Skip to contentUnited States Department of Transportation - Federal Highway AdministrationSearch FHWAFeedback

Geotechnical Engineering

 

Behavior of Fiber-Reinforced Polymer (FRP) Composite Piles under Vertical Loads

CHAPTER 4. EVALUATION OF FRP COMPOSITE PILING CAPACITY USING WAVE EQUATION ANALYSIS

The engineering use of FRP piles on a widespread basis presently raises the need to establish testing procedures to determine the dynamic properties of FRP composite materials. In particular, PDA tests and full-scale loading tests on FRP piles must be conducted to establish site correlations for evaluating the use of wave equation analysis in predicting static piling capacity and evaluating pile drivability and integrity. The main objective of the study described in this chapter was to conduct a full-scale experiment including dynamic and static load tests on FRP piles to address these engineering needs.

PDA and PIT test results were analyzed using CAPWAP(1,45-46) and GRLWEAP(47) programs to establish the dynamic properties of the FRP piles. The PDA was also used to evaluate the feasibility of installing FRP piles using standard pile driving equipment. Pile-bearing capacities were assessed using the CAPWAP program with the dynamic data measured by the PDA, and the calculated pile capacities were compared to the results of SLTs performed on the four FRP piles.

Testing Program and Site Observations

Test Piles and Dynamic Instrumentation

The test piles were driven with an ICE 70 single-acting, variable stroke hydraulic hammer. According to the manufacturer's literature, this hammer has a 31.1-kN (7.0-kip) ram with a maximum rated energy of 28.5 kJ (21 ft-kips) at a maximum stroke of 91.4 cm (3 ft). This hammer has a pump-controlled stroke, and during the initial driving and restrike, the stroke was varied from 30.5 to 91.4 cm (1 to 3 ft). Plywood cushions were used to protect the pile top. In most cases during driving and restriking, the cushion was nominally 24.8 cm (9.75 inches) thick. 15.2-cm- (6-inch-) thick cushions were used to drive the Lancaster piles.

Dynamic measurements of strain and acceleration were taken 91.4 cm to 1.5 m (3 to 5 ft) below the head of the test piles. During initial drive, two strain transducers and two accelerometers were bolted at opposite sides of the pile to monitor strain and acceleration. During redrive, four strain transducers and four accelerometers were attached to the pile. Strain and acceleration signals were conditioned and converted to forces and velocities by the PDA. During driving, the PDA calculated values of the maximum transferred hammer energy and the maximum compression stress at the gauge location, and an estimate of the pile capacity by the Case Method.(8) These results were displayed on the PDA monitor for every blow.

Force and velocity records from the PDA were also displayed on a graphic liquid crystal display (LCD) screen to evaluate data quality, soil resistance distribution, and pile integrity. The force and velocity records were also digitally stored on a disc for subsequent laboratory analysis.

Site Observations on Drivability and Durability During Pile Installation

Figures 68 through 71 show the FRP piles before and after driving. Site observations on drivability and durability during installation of the test piles are summarized below.

The steel pile was driven with the lowest blow counts (36 blows/m (11 blows/ft)) and the highest transfer energy of 23.1 kJ (17 ft-kips); this pile, however, was driven without a plywood pile cushion and with a helmet specifically designed for steel pipe piles. Adding 15.2 to 24.8 cm (6 to 9.75 inches) of pile cushion may explain some of the energy losses and increased blow counts observed with the other piles.

Among the FRP piles, the SEAPILE pile was driven with the lowest blow counts (62 to 118 blows/m (19 to 36 blows/ft)), and showed relatively high transfer energies of 13.6 to 17.6 kJ (10 to 13 ft-kips) at the end of driving. The pile cushion for these piles showed very little degradation after driving.

The PPI piles were driven with blow counts ranging between 1 to 1.4 blows/mm (27 to 37 blows/inch), and transfer energies at the gauge location of 8.1 to 10.8 kJ (6 to 8 ft-kips), significantly lower than those obtained for the SEAPILE and Lancaster Composite, Inc., piles. After driving, it was observed that the PPI piles experienced severe pile top degradation; the plastic matrix covering the rebar had melted and some pieces of the steel rebar had sheared off. The heat and pile top damage could have been the reason for the losses in the transfer energy. After driving, the upper foot of each PPI pile was removed, and no damage or separation between bars and the plastic material was observed.

Click for text description
Click for text description

Figure 68. Photo. PPI pile.

Click for text description
Click for text description

Figure 69. Photo. SEAPILE pile.

Click for text description
Click for text description

Figure 70. Photo. Lancaster Composite, Inc., pile.

Click for text description
Click for text description

Figure 71. Photo. American Ecoboard pile.

At the end of installation (EOI), the Lancaster Composite, Inc., piles had blow counts ranging from 154 to 305 blows/m (47 to 93 blows/ft), with calculated transfer energy values of 12.2 to 17.6 kJ (9 to 13 ft-kips) at the end of drive. It should be noted that to reduce the estimated tension stresses, the Lancaster Composite, Inc., SLT pile was driven with a reduced stroke of 61.0 cm (2 ft) for the last few feet of penetration; this driving process resulted in a higher blow count of 305.1 blows/m (93 blows/ft). By the EOI of these piles, it was observed that the pile cushion had experienced initial burning.

