Interpreting Triaxial Results on Remolded Samples

[cpdonahue]

General question regarding using the results of triaxial testing on remolded samples for slope analysis. We are generally testing silty sands or sandy silts. Not really any clays to or even elastic silts in our area. The lab we use often reports effective values in the neighborhood of 34 degrees friction and 200 psf for cohesion. My understanding is that in a remolded situation (trying to model a slope constructed from fill) is that cohesion should be ignored, that only the friction angle of the particles is providing shear strength to the embankment. Why does the lab report back a C value? Would it be more accurate to redraw the line through C=0?

 

[TigerGuy]

If the tangent line to the curves passes through the axis at 200 psf, don't draw it through the zero. Most soils have some cohesion. Similarly, I've conducted many triaxial tests on clay soils and have yet to see an angle of zero. Yet, all the literature says clay has no angle of friction.

 

[EireChch]

I think all thats happening is that silt particles are providing some cohesion.

TG - you are talking about undrained triaxial tests that should have a phi of zero?

OP- Are you doing them at a targeted relative density?

 

[TigerGuy]

Even with undrained, I've never seen a phi of zero. That is likely due to the clay, silt, sand blend of the soils I tested. They weren't a pure clay or sand, so that's the real world, at least where I was.

I always hated trying to remold samples in clay-rich soils. I don't have the patience for them.