GEOTECHNICAL PERFORMANCE OF EMBEDDED CAST-IN-SITU DIAPHRAGM WALLS FOR DEEP EXCAVATIONS

Transcription

1 25/3 PAGES 3 38 RECEIVED ACCEPTED A. RAHMAN, M. TAHA GEOTECHNICAL PERFORMANCE OF EMBEDDED CAST-IN-SITU DIAPHRAGM WALLS FOR DEEP EXCAVATIONS Abdel-Rahman, Mohamed Taha Assoc.Prof.Eng. Abdel-Rahman, Mohamed, Taha, PhD., Ain shams University, Faculty of Engineering, Structural Engineering Dept., Cairo, Egypt. Tel.: , abderahman_m_t@yahoo.com. Research field: soil mechanics, foundation engineering ABSTRACT KEY WORDS An underground multi story garage has been constructed in El Tahreer Square near Omar Makram mosque, Cairo. The performance of a 13.2 m deep multi- propped excavation in medium dense to very dense sand layers is investigated. The vertical side walls of the excavation were supported by cast-in-place fully reinforced concrete diaphragm walls (27. m depth and.8 m thick). Field measurements of any wall movements and groundwater levels were compiled using inclinometers, deep settlement points, and peizometers. The results of the in-situ monitoring are compared with those obtained from the numerical analysis. The case history presented herein focuses on the behavior of the soil interaction of the diaphragm wall and the general ground deformation regimes in the vicinity of the site. Deep inclinometers are good tools to measure and follow the lateral deformation of any construction activities at the site. They must be deep enough to obtain reliable results. The mathematical results show that the measured soil movements tend to be less than the predicted values by about 2%. diaphragm wall, deep excavation, field monitoring, numerical modeling 1. INTRODUCTION Economic growth in Egypt during recent decades has forced the construction industry to improve the quality of our urban environment, including moving services below the ground surface. Several large projects involving underground structures have been executed during the past 2 years. Most of these projects represent elements of an overall plan to implement the first two lines of the Greater Cairo Metro or replace the existing wastewater networks for the largest urban areas of Greater Cairo and Greater Alexandria (El Nahhas, 1999). More active utilization of the underground spaces has started recently in order to solve traffic problems through the construction of road tunnels along with several underground car parking projects within the congested urban areas of Greater Cairo (El Nahhas, 23). This paper is concerned with the case history of an underground multi story garage, referred to as El-Tahreer Garage No. 2. This garage was constructed in El Tahreer Square near Omar Makram mosque, Cairo, over an area of about 5 m 2 with four underground floors. The northern boundary of the garage faces the El-Sadat station of the First Line of the Cairo Metro. The nearest point of the El-Sadat station is about 6.2 m as depicted on the El-Sadat station drawings. This project represents one of two underground garages to be constructed at the El-Tahreer Square. Figure 1 shows the general layout of El-Tahreer Garage No. 2, while Fig. 2 shows the cross section of the El- Sadat Metro station. The analysis and design of the supported deep excavations is considered one of the most difficult tasks facing geotechnical engineers. The effects of the construction procedure and the characteristics of the soil layers and groundwater should all be taken into consideration. A very important task is to address the structural effects of the garage construction on the El-Sadat Station. The first phase of the study was based on the results of a two-dimensional 3 25 SLOVAK UNIVERSITY OF TECHNOLOGY rahman.indd :28:43

