Fender Pile Design
Fender Pile Design
Fender Pile Design
ENVIRONMENT
Christopher D. P. Baxter, Antonio Marinucci,
Aaron S. Bradshaw and Russell J. Morgan
University of Rhode Island
July 2005
URITC PROJECT NO. 536153
PREPARED FOR
UNIVERSITY OF RHODE ISLAND
TRANSPORTATION CENTER
DISCLAIMER
This report, prepared in cooperation with the University of Rhode Island
Transportation Center, does not constitute a standard, specification, or regulation.
The contents of this report reflect the views of the author(s) who is (are)
responsible for the facts and the accuracy of the data presented herein. This
document is disseminated under the sponsorship of the Department of
Transportation, University Transportation Centers Program, in the interest of
information exchange. The U.S. Government assumes no liability for the contents
or use thereof.
1. Report No
URITC FY01-03
N/A
N/A
5. Report Date
July 2005
6. Performing Organization Code
N/A
7. Authors(s)
N/A
10. Work Unit No. (TRAIS)
N/A
11. Contract or Grant No.
URI 536153
13. Type of Report and Period Covered
Final
14. Sponsoring Agency Code
Kingston, RI 02881
15. Supplementary Notes
N/A
16. Abstract
The regional state-of-the-practice for the construction of pile foundations, fender systems and earth
retention systems in the marine environment is to use materials such as timber, steel and concrete. These
materials are highly susceptible to attack by marine borers, corrosion, and decay. A possible alternative to
traditional piling systems is the use of composite piles constructed of fiber-reinforced polymers (FRP) or highdensity polyethylene (HDPE). Composite piles have advantages over traditional piles including complete
resistance to marine borer attack and corrosion. The primary objectives of this research were to improve the
understanding of the performance of composite piles as a fendering system in the marine environment (1)
during installation and (2) during normal fendering conditions. Technical issues studied include the short-term
stresses generated during hard driving conditions and short-term behavior and displacement due to lateral
impact loading. This was accomplished through two separate field studies in which concrete-filled FRP pipe
piles and steel reinforced plastic piles were installed at a residential site in Old Greenwich, CT and along a pier
at Fort Wetherill in Jamestown, RI. The piles in Old Greenwich were driven to failure with a hydraulic
hammer and PDA and CAPWAP analyses were performed. The piles at Fort Wetherill were impacted with an
85 ton vessel at low speeds and the dynamic response of the piles was measured using accelerometers and
displacement transducers. The results of this research provide useful field data for designers and researchers
who want to evaluate the effectiveness of composite piles as fender piles.
17. Key Words
Kingston, RI 02881
19. Security Classif. (of this report)
Unclassified
Unclassified
Form DOT F 1700.7 (8-72) Reproduction of completed page authorized (art. 5/94)
ii
68
22. Price
N/A
Table of Contents
1.0 Introduction..............................................................................................................1
2.0 Background on the Types of Composite Piling .......................................................2
2.1 Types of Composite Materials .....................................................................2
2.1.1 Steel pipe core pile........................................................................2
2.1.2 Structurally Reinforced Plastic (SRP) Pile ...................................3
2.1.3 Concrete-filled Fiberglass Pipe (FRP) Pile...................................3
2.1.4 Fiberglass Pultruded Pile ..............................................................3
2.1.5 Fiber Reinforced Plastic Piling (Plastic Lumber) .........................3
2.2 Applications .................................................................................................5
2.2.1 Foundation Systems ......................................................................5
2.2.2 Marine Piling ................................................................................5
2.3 Advantages and Disadvantages for Marine Applications............................5
2.3.1 Interface Bond Effects ..................................................................6
2.3.2 Drivability .....................................................................................6
3.0 Analysis of Driving Stresses and Pile Integrity During Hard Driving ....................8
3.1 Background ..................................................................................................8
3.2 Site Conditions.............................................................................................8
3.3 Description of Test Piles............................................................................10
3.4 Field Testing Methodology........................................................................10
3.4.1 Pile Integrity Testing (PIT).........................................................10
3.4.2 Pile Driving Equipment ..............................................................11
3.4.3 Dynamic Pile Testing..................................................................11
3.5 Pile Driving Results ...................................................................................11
3.5.1 Observations during Pile Installation..........................................11
3.5.2 Instrumentation Problems during Installation.............................13
3.5.3 Pile Integrity................................................................................14
3.5.4 Hammer Efficiency.....................................................................15
3.5.5 Pile Driving Stresses ...................................................................15
3.5.6 Dynamic Pile Capacities/Parameters ..........................................15
4.0 Field Impact Tests of Composite Fender Piles ......................................................17
4.1 Background ................................................................................................17
4.2 Site Conditions...........................................................................................18
4.3 Pile Installation ..........................................................................................19
4.4 Field Testing Methodology........................................................................21
4.4.1 Test Configuration ......................................................................21
4.4.2 Instrumentation ...........................................................................23
4.5 Data Processing..........................................................................................24
4.6 Dynamic Response of Impact Tests...........................................................27
4.6.1 Corrections to Accelerometer Data.............................................29
4.6.2 Impact Energy and Evaluation of Fender Pile Stiffness .............30
5.0 A New Dynamic Model for Analysis and Design of Fender Piles ........................32
5.1 Description of the Dynamic Model............................................................32
5.1.1 Mass Matrix ................................................................................36
5.1.2 Stiffness Matrix...........................................................................36
iii
5.1.3 Damping......................................................................................38
5.2 Comparison of Model and Field Impact Test Results ...............................39
5.2.1 Test Configuration and Results...................................................39
5.2.2 Mass Parameters .........................................................................40
5.2.3 Stiffness Parameters....................................................................40
5.2.4 Damping Parameters...................................................................42
5.2.5 Comparison of Field Test and Model Data.................................42
6.0 Conclusions............................................................................................................45
7.0 Acknowledgements................................................................................................46
8.0 References ..............................................................................................................47
Appendix A: MATLAB code ......................................................................................49
Appendix B: Acceleration and Displacement Results .................................................57
iv
List of Tables
Table 2.1. Common types of composite piling (after Iskander and Hassan 1998)........2
Table 2.2. Composite piles used in foundation applications. ........................................5
Table 2.3. Composite piles used in marine applications................................................5
Table 3.1. Summary of test pile properties. .................................................................10
Table 3.2. CAPWAP results for the PPI pile. ..............................................................16
Table 4.1. Refusal depths determined by Split-spoon Sample Tests...........................18
Table 4.2. Maximum displacements for FRP and PPI pile tests for each
accelerometer (see Figure 4.7 for location). ....................................................27
Table 4.3. Summary of average vessel and fender pile parameters.............................31
Table 5.1. Summary of maximum lateral pile displacements (in cm)
recorded during the impact tests. .....................................................................40
Table 5.2. Summary of velocities measured at the vessel impact point (ACC 2).
The damping ratio was calculated from the velocities using Equation (5-18).40
List of Figures
Figure 2.1. Steel core piling manufactured by Plastic Pilings, Inc. ...............................3
Figure 2.2. a.) Steel reinforced plastic pile (Plastic Pilings, Inc.) and
b.) Concrete-filled FRP pipe pile (Lancaster Composite). ................................4
Figure 2.3. Different composite pile types (from Iskander and Stachula 2001)............4
Figure 2.4. Delamination of a concrete-filled FRP pipe pile during a four point
bending test (Lampo et al. 1998). ......................................................................7
Figure 3.1. Site location. ................................................................................................9
Figure 3.2. Profile of Boring No.6 (obtained from Heller & Johnson). ........................9
Figure 3.3. Michael Sutyla of Heller and Johnsen performing PIT.............................10
Figure 3.4. a.) Deformed pile top and b.) exposed steel at the pile tip due to
hard driving conditions. ...................................................................................12
Figure 3.5. a.) Setup and installation of FRP pile, and b.) Damage to the pile
top after driving................................................................................................13
Figure 3.6. PIT plots of the FRP pile (a) before driving and (b) after driving. ...........14
Figure 3.7. PIT plots of the PPI pile (a) before driving and (b) after driving..............15
Figure 3.8. Pile driving results from the PPI pile. .......................................................16
Figure 4.1. Single fender pile system modeled by Maher et al. (1996).......................17
Figure 4.2. a.) Location of test site and b.) layout of test piles....................................18
Figure 4.3. Sediment sampling at Fort Wetherill using the
URI Large-Diameter Gravity Corer.................................................................19
Figure 4.4. Predriving using a Greenheart timber pile.................................................20
Figure 4.5. Use of W8x28 section to prevent bending of FRP pile
during handling. ...............................................................................................20
Figure 4.6. Damage to the a.) PPI pile and b.) FRP pile during installation. ..............21
Figure 4.7. Cross-section of test layout, including location of instrumentation. .........22
Figure 4.8. Setup for impact load tests. .......................................................................22
Figure 4.9. a.) Kistler 8305A differential (top) and 8310A
single-ended (bottom) accelerometers, and b.) mounting an
accelerometer on the pile (Kistler, 2004).........................................................23
Figure 4.10. Massa M-5000 Smart Ultrasonic Sensor (Massa, 1998). ........................23
Figure 411. Filtered and unfiltered acceleration signals from FRP Pile
impact test 4. ....................................................................................................25
Figure 4.12. Filtered acceleration and displacement graphs for FRP pile
impact Test 4....................................................................................................26
Figure 4.13. Acceleration, velocity and displacement response for FRP
pile impact test 4. .............................................................................................26
Figure 4.14. Acceleration and displacement response of the FRP pile
(a and b) and the plastic pile (c and d) during impact......................................28
Figure 4.15. Acceleration signal (top) and correction function (bottom). ...................29
Figure 4.16. Corrected velocity (top) and displacement (bottom) time histories. .......30
Figure 5.1. Schematic of the dynamics of a vessel impacting a fender pile:
(a) prior to impact, (b) at time of impact, (c) after impact, and
(d) after decoupling..........................................................................................33
Figure 5.2. Schematic of the dynamic fender pile model. ...........................................33
vi
vii
1.0 Introduction
The regional state-of-the-practice for the construction of pile foundations, fender systems
and earth retention systems in the marine environment is to use materials such as timber, steel
and/or concrete. These materials are highly susceptible to attack by marine borers, corrosion,
and decay. The recent environmental improvement of Americas harbors has actually
accelerated the damage done to timber piles by improving conditions for species of marine
borers such as Limnoria and Tordo Novalis. Most timber piles are chemically treated with
creosote or copper-chrome-arsenic (CCA) to resist such attack, but these chemicals themselves
can pollute the environment and harm marine life (Iskander and Hassan, 1998; Iskander and
Stachula, 1999).
