A study of Abnormal Bridge Loading

Section: Asset Management

Project Contractor: TRL

Introduction:

Project Objectives:

In the UK, the current highway bridge loadings comprise HA, a Uniformly Distributed Load (UDL)
and knife edge load designed to represent the normal traffic stream, and HB which was designed to
model an extreme load caused by a single heavy vehicle. The Eurocode 1 part 3 ENV 1991-3: 1995
does not include HB loading but accounts for heavy vehicle loading by including a list of heavy
vehicle specifications that is intended to represent the current modular multi-axle vehicles in use
throughout Europe. Many similar vehicle types are in use in the UK and this project was set up to
examine the relationship between the effects of 45 units of HB loading and the Eurocode garage of
heavy vehicles.

Summary:

The comparison study utilised a structural analysis program incorporating an autoloading module to
calculate the maximum load effects for particular heavy vehicles. The vehicles were applied to a
number of sample bridge spans. The vehicles used were the Eurocode heavy vehicle garage, British
heavy vehicles identified form previous work and 45 units of HB. The vehicles were applied to the
sample structures in association with HA loading and the normal rules regarding displacement of
traffic on adjacent lanes were applied to distinguish between the effects of width with differing
vehicle types.

For the study the vehicles were run over a range of bridge spans, from 2m to 50m. The results indicate
that the relationship between the load effects of HB with associated HA loading and the load effects
of the existing and Eurocode heavy vehicles is complex. This relationship is however, capable of
simplification and could be defined graphically. Some output from this work shows the relationship
between HB loading, Eurocode heavy vehicle loading and modern heavy load trailers existing in this
country. It also highlights the difference in overall loading caused by vehicles that are wide enough to
displace the adjacent lane of traffic compared with those which are not.

Development of a Bridge Live Load Model for Serviceability Limit States

by Mayrai Gindy and Hani Nassif

Document type: Conference Proceeding Paper

Abstract: In 1994, the American Association of State Highway Transportation Officials

(AASHTO) introduced the Load and Resistance Factor Design (LRFD) Bridge
Specifications. The LRFD, based on a probabilistic design philosophy, was
calibrated based only on the ultimate strength limit state (ULS). Other limit
sates such as serviceability and fatigue were determined according to the
performance of existing bridges designed using previous design codes.

Live load deflection limits, historically carried over from previous code editions
dating back to the 1930s, are now made optional (except for orthotropic decks).
Such limits were first introduced in response to complaints of objectionable
vibration of some bridges. In an investigation by the Bureau of Public Roads,
bridges with objectionable vibrations, subjectively determined by human
 tolerance levels, were linked with deflections in excess of Span/800.

Although not strictly enforced by the LRFD, many state highway departments
specify their own deflection limits based on a rather arbitrary criteria. Wisconsin,
for example, has a relatively low legal limit of 80 kips (356 kN) yet enforces one
of the strictest deflection limits of Span/1600 whereas Michigan with the highest
legal load tolerance of 164 kips (730 kN) has a deflection limit of Span/800. In
addition, the live load used in the calculation of girder displacement varies
considerably from one state to another. For example, Idaho uses the HS20 truck
with dynamic load allowance, whereas California uses the HS20 truck plus lane
load plus the dynamic load allowance.

The use of load factors is another source of variation among state design
practices; Arkansas uses factored live loads while Colorado does not for the same
live load type. These variations among states can ultimately lead to non-
uniformity in design and result in a wide range of girder stiffness for the same
load demand.

The objective of this study is to present results of a reliability-based analysis of

live load deflections in steel girder bridges. The analysis uses actual truck traffic
load data to calculate maximum deflections. Girder displacement values are
forecast for different return periods (i.e. 1-, 2- ... 75-year) using extreme value
theory (EVT). Results presented herein provide an alternate approach for
developing a deflection limit state.

Bridge Live-Load Models

by Andrzej S. Nowak, M.ASCE, (Prof. of Civ. Engrg., Univ. of Michigan) and

Young-Kyun Hong, (Res. Assoc., Dept. of Civ. Engrg., Univ. of Michigan)

Document type: Journal Paper

Abstract: Live load is an important load component for highway bridges. The paper
summarizes the available data base and formulates the approach to
calculating maximum moments and shears for various time periods. The
live-load data base consists of truck-survey results. Moments and shear
forces are calculated for surveyed trucks and various spans.

The maximum load effects for time periods from one day to 75 years are
derived by extrapolations and simulations. Calculations are performed for
single-lane and two-lane bridges. For single lanes, the maximum moment
and shear are caused by a single truck for spans up to about 100 ft (30 m).
For longer spans, two trucks following behind each other govern. For two
lanes, the maximum effect is obtained for two trucks, side-by-side, with
fully correlated weights. For the maximum 75-year moment or shear, each
side-by-side truck is represented by the maximum 1.5-month truck. The
effect of a 1.5-month truck is about 15% smaller than that of a 75-year
truck.
CHAPTER THREE* BRIDGE LOADS
This section will identify and define what the NMDOT fundamentally recognizes and uses
as the loads and load combinations in its bridge design procedures. In using the Load and
Resistance Factor Design (LRFD) design method factored loads are compared to factored
resistances. The load factor applied to an individual load depends both upon the type of load
and the limit state under consideration. The Load Combinations and Factors presented in
Table 3.4.1.1 of the AASHTO LRFD Specs are to be used in New Mexico.

3.1* *LOAD TYPE DEFINITIONS AND LOAD MODIFIERS

Definitions of what normal and/or project specific loads and loading combinations are
generally included by NMDOT in the design process follow. In addition, practices currently
in use are described. The reader is referred to the AASHTO LRFD Bridge Design Specs,
Section 3 (Loads and Load Factors) for more thorough definitions of these terms. In
addition, the LRFD Specs provide direction and guidance concerning all other items or
matters not specifically covered below.

