Eurock06 Stress-released Slope Movement

EUROCK 2006 – Multiphysics Coupling and Long Term Behaviour in Rock Mechanics – Van Cotthem, Charlier, Thimus & Tshibangu (eds) © 2006 Taylor & Francis Group, London, ISBN 0 415 41001 0

Stress-released slope movement induced by excavation in fault zone Z.Y. Yang Department of Civil Engineering, Tamkang University, Taipei, Taiwan

J.Q. Hsiao Hsiao-Jong-Quan Geotechnical Consultant, Tauyuan, Taiwan

H.M. Chen Kung-Sing Engineering Corporation, Taipei, Taiwan

ABSTRACT: A 5 m shallow excavation in Hsinchuang fault zone in Taiwan causes a serious slope movement. Numerous tensile cracks indicating slope instability appear at the upper ground surface to 80 m far away from the open cut. A trial excavation in 3 m depth is carried out in field to investigate the lateral displacement behavior of the high-stressed material in fault zone. To monitor the free displacement of the cut, no support is applied and an inclinometer is pre-installed close to the wall of excavation. After excavating, a surface tension-crack opening 5 cm in width is rapidly developed in 2 hours. Finally, the squeezing wall of excavation failed in toppling by the tension crack. This field observation indicates that the faulted material is in a highly stressed condition and is quickly squeezing by the released stress.

1

INTRODUCTION

A shallow excavation with 5 m depth of dormitory building in Hsinchuang fault zone for MRT maintenance in northern Taiwan causes a rapidly lateral slope movement. Building near the excavation was tilted and several tensile cracks extended far from the ground surface of upper slope of dormitory cut. The range of surface cracking progressively reaches to 80 m far away from the cut in five days. It reveals an instability phenomena or rapid landslip in this slope. However, this angle of this slope in landscape (see Figs. 1–3) is smaller than 6◦ . This is very safe in slope stability. According to the experimental data and support design, the excavation and support system actually are on the safe side. In the other hand, the field displacement measurements of inclinometers show that the possible slip surface of landslip in this slope is not a circular, but limited within a certain distance. This implies that the slip surface could be cut off by a vertical tension crack in a certain distance. This study is aimed to explore the failure mechanism of this slope movement in fault zone due to excavation. A trial cut (10 m × 4 m in area) of vertical opening with 3 m depth, no applied support, is performed beside by the dormitory in the field. In order to observe and measure the wall lateral displacement by pre-setup inclinometer close to the face of cut in 50 cm.

Figure 1. Section of soil formation of the dormitory and retaining wall in the slope.

Figure 2. Front view of the gentle landscape and dormitory site.

671

0.5 fault breccias

Axial stress (Kg/cm2)

0.4

Figure 3. Location of dormitory excavation related to retaining wall.

0.3

0.2

0.1 water content : 32.5%

0 0

0.02 0.04 0.06 Axial strain

0.08

Figure 5. Stress-strain behavior of fault breccias and the sheared fracture plane.

Lateral Displacement (mm)

24

Figure 4. Topography of the MRT plan and the fault zone distribution.

2

GEOLOGICAL CONDITION

This location of MRT dormitory buildings is close to a reversal fault -Hsinchuang fault (see Fig. 4). Actually, the formation material of ground within cut is the fault breccias (with mudstone, shale and sandstone) in faulted zone which is highly stressed. As shown in figure 3, the stick-slip in the axial stress- strain curve of mudstone (see Fig. 5) indicates the inhomogeneity of fault breccias. The coarse grain of fault breccias also appears in the sheared surface of mudstone under axial compression. The grain shape of these fault breccias is sub-rounded and the maximum size is about 2.6 cm. The properties of the mudstone are: residual cohesion (C) = 0.4 kg/cm2 , the residual friction angle = 25◦ (from the direct shear test), unit weight = 23 kN/m3 , and moisture content = 8%. The amount of free swelling for mudstone in water content of 15% is up to 3∼12%. The swelling potential of the mudstone impacting on slope stability is the most concern in previous design suggestion for this MRT excavation project. However, the mudstone sample (with fault breccias) contains 84% finer of clay. From, the experimental result, it shows that the permeability of groundwater in this ground formation is very small and the seepage of groundwater is very slow. Thus, the swelling could not be the possible for this rapid slope movement.

inclinometer (at 1 m depth) 16 8 0

2nd Excavating

-8

1st Excavating & Backfill

-16 0 50 (2004 /1/ 1)

100

150

200 250 Time (day)

300

350 400 (2005 /1/7)

Figure 6. Measurement of lateral displacement by inclinometer (at 1 m depth).

