SECTION 17 - EXTREN DWB® DESIGN GUIDE

Table of Contents

Introduction .............................................................17-3

Independent Testing Certification............................17-4

Physical and Section Properties..............................17-5

Materials..................................................................17-7

Statement of Approach............................................17-8

Beam Load Tables................................................17-10

Web Buckling.........................................................17-30

Connections...........................................................17-31

Appendix - Existing Bridge Projects......................17-40

Look for this blue line in the left margin of the

Design Manual documents. This line shows
you where the latest update has been made.
 Section 17 Metric
EXTREN DWB® Design Guide

SECTION 17

EXTREN DWB® DESIGN GUIDE

8” x 6” (203.2 mm x 152.4 mm)

EXTREN DWB®
Hybrid and All-Glass Material Configurations

36” x 18” (914.4 mm x 457.2 mm)

EXTREN DWB®
Hybrid Material Configuration

Copyright © 2013 Strongwell Corporation

All Rights Reserved 17-1 Rev. 0910
 Section 17 Metric
EXTREN DWB® Design Guide

EXTREN DWB® DESIGN GUIDE

203.2 mm x 152.4 mm EXTREN DWB® Hybrid and All-Glass Material
Configurations

Based on Independent Research, Testing and Analysis Under the Direction of:

Professor John J. Lesko, Ph.D.

Department of Engineering Science & Mechanics
Virginia Polytechnic Institute and State University
&
Professor Thomas E. Cousins, Ph.D.
Via Department of Civil and Environmental Engineering
Virginia Polytechnic Institute and State University

©Copyright 2008 by Strongwell Corporation

Bristol, VA, U.S.A.
All Rights Reserved

Copyright © 2013 Strongwell Corporation

Rev.0910
17-2 All Rights Reserved
 Section 17 Metric
EXTREN DWB® Design Guide

INTRODUCTION FROM STRONGWELL

Strongwell’s EXTREN DWB® (Double Web Beam) was developed with the assistance of
the U.S. Department of Commerce’s Advanced Technology Program (ATP). This involved
a three year cooperative research and development program between Strongwell and
the Advanced Technology Program.
The goal of Strongwell’s ATP project was to design, develop and produce an optimized
fiber-reinforced polymer (FRP) structural shape for use in heavy structures such as
vehicular bridges and offshore drilling platforms. The program included the development
of manufacturing processes and equipment to produce such a product. The result of
Strongwell’s efforts is a double web beam with carbon fibers in the top and bottom flanges
for increased stiffness.
The carbon/glass “hybrid” (hybrid refers to the combination of the dual carbon and glass
reinforcements) beam has a flexural modulus of elasticity greater than 41,369 N/mm2.
The flexural modulus of elasticity of a standard EXTREN® WF beam is 17,926 to 19,305
N/mm2. Additionally, the double web shape has significantly improved the lateral torsional
stability of the beam under load. This increased stability is very significant and reduces
the beam’s need for lateral bracing.
Strongwell presently produces an 203.2 mm x 152.4 mm and a 914.4 mm x 457.2 mm
DWB. Both of these sizes have undergone extensive laboratory testing and the 203.2
mm x 152.4 mm was installed on a short span bridge in Blacksburg, Virginia, in June
19971. The 914.4 mm x 457.2 mm DWB was installed in an AASHTO HS-20 vehicular
bridge in Sugar Grove, Virginia in September 20012.
The EXTREN DWB® design data presented herein is the result of extensive testing
and evaluation work by two engineering departments of Virginia Tech - Via Department
of Civil & Environmental Engineering and the Department of Engineering Sciences
and Mechanics. The availability of Virginia Tech’s heavy structures laboratory and the
recognized expertise of its engineering professors provided Strongwell with an excellent
source for independent testing of the EXTREN DWB®.
Strongwell’s mechanical testing laboratory has created a data bank of coupon test
properties for the EXTREN DWB® structural shape. This data bank enables Strongwell to
compare the coupon properties of each manufactured lot of EXTREN DWB® shapes and
to certify that each lot meets the performance characteristics and criteria identified in this
design guide. Strongwell’s certification of these properties provides structural engineers
with the confidence that EXTREN DWB® structural shapes meet the performance
requirements listed herein.
This guide includes design information for the 203.2 mm x 152.4 mm and 914.4 mm x
457.2 mm shapes. The enclosed information allows structural engineers to design their
projects with EXTREN DWB® beams with confidence.
Strongwell received raw material support in manufacturing the 914.4 mm DWB from
Dow Chemical, Owens Corning and Fortifil Fibers. Their assistance in manufacturing
the beams used for performing the various destructive testing is very much appreciated.
1
Hayes, M.D., Lesko, J.J., Haramis, J., Cousins, T.E., Gomez, J., Massarelli, P., “Laboratory and Field Testing of Com-
posite Bridge Superstructure,” ASCE, Journal of Composites for Construction, Vol. 4, No. 3, 2000, pp. 120-128.
2
Hayes, M.D., Lesko, J.J., Cousins T., Waldron C., Witcher D., Barefoot G., & Gomez J., “Design of a Short Span Bridge
Using FRP Girders,” Composites in Construction International Conference, October 10-12, 2001, Porto, Portugal.
Copyright © 2013 Strongwell Corporation
All Rights Reserved 17-3 Rev. 0910
 Section 17 Metric
EXTREN DWB® Design Guide

INDEPENDENT TESTING CERTIFICATION

MEMORANDUM

To: Users of the EXTREN DWB® Design Guide

FROM: JOHN J. LESKO, PHD.

DEPARTMENT OF ENGINEERING SCIENCE & MECHANICS

THOMAS E. COUSINS, PHD.

VIA DEPARTMENT OF CIVIL AND ENVIRONMENTAL ENGINEERING

Subj: The development of the EXTREN DWB® Design Guide

Date: 10 February 2003

The basis for this design manual comes from independent stiffness and strength
characterizations carried out in our laboratories. These tests were completed on 8"
x 6" all-glass and hybrid DWB and the 36" x 18" hybrid DWB manufactured under
production conditions and supplied to us by Strongwell Corp. The final design values
presented in the EXTREN DWB® Design Guide are the direct result of these tests
and interpretations described within the guidelines.

A LAND-GRANT UNIVERSITY THE COMMONWEALTH IS OUR CAMPUS

AN EQUAL OPPORTUNITY / AFFIRMATIVE ACTION INSTITUTION

Copyright © 2013 Strongwell Corporation

Rev.0910
17-4 All Rights Reserved
 Section 17 Metric
EXTREN DWB® Design Guide

PHYSICAL AND SECTION PROPERTIES

Dimensions specified are nominal and apply for both the all-glass and hybrid forms of this
beam (shown in Figure 2 and Figure 3). Standard tolerances for the as-pultruded shape
(section dimensions and straightness) are also listed in Table 1:

NOMINAL SECTION PROPERTIES

Ixx = 53,693,854 mm4

Sxx = 527,663 mm3

rxx = 77.98 mm

A = 8,839 mm2

A2 webs = 3,458 mm2

A2 flanges = 4800 mm2

Iyy = 13,236,159 mm4

Syy = 173,703 mm3
ryy = 38.1 mm
Weight = 16.7 kg/m

Figure 1. Nominal Section Properties and Dimensions (in mm) for the 203.2 mm DWB

Copyright © 2013 Strongwell Corporation

All Rights Reserved 17-5 Rev. 0910
 Section 17 Metric
EXTREN DWB® Design Guide

PHYSICAL AND SECTION PROPERTIES

NOMINAL SECTION PROPERTIES

Ixx = 6.36 x 109 mm4

Sxx = 13,912,617 mm3

rxx = 327.6 mm

A = 58,839 mm2

A2 webs = 32,323 mm2

A2 flanges = 21,935 mm2

Iyy = 1.09 x 109 mm4

Syy = 4,785,023 mm3
ryy = 136.4 mm
Weight = 104.2 kg/m

Figure 2. Nominal Section Properties and Dimensions (in mm) for the 914.4 mm DWB

TABLE 1
STANDARD TOLERANCES

Condition Tolerance
Wall Thickness  15%
Outside Dimension  1.5%
Straightness .060" x length in ft.
Flatness .040" per inch of outside dimension
Twist 1/2o x length in ft., 5o maximum
Cut Lengths -0”, +3.00"
Squareness of end cut  1o

Copyright © 2013 Strongwell Corporation

Rev.0910
17-6 All Rights Reserved
 Section 17 Metric
EXTREN DWB® Design Guide

MATERIALS

203.2 MM X 152.4 MM EXTREN DWB® — ALL-GLASS

The 203.2 mm x 152.4 mm EXTREN DWB® - G (203.2 mm DWB-G), all-glass beam, is a
pultruded structural shape composed of four different types of E-glass reinforcements in a
vinyl ester resin matrix. The all-glass laminate includes 0° longitudinal rovings, continuous
strand mat, 0°/90° stitched fabric, and 45° stitched fabric. The approximate fiber volume
fraction is 55%. The DWB shape improves the apparent (or effective) modulus of elasticity
and the stability of the structure under load versus traditional FRP WF or I shapes. The
shape weighs 11.2 pounds per linear foot (11.2/lf).

