Tag Archives: stiff fissured clay

How long can stiff fissured clay slopes stand?

Coming back to the issue of stiff fissured clay slopes, one very important question is the standup time of a slope excavated or formed (natural erosion etc) with higher inclination than the one that would be the outcome using fully softened shear strength.

Skempton (1970) in his paper “First time slides in over consolidated clays”, provided a graph presenting the decay of strength with time to almost fully softened where failure of slopes in London clay occurred (pic 1).

Strength changes with time for London Clay (from Skempton, 1970)

Mesri et al (2003) in his paper provides a graph (pic 2) compiled from numerous case studies of failed slopes in stiff fissured clays in which the vertical axis presents the percent of failure surface length (Lr) in residual condition to the total observed failure surface length (Lt)  and on the horizontal axis the age when the slopes failed. The far left points are of unidentified time. Time to failure range from a couple of years to decades.

Ratio of slip surface segment assumed at residual condition to total slip surface length ploted against age of slope

VandenBerge et al (2013), mention that “The time required to reach fully softened strength might be as little as ten years in some cases, and as long as 60 years in others”.

Potts et al (1997) in the paper “Delayed collapse of cut slopes in stiff clay” presented complex coupled finite element analyses assuming strain softening soil behavior capable of swelling. The analysis requires among other parameters the coefficient of permeability which varies with depth, and coefficient of earth pressure at rest. With this complex parametric analysis they conclude that 3:1 (H:V) slopes produced failures between 11 and 45 years. And steeper slopes fail in less than a decade. 15m high slopes failed after 145 years. All these slopes produced deep seated progressive failure.

It is very interesting to note that they evaluated deep seated failures although the problem of investigation was recent shallow failures in highway cuts and embankments.

Another attempt from Kovacevic et al (2007) presented in the paper “Predicting the stand up time of temporary London Clay slopes at Terminal 5, Heathrow Airport” in which again a complex finite element analysis with swelling behavior and differing Ko was executed. The investigation was for both shallow and deep seated failure. The conclusion of the study was that shallow failures occurred earlier than deep seated and the time to failure was influenced most by the assumptions of permeability and the effect of suction.

Cuts in stiff fissured clays will stand up for an undetermined time before failure occurs when inclined steeper than fully softened strength requires. The time may be from six months to hundred or more years. Investigating the stand up time is very complex and requires sophisticated numerical analysis which also utilizes assumptions in many of the parameters used. Very important parameter is the permeability of the clay in the development of swelling and softening.

This is captured in the classical Terzaghi, Peck and Mesri, 1996 book in which it is stated that “If the surfaces of weakness subdivide the clay into fragments smaller than about 25mm, a slope may become unstable during construction or shortly thereafter. On the other hand, if the spacing of the joints is greater, failure may not occur until many years after the cut is made.”

In my opinion the spacing and orientation of joints is the controlling parameter of permeability in such materials. So highly fissured materials with very close spacing may develop shallow slope failures very quickly. More widely spaced  stiff fissured clays with fewer joint systems can provide significant delay time to failure.

A second important issue is the evaluation of shallow or deep seated failure and how a deep seated failure can be defined. This is something very important that will be covered in another entry.

Slope design in stiff fissured clays

A very difficult issue is the design of slopes or cuts in stiff fissured clays (pic 1). The difficulty lies in the evaluation of shear strength for stability calculations. Much work has been done on this issue especially by Skempton (1964) with his excellent Forth Ranking Lecture titled: “Long-term stability of clay slopes” and many others have contributed significantly on this issue.stiff fissured clay

In the classical Terzaghi, Peck and Mesri, 1996 book the following is mentioned regarding this issue: “Almost every stiff clay is weakened by a network of hair cracks or slickensides.” (pic.2). “If the surfaces of weakness subdivide the clay into fragments smaller than about 25mm, a slope may become unstable during construction or shortly thereafter. On the other hand, if the spacing of the joints is greater, failure may not occur until many years after the cut is made.”Stiff fissured clay exposed next to a sliding soil mass

The reduction in strength with time, due to the presence of fissures, has been attributed to swelling and softening due to water infiltration in this hairline cracks especially when stress relaxation and crack opening occurs in excavated slopes.

Laboratory shear strength evaluation of such stiff fissured clays is difficult because large samples are required in order to include significant number of hairline cracks and even if such samples can be tested, the long term swelling and softening cannot be fully developed in the laboratory.

Duncan and Wright (2005) propose, based also on the work of Skempton (1970) to use the fully softened strength for long term slope stability evaluation of stiff fissured clays that have not undergone any prior movement or failure. This fully softened strength can be correlated to the peak strength of normally consolidated clays. In the laboratory the fully softened strength is evaluated on remolded samples of stiff fissured clays.

A very recent paper by VandenBerge, Duncan and Brandon, 2013 presents the outcome of a workshop that took place in 2011 at Virginia Tech, regarding the fully softened shear strength for stability of slopes in highly plastic clays.

In this paper the most recent views regarding the softening process, the way to measure or estimate the fully softened strength and how and when to use it in stability analysis are presented. Together with the paper of Vanderberge et al, it is worth reading the Lade paper in Engineering Geology (2010), titled “The mechanics of surficial failure in soil slopes” where a power function failure model is proposed for the shear strength of clay for shallow stability evaluation. This criterion is mentioned also in the paper by VandenBerge et al (2013).

I would like to bring attention on some issues which are mentioned also in the papers but relate more to practical issues of the subject:

  1. Great care should be given when site investigation is executed and evaluated in such stiff fissured caly materialsFissured clay. If core samples are not collected then it is very difficult to distinguish between stiff clay and stiff fissured clay. Even when samples are collected, great care should be given to break up some core samples because the fissuring will not be observed as can be seen in the picture 3.
  2. In situ tests such as SPT will produce high values, misleading the investigator to think that a very strong (even cemented) material is found.
  3. Shear strength tests on intact samples will produce high values of cohesion, again misleading the investigator to believe that a very strong material is present. In the laboratory the fissuring may not be reported during sample preparation due to the small sizes required.
  4. The additional difficulty comes on how to persuade the Owner or Contractor about the problems (failures) that Steep stable excavation in stiff fissured clay  may be formed after the slope has been excavated (maybe after very long time). They will evaluate the data from the investigation which show high values and if they are not fully aware about the behavior of stiff fissured clays they will push for a more optimistic design in order to reduce excavation volumes. The situation becomes even more difficult in design – built projects where during excavation the contractor may need to use hydraulic hammers to break up the material and steep slopes are  stable (pic. 4). In such situation everybody “blames” the designer for a very conservative design if fully softened shear strength has been used.
  5. The evaluation of existing stable slopes not designed with fully softened shear strength is another difficult situation. If the fully softened shear strength is used the FS may be found to be even below unity FS<1.0 but the slope is stable for a couple of years after excavation. It is difficult to persuade the Owner of such slopes (usually highway or railway) that in the future stability problems may occur.
  6. Finally it is very difficult to evaluate in what part of the slope and how deep you will use fully softened values especially for high slopes.
  7. Another critical issue is the evaluation of the long term pore pressures to be used in the analysis. In my opinion this is the most difficult issue but I will get back on this in another entry because much could be said.

As a closing remark I would like to state the final conclusion of the Vanderberge et al, 2013 paper: “Consideration of local experience with regard to slope performance, recognition of the possible consequence of slope failures, and application of sound engineering judgment are all essential elements of a comprehensive approach to geotechnical engineering of slopes.”