You are on page 1of 24

JOURNAL OF

COMPOSITE
Article M AT E R I A L S
Journal of Composite Materials
47(2) 207–229
! The Author(s) 2012
Analysis and design procedures for the Reprints and permissions:
sagepub.co.uk/journalsPermissions.nav
flexural behavior of glass fiber-reinforced DOI: 10.1177/0021998312438721
jcm.sagepub.com
polymer composite poles

Slimane Metiche and Radhouane Masmoudi

Abstract
Earlier research progress on using fiber-reinforced polymer (FRP) composite poles for transmission lines revealed that
the use of these poles is economically viable and industrially reliable. While this promising alternative technique is coming
to be widely accepted in practice and evolving to become the future of the construction in transmission lines, the
development of appropriate design guidelines that deal with detailed analysis of such poles has lagged behind. Design
approaches for conventional steel/wood poles were influential in providing safety and robustness to these poles.
However, little advanced approaches focused on the design of fiber-reinforced polymer poles. This is the focus of the
present paper. The available design procedures of fiber-reinforced polymer poles are based on the allowable stress design
theory of composite material under various states of stress. However, various experimental results show that cracking
and early failure of fiber-reinforced polymer poles are normally controlled by the relative location of the opening from
the base of the pole. Most of the design guidelines ignore such effect and accounts only for the effect of these holes by
considering their influence on the abrupt reduction in the cross-sectional area of the poles. These guidelines do not give
a specific attention to the impact of the opening on the generated stress concentrations in their vicinity. Furthermore,
local buckling at a nearby area of the opening generally dominates the mode of failure of such poles that requires more
attention to the relative locations of the openings. A new design approach is introduced to accurately evaluate the design
equations for the flexural behavior of fiber-reinforced polymer -composite poles. The accuracy of both the developed
design procedures and that of existing design approaches are verified by comparison with documented test results of an
early experimental program of the authors.

Keywords
Fiber reinforced polymers, glass fiber-reinforced polymer poles, flexural behavior, filament winding, maximum deflection,
allowable stress method, design procedure

Composite structural elements are used today in a


Introduction variety of components for automotive, aerospace,
The mechanical behavior of fiber-reinforced polymer marine, architectural structures, and sports equipment.
(FRP) materials is a topic that has attracted several The majority of the existing electrical poles in
researchers in recent years. These materials have very Canada as well as in the world are made from tradi-
high strength-to-weight ratios, and because of this they tional materials such as wood, concrete, or steel. The
are steadily being selected over metal alloys in a variety limitation in length of wooden poles and the vulnera-
of structural applications.1 bility of steel or concrete poles to climate aggressions
The resulting material is referred to as a composite
material when two or more distinct materials are Department of Civil Engineering, University of Sherbrooke, Sherbrooke,
combined in a macroscopic scale. The basic constituents Québec, Canada
of such a material are usually combined to enhance the
Corresponding author:
mechanical characteristics of the materials. Fiber- Radhouane Masmoudi, Department of Civil Engineering, University of
reinforced polymer technology gained its roots during Sherbrooke, Sherbrooke, Québec, J1K2R1, Canada.
World War II with the need of light structural products. Email: radhouane.masmoudi@usherbrooke.ca
208 Journal of Composite Materials 47(2)

have motivated the manufacturers and researchers to Table 1. Properties of fibers and resin
find alternatives.
Glass fibers Glass fibers Epoxy
While studies have addressed the material failure
Properties type A type B resin
and buckling of thin-walled sections such as I-beams,
box beams, and so on, made from composite mate- Linear mass, a (g/km) 1100 2000 –
rials,2 very few studies have been conducted on the Density (gm/cm3) 2.6 2.6 1.2
behavior of tapered sections.3 Modulus of elasticity (GPa) 80 80 3.38
The behavior of scaled FRP models of transmission Shear modulus (GPa) 30 30 1.6
poles under cantilever loading conditions was investi- Poisson’s ratio 0.25 0.25 0.4
gated by Lin.4 The four specimens tested were of pris-
a
matic circular hollow cross-section. The outside In the fiber industry, the linear mass () of the fibers is known as Tex,
which indicates the weight in gram of a 1000-m long single fiber (linear
diameter of the poles was 76 mm and the wall thickness
density).
was 6 mm. These were fabricated using pultruded sheets
arranged in circular pattern. According to the test
results, a linear behavior of the FRP poles was experimental results. The design method proposed
observed up to failure. allow to determine for a given ratio (E I)/(L ) at the
Shakespeare Company5 provided a report on lateral base of an FRP pole, the ultimate bending moment, the
loading test of a 10.65 -m length fiberglass pole. This maximum pole top deflection, the ultimate longitudinal
report shows the minimal effective loss of strength due compression and tension strains at the base, and the
to the row of 22.23-mm diameter holes drilled in the ultimate longitudinal tension strain at the service open-
side wall to climb the pole. The inner diameters of the ing level. In the ratio (E I)/(L ), E is the longitudinal
specimen at the base and the top end were 216 mm and modulus of elasticity, I is the moment of inertia, L is the
127 mm, respectively. A total of 22 holes were drilled cantilever height of the pole and  is the linear mass of
and spaced by 305 mm. The first one was drilled at the fibers. This paper presents also a design verification
2.13 m far from the base-end of the pole. analysis according to the AASHTO LTS-5-2009 stan-
Experimental and analytical studies were carried dard of the full-scale manufactured and presented
out3 to validate the predicted ultimate loads for tapered GFRP poles.
filament wound FRP scaled poles subjected to cantile-
ver bending. The specimens were 2500 mm in length;
the inner diameters at the base and the top end were Experimental program
100 mm and 74 mm, respectively. The wall thickness
varied depending on the number of layers, the results
Test prototypes
of this study show that the stiffness and the strength of Mechanical bending tests under lateral loading were
FRP poles as well as the mode of failure depend mainly performed on 22 full-scale prototypes of FRP poles
on their wall thicknesses. While a local buckling failure with length ranging from 5 to 12 m. The FRP poles
is observed for thin-walled samples, compression and having hollow circular cross-section and variable wall
tension failures were observed for samples with more thickness were produced as single units with the fila-
significant thicknesses. ment winding process, using epoxy resin reinforced
The fiber orientation has a significant effect on the with E-glass fibers. Each type of the poles tested in
performance of FRP poles. The same performance of this study is constituted by three zones where the geo-
FRP poles having high fiber volume fraction can be metrical and the mechanical properties are different in
achieved using less fiber volume fraction but with each zone. The difference of these properties is due to
changing the fiber orientation toward the longitudinal the number of layers used in each zone and the fiber
direction3 and the incorporation of circumferential orientation of each layer. The mechanical and physical
layers tend to increase the critical ovalization load.6 properties of the fibers and the epoxy resin are pre-
More tests, however, are required to determine the opti- sented in Table 1.
mum values of fiber angle and fiber volume fraction. The characteristics and configuration of the tested
A research project is currently carried out at poles are presented in Table 2. All test prototypes
the University of Sherbrooke (Quebec, Canada). The have been tested in flexural bending up to failure.
main objective of this research project is to study the Two types of fibers (types A and B) are used to evaluate
full-scale flexural behavior of FRP tapered poles man- their effects on the flexural behavior. Note that the only
ufactured by the filament winding process in order to difference between both types is the linear density, as
optimize the design and to propose improvement of the shown in Table 1.
manufacturing process.7–10 This paper presents a design All the prototypes are identified as follows: From the
method for glass FRP (GFRP) poles based on left to the right: the first number indicates the total
Metiche and Masmoudi 209

Table 2. Characteristics and configuration of the tested poles

Principal hole (AGL)b

Pole Id. hTotala hSupporteda Bottom-top Dimensions Location


and samples (mm) (mm) diameters (mm) (mm) (mm) Positioning

17-B-3-C 2 5093 914 150-76 64  127 610 Comp.


17-A-3-C 2 5093 914 150-76 64  127 610 Comp.
18-B-3-C 2 5398 914 155-76 102  305 762 Comp.
18-B-3-T 2 5398 914 155-76 102  305 762 Tension
20-B-4-C 2 5994 1219 164-76 102  305 1372 Comp.
33-B-5-C 2 10 058 1524 261-114 102  305 1219 Comp.
35-B-5-C 2 10 566 1524 270-114 102  305 1219 Comp.
40-A-5-C 2 12 090 1524 291-114 102  305 1219 Comp.
40-B-5-C 1 12 090 1524 291-114 102  305 1219 Comp.
40-B-5-T 1 12 090 1524 291-114 102  305 1219 Tension
29-B-5-C 1 8738 1524 247-114 64  127 457 Comp.
18-B-4-C 1 5398 1219 155-76 102  305 457 Comp.
18-B-4-T 2 5398 1219 155-76 102  305 457 Tension
a
hTotal: Total length of the pole; hSupported: supported length; bAGL: located above ground line.