203.2 MM X 152.4 MM EXTREN DWB® — HYBRID BEAM

The 203.2 mm x 152.4 mm EXTREN DWB® - H (203.2 mm DWB-H), hybrid beam, is a
pultruded structural shape comprised of carbon fiber tows and four different types of glass
reinforcements in a vinyl ester resin matrix. The 0° carbon tows replace some of the 0°glass
rovings in the top and bottom flanges of the shape. The remainder of the laminate is identical
to the all-glass beam. The carbon tows improve the apparent (or effective) modulus of
elasticity at least 30% versus the all-glass beam. The approximate fiber volume is 55%
(including glass and carbon). The shape weighs 11.2 pounds per linear foot (11.2/lf).

914.4 MM X 457.2 MM EXTREN DWB® BEAM — HYBRID BEAM

The 914.4 mm x 457.2 mm EXTREN DWB® (914.4 mm x 457.2 mm DWB-H) is only
produced as a hybrid beam. It is a pultruded structural shape composed of carbon fiber
tows in the top and bottom flanges and the same four types of E-glass reinforcements
as the 203.2 mm DWB-G and 203.2 mm DWB-H in a vinyl ester resin matrix throughout
the entire structural shape. The carbon tows improve the apparent (effective) modulus
of elasticity. The approximate fiber volume is 55% (including glass and carbon) and the
shape weighs 70 pounds per linear foot (70 lbs/lf). The 914.4 mm DWB-H was designed
specifically for use in vehicular bridges.

ANTICIPATED APPLICATIONS FOR EXTREN DWB®

This guide is intended for assistance in the design of structures such as bridges, buildings,
offshore structures, and miscellaneous heavy structural fabrications.
• Bridges — Primary and secondary stringers and floor beams
• Buildings — Primary and secondary structural members for building components
including floor beams, roof beams, purlins, etc.
• Offshore Structures — Floor beams, deck beams, and primary decking structure
• Miscellaneous Structures — Towers, heavy industrial platform and floor beams, pipe
racks, etc.

Copyright © 2013 Strongwell Corporation

All Rights Reserved 17-7 Rev. 0910
 Section 17 Metric
EXTREN DWB® Design Guide

STATEMENT OF APPROACH

The design guide for the Strongwell DWB is presented as a material specification where the
material system and its manufacturing process are well defined and controlled. Given these
tolerances on the FRP product, guidelines for its use in a structure are defined.
As a guide, the Load Resistance Factor Design (LRFD) approach is used to define these
operating limits.3 In this approach, the probability distribution of load/stress (Loads) is
compared to the probability of failure strength of the material (Resistance), as illustrated in
Figure 4. Selecting the form and size of the structure determines the desired overlap of the
two distributions, thus defining the stated allowable risk.

Figure 3. LRFD Conceptual Representation for Design

For the purposes of this design guide, we, however, only define for the engineer the Resistance
side of the problem. Therefore, the engineer of record is required to define the Loads side of
the particular design application based on the variability of loads and operating environment.
These details will define the level of reliability required for the application.
In determining the Resistance element of the design problem, based on this material
specification, Weibull statistics are employed to describe the variability of the material. The
Weibull statistical distribution is widely accepted in the composites community for describing
the variability of failure for these material systems.4,5

3
AASHTO, “LRFD Bridge Design Specification,” 2nd Edition, American Association of State Highway and Transportation
Officials, Washington, D.C., 1998.
4
Weibull, Waloddi, “A Statistical Distribution Function of Wide Applicability,” J. of Applied Mechanics, 1951, pp. 293-297.
5
Weibull, Waloddi, “A Statistical Representation of Fatigue Failures in Solids, Transactions of the Royal Institute of
Technology,” No. 27, Stockholm, 1949.

Copyright © 2013 Strongwell Corporation

Rev.0910
17-8 All Rights Reserved
 Section 17 Metric
EXTREN DWB® Design Guide

STATEMENT OF APPROACH

A reliability based approach is used to define A- and B-basis allowable levels of resistance
(described further in the Commentary, page 21). These values define for the engineer
the level of risk allowed in operating the structure based on a determined design load.
Figure 5 illustrates the margin between a design load (supplied by the engineer) and
the A-basis or the B-basis resistance listed in this guide. This margin is identified as the
level of risk, or inversely, the margin of safety for the design.

Figure 4. Margin of Safety (based on the selection of working

loads/stress relative to the A- and B-basis allowables.

This margin or factor of safety should take into account the variability in loads as defined
by the engineer for the particular structure. An extensive presentation of load factors
is available from the American Society of Civil Engineers.6 In addition, as the A- and B-
basis allowables (resistance) can change over time due to environmental exposure and
fatigue. The selected margin of safety must also consider the potential effects of the service
environment on the performance of the structure. As FRP structural shapes are new to
the industry, definitive criteria for reasonable factors of safety based on durability are not
presently available. However, the engineer is referred to several sources for guidance in
selecting reduction factors for the stated A- and B-basis allowables.7 The engineer is also
referred to several sources of ongoing research on the durability of the DWB in service and
laboratory testing which are not as yet in the form of criteria for the selection of reduction
factors. 8, 9 While this guide does not provide load reduction factors, the referenced
documents and codes do refer the engineer to appropriate load factors.

6
Minimum Design Loads for Buildings and Other Structures, ASCE 7-02, American Society of Civil Engineers,
Reston, VA, 2002.
7
EUROCOMP Design Code and Handbook, “Structural Design of Polymer Composites,” Ed Clarke, J.L., E&F
Spon, London, UK, 1997, pp. 37-41.
8
Neely, W. D., “Evaluation of the In-Service Performance of the Tom’s Creek Bridge,” M.S. Thesis Via Department
of Civil & Environmental Engineering,Virginia Tech, May 2000, electronic thesis available on-line at: [Link]
[Link]/theses/[Link].
9
Senne, J.L., “Fatigue Life of Hybrid FRP Composite Beams,” M.S. Thesis, Department of Engineering Science &
Mechanics, Virginia Tech, July 2000, electronic thesis available at [Link]

Copyright © 2013 Strongwell Corporation

All Rights Reserved 17-9 Rev. 0910
 Section 17 Metric
EXTREN DWB® Design Guide

BEAM LOAD TABLES

The following load tables have been developed using the Load Resistance Factor Design
(LRFD) approach and are further defined later in this section.
TABLE 2
8" x 6" EXTREN DWB® - G (All-Glass)
Major Axis A-Basis Properties
Ezz =4.01 x 10 psi kGzyAv =1.0 x 106 psi-in2 Mmax= 96.1 kip ft.
6

A-Basis Allowable Distributed Loads in Pounds Per Lineal Foot

Span Capacity Deflection

in Ft. Moment L/180 L/240 L/300 L/360 L/420 L/500 L/600 L/800
8 12013 1945 1459 1167 973 834 700 584 438
10 7688 1140 855 684 570 488 410 342 256
12 5339 716 537 429 358 307 258 215 161
14 3922 475 356 285 237 204 171 142 107
16 3003 330 247 198 165 141 119 99 74
18 2373 238 178 143 119 102 86 71 53
20 1922 176 132 106 88 76 63 53 40

