Remerciez-le!

Remerciez @Admin pour avoir partagé cet document gratuitement, de la manière la plus simple, en partageant sur les réseaux sociaux.

A comparison of jacked, driven and bored piles in sand Comparaison des pieux fo

A comparison of jacked, driven and bored piles in sand Comparaison des pieux fonçés, battus et forés dans les sables A.D. Deeks, D.J. White & M.D. Bolton University of Cambridge, UK ABSTRACT Pile jacking technology allows displacement piles to be installed without noise and vibration. The ‘press-in’ method of pile jacking uses previously-installed piles for reaction, so the piles must be installed at close centres. Axial load tests have been conducted to in- vestigate whether existing design guidance based on driven and bored pile behaviour can be applied to closely-spaced jacked piles. The observed axial response was notably stiff, and failure was reached at a load equal to the installation force. This high stiffness is attributed to pre-loading of the pile base during installation and the presence of residual base load. Load transfer back-analysis was used to establish simple parameters for the modelling of single pile stiffness. These parameters predicted the pile group response well using elastic superposition to account for interaction. This high stiffness could lead to more efficient design if jacked piles are used. RÉSUMÉ La mise en œuvre des pieux par fonçage permet d’installer des pieux sans immissions sonores et sans vibrations. Le fonçage des pieux utilise la réaction des pieux installés, de sorte que l’entre-axe des pieux doit être rapproché. Des essais de chargement axiaux sur des pieux et sur des groupes de pieux foncés ont étés effectués afin de déterminer si les codes de dimensionnement pour les pieux battus et forés peuvent être appliqués aux pieux foncés proches les uns des autres. Une grande rigidité des pieux sous charge axiale a été mesu- rée. Cette rigidité est attribuée au préchargement de la base du pieu lors de son installation et à la présence de charges résiduelles à la base du pieu. Des analyses du transfert des charges ont étés effectuées pour identifier des paramètres pour la prédiction de la rigidité des pieux isolés. Ces paramètres ont permis une bonne évaluation du comportement du groupe de pieux, moyennant modifications par superposition élastique pour tenir compte des interactions entre les pieux. 1 INTRODUCTION The strength and stiffness of a pile foundation is influenced by the installation method. Modern techniques of pile construction have led to improved foundation performance. To benefit from this improved performance, design methods must be modified to account for the influence of construction method on strength and stiffness. If a foundation can be constructed from a smaller number of stiffer piles, economies of cost, construction time and environmental impact through reduced material use can result. This paper describes an investigation into the response of jacked displacement piles in sand. One pile jacking technique is the ‘press-in’ method, in which reaction force for the jacking machine is obtained from previously installed piles. The ‘press- in’ piling machine shown in Figure 1 installs tubular piles of di- ameter 1000-1200 mm with a jacking force of up to 3 MN. Pile jacking technology allows pre-formed displacement piles to be installed without the environmental impact of dy- namic methods. The use of static jacking force applied using hydraulic rams avoids the noise and ground vibration associated with conventional dynamic methods. Previous research has demonstrated that pile jacking reduces ground-borne vibrations by an order of magnitude compared to traditional percussive and vibro-hammer installation techniques (Rockhill et al 2003). Pile jacking machines with capacities of up to 4 MN are cur- rently in operation (White et al 2002, Lehane et al 2003). Since ‘press-in’ piling machines ‘walk’ along the pile wall as construction advances, the piles must be installed at a nominal centre-to-centre spacing of one diameter. This geometry con- flicts with conventional design guidance, which advises a minimum pile spacing of 2 or 3 diameters (BS8004, 1986; GEO, 1996). This advice aims to eliminate interaction between the piles, to avoid reduced pile stiffness or strength. Existing design methods may be inadequate for predicting the axial re- sponse of piled foundations installed using pile jacking technol- ogy for two reasons: 1. The axial stiffness of the pile may differ from con- ventional piles due to the jacked installation. 