Comparative Analysis of Design Codes
for Timber Bridges in Canada,
the United States, and Europe
James P. Wacker and James (Scott) Groenier
The United States recently completed its transition from the allowable its companion code commentary (3) for Canada. For Europe, several
stress design code to the load and resistance factor design (LRFD) design standards were used, including EN 1995-1-1 Eurocode 5,
reliability-based code for the design of most highway bridges. For an Design of Timber Structures, Part 1-1: General—Common Rules
international perspective on the LRFD-based bridge codes, a comparative and Rules for Buildings; EN 1995-1-2 Eurocode 5, Design of Timber
analysis is presented: a study addressed national codes of the United States, Structures, Part 2: Bridges; EN 1990:2002 Eurocode, Basis of Struc-
Canada, and Europe. The study focused on codes related to timber tural Design; and EN1990:2002/A1 Eurocode, Basis of Structural
bridges and involved the following parameters: organization format, Design/Amendment A1, Annex A2: Application to Bridges.
superstructure types, loads, materials, design for bending, design for
shear, deflection criteria, and durability requirements. The investigation
found many similarities and some distinctive differences between the Design Philosophy
three bridge codes. Although the United States and Canada have differ-
ent design load configurations, these result in similar bending moments Canada, the United States, and the European Union all have unique
and shear effects over a typical span range. However, the design load con- structural codes for the design of timber highway bridges. The
figuration in the European code produces bending moment and shear Canadians adopted the limit states design (LSD) approach many years
effects that are two to three times greater than the U.S. and Canadian ago, and the United States recently transitioned to the LSD format
levels. The comparative design of a glulam girder bridge revealed that the for its bridge design code. The Eurocode was scheduled to convert to
smallest beam size was required by the Canadian code and the largest was the LSD format in early 2010. The U.S. and the European Union LSD
required by the European code. design codes use a calibration coefficient to convert from tabulated
strength properties in allowable stress design (ASD) to reference
design values in LRFD.
The Canadian bridge design code was one of the first to adopt the
limit-states design philosophy several years ago. In the United States, All codes use the same basic structural equations for flexure and
significant changes related to timber bridges recently have been shear, but they use different adjustment factors to modify reference
adopted into the AASHTO-LRFD Bridge Design Specifications (1). design values or specified strength values. All three codes consider
In addition, FHWA has required the load and resistance factor the ultimate limit states (ULS) and service limit states (SLS) but do
design (LRFD) code for all new bridges in the United States since not require fatigue limit states (FLS) to be considered for the design
2007. In Europe, the LRFD-based Eurocode for bridge design is of timber bridges.
scheduled for implementation in 2010. This paper presents an over-
view of the Canadian, U.S., and European bridge design codes,
highlighting similarities and differences that relate to the design of Organizational Format
timber highway bridges.
All three codes have dedicated sections for each type of material
(wood, steel, and concrete, including resistance values) and design
COMPARISON PARAMETERS factors (loads, load factors, and analysis methods). In Europe, how-
ever, information on the design of timber structures is located in a
The analysis used the fourth edition of the AASHTO-LRFD Bridge separate code, Eurocode 5. The Canadian code is the only one of
Design Specifications (1) for the United States and the 10th edition of the three with a section dedicated to fiber-reinforced-plastic (FRP)
the Canadian Highway Bridge Design Code CAN/CSA S6-06 (2) and materials. The AASHTO code includes companion versions in
metric and customary units, whereas the Canadian code uses only
metric units. The AASHTO code includes commentary in a side-
J. P. Wacker, Forest Products Laboratory, USDA Forest Service, One Gifford Pinchot
Drive, Madison, WI 53726. J. Groenier, Missoula Technology and Development by-side, two-column format; Canada and Europe include their code
Center, USDA Forest Service, 5785 Highway 10 West, Missoula, MT 59808. commentary in companion publications. The Eurocode is available in
Corresponding author: J. P. Wacker, jwacker@fs.fed.us. many languages but is based solely on metric units.