3.1.1 Load Modifiers

AASHTO LRFD article 1.3.2.1 required that *loads* be multiplied by both a load factor and
load modifiers. There are three modifiers related to the ductility, redundancy and
importance of the structure. The application of the modifiers related to ductility and
redundancy are straightforward. Some direction in the application of the modifier related to
the importance of the structure is however necessary. In the application of that modifier,
bridges crossing major rivers and all bridges carrying interstate traffic shall be considered
important bridges. All others shall be considered typical bridges.

3.1.2 Dead *Loads**

Dead *loads* consist of the weights of all permanent portions of an entire structure and
include weights of any anticipated future additions. Provisions are to be made in the design
calculations to add 30 psf for future wearing surface overlays to the deck. In addition, 15 psf
for use of metal stay-in-place *bridge* deck forms needs to be added. Also, future utility and
planned future *bridge* expansion effects shall be accommodated in the *bridge* design, if
known. The weight of the concrete wall barrier can be distributed equally to a maximum of
the three outside girders. In computing dead *loads*, Table 3.1A gives densities for various
materials. *

3.1.3 Live Loads

Live loadings that bridges must carry consist of the moving dynamic weights of motor
vehicles, pedestrians, equestrians, cyclists, and all others crossing the bridges. The
discussion below is limited to highway motor vehicle loads. The Design Live Load (HL93)
consists of a design truck or design tandem and a design lane load. With a few exceptions
the design truck or the design tandem is applied simultaneously with the lane load.

3.1.3.1 Standard Trucks and Lane Loads

The design truck and design tandem loads are shown in Figure 3.1A. The design lane load
consists of a load of 64 klf. The HL93 shall be applied as follows.

1. The design truck in combination with the design lane load.

2. The design tandem in combination with the design lane load.

3. For negative moments and pier reaction only, 90% of 2- design trucks in combination with
90% of a design lane. Rules for placing the trucks are in Article 3.6.1.3. The combination
producing the largest effect shall be used.
Figure 3.1A* *Design Truck and Design Tandem* DESIGN TRUCK DESIGN TANDEM 6
'-0 " 4'-0" 25 k 25 k 12.5 k 12.5 k 12.5 k 12.5 k ** 2'-0" Curb 0 .8 W 0 .8 W V (14'-0" to 30'-
0") 14'-0" 0 .2 W 0.1W 0.4W 0.4W 0.4W 0.4W 32 k 32 k 8 k 0.1W 2'-0" DESIGN TRUCK
6'-0" load lane width Clearance and 10'-0" V = Variable spacing - 14'-0" to 30'-0" inclusive.

3.1.3.2 Design Traffic Lanes

For the purposes of design, the bridge roadway is divided into longitudinal design traffic
lanes 12 feet wide. Fractional parts of design lanes shall not be used. Roadway widths from
20 feet to 24 feet shall have two design lanes, each equal to one-half the roadway width.
Both the standard truck and lane load use a width of 10 feet). The 10 foot loads may move
within the 12 foot wide lanes which, in turn, may move between curbs. Traffic lanes and
loads within lanes shall be positioned to produce maximum stresses in the structural
components being analyzed.

3.1.4 Dynamic Load Allowance

The highway live loads applied to structural bridge elements which are above the ground
surface are increased by a specified allowance to account for various dynamic, vibratory, and
impact forces. For components other than wood and structures other than culverts, the factor
to be applied to the static live load to account for dynamic effects (IM) is Component
Allowance Deck Joints- - All Limit States 1.75

All other components - Fatigue and Fracture 1.15

Limit States - All other Limit States 1.33

For Culverts the allowance is IM = 1 + 0.33 (1-0.125 De)> 1.0 De = Depth of cover in feet.

For timber components the allowance shall be half of the values specified above.

3.1.5 Centrifugal Forces

Bridges on curves are designed for a horizontal radial force, taking into effect any
superelevation. The centrifugal force is equal to the products of the truck or tandem axle
weights and the factor C: /gR/ /V/ /C/ 3 4 2 × = Where C is a factor, V is the design speed
in Ft/Sec., R is the radius of the curve in feet and g is the gravitational accelerator. Neither
lane *loads* nor impact on live *loads* shall be used in the computation of centrifugal
forces. The force shall be applied at 6 foot above the roadway surface measured along the
center- line of the roadway. Multiple presence factor shall apply.

3.1.6 Braking Forces*

The braking force per lane is to be taken as the larger of 25% of axle *loads* or 5% of axle
*loads* plus the lane load. All lanes carrying same direction traffic are to be considered
loaded. The load is applied at a center of gravity point 6 feet above the floor slab. The
longitudinal force is transmitted to the substructure through the superstructure. Multiple
presence factors shall apply. *3.1.7 Temperature, Shrinkage and Creep* *

3.1.7.1 Temperature*

Forces, stresses, and movements in *bridge* components and members due to temperature
variations (both atmospheric and in construc- tion materials) must be calculated and
provided for in such items as elastomeric bearing pads, expansion bearing devices, deck joint
sealing systems, and deck joint openings. New Mexico temperature ranges used in
movement and force calculations are listed in Table 3.1B. *Table 3.1B* *Temperature
Ranges* *South of I-40 (Moderate Climate)* Structure Type Temperature Range Min/Max
design temperature Steel 120°F (0° to 120°F) Concrete 80°F (10° to 90°F) *I-40
and North (Cold Climate)* Structure Type Temperature Range Min/Max design temperature
Steel 130°F (-20° to 110°F) Concrete 80°F (0° to 80°F) The full temperature
range is used in design of the superstructure because the structure is anticipated to have these
full movements during it’s life. The thermal movement used in the design of elastomeric
bearing pads shall be not less that 75% of the total anticipated movement due to temperature.
The assumption made is that the girders will not be placed on the pads at the upper or lower
end of the temperature range. The designer shall specify the range of temperature in which
the girders shall be placed. The size of deck joint seals and required deck openings may be
found in the appropriate standard drawings issued by the Department. These standards
consider factors in addition to temperature, such as creep and shrinkage. If the movement
length of a structure exceeds that given in the standard drawings, a special joint seal may be
required. If situations arise which require special consideration, contact the State *Bridge*
Engineer for assistance. Thermal expansion coefficient conversions are given by: Steel:
/T/ /LL/ Δ × × = Δ 000078 .0 Concrete: /T/ /LL/ Δ × × = Δ 000072 .0 where ΔL
is expressed in inches and L is expressed in feet and ΔT, the change in temperature, is in
°F. Generally the effect of temperature gradient need not be considered. *

3.1.7.2 Shrinkage

The coefficient of shrinkage for normal weight concrete shall be taken as 0.0002.
Alternately, LRFD Code Equations may be used.

