Modeling and monitoring of an excavation support using CSM António Capelo 1 ; A. Gomes Correia 2, Luís F. Ramos 3, Alexandre Pinto 4 and Rui Tomásio 5 1 MSc, Casais Engenharia e Construção, S. A., Mire de Tibães, Portugal, antonio.c.capelo@gmail.com 2 Professor, Department of Civil Engineering/C-TAC; Campus de Azurém, Guimarães, Portugal; agc@civil.uminho.pt 3 Assistant Professor, Department of Civil Engineering/ISISE, Campus de Azurém, Guimarães, Portugal; lramos@civil.uminho.pt 4 MSc, Head of Office, JetSJ Geotecnia, Lisboa, Portugal, apinto@jetsj.pt 5 MSc, JetSJ Geotecnia, Lisboa, Portugal, rtomasio@jetsj.pt ABSTRACT: Cutter Soil Mixing (CSM) is a soil improvement technique with an accentuated growth in the last years satisfying geotechnical and environmental construction s needs and in many cases with economical advantages. This technology uses the soil in situ to create a composite material with improved mechanical and hydraulic properties. This technology is recently being applied in Portugal for improving soil foundations and for retaining structures. This work presents the application of an excavation support for a shaft with a mean depth of 18 m and an internal diameter of 15 m. This support was materialized with panels of 2.4 0.8 m 2 reinforced with IPE300 steel profiles placed tangentially to the internal face of the CSM panel. A numerical analysis was carried out using axy-symmetric and tridimensional modeling and the results obtained were compared with the design values, as well as with the monitoring results during construction, showing a very good performance. Furthermore, it was also evaluated the safety factor of the panels and of the steel profiles. INTRODUCTION The soil treatment techniques are becoming increasingly essential to overcome the today s construction needs reusing in situ soil. Deep Mixing is an in situ soil treatment method that makes use of a technology in which the soil is mechanically mixed with other materials, mainly binders. The composite material will have improved benefits in terms of resistance, compressibility and permeability (Larsson, 2003 and Bruce, 2000). One of the variants of Deep Mixing is the Cutter Soil Mixing (CSM) technique which originates panel elements with accurate geometry, verticality and direction. Furthermore to these advantages, this technique causes low disturbance on the soil and nearby structures (related to the low pressure binder injection) making their use appropriated in urban areas. This work describes a case study using Cutter Soil Mixing applied as earth retaining structure for the construction of two shafts at the River Lima banks, at Ponte de Lima, Portugal. Those shafts with about 18 m depth and 15 m diameter were built in order to allow the installation of a water supply 2400 mm reinforced concrete pipe, under the river using micro tunneling technology. The geological conditions were very heterogeneous, with ground water table located near the Page 1
surface. The excavation works intersected, from the surface, heterogeneous landfills, sandy and gravel soils and weathered schist (W5 and W4). DESCRIPTION OF THE CONSTRUCTION WORKS In order to allow the excavation works, soil - cement panels with an overall depth of about 24 m and a crosss section of 2,4 x 0,8 m2, including 0,30 m of overlapping, were built using the CSM technology. The panels were reinforced with vertical IPE300 steel profiles, in order to resist to all the earth and water pressures. The profiles were braced, at the top, by a reinforced concrete capping beam and by three levels of steel ring beams, in depth. The soil - cement panels were design in order to transmit horizontally the earth and water pressures to the vertical profiles. As already stated, the shaft has an average depth of 18 m and an internal diameter of 15 m (Fig. 1). Figure 1 Layout and section of the shaft (JetSJ, 2009) The execution works of the excavation support were the following: - Top reinforced concrete beam with 1 m depth made of C20/25 Class concrete; - Execution of 2.4 0.8 m 2 CSM panels until 24 m depth, reinforced with IPE300 steel columns tangentially to the internal face of the CSM panel; - Excavation and connection of the circular distribution beam (HEB200) with the reinforced panel each 3.5 m depth; - Bottom reinforced concrete slab with 0.35 m thickness, made of C25/30 Class concrete. Figure 2 shows a general view of the shaft excavation support and a detail of the connection of the circular distribution beam (HEB200) with the reinforced CSM panel. Page 2
