Behavior of Fiber-Reinforced Polymer Composite Piles Under Vertical Loads

Previous | Table of Contents | Next

Chapter 5. Conclusions and Recommendations

The main purpose of this project was to address the need for a feasibility assessment of FRP composite-bearing piles for highway and related facilities substructures, replacing traditional materials such as timber, concrete, or steel, specifically in the construction of waterfront structures in hostile marine environments. The engineering use of FRP composite-bearing piles raised the need to investigate their field performance and to develop and evaluate reliable testing procedures and design methods to assess:

• Mechanical short-term behavior of FRP composite materials under axial compression loads.
• Behavior of FRP composite piles under vertical loads.
• Evaluation of FRP composite piling capacity, drivability, and constructability.

This project report summarizes the results of the full-scale experiment conducted at Port Elizabeth, NJ, and the companion laboratory tests. The main conclusions are summarized below.

Mechanical Short-Term Behavior of FRP composite materials under axial compression loads

Chapter 2 presented an engineering analysis approach for establishing the equivalent mechanical properties of the composite material, including elastic modulus for the initial loading quasilinear phase, axial compression strength, inertia moment, and critical buckling load. The conclusions of this chapter were as follows:

• The laboratory axial compression tests conducted in this study on the SEAPILE composite sample and its component materials (the fiberglass bars and the recycled plastic), illustrated that the recycled plastic has a significant effect on the composite material stability regarding buckling. However, due to its low stiffness and compressive strength compared with the reinforcing fiberglass bars, the recycled plastic does not appear to effectively prevent the peripheral disintegration of the fiberglass bars and, therefore, has a limited contribution to the axial compression strength of the composite material. No delamination between the rebar and the plastic was observed during the axial compression loading.
• The experimental results obtained for the SEAPILE composite materials illustrated that the nonlinear response of the recycled plastic to the axial loading is strain-rate dependent. However, because the quasilinear response of the fiberglass bars to the axial loading does not seem to be strain-rate dependent, the strain-rate effect on the mechanical behavior of the SEAPILE composite is expected to be relatively small.
• A composite material model has been developed for establishing the equivalent mechanical properties of the SEAPILE composite material, including elastic modulus for the initial quasilinear loading phase, axial compression strength, moment of inertia, and critical buckling load. The equivalent material properties of the composite are related to the mechanical properties of the component materials, assuming strain compatibility between the plastic and the fiber reinforcement bars during the axial compression loading.
• As illustrated in table 1, which summarizes the comparison between the predicted and experimental values of the SEAPILE composite material properties, the proposed model appears to predict the experimental results of the axial loading tests on the composite material sample fairly well. Further, this model seems to provide an effective framework for analyzing the effect of the recycled plastic on the mechanical behavior of the FRP composite.
• Further research is required to better understand the failure mechanisms involved in the SEAPILE composite material. Both laboratory and full-scale loading tests are required to provide a relevant database to evaluate and develop reliable design methods for the engineering applications of composite construction materials in load-bearing piles that are used for waterfront and highway structures.
• A compression axial test was performed on a 38.7-cm- (15.25-inch-) diameter, 80-cm- (31.5-inch-) long PPI pile sample containing 16 steel reinforcing rods of 2.54-cm (1.0‑inch) nominal diameter to establish the composite material properties. The test results on the PPI sample showed that this material has a behavior similar to that of steel.
• Axial compression tests were performed on the Trimax pile recycled plastic at several strain rates. The dimensions of the sample were 25.4 cm (10 inches) in diameter and 50.8 cm (20 inches) long. The test results showed that the nonlinear response of the Trimax material to the axial loading is strain-rate dependent. The Young's modulus of this material obtained for the linear portion of the stress-strain curve at the strain rate of 1.7 mm (0.07 inch) per min (0.33 percent/min) is 370,000 kPa (53.7 ksi). Poisson's ratio calculated for this strain rate at 1000 kPa (0.145 ksi) is equal to 0.35, and the corresponding shear modulus value for the linear portion of the stress-strain curve is 1.37 x 10⁵ kPa (19.9 ksi).