A possible alternative to traditional piling systems is the use of composite piles
constructed of fiber-reinforced polymers (FRP) or high-density polyethylene (HDPE). Two
common configurations for these piles are a FRP pipe pile filled with concrete or a recycled
plastic pile reinforced with steel or fiberglass. A recent analytical study on the drivability of
composite piles concluded that several technical issues must be overcome before these piles are
widely used in practice (Iskander et al. 2001). These issues included the instrumentation,
installation, and loading of composite piles under field conditions, analysis of driving stresses,
and the durability of composite piles in the field. Presently, there are no well-documented field
studies in the literature of the dynamic response of composite fender pile systems.
The primary objectives of this research study were to improve the understanding of the
performance of composite piles as a fendering system in the marine environment. Specific areas
of study included an evaluation of driving stresses during hard driving conditions (i.e. bedrock)
and the response of composite piles to lateral impact loads.
This was accomplished through separate field studies in which concrete-filled FRP pipe
piles and steel reinforced composite piles were installed at a residential site in Old Greenwich,
CT and along a pier at Fort Wetherill in Jamestown, RI. The piles in Old Greenwich were driven
to failure with a hydraulic hammer and PDA and CAPWAP analyses were performed. The piles
at Fort Wetherill were impacted with an 85 ton vessel at low speeds and the dynamic response of
the piles was measured using accelerometers and displacement transducers. This data was used
to develop a dynamic approach to the analysis of fender piles where the impacting vessel
coupled with the fender pile is modeled as a freely vibrating, multi-degree of freedom structure
with lumped masses, stiffness, and damping. The response of the coupled vessel and fender pile
was evaluated using a modal approach along with analytical solutions to the equation of motion.
Mass, stiffness, and damping parameters were derived for the fender pile system.
This report is divided into six chapters. Because these piles have not been used
extensively in New England for marine applications, chapter two presents a review of the
literature on composite piles. This includes the different types of composite piles that are
commercially available as well as their potential advantages and disadvantages.
The evaluation of the stresses induced during hard driving conditions is presented in
Chapter 3, and Chapter 4 presents the results of the field impact load tests. Chapter 5 presents the
new dynamic model for fender pile analysis and design, and a summary of the results are
presented in Chapter 6.
Manufacturer
Description
Structurally Reinforced
Plastic Pile
Concrete-filled Fiber Reinforced Pipe (FRP) Pile
Fiberglass Pultruded Pile
U.S. Plastics
Steel Core
HDPE Plastic
Figure 2.1. Steel core piling manufactured by Plastic Pilings, Inc.
2.1.2 Structurally Reinforced Plastic (SRP) Pile
Structurally reinforced plastic (SRP) piles consist of HDPE plastic reinforced with either
fiberglass rods or steel rebar (Iskander and Stachula 1999), as shown in Figure 2.2a. The outer
surface of SRP piles is typically treated to retard UV degradation. SRP piles are available in diameters between 10 and 17 inches and are reinforced with 6 to 16 rods or rebar, with diameters
ranging from 1 to 1.4 inches. These piles can be produced in a variety of lengths.
2.1.3 Concrete-filled Fiberglass Pipe (FRP) Pile
Concrete-filled fiberglass pipe (FRP) piles, as the designation implies, are pipe piles
comprised of a fiberglass shell and concrete infill, as shown in Figure 2.2b. The fiberglass shell
provides the pile's tensile strength while the concrete infill provides flexural rigidity and resistance to buckling. The outer coating retards UV degradation and protects against chemical damage and abrasion. Hardcore piles are filled with concrete after installation, whereas Lancaster
Composite piles are filled with concrete prior to installation. These piles are available in a variety of lengths with outside diameters ranging from 8 to 18 inches.
2.1.4 Fiberglass Pultruded Pile
Fiberglass pultruded piles are comprised of a fiberglass pipe pile, fitted with fiberglass
grid inserts. Figure 2.3 presents a typical cross section for the pultruded pile along with other
composite pile types. In fendering applications, the HDPE shell and fiberglass inserts are used,
among other things, to absorb vessel impact (Iskander and Stachula 1999).
2.1.5 Fiber Reinforced Plastic Piling (Plastic Lumber)
Plastic Lumber is comprised of recycled plastic and fiberglass reinforcement (Iskander
and Stachula 1999). The outer portion of the pile's cross-section is dense and solid, whereas the
inner portion of the cross-section is foam-filled to reduce the pile's weight. These piles are typically 10 inches in diameter and 25 feet in length.
3
(a)
(b)
Figure 2.2. a.) Steel reinforced plastic pile (Plastic Pilings, Inc.) and b.) Concrete-filled FRP pipe
pile (Lancaster Composite).
Figure 2.3. Different composite pile types (from Iskander and Stachula 2001).
4
2.2
Applications
2.2.1
Foundation Systems
According to published literature, composite piling has been rarely used in structural
foundation applications. Table 2.2 provides a select listing in which composite piles have been
installed and tested under typical foundation conditions.
Table 2.2. Composite piles used in foundation applications.
Composite Pile Type
Concrete-filled FRP Pile
Hardcore Fiberglass shell;
concrete infill
Concrete-filled FRP Pile
Structurally Reinforced
Plastic (SRP) Piles
Concrete-filled FRP Pile
Application
Studied Driving Stresses
and Bearing Capacity
Dynamic Analysis and
Driving Cond. Study
Dynamic Analysis and
Driving Cond. Study
Dynamic Analysis and
Driving Cond. Study
Bridge Bent Foundation
Support
Location
Asbury Park, New Jersey
New Castle, Delaware
Reference
Goble et al. 2000
Kozera 1997
2.2.2
Marine Piling
Unlike the limited use in structural foundation systems, composite piling has been used
increasingly in marine applications, most notably in fender systems/piling. Table 2.3 provides a
select listing in which composite piles have been installed in marine applications.
Table 2.3. Composite piles used in marine applications.
Composite Pile Type
Steel Pipe Core Piling
Application
Anchor Floating Docks
Location
Ferry Docks, Newport, RI
Reference
N/A
Fender Piling
Fender Piling
Fender Piling
Fender Piling
2.3
When used in fendering applications, some composite pile types have been reported
to absorb as much as 40 times more energy than traditional timber piles.
Conversely, composite piles also have inherent disadvantages, including the following
(from Iskander and Stachula 1999, 2001; Iskander and Hassan 1998; Lampo et al. 1998):
The initial cost of composite piles is two to three times more expensive.
Due to their low stiffness, composite pile installation is typically less efficient and
more difficult than traditional piling, as explained in a subsequent section.
Composite piles having low stiffness can cause handling and installation problems.
Interface bonding and delamination is an ongoing concern, as explained in a subsequent section.
Another potential disadvantage is illustrated by the Tiffany State Pier case study referenced in Table 2.3. In 1996, the pier was destroyed by a major fire. The high density polyethylene piles were severely damaged by the fire, and the pier was closed the following year. This
illustrates the need for adequate fire protection of composite pile materials.
2.3.1 Interface Bond Effects
As reported in numerous published works, ineffective interface bonding between the different materials (e.g. steel and plastic, FRP and concrete) in composite piles was a significant
problem in early designs (e.g. Iskander and Stachula 1999; Iskander and Hassan 1998). Figure
2.4 provides an example of interface delamination between the concrete core and the FRP shell
during a flexure test. The concrete-filled fiberglass pipe pile was tested in four point bending
and, during the test, the squeezing out of concrete occurred. This delamination clearly indicates
debonding at the interface between the FRP shell and the concrete core. In the recent past, it has
been reported by various authors referenced throughout this text that manufacturers have altered
the fabrication process to minimize the occurrence of delamination. For example, in some instances, the inside surface of the FRP shell is roughened to improve the mechanical interface
bond. In addition, bonding agents have been used on the inside surface of the shell prior to infilling with concrete. Expansive concrete has also been used in the core (Rizkalla and Fam 1999).