a) b) Figure 2 a) Retaining structure after soil excavation; b) Connection between circular distribution beam (HEB200) and CSM reinforced panels MODELING AND DESIGN Simplified Approach A first simplified modeling work was carried out using Robot Structural Analysis 2010, assuming an axy-symmetric problem in a homogeneous soil with a constant groundwater level. This structural analysis was considered reasonable since the deformation of this type of structure is small and consequently compatible with the use of linear models. However, the panel-type elements of CSM reinforced with steel columns needs the consideration of two materials (fig. 3): (1) soil cement and (2) composite material (soil cement with IPE300). Figure 3 Cross-section of CSM reinforced panel-type The constitutive material model chosen for the slab, top concrete beam, CSM panel and CSM composite material was the elastic linear model. The main input material parameters are summed up in Table 1. In this structural model, the bending moments are very important and consequently it is essential that the composite material reproduced correctly the bending behavior of the original section. Using a simple homogenization process, the deformation modulus and the unit weight of the composite material were calculated and the obtained values are presented in Table 1. Page 3
Table 1. Main material properties of the input parameters Material Deformation Modulus (GPa) Weigth (kn/m 3 ) C25/30 30.0 24.53 CSM 1.0 16.03 Composite material 1.5 16.55 In which concern the boundary conditions, the circular distribution beams were simulated by elastic supports, with a stiffness of 4000 kn/m. Furthermore it was assumed a fixed support at the bottom of the panel. Since the used structural software didn t allow the soil modeling as a continuous medium, the soil-wall interaction problem was solved outside the program and simulated by a set of independent horizontal springs characterized by bilinear elasticplastic pressure-displacement relation (Clayton et al., 1993). These springs below the excavation were spaced 1 m of each other in depth. The limit earth pressures were calculated using soil parameter values presented in Table 2. Furthermore, the pressures caused by the overburden (4.3 kn/m 2 ) and groundwater level were also calculated. Table 2. Estimated soil parameters (JetSJ, 2009) Layers φ c (kn/m 2 ) γ (kn/m 3 ) Fill materials and vegetable soil 25 0 19 Sandy material and gravels 30 0 20 Very weathered schist (W5) 30 20 20 Weathered schist (W4) 35 40 21 For the structural analysis, load combinations for the Ultimate Limit State and for the Serviceability Limit State were defined. The obtained results in terms of deformations and stresses are illustrated in Figure 4. It is observed that deformation increased with depth and with significantly increasing after the last distribution beam. A higher deformed zone can be observed between the last distribution beam and the wall s bottom (fig. 4a). The maximum horizontal deformation attained was equal to 18 mm around a depth of 22 m. For the Ultimate Limit State, a maximum compressive stress of 1 MPa was observed on the soil-cement material (fig. 4b), which represents around 25% of the material s compressive strength. Furthermore, it was possible to notice a maximum compressive stress of 1.13 MPa in the last level of the circular steel beam (fig. 4c). Since this HEB beam is in a compression state, it was important to check if the Strength Limit State was verified. Therefore, the compressive force was calculated, and the buckling resistance was then verified ( N / =1. 3). Rd N sd Page 4
a) b) c) Figure 4 CSM reinforced panel modeling results: a) lateral displacement for the Service Limit State, b) stresses for the Strength Limit State, c) maximum compressive stress at the last level of the circular steel beam Enhanced Approach For a more complete and detailed analysis a 3D model was also used (fig. 5a). In this case the CSM Panel could be analyzed individually, as a unit cell (fig. 5b). The composite material was replaced by panels of a material defined with the soilcement parameters and by IPE300 steel columns. The bilinear elastic supports were also replaced by the HEB200 circular beams. Figure 5c shows a cross-section of CSM reinforced panel, allowing the evaluation of the maximum compressive stress. This value (1.08 MPa) was less than the compressive strength of the material (4 MPa). The induced shear stresses were also calculated, being lower than the shear strength of the material. Comparing the simplified axi-simetric model with 3D model results it was concluded that the deformed shape of CSM reinforced structure is very similar in both models (fig. 4a and fig. 6), as well as the level of safety against the ultimate limit states. Therefore, despite the simplifications of the axi-simetric model using a composite material and discrete supports, it seems to be a practical numerical tool to be possible to apply in similar case studies at design level. Page 5