The Lancaster Composite, Inc., pile that was driven to refusal (almost 0.4 blows/mm (10 blows/ inch)) into the weathered shale at the depth of 26.5 m (87 ft) below grade was spliced. Driving resistance ranged from 164 to 299 blows/m (50 to 91 blows/ft) for a penetration of 18.8 to 26.5 m (60 to 87 ft).

The American Ecoboard piles were driven into the upper sand layer at approximately 9.1 m (30 ft) below the ground surface. They were spliced at the depth of 6 m (19.7 ft). After splicing, the American Ecoboard dynamic testing results evidenced a toe signal, which indicated that the stress wave propagated through the splice. A significant impedance increase followed by a sharp decrease was observed at the splice location. This wave reflection is likely to be the result of the additional steel bolts and sleeve, as well as of a slight gap between the two sections.

Dynamic Properties of the FRP Composite Pile Materials

Methods for Evaluating Dynamic Properties

Before dynamic testing, several material properties at the gauge location have to be determined. These data include the pile length, cross-sectional area, specific weight, wave speed, and elastic modulus. Testing procedures to establish the dynamic elastic modulus of FRP composite piles have not yet been established or evaluated. For the purpose of this study, the dynamic testing procedures commonly used for steel and concrete piles have been followed to evaluate the dynamic properties of the composite piles, including:

  • PIT—low-strain integrity testing (ASTM D5882).(48)
  • PDA testing—high strain (ASTM D4945).(49)
  • PDA testing—considering the initial portion of the high-strain records (ASTM D4945).(49)

These testing procedures are briefly summarized below.

Pile Integrity Tester (PIT)

Before initial installation, low-strain integrity testing was performed with a PIT. The PIT testing (ASTM D5882) consists of attaching a small accelerometer to the pile top sample and using a small handheld hammer to lightly impact the pile sample top. This impact induces a low-strain stress wave in the pile, which travels to the pile toe and reflects back toward the pile top. With the length of the pile sample known, a wave speed, c, can be calculated from the time measured between the impact and the toe response. The wave speed can, in turn, be used with a known material density, r, to calculate the low-strain dynamic modulus using the relationship in figure 72.

Click for text description

Figure 72. Equation. Dynamic modulus E.

PDA Testing—High Strain

PDA testing yields wave speed following a procedure similar to that used for low-strain PIT tests. For a specified length of the pile, the time between impact and toe response can be used to estimate an overall wave speed. The dynamic modulus, E, can be determined using the equation in figure 72.

PDA Testing—Initial Portion of the High-Strain Records

This method involves the initial portion of the high-strain records and the expected proportionality between measured strain and wave speed. Assuming that for low strains, the FRP piles behave as a linear elastic material and no changes in resistance or pile properties occur during the pile installation for the initial part of the high-strain record, this assumption implies that the wave speed is directly proportional to the measured strain and can, therefore, be calculated from the equation in figure 73.

Click for text description

Figure 73. Equation. Pile particle speed v.

Where:

v = Pile particle speed calculated from the acceleration record.

e = Measured strain.

c = material wave speed.

The dynamic modulus, E, then can be determined using the equation in figure 72.

Dynamic Properties—Test Results

PIT tests were performed before driving on American Ecoboard, Lancaster Composite, Inc., PPI, and SEAPILE full-scale piles. For the FRP piles, the foam (i.e., outer layer) is stiffer than the core (i.e., inner layer). Therefore, for the American Ecoboard piles, the PIT tests were conducted on the outer layer, and for the SEAPILE and PPI piles, tests were conducted on the fiberglass bars and the reinforced foam, respectively.

In a linearly elastic pile material, such as steel or concrete, the wave speed determined from the proportionality condition (figure 73) is equal to the overall wave speed calculated from the PDA results. In the case of FRP piles, however, the overall wave speeds obtained with PIT, cl, PDA, ch, and the proportional wave speed, cp, values were consistently different. Table 11 summarizes these results.

Table 11. Elastic modulus of FRP piles estimated from PIT and PDA tests.
Pile Type Specific Weight(1) Low Strain (PIT) High Strain (PDA)
kN/m3 (lb/ft3) Measured Wave Speed (Test I) cl , m/s (ft/s) Estimated Elastic Mod.El, kPa (ksi) Proportional Wave Speed (Test II) cp, m/s (ft/s) Estimated Elastic Mod. Ep, kPa (ksi) Overall Wave Speed (Test III) ch, m/s (ft/s) Estimated Elastic Mod. Eh, kPa (ksi)
American Ecoboard 7.9 (50) 1524 (5000) 1.86 x 106 (270) 1828.8 (6000) 2.68 x 106 (388) 1371.6 (4500) 1.50 x 106 (218)
Lancaster Composite 22.0 (140) 4114.8 (13500) 37.94 x 106 (5503) 4175.76 (13700) 39.07 x 106 (5667) 4023.36 (13200) 36.27 x 106 (5261)
PPI 8.0 (51) 3505.2 (11500) 10.03 x 106 (1455) 3810 (12500) 11.85 x 106 (1719) 3322.32 (10900) 9.01 x 106 (1307)
SEAPILE 8.5 (54) 2743.2 (9000) 6.50 x 106 (943) 3048 (10000) 8.03 x 106 (1165) 2529.84 (8300) 5.53 x 106 (802)

(1) Reported by the manufacturers.

