"Simulating" an Su/P' with a Phi value in PCStabl

In newer stability programs, an undrained strength ratio (Su/P') can be assigned to a soil layer. But since I have a lot of problems already set up in PCstabl files, I still use PCstabl (w/Sted) most of the time.

Now the question:

If I have a construction or other quick embankment loading on a fine grained foundation, I usually approximate the Su/P' in PCStabl by breaking the material into layers and zones with a separate C value assigned to each layer based on the P' calculated for the layer or zone. But I have also encountered engineers who simply convert the Su/P' to an "equivalent" phi value (arctan Su/P') for the fine grained layer. While not theoretically correct, the "equivalent phi strength" gives, and unless I'm missing something, should give, the same results. Is this correct, or am I missing something important?

 

I answered my question.

The engineers' "treat as a phi" approach will produce erroneous, too-low results for the resisting reaction at the base of all slices in the analysis, except those with horizontal bases. At least their errors have been on the conservative side.

 

[dgillette]

I don't think it is necessarily conservative for the slices on the passive side. Those slices are in something more like plane-strain extension, and the pseudo-drained analysis can give them a normal stress that's higher than the vertical preconsolidation stress.

Refer to Chuck Ladd's Terzaghi Lecture: Ladd, C.C. (1991), "Stability Evaluation During Staged Construction," the 22nd Terzaghi lecture, ASCE Journal of Geotechnical Engineering, vol.117, no. 8, pp. 1540-615.

That's probably the best general ref available on undrained shear strength. Been a few years since I read it, but I do recall that there is a discussion of using Phi-cu = arctan (Su/p). Su/p is not a constant - it varies with stress path: CIUC vs CAUC vs CAUE vs plane-strain extension, vs etc.

You work at MSHA with George G.?