2 25/3 PAGES 3 38 finite element analysis conducted to model the interaction between the underground garage and the metro station. Implementation of in-situ monitoring programs during the construction of the supported deep excavations was necessary to verify the design assumptions and assure the safety requirements N Omar Makram Mosque Tahreer Street 6,12 El-Sadat Metro Station 6,43 6,87 El-Tahreer Garage (2) M M Platform of Tahreer Mogama a Fig. 1. Location and general layout of El-Tahreer Garage No. 2. (21.5 m) Top Slab (18.5 m), 1 cm Thickness Second Slab (14.5 m), 5 cm Thickness Diaphragm wall Thickness = 8 cm Length = 3 m Raft Foundation (7.63 m), 12 cm Thickness 1.14 Diaphragm wall Thickness = 8 cm Length = 3 m Fig. 2. Schematic diagram of the El-Sadat Metro station s cross section. (Burland & Hancock, 1977; Kovari, 1984; El-Nahhas, 1987; Sakurai, 1987; El-Nahhas, et al., 1989; Finno, et al., 1989; Hansmire, et al., 1989; Urlich, 1989; Di Biango, 1991; El-Nahhas, 1992; Boone, et al., 1992; Sharma, et al., 21; El-Nahhas & Morsy, 22). Without in-situ monitoring and feed-back analysis, it is not qualitatively possible to assess how conservative the design of the supported deep excavation is. In this paper, the in-situ geotechnical performance of a 13.2 m deep multi-propped excavation in medium dense to very dense sand layers is investigated. Field measurements of the lateral wall deformations and groundwater levels were compiled using inclinometers, deep settlement points, and peizometers. The results of the in-situ monitoring are compared with those obtained from the finite element analysis. The case history presented herein focuses on the behavior of the soil interaction with the diaphragm wall and the general ground deformation regimes in the vicinity of the site. 2. THE SITE S SUBSURFACE CONDITIONS The geotechnical condition at the project site was determined from subsurface investigations prepared by (Ardaman Ace, 1999). The investigation was comprised of 12 boreholes of a depth ranging between 48 and 5 m below the existing ground surface with field testing, including Standard Penetration Tests. Four boreholes were drilled at the site selected for this paper, and eight boreholes were located at the site of Garage No. 1, which is currently under construction adjacent to the Nile Hilton Hotel. The soil formation at the site consists of surface man-made clayey fill (3. to 6. m thick) followed by a sand layer which extends to the end of all the boreholes. The particle size of the sand is fine to medium, with a relative density ranging between medium to very dense. A clay layer appeared at depths ranging between 3. to 39. m in all the boreholes sunk at the garage. The thickness of this clay layer ranged between 1. to 4. m. The water table within the site is located at about 3.6 m below ground level. Fig. 3 illustrates the average subsurface condition at the site and the results of the Standard Penetration Tests. The challenging aspects of this project that required special attention were: 1. The stability of the earth-retaining structures and the associated ground movements, especially during construction, particularly the effect of the ground movement on the nearby Metro Station; 2. The high permeability of the soil and the nearby source of groundwater flow from the Nile. GEOTECHNICAL PERFORMANCE OF EMBEDDED CAST-IN-SITU DIAPHRAGM WALLS rahman.indd :28:46

3 25/3 PAGES End of Stratum Legend Soil Description FILL, Sandy Silty Clay, Lime Stone Fragments N Stage 1: Excavation to Depth 3.65 m 3.65 m Yellow, Medium dense, MEDIUM GRADED SAND, Traces of Silt and Calcareous Materials Stage 2: Installation of Top Two Slabs and Excavation to Depth 6.35 m From G.S. 3. m.65 m Top Slabs Second Slab 2.7 m m Top Slabs Yellow, Dense to Very Dense,MEDIUM GRADED SAND, Traces of Silt. -2 Stage 3: Installation of The Third Slab and Excavation to Depth 9.5 m From G.S. 3. m 2.7 m 2.7 m Second Slab Third Slab m Top Slabs Grey, Very Stiff,SILTY CLAY, Traces of Sand -35 Stage 4: Installation of Fourth Slab and Excavation to Depth 13.2 m From G.S. 3. m Second Slab Yellow, Dense to Very Dense,MEDIUM GRADED SAND, Traces of Silt.. Intercalations with Silty Clay m 2.7 m Third Slab Fourth Slab 4.5 m 5. Fig. 3. The average subsurface condition at the site and the SPT results Fig. 4. A schematic presentation of the construction stages of Garage No CONSTRUCTION STAGES OF THE GARAGE The garage was constructed following the cut-and cover technique, utilizing the laterally supported diaphragm wall to retain earth and water pressures. The excavation depth within the garage reaches a maximum of 13.2 m from the ground surface. The diaphragm walls designed for the outer perimeter of the garage are 27. m in depth and.8 m in thickness and are fully reinforced from top to bottom. The hydro-phrase machines used for drilling panels