Dynamic Pile Testing Data Analysis

The PDA data analysis was conducted to obtain the dynamic properties, including Smith damping and soil quake, and evaluate the load-set response and the static bearing capacity of the FRP piles. The dynamic analysis involved:

  • Case method analysis (PDA)—for onsite measurement of the transfer energy and evaluation of the driving system performance, pile stresses, structural integrity, and static capacity.
  • CAPWAP analysis—using the force and velocity data recorded in the field to obtain the dynamic properties, including Smith damping and soil quake, and to evaluate the load-set pile response.
  • GRLWEAP—using CAPWAP analysis results to evaluate the dynamic Young's modulus of each FRP composite pile.

The dynamic analysis results were compared to the results of SLTs performed on four FRP piles.

Case Method Analysis (PDA)

During the PDA testing, the Case Method is used to interpret the measured dynamic data to assess the hammer and driving system performance, determine pile head compression stresses, evaluate structural integrity, and estimate the static pile capacity. Table 12 summarizes the PDA dynamic test results, including the maximum measured pile top strain, maximum compressive stress, maximum estimated tensile stress, transfer energy to pile, hammer operating rate, and the pile capacities calculated by the case method and by CAPWAP analysis. The maximum compressive and tensile stresses were calculated using the dynamic elastic modulus obtained with the PDA testing, considering the wave speed-strain proportionality condition of the initial portion of the high-strain records.

CAPWAP Analysis

The CAPWAP Version 2000-1 was used with the data recorded in the field to evaluate the ultimate pile capacity of the tested piles. CAPWAP analyses compute soil resistance forces and their approximate distribution using the velocity and speed data recorded in the field during the dynamic pile testing. The CAPWAP results include an evaluation of the soil resistance distribution, soil quake and damping factors, and a simulated static load-set graph. The static load-set simulation is based upon the CAPWAP calculated static resistance parameters and the elastic compression characteristics of the pile. For each pile, the three elastic modulus values calculated from PDA and PIT tests as reported in table 11 were considered. The CAPWAP analyses using the elastic modules calculated from PDA testing and considering the initial portion of the high-strain records (1) yielded the most reasonable energy values for the ICE 70 hammer and (2) predicted load-set curves that yielded the best correlation with SLT results. CAPWAP static capacities calculated using the dynamic elastic modulus values are reported in table 13.