3.1.7.3 Creep

The creep coefficient is based on AASHTO LRFD *Bridge* Design Specification, Article
5.4.2.3. The creep coefficient shall be estimated by AASHTO equation 5.4.2.3.2-1. Ψ(t,t/i/) =
1.9 kvs khc kf ktd ti -0.118 The following values for H, ti and t shall be used. For all other
parameters, the designer is referred to the AASHTO LRFD *Bridge* Specifications. H = 25
percent (Relative Humidity) ti = 7 days (age of concrete when load is initially applied) t =
1,000 days (maturity of concrete)

The prestressed concrete design strain shall be calculated as follows: where The axial force,
“F”, shall be based on the prestressing force after losses. The cross sectional area shall
be the area of the prestressed beam. The design creep shall be based on 75 percent of the
creep coefficient calculated by LRFD equation 5.4.2.3.2-1. It is assumed that approximately
25 percent of the prestressed beam creep will occur before construction. Design creep design
length ( ) = ⋅ 0 75. , Ψ /t ti /ε ⋅ *

3.1.8 Wind *Loads**

Wind *loads* applied to bridges are to be calculated according to the AASHTO LRFD
*Bridge* Design Specifications. The structural supports of signs, luminaires, and traffic
signals are designed according to the AASHTO Standard Specification for Structural
Supports for Highway Signs, Luminaires and Traffic Signals. Sound walls are designed
according to AASHTO’s Guide Specifications for Structural Design of Sound Barriers.
The wind load for structural supports and for sound walls shall be calculated using a design
wind velocity of 85 mph. The wind *loads* and forces applied to bridges are given in the
AASHTO LRFD *Bridge* design Specifications. These *loads* and forces are computed
from a base wind velocity of 100 mph. If a change is made to the base wind velocity due to a
higher known maxi- mum wind velocity, the new base wind velocity is shown on the plans
after approval by the State *Bridge* Engineer. Changes in the design wind velocity may be
made if perma- nent terrain features make such changes safe and advisable. *

3.1.9 Stream Pressure and Buoyancy

Forces* The structure and it components are designed to withstand the maximum stresses
induced by stream flow and buoyancy. The structural elements which may be affected by
these forces include substructure piers, footings, abutments, pilings, and walls to the super-
structure deck, girders, railings, connections, and others. /A/ /F/ = σ /E/ σ ε = The
NMDOT *Bridge* Section’s policy is to design bridges over waterways for the design
flood event, and to check the bridges to ensure they can withstand the 500-year flood event.
In checking for 500-year event survival, the load and resistance factors should be modified
so that a minimum “factor of safety” of 1.0 is attained. Scour, floatation, and the
connection between the superstructure and substructure are of particular concern for the 500-
year event. During a 500-year flood event, the high water elevation may overtop the
*bridge*. *Bridge* design should consider that in some cases this flooding could cause the
roadway embank- ment to fail. If this occurs the water level may drop and result in lower
stream veloci- ties. This actually may produce lower stream forces than the design flood. *

3.1.10 Earth Pressures

The minimum active earth pressure to be designed for in structures that retain fill such as
abutments and retaining walls shall be not less than an equivalent fluid pressure (for the soil)
36 lb/ft2 per foot of height. Although The Columb method is acceptable, Rankine's formulas
are normally be used to make the earth pressure determinations needed to appropriately
proportion retaining structures. Adequate drainage for the backfill shall also be provided in
the structure design to mini- mize any hydrostatic pressure buildup.

The various Rankine formulas to be used for earth pressure calculations are given in Figure
3.1B and C. The formulas provide values for active or passive pressures with or without
cohesion on flat or inclined backfill slopes. Note that one should not use the Rankine method
for Kp when α > 0. In addition to the active and passive earth pressures values presented in
Figure 3.1B, the at rest pressure Ko, should be used in rigid structures that do not yield from
earth pressure. The formula for Ko can be found in any soil mechanics text. Live load
surcharges are specified in AASHTO LRFD article

3.11.6.2. At *bridge* abutments which have approach slabs, a live load surcharge equal to 2
feet of soil is to be applied in both abutment and wing wall design. The surcharge is applied
to account for the pressures generated by the high compaction effort used in the backfill
behind the abutment.