Behavior of FRP Composite Piles Under Vertical Loads

The testing program included four SLTs on instrumented piles, which were driven in the selected site at Port Elizabeth, NJ. In chapter 3, the experimental results were compared with the methods commonly used for evaluating the ultimate capacity, end bearing capacity, and shaft frictional resistance along the piles. This engineering analysis led to the following conclusions:

• The full-scale experiment confirmed that the FRP composite piles can be used effectively as load-bearing piles and represent an alternative for deep foundation construction, especially in waterfront environments and aggressive soils.
• Distinct plunging failure occurred during the SLTs on PPI and SEAPILE piles as the applied loads reached 115 and 90 t (253 and 198 kips) and the measured pile top settlements were 1.64 cm (0.65 inch) and 1.16 cm (0.46 inch), respectively.
• The Lancaster Composite, Inc., pile did not experience a distinct plunging failure, and the maximum load applied on the pile reached 128 t (282 kips) with a measured settlement of 1.73 cm (0.68 inch).
• The maximum load applied on the American Ecoboard pile, which was driven to the sandy layer, was 60 t (132 kips). At this load, the pile top settlement was 9.34 cm (3.7 inches), and no distinct plunging failure was observed. The pile top settlements of this pile, which contained only recycled plastic with no reinforcement bars, were significantly greater than the settlements measured during the tests on the other FRP piles.
• Several methods commonly used for evaluating the ultimate capacity were compared with the SLT results. In general, the loads calculated for FRP piles using the Chin-Kondner method are greater than the maximum loads that were applied at the field test. The DeBeer yield load method yields conservative loads compared to the maximum loads applied at the SLTs. The Davisson offset limit load method, using an equivalent elastic modulus that is experimentally derived from the quasilinear load-settlement relationship of the unloading-reloading cycle, yields limit loads that are within the range of loads obtained with the above-mentioned methods and settlement estimates that are consistent with the field test results.
• The experimental results are compared with several codes^(5-7) and analytical methods^(4,39‑40) commonly used for evaluating the shaft friction and end bearing capacity of the piles. The maximum end bearing capacities measured at the Port Elizabeth site were relatively small in comparison to the applied loads, indicating that the piles were frictional piles. The American Petroleum Institute (API) method yields end bearing capacities that are significantly lower than the FHWA and Meyerhof methods, and are in fairly good agreement with the end bearing capacity measured in the field tests.
• Several methods commonly used for evaluating the shaft friction along the piles were compared with the experimental results. Burland and Meyerhof's methods and the AASHTO code yield shaft friction values that are in good agreement with the average shaft friction measured on the PPI and SEAPILE piles in the upper soil (fill and sand) layer. The FHWA method yields the best correlation for the shaft friction values obtained in the lower silt and clayey soil layer.
• The engineering use of FRP piles on a widespread basis requires developing and evaluating design methods through extensive comparison of predictions with reliable data measured during full-scale loading tests. For this purpose, the FRP piles were instrumented by strain gauges that were specifically designed for strain measurements in these piles. This instrumentation allowed measuring the shaft friction distribution along the piles and the end bearing capacities in the saturated soils of the selected site during the SLTs on the FRP piles.
• This full-scale experiment demonstrated the feasibility of effectively using FRP piles as vertical load piles. However, because soil-pile interaction depends largely on prevailing soil profile and site conditions, further full-scale testing in different soil profiles is required to establish a reliable database for developing and evaluating codes and methods of analysis for designing FRP piles.

Evaluation of FRP Composite Piling Capacity, Drivability, and Constructability

In chapter 4, the authors described the full-scale experiment, the dynamic pile testing results, and the engineering analysis of the SLTs on the FRP piles. The main objectives of this demonstration project were to:

• Assess the drivability and durability of FRP piles during installation using PDA.
• Evaluate the currently available dynamic testing methods for establishing the dynamic properties of FRP piles and the dynamic soil-pile interface parameters during driving.
• Evaluate the dynamic methods currently used in predicting the load-set response and ultimate static capacity of FRP piles and, more specifically, the correlations between CAPWAP analysis and SLT results.
• Develop appropriate design criteria for the allowable dynamic stresses in the composite pile material and its basic components during pile driving.
• Obtain dynamic data records of the FRP pile driving that, in the absence of design criteria and field data for these piles, can be used to establish allowable tension and compression stresses for performing dynamic analyses and evaluating drivability and integrity during driving.

The engineering analysis of the dynamic pile testing results and the SLTs led to several conclusions.