Figure 1. Zones and thickness of FRP poles. FRP: fiber-reinforced polymer.

Table 3. Stacking sequence for the three zones of the poles

Pole ID Zone I (Degrees) Zone II (Degrees) Zone III (Degrees)

17-B-3-C [70/70/60/60/(15/15)2/60/70] [(60/60)2/(15/15)2/60/70] [15/15/30/60/70]


17-A-3-C [60/70/60/30/60/70] [60/60/(15)2/60/70] [15/70/70]
18-B-3-C [60/60/30/30/60/70] [90/(15/15)2/60/90] [25/25/75/75]
18-B-3-T [60/60/30/30/60/70] [90/(15/15)2/60/90] [25/25/75/75]
20-B-4-C [60/60/25/25/70/70] [90/60/60/15/15/60/60] [15/60/70]
33-B-5-C [70/80/20/70/80] [90/15/70/80] [15/90]
35-B-5-C [70/75/20/70/80] [90/70/15/70/80] [70/10/10/þ10/70/80]
40-A-5-C [70/30/70] [90/20/20/70] [70/10/10/70]
40-B-5-T [75/20/30/70/80] [90/75/15/þ15/70/80] [80/70/10/70/80]
40-B-5-C [75/20/30/70/80] [90/75/15/þ15/70/80] [80/70/10/70/80]
18-B-4-C [60/60/30/30/60/70] [90/(15/15)2/60/90] [25/25/75/75]
18-B-4-T [60/60/30/30/60/70] [90/(15/15)2/60/90] [25/25/75/75]
210 Journal of Composite Materials 47(2)

Table 4. Average thickness and length for the three zones of the poles

Zone I Zone II Zone III

Length Average Length Average Length Average


Prototypes Vf (%) (mm) thickness (mm) (mm) thickness (mm) (mm) thickness (mm)

17-B-3-C 49 2133 4.78 2219 4.30 740 5.18


17-A-3-C 49 2133 4.78 2219 4.30 740 5.18
18-B-3-C 51 1220 3.27 980 7.20 3198 3.04
18-B-3-T 51 1220 3.27 980 7.20 3198 3.04
20-B-4-C 50 2000 2.81 1200 6.73 2794 2.84
33-B-5-C 59 2200 4.56 1000 8.35 6858 5.97
35-B-5-C 57 2200 3.87 1000 9.69 7366 5.37
40-A-5-C 51 2200 4.72 1000 9.80 8890 6.70
40-B-5-T 55 2200 5.54 1000 10.44 8890 7.73
40-B-5-C 55 2200 5.54 1000 10.44 8890 7.73
18-B-4-C 51 1220 3.27 980 7.20 3198 3.04
18-B-4-T 51 1220 3.27 980 7.20 3198 3.04

height of the pole in feet, the first letter indicates the This test setup consists mainly of three parts: a
type of fiber, the second number indicates the sup- ‘‘ground-line support,’’ a ‘‘butt support,’’ and a ‘‘lifting
ported length in feet and the second letter indicates jaws.’’ This fixture provides a practical way to test all
the hole positioning (in compression or in tension, the types of utility poles. The ground-line support or front
number zero means that the pole was tested without support is used with wooden saddle to support the pole
any hole above ground line), and the latest number at ground line and is designed to allow a vertical and/or
(in parentheses) indicates the test number under the horizontal translation to anchor the various possible
same parameters. diameters of the poles. The pole butt support or rear
Each type of pole tested in this study is constituted support is used with wooden saddle to support the
by three zones, where the geometrical and the mechan- lower end of the pole and is designed to allow longitu-
ical properties are different in each zone. The difference dinal translation to test various burial lengths of the
of these properties is due to the different number of poles. The lifting jaws constitute the load application
layers used in each zone and the fiber orientation of point on the pole and consist of two quarters of a
each layer (Figure 1). The stacking sequence as well metallic tube assembled so as to form two jaws
as the average thickness and length for the three (Figure 2(a)). After the pole were mounted and leveled
zones of poles are presented in Tables 3 and 4, respec- on the test fixture, a bridge crane was positioned with
tively. The fiber content of each prototype, expressed in its hook centered above the lifting jaws, 305 mm far
volume ratio Vf was determined experimentally by from the top of the pole.
pyrolysis tests10 and is presented in Table 4.
It should be mentioned that all the prototypes pre-
sented in this study are single segment and were fabri-
Instrumentation
cated with extra reinforcing provided around the A 225-KN load-capacity cell was used while the displace-
principal holes except for the prototypes 17-B-3-C ment rate of the bridge crane was 12 mm/s (Figure 2(b)).
and 17-A-3-C. On the other hand, there was no extra The deflection of the FRP poles was measured with a
reinforcing provided around the hole located under the draw wire transducer (DWT) at hc/4; hc/2 as well as
ground line. All the holes were cut at the manufacturer under the load application point (Figure 2(b)), where hc
site, after the poles were fabricated. is the cantilever length or free length of the pole.
Electrical strain gages were mounted on the two faces
(compression and tension) near the ground line support,
Test setup at hc/4; hc/2; 3/4hc as well as around the hole. The strain
A new test setup (Figure 2(a)) was designed and built gages were used to monitor the deformations in the lon-
according to the recommendations of the Standards gitudinal, circumferential directions and at 45 from the
ASTM D 4923-0111 and ANSI C 136.20-199012 as longitudinal axis of the pole. Two linear variable differ-
well as the Proposed California Test 683-1995.13 ential transducer (LVDTs) used to measure displacement
Metiche and Masmoudi 211

Load direction

Chain
FRP prototype

Load cell Web strap

Pole butt support


Lifting jaws

Rubber lined
wooden support
Ground line saddle bloc
support
DWT

Winch binder

Lifting
Ground line
jaws
support

FRP pole

Pole butt
support

Figure 2. (a) Schematic drawing of the full-scale test setup. (b) Full-scale test setup.

at the pole base were positioned against either the test moment of inertia, L is the cantilever height of the
fixture or the lower wall of the pole. LVDT # 1 was cen- pole, and  is the linear mass of the fibers. The follow-
tered on the underside of the pole at the ground line. ing relationships (equations (1) to (3))14 were used to
LVDT # 2 was centered on the topside of the FRP pole determine the modulus of elasticity E in the longitudi-
above the wooden support on the rear pole butt support. nal direction at the base of the pole.
Two other LVDTs were positioned laterally in order to
measure the possible ovalization of the pole near the X
N  
ground line support. An automatic data acquisition E¼ ðPi ÞExi ð1Þ
system was used to collect the load, LVDTs, DWTs, i¼1
and strain gages data.
where
Proposed design procedures
A new design approach is introduced based on the 1
Exi ¼ 4 4   ð2Þ
overall experimental results. The design method cos i sin i 2 2 1 tl
þ þ cos i sin i 2
allows to determine for a given ratio (E I)/(L ) at El Et Glt Et
the base of an FRP pole, the ultimate bending
moment, the maximum pole top deflection, the ultimate and
longitudinal compression, and tension strains at the
base of the FRP pole. In the ratio (E I)/(L ), E is tl lt
¼ ð3Þ
the longitudinal modulus of elasticity, I is the Et El
212 Journal of Composite Materials 47(2)

Figure 3. Design curve – ultimate bending moment (Mu,b) at the base – failure at the base.

