U.S.

Department of Transportation

Federal Railroad Administration

ASLRRA/FRA STUDY
CAPABILITY EVALUATION
OF
SHORT LINE TIMBER TRESTLES
TO
HANDLE 286K LOADS

AREMA CONFERENCE
PRESENTATION
AUTHOR:

John Horney, P.E.

Carter & Burgess, Inc.
AREMA Member
AREMA Committee 10 Member

ABSTRACT:

This report will be a joint presentation between the Federal Railroad Administration
(FRA) and Carter & Burgess, and will provide the results of a study to evaluate the
capability of short line railroad timber trestles to handle 286k freight cars. This study
was funded by the FRA, sponsored by the American Short Line and Regional
Railroad Association (ASLRRA), completed by Carter & Burgess, and is available
through the ASLRRA.

The presentation will initially identify the background, need, financing, support and
availability of the study followed by a discussion of the content of the formal report
and the computer program used to evaluate the timber trestles. The formal report
includes a Project Understanding; Cooper Loading Information; Equivalent Cooper
Loading for 286k Cars; Timber Bridge Evaluation Discussion; Bridge Condition
Considerations; Analysis Methodology; Computer Program Matrix Entry Instructions;
Compliance Criteria; and Bridge Nomenclature. The presentation will then discuss
the computer program matrix, how the user will enter the information, and the
program’s output analysis. In summary, the program will guide and direct the user to
enter specific detailed information about a timber trestle. Using this data, the
program will analyze the structure and advise the user if the bridge is either capable,
marginal, or not acceptable for 286k cars. If marginal or not acceptable, the program
will provide a list of exceptions as well as a stress summary as developed in
AREMA’s Chapter 7.
INTRODUCTION:

This project has been funded by the Federal Railroad Administration (FRA) with the
American Short Line and Regional Railroad Association (ASLRRA) performing the
contract administration. The ASLRRA initiated a committee with representatives from
five short line railroads to oversee, review, and provide comments throughout the
development of the program. Carter & Burgess, Inc. was selected to develop the
program. On behalf of Carter & Burgess, Inc., I would like to thank:
Mr. Don Plotkin with the FRA,
Mr. Steve Sullivan with the ASLRRA,
And the following members of the ASLRRA Committee:
Mr. Chris Dodge with Omnitrax,
Mr. John Porter with Rail America,
Mr. Bill McKillp with Dakota, Minnesota and Eastern,
Mr. P. Sheldon with Iowa Interstate, and
Mr. R. Buchan with Paducah & Louisville.

The objective of this project is to develop a program for the Short Line Railroad
industry that will provide a consistent methodology for evaluating their timber bridge
inventory. One of the realities of Short Line Railroad operations is that their operating
staff may not include an experienced bridge engineer. This method of evaluation is
in the form of a questionnaire and computer program that will provide the user with
instructions and diagrams to help explain the detailed information that is needed and
how to enter the data necessary for the program to properly analyze the timber
structures. The program has a computer matrix analysis that will characterize a
timber bridge and determine its capability to safely support the 286,000-pound cars.
To accomplish this, the program must consider the method of bridge construction,
environment, train speeds, and both mechanical and organic decay. With this
program, the FRA, ASLRRA, and railroad industry will have a system that will form
the basis for a bridge management program.
WHAT’S OUT THERE NOW?

According to a 1992 survey conducted by the Federal Railroad Administration (FRA),

there are approximately 36,171 timber trestle railroad bridges throughout the United
States. On an average, each bridge is about 120-feet long and made up of 8 spans.
About half of these timber bridges are located on Class I railroad main line track
sections leaving at least half of the remaining timber bridges located across
secondary branch lines or on Short Line Railroads. Based on the average age of the
timber railroad bridges, 55% are in need of replacement or major repairs and another
20% will join this category within the next 10 years. In conjunction with the age and
condition of the timber railroad bridges, there is increasing pressure for the railroads
to ship 286k capacity rail cars across these structures. With the combination of these
events, there is a growing concern to find a method for the Short Line Railroad
industry to make an initial evaluation of their timber bridge inventory and determine if
they may have the capability to safely ship 286k revenue cars.