In concretefilled FRP piles, the concrete core must be well connected to the shell material. This is accomplished by using a non-smooth FRP interface surface or by using bonding
agents. If there is insufficient interface bond between the shell and the core, the concrete fill will
delaminate from the composite shell, thereby resulting in independent material behavior rather
than composite.
2.3.2 Drivability
Composite materials, in general, have been reported to have a higher damping coefficient
and lower stiffness than traditional materials. Highly damped, low stiffness piles are more difficult to drive due to the difficulty in transferring driving energy to the pile. Iskander and Hassan
(1998) reported that the modulus of elasticity, (typically not provided by the manufacturer) and
the pile's specific weight has a profound influence on the drivability of HDPE reinforced piles.
Conversely, concrete-filled FRP pipe piles have been reported to have a higher stiffness and a
low damping coefficient due to its concrete core. According to Iskander and Hassan (1998),
FRP piles are easier to drive and have an overall driving behavior similar to purely concrete piles
(Iskander and Hassan 1998). However, there is limited published information pertaining to
composite pile drivability.
Figure 2.4. Delamination of a concrete-filled FRP pipe pile during a four point bending test
(Lampo et al. 1998).
3.0 Analysis of Driving Stresses and Pile Integrity During Hard Driving
3.1 Background
Numerous studies have been performed on the drivability of composite piles. However,
many of these studies are theoretical in nature and do not evaluate actual pile driving in the field.
For example, Iskander et al. (2001) and Iskander and Stachula (2002) used a wave equation
model (WEAP) to study the influence of the modulus of elasticity, damping, and unit weight on
the drivability of three types of composite piles. These studies report that driving resistance
(hammer blows) decreases appreciably with either a decrease in (1) the pile's elastic modulus or
(2) the pile's unit weight, especially below a unit weight of about 110 pcf. Conversely, damping
was shown to have a negligible effect on drivability.
Iskander et al. (2001) used WEAP to compare the drivability of short (60 ft), low capacity piles and long (90 ft), high capacity composite piles in a typical marine soil profile. The results indicate that the drivability of reinforced plastic (plastic lumber) piles, concrete-filled fiberglass pipe (FRP) piles, and timber piles was not a problem for the short, low capacity piles.
However, the drivability (i.e. ease of installation) of these pile types is very different for the long,
high capacity piles.
Iskander and Stachula (2002) back evaluated WEAP parameters (modulus of elasticity,
damping and unit weight) by matching the results to PDA results obtained during driving of the
plastic lumber and the FRP piles. Based on this analysis the authors recommend the following
parameters for the plastic lumber piles: an elastic modulus equal to 2/3 of the manufacturer's reported composite modulus, the manufacturer's reported unit weight, and a pile damping factor of
9. Typical WEAP parameters published for traditional prestressed concrete piling provided a
good match to measured results for the FRP pile.
Wave equation analyses such as PDA, CAPWAP and WEAP have been used in practice
for the design of composite piles (Kesavanathan and Kozera, 1997; Goble 2000). However,
there is limited data supporting the reliability of these methods to model the non-linear behavior
of composite piles. Clearly, there is a need for continued study of pile driving in the field.
This chapter presents the results of pile driving on two types of composite piles in Old
Greenwich, Connecticut. The piles were driven to failure to measure the driving stresses and
evaluate methods for detecting damage. An expanded treatment of this portion of the study can
be found in Gummert (2003).
3.2 Site Conditions
The composite piles used in this study were installed using an impact hammer at an existing residential construction site in Old Greenwich, Connecticut, shown in Figure 3.1. Due to the
proximity to the waterfront, the high groundwater table and soft soils, timber piles were installed
to support the structure. Since the pile driving equipment was already on site, the project provided a unique opportunity to drive composite piling for this study.
The subsurface conditions were evaluated from 6 borings previously performed at the site
by Heller and Johnsen. A typical boring log (No. HJ-6) from the site is shown in Figure 3.2.
The subsurface conditions consist of alternating layers of silt and sand overlying bedrock. Bedrock was encountered at various depths across the site ranging from 13.5 to 16.5 feet, and the
upper 1.5 feet of the bedrock was in a weathered condition. The composite test piles used in this
study were installed adjacent to Boring No. HJ-6.
Interstate I- 95
Riverside
Old
Greenwich
Site
Figure 3.2. Profile of Boring No.6 (obtained from Heller & Johnson).
9
Pile Type
PPI
FRP
Notes:
1. Values reported by Manufacturer.
11
(a)
(b)
Figure 3.4. a.) Deformed pile top and b.) exposed steel at the pile tip due to
hard driving conditions.
The concrete-filled fiberglass pipe (FRP) pile was then instrumented and installed. However, there were problems attaching the PDA accelerometers and strain gages, as will be discussed in section 3.4.2. Prior to driving, PIT was performed to verify the integrity of the concrete core. A four-inch thick plywood cushion was used to prevent the ram from damaging the
top of the pile. In order to avoid the same problems encountered when driving the PPI pile, the
FRP was installed away from the PPI pile location.
The first 7 ft of the FRP pile pushed into the soil under the combined self-weight and
driving equipment weight. At an embedment of approximately 8 ft, the driving commenced us-
12
ing a rated energy of 20,000 ft-lb. By a depth of approximately 9 ft, the blow counts exceeded
20 blows/ft. Figure 3.5a shows the installation setup for the FRP pile. Soon after driving commenced, problems with the PDA instrumentation were realized; however, this is explained in
greater detail in a subsequent section. Due to the instrumentation problems and the ineffectiveness of the PDA, the pile was extracted to determine whether the pile tip was damaged. After
inspecting the pile and finding no appreciable damage, the pile was reattached and redriven. At
approximately 12 ft, the energy of the hammer was increased to 52,000 ft-lb. Due to this high
energy, the pile cushion broke and the concrete core at the pile top began to crack and spall. At
approximately 12.75 ft of embedment, with blow counts equal to 48 blows/ft, driving was
stopped in order to extract and inspect the pile for damage. Unfortunately, extraction efforts
were unsuccessful. Upon inspection, the pile top was visibly broken, as shown in Figure 3.5b.
Since the pile tip could not be inspected, the FRP pile was cut approximately 11.4 ft from the top
and PIT was performed on the embedded 12.35 ft portion of the pile to assess pile damage.
However, when the FRP pile was extracted after the initial driving in order to reattach the PDA
gages, the pile tip did not show appreciable damage.
(a)
(b)
Figure 3.5. a.) Setup and installation of FRP pile, and b.) Damage to the pile top after driving.
3.5.2 Instrumentation Problems during Installation
As a requisite part of the Pile Driving Analyzer (PDA) setup, the composite piles were
instrumented with two acceleration transducers and two strain gages. GZA Geoenvironmental,
Inc. in Norwood, Massachusetts, installed the devices and performed the dynamic testing. The
acceleration transducers and strain gages used in conjunction with the PDA were attached approximately 5.0 feet below the top of each composite pile.
Problems with the instrumentation were realized during the installation of the devices and
during the installation of the FRP pile itself. The first concern pertained to the location of the
devices. In order to obtain accurate dynamic measurements, the instrumentation needs to be in-
13
stalled as close to the top of the pile as practical. Therefore, the instrumentation was installed
approximately 5.0 feet below the top of the pile. However, at this position, the instrumentation
was located approximately 4 inches below the FRP shell top. Unfortunately, the instrumentation
was attached at the end of adhesive/binding material that is used to create the bond between the
FRP shell and the concrete core. During pile installation, the instrumentation became detached,
and, as a result, the PDA did not record any measurements. The instrumentation was attached at
a new location on the same pile. However, shortly after commencement of pile driving, the instrumentation once again became detached and no measurements were obtained. After analyzing
the problem with the instrumentation attachment to the composite pile, it was assumed that the
anchors attaching the devices could not expand properly within the concrete core, and therefore
kept detaching. No PDA measurements were obtained for the FRP pile.
3.5.3 Pile Integrity
The results from the PIT performed on the FRP pile is shown in Figure 3.6 for conditions
(a) before driving and (b) after driving. The signals obtained through the FRP pile were relatively easy to obtain. The wave trace before driving shows a distinct return signal from the pile
toe as expected, along with a reflection approximately halfway down the pile. It is anticipated
that the latter reflection was attributed to a crack that may have been made during transport or
handling of the pile. The calculated wave speed of 14,000 ft/sec is slightly higher than published
values of approximately 10,000 ft/sec for concrete (Kindsler et al. 1982). The PIT results obtained after driving shows a highly irregular wave trace suggesting significant cracking and damage to the pile from driving. This is consistent with the very hard driving that the piles were subjected to at the end of installation, as explained in a previous section.