TABLE 3
8" x 6" EXTREN DWB® - G (All-Glass)
Major Axis B-Basis Properties
Ezz =4.25 x 10 psi kGzyAv =1.6 x 106 psi-in2 Mmax= 108 kip ft.
6

B-Basis Allowable Distributed Loads in Pounds Per Lineal Foot

Span Capacity Deflection

in Ft. Moment L/180 L/240 L/300 L/360 L/420 L/500 L/600 L/800
8 13463 2338 1754 1403 1169 1002 842 701 526
10 8616 1322 992 793 661 567 476 397 298
12 5983 811 609 487 406 348 292 243 183
14 4396 530 398 318 265 227 191 159 119
16 3366 364 273 218 182 156 131 109 82
18 2659 260 195 156 130 112 94 78 59
20 2154 192 144 115 96 82 69 58 43

Failure Mode: The controlling failure mode observed for the all-glass beams was
delamination within the compression flange. The tension flange typically was left intact
and able to support load. In some cases, interply damage was observed in the tension
flange and less able to carry bending loads.
Lateral Torsional Stability: Flexural stiffness and strength characterizations were
carried out with no lateral supports for spans to 20'. Thus, for spans to 20', flexural
strength is the controlling limit state for both all-glass and hybrid beams. In subsequent
flexural tests on laterally unsupported all-glass spans of 20' to 40', lateral-torsional
buckling was not observed at deflections to L/90.

Copyright © 2013 Strongwell Corporation

Rev.0910
17-10 All Rights Reserved
 Section 17 Metric
EXTREN DWB® Design Guide

BEAM LOAD TABLES

Long-term Performance, Fatigue & Durability: Presently there is limited information

regarding the long-term performance of the beam in combined hygro-thermal mechanical
service environments. Fatigue loading of the hybrid beam has revealed no loss in stiffness
and no failure after 10 million cycles at an applied moment of 37.3 kip ft.,9 slightly above
the A-basis allowable single cycle moment capacity of 36.1 kip ft. Further field work with
the beams in the Tom’s Creek Bridge, Blacksburg, VA, has demonstrated that the beam
can withstand 15 months in service with no loss in stiffness and strength under a modest
service environment.8 Moreover, no residual creep deflection was observed following the
15 months in service.8 TABLE 4
8" x 6" EXTREN DWB® - H (Hybrid)
Major Axis A-Basis Properties
Ezz= 5.66 x 106 psi kGzyAv= 1.8 x 106 psi-in2 Mmax= 36.1 kip ft.
A-Basis Allowable Distributed Loads in Pounds Per Lineal Foot
Span Capacity Deflection
in Ft. Moment L/180 L/240 L/300 L/360 L/420 L/500 L/600 L/800
8 4513 2970 2228 1782 1485 1273 1069 891 668
10 2888 1703 1277 1022 851 730 613 511 383
12 2006 1054 791 632 527 454 379 316 237
14 1473 693 520 416 346 297 249 208 156
16 1128 478 358 287 239 205 172 143 107
18 891 342 257 205 171 147 123 103 77
20 722 253 190 152 127 109 91 76 57
TABLE 5
8" x 6" EXTREN DWB® - H (Hybrid)
Major Axis B-Basis Properties
Ezz= 5.97 x 10 psi kGzyAv= 2.2 x 106 psi-in2 Mmax= 51.6 kip ft.
6

B-Basis Allowable Distributed Loads in Pounds Per Lineal Foot

Span Capacity Deflection
in Ft. Moment L/180 L/240 L/300 L/360 L/420 L/500 L/600 L/800
8 6450 3266 2449 1960 1633 1400 1176 980 735
10 4128 1850 1388 1110 925 793 666 555 416
12 2867 1136 852 682 568 487 409 341 256
14 2106 743 557 446 372 318 268 223 167
16 1613 511 383 306 255 219 184 153 115
18 1274 365 274 219 182 156 131 109 82
20 1032 270 202 162 135 116 97 81 61

Failure Mode: The controlling failure mode observed for the hybrid beam was
delamination within the compression flange, leaving the tensile flange essentially
undamaged.
Lateral Torsional Stability: Flexural stiffness and strength characterizations were carried
out with no lateral supports for spans to 20'. Thus, for spans to 20', flexural strength is the
controlling limit state for both all-glass and hybrid beams. In subsequent flexural tests on
laterally unsupported hybrid spans of 20' to 40', lateral-torsional buckling was not observed
at deflections to L/90.

Copyright © 2013 Strongwell Corporation

All Rights Reserved 17-11 Rev. 0910
 Section 17 Metric
EXTREN DWB® Design Guide

BEAM LOAD TABLES

TABLE 6
36" x 18" EXTREN DWB® - H (Hybrid)
Major Axis A-Basis Properties
Ezz= 5.76 x 106 psi kGzyAv= 44.5 x 106 psi-in2 I= 15291 in4
Mmax= 964.0 kip-ft. @ 30’ Span & 635.6 kip-ft. 40-60’ Span
A-Basis Allowable Distributed Loads in Pounds Per Lineal Foot

Span Capacity Deflection

in Ft. Moment L/180 L/240 L/300 L/360 L/420 L/500 L/600 L/800
30 8569 1960 1764 1411 1176 1008 882 784 705
35 5223 1465 1319 1055 879 754 659 586 527
40 3178 1117 1006 804 670 575 503 447 402
45 2511 867 780 624 520 446 390 347 312
50 2034 683 615 492 410 351 307 273 246
55 1681 546 491 393 328 281 246 218 197
60 1412 442 398 318 265 227 199 177 159

TABLE 7
36" x 18" EXTREN DWB® - H (Hybrid)
Major Axis B-Basis Properties
Exx =6.10 x 106 psi kGxyA =46.2 x 106 psi-in2 I=15291 in4
Mmax= 1139 kip ft. @ 30’ Span & 916.7 kip-ft. 40-60’ Span
B-Basis Allowable Distributed Loads in Pounds Per Lineal Foot

Span Capacity Deflection

in Ft. Moment L/180 L/240 L/300 L/360 L/420 L/500 L/600 L/800
30 10124 2051 1846 1477 1231 1055 923 820 738
35 6712 1536 1382 1106 921 790 691 614 553
40 4584 1173 1055 844 704 603 528 469 422
45 3622 911 820 656 547 468 410 364 328
50 2933 719 647 517 431 370 323 287 259
55 2424 575 517 414 345 296 259 230 207
60 2037 466 419 335 279 239 210 186 168

Failure Mode: The controlling failure mode for all beams was delamination within the
compression flange, leaving the tensile flange undamaged.

Bearing Conditions: The values noted are valid for full width elastomeric bearing.

Copyright © 2013 Strongwell Corporation

Rev.0910
17-12 All Rights Reserved
 Section 17 Metric
EXTREN DWB® Design Guide

BEAM LOAD TABLES

ANALYTICAL METHODOLOGY

The preceding charts are allowable load tables for the hybrid and all-glass EXTREN DWB®
when used as flexural members (beams).
These allowable loads are based upon:
1. Flexural testing conducted at ambient conditions under four-point loading
2. Laterally unsupported beams
3. Single span with simply supported ends
4. Allowable distributed loads in the plane of the web, based on strength (ultimate
flexural moment capacity) and deflection determined from A- and B-basis statistics
for a shear deformable beam where,
5. Because k and Av are difficult to quantify in some cases, the full value of kGzyAv is
experimentally determined for the purposes of this design guide. See the details
presented in the Commentary (pages 17 -18) regarding the determination of kGzyAv
and it’s A- and B-basis values.