*3.1.11 Seismic Forces*

NMDOT *Bridge* Section has set fourth the following Policy Statements for determining
seismic load factors fro the LRFD Code. Bridges crossing The Rio Grande shall be
considered critical bridges. Bridges along interstate highways shall be considered essential.
All others shall be considered “other” bridges. Structures subjected to earthquake
forces shall be designed to survive the strains resulting from the design earthquake motion.
Factors that are considered when designing to resist earthquake motions are:
1. The proximity of the site to known active faults.
2. The seismic response of the soil at the site.
3. The dynamic response characteristics of the total structure.
The design to survive earthquake forces shall be in accordance with 3.10 of the AASHTO
LRFD *Bridge* Design Specifications. Figure 3.1D is a map of New Mexico showing
contours of horizontal acceleration that are to be used in determining an applicable accelera-
tion coefficient (A) which then is used to decide the seismic performance zone of any
particular *bridge* site. Then depending on the complexity of design and the number of
*bridge* spans proposed, the method (procedure) of seismic analysis is selected. Referring
to Figure 3.1D, all bridges in New Mexico will only be subject to either seismic performance
zone 1 or 2 criteria. *

3.1.12 Vehicular Collision*

The provisions of Article 3.6.5. of the LRFD specifications shall apply as applicable. *3.1.13
Permit *Loads* * The current NMDOT Permit Vehicle, P327- 13, shall be used for
superstructure design only excluding deck slab. Refer to Figure 3.1E for a definition of the
vehicle. This is applied in AASHTO LRFD 3.4.1 Strength II. The Permit Load shall be used
for the following:
1. Interstate Bridges.
2. U. S. Route Bridges
3. N.M. 6, N.M. 26, N.M. 68, N.M. 522, N.M. 501, N.M. 502 Bridges and other N.M.
Bridges as designated by the State *Bridge* Engineer.
3.2* *LOAD COMBINATIONS*

Load combinations are presented in 3.4 of the LRFD *Bridge* Design Specifications. Each
component of the structure shall be designed to safely withstand all applicable load
combinations.

Figure 3.1B* *Rankine’s Formulas* a P p Kc2+H p K p 2c K H ) Passive Earth Pressure Coefficient: Kp =

tan 2 o 2 (45 ) 2 o (45 2 Active Earth Pressure Coefficient: Ka = tan Rankine's Earth Pressure Coefficients: H/3 H
z Pa p (or P ) a (or )p a K H H p K(or ) where: = Angle of friction of soil o p Por( ) Ka H- 2c Ka or PASSIVE
ACTIVE CASE CASE 2c Ka Forces due to Earth Pressure (Without Soil Cohesion): Forces due to Earth
Pressure (With Soil Cohesion): Wall height H = = Density of soil = Active earth pressure at depth z a p = Passive
earth pressure at depth z where: = z Ka p Kz = Active Earth Pressure Resultant Force: Pa = H Ka 2 2 1 1 2 2 H
Kp =P Passive Earth Pressure Resultant Force: p p z = K = z Ka a = Active earth pressure at depth z = Passive
earth pressure at depth z where: p = Cohesion of soil c 2c a K p K 2c + - 1 2 2 H K =P Active Earth Pressure
Resultant Force: - 2cH 2cH - P = H K 2 2 1 Passive Earth Pressure Resultant Force: p p K Kpa

Figure 3.1C* *Rankine’s Formulas for Sloping Backfill* z H H/3 Forces due to Eartth Pressures (for Inclined Backfill)
where: = Angle of friction of soil Rankine's Earth Pressure Coefficients for Inclined Granular Backfill: Active Earth Pressure
Coefficient: Ka = cos Passive Earth Pressure Coefficient: Kp = cos o cos cos - + cos 2 2 cos - o o - 2 cos 2 cos cos + cos - cos 2
cos 2 - o o - 2 cos 2 cos = Angle of Incline Pa p (or P ) a (or )p pz = K = z Ka a = Active earth pressure at depth z = Passive
earth pressure at depth z where: p = Density of soil = H Wall height Passive Earth Pressure Resultant Force: Active Earth
Pressure Resultant Force: 2 H Kp = P p 1 2 Pa = H Ka 2 2 1

Seismic Accelerations

Figure 3.1E* *Permit Design Live Load P327-13

Chapter 11. Loads

Covered bridges are exposed to load issues unique to their form and material of construction.
This chapter focuses on loads as they relate to evaluating and rehabilitating existing bridges,
but also includes some guidance relevant to designing new covered bridges.

In simple terms, the loads are categorized as dead (nontransient, often primarily based on the
bridge's self-weight) and live (transient, such as snow, and moving, such as vehicular and
pedestrian). Combinations of loads involving snow are especially important to covered
bridges, due to the lack of guidance in conventional bridge specifications.

Dead Load (Self-Weight)

In most short- to mid-span concrete or steel bridges, the stresses induced by the weight of the
design vehicle represent a large portion of the total stress in the primary components.
Stresses from the weight of these bridges, while not small, are usually much smaller than
those caused by the design vehicle.

Covered timber bridges, however, are unusual in that the stresses from their own weight
represent a significant part of the total stresses. The chord forces due to dead load commonly
are equal to, or even exceed, those caused by vehicular loads. This is due, in part, to the
weight of the roof and siding that conventional bridges do not have. The heavy timber
trusses also have relatively high weight-to-strength ratios, compared with the efficient beam
cross sections used in steel and concrete. This means that it is very important to predict a
timber-covered bridge's self-weight as accurately as possible.
The AASHTO bridge specifications suggest using 800 kg/m3 (50 pcf) for the density of
wood when determining dead load for a timber bridge. This value is conservatively high and
prudent for most common (i.e., uncovered) bridges and with most (but not all) timber
species. (For comparison, many building designs use a density of 560 kg/m3 (35 pcf) as a
default density.) The 800 kg/m3 (50 pcf) value was established when the specifications were
first published in 1935. It is strongly influenced by open timber structures, which are prone
to high moisture content. In addition, creosote wood preservative commonly was used,
which could add as much as an additional 160 kg/m3 (10 pcf) to the weight of the wood.

Therefore, a prudent first step in analyzing a covered bridge is to use 800 kg/m 3 (50 pcf) for
the assumed weight of the wood, based on the AASHTO guidance. If the results of that
assumption are acceptable, as evidenced by a calculated satisfactory live load capacity, then
no further estimate regarding the weight of wood is necessary.

However, if the results of the analysis indicate an unacceptably low capacity for live load,
then further investigation in determining the weight of wood may be warranted. The
following discussion provides guidance on those next steps.