Figure 4. Design curve – ultimate bending moment (Mu,o) at the opening – failure at the opening.

where Exi is the Young’s modulus in the longitudinal ultimate load (Fu). Figure 3 presents the prototypes
direction of the ith layer, (N) is the total number of that failed at the base. Equation (4) was obtained
layers at the base of the pole, tl and lt are the from the curve presented in the Figure 3 with a coeffi-
Poisson’s ratios, i is the fiber angle of the ith layer cient of regression (R2) of 0.99. R2 indicates the rate of
evaluated experimentally by a pyrolysis test, Pi is the correspondence between the trend curve and the exper-
thickness contribution of the ply with respect to the imental results. Figure 4 presents the curve of the ulti-
overall thickness of the laminate constituting the base mate bending moment (Mu,o) at the principal opening
zone of the pole. Pi was evaluated by determining the of an FRP pole for a given ratio (E I)/(L ) at the base
thickness of each layer using scanning electron micro- of the pole. The ultimate bending moment (Mu,o) was
scope. Figure 3 presents the curve of the ultimate bend- induced by the ultimate load (Fu). Figure 4 presents the
ing moment (Mu,b) at the base of an FRP pole for a prototypes that failed at the principal opening.
given ratio (E I)/(L ) at the base of the pole. The ulti- Equation (5) was obtained from the curve presented
mate bending moment (Mu,b) was induced by the in the Figure 4 with an R2 of 1.00.
Metiche and Masmoudi 213

Figure 5. Design curve – ultimate bending moment.

Figure 6. Design curve – pole top deflection.

EI earlier experimental results, which show that the use


Failure at the base : Mu,b ¼ 1496 ð4Þ of fibers with low linear density leads to better
L
mechanical performances.7 The low linear density
Failure at the principal opening (linear mass) of the fibers confers to them a better
 3  2   capacity of resin absorption at the wetting phase of
EI EI EI
Mu,o ¼ 13:9 390:59 þ2957:5 ð5Þ the filament winding process. The resin being better
L L L
impregnated makes it able to create bonds between
Figure 5 presents the prototypes that failed at the fibers and reduce the rate of voids giving thus better
base and the prototypes that failed at the principal woven and more resistant laminates. By the same
opening as well as the respective trend curves. manner, the design curves of the maximum deflec-
Figure 5 shows that for a ratio (E I)/(L ) greater tion (max) at the loading position of an FRE pole
than 23.5 kNm2/g, the failure occurs at the base of were determined for a given ratio (E I)/(L ) at the
the pole. Figure 5 shows that the ultimate moment base of the FRP pole. Figure 6 presents the proto-
at the base is inversely proportional to the linear types that failed at the base and the prototypes that
fiber mass . This result is in good agreement with failed at the principal opening as well as the
214 Journal of Composite Materials 47(2)

Figure 7. Design curve – ultimate longitudinal compression strain ("Cx,b) at the base.

Figure 8. Design curve – ultimate longitudinal tension strain ("Tx,b) at the base.

respective trend curves and R2. Equations (6) and (7) Figure 7 presents the ultimate longitudinal compression
were obtained from the curves presented in the strain ("Cx,b) at the base of an FRP pole for a given
Figure 6, respectively, for au failure at the base ratio (E I)/(L ) at the base of the pole. The ultimate
and a failure at the principal opening with an R2 bending moment (Mu,b) will be determined from the
of 0.89 and 0.96, respectively. curve of Figure 5 or using the equation (4). Figure 7
presents the prototypes that failed at the base. Equation
(8) was obtained from the curve presented in the
  Figure 7.
EI
Failure at the base : max ¼ 715:07Ln
L ð6Þ
þ 187:79  3  2
"C
x,b EI EI
Failure at the base : ¼ 0:054 4:6647
Mu,b L L
Failure at the principal opening  
EI
 2   ð7Þ
þ 129:13  1342:7
EI EI L
max ¼ 16:093 396:2 þ 2640:7
L L ð8Þ
Metiche and Masmoudi 215

Figure 9. FRP pole, dimensions, and geographical location.16 FRP: fiber-reinforced polymer.

Figure 8 presents the ultimate longitudinal tension the base of the pole. The few existing design guidelines
strain ("Tx,b) at the base of an FRP pole for a given ignore such effect and accounts only for the effect of
ratio (E I)/(L ) at the base of the FRP pole. these holes by considering their influence on the
The ultimate bending moment (Mu,b) will be deter- abrupt reduction in the cross-sectional area of the
mined from the curve of Figure 5 or using equation poles. These guidelines do not give a specific attention
(4). Figure 8 presents the prototypes that failed at the to the impact of the opening on the generated stress
base. Equation (9) was obtained from the curve pre- concentrations in their vicinity. Furthermore, local
sented in the Figure 8. buckling at a nearby area of the opening generally
dominates the mode of failure of such poles that
 3  2
"Tx,b EI EI requires more attention to the relative locations of
Failure at the base : ¼ 0:0483 þ4:5078 the openings. The new proposed design verification
Mu,b L L
  procedure is introduced to provide a comprehensive
EI
 136:04 þ 1477:1 ð9Þ method of design verification based on the synthesis
L
of existing approaches, specifications, and guidelines
and to accurately evaluate the available design equa-
The design curves presented in Figures 7 and 8 show tions for the flexural behavior of FRP-composite
that the ration ("Cx,b)/(Mu,b) (respectively the ratio poles. The results of the verification of the current
("Tx,b)/(Mu,b)) decreases when increasing the ratio (E design of the tested prototypes 33-B-5-C-(1), 35-B-5-
I)/(L ) for values of (E I)/(L ) less than 17 kNm2/g. C-(1), and 40-B-5-C-(1) according to the AASHTO
For the values of (E I)/(L ) greater than 17 kNm2/g standard15 are presented in this section. The procedure
the ration ("Cx,b)/(Mu,b) (respectively the ratio ("Tx,b)/ remains the same for the other prototypes presented in
(Mu,b)) is almost constant. this paper. The assumed geographical location for the
implementation of these FRP poles was Montreal
(Quebec), Canada, implying a wind pressure of refer-
Design verification procedures ence (q) equal to 365 Pascal, and a thickness of accu-
The very few available design procedures of FRP mulated ice equal to 31 mm,16 the considered density
poles are based on the allowable stress design theory of ice was 9800 N/m3. The FRP prototypes 33-B-5-C-
of composite material under various states of stress. (1), 35-B-5-C-(1), and 40-B-5-C-(1) were supposed to
However, various experimental results show that support a light fixture placed at the pole top
cracking and early failure of FRP poles are normally (Figure 9). The geometrical, physical, and mechanical
controlled by the relative location of the opening from properties of the poles and light fixture considered for
216 Journal of Composite Materials 47(2)

Table 5. Mechanical, physical and geometrical properties at level (0) (ground level)

Prototypes

Symbol Unit 33-B-5-C-(1) 35-B-5-C-(1) 40-B-5-C-(1)

Physical and geometrical properties


Total height hT mm 10058 10566 12090
Embedded height he mm 1524 1524 1524
Cantilever height hc mm 8534 9042 10566
Wall thickness t mm 4.31 3.87 5.20
External diameter D mm 248 255 279
Internal radius rint mm 120 124 134
External radius R1 mm 124 128 139
Cross section area x 1003 A mm2 3.3 3.1 4.5
07 4
Moment of inertia x 10 I mm 2.44 2.40 4.18
Radius of gyration r mm 86 89 97
Width of light fixture bL mm 400 400 400
Outer area of the pole (tube) x 1006 Aext.P mm2 4.73 5.13 6.34
06 2
Outer area of light fixture x 10 Aext.L mm 0.96 0.96 0.96
Projected area of the pole (tube) x 1006 AP mm2 1.63 1.76 2.17
Projected area of light fixture x 1006 AL mm2 0.79 0.79 0.79
Projected area of ice around the pole x 1006 AGP mm2 0.53 0.56 0.66
Projected area of ice around light fixture x 1006 AGL mm2 0.05 0.05 0.05
Centroid height of the pole (tube) h1 mm 3760 3970 4570
Centroid height of light fixture h2 mm 9030 9540 11070
Weight of pole (tube) MP kg 66.7 68.6 96.2
Weight of the light fixture ML kg 15 15 15
Weight of ice on pole (tube) WGP N 1438 1558 1926
Weight of ice on light fixture WGL N 292 292 292
Mechanical properties
Experimental failure load Fu N 4458 3465 4293
Ultimate bending moment Mu N.m 36688 30274 44051
Longitudinal modulus of elasticity E1 GPa 16.24 13.64 15.07
Transversal modulus of elasticity E2 GPa 19.75 16.38 11.23
Shear modulus G GPa 4.01 3.42 4.00
Modulus of elasticity in compression Ec GPa 26.00 23.00 21.00
Ultimate bending stress Fbu MPa 331 290 275
Ultimate compression stress Fau MPa 181 190 169
Ultimate tension stress Ftu MPa 185 164 139
Ultimate shear stress Fvu MPa 185 178 140
Effective buckling length factor k – 2 2 2
Poisson’s ratio in the longitudinal direction 12 – 0.30 0.27 0.34
Constant relating Poisson’s ratio  – 0.89 0.91 0.91
Orthotropy factor K1 – 0.85 0.84 0.77

the design verification at (0) level (ground level) and at by pyrolysis tests and scanning electron microscope
the service opening are presented in Tables 5 and 6, analysis using micromechanical models.10
respectively. The mechanical properties (Ec, Fbu, Fau,
Ftu, Fvu) presented in Tables 5 and 6 were determined
Design assumptions
by mechanical tests performed on specimens cut-out
from the FRP poles.8,17 The mechanical properties The design of FRP members according to AASHTO
(E1, E2, G) presented in Tables 5 and 6 were evaluated standard15 is based on the allowable stress design
Metiche and Masmoudi 217

