Mohr Coulomb 3d tunneling: uplift issue

[manfro88]

Hello, I'm grateful to you because I read all your usefull posts for my work. So, I'm modelling a classical 3d step by step excavation of a circular shallow tunnel with concrete lining installation after the tbm moving step.
All of this happens in a non cohesive soil wich is set up in abaqus with:
density: 25000 kg/m3
elastic-young modulus: 30000000Pa (30MPa)
elastic-poisson's ratio:0.3
mohr coulomb -friction angle:25°
mohr coulomb -dilatation angle:1°
mohr coulomb -cohesion: 0.1 Pa
no other material properties are specified.

Geostatic step goes ok
but when the excavation procede with the model change that switch off part of the soil, all the near area has positive displacement! Because seems that the soil under the excavation area has a relaxing movement! My soil has up lift also at the crown of the tunnel! I can't explain that behavior! also giving an exagerated density to the lining, than the soils seems to be too stiff! (see image attached)

Any ideas on it could help me
thanks
Luca

 

[dBasement]

I think your density value is off by a power of 10. Even 2500 kg/m3 sounds high for a non-cohesive material. Sorry, I know nothing about the rest of your question.

 

[manfro88]

yes, I only wrote wrong on my previous post. But in every case an higher density could only help my analysis.

 

[ntschwanz]

Luca,

From your initial post you recognize that there will be rebound associated with unloading but you question why there is upward displacement at the crown. How is the model built? It appears you have a finer mesh around the tunnel opening, do you have a liner installed? If so, is it installed immediately following excavation and prior to rebalancing the model? If that's true, the stiffness of the liner could transmit rebound to the ground surface. If you don't have a liner I would have expected collapse unless you are allowing tension in surrounding soil. It looks like your modulus properties are not stress depedant and you may want to move towards a more advanced constitutive model.

 

[manfro88]

Hello,
yes you are right! I've solved my problem changing the model property.
You know that in tunnelling with tbm there is always a loss of volume because the excavation diameter is major than the lining external diameter. In my model i filled that volume with a grout that is completely in contact with the lining and with the soil excavated surface. I was wrong because I was putting a realistic elastic modulus of the grout but that high stiffness did not permit to have any loose of volume. that was irrealistic.
Now I lowered the grout elastic modulus, so it has big strains and negative displacement on the surface appears. Solved!
Luca

The attached image shows a correct deformative pattern