The Canadian code includes the evaluation (load rating) of bridges
Transportation Research Record: Journal of the Transportation Research Board,
No. 2200, Transportation Research Board of the National Academies, Washington,
within their bridge design code; this topic is covered in a separate
D.C., 2010, pp. 163–168. publication in the United States code. The Manual for Condition Eval-
DOI: 10.3141/2200-19 uation and Load and Resistance Factor Rating (LRFR) of Highway
163
164 Transportation Research Record 2200
Bridges is in use by engineers to evaluate and load rate bridges in the
United States (4). In a unique aspect of the Eurocode, each country can
produce a national annex that includes nationally determined param-
eters (NDPs) that modify EN 1995-2 with respect to load duration,
partial factors for material properties, deflection limits, damping
ratios, and other geographically specific data (i.e., climate, snow loads).
Although several states in the United States have design requirements
that are more stringent than those given by AASHTO, the requirements
must exceed the minimum requirements set by AASHTO. However,
European countries have more latitude for setting their NDPs above or
below Eurocode design recommendations.
FIGURE 1 Bridge design loading configuration used
in United States.
Superstructure Types
All three codes include design specifications for a variety of bridge (Canada), Strength I (United States), and Strength (European Union).
superstructures that use timber structural components. The lon- The basic load combinations used for the normal design vehicle and
gitudinal deck systems included in all codes are spike-laminated, a dead load without wind follow:
glued–laminated (glulam), and stress-laminated superstructures.
The transverse deck systems on beam girders included in both codes Canada:
are planks, nail-laminated, glulam panels, and concrete slabs. Specific
differences in the design of each superstructure systems in the various Q = 1.2 ⴱ dead _ load + 1.7 ⴱ live _ load
countries include the following:
United States:
• The U.S. and Canadian codes do not permit longitudinal (con-
tinuous) nail-laminated decks but do allow panelized nail-laminated Q = 1.25 ⴱ dead _ load + 1.75 ⴱ live _ load
decks, commonly referred to as spike-laminated decks.
• The Canadian code includes a composite nail-laminated con-
The design live loading for each code has different vehicle and
crete longitudinal deck system and permits mechanically spliced, uniform distributed load configurations. The AASHTO code uses an
butt-jointed deck laminations. HL-93 design load (Figure 1), the Canadian code uses the CL-625
• The Canadian code permits the use of FRP prestressing strands
design load (Figure 2), and the Eurocode uses Load Model One
for stress-laminated decks. (LM1; Figure 3) for the design vehicle or uniform distributed load
• The Eurocode contains design provisions for cross-laminated
combinations. At first glance, the U.S. and Canadian design vehicles
slab bridge designs consisting of several (flatwise) lamination layers look quite different because of their wheel spacing and axle loads.
that are glued or mechanically fastened into deck slabs with each layer Despite these differences, they yield about the same design moments
having a different grain direction (crosswise or at different angles). and shear (Figure 4). The Canadian design vehicle (CL-625) has a
gross weight of about 70 tons (625 kN). The AASHTO design
vehicle (HL-93) plus its lane load for a 50-ft bridge has a combined
Loads weight of 52 tons (463 kN). The Eurocode (LM1) design loading has
a combined weight of approximately 135 tons (600 kN) evenly spread
AASHTO and the Canadian code have dedicated chapters for load over two axles for spans of less than 10 m with the dual axles replaced
and load effects within their specifications. The Eurocode has separate with a single axle for spans greater than 10 m. A superimposed uniform
documents for bridge loadings with associated load factors provided distributed load of 9 kN/m2 is applied simultaneously.
in national annexes. These sections cover load combinations for
ULS, SLS, and FLS. The load factors and load combinations for the
Canadian code are typically less than those in AASHTO. AASHTO
has seven ULS load combinations, which it refers to as Strengths I
through V and Extremes I and II, whereas the Canadians have nine
ULS load combinations. AASHTO has four SLS load combinations,
and the Canadians have two. The Eurocode has four ULS—equilibrium
(EQU), strength (STR), geotechnical, and fatigue—and has two
SLS—vibration and deformation. For the design of timber bridge
superstructures via Eurocode, the ultimate limit states for EQU and
STR and the service limit states for vibration and deformation are
usually checked. Load factors for dead load and (bridges) live load
are provided in Eurocode by each member country in its National
Annex 2 to EN 1995-2.