a) b) c) Figure 5 3D modeling: a) CSM reinforced retained structure, b) cell unit, c) maximum compressive stress at the cross-section of CSM reinforced panel Figure 6 Deformed shape of CSM reinforced structure Page 6
COMPARISON OF NUMERICAL AND EXPERIMENTAL RESULTS A monitoring plan was installed to observe the structure performance during excavation. This was materialized by three inclinometers in the soil nearby the CSM wall and by eight on the first and second circular beams levels (fig. 7). Figure 7 Monitoring system The topographic results show a large scatter that can mainly be attributed to external actions induced by machinery operations. Consequently, these results were not considered in the analysis. The inclinometer results confirm an overall good performance of the structure during excavation, showing relative reduced displacements at the level of the circular beam (supports). Furthermore, it was observed a maximum horizontal displacement value of 10.76 mm at 17 m depth, which is reasonable close to the modeling results, particularly in which concerns the depth (difference of 20%). In fact the excavation works and measurements were carried out during the summer time, where the groundwater level was lower, as assumed in the modeling, inducing fewer displacements. This result also reveals a small horizontal strain in relation to the excavation depth (0.06 %), supporting the linear elastic modeling approach. It should be mentioned that the results of the three inclinometers show a certain scatter between them. This can be explained by the fact that structure behavior was not fully axy-symmetric. The reasons for the scatter results can be justified by thickness variations of soil layers around the shaft, as well as the lack of symmetry during the soil excavation. A more detailed geotechnical site investigation and the full record of the excavation process could allow a better modeling. CONCLUSIONS AND FUTURE DEVELOPMENTS In this study the Cutter Soil Mixing technology, reinforced with vertical IPE300 steel profiles, was applied successfully in the construction of two circular shafts with Page 7
about 18 m depth and 15 m diameter at the river banks, in a heterogeneous formation with ground water table at surface. An axi-symetric and tridimensional modeling were carried out, using linear elastic models for all the materials. Based in the monitoring results it was found that this modeling work performed at design stage was conservative. This was intentionally done, mainly because of the lack of geotechnical and hydrogeological information. The monitoring results also confirm that the maximum horizontal displacements induced by excavation were in the domain of small strains (0.06 %), showing that the assumption of linear elastic material behavior was reasonable. The modeling results presented in this paper, though relative to a case study, shows that the use of tridimensional modeling for the Ultimate Limit State evaluation of a CSM reinforced panel is more realistic, since it allows the separate study of the unit cell and the steel elements. However, the axi-symetric model could be a quick and practical solution at an early design stage giving conservative results for the Service Limit State evaluation. Of course these conclusions need to be validated on other real case studies. For future works is recommended a more detailed site investigation in order to obtain the required geotechnical parameters for design, both for the Ultimate Limit State (strength parameters) and for the Service Limit State (deformability parameters). Furthermore, for the modeling work it is very important to register the excavation sequences associated with the monitoring survey, as well as the water table nearby the excavation. ACKNOWLEDGMENTS The authors thank the Owner of the described case history the permission for presentation of the present paper. REFERENCES Borel, S. (2007). Soil mixing innovations: geomix, springsol and trenchmix. In: BGA CFMS Conference, London. Bruce, D.A. (2000) An introduction to the deep soil mixing methods as used in geotechnical applications. Federal Highway Administration, Georgetown. Clayton, C.R.I., Milititsky, J. & Woods, R.I. (1993). Earth pressure and earthretaining structures. Chapman & Hall. JetSJ (2009). Excavation and CSM retaining walls for shafts. Design. (in portuguese). Larsson, S. (2003). Mixing processes for ground improvement by deep mixing. Department of Civil and Environmental Engineering, Royal Institute of Technology in Stockholm. Page 8