How to increase the undrained shear strength? 1

[pietro82]

Hi all,

I'm analyzing the stability of a slope with SSAP. My colleague provided me the soil shear strength data of the site and the slope is rather stable for the long-term analysis (i.e. using c'-phi') but not for the short-term analysis. (i.e. using cu). The reason is the slope is long and the cu is rather small. Is there any way to increase the undrained shear strength?

Thanks.

Best regards,

Pietro

 

[EireChch]

Hi Pietro,

There are any amount of ways to improve the undrained shear strength such as ground improvement. Other options to improve stability without improving shear strength such as soil nails, regrading etc. If you give more details (i.e. cross section with soil parameters etc)i am sure you would get better, more detailed answers.

 

[oldestguy]

Many slope stability situations are due to presence of ground water of many types or anther. Getting rid of water is a primary treatment in most cases. Again, as stated above, with details you provide we might help.

 

[pietro82]

The average gradient is 23%, the length is 200m and the total difference in altitude is 82m, soil shear parameters are cu=40.10kPa, dry bulk density=17.40kN/m^3 and wet bulk density=19.10kN/m^3. There is water table at 20m of depth. With my simulations, I get a very small FS (around 0.3). How can I increase it? The presence of the water table doesn't affect the simulation.

 

[EireChch]

Thats quite a big slope. What kind of failure is it? Shallow or deep? Can you post an image of critical failure plane. Soil nails etc might be an option but if its deep seated they will do nothing as they need to extend beyond your failure plane.

Also, have you done a sensitivity analysis. What shear strength do you need to achieve a FoS of 1.3 (or what ever your chosen FoS is).

I am out of the office at the minute but i might do a quick analysis when i get back to the office.

 

[pietro82]

it is a deep failure.

Here a screenshot of the failure plane:

https://s8.postimg.org/3olgom3c5/Capture2.png

As I did not expected, the failure plane goes to the second soil layer, that is a bit stronger cu=27kPa.

I tried to carry out a sensitivity analysis but only the cu significantly affect the FS.

 

[EireChch]

I havent got a chance to do an analysis myself but it does look like you have a bad slope if that failure plane is correct. 40kpa layer on top of a 20kPa is giving you the problem. Its a very big slope and anything you do will be expensive.

I would advise that you get your analysis sorted. Are all your inputs correct, have your questions the Cu data?

I would also try a different software too as a comparison. You say above that the watertable is at 22mm depth however you analysis pic shows a blue line (watertable?) very close to the surface. Is this correct? Lowering the groundwater level using horizontal drains or similar should make a massive improvement. This wont help your analysis now but it would be a recommendation in my report, direct all stormwater from above the slope to the stormwater network.

Can you analyise groundwater level as it currently sits. This analysis should give you a FoS of greater than one because the slope is stable. But that is my next question, are there signs of instability already. Cracks etc.

 

[HENRYZAU]

Is it a cut slope, a embankment slope or an existing slope? Or what's the dominant stress path?

 

[Okiryu]

I tried to model the same slope that the OP posted above. I got FOS of 0.18 by Spencer method, see attached. If the slope is currently stable, you will need to double check the source of your shear strength parameters. Anyways, to reinforce that slope will be very costly, so I will spend some money for doing a proper investigation before spending $$$ for reinforcing that slope.

 

[Mad Mike]

I can only imagine it's a natural slope- a cut/fill embankment nearly 100m high is unlikely.

I have dealt with several landslides of a similar geometry- the first option is to avoid the site completely, since stabilising these features is never economic.

Failing the above, the choice needs to be made whether a structural solution or an earthworks solution- based on what you've said, I assume there is no indication of the slope being unstable (look for tension cracks along the crest). If the slope appears stable on site, I would check all soil parameters and rerun your analysis. As another poster mentioned, your very elevated water table is suspicious and if your software is functioning, will affect the analyses dramatically.

As a final note- your profile shows 3 layers- I'm assuming the lowermost layer [3] is stable bedrock, the middle layer [2] is a ?40m thick? uniform soil deposit of relatively low strength and the upper layer [1} which is nominally thick will not affect your analysis at all. Are you sure this profile is correct? Your entire failure surface is located in layer [2], which your analyses show to be completely saturated. Check this carefully!

All the best,
Mike

 

[BigH]

There are, at least in my view, a number of issues.

1. Su values of 40 kPa - then he talks about a lower layer with 27 kPa? Does this seem reasonable? You have, say 40 m of slope above the failure plane - using Su/p' = 0.23, then Su should be in the order of 150++ kPa if the clays are normally consolidated. If this is a natural slope as it appears, it would seem to me that the clays are at least normally consolidated - i.e., they are not undergoing continued primary settlements. I find the Su values highly unrealistic - UNLESS, there had been a previous slide and this is a "residual" Su value ona pre-existing failure plane . . .

2 Secondly, if you have the c'-phi' values AND you know the groundwater/porewater pressure values, you should be using the effective stress analysis for both short and long term. Undrained strength values are used when you do not have a good handle on the actual porewater pressures.