ω = distributed load in pounds per

foot of beam length
L = length of the simple span
5ωL4 + ω L2
Ezz = flexural modulus of the beam
δMAX = 384EZZIXX 8kGzyAv section about its major axis
k = shear non-dimensional correction
factor for the cross section
Gzy = shear modulus of the beam
Av = shear area of the beam
Determination of Allowables
Five or more replicates at each of the spans were used to compute the “allowable” design
values for stiffness and strength. The allowable values are a measure of the confidence in
the data and the reliability at which one desires to operate a structural system*. Allowable
values are prescribed here as opposed to arbitrary factors of safety because they represent
a level of confidence in the data and a desired level of reliability prescribed for the structure.
For instance, the A-basis allowable is based on a level of confidence of 95% in the data
(that is, 95% of the data falls above a prescribed value) and ensures 99% reliability (only
1% of the derived values will fall below this value) in the value chosen as the design value.
Likewise, the B-basis allowable prescribes a level of confidence of 95% (that is, 95% of the
data falls above a prescribed value) and ensures 90% reliability (10% of the derived values
will fall below this value). Both bases are presented in this design guide.

Copyright © 2013 Strongwell Corporation

All Rights Reserved 17-13 Rev. 0910
 Section 17 Metric
EXTREN DWB® Design Guide

BEAM LOAD TABLES

Weibull Statistics
The basis for these calculations lies in Weibull statistics, where the cumulative probability
distribution function describing the distribution of measured values is derived from,
α
-(x/β)
F(x) = 1- e

where α and β are the two parameters used to fit the data. The value of α (the shape
parameter) determines the breadth of the distribution while β (the location parameter) defines
the value most closely representing the center of the distribution. Based on the concept of
reliability, R(x), the probability of failure, F(x), is related to the reliability by,

R(x) = 1 - F(x)

This relationship can be rearranged to form an expression for the A- and B-basis values
from the following expressions,

1/
α 1/
α
A allowable = βlower [ ( 1
Ln 0.99 )] B allowable = βlower [Ln( 1
0.90 )]

* MIL Standard: MIL-HDBK-17 (2001). Composite Materials Handbook. Available at [Link]

Copyright © 2013 Strongwell Corporation

Rev.0910
17-14 All Rights Reserved
 Section 17 Metric
EXTREN DWB® Design Guide

BEAM LOAD TABLES

Lateral Torsional Stability: Flexural stiffness and strength characterizations were carried
out with no lateral supports for spans to 60'. Thus, flexural strength is the controlling limit
state in these conditions. Subsequent flexural tests on laterally unsupported spans @ 60'
demonstrated that lateral-torsional buckling does not occur at deflections of L/180. It is,
however, recommended that the beam only be loaded to L/360, allowing for a factor of safety
of 2.
Long-term Performance, Fatigue & Durability: Fatigue testing of the girder is presently
underway to assess the flexural durability of the section. Failure mode and number of cycles to
failure under design loads will be determined for limited conditions. The girder has also been
installed (September 2001) in the Dickey Creek bridge of Route 601 in Sugar Grove, VA.2
Monitoring and field work are underway to examine the performance of the bridge and the
girders under service conditions.

OTHER SECTION PROPERTIES:

Shear Deformable Beams
The elastic shear properties of the section are represented by the value kGxyAv, where k is the
shear correction factor (which accounts for the non-uniform shear stress distribution through
the depth of the beam), Gxy is the shear modulus and Av is the shear area. Because k and Av
are difficult to quantify in some cases, the full value of kGxyAv is experimentally determined for
the purposes of this design manual. The A & B basis values for kGxyAv are presented in the
tables preceeding this section.
The average10 shear values (kGxyAv) have been determined as:

TABLE 8

Beam Type (kGxyAv)

(Msi-in4)
8” DWB Hybrid 2.8
8” DWB All-Glass 3.1
36” DWB Hybrid 46.5

Copyright © 2013 Strongwell Corporation

All Rights Reserved 17-15 Rev. 0910
 Section 17 Metric
EXTREN DWB® Design Guide

BEAM LOAD TABLES

Torsional
Using the relationship for torsion,
GJeff=TL / φ

the torsional section stiffness, GJeff of the hybrid beam are reported as averages:10

TABLE 9

Beam Type GJeff

(Msi-in4)
8” DWB Hybrid 3.1
8” DWB All-Glass 3.4
36” DWB Hybrid 3170

where T is the applied torque, L is the span and φ is the angle of rotation in radians.

Minor Axis Bending Modulus of Section

Minor axis flexural moduli were computed via laminated beam theory11, 12 for bending about the
yy axis. Validation of these computed values was undertaken for the 8” DWB for bending about
the major axis and found to be in good agreement with the experimentally determined values
discussed above. Confirmation of the 36” DWB prediction has not been completed.10

TABLE 10

Beam Type Eyy

(Msi)
8” DWB Hybrid (Iyy=31.8 in4) 5.96
8” DWB All-Glass (Iyy=31.8 in4) 3.58
36” DWB Hybrid (Iyy=2626 in4) 4.35

11
E. J. Barbero, R. Lopez-Anido, and J.F. Davalos, “On the Mechanics of Thin-Wall Laminated Composite Beams,”
Journal of Composite Materials, v27 n8 (1993), pp. 806-829. AND,
12
J. F. Devalos, H.A. Salim, P. Qiao, R. Lopez-Anido, and E.J. Barbero, “Analysis and Design of Pultruded FRP
Shapes Under Bending,” Composites Part B: Engineering, v27 n3-4 (1996) pp. 295-305.

Copyright © 2013 Strongwell Corporation

Rev.0910
17-16 All Rights Reserved
 Section 17 Metric
EXTREN DWB® Design Guide

BEAM LOAD TABLES

COMMENTARY:

Flexural Stiffness Properties and Moment Capacity Determination

Major axis bending stiffness and moment capacity was assessed at spans of 8’, 14’
and 20’ for the 8” DWB and nominally 30’, 40’, 60’ for the 36” DWB. A four-point bending
test configuration was used to assess these performance attributes, as shown below.
Measurements of load deflection and strain were taken continuously during the course
of each test. Quantities recorded during the tests and their location are also noted in
Figure 6. Load was applied using open loop servo hydraulic actuators. The duration of
the test (from zero load to the failure load) was less than five minutes to avoid creep
induced damage.
F F = Load
L/3 L/3 L/3
y Di = Displacement, quarter & midspan
S = Shear strain bridge
B
z x S T S
B T = Torsional strain bridge
Dqtr Dmid
B = Bending strain bridge
L/4 L/2

Figure 5. Test Set-up for the Determination of Strength and Elastic Constants

Bending strain gauges positioned within the constant moment section of the beam (see
Figure 5) were used to determine the flexural modulus using the classical relationship,
Mxxc
Ezz =
εzzIxx
where c, is the outer fiber distance from the neutral axis (4" for the 8” DWB and 18” for the
36” DWB). Because there are some differences in the tension and compression strains
due to differences in the material response in these modes (typically only a few percent),
the top and bottom strains were averaged. Again, this value only represents the strains
due to bending and does not include shear effects.
Extraction of the shear contribution to deflection is accomplished by treating the section
shear properties kGzyAv, as a single quantity. Again, k is the non-dimensional shear
correction factor (which accounts for the non-uniform shear stress distribution through
the depth of the beam), Gzy is the shear modulus and Av is the shear area. Because k and
Av are difficult to quantify in some cases, the aggregate value of kGzyAv is experimentally
determined for the purposes of this design guide. This was accomplished by taking the
shear deformable expression for a four-point loading case,
7PL3 PL
δmax = 216EzzIxx + 3kG A
zy v

and solving for kGzyAv. For each span and replicate tested, the value for kGzyAv was
determined at a nominal moment. This was done for each beam tested in conjunction with
its individually determined Ezz.

