Disturbed Soil Properties and Geotechnical Design

Abstract

This book originated with seminars that I gave in November 1999 at Georgia Institute of Technology in Atlanta. I outlined their intended content in the following three paragraphs: There is a fundamental error in the Mohr–Coulomb equation. The proposition that opens Coulomb’s 1773 Essay supposes that a pier is cut by an inclined plane in such a way that the two portions are connected at the cut by a given cohesion, while all the rest of the material is of perfect strength. The pier is loaded by a weight, which makes the upper portion of the pier slide along the inclined plane. Coulomb resolves the load components along and normal to the inclined plane and determines the inclination of the plane for which cohesion and friction combine to give the greatest load. The same result is obtained if Mohr’s circles have a limiting envelope with constant cohesion and friction. The error in this simple analysis is that it omits a component of strength that is due to ‘interlocking’. Taylor in 1948 reported shear box tests on dense Ottawa standard sand. When the upper part of his shear box was displaced laterally by dx it rose up vertically by dy as his dense sand dilated. This is the phenomenon that he called ‘interlocking’. Peak strength t in dense sand occurred at a point where dy=dx was a maximum. Taylor calculated what happened to the work t dx at peak strength. Part went into friction ms0 dx and part went to lift the weight s0 dy on the normal load hanger. This led to friction and interlocking components in the peak strength of dense sand ðt=s0 Þ¼ðm þ ðdy=dxÞÞ. The Mohr–Coulomb equation omits interlocking. After the 1948 publication of Taylor’s book, Terzaghi should have reviewed his interpretation of data of load-controlled drained tests of saturated reconstituted clay soil in a shear box. Terzaghi and Hvorslev had fitted peak strength data to a line with ‘true’ friction and ‘true’ cohesion, but there was an increase of water content in the region of failure and hence a volume increase. This effect is found both in laboratory shear box tests, and in slickenside gouge material in failure planes in the field. Terzaghi and Hvorslev did not have a component of peak strength due to interlocking, hence part of the strength they attributed to bonds among fine soil grains was not due to ‘cohesion’ but to the high relative density of stiff clay soil.