Stabilising effect of the failed mass

In this part, the stabilizing effect of the slumped debris on slope stability is examined. To this end, slope models were built in the transparent flume and were subjected to heavy rainfall. Two different cases are presented here; in one case debris is manually removed after the occurrence of each, while in the other one, debris is allowed to accumulate at the slope toe.

Experimental results of two sets of experiments under the same rainfall

intensity are presented. The experiments presented below are

a.

161 S90_H15_L35_14 and S90_H15_L35_16 (see Table 5.2). A steady rainfall intensity of Ir=45mm/h was applied on the slopes. In the first batch of

experiments the debris accumulating at the slope toe after each failure is manually removed to mimic the effect of strong atmospheric agents or fluvial or marine erosion that remove all the debris before a new landslide develops. The debris is carefully removed just after a landslide has occurred using a scoop, following the line of the failure mechanism (Figure 5.33), in order to avoid any changes to the new slope profile.

a. b. c.

Figure 5.33. Removal of slumped debris after the occurrence of each failure (test: S90_H15_L35_14)

In Figure 5.34 and Figure 5.35, the evolution of the slope is illustrated for the case that debris is removed from the toe and where debris is left in situ respectively. Figure 5.34b and Figure 5.35b show the slope during the first failure that occurred approximately 16 minutes after the rainfall started. In Figure 5.34c and Figure 5.35c the second failure is plotted. In the case of the removed debris from the slope toe (Figure 5.34c), the second failure takes place earlier at t2=22min than in the case of debris left at the slope toe (Figure

5.35c) where the second failure occurs at t2΄=25min. Moreover, the wedge

coming off the slope face is larger with the failure mechanism being deeper in case of removed debris (see Figure 5.34c and Figure 5.35c). Also note that at

Chapter 5: Experimental study

162 the onset of the second failure the position of the waterfront is shallower in the case of removed debris than the position in the case of debris left in situ, because it occurs much before the second failure in the case of debris left in situ allowing less time for water to infiltrate.

In Figure 5.34d the slope at t3=33min after initiation of rainfall is plotted

when the third failure takes place in the slope with the debris being removed from the toe. Instead, in case of debris left in situ on the slope toe no third failure is observed (see Figure 5.35d). The experiment lasted for 45 minutes in total until the rainfall was stopped. One more failure took place at t4=44min in

the case of debris removal, whereas no further failure was observed on the slope where the debris was left on the slope toe.

In Figure 5.36a and Figure 5.36b the slope is photographed after 45 min for the case of debris removal and debris left in situ respectively. It can be observed that when the debris is removed from the slope toe, the slope front retreats further inwards since more failures occur for the same rainfall intensity and duration. This is in agreement with analytical and numerical models of cliffs subject to weathering (Utili and Nova, 2008, Utili and Crosta, 2011a, Utili and Crosta, 2011b) where it is shown that the weight of the soil debris accumulating on the slope toe contribute to resist subsequent failures and decrease the total final inward retreat of the slope.

163

b.

c.

d.

Figure 5.34. Initiation of three successive failure when debris is removed from the slope toe (test: S90_H15_L35_14), a. initial state (t0=

0 min) b. first failure (t1= 16 min) c.

second failure (t2=22 min) d. third

failure (t3= 33min)

Chapter 5: Experimental study 164

d.

c.

b.

a.

Figure 5.35. Initiation of successive failure without debris removal (test: S90_H15_L35_16) a. initial state (t0’=

0 min) b. first failure (t1’= 16 min) c.

second failure (t2’= 25min) d. at the

165 The aim of this part was to investigate progressive rainfall-induced slope failures and the stabilizing effect of debris accumulated at the slope toe and the following conclusions can be drawn:

• The results suggested that when the debris that accumulated at the slope toe is removed before a new landslide develops; the new failure takes place earlier than in the case where the soil remains at the slope toe. • It was illustrated that the cliff retreat is faster when the accumulated debris

is removed and the amount of soil slides away in the case of soil removal is larger.

• The slumped debris contributes to the stability of the slope, acting as a stabilizing force. This, if not accounted for, can alter the stability number and the failure mechanisms when performing slope stability analyses.

Figure 5.36. Slope profile at t=45min from the initiation of rainfall after the occurrence of a. four failures (debris removal) b. two failures (no debris

removal)

a.

Chapter 5: Experimental study

166