Copyright © 2013 Strongwell Corporation

All Rights Reserved 17-17 Rev. 0910
 Section 17 Metric
EXTREN DWB® Design Guide

BEAM LOAD TABLES

Note that in our use of the expression, kG zyAv is dependent on Ezz and can not be
determined independently as was accomplished with the bending modulus. For this reason,
the determination of A- and B-basis allowable shear properties could not be determined
independently. The means to determine these allowables are discussed next.
Possessing Ezz, i and the (kGzyAv)i for each beam, the distributed load, ωi , for each simply
supported beam corresponding to a given deflection was determined using,

1
ωi =
κ L{ }
5L 2
1
384Ezz, i I xx + (kGzyAv )i

where κ defines the basis for the deflection criteria noted in the design tables (that is, δ=L/κ).
A- and B-basis allowable distributed loads, ωa, were determined through the Weibull based
approach discussed in this design guide. Using the A- and B-basis allowable major axis
flexural modulus for the beams, Ezz,a the A- and B-basis allowable (kGzyAv)a’s were determined
from,

384ωaκ LE zz,aIxx
kGzyAv =
8(384Ezz,a Ixx - 5 κ ωa L3)

βlower is the value representing the lower bound of the 95% confidence interval above which
95% of the data is predicted to occur in the distribution. This is computed from,

[ ]
1/α
2n
βlower = β χ(2n)20.05

where χ(2n)20.05 is the Chi-Squared of the one-sided confidence interval at 5% for n degrees
of freedom (n = the number of samples or replicates). This value can be obtained from
standard math tables or text on statistics.

Copyright © 2013 Strongwell Corporation

Rev.0910
17-18 All Rights Reserved
 Section 17 Metric
EXTREN DWB® Design Guide

BEAM LOAD TABLES

COMMENTARY:

8" x 6" EXTREN DWB® Lateral Torsional Stability

To investigate the lateral torsional buckling characteristics of the 8" x 6" EXTREN DWB®,
multiple tests at varying spans were performed. These multiple tests were conducted on
an unbraced, simply supported beam subjected to a single vertical load at mid-span and
allowed to freely torque about the long axis of the beam and bend horizontally (see Figure
6). Simply supported ends were augmented with torsional constraints to prevent twisting.13
The loading point allowed for the rotation and translation of the beam (see Figure 7). Four
beams, 42’ long, (two hybrid and two all-glass beams) were loaded to a deflection of L/90 at
spans of 20’ to 40’, in 4’ increments. Instrumentation was applied (including strain gauges
and deflectometers) to measure the tendency of the beam to rotate and/or deflect out-of-
plane as load was applied. In all cases, lateral torsional buckling was not observed at the
L/90 deflection.
4x Lateral Supports
@ 9' off-center w/ 1" clearance

2x Lateral-Torsional restraints to
satisfy Mottram boundry conditions

Figure 6: Plan view of lateral-torsional buckling test configuration.

Figure 7: Loading point configuration to allow rotation and

translation of the midspan.
13
Mottram, J.T., “Lateral-Torsional Buckling of a Pultruded I-beam,” Composites, V.23, No.2, 1992, pp. 81-92.

Copyright © 2013 Strongwell Corporation

All Rights Reserved 17-19 Rev. 0910
 Section 17 Metric
EXTREN DWB® Design Guide

BEAM LOAD TABLES

COMMENTARY:

36" x 18" EXTREN DWB ® Lateral Torsional

A similar arrangement to that of the 8” DWB stability test was conducted for the 36" DWB.
Load was applied at midspan using a conventional hydraulic ram. Load was transferred
to the beam using a system of rollers that released the beam from any torsional or lateral
restraints (see Figure 8). The ends of the beam rested on elastomeric bearing pads. These
tests were deflection controlled, and investigated beam stability up to a deflection of L/180,
or approximately 4" deflection at midspan for a test span of 60’.

Figure 8: Loading point configuration

to allow rotation and translation of the
mid-span for the 36" DWB.
Simply supported ends (on full width bearing pads) were augmented with torsional
constraints to prevent twisting at the ends (preventing movement of the top flange laterally
either direction). Lateral guards were used for the 36" DWB as was done for the 8" DWB
(See Figure 6).
The beam was instrumented to detect lateral, vertical, and torsional movement in the beam
using a system of wire pots and strain gauges. Wire pots were attached to the bottom
flange to measure vertical movement, attached horizontally to the top and bottom flanges
to measure twist and lateral movement, and attached to a bar projecting away from and
clamped to the top flange to detect rotation. Strain gauges were placed on the bottom
flange along the centerline of the beam and on the underside of each flange on each side
of the beam, near the edges of the flange. Any warping of the cross section was expected
to show up as a difference in the strain values.
The beam was cycled to 10 kips three times to ensure that the test set-up was both safe and
working properly. The beam was then cycled to roughly 50 kips (or 4" of deflection) three
[Link] beam did not buckle laterally or torsionally in any of the cycles. The beam did
slightly deflect laterally as it was loaded, but showed no signs of decreased load capacity.
For further details on shear deformable beams, please see: Cowpers 196614, Bank 198715
and Hayes 199816.
14
Cowper, G.R., “The Shear Coefficient in Timoshenko’s Beam Theory,” Journal of Applied Mechanics, June 1966, pp. 335-340.
15
Bank, L.C., “Shear Coefficients for Thin-Walled Composite Beams,” Composite Structures, Vol. 8, 1987, pp. 47-61.
16
Hayes, M.D., “Characterization and Modeling of a FRP Hybrid Structural Beam and Bridge Structure for Use in the
Tom’s Creek Bridge Rehabilitation Project,” M.S. Thesis, Department of Engineering Science & Mechanics, Virginia Tech,
February 1998, electronic thesis available at [Link]
Copyright © 2013 Strongwell Corporation
Rev.0910
17-20 All Rights Reserved
 Section 17 Metric
EXTREN DWB® Design Guide

BEAM LOAD TABLES

ANALYTICAL METHODOLOGY:

Shear Deformable Beam Formulas

Single span beam simply supported under a uniformly distributed load.

L
Figure 9.

δ(z,L)= ω
[
z4 - Lz3 + L3z
12EzzIxx 2 2 ] -2kGω A [z -Lz ]
zy v
2

δ = δ(z = L/2) = 5ωL4 + ωL2

Max
384EzzIxx 8kGzyAv

Single span beam simply supported under a mid-span point load.

L/2 L/2
Figure 10.

Pz(24Ezzlxx + 3kGzyAvL2 - 4kGzyAvz2)

For z < L/2: δ(z,L) =
48EzzIxxkGzyAv

P(z-L)(-6EzzIxx+1/4kGzyAvL2 - 2kGzyAvLx + kGzyAvz2)

For z >
– L/2: δ(z,L) =
12EzzIxxkGzyAv

PL3 PL
δ = δ(z = L/2) = +
48EzzIxx
Max
4kGA

Copyright © 2013 Strongwell Corporation

All Rights Reserved 17-21 Rev. 0910
 Section 17 Metric
EXTREN DWB® Design Guide

BEAM LOAD TABLES

Simply supported beam under four-point loading.

P P
b b

L/2 L/2

Figure 11.

Pz(24EzzIxx+ kGzyAv(-12b2 + 3L2 - 4z2)

For z< L/2 - b : δ(z,L)=
24EzzIxxkGzyAv

P(2b - L)[- 24EzzIxx+ kGzyAv(4b2 - 4bL + L2 -12Lz+12z2)]

For L/2 - b < z< L/2 + b : δ(z,L) =
48EzzIxxkGzyAv

P(L-z)[- 24EzzIxx+ kGzyAv( 12b2 + L2 - 8Lz + 4z2)]

> L/2 + b :
For z – δ(z,L) =
24EzzIxxkGzyAv

P(2b - L)[ -24EzzIxx+ kGzyAv(4b2 - 4bL - 2L2)]

δ = δ(z =L/2) =
48EzzIxxkGzyAv
Max

Cantilevered beam under a uniform load.

Figure 12.

δ(z,L) = 2Eω I [ - +
kGzyAv]
z4 Lz3 L2z2 - ω [ Lz - z2 ]
2
zz xx 12 3 2

δ = δ(z = L) =
ωL4 + ωL2
Max
8EzzIxx 2kGzyAv

Copyright © 2013 Strongwell Corporation

Rev.0910
17-22 All Rights Reserved
 Section 17 Metric
EXTREN DWB® Design Guide

BEAM LOAD TABLES

COMMENTARY:

Designing For Concentrated Loads

As previously noted in the load tables, compression flange failure was the controlling failure
mode for all spans recommended for safe use (i.e. 8’-20’ for the 8" DWB and 30'-60' for
the 36" DWB. Typical failure modes for both the 8" DWB and the 36" DWB are shown in
Figure 13.