Both codes have tables for permanent load factors that give the
maximum and minimum values used to produce the more critical
combinations for design loads. Only one load combination is discussed FIGURE 2 Bridge design loading configuration used
here, the main load combination for all codes, ULS Combination 1 in Canada.
Wacker and Groenier 165
300 kN 300 kN
(67.44 kips) (67.44 kips)
1.2 m
(3.94 ft)
Uniform load of 9 kN per square meter (187.97 psf)
(a)
600 kN
(134.89 kips)
Uniform load of 9 kN per square meter (187.97 psf)
(b)
FIGURE 3 Bridge design loading configuration used in Europe:
(a) spans of less than 10 m and (b) spans of 10 m or more.
Although the AASHTO design vehicle weighs less than the Cana- AASHTO does not require a dynamic load allowance for timber
dian design vehicle, the variable wheel spacing of the AASHTO design bridges on the assumption that wood is stronger for short-duration
truck develops more concentrated load than does the Canadian design loads than it is for long-duration loads. This increase in strength
vehicle. The Eurocode (LM1) design loading is significantly higher (by cancels the increase in force of dynamic loads. Canada requires a
a factor of 2 to 3) than its U.S. and Canadian counterparts. dynamic load allowance for wood, but that allowance is only 70% of
the allowance required for steel and concrete bridges. It is not yet clear
4000
whether individual countries will provide dynamic load allowance
3500 Canadian CL-625
as part of their Eurocode national determined parameters.
Live Load Moment (m-kN)
3000 AASHTO HL-93 The multiple presence factor (United States) and multilane loading
(Canada) are included in the codes to account for the probability of
2500 Eurocode LM1 more than one lane being loaded at a time (Table 1). The Eurocode
2000 uses a different approach to multiple-lane loading, whereby truck
1500 axle loads and uniform distributed loads are reduced in the second
and third lanes, with remaining areas having a 2.5 kN/m2 uniform dis-
1000
tributed load applied as well. For the U.S. code, the multiple presence
500 factors are integrated into the approximate equations for distribution
0 factors for bending and shear.
3 4 5 6 7 8 9 9 10 11 12 13 14 15 16 17
Span (m)
(a) Materials
900 All three bridge codes base their strength and design values on other
referenced codes. AASHTO references the 2005 National Design
800
Specification for Wood Construction for sawn-lumber reference
Live Load Shear (kN)
700
design values and the American Institute of Timber Construction’s
600
117-2004: Standard Specifications for Structural Glued Laminated
500
Timber of Softwood Species for glulam timber reference design
400
300
200 TABLE 1 Comparison of Factors
100
0 Number of U.S. Multiple Canadian Multilane
3 4 5 6 7 8 9 9 10 11 12 13 14 15 16 17 Lanes Loaded Presence Factors Loading Factors
Span (m)
1 1.2 1.0
(b) 2 1.0 0.9
FIGURE 4 Comparison for unfactored design loads: 3 .85 0.8
(a) live load moment and (b) live load shear. (Note: Graphs 4 .65 0.7
designed based on 3-ft span intervals and then converted 5 .65 0.6
to metric scale. Duplicate number 9 in each x-axis is due 6 or more .65 0.55
to round-off error.)
