Novel Toe Driving for Thin-Walled Piles And

Western Ontario University, Canada – Research Study

Novel toe driving for thin-walled piles and performance of fiberglass-reinforced polymer (FRP) pile segments

Mohammed Sakr, M. Hesham El Naggar, and Moncef Nehdi

Abstract: Despite the rapidly growing use of pile foundations, it is presently difficult to assure the integrity and uniformity of the cross-sectional area of cast-in-place piles when using normal concrete. Cavities and soil encroachments leading to soil pockets can jeopardize their load-bearing capacity. Moreover, corrosion in reinforced concrete and steel shell piles has been very costly, exceeding US$2 billion in annual repair costs in the United States alone. To address these two challenges, extensive research has beenunderway at the University of Western Ontario to develop novel technology for the construction of piles. Self-consolidating concrete (SCC), a material that flows under gravity and assures the integrity of piles, is cast into fiberglass-reinforced polymer (FRP) tubes that provide corrosion-resistant reinforcement. A toe driving technique was developed to install the empty FRP shells into the soil, and SCC is subsequently

cast into the shells. Driving tests using this new technique were carried out on large-scale model FRP and steel pipe piles installed in dense dry sand enclosed in a pressure chamber. FRP–SCC and steel closed-end piles were also driven using conventional piling at the pile head.Static load tests were conducted on the various pile specimens under different vertical and horizontal confining pressures. The pile specimens were instrumented to investigate their dynamic behaviour under driving and their response to static compressive, uplift, and lateral loading. It is shown that the toe driving technique is very suitable for installing FRP piles in dense soils.

Results from the driving tests and static load test indicate that FRP–SCC hybrid piles are a very competitive and attractive option for the deep foundations industry.

Key words: FRP, self-consolidating concrete, piles, pile drivability, toe driving, axial load, uplift load, lateral load,

large-scale modeling, shaft resistance, dense sand.

Résumé : En dépit de l’utilisation croissante des fondations de pieux, il est présentement difficile de s’assurer de l’intégrité et de l’uniformité de la section en travers des pieux coulés en place lorsqu’on utilise le béton normal. Des cavités et des empiétements du sol conduisant à des poches de sol peuvent compromettre la capacité portante de ces pieux. De plus, la corrosion des armatures du béton armé et des enveloppes d’acier des pieux a causé des coûts très importants excédant 2 milliards de dollars en coûts de réparation annuelle aux États-Unis seulement. Pour s’attaquer à ces deux défis, une recherche intense a été entreprise à la University of Western Ontario pour développer une nouvelle

technologie pour la construction de pieux. Un béton auto-consolidant (SCC), matériau qui s’écoule sous l’effet de la gravité et assure l’intégrité des pieux, est coulé dans des tubes de polymère renforcé de fibre de verre (FRP) qui four-nit une armature résistante à la corrosion. Une technique de fonçage par le pied a été développée pour installer dans le

sol les tubes vides de FRP, et par la suite, le SCC est coulé dans ces tubes. Des essais de fonçage au moyen de cette nouvelle technique ont été réalisés sur des modèles à grande échelle de pieux avec tubes FRP et tubes d’acier mis en place dans un sable dense sec à l’intérieur d’une chambre à pression. Des pieux de FRP–SCC et d’acier à bout fermé

ont aussi été foncés au moyen de fonçage conventionnel sur la tête du pieu. On a fait des essais de chargement statique sur les divers spécimens de pieux sous différentes pressions de confinement verticales et horizontales. Les spécimens de pieux ont été instrumentés pour étudier leur comportement dynamique durant le fonçage et leur réaction à des

chargements statiques en compression, en soulèvement et latéraux. On montre que la technique de fonçage en pied est très adéquate pour la mise en place des pieux FRP dans les sols denses. Les résultats des essais de fonçage et de l’essai de chargement statique indiquent que les pieux hybrides FRP–SCC sont une option très attrayante et compétitive

pour l’industrie de fondations profondes.

Mots clés : FRP, béton auto-consolidant, pieux, potentiel de fonçage, charge axiale, résistance au soulèvement, charge latérale, modèle à grande échelle, résistance du fût, sable dense.

[Traduit par la Rédaction]

Received 20 December 2002. Accepted 3 September 2003. Published on the NRC Research Press Web site at on

14 April 2004.

M. Sakr and M.E. El Naggar.1 Geotechnical Research Centre, Faculty of Engineering, The University of Western Ontario, London, ONN6A 5B9, Canada.

M. Nehdi. Department of Civil and Environmental Engineering, The University of Western Ontario, London, ONN6A 5B9, Canada.

1Corresponding author (e-mail: ).

Research significance

The outcome of this research could be very useful to the construction industry, since implementing fiberglass-reinforced polymer (FRP) composites in deep foundationscould result in much longer service life (expected to be three times that of conventional pile materials), which in turn could result in substantial life cycle savings in project costs.

