HPFRC plates for ground anchors

Transcription

1 HPFRC plates for ground anchors M. di Prisco, D. Dozio, A. Galli, S. Lapolla Department of Structural Engineering, Politecnico di Milano M. Alba Surveying Department (DIIAR), Politecnico di Milano ABSTRACT: In order to stabilize a ground slope, special plates of reduced sizes made of High Performance concrete reinforced with straight steel fibres were designed and built. The plates are reinforced also with special high bond steel bars with special steel threaded bushes welded at the ends to guarantee their tensile action on overall the slab size. Their weight is limited in order to assure the transportability by helicopter everywhere in mountain regions. After the experimental characterization of the material aimed to identify a constitutive relationship, a limit design approach was carried out. Ten plates were placed in situ and two of them were instrumented in order to follow the real stress state inside of the anchor plates: the main results are here described. 1 INTRODUCTION In the Pre-Alps region there are wide slopes originated in the glacial era, made of heterogeneous material characterized by a high geological risk. The mitigation of the risk associated to this kind of slopes is not an easy engineering problem for several geotechnical and structural reasons. A new precast retaining structure was designed to take advantage of High Performance Fibre Reinforced Concrete in order to develop a faster and more effective procedure of intervention. The structure geometry was designed to reduce as much as possible the own weight in order to allow the helicopter transport (maximum weight < 8-10 kn), but at the same time the high performances of the material guarantee huge surface hardness and local toughness and a high resistance to local pressures, whose distribution on the structure are difficult to predict due to the roughness and the heterogeneity of the slope. A plate (0.24 x 0.8 x 0.8 m), reinforced by means of four steel B450C bars aligned in the two directions and located in the intrados face, is anchored to the slope by a prestressed cable (7 strands 0.6 each made of 7 wires; diameter 15.2 mm,; A p = 973 mm 2 ; f ptk = 1860 MPa), 14.5 m long (Fig. 1) passing in a central hole suitably designed to accommodate the cable slope-face end. The bar ends are welded to threaded bushes able to guarantee a direct anchorage to the concrete at the extremities and allowing the reinforcement to be active in the whole bar length. Steel fibre reinforcement was used, without the introduction of any further conventional reinforcement. Three more plate elements were cast in order to carry out experimental tests in the laboratory with well known boundary conditions. The paper describes the material characterization and the handling phases of ten plates placed in an environmental laboratory of the Politecnico di Milano located in Caslino d Erba close to Como (di Prisco et al. 2006). A careful monitoring of three plates (the two here described anchored by means of traditional prestressed steel cables and the last one anchored by means of a GFRP anchor) was planned to detect the structural behaviour along a time period of at least five years. 2 MATERIAL CHARACTERIZATION 2.1 Mix design The composite was selected by comparing different solutions starting from the aggregates generally used by the precast producer and limiting their maximum aggregate size to 2 mm. The material presents typical proportioning of a self compacting concrete. The mix design of the HPFRCC material is specified in Table 1. Steel fibres are high carbon straight fibres 13 mm long, with a 0.16 mm diameter; their content is equal to 100 kg/m 3. Table 1. Material mix design C 52,5 I Slag Sand 0/2 Fiber Additive [l/m 3 ] Water [l/m 3 ]

2 A 200 SECTION A-A B B A SECTION B-B Figure 1. Geometry of anchor plate (measures in [mm]). Figure 2. Plate casting (a) (b) (d) Figure 3: Fresh behaviour tests: (a) slump flow; (b) V funnel; L-shape box; (d) J-ring P/2 P/2 LVDTs (CTOD) clip gauge (CMOD) P/2 P/ N [MP a ] 8 4 T 1 T 2 T3 T 4 Average CTOD m [m m ] Figure 4. Notched tests according to UNI 11188: (a) test set-up and (b) nominal stress vs. Crack Tip Opening Displacement.