Figure 13: Typical compression flange failures for

four point bend testing of the 8" DWB (top photo)
and the 36" DWB (bottom photo).

Copyright © 2013 Strongwell Corporation

All Rights Reserved 17-23 Rev. 0910
 Section 17 Metric
EXTREN DWB® Design Guide

BEAM LOAD TABLES

Testing of the 8" DWB below an L/d (span to depth ratio) of 10 continued to exhibit this
compression flange failure mode. However, at an L/d of 6, the 36" DWB failed at the
supports, as shown in Figure 14. The magnitude of the load at which end support failure
occurred is dependent on the bearing pad geometry.

Figure 14: Bearing failure at supports for 36"

DWB tested at L/d of 6.

These data are summarized in an examination of shear capacity versus the span to depth
ratio, Figure 15. The recommended bearing condition for the 36" DWB is noted as condition
3 of Figure 15 where the pad width extends only to the flange tips.

Figure 15: Support bearing capacity and associated failure mode as a function of the span
to depth ratio for the 8" DWB and the 36" DWB.
Copyright © 2013 Strongwell Corporation
Rev.0910
17-24 All Rights Reserved
 Section 17 Metric
EXTREN DWB® Design Guide

BEAM LOAD TABLES

ANALYTICAL METHODOLOGY:

EXAMPLE 1: Design for Concentrated Load

Given: A simply supported beam of length L=6' is loaded at midspan by a concentrated load.
A-Basis Allowables:
Moment Capacity = 96.1 kip- ft
Shear Capacity = 35.6 kip

Determine: If the all-glass 8" DWB is to operate under A-basis allowables, what is the
maximum allowed concentrated load? What is Pmax if limited to deflection of L/180?
Pmax

3' 3' L/2

V V

Figure 16.

PL M 4 (96.1 kip-ft)(4)...
; Pmax= max =
Pmax based on Mmax: Mmax= .........................................................
4 L (6 ft)
Pmax = 64 kips
Reaction at support is 32 kips = Vmax
Vmax, Vallowed = 32 kips, < 35.6 kips
∴ Pmax = 64 kips = Pallowable

If deflection is a constraint for design and

limited to L/180, Pmax =
(not including shear deformation)
L PL3
180 = 48EI

P = 48EI 2
180 L

(48)(4.01x 106 psi)(129 in4)

P=
180 (6 ft x 12 in )2
ft
∴ Pmax = 26.6 kips = Pallowed
Copyright © 2013 Strongwell Corporation
All Rights Reserved 17-25 Rev. 0910
 Section 17 Metric
EXTREN DWB® Design Guide

BEAM LOAD TABLES

ANALYTICAL METHODOLOGY:

EXAMPLE 2: Design for Concentrated Load

Given: A simply supported 8" DWB of length L=13.5' is loaded off center @ L/5 by a
concentrated load.
A-Basis Allowables:
Moment Capacity = 36.1 kip-ft
Shear Capacity = 13.4 kip

Determine: If the hybrid 8" DWB is to operate under A-basis allowables, what is the
maximum allowed concentrated load if limited to deflection of L/180?
Pmax

a b
2.7' 10.8'

V1 V2
Figure 17.
Pmax from Mmax:

Mmax = ab
L
Pmax = Mmax L = (36.1 kip-ft)(13.5 ft) = 16.7 kips
ab (2.7 ft)( 10.8 ft)

Vmax = V1 = Pb = (16.7 kips)(10.8 ft) = 13.4 kips

L (13.5 ft)
∴ Pmax = Pallowable = 16.7 kips

Note: Vmax = Vallowed

Checking for deflection criteria: (not including shear deformation)

L Pa (L2-a2)3/2
=
180 9 3 EI L

a= L
5
Pmax = Pallowed = 12 kips

Copyright © 2013 Strongwell Corporation

Rev.0910
17-26 All Rights Reserved
 Section 17 Metric
EXTREN DWB® Design Guide

BEAM LOAD TABLES

ANALYTICAL METHODOLOGY:
EXAMPLE 3: Design for Concentrated and Distributed Load
Given: A simply supported 8" DWB x 12' long is loaded with a distributed load of 500 lb/ft
and three concentrated loads of unknown value spaced 3' apart.
B-Basis Allowables:
Moment Capacity = 51.6 kip-ft
Shear Capacity = 19.1 kip

Determine: If the hybrid 8" DWB is to operate under B-basis allowables, what is the
maximum allowed loads P that can be applied?

Figure 18.

Pmax based on moment capacity:

Mmax = 9 kip -ft + 6P

51.6 kip -ft = 9 kip-ft + 6P

P = (51.6 kip -ft - 9 kip-ft) = 7.1 kips = Pmax
6
Pmax based on max shear capacity

The reaction at the supports are given by:

V1 = V2 = 3 kips + 3/2 P
19.1 kip = 3 kips + 3/2 P
Pmax = (19.1 kips - 3 kips) 2/3
Pmax = 10.73 kips
Pmax = P allowed
We must choose Pmax = Pallowed = 7.1 kips due to
limits on moment capacity.

NOTE: Also check deflection

Copyright © 2013 Strongwell Corporation

All Rights Reserved 17-27 Rev. 0910
 Section 17 Metric
EXTREN DWB® Design Guide

BEAM LOAD TABLES

ANALYTICAL METHODOLOGY:

EXAMPLE 4: Design for Concentrated Load

Given: A simply supported beam of length L=30' is loaded at midspan by a concentrated
load.
A-Basis Allowables:
Moment Capacity = 964 kip-ft (See Table 6)
Web Buckling = 45.0 kip (See Table 11)

Determine: If the 36" DWB is to operate under A-basis allowables, what is the
maximum allowed load? What is Pmax if limited to deflection @ L/360?

Figure 19.
Pmax based on Mmax:
Mmax= PL ; Pmax= M 4 = (964 kip-ft)(4)....
max
...........................................................
4 L (30 ft)
Pmax = 129 kips
Reaction at support is 64.5 kips = Vmax
Vmax = 64.5 kips, > 45.0 kips
∴ Pmax = (45.0 kips)(2)= 90.0 kips = Pallowable

If deflection is a constraint for design and limited to L/360, Pmax =

(not including shear deformation)

L PL3
360 = 48EI

P = 48EI 3
360 L
(48)(5.76 x 106 psi)(15291 in4)
P=
360 (30 ft x 12 in )3
ft
∴ Pmax = 252 kips> Pallowed; therefore Pmax = 90.0 kips

Copyright © 2013 Strongwell Corporation

Rev.0910
17-28 All Rights Reserved
 Section 17 Metric
EXTREN DWB® Design Guide

BEAM LOAD TABLES

ANALYTICAL METHODOLOGY:

EXAMPLE 5: Design for Concentrated and Distributed Load

Given: A simply supported beam of length 44' is loaded with a distributed load of 1,000
lb/ft and three concentrated loads of unknown value spaced 3' apart.
B-Basis Allowables:
Moment Capacity = 917 kip-ft (See Table 7)
Web Buckling = 50.1 kip (See Table 11)

Determine: If the hybrid 36" DWB is to operate under B-basis allowables, what is the
maximum allowed loads P that can be applied?

Figure 20.

Pmax based on moment capacity:

Mmax = 242 kip -ft + 22P

917 kip -ft = 242 kip-ft + 22P

P = (917 kip -ft - 242 kip-ft) = 30.7 kips = Pmax
22
Pmax based on max shear capacity:

The reaction at the supports are given by:

V1 = V2 = 22 kips + 3/2 P
50.1 kip = 22 kips + 3/2 P
Pmax = (50.1 kips - 22 kips) 2/3
Pmax = 18.7 kips
Pmax = P allowed
We must choose Pmax = Pallowed = 18.7 kips due to
limits on web buckling capacity.