The focus of the program questionnaire is on the individual timber trestle component
sizes, quantities, and method of construction requiring detailed information for each
span of each timber bridge. Specific questions are asked that will determine the
physical condition of key components of the trestle by use of a rating system.

The results of this project will provide the Short Line Railroad Industry with a
consistent method for the Engineering Industry to evaluate the capacity of these
timber trestle structures. The Short Line Railroads could use the results of this
Report as a tool for developing their Planning Programs and Budgets.

AREMA – WHAT DOES IT SAY?

The American Railway Engineering and Maintenance of Way Association (AREMA)

provides the guideline for design, rating and inspection of timber railroad bridges.
Chapter 7 of the AREMA Manual is devoted to timber structures and is made up of
three parts: Material (15%), Design (62%), and Inspection (23%). The design portion
consists of mostly graphs, tables, and charts that makes up about 85% of the section.
This program draws information from the AREMA Manual with a focus on inspection,
rating, and design.

AREMA identifies the inspection as an on-site visual and physical investigation that
includes any timber sounding and drilling as required. The rating is described as
using allowable design stresses to compare against actual stresses that take into
account any organic and/or mechanical deterioration. The AREMA design section
provides design requirements, allowable stresses, and various combinations of
material for both open-deck and ballast-deck timber bridges that are rated for Cooper
Loads of E-60, E-72, and E-80, with the E-80 being the current standard loading for
new railroad structures.

COOPER LOADING DIAGRAMS

[reference?] To place the E-80 live load in better perspective, the two Cooper steam-
locomotives that define an E-80 load represent 1,136-kips distributed over 104 feet.
The trailing freight cars in the Cooper train represent a train of 40-foot long, 315,000-
lb. cars. This is similar to a train of 100-ton ore cars, 32'-7½ " long, over coupler
pulling faces. It should be noted that the specified 315,000-lb. car (and the ore car)
do not meet the present AAR vehicle criterion by exceeding the 263,000-lbs. gross
rail load. Thus, the conservative design load for the Cooper E-80 train calls for
a highly concentrated load compared to the heaviest diesel locomotives in
service today. And, any equivalent cars would be heavier and have closer axle
spacing than is permitted by the current AAR Specifications. In contrast, a
modern six-axle diesel locomotive is approximately 71-feet and may weigh 400,000-
lb; however, its axle spaces would distribute the loads beyond the limit of one timber
trestle span.
EQUIVALENT COOPER RATING

The Cooper Live Load Diagram is a common “yard stick” that railroad bridge
engineers use to measure the magnitude of trainloads on a railroad bridge. One of
the objectives of this project is to determine an equivalent Cooper Rating for 286k
freight cars as they relate to timber railroad bridges. The following represents the
wheel loads and spacing for a 286k freight car as used for this project. Although
there are 286k freight cars that may be longer, this diagram, using a shorter
wheelbase, will provide conservative values for the purposes of this project:
71.5 71.5 71.5 71.5

3.04 5.83 21.5 5.83 3.04

286k Car Diagram

When the Cooper E-80 Load is placed across a bridge, it will generate definable
forces throughout the structure. With this approach, the maximum forces will define
the value needed to size the various bridge components to resist the applied loads.
By moving the load to specific locations across the bridge, the maximum values will
be found to design for bending, shear, and compression stresses.

E-80 on Timber Trestle

286k Car on Timber Trestle

Just as with the E-80 Load Diagram, the 286k Load Diagram can be moved across a
bridge to determine its maximum loading values. If, for instance, the E-80 Load
would produce a value of 100 and the 286k Load produces a value of 50 (exactly half
that of the E-80 Load), then it can be said that the 286k load has an equivalent
Cooper Load of an E-40 (half that of the E-80). Thus, the 286k freight car loads, or
any other freight car, can be defined by its equivalent Cooper E-Rating by
comparing the maximum loads produced by each of the loadings. The 286k
equivalent Cooper Loads will vary for each span length and for the type of force
being considered i.e., bending, shear, etc.

The following tables reflect the load values produced by a 286k freight car, the load
values produced by a standard E-80 Load, and the associated 286k equivalent
Cooper Load for bridge spans ranging from 11-feet to 15-feet.