3

MEASUREMENT OF SLOPE MOVEMENT

Before the excavation of dormitory building, a retaining wall system of concrete grid-beam with bolts 6 m in long (see Figs. 2–3) was constructed in the upper slope to stabilize the surface deposit. Several inclinometers were setup to monitor the slope stability during the dormitory excavation. The excavation of dormitory located near 20 m to the toe of retaining wall. In design stage, the excavation method is to be cut in 45◦ slope, because of the high safety in the gentle landscape. In March 2004, during the excavating of dormitory (named 1st excavation), a remarkable displacement was monitored by the inclinometers (see Figure 6). An urgent backfill into the excavation space in the field is determined and stop this excavating. Moreover, to guarantee the safe in successive excavating of dormitory building, a series of PIP piles in 10 m long were driving near the toe of retaining wall (Hsiao et al. 2005). After four months, for a second time the excavation (2nd excavation) is opening deep to 5 m. However, a violent displacement as shown in Figure 6 as well appears in the measurement of inclinometer. The total

672

Figure 7. Monitoring inclinometer is pre-setup close to the trial vertical open cut.

lateral displacement at 1 m depth is about 16 mm in a short period. During this period, no remarkable precipitation of rainfall and excavating work take place near this site. In addition, at the upper ground surface, numerous tensile cracks reach to 80 m far away from the open cut. The abnormal influence distance by this excavation in fault zone is 16 times of the excavating depth. The existence of such tensile cracks indicates that in a certain zone of the potential sliding mass the tension force has exceeded the tension strength of the slope medium (Zhang & Chowdhury 1989). This means that the retaining wall and driving piles are insufficient to stabilize the slope movement. The mechanism of slope movement in this gentle slope is ambiguous and thus a trial excavation is determined.

4 TRIAL EXCAVATION A trial cut of vertical opening with 3 m depth (10 m × 4 m area in size) is carried out beside the dormitory in the field. In order to monitor the real lateral displacement close to the wall of excavation, an inclinometer is set up very close to the cut face in 50 cm (see Fig. 7). For measuring the free displacement behavior, no support is applied to this trial excavation. Four sensors are setup at depth of 1.5, 2.5, 5 and 10 m to record the lateral displacement. Each lateral displacement data in the inclinometer is automatically recorded in each 5 minutes. As shown in Figures 7 and 8, a curve tension-crack in 5 cm width at the ground surface (see Fig. 9) is rapidly developed in 2 hours after excavating. The tension crack gradually vertically extended to deep and became wider. A rock block (in 1 m thickness) within this tension zone is squeezing outward step by step. Finally, this rock block is toppling failure in five minutes. The vertical surfaces of this tension cracks

Figure 8. The failure procedure of tension-crack and toppling failure in 2 hours after excavating.

673

Figure 9. Lateral displacement rate at different depth of faulted material by stress released.

in sidewall are very rough shown in Figure 8(c). This demonstrates the mechanism of cracking is fractured in tension. The lateral movement of excavation wall is mainly contributed by the tension-crack opening and wall squeezing. Figure 9 shows the rate of lateral displacement at different depth. It is found that the lateral displacement is mobilized from 10 minutes after complete cutting. Then, the extension rate of tension cracking is kept constant. Finally, the rate decreases due to the separation of this rock block where the inclinometer is setup. The rate of displacement at 1.5 m depth is faster than that at 2.5 m depth. The rate of tensile-crack opening is about 17.5 mm per hour (i.e. 0.3 mm/min) at the top surface for this fault zone. In the other hand, the rate of tension crack opening at 2.5 m depth is 4.2 mm per hour (i.e. 0.07 mm/min). Therefore, the cracking rate is about 4 times per meter in depth direction. The tension crack rapidly propagates to 1 m in depth within 10 minutes.

5 5.1

DISCUSSIONS AND CONCLUSIONS Creep or squeezing

The time-dependent behavior of lateral displacement of fault material actually is linear during excavating. This time-dependent displacement behavior is very different to the typical creep behavior which has the primary, secondary and tertiary creep stages. It is also different to the S-shaped swelling behavior due to ingress of moisture of mudstone. The displacement behavior of fault material in this faulted zone is mainly dominated by the high-stressed or over-consolidated situation. Barla (1995) defined that squeezing is synonymous of overstressing. Squeezing implies rock mass failure associated with volumetric expansion due to overstressing. The reasons of squeezing in this study are: (a) excessive faulted pressure; (b) the dissipation of

Figure 10. A safe excavation in normally-consolidated clayey soils in Hsinchu of northern Taiwan.

residual stress. The observation of this research indicates that the faulted material in this MRT maintenance site displays the property of a squeezing ground. 5.2 Critical tension crack A normally-consolidated formation of cohesive soil in northern Taiwan being excavated deep into about 3 m is shown in Figure 10. This figure demonstrates that the excavation is very safe. Actually, the excavation is smaller than its critical excavation depth in a cohesive soil. Therefore, the critical depth of excavation to form a tension crack in this trial excavation is more than 3 m. However, a visible tension crack and remarkable displacement appeared rapidly. The result shows that the slope movement is primarily caused by progressively stress releasing. The displacement rate of faulted material is much faster as we know. As a result of stress-released, additional earth pressure released from the in-situ stress will applied to the man-made supporting system. To ensure the safe of excavating within fault zones, this additional earth pressure must be taken into consideration. The approach of small-region excavating in turns to reduce the rate of released-stress is preferential. In practice, a quick supporting scheme or pre-cast retaining system is suggested. However, the quantity of released stress or displacement rate needs greater study. REFERENCES Barla, G. 1995. Squeezing rocks in tunnels. ISRM News Journals, 3/4: 44–49. Hsiao, J.Q.,Yang, Z.Y., Chu, C.C. & Chen, H.M. 2005. Lateral slope flow induced swelling in fault gouge excavation. 11th Conference on current researches in Geotechnical Engineering in Taiwan, paper No: 06. Zhang, S. & Chowdhury, R.N. 1989. Identification of critical slope failure surfaces with critical tension cracks. Rock Mechanics as a Guide for Efficient Utilization of Natural Resources. Rotterdam: Balkema.

674