Table 12. PDA results.
Pile Designation Test Date Pile Penetration,(1) m (ft) Blow Count,(1) blows/m (blows/ft) Average Maximum Measured Strain (me) Average Maximum Compressive Stress,2 kPa (ksi) Maximum Estimate Tensile Stress,3 kPa (ksi) Average Transfer Energy,2 kJ (ft-kips) Hammer Rate, blows/min Case Method Capacity,4 kN (kips) CAPWAP Estimated Capacity, kips Remarks5
Steel Reference 11/14/01 18.9 (62) 36.1 (11) 714 147,547.8 (21.4) NA 23.1(17) 47 413.7 (93) 489.3 (110) EOI
5/24/02 18.6 (61) 0.31 b/mm (8/inch) 499 103,421.4 (15.0) NA 13.6 (10) 45 1267.7 (285) 1156.5 (260) BOR
Ecoboard 5/22/02 8.8 (29) 98.4 (30) 1540 4,136.9 (0.6) 689.5 (0.1) 12.2 (9) 76 435.9 (98) 382.5 (86) EOI, Section 2; 9.75-inch cushion
Lancaster 1 11/14/01 17.7 (58) 210.0 (64) 392 151,684.7 (2.2) 3447.5 (0.5) 14.9 (11) 50 186.8 (42) - ID, Section 1; 9.75-inch cushion
11/14/01 26.5 (87) 298.6 (91) 523 199,948.0 (2.9) 5516 (0.8) 17.6 (13) 49 1450 (326) - EOI, Section 2; 7.5-inch cushion
5/24/02 26.2 (86) 1.5 b/mm (77/ 0.5 inch) 448 16,547.4 (2.5) 4137 (0.6) 16.3 (12) 43 2375.2 (534) 2090.6 (470) EOR
Lancaster 4 11/15/01 17.7 (58) 154.2 (47) 377 14,479.0 (2.1) 5516 (0.8) 12.2 (9) 52 346.9 (78) - EOI; 6-inch cushion
5/24/02 17.1 (56) 0.96 b/mm (25/inch) 324 12411 (1.8) 1379 (0.2) 9.5 (7) 52 1072 (241) - BOR
Lancaster 3 (SLT) 11/15/01 17.7 (58) 154.2 (93) 399 15858.5 (2.3) 8274 (1.2) 13.6 (10) 56 480.4 (108) 516 (116) EOI; 6-inch cushion; 2-ft stroke
5/24/02 17.1 (56) 1.20 b/mm (31 / inch) 309 11721.5 (1.7) 1379 (0.2) 9.5 (7) 44 1049.7 (236) 1000.8 (225) BOR
PPI 5 11/15/01 19.2 (63) 121.4 (37) 436 5171.25 (0.75) 2068.5 (0.3) 9.5 (7) 49 453.7 (102) - EOI; 9.75-inch cushion
5/23/02 17.7 (58) 0.94 b/mm (24 / inch) 514 6205.5 (0.9) 689.5 (0.1) 6.8 (5) 45 756.2 (170) - BOR
PPI 7 11/15/01 19.05 (62.5) 111.5 (34) 535 6205.5 (0.9) 2068.5 (0.3) 10.9 (8) 48 484.8 (109) - EOI; 9.75-inch cushion
5/23/02 17.7 (58) 0.82 b/mm (21 / inch) 529 6205.5 (0.9) 1379 (0.2) 6.8 (5) 44 791.7 (178) - BOR
PPI 6 (SLT) 11/15/01 18.6 (61) 88.6 (27) 373 4137 (0.6) 2068.5 (0.3) 8.1 (6) 47 387 (87) 409.2 (92) EOI; 9.75-inch cushion
5/24/02 17.7 (58) 0.86 b/mm (22 / inch) 654 7584.5 (1.1) 1379 (0.2) 10.9 (8) 44 938.5 (211) 978.6 (220) BOR
Seaward 8 11/16/01 18.6 (61) 78.7 (24) 875 6895 (1.0) 2758 (0.4) 13.6 (10) 57 382.5 (86) - EOI; 9.75-inch cushion
5/22/02 17.7 (58) 0.86 b/mm (22 / inch) 858 6895 (1.0) 1379 (0.2) 12.2 (9) 56 885.2 (199) - BOR
Seaward 10 11/16/01 18.6 (61) 62.3 (19) 855 6895 (1.0) 2758 (0.4) 17.6 (13) 54 516 (116) - EOI; 9.75-inch cushion
5/22-3/02 17.7 (58) 0.39 b/mm (10 / inch) 926 7584.5 (1.1) 1379 (0.2) 16.3 (12) 43 978.6 (220) - BOR
Seaward 9 (SLT) 11/16/01 18.7 (61.3) 118.1 (36) 758 6205.5 (0.9) 2758 (0.4) 12.2 (9) 57 516 (116) 520.4 (117) EOI; 9.75-inch cushion
5/23/2002 17.7 (58) 0.51 b/mm (13 / inch) 1074 8274 (1.2) 1379(0.2) 14.9 (11) 45 978.6 (220) 1067.5 (240) BOR-Original CW


(1) Pile penetrations from ground surface for initial installation and bottom of excavation for restrike. 1 inch = 2.54 cm, 1 ft = 0.305 m

2 Measured at the location of the transducers.

3 Maximum tension stress estimated over all blows.

4 Using RX9 (maximum case method with Jc = 0.90) equation.

5 Unless otherwise noted, 3-ft hammer stroke.

BOR—Beginning of redrive.

EOI—End of installation.

CAPWAP analyses were performed on the records obtained for both the EOI and the beginning of restrike (BOR) of each SLT pile, the steel reference pile, and the Lancaster Composite, Inc., pile (No. 1), which was driven to refusal. The results are summarized in table 13.

Table 13. CAPWAP program calculation results.
Pile Designation Penetration Depth in Soil, m (ft) Driving Resistance, blows/set CAPWAP Remarks
Static Capacity Estimate Smith Damping Soil Quakes
Shaft, kN (kips) Toe, kN (kips) Total, kN (kips) Shaft, s/m (s/ft) Toe, s/m (s/ft) Shaft, mm (inch) Toe, mm (inch)
Steel Pile 18.9 (62) 11/ft 369.2 (83) 120.1 (27) 489.3 (110) 0.26 (0.08) 0.16 (0.05) 2.8 (0.11) 26.9 (1.06) EOI
18.3 (60) 8/inch 1049.7 (236) 106.8 (24) 1156.5 (260) 0.69 (0.21) 0.66 (0.2) 2.3 (0.09) 6.9 (0.27) BOR
American Ecoboard (SLT) 8.8 (29) 30/ft 102.3 (23) 280.2 (63) 382.5 (86) 0.95 (0.29) 1.05 (0.32) 2.8 (0.11) 16.0 (0.63) EOI; after splicing
Lancaster 1 25.9 (85) 77/0.5 inch 702.8 (158) 1387.8 (312) 2090.6 (470) 0.66 (0.2) 1.05 (0.32) 2.5 (0.10) 1.27 (0.05) EOR; pile driven to refusal
Lancaster 3 (SLT) 17.7 (58) 93/ft 240.2 (54) 275.8 (62) 516 (116) 0.66 (0.2) 0.1 (0.03) 3.0 (0.12) 18.3 (0.72) EOI
18.3 (60) 32/inch 836.2 (188) 164.6 (37) 1000.8 (225) 1.12 (0.34) 1.25 (0.38) 2.0 (0.08) 3.8 (0.15) BOR
PPI 6 (SLT) 18.6 (61) 27/ft 329.2 (74) 80.1 (18) 409.2 (92) 0.39 (0.12) 0.69 (0.21) 3.6 (0.14) 16 (0.63) EOI
17.7 (58) 22/inch 818.4 (184) 160.1 (36) 978.6 (220) 0.75 (0.23) 1.28 (0.39) 3.3 (0.13) 3.8 (0.15) BOR
Seaward 9 (SLT) 18.7 (61.3) 36/ft 400.3 (90) 120.1 (27) 520.4 (117) 0.66 (0.2) 0.79 (0.24) 2.3 (0.09) 15.5 (0.61) EOI
17.4 (57) 130 969.7 (218) 97.9 (22) 1067.5 (240) 0.98 (0.3) 1.25 (0.38) 2.5 (0.10) 2.8 (0.11) BOR
17.4 (57) 130 809.5 (182) 84.5 (19) 889.6 (200) 0.89 (0.27) 1.45 (0.45) 2.5 (0.10) 2.0 (0.08) BOR; revised with SLT