NOTE: Also check deflection

Copyright © 2013 Strongwell Corporation

All Rights Reserved 17-29 Rev. 0910
 Section 17 Metric
EXTREN DWB® Design Guide

WEB BUCKLING
WEB BUCKLING:

This section deals with the web buckling capacity of the double web member. Loads and/or
reactions applied to the beam can fail the webs of the beam by crippling at points of high
stress concentrations. The load resistance limit of the beam at these areas is referred to as
the web buckling capacity of the member. Web buckling capacity for the DWB is generally
critical in areas of support reactions.
TABLE 11

Beam Type Web Buckling (kips)

A-Basis B-Basis
8" DWB Hybrid or Glass 34.1 47.3
36" DWB Hybrid 45.0 50.1

To quantify the web buckling failure mode of the 8" DWB and 36" DWB, a series of full section
tests was conducted. Allowable capacities are summarized in the accompanying table.

The web buckling tests consisted of loading the top flanges of various lengths of beams
through 4" x full width plates. The 4" plates simulated bearing pads in erected field conditions.
It is noted that short segments were utilized in the web buckling test as opposed to full-
length beam conditions seen outside of the laboratory. The tests did not incorporate vertical
bearing stiffeners. It is predicted that external vertical bearing stiffeners will increase the
web buckling capacity of the members. Due to the inherent design of the continual internal
horizontal stiffeners, external stiffeners will only be required in extreme loading conditions.

For most applications, other design considerations

(web shear, flexural stress, deflection, etc.) will
ultimately control the particular use of the beam.

Web buckling tests for the 36" DWB were

performed on 12" lengths cut from full sections.
These samples were centered in a test frame and
Crack in Fillet
fully supported by two (2) 9" x 18" elastomeric
bearing pads. Load was applied centrically to the
top flange via 1" x 4" x 1' - 6" steel plates under
load controlled conditions.
It is noted that all specimens were loaded until the
section would take no more additonal load (i.e.
additional application of load head only continued
to buckle the web with no increase in load). In all
samples the web buckled as shown in Figure 22.
Figure 21. Buckling of Web

Copyright © 2013 Strongwell Corporation

Rev.0910
17-30 All Rights Reserved
 Section 17 Metric
EXTREN DWB® Design Guide

CONNECTIONS
COMMENTARY:

This section deals with web-to-web framing connections of 8" DWB members. Connections
of this configuration (see Figure 22 and 23) are controlled by rotation and shear through
the elements, bolt bearing of the fasteners and any related distortion, and shear of the
fasteners.
Due to the performance of the EXTREN DWB® composite material and configuration, it is
recommended that steel clip angles and fasteners be utilized. Testing and methodology
described has been verified using stainless steel bolts and clip angles.

Test Series
Connection tests consisted of the following series of tests:
• Bolt Bearing Capacity
This initial battery of tests established pin bearing capacities and end/edge distances
for web/pin fastening. Steel pins were passed through holes drilled in the webs and
loaded perpendicular to the longitudinal axis of the pins until crushing was initiated in
FRP bearing area around the pins.
• Short Beam Connection Test
Utilizing the criteria developed in the above tests, full scale bolted assemblies were
fabricated and tested (see Figure 22 and 23). Short lengths of beams less than 1'-6"
were incorporated to isolate shear through clip angles, shear through fasteners, and
bolt bearing on fasteners/FRP.
• Long Beam Connection Test
Utilizing the criteria developed in the above tests, full scale bolted assemblies were
fabricated and tested (see Figures 22 and 23). Lengths of the beams were sufficient to
develop full shear and rotation through clip angles, shear and rotation through fasteners,
and bolt bearing on fasteners/FRP.

Testing has demonstrated that when stainless steel fasteners and stainless steel clip angles
are incorporated, the controlling element of the connection capacity is the bolt bearing into
or crushing the region around the fastener.

Copyright © 2013 Strongwell Corporation

All Rights Reserved 17-31 Rev. 0910
 Section 17 Metric
EXTREN DWB® Design Guide

CONNECTIONS
ANALYTICAL METHODOLOGY:

Maximum bolt bearing capacities from the series of connections may be estimated by the
following equations:

FPCr = FU / 3.0
P = (FPCr) (tw) (d)
where:
FPCr = Critical Bearing Stress (psi)
FU = Ultimate Compressive Web Bearing Stress (psi)
P = Pin Bearing Capacity (lbs)
tw = Total Web Thickness (in)
d = Diameter of Fastener (in)

Table 12
Allowable Bearing Capacities In Web Area of Section (in lbs.)
8" DWB (Web Thickness = .36") and Single Pin Fastener

Diameter of
Fastener Single Web Double Web
(in.) (lbs.) (lbs.)
1/4 900 1800
3/8 1350 2700
1/2 1800 3600
5/8 2430 4860
3/4 2700 5400
7/8 3150 6300
1 3600 7200

• Fastener Edge Distances (Web Area) - 2 diameters or 1" minimum, whichever is greater
• Fastener Pitch - 4 diameters or 3" minimum, whichever is greater

Copyright © 2013 Strongwell Corporation

Rev.0910
17-32 All Rights Reserved
 Section 17 Metric
EXTREN DWB® Design Guide

CONNECTIONS
ANALYTICAL METHODOLOGY:
Table 13
Allowable Bearing Capacities in Web Area of Section (in lbs.)
36" DWB (Web Thickness = .69") and Single Pin Fastener

Diameter of
Fastener Single Web Double Web
(in.) (lbs.) (lbs.)
1/4 1725 3450
3/8 2590 5180
1/2 3450 6900
5/8 4310 8620
3/4 5175 10350
7/8 6040 12080
1 6900 13800

• Fastener Edge Distance (Web Area) - 2 diameters or 1" minimum, whichever is greater.
• Fastener Pitch - 4 diameters or 3" mininimum, whichever is greater.

Example 1
Given: 8" DWB with ultimate compressive web bearing stress of 30,000 psi* and (l) 3/4"
diameter steel pin passing through 13/16" diameter holes aligned in the webs.
Find: Pin bearing capacity at one web and on both webs.
FPCr = FU / 3.0 = 30,000 psi / 3.0 = 10,000 psi

P = (FPCr) (tw) (d)

= (10,000 psi) (.36") (.75")
= 2700 lb. one web

Or 2700 lb. x 2 = 5,400 lb. two webs

* Minimum coupon properties and may be affected by enviro/mechanical conditions

Copyright © 2013 Strongwell Corporation

All Rights Reserved 17-33 Rev. 0910
 Section 17 Metric
EXTREN DWB® Design Guide

CONNECTIONS
ANALYTICAL METHODOLOGY:

Example 2

Given: Figure 20 – Clip angles 1/4" stainless steel (Fy = 48,000 psi) and 3/4" diameter
stainless steel bolts (Fv = 30,000 psi). Where Fy is stress of stainless steel and Fv is
bearing stress in FRP beam web.

Find: Capacity of connection assuming 30,000 psi* ultimate compressive bearing stress
of 8" DWB.

1. Find bearing capacity of 2 bolts in

8" DWB web area
FPCr = FU / 3.0
= 30,000 psi / 3.0 = 10,000 psi

P = (FPCr)(tw)(d)
= (10,000 psi)(.36" x 4)(.75")
= 10,800 lb.

2. Shear through stainless steel clips

ANet/Clip = 5.62" x .25" – (2)(.25")(.81") = 1.0"

Fv = .40 Fy = .40(48,000 psi) = 19,200 psi

PvSS= FvA = (19,200 psi)(1" x 2 clips) = 38,400 lb.

3. Shear of stainless steel bolts in double shear

Pbolts = FvANom
= (30,000 psi)(.442") x 2
= 26,520 lb.