Wood density is strongly influenced by its moisture content, which can vary widely with
environmental conditions. AASHTO's suggested density recognizes the potential for higher
moisture contents (and densities) when timber components are exposed to direct wetting
from rain. Covered bridge timbers, however, have lower timber moisture contents (this is the
very purpose of the bridge covers) and associated reduced timber densities. The bridges
inspected in the statewide Vermont study (Chapter 1) contained timbers with moisture
contents much lower than 19-20 percent (a commonly cited upper limit of dry wood in the
timber industry.) The aged and protected air-dried softwood timber found in most covered
bridges typically has unit weights ranging from 417-609 kg/m3 (26-38 pcf).

This issue is extremely important to accurately evaluate covered bridge capacities. Analyses
prepared using the standard 800 kg/m3 (50 pcf) density commonly would indicate the need to
rehabilitate the bridge to replace existing elements with high strength grade timber,
unusually large timber components, or even nontimber components. In some instances, to
provide the necessary factor of safety, the bridge would be unable to support any design live
load. However, use of site-specific wood densities usually leads to a substantial reduction (as
much as 30-40 percent) in the dead load forces and stresses.

This discussion of design wood density is intended to encourage use of site-specific unit
weights for the evaluating and rehabilitating of historic covered bridges. The selection of
unit weights should be based on standard timber references, such as:
 Timber Construction Manual.[9]

 Timber Bridges-Design, Construction, Inspection, and Maintenance.[4]

 Wood Handbook: Wood as an Engineering Material.[10]

The selection of an appropriate unit weight for components requires that a wood scientist or
other qualified evaluator examine the in situ bridge or small samples removed for such
purposes to determine the species of the various elements; these also must be determined for
allowable stress purposes. If some components (e.g., the floor planking) are made of dense
species, then unit weights appropriate to the component should be used, resulting in an
overall weight of the bridge as a summary of the individual components.

Further, the site-specific unit weight should be based on a reasonable estimate of moisture
content. A moisture meter can be used to determine the actual moisture content of
representative elements of the bridge. Those elements below the surface of the deck may
have higher values, especially if the bridge is relatively close to water. As an alternative,
actual samples of the wood from the bridge can be obtained and tested in a laboratory to
determine the actual moisture content. As noted above, it is common to find moisture
contents below the 19-20 percent threshold; this threshold often is cited as the difference
between dry and wet wood.
The 800 kg/m3 (50 pcf) density may be used safely in lieu of site-specific densities, if that is
desirable.

A related topic involves installing dry, versus green, new timber components during
rehabilitation of historic bridges or for the design of new covered bridges. Conventional
practice requires installing only dried primary structural components that would have unit
weights as discussed above. However, the relatively nonstructural components (siding, roof
boards, and some of the bracing) might reasonably be installed using green timber
components. Although this could represent a heavier load than that assumed by the designer,
many believe this option is acceptable, because those components will dry quite rapidly and
will soon reach the reduced unit weight, in most cases before the bridge is opened to traffic.
If green timber is to be used for bigger, primary components, then the conscientious designer
will make appropriate modifications to the unit weight, because the larger members will take
longer to dry. Another minor wood density consideration relates to the extra weight
associated with most modern pressure-applied preservative treatments. Although usually
quite small, common treatments could add another 16 or 32 kg/m3 (1 or 2 pcf), with the
exception of creosote, which adds substantially more, as noted above.

Vehicular Loads (Live Load)

Standard Design Vehicles

When working with any traffic-carrying bridge, a primary design issue is the selection of the
design vehicle. This initial subsection is not directly relevant to many covered bridges, but it
is presented here as appropriate background information.

The AASHTO Standard Specifications for Highway Bridges identifies three types of design
vehicle loads. The first two represent categories of individual vehicles and are routinely
referred to as the H or HS truck. The H truck configuration includes only two theoretical
axles and represents dump truck vehicles. The AASHTO specifications present information
related to two sizes of H-type vehicles: the standard 20-ton (18-metric ton (MT)) (i.e., H20
(M18) truck, as in figure 82) or a smaller 15-ton (13.5-MT) vehicle (the H15 (M13.5)).

Figure 82. H20 design truck vehicle (after AASHTO standard specifications) [8,000 lbs
= 3,632 kg; 32,000 lbs = 14,528 kg].

AASHTO identifies the conventional semi- or tractor-trailer vehicle as an HS truck

configuration. It is identical to the H truck, but with an extra 32,000-pound (14,528-kg) axle
representing the rear axle of the trailer. The standard HS20 vehicle weighs 36 tons (32.7
MT), and the smaller HS15 weighs 27 tons (24.5 MT) (see figure 83).
 Figure 83. HS20 design truck vehicle (after AASHTO standard specifications) [8,000
lbs = 3,632 kg; 32,000 lbs = 14,528 kg].

The third type of design-vehicular load is what AASHTO terms lane load. This uniform load
scheme represents a string of closely spaced H15 single trucks (with 9.15 m (30 ft) between
the rear axle of one vehicle and the front axle of the following vehicle), with a heavier H20
truck in the middle of the string. This type of vehicular load is important for long-span
structures, where slow traffic can lead to a bunching effect, with heavier loads than those
generated by higher speed traffic and traveling with more space between vehicles (see figure
84).

Figure 84. Lane load configuration (after AASHTO standard specifications) [18,000 lbs
= 8,182 kg; 11,818 kg, 640 lbs/ft = 954 kg/m].

These three vehicular load types evolved from the initial AASHTO specifications, published
in 1935. While each load type is a simplified representation of the diverse vehicle
configurations and weights that actually travel the roadways, this trio of AASHTO loads is
acceptably accurate for the purposes of designing most bridge components. Consistent
design load, unless it is completely unrealistic or radical, serves the structural design
profession well.