FRP piles can also eliminate many durability problems of deep foundations located in harsh environments, such as marine borer attack on wooden piles and corrosion of steel reinforcement in concrete or steel casing piles. Self-consolidating concrete (SCC) can eliminate the lack of cross-sectional integrity and soil encroachments often experienced

in deep foundation construction due to the lack of visibility and accessibility.

A new pile driving technique at the pile toe is examined in this study.It can result in substantialimprovements in the pile installation process. It allows driving FRP piles in difficult soil conditions such as dense sand. Moreover, when used with conventional steel shell piles, this technique allows reducing the pile wall thickness and improving the driving efficiency. This study also provides designers and contractors with a demonstration case and a database on the feasibility and advantages of FRP–SCC composite piles.

Introduction

Pile foundations are generally used to support structural loads in situations where shallow foundations cannot provide the required bearing capacity or where soil settlement is a major concern. Typical pile materials for deep foundations include steel, concrete, and wood. Wooden piles, precast or reinforced concrete piles, cast-in-place concrete piles, and steel piles have been used in practice for a long time. These pile materials have limited service life, however, and are associated with high maintenance costs when used in harsh environments, for instance because of corrosion degradation and marine borer attack (Lampo et al. 1998). Moreover, lack of cross-sectional integrity through the formation of air pockets and soil encroachments is often experienced in the construction of drilled shafts due to the lack of visibility and accessibility.

A relatively recent trend in deep foundation design is using FRP composite materials in piles because of their light weight, high specific strength, corrosion resistance, chemical and environmental resistance, and low maintenance cost.Composite FRP piling has been used in practice in waterfront barriers, fender piles, and bearing piles for light structures

(Iskander and Hassan 1998). Most composite piling products are made of FRPs or high-density polyethylene (HDPE), with fiberglass reinforcement and additives to improve mechanical properties, durability, and ultraviolet light protection. However, FRP composite piles have not yet gained wide acceptance in the deep foundation industry because of the lack of proper design guidelines for predicting their drivability and load-carrying capacity, and the need for construction projects that demonstrate the concept. Implementing the use of FRP materials in piling applications may result in substantial benefits to the deep foundation industry, which is the underlying motivation for the research reported herein.

Few studies are available in the literature on FRP piles, probably due to their novelty. Iskander and Hassan (1998) reviewed the available FRP composite products for fender applications. Frost and Han (1999) studied the interface friction between FRP and sand and showed that the interface friction angle between sand and FRP depends on the normal stress level, the shape of sand particles, and the relative roughness of FRP.

Iskander et al. (2001) conducted a parametric study using wave equation analysis on FRP piles. They concluded that the drivability of composite materials depends on the specific weight and elastic modulus of the composite section. Ashford and Jakrapiyanun (2001) compared the drivability of FRP composite piles and steel and concrete piles also using wave equation analysis. They found that FRP piles compared favorably with steel pipes and precast prestressed concrete piles. Mirmiran et al. (2002) conducted a field installation of empty FRP tubes, FRP–concrete piles, and spliced tubes using the conventional top driving technique and concluded that empty FRP piles cannot be driven to great depths and cannot be driven in hard soils such as dense sand. Han and Frost (1999) examined buckling loads of FRP piles during installation and service life. Their study showed that critical buckling loads of FRP piles depend on the shear effect coefficient, the lateral soil resistance, the embedment ratio, the overall boundary conditions, and the critical length. Shear deformation plays an important role in reducing critical loads for FRP piles and steel piles. Critical loads increase with an increase in the shear effect coefficient, lateral soil resistance, and embedment ratio. Han and Frost

showed also that the skin friction between the pile and soil plays a very limited role. They concluded that buckling of FRP piles is unlikely to happen except for very long piles or for piles in very soft soils.

Pando et al. (2000) performed a full-scale pile load test at the Route 40 bridge (Nottoway River, Virginia) using FRP tubes filled with concrete and prestressed concrete piles. Their experimental program consisted of pile driving analyses, Statnamic axial compression load tests, and Statnamic lateral load tests on both piles. They found that both pile materials exhibited a similar response with respect to their drivability and axial capacity in compression. The response of FRP composite piles for lateral loads was significantly softer than that of the prestressed concrete piles, however.

Objectives

The primary objectives of this research are to develop an efficient technique for driving FRP piles and thin-walled piles; to experimentally evaluate the performance of FRP–SCC hybrid piles under compressive, uplift, and lateral loading; and to compare the behaviour of FRP–SCC hybrid piles with that of conventional steel piles.The drivability of FRP

tubes and the behaviour of FRP–SCC hybrid piles under various loading conditions were simulated in a pressure chamber specially designed to simulate field vertical and radial soil confinement pressures. The performance of FRP pile segments was compared to that of steel pile segments driven using the toe driving technique and also using conventional

driving at the pile head. The results of this extensive research program are reported in the following sections.