BOR—Beginning of restrike.

EOI—End of installation.

EOR—End of redrive.

Table 14 shows the quake and damping values recommended by the consulting firm GRL Engineers, Inc., for various types of soils and piles.

Table 14. Quake and damping values recommended by GRL.
  Soil Type Pile Type or Size Quake mm (in) Damping Factors/m (s/ft)
Shaft Quake All soil types All Types 2.5 (0.10) -
Toe Quake All soil types, soft rock Open ended pipes 2.5 (0.10) -
In dry soils, or in very dense or hard soils Displacement piles of diameter D or width D D/120 -
In submerged soils or in loose or soft soils Displacement Piles of diameter D or width D D/60 -
Shaft Damping Noncohesive soils - - 0.16 (0.05)
Cohesive soils - - 0.65 (0.20)
Toe Damping In all soil types - - 0.50 (0.15)

Comparison of the CAPWAP analysis results reported in table 13 and the GRL recommendations summarized in table 14 leads to the following conclusions:

  • The recommended soil shaft quake values appear to be consistent with the CAPWAP results obtained at the beginning of restriking for all the piles tested, with the exception of the PPI piles. The soil quake values obtained for the PPI piles are about 30 percent greater than those recommended by GRL. As soil shaft quake values are generally assumed to be independent of the pile type, the values obtained for PPI piles are most probably affected by local damages at the pile head observed during the pile driving.
  • For the soil type of the site, submerged soils, recommendations yield soil toe quake values of D/60 = 6.85 mm (0.27 inch), where D is the pile diameter. This value agrees with the CAPWAP analysis results obtained at the beginning of restriking for the reference steel pile. The soil toe quake values obtained for the Lancaster Composite, Inc., PPI, and SEAPILE piles are significantly lower (about 45 percent) than the values recommended by GRL.
  • Comparing the values of the soil toe quake obtained for all the test piles at the EOI with those obtained at the BOR, the dissipation of excess pore water pressure generated during pile driving resulted in a significant decrease (up to 80 percent) of the soil toe quake values.
  • The Smith shaft damping values recommended by GRL for the soil profile at this test site, which consists of cohesive and noncohesive soil layers, are within the range of 0.16 to 0.65 second (s)/m (0.05 to 0.20 s/ft). These values appear to be consistent with the CAPWAP results obtained at the EOI for all the piles tested. It should, however, be noted that the Smith shaft damping values obtained at the EOI for the FRP piles are about 250 percent greater than those obtained for the reference steel pile.
  • For all soil types, GRL recommendations yield a toe damping value of 0.49 s/m (0.15 s/ft). This value is consistent with the CAPWAP analysis results obtained at the beginning of restriking for the reference steel pile, indicating a toe damping value of 0.65 s/m (0.2 s/ft). The toe damping values obtained for the American Ecoboard, Lancaster Composite, Inc., PPI, and SEAPILE piles are consistently greater (about 250 percent) than the values recommended by GRL.
  • Comparing the values obtained for all the test piles at EOI with those obtained at the BOR, 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.

GRLWEAP Analysis

To evaluate the dynamic testing procedures used to obtain the dynamic elastic modulus of the FRP piles (i.e., PIT, the PDA testing—high-strain record, and the PDA testing considering the proportionality condition), a parametric study was conducted using the GRLWEAP Version 2002-1 program. GRLWEAP computes the number of blows per foot and the energy at the pile top. The parametric study consisted of calculating, for each FRP pile, the transfer energy obtained by PDA, the static capacity calculated by CAPWAP, and the number of blows per foot using the different elastic modulus values obtained with these testing procedures (table 15). The calculated transfer energy was compared to the PDA results, and the calculated blow counts were compared to the measured number of blows per inch at the beginning of restriking of each pile. CAPWAP results, including ultimate pile capacity, soil quake, and toe and shaft damping are used as GRLWEAP input data (table 13).