Table 6. Mechanical, physical, and geometrical properties at the service opening level

PROTOTYPES

Symbol Unit 33-B-5-C-(1) 35-B-5-C-(1) 40-B-5-C-(1)

Physical and geometrical properties


Total height hT mm 10058 10566 12090
Embedded height he mm 1524 1524 1524
Cantilever height hc mm 8534 9042 10566
Wall thickness t mm 9.20 11.32 11.59
External diameter D mm 239 252 274
Internal radius rint mm 111 115 126
External radius R1 mm2 120 126 137
Cross section area x 1003 A mm 2.8 3.6 4.1
Moment of inertia x 1007 I mm4 2.94 4.61 5.86
Radius of gyration r mm 81 85 93
Width of light fixture bL mm 400 400 400
Outer area of the pole (tube) x 1006 Aext.P mm2 4.73 5.13 6.34
06 2
Outer area of light fixture x 10 Aext.L mm 0.96 0.96 0.96
Projected area of the pole (tube) x 1006 AP mm2 1.63 1.76 2.17
Projected area of light fixture x 1006 AL mm2 0.79 0.79 0.79
Projected area of ice around the pole x 1006 AGP mm2 0.53 0.56 0.66
Projected area of ice around the light fixture x 1006 AGL mm2 0.05 0.05 0.05
Centroid height of the pole (tube) h1 mm 3760 3970 4570
Centroid height of light fixture h2 mm 9030 9540 11070
Weight of pole (tube) MP kg 66.7 68.6 96.2
Weight of the light fixture ML kg 15 15 15
Weight of ice on pole (tube) WGP N 1438 1558 1926
Weight of ice on light fixture WGL N 292 292 292
Mechanical properties
Experimental failure load Fu N 4458 3465 4293
Ultimate bending moment Mu N.m 31252 26051 38819
Longitudinal modulus of elasticity E1 GPa 14.49 12.28 13.35
Transversal modulus of elasticity E2 GPa 32.60 36.27 33.87
Shear modulus G GPa 3.70 3.74 3.65
Modulus of elasticity in compression Ec GPa 22.00 20.00 20.00
Ultimate bending stress Fbu MPa 202 237 257
Ultimate compression stress Fau MPa 210 195 207
Ultimate tension stress Ftu MPa 134 114 113
Ultimate shear stress Fvu MPa 121 102 109
Effective buckling length factor k - 2 2 2
Poisson’s ratio in the longitudinal direction 12 - 0.17 0.13 0.16
Constant relating Poisson’s ratio  - 0.93 0.95 0.94
Orthotropy factor K1 - 0.98 1.13 1.05

approach. FRP materials are generally anisotropic fibers, the material may be considered orthotropic
in character with varying directional properties. or isotropic.15 Although FRP qualifies as a viscoelas-
The mechanical properties vary according to the tic material, which is temperature and time depen-
fiber content, orientation of the fibers, and the dent, its behavior can be considered as linearly
mechanical characteristics of the fibers and resin. elastic that obeys Hooke’s law. A basic assumption
Depending upon the particular placement of the is that ‘‘plane sections remain plane after bending.’’
218 Journal of Composite Materials 47(2)

This is primarily because of the fairly linear stress– Since the fiber orientation changes along the
strain behavior up to failure.7,10,15 prototypes, the samples were taken at locations of
critical stresses. The proposed safety factors shown in
Table 8 are minimum values based on common indus-
Group load combinations
try practice.15
The loads acting on the structure were combined into
appropriate group load combinations15 as shown in
Table 7. Each part of the structure shall be propor-
Evaluation of the maximum wind load
tioned for the combination producing the maximum The wind loads acting on the pole (the tube), the light
effect, using allowable stresses as indicated for the fixture, and on the surface of the ice around them were
group load. The given percentages of allowable stress evaluated according to CAN/CSA-S6-0616 and CAN/
are applicable for the allowable stress design method. CSA-S37-0118 standards. Equation (10) was used16 to
No load reduction factors were applied in conjunction evaluate the maximum wind load (Fw) where (q) is the
with these increased allowable stresses.15 wind pressure of reference (Pa), (Ce) is the exposure
factor, (Cg) is the gust factor, (Ch) is the drag factor,
Mechanical properties and safety factors of the FRP and (Ap) is projected area in (m2).
material Fw ¼ q Ce Cg Ch Apr ðNÞ ð10Þ
The mechanical properties of FRP laminates were
determined by testing according to ASTM standards.
Verification of the maximum bending moment and
Table 7. Group load combinations the maximum pole top deflection
Percent of The FRP pole shall be capable of sustaining a mini-
Group Load combinations allowable stress mum strength equal to two times the maximum bend-
I DL 1.00 ing strength induced by the wind. The bending strength
II DL þW 1.33 of FRP poles shall be determined in accordance to
ASTM D492311 procedure. A safety factor of 2.0
III DL þ Ice þ 1/2(W) 1.33
against failure in bending is specified for the test by
AASHTO LTS-5 standard15 in order to account for
the variability in mechanical properties of FRP mate-
Table 8. Safety factors for different mechanical properties rials. The maximum pole top deflection of an FRP pole
shall not exceed 15% of the pole height above ground
Minimal
when the pole is submitted to the maximum wind loads.
safety
Property Symbol Unit Standard factor Table 9 and Figures 10 to 12 show a comparison
between the maximum bending moment induced by
Bending strength Fbu MPa ASTM D 790 2.5 the maximum wind loads (Mw) and the allowable bend-
Tensile strength Ftu MPa ASTM D 638 2.0 ing moment (Ma), (Ma ¼ Mu/2), respectively, for the
Compressive Fau MPa ASTM D 695 3.0 prototypes 33-B-5-C-(1), 35-B-5-C-(1), and 40-B-5-C-
strength (1). The ultimate bending moment (Mu) was computed
Shear strength Fvu MPa ASTM D 732 3.0 from the experimental failure load (Fu) presented in
Table 5, the bending moment induced by wind loads

Table 9. Verification of the maximum bending moment

Prototypes

Symbol Unit 33-B-5-C-(1) 35-B-5-C-(1) 40-B-5-C-(1)

Maximum bending Mw N.m 17138 18794 23887


moment (max. wind loads)
Ultimate bending moment Mu N.m 36688 30274 44051
Allowable bending moment Ma ¼ Mu/2 N.m 18344 15137 22026
Ratio (Mw/Ma) Mw/Ma – 0.93 1.24 1.08
Comments – – Safe, economic unsafe unsafe
Metiche and Masmoudi 219

50000

Allowable deflection (Δ a)
PROTOTYPE 33-B-5-C-(1)

40000
Mu
Bending moment at the base (N.m)

Mu (36688 N.m)

30000
EXPERIMENTAL CURVE
MAXIMUM WIND LOAD

20000 Allowable bending moment (Mu/2)


Mu/2 (18344 N.m)
Mw = 17138 N.m Safety factor = 2
Δw = 997 mm

Δ a (1280 mm)
10000

0
0 500 1000 1500 2000 2500 3000
Pole top deflection (mm)

Figure 10. Verification of the maximum bending moment and the maximum pole top deflection – experimental curve – prototype
33-B-5-C-(1).

