. Same is true for a simple unloading problem simulated by removing some soil at the top.
Like you said, I am likely oversimplifying a complex problem with simple soil mechanics principles. The change in total stresses (or the change in effective vertical stresses not being equal to the excess pwp stresses) is possibly due to the soil-structure interaction.
Would be interested to hear the Plaxis opinion on this.
Dear Mihir,
When an area in the soil has excess pore pressures that are being dissipated, the excess pore pressures first dissipate along the edges of the area with the excess pore pressure. So along the outside of the area with excess pore pressure the effective stresses increase, which makes the soil shrink (higher effective stresses = negative volume strain). This shrinkage happens all around the outer perimeter of the area with excess pore pressures - and if the outer perimeter shrinks, it will try to compress the soil inside which causes extra excess pore pressures in the middle of the area: , the total stresses there will increase temporarily. This is known as the Mandel-Cryer effect.
With kind regards,
Dennis Waterman
Thank you Dennis for the explanation, that was very helpful.
Just need a bit of clarification on what is meant by "the total stresses there will increase temporarily". Does it mean the increase in total stresses should not be seen once all the excess PWP in the model dissipates to 0?
Thanks and regards,
Mihir
In the perfect case the Mandel-Cryer effect is demonstrated by a sphere of soil with uniform excess pore pressures and then when the excess pore pressures start dissipating from the surface of the sphere inwards, the surface of the sphere shrinks creating additional excess pore pressures inside the sphere that will indeed later dissipate again. Hence, the increase of total stress is temporary. In your case however this area of excess pore pressures does not have a free surface, but is surrounded by soil. So if the area shrinks it will unload the surrounding soil and this will cause some stress redistribution due to which the increase of total stresses may not be completely undone when all pore pressures are dissipated; there can be a residual increase.