Table 15. Comparison of GRLWEAP results with measured elastic modulus, number of blows, and energy.
Pile Manufacturer Measured Elastic Modulus, kPa (ksi) Measured Number of Blows, blows/m (blows/ft) Calculated Number of Blows (GRLWEAP), kN/m (kips/ft) Measured Energy (PDA), kN/m (kips/ft) Calculated Energy (GRLWEAP), kN/m (kips/ft)
SEAPILE PIT—low strain 6.5 x 106 (943) 515.1 (157) 18427.2 (1263) 160.5 (11) 145.9 (10)
PDA—high strain 5.5 x 106 (802) 4537.5 (311) 145.9 (10)
PDA—early portion of high-strain records 8 x 106 (1165) 2290.6 (157) 145.9 (10)
PPI PIT—low strain 10 x 106 (1455) 872.7 (266) 21899.6 (1501) 116.7 (8) 116.7 (8)
PDA—high strain 9 x 106 (1307) 7893.2 (541) 116.7 (8)
PDA—early portion of high-strain records 11.9 x 106 (1719) 3881 (266) 116.7 (8)
American Ecoboard PIT—low strain 1.9 x 106 (270) 98.4 (30) 525.2 (36) 131.3 (9) 189.7 (13)
PDA - high strain 1.5 x 106 (218) 481.5 (33) 189.7 (13)
PDA—early portion of high-strain records 2.7 x 106 (388) 437.7 (30) 189.7 (43)

The results of this parametric study (figures 74, 75, and 76) illustrate (for the same transfer energy level) the effect of the elastic modulus on the number of blows per foot for the American Ecoboard, PPI, and SEAPILE piles. The GRLWEAP calculations show that the elastic modulus obtained from PDA testing—considering the proportionality condition for the initial portion of the high-strain records—yields the best agreement 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. Figure 74 shows that for the American Ecoboard pile, the effect of Young's modulus on the blow count was relatively small. This may be because Young's modulus values measured by PDA and PIT tests for the American Ecoboard pile were relatively small, compared to those obtained for the PPI and SEAPILE pile materials.

Click for text description

Figure 74. Graph. American Ecoboard pile—blows per foot versus elastic modulus.

Click for text description

Figure 75. Graph. PPI pile—blows per foot versus elastic modulus.

Click for text description

Figure 76. Graph. SEAPILE pile—blows per foot versus elastic modulus.

In general, materials exhibit stiffer behavior as the rate of loading increases. Thus, when comparing dynamic test results with those obtained from static laboratory or field load tests, the dynamic modulus are generally greater than the static elastic modulus.

In table 16, the dynamic elastic modulus, calculated from PDA testing—early portion of the high-strain records, are compared with static elastic modulus for steel and FRP piles, including the PPI and SEAPILE piles, obtained from laboratory compression tests. For steel, which is a rigid material compared to the FRP materials, the difference between static and dynamic modulus is 3 percent. The ratio of dynamic-to-static elastic modulus for FRP materials is 1.64 and 2.38 for the PPI plastic with steel bars and SEAPILE plastic with fiberglass bars, respectively.

Table 16. Comparison, static and dynamic elastic modulus of SEAPILE, PPI, and steel piles.
Pile Manufacturer Static Elastic Modulus, kPa (ksi) Dynamic Elastic Modulus, kPa (ksi) Dynamic Elastic Modulus/Static Elastic Modulus
SEAPILE 3.38 x 106 (490) 8.03 x 106 (1165) 2.38
PPI 7.22 x 106 (1048) 1.185 x 107 (1719) 1.64
Steel 1.999 x 108 (29000) 2.068 x 108 (30000) 1.03

Static Load Test Results—Correlations with CAPWAP Analysis

Pile Capacity and Setup

One of the main objectives of the field demonstration project was to evaluate the dynamic method currently used for predicting the load-set response and ultimate static capacity of FRP piles. To simulate the field conditions as consistently as possible, the static load results were used to evaluate the maximum ratio of the toe bearing capacity to the shaft resistance as an input data for the CAPWAP analysis. The CAPWAP analysis results for the total static capacity, the shaft resistance, and the toe bearing capacity for all the test piles at the EOI and BOR are indicated in table 13.

The PDA and CAPWAP analysis results summarized in tables 12 and 13 show that all the piles experienced a long-term increase in pile capacity from the EOI to the BOR. The capacity gains ranged from 311.4 to 854.1 kN (70 to 192 kips) for the piles installed to 17.7-m (58-ft) penetrations. For most of the tested piles, the measured setup factor, defined as a ratio of the BOR to EOI capacities, is in excess of 2.0. For all the SLT piles, the CAPWAP capacities ranged from 938.6 to 1,267.7 kN (211 to 285 kips) during the BOR and from 186.8 to 556.0 kN (42 to 125 kips) at the EOI.

In general, the CAPWAP analysis showed that at the BOR for the piles driven to penetrations of 17.7 m (58 ft), most of the ultimate capacity was due to the shaft resistance, which ranged from 827.4 to 1,049.8 kN (184 to 236 kips). The toe resistance ranged from 97.9 to 164.6 kN (22 to 37 kips). For the American Ecoboard pile, which was driven to a penetration depth of 8.8 m (29 ft), the analysis yielded a shaft resistance of 102.3 kN (23 kips) and a toe resistance of 280.2 kN (63 kips). The Lancaster Composite, Inc., pile, which was driven to refusal at a depth of 26.5 m (87 ft), had a mobilized ultimate capacity of 2,090.7 kN (470 kips) at the end of redrive, with 1,387.8 kN (312 kips) in toe resistance.