(a)
(b)
Figure 3.6. PIT plots of the FRP pile (a) before driving and (b) after driving.
Unlike the FRP pile, signals from the PPI pile were difficult to obtain. The best traces for
the PPI pile are shown in Figure 3.7 for conditions (a) before driving and (b) after driving. Both
traces show a large reflection at the pile toe and at about 5 ft and 6 ft below the pile top. It is difficult to ascertain if the latter reflection is attributed to air bubbles that are typically found within
the plastic matrix or to a defect of concern. The calculated wave speed of 10,500 ft/sec is similar
to the published values for concrete. The PIT traces obtained before and after pile driving are
very similar, suggesting that driving did not cause fracturing or delamination of the pile.
14
Transmitted signal
Received signal
Reflections
(a)
(b)
Figure 3.7. PIT plots of the PPI pile (a) before driving and (b) after driving.
3.5.4
Hammer Efficiency
Hammer efficiency is a measure of the percentage of the rated hammer energy that is
transferred into the pile. The energy transfer characteristics are therefore related to the type of
hammer and type of pile. Typical efficiency values range from 25% for a diesel hammer on concrete/timber piles to 50% for a single-acting air/steam hammer on steel piles (FHWA 1998). The
efficiency of the hydraulic hammer, Junttan HHK6, on the PPI pile was estimated. No pile cushion was used during driving. Throughout driving the hammer maintained a stroke of 1.5 feet.
The rated energy of 19.85 kip-ft was calculated from product of the hammer weight (13.23 kips)
and measured stroke (1.5 ft). The average transferred energy, measured with the PDA over a
depth of 4 to 6 feet, was 10.28 kip-ft yielding an average efficiency of 52%, standard deviation
of 10%.
3.5.5
15
PDA Capacity
Blow Count
Depth (ft)
10
100
1000
200
400
Shaft
(kips)
(kips)
(kips)
(in)
(in)
(kip-sec/ft)
(kip-sec/ft)
300
303
0.1
0.56
0.09
0.008
16
Figure 4.1. Single fender pile system modeled by Maher et al. (1996).
17
Site Location
(a)
(b)
Figure 4.2. a.) Location of test site and b.) layout of test piles.
A site investigation was performed in the small inlet adjacent to the pier to determine the
depth of rock and to the types of soils present at the site. Two large-diameter gravity (LGC)
cores and three split-spoon samples (SS) were obtained. The depth of refusal/bedrock was estimated from the sampling and testing, as shown in Table 4.1. The URI LGC consists of a sampling tube with a driving weight and stabilizing fins at the top, as shown in Figure 4.3. This assembly is lowered on a cable until it is approximately 10 ft (3m) above the sediment surface.
Once in position, the LGC free falls into the sediment. Once the sample is obtained, a check
valve at the top of the apparatus and a special core catcher in the bottom of the sampler hold the
sediment in the tube. The LGC apparatus can obtain 4-inch (10.2cm) diameter samples in a PVC
or steel core barrel up to 10ft (3m) in length. The URI research vessel, CT-1, was used to perform both the LGC and SS tests in the marina sediments.
Table 4.1. Refusal depths determined by Split-spoon Sample Tests
Location
Inner Bay
Corner of Pier
Outer Bay
Water Depth
9.6ft (2.92m)
6.3ft (1.91m)
12ft (3.67m)
18
Refusal Depth
9.1ft (2.78m)
7.9ft (2.41m)
11.4ft (3.48m)
19
Figure 4.5. Use of W8x28 section to prevent bending of FRP pile during handling.
Upon visual inspection of the piles after installation was complete, damage to the top portion of the piles was clearly noticeable. The PPI pile, as shown in Figure 4.6a, experienced damage due to friction heating caused by the clamping of the vibratory hammer to the pile itself. The
FRP pile, as shown in Figure 4.6b, experienced damage to the fiberglass exterior due to friction
and slippage caused by the clamping of the vibratory hammer during installation.
20
(a)
(b)
Figure 4.6. Damage to the a.) PPI pile and b.) FRP pile during installation.
4.4 Field Testing Methodology
4.4.1 Test Configuration
The fender piles were laterally impacted using a ship called the Beavertail from Jamestown, RI. The vessel, which was built by the U.S. Army in 1940 for mine distribution, has an
overall length of 64ft (19.51m), a beam of 18ft (5.49m), and a draft of 7ft (2.13m). The vessel
was originally designed for a displacement of 68 tons (605 kN), however, the current displacement is larger considering recent installation of larger fuel tanks and ballast. The estimated displacement is approximately 85 tons (756 kN) (Fred Pease, URI Ocean Engineering ship captain
and former owner of the Beavertail, personal communication). The impact loads were calculated
using a kinetic energy formulation based on the estimated mass and approach velocity of the vessel.
The piles were instrumented with acceleration and displacement transducers to measure
their dynamic response to impact, as shown in Figure 4.7. The displacements at four locations
along the pile were determined by double integrating the obtained acceleration signals. The accuracy of the integration was assessed at one location by comparing the integrated accelerometer
displacements with displacement measurements obtained by a separate displacement transducer.
In order to perform the impact test, the boat was oriented perpendicular to the test pile, as
shown in Figure 4.8. Dock lines were attached to an adjacent pier to maintain the orientation of
the boat such that its stern would strike the pile head-on. At approximately 10 to 15 ft from the
pile, the boat was put in gear and accelerated to a velocity of about 0.4 to 0.8 ft/sec (0.13 to 0.25
m/sec). A few feet before impact, the boat was put into neutral allowing the boat to drift into the
fender pile. During impact, the acceleration time history was recorded using acceleration transducers mounted at four locations along the pile. A total of 6 impact tests were performed on the
PPI piles and 9 tests were performed on the FRP piles.
21
Z1
Z2
Z3
Z4
Z5
Z6
Z7
Z8
Description
Deck to Displacement Transducer
Deck to Accelerometer 1
Deck to Accelerometer 2
Deck to Impact Location
Deck to Water Level
Deck to Accelerometer 3
Deck to Accelerometer 4
Embedment Length
FRP Pile
3'-4"
3'-6"
4'-7"
3'-9"
6'-8"
7'-3"
13'-10"
19'-0"
PPI Pile
3'-7"
3'-6"
4'-9"
3'-7"
6'-8"
7'-2"
13'10"
19'-0"
22
4.4.2 Instrumentation
Two different types of accelerometers and one acoustic displacement transducer were
used in this study. The accelerometers, shown in Figure 4.9, are manufactured by Kistler Instrument Corporation (models 8305A and 8310A). The acceleration transducers were fastened
to the face of the composite piles at various vertical positions to measure the dynamic response
of the composite piles.
(a)
(b)
Figure 4.9. a.) Kistler 8305A differential (top) and 8310A single-ended (bottom) accelerometers,
and b.) mounting an accelerometer on the pile (Kistler, 2004).
Prior to performing the field experiments, all four accelerometers were calibrated separately employing the same procedure. Each accelerometer was placed in three orientations that
corresponded to three different values of acceleration. The positive (horizontal, face up) position
corresponds to an acceleration of 1 g. The vertical position and the negative (horizontal, face
down) position correspond to accelerations of 0 and 1 g.
Displacements were measured at one point on the piles using a Massa M-5000 Smart Ultrasonic Displacement Transducer, as shown in Figure 4.10. The transducer was mounted on a
tripod and positioned in close vertical proximity to the uppermost acceleration transducer (Figure
4.7) in order to provide a redundant measurement with one of the accelerometers.
Data was acquired using a laptop personal computer with a National Instruments data acquisition card and the program LabView6.
23
1
(at + at 1) + vt 1
2fs
(Eq. 4.1)
st =
1
(vt + vt 1) + st 1
2fs
(Eq. 4.2)
24
where:
at
at-1
fs
st
st-1
vt
vt-1
Figure 4.11. Filtered and unfiltered acceleration signals from FRP Pile impact test 4.
The displacements obtained from double integrating the acceleration signal from the uppermost acceleration transducer were then compared to the displacements obtained from the displacement transducer. Figure 4.13 shows excellent agreement between the measured and calculated displacement for most of the time history. However, the calculated displacement incorrectly increases linearly at the end of each signal. This error is addressed in section 4.6.1.
25
Figure 4.12. Filtered acceleration and displacement graphs for FRP pile impact Test 4.
Fiber Reinforced Pile Test 4
10
Acceleration
ft/sec2
5
0
-5
-10
1
ft/sec
0.5
Velocity
0
-0.5
-1
0.2
Displacement
ft
0.1
0
-0.1
Acceleration Derived
Sonic Range Finder
-0.2
0
0.5
1.5
2
sec
2.5
3.5
Figure 4.13. Acceleration, velocity and displacement response for FRP pile impact test 4.