Equivalent
286k E-80
Span Value Value Cooper E
11 105.0 131.0 64.1
12 108.3 140.0 61.9
13 113.5 147.7 61.5
14 120.7 154.3 62.6
15 127.0 160.0 63.5

286k Equivalent Loads for End Reaction

TABLE 1

For each of the span lengths listed in Table 1, there will be a unique value for the end
bent reaction generated from the 286k freight car and a Cooper E-80 load. These
values are shown in Table 1 with the respective equivalent Cooper Load generated
by the 286k freight car. The values shown in Table 1 represent the reaction
supported by the end of a single span, or at the end bent of a multi-span bridge.
 286k E-80 Equivalent
Span Value Value Cooper E
11 123.0 174.6 56.4
12 137.0 186.7 58.7
13 148.0 197.0 60.1
14 158.0 208.6 60.6
15 166.5 218.7 60.9

286k Equivalent Loads for Intermediate Bent Reaction

TABLE 2

Just as described in Table 1, the values shown in Table 2 represent the loads and
equivalent rating generated by a 286k freight car and a Cooper E-80 load for an
intermediate bent reaction with the same length span on each side of the bent.

Table 3 represents the values for the bending moment of a beam. The bending
moment is the amount of bending that a beam must support.
Equivalent
286k E-80 Cooper E
Span Value Value Value
11 212.0 275.1 61.6
12 245.5 320.0 61.4
13 279.4 380.2 58.8
14 325.2 440.0 59.1
15 379.0 500.2 60.6

286k Equivalent Loads for Bending Moment

TABLE 3

It can be seen in the above Tables 1 through 3 that the 286k freight cars will produce
an equivalent Cooper Load between E-57 and E-64 with a typical value of an E62.
These tables represent an equivalent Cooper’s loading for bridge spans ranging
between 11-feet and 15-feet, and these span lengths represent the typical span
range for timber railroad bridges.
Similar equivalent Cooper Load Tables have been developed for the following types
of equipment:

6 – Axle, 200-Ton Locomotives

4 – Axle, 130-Ton Locomotives
6 – Axle, AC Traction, 210-Ton Locomotives
4 – DASH8, 402k Locomotives
4 – SD-60, 402k Locomotives

The above listed equipment represent typical loadings that will be found on most of
the Short Line Railroads, and none of them produced equivalent Cooper loads
greater than the 286k freight cars for the span lengths represented by this project.
For the purpose of this project and for the evaluation of timber bridges, the
286k freight cars will produce the most severe loads that timber bridges are
likely to support.

THE MATRIX EVALUATION

This program requires the user to respond to a series of questions that ultimately
identify the physical properties of the timber bridge. The questions address the size
and number of each of the bridge components that identify the physical shape and
size of the bridge. The user will be asked about the status of the bridge’s mechanical
and organic decay. Additional questions address the environment such as the bridge
height, scour, and the authorized train speed.

Bridge Description

Although this program will evaluate only timber bridges, the user will be asked to
enter the bridge description that include any steel or concrete spans. The program
recognizes the following types of bridge segments:
 Timber Trestle – Open Deck
Timber Trestle – Ballast Deck
Concrete Spans
Steel Spans

If, for example, a bridge is a single span open deck timber trestle, it will be identified
as One Segment. If the bridge has several open deck timber spans to a steel span
(or spans) it will be identified as having Two Segments. And, if the bridge had open
deck timber spans to a series of steel spans followed by concrete spans concluding
with ballast deck timber spans, it would be identified as having Four Segments. The
user will be asked to identify the overall bridge length and number of bridge
segments. Information will be asked about each segment. Detailed information will
be asked for all timber segments, but, general information will be requested for steel
and concrete segments. For steel and concrete segments, the user will be asked for
the number of spans, span length(s), and height of the span(s). The program will ask
for the type of steel span such as DPG (deck plate girder), TPG (through plate
girder), Bm (beam), or TT (through truss). This information about the steel and
concrete spans will be utilized to develop the owners bridge inventory. The user
could enter the general information on steel and/or concrete bridges that do not have
any timber spans for the purpose of generating a complete bridge inventory.