of the diaphragm walls utilized bentonite slurry. Excavation of panels using these machines induces fewer ground movements compared with the grab bucket technique of excavation. After installing the diaphragm walls, the sequence of excavation and construction within the garage structure was carried out in a top-down sequence of work, i.e., casting slabs and then excavating underneath them to reach the level of the following slabs. With this technique, the slabs act as internal lateral supports to the main diaphragm walls during excavation to the finish level (Elevation +8. m). Fig. 4 shows a schematic presentation of the construction stages that was followed at the El- Tahreer garage (2). The construction of the garage passed through 3 main stages, namely: 1. Diaphragm wall installation. The vertical sides of the 13.2 m deep excavations were supported using fully reinforced cast-inplace concrete diaphragm walls (27. m deep and.8 m thick). 2. A grouted plug was injected between the diaphragm walls from the elevation (.8 m) to (-1.8 m) to control the groundwater level outside the deep excavation. 3. Excavation and strutting sequences to reach the finish level of the garage with simultaneous lowering of the groundwater levels within the deep excavation. Fig. 5 shows a schematic cross section of the constructed garage (2). 32 GEOTECHNICAL PERFORMANCE OF EMBEDDED CAST-IN-SITU DIAPHRAGM WALLS... rahman.indd :28:5

4 25/3 PAGES 3 38 Tahreer Street G.S. (21.2 m) Top Slab (2.7 m) 5 cm Thickness Second Slab (17.7 m) 3 cm Thickness Third Slab (15. m) 3 cm Thickness Fourth Slab (12.3 m) 3 cm Thickness Inclinometer Deep Settlment Piezometer PZ6 I(5) DS(1) 2,4 m 1,45 m PZ7 DS(2) I(3) I(4) 3, m 4,31 m 4,47 m I(6) El-Sadat Metro Station PZ2 PZ8 I(2) 8,7 m 1,25 m DS(3) PZ1 PZ3 PZ9 Raft Foundation (9.6 m) Excavation Level (+8. m) PZ5 PZ4 2,53 m PZ1 I(1) Fig. 6. Location of the inclinometers, settlement points, and piezometers Fig. 5. A schematic cross section of the constructed garage (2). 4. FIELD MONITORING Grouted Plug Elev. ( 1.8 m) Elev. ( 5.8 m) In order to monitor the performance of the supported deep excavation of the El-Tahreer (2) multistory underground garage project, an in-situ instrumentation program was implemented as shown in Fig. 6. It consisted of the following: 1. Six inclinometers (21. to 24. m long) for measuring lateral displacement. Five of them were installed adjacent to the diaphragm wall near the El-Sadat Metro station (21. to 23. m long). The sixth was installed inside the diaphragm wall (24. m long). 2. Six deep settlement points were also installed surrounding the garage to measure the settlement at different levels. Also, surface settlement measurements were conducted using precise level surveying equipment. 3. Ten piezometers, five of them located within the excavation and five installed outside the diaphragm wall near the El Sadat Metro Station. These instruments were used to monitor the groundwater levels inside and outside the excavation. There was no change in the elevations of the groundwater table around the station itself. Periodical readings were taken for each instrument during all stages of the construction from September 21 to April 23. The readings of these inclinometers are used in this study for comparing the field measurements and the numerical modeling results. 5. NUMERICAL MODELING Finite element analysis was utilized to model the excavation sequences and calculate the lateral movement of the diaphragm walls as well as the vertical displacement of the soil surface. PLAXIS TM, a commercially available finite element program, was used to conduct this C-type prediction. The two-dimensional mesh of the El-Tahreer garage (2) and the adjacent Metro station is shown in Fig. 7. Six-node triangular isoparametric elements were used with a total of 2331 nodes and 119 elements to model the soil strata and the grouting. Three-node beam elements were used to model the diaphragm wall, basement floors, and Metro station. Six node interface elements were utilized to model the interface between the soil and the diaphragm walls. 5.1 Soil Modeling The hardening soil model developed by Schanz & Vermeer (1997) was implemented in the analysis to model the soil strata. This model combines the merits of plasticity theory with the logic of the Duncan-Chang model (El-Nahhas & Morsy, 22). Some basic characteristics of the model are: GEOTECHNICAL PERFORMANCE OF EMBEDDED CAST-IN-SITU DIAPHRAGM WALLS rahman.indd :28:55