26
Test Number
FRP
FRP
FRP
FRP
FRP
FRP
FRP
FRP
FRP
PPI
PPI
PPI
PPI
PPI
PPI
Acc 1
(ft)
Acc 2
(ft)
Acc 3
(ft)
1
2
3
4
5
6
7
8
9
0.129
0.154
0.162
0.167
0.148
0.147
0.133
0.161
0.150
0.090
0.120
0.129
0.178
0.128
0.118
0.118
0.127
0.126
1
2
3
4
5
6
0.085
0.107
0.128
0.140
0.167
0.209
0.032
0.044
0.032
0.046
0.046
0.123
27
0.059
0.097
0.110
0.170
0.086
0.070
0.092
0.090
0.097
Acc 4
(ft)
0.002
0.029
0
0.029
NA
NA
NA
0
0.018
Sonic
(ft)
0.124
0.145
0.153
0.172
0.145
0.143
0.153
0.157
0.149
0.036
0.002
0.096
0.062
0.019
0.064
NA
0
0
0
0
0.001
0.119
0.127
0.139
0.153
0.198
0.181
(a)
(b)
(c)
(d)
Figure 4.14. Acceleration and displacement response of the FRP pile (a and b) and the plastic
pile (c and d) during impact.
28
0.5
1.5
2.5
3.5
0.5
1.5
2
Time (s)
2.5
3.5
0
-0.05
-0.1
-0.15
-0.2
29
0.3
0.2
0.1
0
-0.1
-0.2
0.5
1.5
2.5
3.5
0.5
1.5
2
Time (s)
2.5
3.5
0.15
0.1
0.05
0
-0.05
Figure 4.16. Corrected velocity (top) and displacement (bottom) time histories.
2
Vn
2g
where:
(Eq. 4.3)
E
g
Vn
=
=
=
=
As a first order approximation, additional vessel masses from hydrodynamic effects have
been neglected. The maximum reaction force on a perfectly elastic fender pile can be calculated
using Equation 4.4 (Gaythwaite 2004).
Rmax =
2E
(Eq. 4.4)
max
30
where:
Rmax
max
The equivalent spring constant, K, for the fender pile can be calculated using Equation 4.5.
R
K = max
(Eq. 4.5)
max
A summary of average values for vessel velocity, vessel energy, maximum displacement,
maximum reaction force, and pile stiffness are summarized in Table 4.3. In principle, these values of stiffness can be used to design future fender pile systems with the composite piles. However, the stiffness of the PPI pile system is too high because of the restraint caused by the stone
block described above, and that value is not representative of typical conditions.
Table 4.3. Summary of average vessel and fender pile parameters.
Pile
Type
FRP
PPI
Vn
(ft/sec)
0.56
0.69
(kip-ft)
0.83
1.26
max
(ft)
1.36
0.64
31
Rmax
(kip)
1.22
3.93
K
(kip/ft)
0.90
6.14
5.0 A New Dynamic Model for Analysis and Design of Fender Piles
The kinetic energy approach described in section 4.6.2 is the most commonly used
method in the design of fender systems (Gaithewaite 2004). In this approach, the kinetic energy
of a berthing vessel is compared to the energy absorbing capacity of the fender system, and the
fender elements are selected or configured to limit stresses below allowable criteria. The energy
absorbing capacity of a fender element is evaluated by calculating the area under the static force
verses displacement curve. Energy concepts have been adopted for the analysis of free-standing
fender piles (Reese et al. 1970), concrete fender piles (Li and Ramakrishnan 1971), and timber
fender piles (USACE 1983).
Since the energy capacity of fender piles are derived from static tests or analyses, the
kinetic energy method does not consider the energy dissipated during vessel impact. Energy
dissipation should be considered in design because it has the effect of reducing the forces on the
vessel and pile. The first half of this chapter presents a description of a dynamic model that is
derived for a typical fender pile configuration. In this model, the fender pile and impacting
vessel are treated as a freely vibrating multi-degree of freedom structure with lumped masses.
The second half of the chapter presents a comparison of the field impact tests performed on the
FRP composite fender pile and the results of the dynamic model.
5.1 Description of the Dynamic Model
The dynamics of a vessel impacting a fender pile are best illustrated by the simple model
shown in Figure 5.1. In this model the pile is represented by a frictionless mass (Mp) attached to
a spring and a dashpot and the approaching vessel is represented by a second frictionless mass
(Mv) having an initial velocity Vvo. The spring represents the stiffness and the dashpot represents
the energy dissipated in the system. Before impact (Figure 5.1a) the pile mass is in static
equilibrium with zero displacement and zero velocity. At impact (Figure 5.1b) the vessel mass
becomes coupled to the pile mass and as the spring is compressed the vessel mass is slowed
eventually stopping at the point of maximum displacement (Figure 5.1c). The strain energy
stored in the spring pushes the vessel back to the equilibrium point where it becomes decoupled
from the pile (Figure 5.1d). The velocity of the vessel at the point when it leaves the pile (Vvf) is
less than the impact velocity due to the dissipated energy.
The dynamic motion of the fender pile system is therefore analogous to a freely vibrating
system for -cycle which has zero initial displacement and an instantaneous velocity. For a
more accurate representation of the various parameters involved, a multi-degree of freedom
approach can be followed in which the structure is modeled as a system of lumped masses with
stiffness and damping. A schematic of the dynamic fender pile model used in this study is
shown in Figure 5.2. The pile model consists of nine lumped masses positioned along a flexible
beam: one at the upper support at deck level (m1), one at the vessel impact location (m2), two
along the submerged part of the pile (m3 and m4), and the remaining five positioned over the
embedded portion of the pile (m5 through m9). Linear springs are used to model the stiffness of
the soil and the rubber fender which provides support near the top of the pile. The system has 18
degrees of freedom including one translation and one rotation at each node.
32
Vvo
Vvo
Mv
Mp
x
a)
b)
V=0
c)
d)
Figure 5.1. Schematic of the dynamics of a vessel impacting a fender pile: (a) prior to impact, (b)
at time of impact, (c) after impact, and (d) after decoupling.
Fender Pile
Rubber Fender
m
h
Vessel
kf
1
m
h2
h2
m
h3
m4
h3
m5
k s1
m6
k s2
m7
k s3
m8
k s4
m9
k s5
In a single degree of freedom system, the equation of motion governing the free response
of the system is based on Newtons second law defined by the following partial differential
equation
m&x& + cx& + kx = 0
(Eq. 5.1)
Where m, c and k are the mass, damping, and stiffness, and &x& , x& , and x, are the acceleration,
velocity, and displacement. For an N-degree of freedom system there is a system of N coupled
differential equations having N number of modes and N number of natural frequencies. A multidegree of freedom system can be expressed in terms of its modal coordinates whose equation of
motion is of similar form to the single degree of freedom system (Chopra 2000)
M n q&&n + Cn q& n + K n qn = 0
(Eq. 5.2)
Where,
Mn= n m n ,
Cn= n c n ,
Kn= n k n
T
The modal quantities qn , q& n , and q&&n in Equation (5.2) are analogous to x , x& , and &x& in
the 1-D equation of motion. The mode shapes n are vectors that describe the deflected shapes of
the lumped masses for each mode of oscillation. Equation (5.2) can be re-written by dividing
through by Mn and simplifying to obtain the following expression
2
q&&n + 2 n n q& n + n qn = 0
(Eq. 5.3)
The solution to Equation (5.3) is of similar form to the solution of a SDOF system given by
n d
(Eq. 5.4)
Where,
q n (0) =
n T mx (0)
Mn
q& n (0) =
n T mx& (0)
Mn
The variables qn (0) and q& n (0) define the initial conditions of the system given an initial
displacement vector x (0) and an initial velocity vector x& (0) . Since the pile is initially at the
equilibrium position (i.e. x (0) = 0 ), Equation (5.4) simplifies to the following
q& (0)
q n (t ) = n
sin nd t exp( n n t )
nd
(Eq. 5.5)
34
The mass at the vessel impact point (m2 in Figure 5.2) is assumed to have an instantaneous
velocity equal to the vessel velocity while all other masses have initial velocities of zero. The
damped natural frequency of each mode of oscillation (nd) is a function of the undamped
natural frequency (n) and the damping ratio (n) given by
nd = n 1 n 2
(Eq. 5.6)
The undamped natural frequency of each mode (n) is determined by solving the characteristic
equation
det[k n m ] = 0
(Eq. 5.7)
And the corresponding mode shapes ( n ) can be evaluated from the following equation
[k n m] n = 0
(Eq. 5.8)
Note that n and n can also be evaluated simultaneously using the eigenvalue problem function
in the program Matlab. The time varying displacements are evaluated by multiplying Equation
(5-5) by the corresponding mode shape and summing the responses of each mode
9
q& (0)
x (t ) = n n
sin nd t exp( n n t )
n =1
nd
(Eq. 5.9)
Differentiation of Equation (5.9) with respect to time also yields the velocity of the system
9
q& (0) n n
x& (t ) = n q& n (0) cos nd t n
sin nd t exp( n n t )
nd
n =1
(Eq. 5.10)
For design it is of interest to determine the forces acting on the pile from the impacting vessel.