Drivability and Integrity During Driving

• PDA during pile driving and restriking of the PPI and SEAPILE piles showed no damage or separation between the bars and the recycled plastic material, except at the upper foot of each PPI pile. Similarly, in the case of the Lancaster Composite, Inc., piles, no damage or separation between the concrete and the FRP shell piles was observed.
• The PDA and CAPWAP analysis results showed that all the piles experienced a long-term increase in pile capacity from the EOI to the BOR. For most of the tested piles, the measured setup factor, defined as a ratio between the BOR and EOI capacities, was in excess of 2.0. The dissipation of excess pore water pressure generated during pile driving resulted in a significant increase, up to about 100 percent, of the shaft and toe Smith damping.

Evaluation of Dynamic Testing Methods

• The dynamic testing procedures commonly used for steel and concrete piles (i.e., PIT, PDA testing—high-strain, and PDA testing—early portion of the high-strain records) were conducted to evaluate the dynamic properties of the composite piles. The parametric study conducted with CAPWAP and GRLWEAP showed that the elastic modulus obtained from PDA testing—early portion of the high-strain records yields the best correlations between (1) the calculated values of the transfer energy level and the energy measured by PDA, and (2) the calculated blow count and the measured field records.

CAPWAP Analysis

• Comparison between the GRL recommendations and the dynamic soil properties obtained from the CAPWAP analysis leads to the following results:
  • The soil shaft quake values recommended by GRL appear to be consistent with the CAPWAP results obtained for all the piles tested with the exception of the PPI piles.
  • The soil toe quake values obtained for the FRP test piles are significantly lower (about 45 percent) than the values recommended by GRL.
  • The toe damping values obtained for the FRP test piles are consistently greater (about 250 percent) than the values recommended by GRLl.
  • The shaft Smith damping values recommended by GRL appear to be consistent with the CAPWAP results obtained for all the piles tested.
• For all SLT piles, CAPWAP capacities ranged from 938.6 to 1,267.7 kN (211 to 285 kips) during the BOR. In general, CAPWAP analysis showed most of the ultimate capacity came from shaft resistance when the piles were driven to penetrations of 17.7 m (58 ft).

Correlations Between CAPWAP Analysis and Static Load Tests

• The CAPWAP analysis yielded load-set curves and ultimate capacities, which correspond fairly well to the SLT results.
• For the SEAPILE and PPI piles, the CAPWAP analysis seemed to indicate that the settlement at failure was close to elastic compression settlement during loading.
• For the low stiffness American Ecoboard pile that was driven to the sandy layer, the settlement, reaching 96 mm (3.7 inches) under a 60-t (132-kips) load, appeared to be due mainly to the elastic compression. The CAPWAP analysis was limited to the settlement range of about 20 mm (0.78 inch). For this range, it seemed to be quite consistent with SLT results.

Design Criteria and Allowable Stresses

• Design criteria for allowable compression and tension stresses in the FRP piles were developed considering the equation of the axial force equilibrium for the composite material and assuming no delamination between its basic components.
• During pile driving, with the exception of the PPI piles, the measured compression and tension stresses did not exceed the allowable stresses. In the absence of design criteria and field data for FRP plies, the maximum stresses obtained for these piles in the site observations can be used to establish allowable tension and compression stresses for pile driving.

R&D Needs Assessment

The dynamic and static loading tests on instrumented FRP piles conducted in this project demonstrated that these piles can be used as an alternative engineering solution for deep foundations. However, their widespread use will require further site testing and full-scale experiments to establish a relevant performance database to develop and evaluate reliable testing procedures and design methods.

The time-dependent stress-deformation behavior of composite recycled plastic materials is of concern, because the FRP piles may undergo an excessive deformation due to an applied sustained loading. The engineering use of FRP piles on a widespread basis requires developing and evaluating reliable testing procedures and design methods to determine the long-term behavior of these composite piles.

Further, research is now required to evaluate the effect of environmental conditions (i.e., soil confinement, groundwater, etc.) on the long-term behavior of recycled plastic composite materials as well as the combined effects of chemical and mechanical degradation processes. Both laboratory and full-scale loading tests are required to provide a relevant database to develop and evaluate the assessment of the long-term performance of composite, time-dependent FRP piles and to determine the limit creep load for their engineering use in waterfront and highway structures.