As the trucking industry has consistently managed to obtain permission from regulatory
agencies for bigger and heavier trucks, there has been commensurate pressure to increase the
design vehicle used for bridge design. Several years ago, States began addressing this issue
by adopting a scaled-up version of the HS20 vehicle, the HS25. Since then, this has become
a common design vehicle, albeit incompatible with the typical types of vehicles traveling
through a covered bridge. AASHTO also has an LRFD specification, in which an HL93
loading is used. The HL93 is an HS20 truck with the lane load added. These very heavy rigs
do not travel very often on the secondary roadways where most covered bridges are located.

Design Vehicle for Covered Bridges

Bridges built in the 19th century were not designed for these modern vehicular loads. Horses
and wagons crossed these bridges, along with whatever load that could be pulled through the
bridge's openings. A load of loose hay was not very heavy, but skids of logs moved on icy
winter roadways could be quite heavy. The commonly anticipated maximum vehicle load at
the time when the original covered bridges were built would have been significantly lower
than the single vehicle H15 load described above. The relatively high proportion of dead
load to total load in covered bridges has worked in these bridges' favor as live loads have
increased over their functional lives.

Most original covered bridges have been upgraded to safely support less weight than the
standard live loads. Planning for more community vehicles is appropriate in many instances
(e.g., oil trucks, loaded snowplows, school buses, and emergency equipment). Design load
H15 often is selected to simulate such vehicles. When fewer heavy vehicles use the bridge,
the design vehicle can be scaled back to H10 or less.
Selecting the design vehicle for use in rehabilitating an historic covered bridge is vital to
minimize the effect on the required work. The design vehicle should represent the absolute
lowest vehicle weight practical for the site-this cannot be overemphasized.

The selection of the live load design vehicle usually depends on the site of the covered
bridge. A covered bridge that provides sole access to a dead-end road must be able to
support a more diverse and complete set of vehicles than a covered bridge that travelers can
bypass easily. Likewise, vehicles of any weight will use a covered bridge immediately
adjacent to a heavily traveled roadway network more than they would a bridge in a remote
location.

Posted Weight Restrictions

Most covered bridges intentionally have been posted with a lower weight restriction than the
bridges could (and often do) support, to limit the number of heavy vehicles using the bridge.
This helps reduce bridge deterioration from overweight vehicles and extends the time
between required rehabilitations. In many cases, load restrictions support community
planning objectives. Regardless of the reason, select the smallest possible design vehicle to
minimize the potential abuse associated with heavy vehicles, even if these heavy loads might
cross the bridge safely.

Because few covered bridges can support the heaviest legal vehicles safely, a warning sign
that identifies the maximum weight vehicle allowed often is posted on these bridges. FHWA
suggests that the lowest weight restriction be 2.7 metric tons (MT) (3 tons). If a bridge
cannot safely support this minimal live load, closing the bridge to traffic may be the best
option.

There are likely to be more weight restriction violations over bridges in remote areas, so
bridge designers and owners in these areas must carefully weigh whether to rely on a load
posting or to close or rehabilitate the bridge.

Few means effectively prevent overweight vehicles from crossing a covered bridge. One
common method is to install a horizontal bar over the roadway at the entrance of the bridge,
positioned in a way to provide a restricted vertical clearance. A restricted clearance (for
example, one at 2.4 m (8 ft)) prevents larger vehicles from crossing the bridge. Another
method is to introduce a restricted horizontal clearance by squeezing the approach guide
railing. At a minimum, this forces traffic to cross the bridge slowly and prevents some larger
vehicles from crossing the bridge.

Other Live Load Issues

Some covered bridges do not support vehicular loads. In those cases, other live loads,
including pedestrian load, bicycles, carriages, and lighter weight recreation vehicles, may
have to be considered and investigated. AASHTO published Guide Specifications for Design
of Pedestrian Bridges in 1997.[12] This publication restates the provisions of the AASHTO
standard specifications regarding live load allowances, although it provides a slight
reduction for certain circumstances. The allowance for pedestrian loads can be quite small in
most covered bridges, unless the bridge hosts large gatherings for some sort of event, in
which case the weight of a large crowd can be substantial. However, because large live loads
are usually vehicular, this section focuses on vehicular loads.

Impact load is another issue. A conventional bridge would be subjected to a combination of

live load and an impact allowance proportioned to the live load. However, bridges built of
timber do not require a consideration of impact load directly, because of wood's unique
ability to absorb normal impact load without distress. This issue is discussed in more depth
later in this chapter in "Load Combinations and Load Duration."

Snow Load

Most newer, conventional, deck-type bridges are not designed for the weight of snow,
because standard practice involves removing snow from the bridge with plows. Typically, it
is assumed that the bridge does not have to carry both heavy snow and heavy vehicles at the
same time.

In contrast, covered bridges might have snow on the roof at the same time that vehicles pass
through it. The bridge would therefore have to support both snow and vehicular loads. In
northern States, design snow loads can become quite heavy 2.4 KPa (50 psf) or more. In
some of those States, covered bridges carry snow load on the deck level; snow is deliberately
plowed onto the deck to allow snowmobiles to pass. This thin layer represents a light load,
which reasonable analysts might neglect. The melting snow, however, may cause some
decay in deck level timber components, so regular inspections are important. Figure 85
shows an example of the significance and consequences of snow load on covered bridges-
this one collapsed from snow without vehicle load, March 8, 2001. A replica bridge has
since been erected.

Figure 85. Snow load on covered bridges can cause failure-Power House Bridge, Johnson, VT.

However, because the AASHTO bridge specifications do not address this issue, it is up to
the covered bridge design engineer to select a prudent snow load and live load combination.
Many engineers who have experience with covered bridges believe that assuming a covered
bridge must support both a full weight design vehicle and full weight snow load
simultaneously is too conservative. If a covered bridge has a common 6:12 pitch or steeper
roof, the vibrations from vehicles that travel across it cause heavy accumulations of snow to
slide off the roof. For bridges with fairly flat roofs, however, designing for both loads could
be prudent. Most building design codes differentiate between flat roof and sloped roofs (in
which the latter has a slope of more than 5-30 degrees, depending on the specification) in
determining design snow loads.