The equivalent static force acting at the impact point (F2) can be evaluated from the product of
the stiffness matrix and calculated displacements
F2 (t ) = [k x (t )]i = 2
(Eq. 5.11)
Likewise, the force applied to the pier structure through the fender support (F1) can be
determined from the spring constant and displacements at the uppermost node
F1 (t ) = k f x(t ) i =1
(Eq. 5.12)
And the maximum moment in the pile at any given time under the given pile configuration can
be readily obtained
35
M max (t ) = F1 (t ) h1
(Eq. 5.13)
To determine the dynamic response of the fender pile, appropriate mass and stiffness matrices
and damping ratio must be established. These parameters are discussed in detail in subsequent
sections.
5.1.1 Mass Matrix
As shown in Figure 5.2, the fender pile model consists of 9 lumped masses (m1 through
m9) each having 18 degrees of freedom including 9 translation and 9 rotation. It is anticipated
that the pile inertia will be primarily translational, and thus the masses associated with rotation
are neglected to yield a diagonal mass matrix (m) where the term mii in the matrix is defined as
mii = mi
(Eq. 5.14)
The mass of the pile itself is distributed among the lumped masses in proportion to the
spacing between nodes. In addition to the pile mass, mass m2 also includes the vessel mass and
vessel hydrodynamic added mass. Masses m3 and m4 also include the hydrodynamic added mass
of the pile moving laterally through the water column. These parameters will be quantified for
the impact test discussed in the next section.
5.1.2 Stiffness Matrix
The stiffness matrix provides the restoring force in the dynamic system and was
evaluated for the fender pile using the direct stiffness method (e.g. Leet 1988). A typical term kij
within the stiffness matrix k is determined as the force at degree of freedom i due to a unit
displacement at j when all other displacements are zero. Each term in the stiffness matrix was
derived for the 18 degrees of freedom for the model shown in Figure 5.2 yielding an 18 by 18
matrix. To be consistent with the size of the 9 by 9 mass matrix but still include rotational
stiffness, the stiffness matrix is condensed by first partitioning the 18 by 18 matrix as follows:
k
k = tt
k ot
k to
k oo
(Eq. 5.15)
And then substituting the submatricies into the following expression (Chopra 2000)
1
k = k tt k to k oo k ot
(Eq. 5.16)
Where,
36
12EI
12EI
0
0
0
0
0
0
3
h 3 + kf
h1
1
3EI
12EI 3EI
12EI
+
0
0
0
0
0
3
3
3
h3
h1
2h2
2h2
1
3EI
3EI 12EI 12EI
+ 3
0
0
0
0
0
3
3
3
h3
h3
2h2
2h2
12EI
12EI
24EI
0
0
0
0
0
3
3
3
h3
h3
h3
12EI
12EI
24EI
+ kS 2
0
0
0
0
0
d3
d3
d3
12EI
12EI
24EI
+ kS 3
0
0
0
0
0
d3
d3
d3
12EI
24EI
+ kS 4
0
0
0
0
0
0
d3
d3
12EI
0
0
0
0
0
0
0
d3
6 EI
6 EI
0
0
0
0
0
0
0
2
h2
h1
1
3EI
6 EI 3EI 6 EI
0
0
0
0
0
0
2
2
2
h2
2h2
2h2
h1
1
3
6
3
6
EI
EI
EI
EI
0
0
0
0
0
0
2
2
2
2
2h2
2h2
h3
h3
6 EI
6 EI
0
0
0
0
0
0
0
2
2
h
h
3
3
6 EI 6 EI 6 EI
6 EI
k =
0
0
0
0
0
to 0
2
2
d2
d2
h3
h3
6 EI
6 EI
0
0
0
0
0
0
0
d2
d2
6 EI
6 EI
0
0
0
0
0
0
0
2
2
d
d
6 EI
6 EI
0
0
0
0
0
0
0
2
2
d
d
6 EI 6 EI
0
0
0
0
0
0
0
d2
d2
6 EI
h2
1
6 EI
h2
1
0
k = 0
ot
6 EI
2
h1
3EI 6 EI
2
2
2h2
h1
3EI
2
2h2
0
0
3EI
2
2h2
6 EI 3EI
2
2
2h2
h3
6 EI
2
h3
6 EI
2
h3
6 EI
d2
6 EI
d2
6 EI
2
h3
6 EI 6 EI
2
d2
h3
6 EI
d2
6 EI
2
h3
6 EI
d2
6 EI
d2
37
6 EI
d2
0
6 EI
d2
6 EI
d2
6 EI
d2
0
12EI
d3
24EI
+
k
S5
3
d
4 EI
h
1
2 EI
h1
k = 0
to
2 EI
h1
4 EI 2 EI
+
h1
h2
EI
h2
EI
h2
2 EI 4 EI
+
h2
h3
2 EI
h3
0
0
2 EI
h3
8EI
h3
2 EI
h3
0
2 EI
h3
4 EI 4 EI
+
h3
d
2 EI
d
2 EI
d
8EI
d
2 EI
d
2 EI
d
8EI
d
2 EI
d
0
0
2 EI
d
8EI
d
2 EI
d
2 EI
d
4 EI
Here the bending stiffness of the pile is the product of the modulus of elasticity (E) and the
moment of inertia (I) of the pile. The spring constant of the fender is kf and the soil spring
constants are kS1 through kS5 as illustrated in Figure 5.2. The dimensions h1, h2, h3, and d are also
shown in Figure 5.2.
5.1.3 Damping
The energy losses that occur when a ship impacts a fender pile can be attributed to any
combination of the following: viscous drag on the vessel and pile, generation of surface waves,
and material losses within the pile, soil, rubber fenders, and vessel hull. The damping
characteristics of a structure can typically be obtained by performing a free vibration test. In this
test the structure is initially forced into a state of free vibration and the damping ratio is
evaluated from the decay in the displacement, velocity, or acceleration amplitude over time. The
damping ratio () can be approximated from the logarithmic decrement having the form (e.g.
Chopra 2000)
Aj
1
ln
2N A j + N
(Eq. 5.17)
Where N is the number of cycles between two peak amplitude values Aj and Aj+N. In a fender
pile system the vessel only remains coupled to the pile for -cycle. Therefore, the damping ratio
can be estimated from the initial vessel velocity (Vvo) and the velocity when the vessel becomes
decoupled from the pile (Vfo) using the following equation
n = =
ln
Vvo
(Eq. 5.18)
Vvf
For simplicity, a single damping ratio is used to characterize damping for all modes and
degrees of freedom for the fender pile system. Damping for most civil structural systems is
typically less than 20% (Chopra 2000).
38
Velocity (m/s)
0.15
0.1
0.05
0
-0.05
-0.1
0.5
0.5
1.5
2.5
1.5
2.5
Displacement (m)
0.04
0.03
0.02
0.01
0
-0.01
Time (s)
Figure 5.3. Typical velocity and displacement time history (Test 3) as integrated from the
accelerometer mounted at the vessel impact point.
The accelerometer that was mounted near the impact point was also used to estimate the
velocity of the vessel during impact. The mass of the vessel is significantly larger than the pile
mass and therefore changes in the vessels momentum at initial impact are anticipated to be
negligible. As shown by the velocity time history shown in Figure 5.3 there are two points at
which the velocity is a maximum; one just after the point of impact representing the initial
velocity of the vessel and the second when the vessel becomes de-coupled with the pile. A
summary of the initial and final vessel velocities from the tests are summarized in Table 5.2.
39
40
EI =
Pa
(
3L2 4a 2 )
48
(Eq. 5.20)
In the load test a was 1.016 m and L was 3.048 m. As shown in Figure 5.4 the stressstrain curve is slightly non-linear. Since the model requires a linear stiffness, a tangent modulus
of 2.66 x 106 N-m2 was calculated at 20% of the ultimate moment. This criteria is consistent
with other composite pile manufacturers guidelines (e.g. Hardcore Composites).
500
Force (kN)
400
300
200
100
0
0
10
15
Displacement (cm)
41
k si = K h d w
(Eq. 5.21)
The spring constant of the node at the mudline was reduced by 50% since the effective depth of
soil influencing this node is d 2 .
5
Force (kN)
0
0
0.5
1.5
2.5
Displacement (cm)
Figure 5.5. Results of the static compression test performed on a section of rubber fender.
42
Velocity (m/s)
0.15
0.1
0.05
0
-0.05
-0.1
0.2
0.4
0.6
0.8
1.2
1.4
0.2
0.4
0.6
0.8
Time (s)
1.2
1.4
Displacement (m)
0.04
0.03
0.02
0.01
0
Figure 5.6. Modeled displacement at the impact point using an average vessel velocity
of 0.12 m/s and 19% damping.
The maximum lateral displacements measured along the length of the pile as averaged
from all impact tests (Table 5.1) are plotted in Figure 5.7. The maximum displacements
calculated with the dynamic model for an average velocity of 0.12 m/sec and a damping of 19%
are also shown in the figure. For comparison, the kinetic energy method was used to calculate
maximum displacements using the same stiffness matrix used in the dynamic model. The kinetic
energy method is included in the comparison because it represents a condition having zero
energy losses (i.e. no damping).
Given the selected parameters, the displacements calculated in both the dynamic and
static models compare reasonably well with the field data. The kinetic energy method, however,
yielded slightly higher displacements relative to both the modeled and the measured values.