5 25/3 PAGES 3 38 q a = asymptote deviatoric stress (= q f /R f ); q f = failure deviatoric stress; R f = failure ratio ε 1 = vertical strain. The distinction between the primary deviatoric loading and unloading/reloading stress path: C = cohesion; E un = unloading and reloading of Young s modulus; E ref un = E ref i. (3) Failure behavior according to the Mohr-Coloumb failure criteria: The yield function of the hard soil model, Fig. 7. Finite element idealization of the diaphragm wall and the El-Sadat Metro station. Stress dependent stiffness according to the power law for the primary loading: γ P = strain hardening parameter; = plastic axial strain; = plastic volume strain. (4) (5) C = cohesion; E i = initial stress dependent on Young s modulus; E ref i = reference to Young s modulus corresponding to reference pressure P ref = 1kPa; m = power presents the stress depending (=.5 for sand); σ ' 3 = effective preliminary stresses; φ = angle of internal friction. The hyperbolic relationship between the strain and deviator stress: (1) In the case of sand, the plastic volumetric strain tends to be relatively small, and this leads to the approximation =.. The total strain variant, (6) (7) (8) (9) q = deviatoric stress in primary triaxial loading; (2) = elastic axial strain = plastic axial strain. 34 GEOTECHNICAL PERFORMANCE OF EMBEDDED CAST-IN-SITU DIAPHRAGM WALLS... rahman.indd :28:57

6 25/3 PAGES 3 38 The stress dilatancy theory: The essential feature of the stress dilatancy theory is that the material compacts for small stress ratios (φ m <φ cv ), while the dilatancy occurs for high stress ratios (φ m >φ cv ), φ m is the mobilized angle of the internal friction and φ cv is the critical state of the internal friction angle. The corresponding volumetric strain ratio to the mobilized dilatancy (ψ m ) is described by: γ P = rate of plastic shear strain; ψ m = mobilized dilatancy angle; ψ = failure of the dilatancy angle. (1) (11) (12) (13) Table (2) summarizes the parameters for the fill and the two main soil strata used for the finite element simulation. 5.2 Modeling the Supporting Elements The components of the laterally supported diaphragm walls were simulated using the capabilities of the Plaxis program. The properties of the reinforced diaphragm walls, horizontal reinforced slabs, raft foundation, and Metro station were simulated using beam elements. 6. RESULTS AND DISCUSSION The compiled field measurements of the three inclinometers at the locations I(3), I(4), and I(5) were selected to represent the measured lateral displacement of the soil. Inclinometer I(3) is about 2. m from the El-Sadat Metro station. Readings were recorded throughout the excavation and construction of the diaphragm walls and throughout the different construction stages that are shown in Fig. (4). Figs. 8 to 1 show a comparison between the measured and calculated lateral deflections of the diaphragm walls. The negative sign in the lateral deformation indicates displacement towards the excavation. The maximum lateral displacement during the construction of the diaphragm wall was predicted to be 8.2 mm, while the maximum measured lateral displacement did not exceed 1.85 mm. This can be attributed to the limited depth of the inclinometers used (21. to 23. m from the ground surface). It should be noted that the data reduction of the inclinometer readings is based on assuming the fixed lower end of the casing. The limit depth of the inclinometers used allowed the toe of the pipe to move laterally. Thus, the measured displacement by the inclinometers is considered to be a relative displacement between the upper point and the lower point of the inclinometer. A detailed comparison between the predicted and observed lateral displacement during the actual construction sequences at the inclinometers I(3), I(4), and I(5) are shown in Figs 8a, b, c, and d. Figure 8 reveals that: During the first and second stages of construction, the field and the computed lateral movement had similar trends and close values, especially those of inclinometer I(3). The measured values of the lateral movement of the inclinometers I(4) and I(5) were less than the computed ones by about 2% to 5% from 5. m up to 5. m Table (2) Parameters used for FEM simulation Stratum γ Identification Type dry γ wet k x K y C ref φ No. (kn/m 3 ) (kn/m 3 ) (m/day) (m/day) (kn/m 3 ) ( ) 1 Fill Drained Medium dense Sand Drained Very dense sand Drained Stratum No. Identification Type ψ ( ) ν υρ ( ) P ref (kn/m 2 ) (kn/m 2 ) (kn/m 2 ) (kn/m 2 ) 1 Fill Drained Medium dense Sand Drained E5 3 Very dense sand Drained E5 GEOTECHNICAL PERFORMANCE OF EMBEDDED CAST-IN-SITU DIAPHRAGM WALLS rahman.indd :29:

7 25/3 PAGES Stage Stage Stage Stage4-2 I(3) Average I(4) Average I(5) Average -2 I(3) Average I(4) Average I(5) Average -2 I(3) Average I(4) Average I(5) Average Modified -2 I(3) Average I(4) Average I(5) Average Modified Fig. 8. and measured lateral movements for different stages in the soil 2. from the El- Sadat Metro station 5 Stage1-2 I(6) Average 5 Stage4-2 I(6) Average Modified Fig 9 and measured lateral movements for the first and fourth stages in the diaphragm wall of garage (2). below the surface. The deeper measured values approximately coincide with the predicted values. During the third and fourth stages of construction, the field and the computed lateral movement were completely different. This is due to the limited depth of the inclinometers used. The limited depth of the inclinometers used allowed the toe of the pipe to move laterally. Thus, the measured displacements by the inclinometers are considered to be the relative displacement between the different points and the lower point of the inclinometer. The modified predicted, deformation was obtained by subtracting the calculated movement at the end of the inclinometer casing from all the upper calculated values. The measured lateral displacements subsequently better matched the modified calculated deformation of these two stages. Fig. 9 shows a comparison between the predicted and observed lateral displacements during the first and fourth construction stages as an example of the obtained values. Also an evaluation was carried out for the calculated and field-measured lateral deformation for inclinometer I(1), which represents the lateral deformation of the soil near the garage diaphragm walls, but is relatively far from the Metro tunnel. This evaluation is shown in Fig. 1, which represents the first and fourth stages of construction as an example of the results obtained. 36 GEOTECHNICAL PERFORMANCE OF EMBEDDED CAST-IN-SITU DIAPHRAGM WALLS... rahman.indd :29:3

8 25/3 PAGES Stage Stage4 Settlement (mm) Stage4-2 I(1) Average -2 I(1) Average Modified Settlement DS(2) Average Thompson,1991 Clough and O'Rourke,199 Fig. 1. and measured lateral movements for the first and fourth stages in the soil, away from the El-Sadat Metro Station. -2 The predicted settlement at point DS(2) and the measured surface and subsurface settlements are shown in Fig. 11. The measured and predicted surface and subsurface settlement profiles do not match, but the field-measured settlements or vertical movements were generally less than the predicted values. This result agrees with Abdel-Rahman, A. & El-Said, S. M. (22). The predicted settlement was 1%, at some points, more than the field-measured values. The measured surface settlement was also less than that proposed by Clough and O Rourke (199), but matched Thompson (1991). 7. CONCLUSIONS This paper discusses the performance of the diaphragm wall of an underground multi-story garage constructed in downtown Cairo based on the geotechnical instrumentation. The field measurements are compared with the predicted soil movements calculated using a two-dimensional finite element analysis. Although the measured soil movements were less than the predicted values, a reasonable concordance is achieved. This can be attributed to the difference between the physical and analytical modeling. Inclinometers are good tools to measure and investigate the lateral deformation of soil due to earthwork. However the casing utilized Fig. 11. and measured vertical movements for the fourth stage in the soil, 2. m from the El-Sadat Metro Station. must be deep enough to obtain reliable results. For this reason inclinometers inside the diaphragm walls