Structural Evaluation

In order to perform a structural analysis, the program will need to know specific
details of each component of the structure. It will be the responsibility of the user to
physically determine all the timber component dimensions needed to perform this
analysis. Based on information that the user provides, the program will perform
detailed calculations that will identify the stresses generated by 286k cars. There will
be numerous questions asked that will need to have a response. To help keep this
exercise as user friendly as possible, every effort is made to provide instant
duplication of responses by the computer for common timber sizes rather than
require the user to make multiple duplicate responses.

Dead Load Above the Stringers:

The program will develop a weight-per-foot-of-bridge for the rail, ties, Inside
Guardrail, tie spacers, hardware, footwalk and handrails. If the bridge is a ballast
deck structure, it will calculate the weight of the rail, ties, hardware, ballast, ballast
stops, ballast deck, and handrails. For ballast deck bridges, the weight of the ties are
included in the volume of ballast since the unit weights are nearly the same. Each
open deck bridge can have different size rail and ties, or the ties may have different
spacing that will all affect the value of the dead load. The answers provided by the
user will identify all of this information and generate a Dead Load that is unique to
that structure.

Dead Load and Stringer Characteristics:

The user will provide the number and size of each stringer in each span of the bridge.
The program will evaluate each span and each bent separately based on the
information unique to that span and bent. That evaluation will incorporate variations
of span lengths and the number or size of stringers in each span. The program will
determine the stringer Dead Load per-foot-of-track and add it to the Dead Load
above the stringers to obtain a total uniform dead load for each span across the
bridge that will be used to obtain the dead load bending moments and reactions. For
ballast deck timber trestles, the computer will generate a modified uniform dead load
that will analyze the main load carrying stringers. The modified uniform load will
neglect the outside stringer on each side of the bridge deck, ballast retainers and the
outside two-feet of timber deck and ballast on each side of the bridge.

Once the number of main stringers and their sizes are identified, the program will
develop the structural characteristics for the main stringers in each span, such as
area, weight, and section modulus, that will ultimately determine the stringer’s
stresses. For Ballast Deck Trestles, the program will use all the stringers to
determine dead loads for the piles. However, when evaluating the stringers, it will
omit the outside stringers and their respective dead load, as identified above, then
use the remaining dead load and all the live loads to be supported on the interior
stringers. Also, only the properties for interior stringers will be used to determine the
ballast deck stringer stresses.

Live Loads and Stresses:

This report will utilize 286k freight cars for its live load analysis. By applying dead
and live loads according to the AREMA guidelines, the respective loads and stresses
will be determined for each of the timber bridge components. The induced stresses
will be compared to the allowable timber stresses as identified by AREMA. Table 7-
2-9 is identified as “Allowable Unit Stresses for Stress Graded Lumber – Railroad
Loading (Visual Grading)” and provides allowable stresses for various species of
timber. For the purpose of this Program, minimum allowable stress values were
selected from the Table for Douglas Fir and Southern Pine. It is believed that these
two species of timber represent the vast majority of timber used throughout the
railroad industry in the United States. From this information, the following Table 4
reflects the allowable stresses utilized in this Program:
Beams Posts
Unit Stress in Extreme Fiber in Bending ………….. Fb 1,400 1,250

Unit Stress in Tension Parallel to the Grain …….… Ft 850 850

Unit Stress in Horizontal Shear …………………….. Fv 110 110
Unit Stress in Compression Perpendicular to Grain .. Fcp 395 395
Unit Stress in Compression Parallel to Grain ………. Fc 825 825
Modulus of Elasticity …………………………………… E 1,200,000 1.200,000

TABLE 4 - Allowable Unit Stress (psi)

AREMA Section 2.7.1c states that “Where timber is treated by creosoting or other
process rendering it decay resistant, the working stresses for continuously dry may
be used except in compression perpendicular to the grain and for joists and planks
continuously submerged.” The exception accounts for stresses in the stringer where
it is bearing on the bridge cap. This area is noted for its ability to retain moisture
because it is protected from the sun and slow to dry.