166 Transportation Research Record 2200
values. The design values in the AASHTO code are based on ASD for the Canadian code. The resistance factor is φ for both codes, but
values, at 19% maximum moisture content with a 10-year load AASHTO uses a resistance factor of 0.85 for flexure whereas the
duration. The Canadians reference CAN/CSA-086 for sawn lumber Canadians use a resistance factor of 0.9. Four adjustment factors are
and CSA 0177 for glulam timber. The design values in the Canadian common to both codes: beam stability factor, CL and kls; time-effect
code are based on LSD values at >20% moisture content with a factor or load duration factor, Cλ and kd; size-effect factor for sawn
1-month load duration. The Eurocode references EN 14081-1:2005 lumber or volume factor, CF or CV and ksb; and the deck factor or
for sawn lumber and EN 14080:2005 for glulam timber. The design load-sharing factor, Cd and km. AASHTO uses a few more adjustment
values in the Eurocode are based on ASD values at approximately factors such as Ckf for the format conversion factor to convert from
12% moisture content with a 5-min load duration. ASD to LRFD, CM for the wet-service factor, Cfu for the flat use factor,
The number of sawn lumber species covered by the codes varies and Ci for the incising factor.
from 11 in the AASHTO code to eight (four species combination The reference design values ( fm,y/z,d) in the Eurocode are determined
groups) in the Canadian code. The Eurocode contains timber strength by multiplying the characteristic (mean) bending strength ( fm,k) of a
classes instead of individual species groups (12 classes for softwoods timber component by the following three factors: kmod (modification
and six classes for hardwoods). The Canadians use only Douglas factor for moisture condition and service classes), for which a value
fir for glulam timbers and have specified strength values for four of 0.7 is appropriate for wet-service conditions and short-term traffic
combinations in bending, one combination in compression, and loading; ksys, the system strength factor; and kh, the modification factor
one combination in tension. AASHTO lists five species of trees and for member size effects. In conversion to a factored bending resistance
includes reference design values for 21 combinations for bending value, the reference design value is divided by the partial coefficient
and 15 combinations for compression and tension. The Eurocode for material properties (γM), for which a value of 1.30 is used for
includes both European lumber species and imported species from lumber and a value of 1.25 is used for glulam timber components.
Central America, the United States, and Canada. Canada uses a true LSD, so they do not require a format conversion
factor to convert from ASD to LRFD. The semiwet condition and
incising factor are already included in the Canadian code’s specified
Design for Bending strength tables.
The approach of both the U.S. and the Canadian codes to bending
strength design is very similar. They both use reference design or
Design for Shear
specified strength values and multiply them by adjustment factors
to calculate an adjusted design value. This adjusted value is used to Both the U.S. and the Canadian codes use a similar approach for shear
calculate a nominal flexural resistance value, which is modified by strength design. Both use reference design or specified strength values
the resistance factor to yield the factored flexural resistance value. and multiply them by adjustment factors to calculate an adjusted
The factored flexural resistance value must be larger than the total design value. This adjusted value is used to calculate a nominal shear
factor load for the beam in bending. resistance value, which is modified by the resistance factor to yield
The AASHTO code uses the following equations to determine
the factored shear resistance value. The factored shear resistance
factored flexural resistance:
value must be larger than the total factor load for the beam in shear.