have to be installed at least to the end of the walls or even deeper. It is recommended to install the inclinometers in a similarly important project to develop feedback analysis for performing more accurate analytical models of soil parameters for the lateral deformation of diaphragm walls or for better design criteria. ACKNOWLEDGEMENT The inclinometer readings described in this paper were performed by the Soil Mechanics and Foundation Unit of Ain Shams University. The author heavily participated in this in-situ monitoring program. The author wishes to thank the Arab Contractor Company, the general project contractor, for their sincere cooperation while conducting the field monitoring program and their permission to use the results obtained in this paper. GEOTECHNICAL PERFORMANCE OF EMBEDDED CAST-IN-SITU DIAPHRAGM WALLS rahman.indd :29:7

9 25/3 PAGES 3 38 REFERENCES Abdel-Rahman, A; El-Said S. M., (22) Building subsidence associated with cut-and-cover excavation in alluvial soils. Ain Shams Uni., Faculty of Eng. Journal, Vol. 37, No.4, pp Boone, S.J.; Westland, J.; and Nusink, R., (1999) Comparative evaluation of the building re-sponses to an adjacent braced excavation, Canadian Geotechnical Journal, Vol. 36, No.2, pp Burland, J.B., and Hancock, R.J.R. (1977) Underground car park at the House of Commons, London: Geotechnical aspects, Structural Engineering, 55(2), 87. Clough, R.W., and O Rourke, T.D. (199) Constructioninduced movements of in-situ walls, Proceedings, ASCE Conference on Design and Performance of Earth-Retaining Structures, Geotechnical Special Publication, Cornell University, Ithaca, N.Y., 25, Di Biagio, E., ed. (1991) Third International Symposium on Field in Geome-chanics, Oslo, Norway. Rotterdam: A.A. Balkema. El-Nahhas, F. (1987), Prediction of ground movement adjacent to a supported deep excava-tion, Proc. Of 9th African Regional Conference on Soil Mechanics and Foundation Engi-neering, Nigeria, Vol. 1, El-Nahhas, F. (1992), Construction monitoring of urban tunnels and subway stations, Tun-neling and Underground Space Technology, Pergamon Press,Vol. 7, No. 4, El-Nahhas, F. (1999), Soft ground tunneling in Egypt: geotechnical challenges and expecta-tions, Pergamon Press, Vol. 14, No. 3, El-Nahhas, F.; Eisenstein, Z.; and Shalaby, A. (1989) Behavior of diaphragm walls during con-struction of the Cairo Metro Proc. Twelfth Int. Conf. on Soil Mechanical and Foundation Engineering, Rio de Janeiro, Brazil, El.Nahhas, F.; Morsy, M. M.(22), Comparison of a measured and computed diaphragm re-taining wall in Egypt, Proc. of Int. Symposium on Numerical Models in Geomechanics, Rome, Italy, April 22, pp El.Nahhas, F. (23), Geotechnical Aspects of Controlling Groundwater Levels in Urban Ar-eas, Proc. of Tenth Int. Colloquium on Structural and Geotechnical Engineering, April 23, Keynote lecture 4/1-4/21. Finno. R.J., Atmatzidis, D.K., and Perkins S.B. (1989), Observed performance of a deep exca-vation in clay, Journal of Geotechnical Engineering, ASCE, 115(8), Fansmire, W.h., Russell, H.A., Rawnsley, and Abbott, E.L. (1989), Field performance of structural slurry wall, Journal of Geotechnical Engineering, ASCE, 115(2), Kovari, K. ed. (1984), First International Symposium on Field in Geomechan-ics, Zurich. Rotterdam: A. A. Balkema. Sakurai, S., ed. (1987), Second International Symposium on Field in Geome-chanics, Kobe, Japan. Rotterdam: A. A. Balkema. Schanz, T., and Bonnies, P.G. (1997), Verification of a soil model with predicted behaviour of sheet pile wall, Computer Methods and Advances in Geomechanics, Schanz, T., and Vermeer, P.A. (1997), On the stiffness of sand, Geotechnique Symposium on pre-failure deformation behaviour of geomaterials, London. Sharma, J.S.; Hefny, A.M; Zhao, J.; and Chan, C.C. (21) Effect of large excavation on de-formation of adjacent MRT tunnels, Tunneling and Underground Space Technology, Pergamon Press, Vol. 16,No. 2, Thompson, P. (1991), A review of retaining wall behavior in over-consolidated clay during early stages of construction, Mphil Thesis, Univ. of London, England. Ulrich, E. J. (1989) Internally braced cuts in over-consolidated soils, Journal of Geotechnical Engineering, ASCE, 115(4) GEOTECHNICAL PERFORMANCE OF EMBEDDED CAST-IN-SITU DIAPHRAGM WALLS... rahman.indd :29:9