Bearing Capacity & Slope Stability

[TerraBoy]

I am working on a project with 30 foot MSE walls on very soft clay. Limit equilibrium tells me the "slope stability" is ok as long as I use some additional geogrids within the rest of the embankment (behind the MSE walls). Traditional bearing capacity calculations tell me the bearing capacity is a problem and the rough checks for lateral squeeze also indicate it is a problem. Finite element analysis also indicate that bearing capacity and lateral squeeze is a problem.

Should a limit equilibrium "slope stability" analysis be able to check for a bearing capacity failure, or is it a different failure mechanism? Even when I constrain my slip surface to look like a bearing capacity failure, it does not seem to be as critical as the finite element and traditional bearing capacity equations tell me it is.

Also, why do the traditional factors of safety for slope stability (1.5 for the drained case) and bearing capacity (about 3) differ so much? Is it all empirical - have we just found what works and we are sticking with that? Is there less uncertainty in slope stability (seems unlikely)?

I have heard a lot of interesting ideas on these questions and thought I would see what the rest of you come up with...

 

[BigH]

A couple of questions - how high is your wall? Is it RE wall or geogrid based?
We are working on such walls here - 8m high walls to 10m high walls on soils with Su = 20kPa. We used wick drains, stage loading, etc. So far have reached the penultimate panels without distress - also about 1/3 the rough estimate of settlement. (very crucial to know preconsolidation pressure even if it is quasi-preconsolidation).
I, too, thought of the various analyses. This is my conclusion - to date. First, I would use slope stability as my primary check. The MSE wall is a flexible wall system that has some integrity, but it is still basically a vertical slope that has just been reinforced to stand vertically. Usually, the backside of the wall is extensive behind the wall - therefore, like a slope. Failure is geared towards the one direction. In traditional bearing, you have the solutions that are based on bi-directional failure surfaces - one with "little" surcharge and the other with "large" surcharge. The bearing capacity is approximately = 2pi x Su. In the slope problem, the sliding is at about 1pi x Su. There is your factor of safety relationship - bearing is 2x that of slope problem. Hence, if you use 1.5 for slope, you would use about 3 for bearing. In reviewing a lot of literature on the net, it seems to be the consensus to use a factor of safety of 2 on traditional bearing capacity analyses. I suggest that you get a copy of BS:8006:1995 which is the British Standard for MSE walls. It is an excellent manual.
Anyway, that is my take at this late hour. Maybe somewhat muddled, but fairly clear in my mind.

ps do a search and this was discussed to a degree previously in other threads.

 

[TerraBoy]

As I mentioned in the first post, the wall is 30' tall. The details of it don't really matter - I am looking for a general discussion. But just so you know, it will probably be geogrid reinforced. Staging isn't a possibility and the soils are heavily OC, so strength gain is pretty small until you get to the NC range. We will use geofoam to make things work.

I am unclear about what you meant by "In the slope problem, the sliding is at about 1pi x Su."

 

[BigH]

1. When you said you were dealing with very soft clays, I wouldn't think that you were so heavily overconsolidated - why I discussed the project we are currently involved with that has soft clays.
2. The slope problem, to a degree, is a one way sliding - rather than the two way (bearing is 2pi in two way, in one way it is 1pi - was trying to show the "correlation" of the two factors of safety generally taken. Of course, in the overall slope analysis you have the strength within the fill itself. In our case, the slip surfaces were "behind" the reinforced system. I also believe that the general consensus of using 2 for bearing computations takes into account the "one way" movement that would develop.
3. I would suggest the limit equilibrium slope analyses as a better indication of the material's behaviour in shear.
Hope this was better.