50000
Allowable deflection (Δ a)

PROTOTYPE 35-B-5-C-(1)

40000
Bending moment at the base (N.m)

Mu
30000 Mu (30274 N.m)

MAXIMUM WIND LOAD


EXPERIMENTAL CURVE

20000 Mw = 18794

Allowable bending moment (Mu/2)


Mu/2 (15137 N.m)
Safety factor = 2
Δw = 1180 mm

Δ a (1356 mm)

10000

0
0 500 1000 1500 2000 2500 3000
Pole top deflection (mm)

Figure 11. Verification of the maximum bending moment and the maximum pole top deflection – experimental curve – prototype
35-B-5-C-(1).
220 Journal of Composite Materials 47(2)

50000
PROTOTYPE 40-B-5-C-(1)

Allowable deflection (Δ a)
Mu
Mu (44051 N.m)

40000
Bending moment at the base (N.m)

EXPERIMENTAL CURVE
MAXIMUM WIND LOAD
30000

Mw = 23887
Allowable bending moment (Mu/2)
Mu/2 (22026 N.m)
20000
Safety factor = 2

Δw = 1392 mm

Δ a (1585 mm)
10000

0
0 500 1000 1500 2000 2500 3000
Pole top deflection (mm)

Figure 12. Verification of the maximum bending moment and the maximum pole top deflection – experimental curve – prototype
40-B-5-C-(1).

Table 10. Verification of the maximum pole top deflection

Prototypes

Symbol Unit 33-B-5-C-(1) 35-B-5-C-(1) 40-B-5-C-(1)

Maximum pole top deflection (max. wind loads) w mm 997 1180 1392
Allowable pole top deflection a ¼ 15% hc mm 1280 1356 1585
Ratio (w/a) w/a – 0.78 0.87 0.88
Comments Safe Safe Safe

(Mw) was calculated using the equation (11). Bending induced by wind loads (Mw), (Dw) is read directly
moments (Mu) and (Mw) were evaluated at the base of from the experimental curves (moment at the base –
the pole (level 0, Figure 9). pole top deflection; Figures 10 to 12).
Table 9 and Figure 10 show that the bending
Mw ¼ Fw1 h1 þ Fw2 h2 ð11Þ moment induced by the maximum wind loads (Mw) of
the prototype 33-B-5-C-(1) is below the allowable bend-
ing moment (Ma), Table 10 and Figure 10 show that the
Table 10 and Figures 10 to 12 show a comparison experimental deflection under the maximum wind loads
between the maximum allowable deflection (a) and (w) of the prototype 33-B-5-C-(1) does not exceed the
the experimental deflection under the maximum wind allowable deflection (a). This indicates a safe and eco-
loads (W), respectively, for the prototypes 33-B-5-C- nomical design of the prototype 33-B-5-C-(1).
(1), 35-B-5-C-(1), and 40-B-5-C-(1). The maximum Although the deflection (w) of the prototype
allowable deflection (a) is equal to 15% of the canti- 35-B-5-C-(1) (Table 10 and Figure 11) and the proto-
lever height of the pole, the experimental deflection type 40-B-5-C-(1) (Table 10 and Figure 12) does not
under the maximum wind loads (w) is the value of exceed the allowable deflection (a), the bending
the deflection corresponding to the bending moment moment induced by the maximum wind loads (Mw) of
Metiche and Masmoudi 221

Table 11. Verification of compression stress–local buckling

Case 1: Load combination I (DL)

Prototypes

Symbol Unit 33-B-5-C-(1) 35-B-5-C-(1) 40-B-5-C-(1)

Level (0) (ground level)


Applied compression stress fa MPa 0.25 0.27 0.25
Allowable compression stress Fa MPa 48 35 43
Ultimate compression stress divided Fau/n; (n ¼ 3) MPa 60 63 56
by the safety factor
Service opening level
Applied compression stress fa MPa 0.29 0.23 0.27
Allowable compression stress Fa MPa 107 122 116
Ultimate compression stress divided Fau/n; (n ¼ 3) MPa 70 65 69
by the safety factor
Case 2: Load combination III (DL þ Ice)
Level (0) (ground level)
Applied compression stress fa MPa 0.77 0.88 0.75
Allowable compression stress Fa MPa 48 35 43
Increased allowable compression stress 1.33 Fa MPa 64 46 57
Ultimate compression stress divided Fau/n; (n ¼ 3) MPa 60 63 56
by the safety factor
Service opening level
Applied compression stress fa MPa 0.72 0.60 0.68
Allowable compression stress Fa MPa 107 122 116
Increased allowable compression stress 1.33 Fa MPa 143 162 154
Ultimate compression stress divided Fau/n; (n ¼ 3) MPa 70 65 69
by the safety factor

the prototypes 35-B-5-C-(1) and 40-B-5-C-(1) exceeds "  12 !  12  #12
the allowable bending moment (Ma) (Table 9 and E2 E2 G
With K1 ¼ 1:414 1 þ 12 ð13Þ
Figures 11 and 12). This indicates that the poles 35-B- E1 E1 E1
5-C-(1) and 40-B-5-C-(1) require a new design to satisfy  
E2
the condition of allowable bending moment. And  ¼ 1  212 ð14Þ
E1
Table 11 presents a comparison of the applied com-
Stress verification pressive stress (fa) compared with the lowest value
between the allowable compressive stress considering
The allowable stresses for tubular FRP poles with cir- local buckling (Fa) and the ultimate compressive
cular cross section are given by the AASHTO stan- stress divided by the safety factor (Fau/n) for the poles
dard.15 The applied stresses were determined for a 33-B-5-C-(1), 35-B-5-C-(1), and 40-B-5-C-(1) using
tubular FRP pole with circular cross-section supporting equation (12). The compressive stress (fa) applied to
a light fixture (Figure 9, Tables 5 and 6). the pole’s cross section was calculated from the sum
of the dead load of the pole, the light fixture, and the
Compression stress – local buckling. The local buckling weight of the ice around them (Figure 9 and Tables 5
shall be considered in the design of relatively thin and 6). The allowable compressive stress (Fa) consider-
wall FRP poles,19 the allowable compressive stress con- ing local buckling was calculated using equation (12).
sidering local buckling was calculated as follows15: The ultimate compressive stress (Fau) determined exper-
imentally was presented in Tables 5 and 6, respectively,
for the level (0) (ground level) and the service opening
0:57 E1 K1 Fau
Fa ¼    ð12Þ level. Two cases of load combinations were considered:
D 1 n load combinations I (Case 1) and load combinations III
n 2
t (Case 2).
222 Journal of Composite Materials 47(2)

Table 12. Verification of compression stress – flexural buckling

Case 1: Load combination I (DL)

Prototypes

Symbol Unit 33-B-5-C-(1) 35-B-5-C-(1) 40-B-5-C-(1)

Level (0) (ground level)


Applied compression stress fa MPa 0.25 0.27 0.25
Allowable compression stress Fa MPa 2.17 1.82 1.45
Service opening level
Applied compression stress fa MPa 0.29 0.23 0.27
Allowable compression stress Fa MPa 1.65 1.45 1.27
Case 2: Load combination III (DL þ Ice)
Level (0) (ground level)
Applied compression stress fa MPa 0.77 0.88 0.75
Allowable compression stress Fa MPa 2.17 1.82 1.45
Increased allowable compression stress 1.33 Fa MPa 2.89 2.42 1.92
Service opening level
Applied compression stress fa MPa 0.72 0.60 0.68
Allowable compression stress Fa MPa 1.65 1.45 1.27
Increased allowable compression stress 1.33 Fa MPa 2.19 1.93 1.69

sffiffiffiffiffiffiffiffiffiffiffiffiffi
Table 11 shows that at the service opening level, in
kL 22 Ec
both load combinations cases, the allowable compres- For 
sive stress (Fa) is over than the ultimate compressive r Fau
stress divided by the safety factor (Fau/n). According 2 Ec ð16Þ
Fa ¼  2
to equation (12), the design will be based on the value kL
of (Fau/n) at the service opening level. n
r
Table 11 shows that at the level (0), the applied com-
pressive stress (fa) is much less than the allowable com- Table 12 shows a comparison of the applied com-
pressive stress (Fa) in the case of load combination I. In pressive stress (fa) compared to the allowable compres-
the case of load combination III, the applied compres- sive stress considering flexural buckling (Fa) for the
sive stress (fa) is much less than the increased allowable poles 33-B-5-C-(1), 35-B-5 -C-(1), and 40-B-5-C-(1).
compressive stress (1.33 Fa). Table 11 shows that at the The compressive stress (fa) applied to the pole’s cross-
service opening level, the applied compression stress (fa) section was calculated from the sum of the pole’s dead
is much less than the ultimate compressive stress load, the light fixture, and the weight of the ice around
divided by the safety factor (Fau/n) in both load com- them (Figure 9 and Tables 5 and 6). The allowable
bination cases. This indicates a secure but non-eco- compressive stress considering flexural buckling (Fa)
nomic design regarding the local buckling due to the was calculated using equation (16). Two cases of load
compressive stresses. combinations were considered: load combinations I
(Case 1) and load combinations III (Case 2). Table 12
shows that for load combination I the applied compres-
Compression stress – flexural buckling. Compression design sive stress (fa) is less than the allowable compressive
of FRP members is generally controlled by buckling. stress (Fa) and for load combination III the applied
Flexural buckling and local buckling should be consid- compressive stress (fa) is less than the increased allow-
ered. The allowable compression stress considering flex- able compressive stress (1.33 Fa) whether at the level (0)
ural buckling was calculated as follows15: or at the service opening level. This indicates a safe
sffiffiffiffiffiffiffiffiffiffiffiffiffi design regarding the flexural buckling due to the com-
kL 22 Ec pressive stresses.
For 5
r Fau
  ð15Þ
3
Bending stress. For thin-walled FRP sections, local
Fau F2au kL
Fa ¼  pffiffiffiffiffiffiffiffiffiffiffiffiffi buckling is a major parameter that controls the strength
n 2n 22 Ec r of the member in bending. The allowable bending stress
Metiche and Masmoudi 223