3 2.2 Fresh concrete tests The mix design was used to cast 13 reinforced plates without vibration and the specimens, cubes and beams, useful to identify both uniaxial compression and uniaxial tension constitutive relationships (di Prisco et al., 2008). The fresh behaviour was controlled by means of several tests (Fig.3): (a) slump flow; (b) V funnel; L-shape box and (d) J-ring tests. The results (Table 2) confirmed a SCC consistency (Fig. 2). Table 2. Fresh behaviour characterization. The indication between brackets are referred to the test type: (a) slump flow; (b) V-funnel; L-box and (d) J-ring test. D m (a) t m (a) t m (b) h 1 h 2 t 200 t 400 D m (d) [mm] [s] [s] [mm] [mm] [s] [s] [mm] Hardened concrete tests An average cubic compressive strength of 140 MPa and an elastic modulus close to 40 GPa characterized the material in the preliminary qualification. Cubic compression tests and four point bending tests on notched specimens gave first cracking and residual strengths (Fig.4a, Table 3) significantly smaller than those obtained in the preliminary tests (di Prisco et al., 2008). The residual strengths in uniaxial tension were measured by using a four point bending tests on notched specimen according to UNI Italian recommendations. The geometry of the notched specimen and its set-up are described in Figure 4a; the nominal stress σ N vs. crack tip opening displacement (CTOD) are shown in Figure 4b. The recommendations suggest to compute three strengths in order to classify the bending behaviour of the material: the first crack strength (f IF ) that is correlated to the tensile strength of the material, and the residual strengths respectively in the range (f eq0-0.6 ) and and 3.0 mm (f eq0.6-3 ) corresponding to serviceability and ultimate limit states. In the table also standard deviations are indicated. Table 3. Hardened behaviour characterization Test n Age [days] 4 28 R c28,m (8.78) f If,m (1.22) 3 STRUCTURAL DESIGN f eq(0-0,6),m (1.48) f eq(0,6-3),m 9.76 (1.78) The design of the retaining structure is mainly oriented to emphasize the SFRC structural behaviour at the serviceability limit state, because the prestressed anchors are chosen as critical ring in the construction design. In the EuroCode 7 framework, the retaining structure designed belongs to the third category, because it can be regarded as a new structure due to its conceptual design as well as to the material used. The active confinement action applied to the ground slope by means of prestressed anchor strands, allows us to consider the limit state analysis in the case C, that means to regard the slope instability associated to total or partial collapse of the tendons as the critical limit state. The safety check was performed by means of predictive calculations and by adopting the observational method. The research here described is aimed to investigate the structural behaviour of the reinforced SFRC plate during placing. It is conceived to react to a prestressing action of about 1150 kn originating by the stretching of the cable made of seven 7W strands. The plate design follows the overestimation rule which is forced by the impossibility to predict the real boundary conditions due to the roughness of the morenic slope face and the need to prevent any local cracking due to concentrated loads. About the computation of the ultimate limit state, limit analysis was used associated to a kinematic mechanism of 2 yield lines aligned to the symmetric axes of the plate and justified by the unilateral constraint of the ground reaction with the consequent rotation around the diagonal of each forth square of the plate. The specific bending moment was computed adding the contribution of each steel bar smeared along the half of each yield line to the contribution of the residual tensile strength associated to a crack opening of 2 mm, which corresponds to a steel strain of 1% when the distance of the bar axis from the top fibre (equal to 200 mm) is considered as characteristic length. 4 ON SITU PLACING PHASE 4.1 Settlement description The plates were placed in the established location on the slope as briefly shown in Figure 5. No mortar layer was introduced between the plate and the slope face to accelerate the placing procedure. Due to this choice, the rigid motion of the anchor plates was significant as detected by laser scanning technique and shown in Figure 7 for the two instrumented plates. Also the displacements of the four corners of these plates were measured and the relative displacements associated to each load step were computed. According to these results, the relative displacement of the barycentre of each plate, assumed as completely rigid, was also identified in order to depurate this value by the total sliding measured between the jacket reference and the stretched strands. The results are proposed for the top instrumented anchor plate in Figure 8.

4 Figure 5. Anchor plate location on slope. Figure 6. Placing of anchor plates on slope. Figure 7a. Laser scanning displacements of top plate. Figure 7b. Laser scanning displacements of bottom plate. 4.2 Steel anchor stretching The force activated in the anchor cables was measured by means of suitable loading cells placed in order to measure the force in only one strand for each cable. The total force was measured by the pressure sensor of the loading system. Moreover, the relative displacement between the jacket reference and the stretched strand read at the end of each step for all the strands allowed us to deduce the real stiffness of the anchor cables. In fact, the real stiffness of each cable can be computed by subtracting from the total relative displacement the rigid displacement of the plate barycentre that takes into account also the local punching of the plate in the slope face (Figs. 7a,b). The relative displacement computed can be finally compared with the theoretical prediction based on the different assumptions on the cable length. In Figure 8 the load corresponds to that measured by the loading cell and it is referred to only one strand. The real stiffness of the strand is not known because it depends by the sliding between the strand and the mortar and also by the sliding between the mortar and the ground. The former usually is small if the injection is well done, while the latter can be

5 Load [kn] 60 Plate barycentre rigid motion Calculated strand elongation Measured strand wire/jacket sliding 40 Theoretical strand elongation (cable length assumed = 10 m) Theoretical free length strand elongation Displacement [mm] Figure 8. Load vs displacement of top anchor plate during placing (normal component with respect to slope face). 100 Top plate Nup 80 Load [kn] Nup εcls [x 10-3 ] Figure 9a. Top plate placing: load vs deformation. 160 Bottom plate 120 Load [kn] Sup εcls [x 10-3 ] Sup Figure 9b. Bottom plate placing: load vs deformation.