4. Check pin bearing of stainless steel

PSS Pin Bear = .45 FyA
= .45(48,000 psi)(2)(.25" x .75")
= 18,000 psi
By inspection, the connection is controlled by bearing capacity of bolts on 8" DWB or 10,800
lb.
* Minimum coupon properties and may be affected by enviro/mechanical conditions

Copyright © 2013 Strongwell Corporation

Rev.0910
17-34 All Rights Reserved
 Section 17 Metric
EXTREN DWB® Design Guide

CONNECTIONS
ANALYTICAL METHODOLOGY:

Example 3

Given: Figure 20 – Clip angles 1/4" stainless steel (Fy = 48,000 psi) and 3/4" diameter
stainless steel bolts and rods (Fv = 30,000 psi).

Find: Capacity of connection assuming 30,000 psi* ultimate compressive bearing stress
of 8" DWB.

1. a. Find bearing capacity of 2 bolts in 8" DWB

web area
FPCr = FU / 3.0
= 30,000 psi / 3.0 = 10,000 psi

P8” = (FPCr)(tw)(d)
= (10,000 psi)(.36" x 4)(.75")
= 10,800 lb.

1. b. Find bearing capacity of 2 rods in 36" DWB

web area

P36” = (FPCr)(tw)(d)
=(10,000 psi)(.69” x 2)(.75”)
=10,350 lb.

2. Shear through stainless steel clips

PvSS= 38,400 lb. (See Example 2)

3. Shear of stainless steel bolts/rods

Pbolts = 26,520 lb. (See Example 2)

4. Check pin bearing of stainless steel

PPin Bear = 18,000 lb. (See Example 2)

By inspection, the connection is controlled by bearing capacity of bolts on 36" DWB or 10,350 lb.

* Minimum coupon properties and may be affected by enviro/mechanical conditions

Copyright © 2013 Strongwell Corporation

All Rights Reserved 17-35 Rev. 0910
 Section 17 Metric
EXTREN DWB® Design Guide

CONNECTIONS

NOTES: These details are framing and bearing connection details and are provided as
suggested construction details. The designer is cautioned that particular specific site
conditions may affect or require the altering of these details.

Figure 22. Web-to-Web Framing Connection

Figure 23. Web-to-Web Framing Connection

8" DWB to 36" DWB

Copyright © 2013 Strongwell Corporation

Rev.0910
17-36 All Rights Reserved
 Section 17 Metric
EXTREN DWB® Design Guide

CONNECTIONS

Figure 24

Figure 25. Section A—A; Bridge Bearing

Copyright © 2013 Strongwell Corporation

All Rights Reserved 17-37 Rev. 0910
 Section 17 Metric
EXTREN DWB® Design Guide

CONNECTIONS

Figure 26. Bridge Diaphragm

Figure 27. Deck Connection

Copyright © 2013 Strongwell Corporation

Rev.0910
17-38 All Rights Reserved
 Section 17 Metric
EXTREN DWB® Design Guide

CONNECTIONS

Figure 28. Steel Shelf Lug

Copyright © 2013 Strongwell Corporation

All Rights Reserved 17-39 Rev. 0910
 Section 17 Metric
EXTREN DWB® Design Guide

APPENDIX
Tom’s Creek Bridge
8" x 6" EXTREN DWB® Demonstration Project

The Tom’s Creek Bridge is a small-scale demonstration project involving the use of 8" x 6” EXTREN
DWB® hybrid beams as the main load carrying members in a short-span bridge. The Tom’s Creek
Bridge is located in Blacksburg, VA and was built during the Summer of 1997.
The project is intended to serve two purposes. First, by calculating bridge design parameters
such as the dynamic load allowance, transverse wheel load distribution and deflections under
service loading, the Tom’s Creek Bridge will aid in modifying current American Association of
State Highway and Transportation Officials (AASHTO) bridge design standards for use with
FRP composite materials. Second, by evaluating the FRP girders after being exposed to service
conditions, the project will begin to answer questions about the long-term performance of these
advanced composite material beams when used in bridge design.
This project involved replacing the superstructure in the Tom’s Creek Bridge, a rural short-span
(18 feet) medium volume vehicular traffic bridge with corroded steel girders and a timber deck.
Twenty-four (24) 8" DWB hybrid beams and a glulam timber deck with asphalt surface were used
to rehabilitate the bridge. In order to verify the composite girder design and to address construction
issues prior to the rehabilitation, a full-scale mock-up of the bridge was built and tested in the
laboratory. This set-up utilized the actual composite beams, glulam timber deck panels, and
geometry to be implemented in the rehabilitation. (Figure 29)
After the rehabilitation was completed, the bridge was field tested under a known truck load. Five
load tests nominally, at six-month intervals, were conducted. Using midspan strain and deflection
data gathered from the FRP composite girders during these field tests, the above mentioned bridge
design parameters were obtained. The Tom’s Creek Bridge was determined to have a dynamic
load allowance of 0.90, a transverse wheel load distribution factor of 0.101 and a maximum live
load deflection of L/490. Also, no significant long-term change in these parameters for the bridge
were noted over the 3 year duration.
Two 8" DWB bridge girders were removed from the Tom’s Creek Bridge after 15 months of service.
These FRP composite girders were tested at the Structures and Materials Research Laboratory
at Virginia Tech for residual stiffness and ultimate strength and compared to pre-service values
for the same beams. This analysis indicates that after 15 months of service, the FRP composite
girders had not significantly changed in stiffness or ultimate moment capacity.
For complete details about this project see the theses of Michael David Hayes and William Douglas
Neely at [Link]

Figure 29. Full scale mock-up Figure 30. Installation of the

of Tom’s Creek Bridge. Tom’s Creek Bridge.
Copyright © 2013 Strongwell Corporation
Rev.0910
17-40 All Rights Reserved
 Section 17 Metric
EXTREN DWB® Design Guide

APPENDIX

Route 601 Dicky Creek Bridge

36" x 18" EXTREN DWB® Demonstration Project

The Virginia Route 601 Bridge, spanning 39 feet over Dickey Creek in Sugar Grove, VA, is the first
use of Strongwell’s 36" x 18" EXTREN DWB® hybrid beams as the main load carrying members in a
low volume vehicular traffic bridge. The bridge was designed with the aid of the American Association
of State Highway and Transportation Officials’ (AASHTO) Standard Specification for Highway Bridges
for an AASHTO HS20-44 and alternate military loading with a targeted deflection limit of L/800. To
meet the deflection target, eight beams were required and spaced transversely at 3.5 feet. A glulam
timber deck was used with an asphalt overlay and the guard rail was a crash tested glulam system.
The photos below show the Route 601 bridge.
The experimental research related to the Route 601 Bridge consisted of two phases. The first phase,
completed in July of 2001, consisted of testing eleven 36" DWB beams (eight of these beams were
used in the bridge) to determine their stiffness properties (E and kGxyAv) to insure that these properties
were above the values assumed in the preliminary design. One of these eleven girders was then
tested to failure to determine the failure mode and flexural strength of the 36" DWB. The test of the
beam to failure revealed a safety factor of over 7 against the AASHTO service load.
The second phase began in October of 2001 after construction of the Route 601 Bridge was
completed and consisted of field testing the bridge to determine girder distribution factors, dynamic
load allowance, and service load deflections for the structure. To evaluate the in-service behavior of
the bridge, mid-span deflections and strains were continuously recorded during live load tests with
a vehicle slightly above the legal load limit for the bridge. The wheel load distribution factors in the
AASHTO Standard Specification for Highway Bridges for glulam timber decks on steel stringers were
found to apply to this bridge. A dynamic load allowance was determined to be 0.36 (slightly larger than
that specified in AASHTO), and the maximum deflection of the bridge was L/1100. This improvement
in deflection performance is attributed to partial composite action of the deck-to-girder connections,
bearing restraint at the supports, and contribution of guardrail stiffness. It was also found that the
absence of a midspan diaphragm had a minimal effect on the wheel load distribution factor.
For complete details about this project see the theses of Christopher J. Waldron and Edgar Salom
Restrepo at [Link]

Figure 31. Route 601 Bridge Superstructure

Copyright © 2013 Strongwell Corporation

All Rights Reserved 17-41 Rev. 0910