Most importantly, by incorporating damping into the system the maximum displacement at the
impact point is reduced by about 25 percent. For a linear system this is equivalent to a 25
percent reduction in the forces and moments on the pile. From a design perspective, lower
stresses may justify the use of smaller piles (or fewer piles) resulting in cost savings.
43
Measured
10
12
14
Figure 5.7. Comparison of modeled and average measured displacements along the length of the
pile for an average vessel velocity of 0.12 m/s.
44
6.0 Conclusions
The primary objectives of this research study were to improve the understanding
of the performance of composite piles as a fendering system in the marine environment.
Specific areas of study included an evaluation of driving stresses during hard driving
conditions (i.e. bedrock) and the response of composite piles to lateral impact loads.
This was accomplished through two separate field studies in which concrete-filled
FRP pipe piles and steel reinforced plastic piles were installed at a residential site in Old
Greenwich, CT and along a pier at Fort Wetherill in Jamestown, RI. The piles in Old
Greenwich were driven to failure with a hydraulic hammer and PDA and CAPWAP
analyses were performed. It was found that the Pile Integrity Tester (PIT) was effective
in measuring damage of the FRP pile due to driving. It was difficult to obtain PIT results
with the plastic pile. Definitive conclusions regarding CAPWAP and driving stresses
could not be made because of problems installing the piles to sufficient depths. It was
particularly difficult to anchor accelerometers and strain gauges in the FRP pile.
The piles at Fort Wetherill were impacted with an 85 ton vessel at low speeds and
the dynamic response of the piles was measured using accelerometers and displacement
transducers. The acceleration time histories were integrated twice to obtain the
displacements using the program MATLAB. The results indicated that the piles absorbed
the impact energy through both translation and bending, and the overall stiffness of the
fender system was estimated using the kinetic energy approach.
A new dynamic analysis approach was developed for the design of these flexible
fender piles. Unlike the traditional kinetic energy method, the dynamic approach
considers the energy losses in the system (i.e. damping) that occurs during vessel impact.
Damping in a fender system is a useful parameter for quantifying its energy dissipation
characteristics. The model accounts for the stiffness contributions of all fender
components including rubber fender supports, pile, and soil. Impact test results
performed on an FRP composite pile were used to evaluate the damping properties of this
fender pile system, and to validate the dynamic model. An average damping of 19% was
estimated using the velocity time history of the vessel recorded during impact.
Incorporating this value of damping the displacements obtained from the dynamic model
compare well with the measured data in terms of the impact duration and maximum
displacements. As compared to the kinetic energy approach, the dynamic approach
reduced the displacements, forces, and moments by about 25 percent. Additional studies
of the damping characteristics of other pile types such as timber and plastic piles may
provide further insight on the relative design benefits of using these pile types for
fendering applications.
45
7.0 Acknowledgements
This study was made possible by a grant from the University of Rhode Island
Transportation Center and in-kind support from the following organizations: G.
Donaldson Construction Co., Inc., GZA Geoenvironmental, Inc., Heller and Johnsen,
Lancaster Composite, Plastic Pilings, Inc., and the Rhode Island Department of
Environmental Management. Their assistance is greatly appreciated. Special thanks to
Francois Enet, Jason Ressler, and Fred Pease for help with the instrumentation and field
impact tests. The authors would also like to acknowledge the insightful comments of the
anonymous reviewer in the development of the new analysis and design method
presented in Chapter 5.
46
8.0 References
Bowles, J. E. (1988). Foundation Analysis and Design, McGraw-Hill Publishing Company, New
York.
Chopra, A.K. (2000). Dynamics of Structures: Theory and Applications to Earthquake
Engineering, Prentice Hall, New Jersey.
Fellenius, B. H., Riker, R. E., OBrien, A. J., and Tracy, G. R. (1989). Dynamic and Static
Testing in Soils Exhibiting Setup. American Society of Civil Engineers, Journal of
Geotechnical Engineering, 115(7), pp. 984-1001.
FHWA (1998). Design and Construction of Driven Pile Foundations, Vol. II, U.S. Department
of Transportation Federal Highway Administration, National Highway Institute, Publication No.
FHWA HI 97-014.
Gaythwaite, J.W. (2004). Design of Marine Facilities, ASCE Press, Reston, Virginia.
Goble, Rausch, Likins, and Associates, Inc. (2000). PDA Results - Pipe Extension No.4 Beach
Erosion Control Project, Hardcore Composite Report, Report No.2000HCR0030.
Gummert, M. N. (2003). Evaluation of the Drivability of Composite Piles. Masters Thesis,
University of Rhode Island, 220p.
Iskander, M.G. and Hassan, M. (1998). State of the Practice Review in FRP Composite Piling,
Journal of Composites for Construction, August, pp. 116-120.
Iskander, M.G. and Stachula, A. (1999). FRP Composite Polymer Piling: An Alternative to
Timer Pilling for Water-Front Applications, Geotechnical News, pp. 27-29.
Iskander, M.G., Hanna, S., and Stachula, A. (2001). Drivability of FRP Composite Piling,
Journal of Geotechnical and Geoenvironmental Engineering, February, pp.169-176.
Iskander, M.G., and Stachula, A. (2001). Drivability of FRP Composite Piling, Journal of
Geotechnical and Geoenvironmental Engineering, 127(2), pp.169- 176.
Iskander, M.G., and Stachula, A. (2002).
Kinsler, L. E., Frey, A. R., Coppens, A. B., and Sanders, J. V. (1982). Fundamentals of
Acoustics, 3RD Edition. Canada, John Wiley & Sons, Inc.
Kistler Instrument Corporation (2004). Website: www.kistler.com.
Kozera, D.W. (1996) Dynamic Pile Testing, Hardcore Composites Report, Report No.
2000HCR0004.
47
Lampo, R., Nosker, T., Barno, D. Busel, J., Maher, A., Dutta, P., and Odello, R. (1998).
Development and Demonstration of FRP Composite Fender, Loadbearing, and Sheet Piling
Systems, USACERL Technical Report 98/123, September.
Leet, K.M. (1988). Fundamentals of Structural Analysis. Macmillan Publishing Company, New
York.
Li, S., Ramakrishnan, V. (1971). Ultimate energy design of prestressed concrete fender piling.
J. Waterways, Harbors, and Coastal Engineering Division, ASCE, 97(WW4), 647-662.
Maher, M.H., Gucunski, N., and Chae, Y. S. (1996). Composite Fender and Sheet Piles in
Marine Front Systems, Proceedings of First International Conference on Composites in
Infrastructure, ICCI 96, Tucson, AZ.
Massa Products Corporation (1998). Website: www.massa.com
Newman, J.N. (1977). Marine Hydrodynamics, MIT Press, Cambridge, Massachusetts.
Pando, M. A., Filz, G. M., Early, C., Hoppe, E. (2003). Axial and Lateral Load Performance of
Two Composite Piles and One Prestressed Concrete Pile. TRB 2003 Annual Meeting CDROM, Paper No: 03-2912.
Pando, M. A., Brown, D., and Filz, G. M. (2004). Performance of Laterally Loaded Composite
Pile at the Nottoway River Bridge, GeoTrans Conference Proceedings, Geotechnical
Engineering for Transportation Projects, ASCE GSP No. 126.
Pile Dynamics, Inc. (2002). GRLWEAP Wave Equation Analysis of Pile Driving Procedures
and Models Version 2002.
Rausche, F., Likins, G. E., Goble, G. G., and Miner, R. (1985). The Performance of Pile
Driving Systems. Main Report, U.S. Department of Transportation, Federal Highway
Administration, Office of Research and Development, Washington, D.C.
Reese, L.C., ONeill, M.W., and Radhakrishnan, N. (1970). Rational design concept for
breasting dolphins. J. Waterways and Harbors Division, ASCE, 96(WW2), 433-450.
Rizkalla, S.H., Fam, A.Z. (1999). State-of-the-art Report on Stay- In-Place FRP Formwork,
submitted to ACI Committee 440 J, Lancaster Composite, Inc. Technical Reference Part 1.
Rutgers (1996). Flexure tests of 12.75 diameter Composite Post 40 piles-preliminary report.
Department of Civil and Environmental Engineering, Rutgers University.
USACE (1983). Engineering and Design of Military Ports. TM 5-850-1, US Army Corps of
Engineers.
48
APPENDIX A:
MATLAB code used to filter, integrate, and plot the data from the field impact tests.