Table 13. Verification of bending stress

Load combination II (DL þW)

Prototypes

Symbol Unit 33-B-5-C-(1) 35-B-5-C-(1) 40-B-5-C-(1)

Level (0) (ground level)


Applied bending stress fb MPa 87 100 80
Allowable bending stress Fb MPa 77 55 68
Increased allowable bending stress 1.33 Fb MPa 102 73 90
Ultimate bending stress divided by the safety factor Fbu/n; (n ¼ 2.5) MPa 132 116 110
Service opening level
Applied bending stress fb MPa 59 44 49
Allowable bending stress Fb MPa 169 192 183
Increased allowable bending stress 1.33 Fb MPa 225 256 243
Ultimate bending stress divided by the safety factor Fbu/n; (n ¼ 2.5) MPa 81 95 103

is defined as a function of the critical buckling stress of 33-B-5-C-(1), 35-B-5-C-(1), and 40-B-5-C-(1) using
the section. Equations to obtain the critical buckling equation (17). The applied bending stress (fb) was cal-
stress are based on the plate theory for orthotropic ele- culated from the bending moment due to wind load.
ments, and they are expressed in terms of the aspect The allowable bending stress (Fb) was calculated by
ratio (D/t) of the pole where (D) and (t) are, respec- the equation (17). The ultimate bending stress (Fbu)
tively, the external diameter and the thickness of the determined experimentally was presented in Tables 5
cylinder. It has been shown that the critical compressive and 6, respectively, for the level (0) (ground level) and
stress caused by bending is 30% higher than the critical the service opening level. The load combination II
compressive stress caused by axial compressive loads (D þ W) was considered implying an increase in the
for round tubular sections.15 Therefore, the critical allowable bending stress of 33%. Table 13 shows that
buckling stress for round tubular member under bend- at the service opening, the allowable bending stress (Fb)
ing is taken as 1.3 times the critical buckling stress for is over than the ultimate bending stress divided by
round tubular member under axial compression. the safety factor (Fbu/n). According to equation (17),
Hence, the allowable bending stress for round tubular the design at service opening will be governed by the
sections was calculated as follows15: value of (Fbu/n).
0:75 E1 K1 Fbu Table 13 shows that at the level (0) the applied bend-
Fb ¼    ing stress (fb) is less than the increased allowable bend-
D 1 n ð17Þ
n 2 ing stress (1.33 Fb) except for the prototype 35-B-5-C-
t (1) and at the service opening level the applied bending
Where stress (fb) is less than the ultimate bending stress divided
" by the safety factor (Fbu/n) for all the prototypes. This
 12 ! 12  #12
E2 E2 G indicates a safe design regarding the bending stresses of
K1 ¼ 1:414 1 þ 12
E1 E1 E1 all the prototypes except at level (0) of the prototype
35-B-5-C-(1).
And Table 5 shows that the thickness (t) of the prototype
  35-B-5-C-(1) at the level (0) is equal to 3.87 mm, a pos-
E2
¼1 212 sible solution to satisfy equation (17) would be to
E1 increase this thickness to 4.60 mm, which will decrease
The applied bending stress was calculated as follows: the applied bending stress (fb) to 84 MPa and increase
the increased allowable bending stress (1.33 Fb) to
Mw
fb ¼ y ð18Þ 86 MPa, which satisfies equation (17). Another possible
I solution to satisfy equation (17) would be to increase
Table 13 presents a comparison of the applied bend- the tensile modulus in the longitudinal direction (E1) by
ing stress (fb) compared with the lowest value between orienting the fibers toward the longitudinal direction,
the allowable bending stress (Fb) and the ultimate bend- which will further increase the value of the allowable
ing stress divided by the safety factor (Fbu/n) for poles bending stress (see equation (17)).
224 Journal of Composite Materials 47(2)

Table 14. Verification of shear stress

Load combination II (DL þW)

Prototypes

Symbol Unit 33-B-5-C-(1) 35-B-5-C-(1) 40-B-5-C-(1)

Level (0) (ground level)


Applied shear stress (transverse loads) fvb MPa 1.99 2.26 1.70
Applied shear stress (torsion loads) fvt MPa 0.14 0.14 0.10
Allowable shear stress Fv 6.76 4.12 7.46
Increased allowable shear stress 1.33 Fv MPa 8.99 5.49 9.92
Ultimate shear stress divided by the safety factor Fvu/n; (n ¼ 3) MPa 62 59 47
Service opening level
Applied shear stress (transverse loads) fvb MPa 2.34 1.92 1.85
Applied shear stress (torsion loads) fvt MPa 0.07 0.05 0.05
Allowable shear stress Fv 19.43 23.86 20.50
Increased allowable shear stress 1.33 Fv MPa 25.84 31.73 27.26
Ultimate shear stress divided by the safety factor Fvu/n; (n ¼ 3) MPa 40 34 36

Tensile stress. In the case studied and shown in Figure 9,


there are no tensile stresses acting on the pole inducing And e ¼ 0:15 bL ð24Þ
axial tensile loads, therefore, this verification was not
performed. The applied shear stress due to transverse loads (Fvb)
and the applied shear stress due to torsion (Fvt) calcu-
Shear stress. The allowable shear stress of the FRP poles lated, respectively, by equations (20) and (22) for each
subjected to transverse loads or torsion was calculated pole were compared to the lowest value between the
as follows15: allowable shear stress (Fv) and the shear stress divided
   32 by the ultimate safety factor (Fvu/n) using equation
1 0:533Ftu ð1 þ 12 Þ G t Fvu (19). The allowable shear stress (Fv) was calculated by
Fv ¼  ð19Þ
n  Fvu R1 n equation (19), the ultimate shear stress (Fvu) determined
experimentally is shown in Tables 5 and 6, respectively,
With for the level (0) (ground level) and the service opening
  level. The load combination II (D þ W) was considered
E2
¼1 212 implying an increase in the allowable shear stress of
E1 33%. Table 14 shows that at the level (0) and at the
The applied shear stress due to transverse loads was service opening level the allowable shear stress (Fv) is
calculated as follows15: less than the ultimate shear stress divided by the safety
factor (Fvu/n), according to equation (6.19), the shear
2 VS design will be governed by the allowable shear stress
fvb ¼ ð20Þ
A (Fv). Table 14 shows that the applied shear stress due to
transverse loads (Fvb) and the applied shear stress due
With Vs ¼ Fw1 þ Fw2 ð21Þ to torsion (Fvt) are below the increased allowable shear
stress (1.33 Fv). This indicates a safe design regarding
The applied shear stress due torsion loads was cal- the shear stresses.
culated as follows15:
Combined bending and compression stresses. Equations (25)
Mz and (27) were used to check combined bending and
fvt ¼ ð22Þ
6:28 R2 t compression stresses.15 Equation (25) is intended for
intermediate unbraced locations where the member is
With Mz ¼ Fw2 e ð23Þ susceptible to lateral displacement. Equation (27) is
Metiche and Masmoudi 225

Table 15. Verification of combined bending and compression stresses.