At the end of Table 7-2-9 there are Notes that discuss specific conditions and
exceptions to be considered when using the Table. Note 7 provides for Shear Stress
Modification Factors to be applied to the allowable Fv, Shear Stress. Horizontal
shear, when dealing with a rectangular section such as a timber stringer, will affect
the stringer by causing a horizontal split that runs longitudinally down the center of
the member. If a stringer does not have a horizontal split, the tabulated Fv value can
be multiplied by a factor of 2. If there is a horizontal split that is up to 4-inches for an
8-inch deep stringer, the factor will be 1.67. With a conservative Fv used for Douglas
Fir and a conservative assumption of an existing split with a multiplier factor of 1.5
will produce an allowable Fv of 110-psi. This is consistent with an approach using
Section 2.10.14 that provides for a 10% increased allowable stress for a known
timber species. When the allowable Fv stress for Douglas Fir is 100-psi from Table
7-2-9, it can be increased to 110-psi which is still conservative when there is no
horizontal split.

Timber is grown naturally and does not have the same quality controls during its
growth as steel has during its development from a mill. AREMA recognizes this
important fact and discusses how this issue affects the application of the written
theoretical recommendations. Section 2.3.19 comments about the variability of the
Factor of Safety for timber trestles.
“There are many factors affecting the strength of lumber for which no
satisfactory, commercially applicable methods of evaluating the effects have
been found. These factors produce a variability among pieces which
otherwise seem to be alike. Since the allowable stresses of Table 7-2-9 are
based on the strength of the weakest pieces that may occur in the grade and
assume that each piece just carry its load, it follows that if a load is carried by
several members, not independent of each other, the designer could
reasonably allow somewhat higher stresses. Conversely, if the failure of a
single member would cause unusually great damage, the allowable stress on
 that member should be reduced. An overload of 50-percent will cause
failure in only rare cases, but if the load is doubled, failures will be
frequent.”

BRIDGE CONDITION

AREMA recognizes the importance of evaluating the physical condition of a timber

trestle and incorporating those conditions into the overall evaluation of the structure.
Section 2.10.4 reads:
“The carrying capacity of a bridge shall be determined by the computation of
stresses based on authentic records of the design, details, species and grade
of wood, materials, workmanship, and physical condition, including data
obtained by inspection.”

Regardless of how well and how strong a bridge may have been constructed, its
capacity may be significantly reduced if it is in poor physical condition. Two
independent types of deterioration define the physical condition of a timber trestle:
organic deterioration; and, physical deterioration.

Organic Decay:

Organic timber decay can be found around connections using bolts and drift pins.
These areas can be found at cross braces to piles; cross braces to caps; caps to
piles; stringers to piles; and, deck connections to stringers. Organic decay usually
starts internally and cannot be found unless an inspector “sounds” the wood at these
areas. When a dull “watermelon” type sound is heard, there is a good chance that
the timber has a decay void. In addition to points of connection, organic decay is
predominate around the ground line at piles, stringer to cap bearings, and across the
tops of stringers between the ties. All of these locations are subject to frequent wet-
to-dry conditions. If a void is not discovered through inspection, and the outer shell
thickness grows too thin, it is possible to experience a sudden failure of a cap or pile
under train traffic.
Mechanical Decay:

Mechanical decay is caused from long-term train loads. Examples of mechanical

decay can be found at points of bearing such as stringers to caps and caps to piles.
Crushing of the timber at bearing points, or individual pile settlement is evidence of
mechanical decay and may cause excessive movement in the structure due to the
pumping action. Shims found at the tops of piles or under stringers are good
indications of mechanical decay problems.

To require detailed input from a good physical bridge inspection is beyond the
intention of this Program. General questions will be presented that will require the
user to note a general condition of the structure. For organic decay, the user needs
to determine if there are any obvious bulges along the face of the cap or piles, or if
there are noticeable decay pockets along the top of the stringer between the ties. If
the owner has obtained a detailed inspection report that provides known boring data,
such as shell thickness and void depths, it would be wise to have that information
available while responding to the program questions. For mechanical decay, the
user will be asked to enter the depth and number of shims between the stringer and
cap, or pile and cap, or if there are any gaps and their size.