6 significant due to the relatively high heterogeneity of the ground. In the comparison shown in Figure 8, a total active length equal to 10 m, that corresponds to the free length (5.5m) plus the half of the foundation length (4.5m), is first proposed to estimate the total displacement. Afterwards, a second theoretical curve computed with only the free length is also shown to fit the slope of the measured curve after the application of a significant load. The comparison between the experimental curve and the theoretical ones justifies the reliability of the anchor system. It is worth to note how the important rigid displacement of the plate causes some local plasticity phenomena on the slope face, but this occurrence does not compromise the anchor system efficiency. 4.3 Plate deformation The two instrumented plates were equipped by means of 6 suitable vibrating wire sensors able to measure the strains on a 200 mm gauge length: four sensors were located at 30 mm from the intrados (one for each side; Fig.9a,b) and two were located at 30 mm from the extrados to measure the compressive strains associated to the bending along the direction at right angle with respect to the yield line trace assumed. The values shown (Figs. 9a,b) highlight as the maximum relative displacement can justifies cracking, even if the crack opening remains always admissible according to serviceability limit states. It is also evident how large can be the scattering between the measures performed on the four sides of the plate, thus justifying the rather complex and unpredictable set of boundary conditions to take into account at the design level (Fig.10). Figure 10. Anchor plate during loading phase. 5 CONCLUDING REMARKS The paper is focused on the design of a light retaining structure able to insert anchor cables in a ground slope to increase its safety coefficient by preventing its instability. The main assets of this design solution are the limited weight, the reduced cost and the high speed in its placing. The limited weight allows the designer to use it in quite inaccessible sites like mountain regions by helicopters transport on situ. The reduced cost is related to the optimized volume and to a substantial convenience when compared to a steel solution: in this case the cost ratio is larger than five times and the environmental impact is not comparable. The high speed of its placing is guaranteed by the lack of any mortar layer between the slope face and the plate. To this aim the use of HPFRC is particularly promising because it allows to dispose of a very high local toughness and hardness, which are essential characteristics in such structures where boundary conditions and applied constraints cannot be easily a priori known. The lack of complex reinforcement detailing allows a complete industrialized production and the structure obtained is versatile, because it can be combined with other similar structures to give rise to modular and quite articulate retaining structure. A system of anchor plates can also be designed in combination with conventional steel wire net in order to solve at the same time the surface and the deep landslide. The surveying of these structures activated via GPRS in the environmental laboratory located in Caslino d Erba will give precious information on durability of this new promising retaining structure solution. 6 REFERENCES CNR-DT 204, Instructions for design, execution and control of fibre reinforced concrete structures, Italian Standards. Naaman, A.E. & Reinhardt, H.W., Eds High Performance Fiber Reinforced Cement Composites (HPFRCC4), PRO 30, Rilem publication S.A.R.L.. Reinhardt, H.W. & Naaman, A.E., Eds High Performance Fibre Reinforced Cement Composites (HPFRCC5), PRO 53, Rilem publication S.A.R.L.. di Prisco, C., di Prisco, M., Mauri M.& Scola, M A New Design for Stabilizing Ground Slopes, Proc. of the 2nd fib Congress, Napoli (Italy), June 5-8, ID 4-1 on CD- ROM. di Prisco, M., Lamperti, M., Lapolla, S. & Khurana, R.S HPFRCC thin plates for precast roofing, Proc. of Second International Symposium on Ultra High Performance Concrete, Kassel Germany (in printing). 7 ACKNOWLEDGEMENTS The authors thank Prof. Alberto Giussani and Prof. Marco Scaioni for their technical and financial support in the surveying activity and Dr. Fabio Roncoroni for their precious support in the on situ placing phases and Prof. Claudio di Prisco for the enlightening discussion in the design activity. The authors would like to thank also Fumagalli Prefabbricati, Basf and Halfen Companies for their technical and financial support of the overall research.