AASHTO uses the following equations to determine factored
AASHTO 8.4.4.1-1:
shear resistance:
Fb = Fbo ⴱ Ckf ⴱ C M ⴱ (CF or CV ) ⴱ C fu ⴱ Ci ⴱ Cd ⴱ Cλ
AASHTO 8.4.4.1-2:
AASHTO 8.6.2-1:
Fv = Fvo ⴱ Ckf ⴱ CM ⴱ Ci ⴱ Cλ
Mn = Fb ⴱ S ⴱ CL
AASHTO 8.7-2:
AASHTO 8.6.1-1:
bⴱd
Vn = Fv ⴱ
Mr = φ ⴱ Mn 1.5
The Canadian code uses a single equation to determine factored AASHTO 8.7-1:
flexural resistance:
Vr = φ ⴱ Vn
M = φ ⴱ kd ⴱ kls ⴱ km ⴱ ksb ⴱ fbu ⴱ S
The Canadian code uses a single equation to determine factored
The Eurocode also uses a single equation to determine the flexural shear resistance:
resistance:
A
V = φ ⴱ kd ⴱ km ⴱ ksv ⴱ fvu ⴱ
fm, y z ,d = kmod i ksys i kh i km,k 1.5
The equations’ terms for design values and adjustment factors The Eurocode uses a single equation to determine factored shear
are similar. The design values for AASHTO are Fbo for the reference resistance:
design value and Fb for the adjusted design value. The reference
design value can be compared to specified bending strength, fbu, fv,d = kmod i ksys i fv,k
Wacker and Groenier 167
The equations’ terms for design values and adjustment factors are timbers, or (c) preservatively pressure-treated materials. The allow-
similar. The design values for AASHTO are Fvo for the reference able preservative treatments are similar with each providing preser-
design value and Fv for the adjusted design value. The AASHTO vative alternatives from creosote to chromate copper arsenate. All
reference design value can be compared to specified shear strength, three codes require galvanized metal fasteners and hardware.
fvu, for the Canadian code. The resistance factor is ϕ for both codes,
but AASHTO uses a resistance factor of 0.75 for flexure, whereas
the Canadian code uses a resistance factor of 0.9. The time effect or COMPARATIVE BRIDGE DESIGN
load duration factor, Cλ and kd, is the only adjustment factor for shear
that is common to both codes. AASHTO has a few more adjustment A bridge component was designed by using the different design
factors, such as Ckf for the format conversion factor to convert from loadings along with similar design assumptions so that a general
ASD to LRFD, CM for the wet-service factor, Cfu for the flat use comparison of the bridge design codes could be made. This compar-
factor, and Ci for the incising factor. The Eurocode calculates shear ative design analysis was based on a longitudinal glulam stringer
based on the characteristic value ( fv,k) based on the shear strength bridge with a transverse glulam deck. The single-lane bridge measured
parallel to the grain, which is modified by the factor for moisture 60 ft long and had a span distance of 59 ft (center-center bearings).
condition and service classes (kmod), for which a value of 0.7 is The transverse glulam deck was 5.125 in. thick and measured 18 ft
typically used, and the system strength factor (ksys). wide, providing a 16-ft roadway width. Five glulam beams spaced
The Canadian code uses a true LSD, so it does not require a format at 42 in. (center-center) supported the deck; these were Douglas fir
conversion factor to convert from ASD to LRFD. The semiwet condi- with a nominal bending strength of approximately 2,400 lb/in.2. Design
tion and incising factor are already included in the Canadian code’s loading was as required by each design code, and adjustments were
specified strength tables. The last two adjustment factors, km, modifi- made for wet-use exposure conditions. Live load deflection was
cation for load sharing, and ksv, modification for size effect for shear, limited to approximately L/400 for the glulam girder designs.
are found only in the Canadian code and the Eurocode. Specific design parameters were derived from the national annex of
Portugal for the Eurocode beam design.
The interior beam size required by each national bridge design code
Deflection Criteria is provided in Table 2. The required beam sizes for the United States
and Canada are similar with live load deflection controlling in both
All three bridge design codes set limits for the amount of deflection cases. The slightly deeper beam required in the United States was
at the SLS with the allowable deflection varying from L/400 for the attributed to different load distribution, deflection limits, and deflec-
Canadian code, L/425 for the AASHTO code, and L/300 to L/400 in tion loadings. The required beam size in the Eurocode design spec-
Eurocode 5, where L equals the length of the bridge. The Eurocode uses ifications is much larger at 9.5 × 63 in. for the Class I loading level.
mean values (versus fifth-percentile value for the United States and This requirement in Europe for a wider and deeper beam was attributed
Canada) for stiffness-related properties in the service load limit state. to the significantly higher design loads required for Class I loading
However, each member country of the Eurocode can impose alter- and resulted in a bending controlled design. If Class II loadings with
native deflection limits in their individual national annex document, a reduced axle and lane loading are considered, the Eurocode required
which modifies the requirements of Eurocode EN 1995-2. beam size becomes 9.5 × 45.5 in. and is much closer in beam depth
to that of the United States and Canada. Remaining differences are
related to the live load distribution to the girders and safety factors
Decks between North American and European bridge design.