A theoretical analysis of a timber trestle will typically show that a cause for rejection
will be associated with the bending stress of the stringers, the horizontal shear, or
bearing stresses perpendicular to the grain. With the built-in factor of safety
previously mentioned, any organic decay issues within the stringers or caps that are
not obvious will be considered as acceptable. The Program will automatically
assume a void in the piles with a 3½ -inch outer shell during the evaluation of bearing
stresses. With a typical unbraced pile length of 25-feet, the allowable compressive
stress of the piles will not represent a cause for rejection under these conditions.

This program does not have the capability to determine if any of the bridge
components have serious organic or mechanical decay without the honest
input from the user. The program will ask several questions regarding the organic
and mechanical decay of the bridge. If the user indicates certain conditions exist,
such as multiple shim stacks, gaps at bearings that are missing shims, bulges in pile
caps or piles, etc., the program will identify these problems. The resulting condition
analysis will identify each exception, provide a list of the exceptions, and will place
the bridge in a category of either marginal or not acceptable. The reality of this type
of result is that these types of bridge conditions exist for a reason – there is a
problem. The good news is that these types of problems can be repaired. Unless a
bridge has multiple problems that would justify a new structure, the owner will have a
good maintenance list that, when the exceptions are corrected, the program would
likely move the bride to an acceptable structure to support 286k traffic. It is
important for the owner/user to understand that, even with a timber bridge that
is in good condition, once 286k traffic starts to operate across these
structures, the inspection and maintenance cost will increase in order to
maintain the integrity of timber bridges. If organic or mechanical exceptions
already exist, without being repaired before 286k traffic, there could be serious
problems or failures.

COMPLIANCE CRITERIA

Based on the results of the program’s technical analysis of the bridge components
and the response of the user’s answers regarding the bridge condition, the following
results will be generated by the program:

• Acceptable
• Marginal
• Not Capable

Based on computed analytical stresses of the individual bridge components, the

program will generate the following results based on the actual stress as compared to
the allowable stresses:
 • Not Capable: Any component of the bridge with a stress greater than
15% of the allowable stress. Each exception will be identified and
listed. i.e. Stringer 3 of Span 5 fails in bending (or compression, or
horizontal shear)

• Marginal: Any component of the bridge that is equal to, or up 15%

greater than the allowable stress. Note: If the actual stress is within
this identified range, AND the train speed is 30mph or greater, the
program will identify the condition as “Not Capable.” Each of these
conditions will be identified and listed.

• Acceptable: Any component of the bridge that is less than the

allowable stress. No need to identify or list these conditions.

Based on the user’s responses to “condition” questions. The following will justify a
“Not Capable” or “Marginal” response and will be identified as follows:

• Non-compliant FRA track across the bridge - Not Capable

• Rail size of 115# or greater is Acceptable: 90# or Less is Not Capable;

and rail between 90# and 115# is Marginal

• Ties are not anchored to tie spacers or stringers - Not Capable

• Any deck condition noted as a 4 to be Marginal, or 5 to be Not

Capable

• Stringers with less than 3” of end bearing on a cap - Not Capable

• If there are two secondary caps (making 3 caps: primary and two
secondary) - Not Capable

• Shims: More than 3-inches high and 3 or more separate shims at the
stringers/caps - Not Capable

• If track shims are greater than ½” - Not Capable

• Void (gap) greater than ¼-inch between the cap and any pile - Not
Capable

• Piles with an outside shell thickness less than 1.5-inches - .Not

Capable

• Any Bridge with a 3-pile bent - Not Capable

• Any rating condition entered as a 4 will be listed as Marginal

• Any rating condition entered as a 5 will be listed as Not Capable

• Erosion or Scour condition as “Progressive”, will be listed as Marginal

• Erosion or Scour condition as “Serious” will be listed as Not Capable

• If the stringers are noted as being framed with square notches, a rating
of Marginal will be issued; however, if the horizontal shear stress of the
stringers are within 15% of allowable, this condition will cause a Not
Capable condition.

• Any Bent with two adjacent piles that have been posted, or if there are
three or more posted piles, it will be rated as Not Capable.
EXCEPTION

Any bridge with a bent height (top of tie to ground line) greater than 30-feet to
require detailed inspection by experienced inspector to verify condition and
compliance.