2. Current design methods for pile groups have not been tested against piles installed at spacing ratios as low as unity. Field load tests have been conducted to examine these two uncertainties. A series of maintained load tests (MLT) on jacked-in, open-ended tubular piles are reported. These piles were either alone, in a short wall, or in a group of up to 12 piles. Back-analysis of the load-settlement response is carried out us- ing a load transfer approach. Figure 1. A ‘press-in’ piling machine for installing large tubular piles 2 METHODOLOGY 2.1 Ground conditions This series of pile load tests was conducted during summer 2003 at the Takasu test site located in Kochi, Japan. The ground conditions comprise made ground overlying layers of silt, silty sand and sand (Fig. 2). Prior to installation of the test piles the made ground was excavated and replaced by sand. Figure 2. Ground conditions at test site Figure 3. Arrangement of test piles 2.2 Test piles The test piles were uninstrumented open-ended steel tubes with an external diameter, D, of 4 inches (101.6 mm) and a wall thickness of 5.7 mm. Two lengths, L, of test pile were used, with embedded depths of 5.85 and 6.85 m. A total of 43 piles were tested, either alone, in short walls of 2 or 3 piles, or in cir- cular groups of 6 or 12 (Fig. 3). A Giken AT150 ‘press-in’ piler was used to install the piles in jack strokes of 700 mm. Reaction force was provided by sheet piles located at a minimum distance of 600 mm (∼6D) from the test piles. The maximum jacking re- sistance was encountered during the final stroke, and was re- corded by a load cell between the pile head and the piler. 2.3 Load test procedure A hydraulic jack was used to apply force through a load cell to the head of the single piles, or to a steel cap mounted on the pile groups. Pile head settlement, w, was monitored relative to inde- pendent reference beams. Six equal load increments were ap- plied up to 75% of the installation force of the test pile (or n times the installation force of the first pile for groups of n piles). A further 4-6 smaller load steps took the pile to plunging fail- ure. An unload-reload loop was conducted after a settlement of D/10 (10 mm). Each load increment was maintained until the pile head settlement rate was less than 0.02 mm/minute. 3 BACK-ANALYSIS: LOAD TRANSFER METHOD Back-analysis of the observed load-settlement response has been conducted using the RATZ load-transfer program (Randolph 2003). This program combines parabolic models for the local shaft (τs-z) and base (qb-z) resistance response with elastic compression of the pile to calculate the resulting pile head load–settlement response. The parabolic τs-z model re- quires the initial operative soil stiffness, Goper, to be estimated, in addition to the limiting local shaft resistance, τsf. The initial slope of the parabolic τs-z response is G/2D following the elas- tic solution of Randolph & Wroth (1978). The parabolic base response is defined by the limiting base resistance, qbf, and the settlement required to mobilise this resistance, wbf. The initial slope of the parabolic qb-z response is 2qbf/wbf. The pile groups were modelled using an interaction factor approach. The ‘elastic’ response of a pile element, defined by the initial stiffness of the τs-z and qb-z parabolae, was softened by a settlement ratio, denoted Rs for the shaft and Rb for the base. The ‘plastic’ component of settlement, represented by the parabolic deviation from the initial stiffness, remained un- changed. Rs and Rb were calculated as the proportional increase in settlement of a pile element due to the additional settlement contributions created by the neighbouring piles. It was assumed that the piles within each group carried equal load. Following Randolph & Wroth (1979), the settlement trough around the shaft was estimated from the Randolph & Wroth (1978) elastic solution, whilst the settlement around the base was estimated from an approximation of the elastic rigid punch solution. 4 RESULTS 4.1 Pile installation A rigid plug formed within each pile during jacking. Plug lengths in the range 0.8-2.5 m were recorded after installation. All piles failed in a plugged manner during load testing. The jacking force at the end of the final installation stroke, Qinstall, is shown in Table 1 and on Figure 2. Table 1. Summary of test programme Description [-] ID. [-] n [-] L [m] Qinstall [kN] Qgroup [kN] (per pile) Single pile TS1 1 5.85 78 78 (78) Single pile TL1 1 6.85 79 81 (81) Single pile (rusty) TLR1 1 6.85 72 77 (77) Two piles TL2 2 6.85 73 170 (85) Two piles TL2X 2 6.85 38 97 (48.5) Three piles TL3 3 6.85 69 240 (80) Six piles TS6 6 5.85 55 340 (56.7) Twelve piles TL12 12 6.85 67 820 (68.3) Considerable variation in Qinstall (+/- 35%) uploads/s3/ 164-icsmge-osaka-2005-vol2-2103-2106.pdf