Case 1: Load combination II (DLþW)

Prototypes

Unit 33-B-5-C-(1) 35-B-5-C-(1) 40-B-5-C-(1)

Level (0) (Ground level)

fa fb
þ  1:0 (see Equation 27) MPa 0.86 1.37 0.89
1:33 Fa 1:33 F b

Service opening level

fa fb
þ
 1:0 (see Equation 25) MPa 0.32 0.21 0.26
1:33 Fa 1:33 Fb 1  fa0
Fe

Case 2: Load combination III (DL þ Ice + 1/2 W)


Level (0) (Ground level)

fa fb
þ  1:0 (see Equation 27) MPa 0.44 0.70 0.45
1:33 Fa 1:33 F b

Service opening level

fa 1=2fb
þ
 1:0 (see Equation 25) MPa 0.24 0.15 0.22
1:33 Fa 1:33 Fb 1  fa0
Fe

intended for locations at the end of the member where 1.0. This indicates a safe design regarding the combined
lateral displacement is restrained where (Fa) is the compression and bending stresses of the poles presented
allowable compression stress due to local buckling. in Table 15 except at the level (0) of the prototype
35-B-5-C-(1). Table 5 shows that the thickness (t) of
fa f the prototype 35-B-5-C-(1) at the level (0) is equal to
þ  b   1:0 ð25Þ
Fa fa 3.87 mm, a possible solution to satisfy equation (6.27)
Fb 1  0 would be to increase this thickness to 4.60 mm, hence
Fe
equation (6.27) will yield a value of 0.97 lower than 1.0,
2 Ec which satisfy equation (6.27). Another possible solution
With F0e ¼   ð26Þ to satisfy equation (6.27) would be to increase the ten-
kb L 2
n sile modulus in the longitudinal direction (E1) by ori-
rb enting the fibers toward the longitudinal direction. This
will further increase the allowable compressive and
fa fb
And þ  1:0 ð27Þ bending stresses and hence allow satisfying the equa-
Fa Fb tion (6.27).

Table 15 presents the results of the combined bend- Combined bending, compression and shear
ing and compression stresses verification of the proto- stresses. Equation (28) was used to verify the vertical
types 33-B-5-C-(1), 35-B-5-C-(1), and 40-B-5-C-(1). cantilever FRP poles subjected to axial compression,
Equation (27) was used at the level (0) (ground level) bending moment, and shear loads. Equation (28) is
and equation (25) was used at the opening level. Two given by AASHTO standard15 for the design aluminum
cases of load combinations were considered: load com- poles.
binations II and load combinations III implying in both  2
cases an increasing of the allowable stresses of 33%. fa fb fv
For hc  15 m þ þ  1:0 ð28Þ
Table 15 shows that in both cases of load combinations, Fa Fb Fv
the poles 33-B-5-C-(1), 35-B-5-C-(1), and 40-B-5-C-(1) Table 16 presents the results of the combined bend-
satisfy equations (6.25) and (6.27) except at the level (0) ing, compression, and shear stresses verification of the
of the prototype 35-B-5-C-(1) for the load combination prototypes 33-B-5-C-(1), 35-B-5-C-(1), and 40-B-5-C-
II, where equation (27) yields a value of 1.37 over than (1) using equation (28). Two cases of load combinations
226 Journal of Composite Materials 47(2)

Table 16. Verification of combined bending, compression and shear stresses.

Case 1: Load combination II (DLþW)

Prototypes

Unit 33-B-5-C-(1) 35-B-5-C-(1) 40-B-5-C-(1)

Level (0) (Ground level)


 2
fa fb fvb (see Equation 28) MPa 0.91 1.54 0.91
þ þ  1:0
1:33 Fa 1:33 F b 1:33Fv

Service opening level


 2
fa fb fvb (see Equation 28) MPa 0.27 0.18 0.21
þ þ  1:0
1:33 Fa 1:33 F b 1:33Fv

Case 2: Load combination III (DL þ Ice þ 1/2 W)


Level (0) (Ground level)
 
fa 1=2fb 1=2fvb 2 (see Equation 28) MPa 0,45 0.74 0.46
þ þ  1:0
1:33 Fa 1:33 F b 1:33Fv

Service opening level


 
fa 1=2fb 1=2fvb 2 (see Equation 28) MPa 0.14 0.09 0.11
þ þ  1:0
1:33 Fa 1:33 F b 1:33Fv

were considered: load combinations II (case 1) and load


combinations III (case 2) implying in both cases an
Conclusion
increasing of the allowable stresses of 33%. (fa) and Design approaches for conventional steel/wood poles
(Fa) are, respectively, the applied compressive stress were influential in providing safety and robustness to
and the allowable compressive stress for local buckling. these poles. However, little advanced approaches
(fb) and (Fb) are, respectively, the applied bending stress focused on the design of FRP poles. This is the focus
and the allowable bending stress. (Fvb) and (Fv) are, of the present paper. The available design procedures of
respectively, the applied shear stress due to transverse FRP poles are based on the allowable stress design
loads and the allowable shear stress. In equation (28), theory of composite material under various states of
only the applied shear stress due to transverse loads stress. However, various experimental results show
(Fvb) was considered as this gives a worst case calcula- that cracking and early failure of FRP poles are nor-
tion. Table 16 shows that in both cases of load combi- mally controlled by the relative location of the opening
nations, the poles 33-B-5-C-(1), 35-B-5-C-(1), and 40-B- from the base of the pole. Most of the design guidelines
5-C-(1) satisfy equation (28) except at the level (0) of ignore such effect and accounts only for the effect of
the prototype 35-B-5-C-(1) for the load combination II these holes by considering their influence on the abrupt
where equation (28) yields a value of 1.54 over than 1.0. reduction in the cross-sectional area of the poles. These
This indicates a safe design regarding the combined guidelines do not give a specific attention to the impact
compression, bending, and shear stresses of the poles of the opening on the generated stress concentrations in
presented in Table 16 except at the level (0) of the pro- their vicinity. Furthermore, local buckling at a nearby
totype 35-B-5-C-(1). Table 5 shows that the thickness area of the opening generally dominates the mode of
(t) of the prototype 35-B-5-C-(1) at the level (0) is equal failure of such poles that requires more attention to
to 3.87 mm, a possible solution to satisfy equation the relative locations of the opening. A new design
(6.28) would be to increase this thickness to 4.80 mm, approach is introduced to advance in the design of
hence equation (28) will yield a value of 0.99 lower than FRP poles. The accuracy of both the developed design
1.0, which satisfy equation (6.28). procedures and that of existing design approaches are
Metiche and Masmoudi 227

verified by comparison with documented test results of FRP poles section of the next AASHTO LTS stan-
an early experimental program. Different types of FRP dard edition.
poles, having different geometrical properties and made . The experimental deflection corresponding to the
of two different types of glass fibers, were subjected to maximum wind loads (w) does not exceed the
full-scale flexural static testing. Each type of the poles allowable deflection (a).
tested in this study is constituted by three zones where . The bending moment induced by the maximum wind
the geometrical and the mechanical properties are differ- loads (Mw) of the prototypes 35-B-5-C-(1) and 40-B-
ent in each zone. The difference of these properties is due 5-C-(1) exceeds the allowable bending moment (Ma).
to the different number of layers used in each zone and This indicates that these prototypes require a new
the fiber orientation of each layer. The mechanical prop- design to satisfy the condition of the allowable bend-
erties, the fiber, and the matrix volume ratios as well the ing moment.
stacking sequences of the different pole’s zones were . The verified prototypes meet the limitations of the
determined in this investigation by mechanical tests per- AASHTO LTS-5-2009 standard except at the base
formed on specimens cut out from the FRP poles and/or (level 0) of the prototype 35-B-5-C-(1) for some
evaluated by pyrolysis tests and scanning electron stress and load combinations cases. Solutions have
microscope analysis using micromechanical models. A been proposed in this paper to meet the requirements
design verification analysis according to the AASHTO of AASHTO standard.
LTS-5-2009 standard of the full-scale manufactured and
tested GFRP poles was performed and presented in this The contribution of this research work lies mainly in
paper. the characterization of new FRP composite poles and
The following conclusions can be drawn: describes a new design approach to advance in the
design of FRP poles.
. For a ratio (E I)/(L ) greater than 23.5 kNm2/g, the
failure occurs at the base of the pole. Funding
. The ultimate moment at the base is inversely propor- The research reported in this paper was partially sponsored
tional to the linear fiber mass . This conclusion is in by Hydro-Québec (Distribution-Division, Montreal, Qc) and
good agreement with documented test results of an the Natural Sciences and Engineering Research Council of
early experimental program of the authors. Canada (NSERC, Ottawa, Canada).
. The ratio ("Cx,b)/(Mu,b) (respectively, the ratio
("Tx,b)/(Mu,b)) decreases when increasing the ratio Conflict of Interest
(E I)/(L ) for values of (E I)/(L ) less than The authors declare that they do not have any conflict of
17 kNm2/g. For the values of (E I)/(L ) over than interest.
17 kNm2/g the ratio ("Cx,b)/(Mu,b) (respectively, the
ratio ("Tx,b)/(Mu,b)) is almost constant. Acknowledgements
The authors acknowledge the contribution of the Canadian
The stress verifications performed according to the Foundation for Innovation (CFI) for the infrastructure used
AASHTO LTS-5-2009 standard whether at the level (0) to conduct testing. Special thanks to FRE-Composites Inc,
(ground level) or at the service opening level of the QC, Canada, for providing the FRP Poles.
prototypes 33-B-5-C-(1), 35-B-5-C-(1), and 40-B-5-C-
(1) showed that:
Abbreviations
. Local buckling caused by the applied compressive A Pole’s cross-sectional area (m2)
stress did not govern due to the very low axial loads. Apr Projected area (m2)
. The design was governed by the allowable stress bL Width (or diameter) of light fixture (m)
value. However, the verification of the compressive Ce, Cg, Ch Exposure factor, gust factor and drag
stress – local buckling as well as the bending stress factor, respectively
performed at the service opening level showed D Outer diameter (mm)
that the design was governed by the ultimate stress DL Dead load (N)
value. E Young’s modulus in the longitudinal
. The verification of the combined bending, compres- direction of the lamina constituting
sion, and shear stresses was the worst combination the pole’s base zone (N/m2)
case. This verification was performed using the El Modulus of elasticity in the fiber direc-
design equation given by AASHTO LTS-5-2009 tion (unidirectional layer) (N/m2)
standard for aluminum poles. The authors suggest Et Modulus of elasticity in the transverse
that this equation be considered in the design of direction (unidirectional layer) (N/m2)
228 Journal of Composite Materials 47(2)