Soil samples for pile installation

Dense sand samples were prepared in a pressure chamber to represent difficult field piling conditions. Fanshawe brick sand was used in the tests and consisted of fine, subround to round, air-dried sand. The grain-size distribution for this sand is shown in Fig. 1. The sand used was classified as poorly graded, with particle sizes in the range of 0.075– 2.00 mm. A standard density test showed that the sand had a maximum unit weight of 17.72 kN/m3 and a minimum unit weight of 14.66 kN/m3. Soil samples for pile installation were prepared using a raining (pluviation) technique, which is known to provide uniform soil properties. The relative density of the soil samples used for this testing program was about 90%, with a standard deviation of 2.5%. Table 1 provides details of the soil properties. The density of the soil samples was monitored using three different techniques to ensure that soil properties do not vary from one pile testing

to another: (i) three small soil samples of 106 × 10–6 m3 were taken every 100 mm of soil deposition at three different locations in the pressure chamber to monitor the uniformity of the relative density of the sand in the pressure chamber; (ii) a relatively large sample using a 944 × 10–6 m3 mould was taken at each 100 mm of soil deposition, and at each 200 mm of soil deposition the raining process was stopped and in situ nuclear density measurements were conducted to determine the density and water content of the soil; and (iii) the pressure chamber was weighed both empty and filled with soil to determine the average density of the soil sample. These measurements ensured that a uniform soil sample would be achieved with the desired relative density. Figure 2 shows the density versus depth for the soil samples. It is clear from Fig. 2 that the soil density was reasonably uniform. The disparity between the density measurements obtained from the mould and those obtained from small samples was about 2%, and the disparity between the mould and nuclear measurements was less than 10%. The disparity between unit weight values measured using containers and those from nuclear density measurements can be attributed to decaying radioisotope used in the nuclear density measurements.

Test setup

A large-scale modeling facility was developed at the University of Western Ontario to test large-scale model pile segments of 1.5 m in length at different confining pressures. The key advantage of using large-scale pile modeling over other smaller scale model pile and centrifuge testing is the ability to reasonably simulate vertical and radial effective

stress profiles at any depth using a pile with relatively large dimensions. Modeling the pile thickness is of prime importance for FRP composite piles, since these composites are often nonhomogeneous and exhibit anisotropic viscoelastic behaviour (which is pertinent for piles subjected to large lateral loads); a small model pile with a small thickness such

as in centrifuge testing may not adequately represent the material nonhomogeneity. Another advantage of using large-scale model piles is the ability to correctly present pile–soil interface characteristics. Moreover, model piles can be easily installed using different driving techniques. Lastly, a pressure chamber can be operated and maintained at a fraction of the cost of a centrifuge (Yazdanbod et al. 1984). However, the only limitation for large-scale pile modeling is that it is only applicable to pile segments. Therefore, special caution should be exercised when extrapolating the results of these tests to full-scale piles.

The pile installation setup consists of a pressure chamber, pile driver, toe driving device, model piles, and instrumentation and data acquisition system. Six tests were conducted on FRP tubes, FRP–SCC composites, and steel piles. Figure 3a shows a schematic view of the testing setup, and Fig. 3b shows an oblique view of the pile driving assembly.

Pressure chamber

As shown in Fig. 3a, the soil column is formed in a containment steel cylinder of 1.34 m inside diameter and 1.52 m height. Vertical and radial stresses within the in situ soil at different depths were modeled using vertical and circumferential air bladders with independent control. Therefore, it was possible to establish the ratio of horizontal to vertical effective stresses in the chamber, K0, to be equal to the ambient stress ratio at that depth in situ. A thermoplastic polyfin membrane sleeve was used as a radial pressure bladder. Pressure was applied to the membrane by compressed air through an input port and monitored by a pressure sensor connected to the compressed air supply and two earth pressure cells embedded vertically and horizontally at a depth of 0.6 m to measure the actual radial and vertical pressures in the soil sample. The vertical pressure was applied by a flat gum rubber membrane 4.76 mm in thickness, situated at the top of the chamber. When pressurized, this membrane reacts against a 19 mm steel cover plate. Three ports pass through both the plate and the membrane: a central port with a diameter of 279 mm to accommodate the pile and another two ports with a diameter of 50.8 mm to serve as exit points for an air pressure hose, drainage hoses, and earth pressure cell wires. The rubber sleeve was closed at the bottom by a 50.8 mm thick waffle-type neoprene energy absorber to reduce the energy in waves reflected from the base of the chamber. The rubber tube is impermeable and provides water tightness to the soil column. Field-like conditions were achieved in the pressure chamber by limiting the size of pile specimens to be tested to 168 mm in diameter and installing the pile to a maximum depth of 1.20 m to satisfy a horizontal boundary of eight pile radii, and three vertical radii from the pile tip (Vipulanandan et al. 1989).

Model pile specimens