A bridge's tendency to shed snow load is also a function of the roofing material. A metal
roof sheds snow load much more readily than does a roof with wood shingles. Bridges in
heavy snow areas, therefore, may benefit from a metal roof. In addition, metal roofing
systems are lighter than other types of roofing systems.

If the bridge is closed to vehicular traffic, then full snow load should be anticipated and
evaluated. The snow can become quite heavy and can represent a significant load on the
structure-in some cases much more than the weight of the design vehicle.

The design snow load magnitude is addressed in many specifications, but one quite
commonly cited is the ANSI/ASCE 7, Minimum Design Loads for Buildings and Other
Structures.[8] Many State governments have special snow load maps to provide additional
guidance on appropriate snow load designs; most of these target building design but are
appropriate for covered bridges. As noted above, the designer of any work on a specific
covered bridge should select the appropriate combination of snow load and group load
factors.

Considering uniform snow load is important; however, understanding asymmetrical snow

load is perhaps more important, and often controls the analysis of this type of load.
Asymmetrical loading may occur from wind-drifted snow or when the snow on one side of
the roof slides off before the other side. It is important to address both uniform and
asymmetrical snow loads.

Wind Load

Wind load is important to covered bridge design because the wind's relatively large
projected areas can develop substantial forces in the bridge. The AASHTO standard
specifications do not provide sufficient guidance for developing design wind loads that
address the sloped roofs found on most covered bridges. ANSI/ASCE 7[8] provides more
specific guidance regarding wind load; it contains provisions for wind pressure coefficients
against the sides and roof-both windward and leeward sides.

Wind load is important for the design of several components in covered bridges:
 Transverse and vertical knee braces.

 Upper chord level, horizontal lateral bracing.

 Lower chord level, horizontal lateral bracing (if used).

Figure 86. A covered bridge destroyed by wind-Bedell Bridge between Haverhill, NH,
and Newbury, VT, 1979.

Before it was destroyed, the bridge in figure 86 had served successfully for more than 110
years. The bridge was destroyed shortly after a rehabilitation that made two significant
structural modifications. First, the siding was extended closer to the eaves of the bridge,
thereby increasing the area exposed to wind forces and concurrently reducing the ventilation
strip that could have allowed some of the wind pressure to pass through the bridge. Second,
the two-span Burr arch truss bridge was augmented with arches that had been seated in
pockets cast into the abutments and pier. The rehabilitation had eliminated the pockets by
filling them with concrete to reduce opportunities for rot due to trapped water. The arches
were then butted against a flat surface. Eliminating the pockets removed the lateral support
of the arches, thereby decreasing the bridge's ability to withstand the lateral forces from
wind loads. Although the exact cause of the bridge's failure cannot be established, these two
structural modifications may have contributed to its demise.

A second example of bridge failure caused by wind is shown in figure 87. Figure 88 shows
alterations to the overhead bracing system in this bridge; these were made shortly before its
collapse. The bracing reconnections in figure 88 were quite unusual; given these, it was
likely that problems would occur.
 (photograph by Jim Smedley)

Figure 87. Another example of collapse by wind-Smith Bridge at Brownsville, VT.

(photograph by Ed Barna)

Figure 88. The Smith Bridge before collapse.

These bridge failures demonstrate the importance of considering the ramifications of

proposed structural modifications during rehabilitation projects.

Other Loads

Other load conditions that might reasonably be considered for covered bridges under some
special situations include seismic events (earthquakes), thermal differentials, erection
conditions, longitudinal and centrifugal traffic loads, and loads associated with flowing
streams and ice. Of these loads, only those related to stream flow (including debris loads
during floods) and/or ice forces have much relevance to most covered bridges.

In most instances, covered bridges that have survived for many decades have lasted only
because they are not exposed to such water-related forces. In some cases, bridges have been
able to withstand the forces related to occasional loads imposed by streams or ice floes.

As design conditions, these forces generally are not relevant to sizing members in a covered
bridge. However, these rare but heavy loads do mean that consideration should be given to
potential details to help strengthen a bridge in ways that counteract the rare instances of
these unusual loads; these are primarily aimed at bracing or support connection details.
These issues are discussed in relevant sections of this manual.

As with any structure, erection conditions for rehabilitation of covered bridges warrant
careful evaluation. Timber trusses often are lifted individually to facilitate the work; in some
instances, the entire bridge is lifted. Bracing the compression elements remains vital during
such lifting operations. Appendix B demonstrates a truss skeleton failure due to an erection
overstress during a relocation operation.

Load Combinations and Load Duration

Designing steel or concrete structures involves considering combinations of load; therefore,

most structural engineers are quite familiar with assessing the probability of various load
combinations. As an example, it is unlikely that any bridge will need to resist full design
vehicle load, design wind load, and the structure's self-weight simultaneously. Hence, design
specifications provide for a load reduction factor that depends on the load combination. The
precise load combination reduction factor varies among design specifications; therefore,
designers must be familiar with the specification governing the project in question. In most
cases involving covered bridges, the design is based on the AASHTO specifications; hence,
various combinations of dead, live, snow, and wind are usually involved. Accordingly, load
combination factors from a low of 1.25 to a high of 1.4 are possible. However, combinations
of wind load only rarely control primary load-bearing timber truss components in covered
bridges. The controlling load combination is almost always dead load plus live load, which,
according to the AASHTO specifications, has no load combination reduction factor for
single-lane (or even two-lane) structures. As mentioned in the earlier discussion about snow
load, AASHTO does not directly address this issue, leaving it to the designer to select an
appropriate combination of dead, live, and snow loads, and a corresponding load
combination factor.