49
PHYSICAL CONSTANTS
=
32.2;
(ft/s^2)
=
2*pi;
(d'less)
%
fs
Rate)
EXPERIMENTAL FACTORS
=
200.00;
(Hz)
CONVERSION FACTORS
%
CALIBRATION FACTORS
m1
=
0.9989;
calibration
(g/volts)
b1
=
0.017;
intercept (g)
m2
=
0.9893;
calibration
(g/volts)
b2
=
0.0514;
intercept (g)
m3
=
4.9396;
calibration
(g/volts)
b3
=
-0.0596;
(g)
m4
=
4.9775;
calibration
(g/volts)
b4
=
-0.0203;
(g)
m5
=
16.787;
(inch/volts)
b5
=
-5.1742;
(inch)
Two times pi
accelerometer 1
accelerometer 1 y-
accelerometer 2
accelerometer 2 y-
accelerometer 3
accelerometer 3 y-intercept
%
accelerometer 4 y-intercept
%
50
accelerometer 4
51
(ft/s^2)
(ft/s^2)
(ft/s^2)
(ft/s^2)
(inch)
mm=mean(Acc_Data_1(startone:endone));
mmm=mean(Acc_Data_1(starttwo:endtwo));
m=[mm mmm];
Mean_Acc_1=mean(m);
mm=mean(Acc_Data_2(startone:endone));
mmm=mean(Acc_Data_2(starttwo:endtwo));
m=[mm mmm];
Mean_Acc_2=mean(m);
mm=mean(Acc_Data_3(startone:endone));
mmm=mean(Acc_Data_3(starttwo:endtwo));
m=[mm mmm];
Mean_Acc_3=mean(m);
mm=mean(Acc_Data_4(startone:endone));
mmm=mean(Acc_Data_4(starttwo:endtwo));
m=[mm mmm];
Mean_Acc_4=mean(m);
% Find shift offsett or mean of range data excluding the peaks caused
% by instrument noise
d=1;
for p=startone:1:endone;
if abs(Range_Data(p)-Range_Data(p+1)) <= 0.01
Mean_Range_Data(d)=Range_Data(p);
d=d+1;
end
end
Mean_Range=mean(Mean_Range_Data);
% Subtract mean offsets to shift curves and data around zero
Acc_Data_1b=Acc_Data_1a-Mean_Acc_1;
Acc_Data_2b=Acc_Data_2a-Mean_Acc_2;
Acc_Data_3b=Acc_Data_3a-Mean_Acc_3;
Acc_Data_4b=Acc_Data_4a-Mean_Acc_4;
Range_Datab=Range_Dataa-Mean_Range;
% Create a time vector for the data based on the step size
timesteps=length(Acc_Data_1b);
time=[0:1/fs:(timesteps-1)/fs];
timeacc=[0:1/fs:(timesteps-1)/fs];
timevel=[0:1/fs:(timesteps-2)/fs];
timedisp=[0:1/fs:(timesteps-3)/fs];
%
%
%
%
%
%
%
%
% Plot the four accelerometer data and the range finder data
figure
subplot(5,1,1);plot(time,Acc_Data_1b);
ylabel('ft/sec^2');
set(gca,'xticklabel',[])
title(filenames(i).name);
legend('Acc 1 Not Filtered');
axis([0 4 -10 10]);
52
%
%
%
%
%
%
%
%
%
%
%
%
%
%
%
%
%
%
%
subplot(5,1,2);plot(time,Acc_Data_2b);
ylabel('ft/sec^2');
set(gca,'xticklabel',[]);legend('Acc 2 Not Filtered');
axis([0 4 -10 10]);
subplot(5,1,3);plot(time,Acc_Data_3b);
ylabel('ft/sec^2');
set(gca,'xticklabel',[]);legend('Acc 3 Not Filtered');
axis([0 4 -10 10]);
subplot(5,1,4);plot(time,Acc_Data_4b);
ylabel('ft/sec^2');
set(gca,'xticklabel',[]);legend('Acc 4 Not Filtered');
axis([0 4 -10 10]);
subplot(5,1,5);plot(time,Range_Datab);
ylabel('ft');
xlabel('Time (sec)');legend('Disp Not Filtered');
axis([0 4 -0.2 0.2]);
% Define the data matrix (this is shifted around zero and cropped for
% the impact zone.
accdata=[Acc_Data_1b Acc_Data_2b Acc_Data_3b Acc_Data_4b];
% Filter data. This must be done before the integrations so that the
% noise is not integrated and therefore compunded throughout the
% cumulative integration
order=1;
% this is the order of the butterworth filter
cutoff=0.05; % cutoff frequency (nondimensionnal see help)
[B,A] = butter(order,cutoff);
if i==4;
figure;
subplot(4,2,2);plot(time,accdata(:,1));
axis([0 4 -10 10]);text(2,-8,'Acc 1 Unfiltered');
set(gca,'xticklabel',[]);
ylabel('ft/sec^2');
subplot(4,2,4);plot(time,accdata(:,2));
axis([0 4 -10 10]);text(2,-8,'Acc 2 Unfiltered');
set(gca,'xticklabel',[]);
ylabel('ft/sec^2');
subplot(4,2,6);plot(time,-1.*accdata(:,3));
axis([0 4 -10 10]);text(2,-8,'Acc 3 Unfiltered');
set(gca,'xticklabel',[]);
ylabel('ft/sec^2');
subplot(4,2,8);plot(time,-1.*accdata(:,4));
axis([0 4 -10 10]);text(2,-8,'Acc 4 Unfiltered');
xlabel('sec');
ylabel('ft/sec^2');
end
plc=1;
for m=1:1:4;
53
accfilt(:,m) = filter(B,A,accdata(:,m));
if i==4;
subplot(4,2,plc);
if plc == 5;
plot(time,-1.*accfilt(:,m));
elseif plc == 7;
plot(time,-1.*accfilt(:,m));
else
plot(time,accfilt(:,m));
end
ylabel('ft/sec^2');
axis([0 4 -10 10]);
if m==1; text(2,-8,'Acc 1 Filtered');set(gca,'xticklabel',[]);
text(3,14,'Fiber Reinforced Pile Test 4');
elseif m==2; text(2,-8,'Acc 2 Filtered');set(gca,'xticklabel',[]);
elseif m==3; text(2,-8,'Acc 3 Filtered');set(gca,'xticklabel',[]);
elseif m==4; text(2,-8,'Acc 4 Filtered');xlabel('sec');
end
plc=plc+2;
end
end
% Now that all of the data is corrected and filtered, plot this on a
% subplot includign acceleration from acc 1 - 4 and disp from sonic
figure
subplot(5,1,1);plot(time,accfilt(:,1));
ylabel('ft/sec^2');
set(gca,'xticklabel',[])
title(filenames(i).name);
text(2.5,7,'Accelerometer 1 Filtered');
axis([0 4 -10 10]);
subplot(5,1,2);plot(time,accfilt(:,2));
ylabel('ft/sec^2');
set(gca,'xticklabel',[]);text(2.5,7,'Accelerometer 2 Filtered');
axis([0 4 -10 10]);
subplot(5,1,3);plot(time,-1.*accfilt(:,3));
ylabel('ft/sec^2');
set(gca,'xticklabel',[]);text(2.5,7,'Accelerometer 3 Filtered');
axis([0 4 -10 10]);
subplot(5,1,4);plot(time,-1.*accfilt(:,4));
ylabel('ft/sec^2');
set(gca,'xticklabel',[]);text(2.5,7,'Accelerometer 4 Filtered');
axis([0 4 -10 10]);
subplot(5,1,5);plot(time,Range_Datab);
ylabel('ft');
xlabel('Time (sec)');text(2.5,0.13,'Sonic Range Finder');
axis([0 4 -0.2 0.2]);
54
%
%
figure
plot(time,Range_Datab,timedisp,accdataint2(:,1));
55
%
%
%
%
%
title(filenames(i).name);
text(0.5,-0.18,'Range finder and Accelerometer Displacement Comparison');
xlabel('Time (sec)');
ylabel('Displacement (ft)');
legend('Range Finder','Integrated Accelerometer');
% Plot all displacement data together
figure
subplot(4,1,1);
plot(timedisp,accdataint2(:,1),time,Range_Datab);
title(filenames(i).name);
set(gca,'xticklabel',[]);ylabel('ft');
text(0.25,0.2,'Range Finder and Integrated Accelerometer 1');
axis([0 4 -0.3 0.3]);
subplot(4,1,2);
plot(timedisp,accdataint2(:,2));ylabel('ft');
axis([0 4 -0.3 0.3]);set(gca,'xticklabel',[]);
text(0.25,0.2,'Integrated Accelerometer 2');
subplot(4,1,3);
plot(timedisp,-1.*accdataint2(:,3));ylabel('ft');
axis([0 4 -0.3 0.3]);set(gca,'xticklabel',[]);
text(0.25,0.2,'Integrated Accelerometer 3');
subplot(4,1,4);
plot(timedisp,-1.*accdataint2(:,4));ylabel('ft');
axis([0 4 -0.3 0.3]);
text(0.25,0.2,'Integrated Accelerometer 4');
xlabel('sec');
% Now find the maximum displacement throughout the depth of the pile
% This array will have columns that contain accelerometer data 1
% through 4 integrated twice to find displacement in coulumn 1 through
% 4 and the fifth column will be displacement found via the range
% finder. All acc data has been filtered.
dispmax1=min(accdataint2(:,1));
dispmax2=min(accdataint2(:,2));
dispmax3=max(accdataint2(:,3));
dispmax4=max(accdataint2(:,4));
dispsonic=min(Range_Datab);
56
APPENDIX B:
Filtered acceleration and integrated displacement data for all tests on FRP and PPI piles.
57
58
59
60
61