Correlations of CAPWAP Analysis With Static Load Test Results

Figures 78 through 81 illustrate the comparison between the load-set curves calculated by the CAPWAP program for the BOR and the SLT results for the Lancaster Composite, Inc., PPI, SEAPILE, and American Ecoboard piles. Table 17 summarizes the main experimental and dynamic analysis results, including the maximum applied or ultimate loads and the corresponding pile top settlements. These settlements are also compared with the elastic settlements of the piles due to their elastic compression, which are calculated assuming that the Lancaster Composite, Inc., PPI, and SEAPILE piles are frictional piles. For frictional piles, assuming a constant load transfer rate along the pile, the settlement due to the elastic compression, Selastic, can be approximately calculated by the equation in figure 77.

Click for text description

Figure 77. Equation. Selastic.

Where:

Pmax = Maximum applied or ultimate load.

E = Elastic modulus of the pile.

A = Pile section area.

L = Pile length.

The elastic modulus values indicated in table 17 were obtained from laboratory static compression tests conducted on FRP pile samples, taking into consideration the initial quasilinear portion of the stress-strain curves.

Table 17. Comparison between CAPWAP analysis and static load test results.
Pile Manufacturer CAPWAP Static Load Test EA, kN (kips) Selastic, mm (inches)
Pmax, kN (kips) Slimit, mm (inches) Pmax, kN (kips) Slimit, mm (inches)
Lancaster Composite (end of first loading cycle) 1000 (224.8) 5 (0.2) 1000 (224.8) 7.5 (0.3) 4.1 x 106 (0.9 x 106) 2.3 (0.1)
PPI 980 (220.3) 12 (0.5) 1100 (247.3) 9.6 (0.4) 0.9 x 106 (0.2 x 106) 11 (0.4)
SEAPILE 900 (202.3) 15 (0.6) 800 (179.8) 9.8 (0.4) 0.745 x 106 (1.67 x 105) 16.8 (0.7)
American Ecoboard 382 (85.9) 30 (1.2) 600 (134.9) 93.0 (3.6) - -
400 (89.9) 40 (1.6)

The comparison between the CAPWAP analysis and the SLT results for the FRP piles leads to the following observations.

For the Lancaster Composite, Inc., SLT pile, the CAPWAP analysis yields a load-set curve, which is quite consistent with the results of the first static load cycle. However, it leads to underestimating the total bearing capacity of the pile reached in the second load cycle. The difference between the CAPWAP analysis and the SLT results is likely due to the very high blow counts observed during the restrike (approximately 1 blow/mm (32 blows/inch)). With such low sets, the pile displacement under the dynamic load appears to be too small to fully mobilize the soil resistance and to obtain reliable measurements of the velocity and the force waves as input data for the CAPWAP analysis.

Click for text description

Figure 78. Graph. SLT and CAPWAP analysis—Lancaster Composite, Inc., pile.

Click for text description

Figure 79. Graph. SLT and CAPWAP analysis—PPI pile.

For the PPI SLT pile, the CAPWAP analysis yields an ultimate capacity that corresponds to about 90 percent of the plunging failure load observed during the SLT. The predicted limit settlement at failure is about 25 percent higher than the measured settlement, indicating that the CAPWAP analysis underestimates the stiffness of the pile response to static loading. The calculated elastic settlement due to pile compression is quite close to the limit settlement calculated by CAPWAP and slightly exceeds the experimental value. This seems to indicate that for the PPI pile, due to its low stiffness compared with the Lancaster Composite, Inc., pile, the settlement at failure is mainly the result of the elastic compression of the pile during loading.

Click for text description

Figure 80. Graph. SLT and CAPWAP analysis—SEAPILE pile.

For the SEAPILE SLT pile, the CAPWAP analysis yields an ultimate capacity that corresponds to about 112.5 percent of the plunging failure load observed during the SLT. Similar to the PPI pile, the predicted limit settlement at failure is higher (about 25 percent under the experimental ultimate load) than the experimental settlement, indicating that the CAPWAP analysis underestimates the stiffness of the pile response to static loading. Consistent with the results obtained for the PPI pile, the calculated elastic settlement due to pile compression is quite close to the limit settlement calculated by CAPWAP and exceeds the experimental value. This seems to confirm that for the SEAPILE and PPI tested piles, the settlement at failure is mainly the result of the elastic compression of the pile during loading.

Click for text description

Figure 81. Graph. SLT and CAPWAP analysis—American Ecoboard pile.

For the American Ecoboard SLT pile, the CAPWAP analysis showed a good correlation with the SLT up to 2 cm (0.8 inch) settlement at a load of 25 t (5.6 kips). CAPWAP analysis to predict the behavior of piles during SLTs is limited to settlements that do not exceed approximately 5 cm (2 inches). This might explain the difference between the SLT behavior and CAPWAP prediction curve for the low stiffness American Ecoboard pile, which reached a settlement of 96 mm (3.8 inches) under the static loading of 600 kN (134.9 kips). In addition, the American Ecoboard pile was driven to the sandy layer to assess its behavior primarily as an end bearing pile. The experimental results illustrated that no distinct plunging occurred under the applied static loading. As indicated in table 13, for this pile, the CAPWAP analysis predicted a ratio of end bearing to shaft resistance of 270 percent, which is substantially higher than that predicted for all the FRP piles driven to the clay layer at the depth of about 20 m (65.6 ft).