To clarify, the following depicts the differences in load combinations by the two commonly
cited specifications used in work on covered bridges: AASHTO's Standard Specifications
for Highway Bridges[2] and ANSI/ASCE 7[8]. Those combinations relevant to the review of
the covered bridge superstructure are based on service load (also known as allowable stress)
design philosophy. The following load combinations were extracted from Table 3.22.1A in
AASHTO's Standard Specifications for Highway Bridges:
 Group I - (Dead + Live) at 100% of Allowable Stress (i.e., Load Combination Reduction Factor = 1.0).

 Group II - (Dead + Wind) at 125% of Allowable Stress (i.e., Load Combination Reduction Factor =
1.25).

 Group III - (Dead + Live + 0.3 Wind) at 125% of Allowable Stress (i.e., Load Combination Reduction
Factor = 1.25).

ANSI/ASCE 7[8] (ASCE 7-98 Section 2.4) provides the following "Combining Nominal
Loads Using Allowable Stress Design":
 Combination 1 - Dead

 Combination 2a - Dead + Live

 Combination 2b - Dead + (Live + Snow)*0.75

 Combination 3a - Dead + Wind

 Combination 3b - Dead + (Wind + Live)*0.75

 Combination 3c - Dead + (Wind + Live + Snow)*0.75

Timber components can absorb loads applied over a short time without apparent distress a
characteristic not true of steel or concrete. In fact, timber components can often accept twice
as much load as steel or concrete, if the load is applied suddenly. This material behavior
specific to timber is addressed through a load duration factor, which may range up to 2.0,
as provided in the NDS. Because this load duration concept does not apply when designing
with steel or concrete, it is not familiar to most engineers and can be very difficult to
understand.

A common misconception about load duration factors relates to the concurrent application
of both a load combination factor and a load duration factor. Designers should understand
that the duration of load factor relates to the behavior of timber as a material, while load
combination factors relate to the probabilities of concurrent design loads. Nonetheless, the
topic is confusing because the applicable load duration factor depends on the combination of
individual loads being considered. A good discussion of this topic is presented in the
Commentary on the National Design Specifications for Wood Construction.[13] NDS focuses
on timber and thoroughly addresses load duration, while only referencing the pertinent
design specification for load combination factors (for covered bridges, this is usually
provided in the AASHTO specifications).

The following example is provided to demonstrate how the controlling load combination is
determined. Note that this examination is required for each component in the bridge, and
different combinations of load and duration of load factor may control the design of different
elements.

Consider the following two load scenarios for a timber structure:

(A). A live load (LL) of 100 PSF and a dead load (DL) of 60 PSF

(B). A live load of 20 PSF and a dead load of 80 PSF

The controlling load combination will either be LL + DL or DL alone, and it is determined

by dividing the load combination by the duration factor (CD) of the least-duration loading in
each combination.

For scenario (A), the two combinations are:

LL + DL => 100 PSF + 60 PSF = 160 PSF (1)

For this combination, the CD for LL is 1.15 and for DL is 0.9. The least-duration CD is 1.15.
Therefore, the combination is divided by the CD, which yields:

160 PSF / 1.15 = 139 PSF

DL=> 60 PSF (2)

For this combination, the only CD is that for dead load, which is 0.9. Thus, the following
division is performed:

60 PSF / 0.9 = 67 PSF (3)

One then compares the two results, and the larger quotient indicates the controlling load. In
this case, the quotient for the LL + DL combination is the larger, thus the timber structure
would be designed for a load combination of 160 PSF. (It is important to note that the
quotient of 139 PSF is not used for design or any other purposes, only to determine the
controlling load combination.) In this instance, the stresses caused by the load of 160 PSF
would be compared against the basic allowable stress multiplied by the load duration factor
of 1.15.

Although it may seem obvious that a structure should be designed for the higher total load
combination (as it would for steel or concrete), a review of scenario B indicates how such an
assumption is misleading. Performing the same computations, but with the values indicated
for B, shows that, again, the two combinations are:

LL + DL => 20 PSF + 80 PSF = 100 PSF (4)

For this combination, the CD for LL is again 1.15 and for DL is again 0.9. The least-duration
CD is 1.15. Therefore, the combination is divided by the CD, which yields:

100 PSF / 1.15 = 87 PSF

DL => 80 PSF (5)

For this combination, the only CD is that for dead load, which is 0.9. Thus, the following
division is performed:
 80 PSF / 0.9 = 89 PSF (6)

By comparing the two results and remembering that the larger quotient indicates the
controlling load, one finds that the load combination of dead load alone governs in scenario
B. Accordingly, the entire design would be dictated by using a total governing load of 80
PSF (from just dead load), not the higher combined live plus dead load of 100 PSF. The
stresses caused by the load of 80 PSF would be compared against the basic allowable stress
multiplied by the load duration factor of 0.9.

In most cases in covered bridges, the combination of live plus dead is the controlling load
combination. However, this always must be checked, and it is especially important when (1)
the dead load to live load ratio is high, or (2) when any of the load duration factors are
significantly above 1.0.

Because the issue of snow load in combination with vehicular load is not directly addressed
in any nationally recognized design specification, the following guidance, which has been
used by several engineers when confronted with this issue, may be worth considering:

Proposed load combination: Dead load plus live load (vehicular), plus snow (either uniform
or drifted/unbalanced)-use a group load reduction factor of 1.33 (representing a probability
of occurrence).

This is based on the ASCE 7 discussion of load combinations. This proposal is somewhat
more liberal than a strict review of the AASHTO bridge specifications; these can be
interpreted to proscribe only a 1.25 factor. Proposed load duration factor for this load
combination comes directly from the NDS specifications:

For the combination of dead, live, and snow loads, the load duration factor would be 1.15.

The AASHTO bridge specifications use 1.15 for a two-month duration, which AASHTO
also parenthetically identifies for vehicle live load. NDS has traditionally associated the two-
month duration with snow load on timber structures.

However, some engineers believe this load combination and load duration factor is still
overly conservative when reviewing extant covered bridges, because the results of an
analytical investigation using this combination often indicate major weaknesses in structures
that have successfully supported loads for a long time.