The U.S. code has a separate section for decks, and the Canadian
code includes decks in the wood section. The Eurocode integrates SUMMARY
its deck design provisions within its bridge design code. The decks
covered in all three codes are glulam, stress-laminated, and nail- This study performed a comparative analysis of the national codes
laminated decks. The Canadian code also includes wood–concrete from the United States, Canada, and Europe related to the design of
composite decks in its specifications. The Eurocode includes provi- timber highway bridges. The analysis found many similarities and
sions for cross-laminated decks and stress-laminated decks consist- some distinctive differences among the three bridge codes. Although
ing of glulam beams that are prestressed and glued at the lamination
interfaces.
TABLE 2 Summary of Comparative Interior Beam
Bridge Design Analysis
Durability United States Canada European Uniona
AASHTO and the Canadian bridge code require timber used in Beam size 8.75 × 42 in. 8.75 × 40.5 in. 9.5 × 63 in.
(Class I)
bridges to be treated with preservatives applied by pressure treatment. 9.5 × 45.5 in.
AASHTO follows the AASHTO M 133 standard for allowable (Class II)
treatments and retentions. The Canadian code’s subsection on Deflection limit L/425 L/400 L/400
durability lists allowable preservatives and follows the Canadian Deflection loading 100% truck 90% truck + 100% truck +
Standards Association 080 series of standards. Both the U.S. and loads lane loads lane loads
Canadian codes reference the American Wood Preservative Asso- Controlling factor Deflection Deflection Bending
ciation standards. The Eurocode allows for designer choice from
(a) sufficient flashing or sheltering details, (b) use of naturally durable a
Eurocode calculations are based on bridge design requirements in Portugal.
168 Transportation Research Record 2200
the United States and Canada have different design load configu- ACKNOWLEDGMENT
rations, they produce similar live load effects for bending and shear.
The design load configuration of the Eurocode produces bending The authors thank Alfredo Dias of the University of Coimbra,
moment and shear effects that are significantly higher than U.S. and Portugal, for his generous contributions regarding the Eurocode
Canadian levels. A comparison design was performed for a 60-ft design work.
(Douglas fir) glulam beam bridge by using the design load con-
figuration for each national design code. The largest beam size of
9.5 × 63 in. was required by the Eurocode, whereas the beam size REFERENCES
requirements were smaller in North America. The United States
required a beam size of 8.75 × 42 in., and Canada required a beam size 1. AASHTO-LRFD Bridge Design Specifications, 4th ed. AASHTO,
of 8.75 × 40.5 in. The controlling design parameter was deflection Washington, D.C., 2007.
in the United States and Canada, whereas bending controlled in the 2. CAN/CSA-S6-06. Canadian Highway Bridge Design Code, 10th ed.
Canadian Standards Association, Ontario, Canada, 2006.
Eurocode (Portugal) design. The large differences noted between 3. CSA-S6-1.06: Commentary on CAN/CSA-S6-06. Canadian Standards
required glulam beam sizes in North America and Europe are most Association, Ontario, Canada, 2006.
likely associated with the different design loads, live load distribution, 4. Manual for Condition Evaluation and Load Resistance Factor Rating
and safety factors used in the design of timber highway bridges. (LRFR) of Highway Bridges. AASHTO, Washington, D.C., 2003.
Future work will focus on those key details used in each country
(or national code) in designing for durability. The General Structures Committee peer-reviewed this paper.