Pile Tension and Tensile Stress During Driving

In situ evaluation of the integrity of FRP piles during driving requires: (1) reliable data monitoring for determining the compression and tension stresses in the pile, and (2) development of appropriate design criteria for the allowable dynamic stresses in the composite pile material and its basic components.

Figures 84 through 87 show typical pile stress versus penetration depth profiles obtained for the Lancaster Composite, Inc., PPI, SEAPILE, and American Ecoboard piles from the field PDA testing. The dynamic elastic modulus used for this analysis was determined from the proportional condition of the initial portion of the high-strain records. The results show that the calculated tensile stresses did not exceed 6,894.8 kPa (1 ksi); calculated compressive stresses at the pile top ranged from 4,136.9 to 15,857.9 kPa (0.6 to 2.3 ksi) during the driving of all FRP piles.

The design criteria for allowable compression and tension stresses in the FRP piles were primarily developed considering the equation of the axial force equilibrium for the composite material and assuming no delamination between its basic components (figure 82).

Click for text description

Figure 82. Equation. Applied axial force F.

Where:

F = The applied axial force.

At = Total pile cross section area.

Ar = Total cross section area of the reinforcement.

Ap = Total cross section area of the plastic or concrete materials.

st = Equivalent axial stress, tension or compression, acting on the pile section.

srallowable = Allowable axial stress, tension or compression, acting in the pile reinforcements.

spallowable = Allowable axial stress, tension or compression, acting in the plastic material or the concrete for the Lancaster Composite, Inc., pile.

The equation in figure 82 can be solved for the equivalent allowable axial stress in the composite pile section (figure 83).

Click for text description

Figure 83. Equation. Equivalent axial stress st.

Table 18 summarizes the main criteria used for establishing the allowable stresses for the FRP pile installation. It also shows, for each pile, the actual stresses measured by PDA during pile driving.

Click for text description

Figure 84. Graph. Stress versus penetration depth for Lancaster Composite, Inc., SLT pile.

Click for text description

Figure 85. Graph. Stress versus penetration depth for PPI SLT pile.

Click for text description

Figure 86. Graph. Stress versus penetration depth for SEAPILE SLT pile.

Click for text description

Figure 87. Graph. Stress versus penetration depth for American Ecoboard splice SLT pile.

Table 18. Measured and allowable stresses for FRP piles.
Type of Stress Lancaster Composite Lancaster Composite; Driven to the Bedrock Layer Seaward International PPI American Ecoboard Steel Pile
Compressive stress, kPa (ksi) 85% of the concrete compressive strength 85% of the concrete compressive strength σmallowable=0
Ar/At=0.175
σr=1.81x105(26)
σmallowable=0
Ar/At 0.011
σr=2.89x105(42)
- 90% of yield strength for steel piles
Tension stress σmallowable=0
Ar/At=0.036

σr=3.47x105
(50.3)
σmallowable=0
Ar/At=0.036
σr=3.47x105 (50.3)
σmallowable=0
Ar/At=0.175
σr= NA
σmallowable=0
Ar/At=0.011
σr = 2.89x105
(42)
- 90% of yield strength for steel piles
Allowable compressive stress, kPa (ksi) 37233 (5.4) 37233 (5.4) 31027.5 (4.5) 3171.7 (0.46) NA -
Allowable tension stress, kPa (ksi) 12617.9 (1.83) 12617.9 (1.83) NA 3171.7 (0.46) NA -
Maximum measured compressive stress, kPa (ksi) 15858.5 (2.3) 19995.5 (2.9) 8274 (1.2) 7584.5 (1.1) 4895.5 (0.71) 183407 (26.6)
Maximum measured tension stress, kPa (ksi) 7998.2 (1.16) 5791.8 (0.84) 2827 (0.41) 2206.4 (0.32) 689.5 (0.1) 9790.9 (1.42)
Maximum measured compressive stress during restrike, kPa (ksi) 14824.3 (2.15) 16548 (2.4) 8549.8 (1.24) 8205.1 (1.19) - 238567 (34.6)
Maximum measured tension stress during restrike, kPa (ksi) 1448 (0.21) 4137 (0.60) 3378.6 (0.49) (0.18) - (9.2)

During pile driving, the measured compression and tension stresses did not exceed the allowable stresses, except for PPI pile. During and after pile driving and restriking of PPI and SEAPILE piles, no damage or separation between the bars and the recycled plastic material was observed, except at the upper foot of each PPI pile. In the case of the Lancaster Composite, Inc., piles, no damage or separation between the concrete and the FRP shell piles was observed (figures 68 through 71). Following these site observations, in the absence of experimental data, the maximum stresses obtained for the FRP piles can be used to establish allowable tension and compression stresses for pile driving.

<< Previous Main Next >>
 
FHWA-RD-04-107

Contact:

Updated: 04/07/2011
 

FHWA
United States Department of Transportation - Federal Highway Administration