Exi Young’s modulus in the longitudinal Mz Torsion moment (Nm)


direction of the ith layer (N/m2) N Total number of layers in the pole’s
Ec Modulus of elasticity in compression base zone
(MPa) N Safety factor
E1 Modulus of elasticity in bending in the Pi The rate representing the ith layer of
longitudinal direction of the member the laminate constituting the base
(MPa) zone of the pole
E2 Modulus of elasticity in bending in the Q Wind pressure of reference (Pa)
transverse direction of the member R1 Radius from center to external face of
(MPa) the pole (m)
Fa, Fau Allowable compression stress (MPa) R Radius measured to the mid-thickness
and Ultimate compression stress of the wall (m)
(MPa), respectively R Radius of gyration (mm)
Fb, Fbu Allowable bending stress (MPa) and rb Radius of gyration in the plane of bend-
Ultimate bending stress (MPa) ing (m)
respectively T Wall thickness (mm)
F’e Euler stress divided by the safety factor Vf Fiber volume content (%)
in compression Vs Transverse shear load (N)
Ftu Ultimate tensile stress (MPa) W Wind load (N)
Fu Ultimate applied load (N) Y Distance from center to external face of
Fv, Fvu Allowable shear stress (MPa) and ulti- the pole (m)
mate shear stress (MPa), respectively Q Linear mass of the fibers (g/km). See
Fw Maximum wind load (N) Table 1
Fw1, Fw2 Wind load applied on the pole (tube) "Cx,b Ultimate longitudinal compression
(N) and Wind load applied on the strain at the base of an FRP pole ("/
light fixture (N), respectively kNm)
fa Applied compression stress (MPa) "Tx,b Ultimate longitudinal tension strain at
fb Applied bending stress (MPa) the base of an FRP pole ("/kNm)
fv Applied shear stress (MPa) max Maximum deflection at the loading
fvb Applied shear stress due to transverse position of an FRE pole (mm)
loads (MPa) w Maximum pole top deflection due to
fvt Applied shear stress due to torsion maximum wind loads (mm)
loads (MPa) a Allowable pole top deflection (mm)
G In-plane shear modulus (MPa) yi Fiber angle of the ith layer
Glt Shear modulus (unidirectional layer), lt; tl Poisson’s ratios (unidirectional layer)
(N/m2) 12 Poisson’s ratio in the longitudinal direc-
h1 Centroid height of the pole (tube) (m) tion of the member
h2 Centroid height of light fixture (m) m Constant relating Poisson’s ratio of an
Ice Ice load (N) orthotropic material
I Moment of inertia (m4)
K1 Orthotropy factor
K Effective buckling length factor, k ¼ 2.0
for cantilever beams References
kb Effective buckling length factor in the 1. Labossière P and Neale KW. Macroscopic failure criteria
plane of bending for fibre-reinforced composite materials. Solid Mech Arch
L Cantilever height of the pole (m) 1987; 12: 65–95.
2. Barbero EJ and Raftoyiannis IG. Euler buckling of pul-
Ma Allowable bending moment (Nm)
truded composite columns. Compos Struct 1993; 24:
Mu Ultimate bending moment (Nm)
139–147.
Mu,b Ultimate bending moment at the base 3. Polyzois D, Ibrahim S and Raftoyiannis IG. Performance
of an FRP pole (kNm) of fiber-reinforced plastic tapered poles under lateral load-
Mu,o Ultimate bending moment at the prin- ing. J Compos Mater 1999; 33: 941–960.
cipal opening of an FRP pole (kN.m) 4. Lin ZM. Analysis of pole-type structure of fiber-reinforced
Mw Maximum bending moment due to plastics by finite element method. PhD Thesis. University of
maximum wind loads (Nm) Manitoba, Manitoba, Canada, 1995.
Metiche and Masmoudi 229

5. Gart M and Krambule G. Backyard pole replacement 13. California Department of Transportation. Proposed
using fiberglass poles. Transm Distrib World 1983; 11: California Test 683, method for testing deflection and
57–60. bending strength of fiber-reinforced plastic poles. New
6. Ibrahim S and Polyzois D. Ovalisation analysis of fiber- York, CA.
reinforced plastic poles. Compos Struct 1999; 45: 7–12. 14. Gay D. Mate´riaux composites. Paris, France: Hermès,
7. Metiche S and Masmoudi R. Full-scale flexural testing on 1997, p.672.
fiber-reinforced polymer (FRP) poles. Open Civil Eng J 15. American Association of State Highway and
2007; 1: 37–49. Transportation Officials. Standard specifications for
8. Metiche S, Nacer R and Masmoudi R. Mechanical prop- structural supports for highway signs, luminaires and traf-
erties of FRP laminates produced by the filament winding fic signals, AASHTO LTS-5, 5th edn. Washington, DC:
process. In: 2nd International Structural Specialty AASHTO, 2009.
Conference, Canadian Society for Civil Engineering, 16. Canadian Standard Association. Canadian highway
Québec, Canada, June 10–13, 2008. bridge design code. CAN/CSA-S6-06. Toronto, Ontario,
9. Masmoudi R, Mohamed H and Metiche S. Finite element Canada: CSA International, 2006, p.800.
modeling for deflection and bending responses of GFRP 17. Nacer R. Évaluation des proprie´te´s physiques et me´cani-
poles. J Reinf Plast Compos 2008; 6: 639–658. ques de lamine´s en mate´riaux composites fabrique´s par
10. Metiche S. Évaluation expe´rimentale et the´orique du com- enroulement filamentaires. Master thesis, Université de
portement à la flexion de nouveaux poteaux en mate´riaux Sherbrooke, Sherbrooke, Québec, Canada, 2006.
composites. PhD Thesis. University of Sherbrooke, 18. Canadian Standard Association. Antennas, towers, and
Québec, Canada, 2008. antenna-supporting structures. CAN/CSA-S37-01.
11. American Society for Testing and Materials ASTM D Toronto, Ontario: CSA International, 2001, p.118.
4923 – 01. Standard specification for reinforced thermo- 19. American Society of Civil Engineers. Recommended
setting plastic poles. In: Annual book of ASTM standards. practice for fiber-reinforced polymer products for over-
Vol. 08.02, West Conshohocken, PA: ASTM head utility line structures. ASCE Manuals Reports Eng
International, 2005, pp.726–735. Pract 2003; 104: 76.
12. American National Standard Institute. Fiber-reinforced
plastic (FRP) lighting poles. USA: American National
Standard for Roadway Lighting Equipment, C 136.20.

You might also like