LEHIGH UNIVERSITY
High Cycl~ Fatigue ofWelded Bridge· Details
FRrrZE'~JGfNEERIN('·
l/u30RATORy'LIBRARY
FATIGUE DAMAGEIN THE LEHIGH CANAL BRIDGE
by
John W. Fisher
Ben T. Yen
N'icholas V. Marchica
frits Engi~eering Labol'atoryReport No. 386.1
1. Report No. 2. Government Accession No.
TECHNICAL REPORT STA~DARD TITLE PAGf:
3. Recipient' $ Catalog No.,
4. Title and Subtitle
FATIGUE DAMAGE IN THE LEHIGH CANAL BRIDGE
5. Report Date
November 19746. Performing Organization Code
7. Author( 5)John W. FisherBen T. Yen, Nicholas V. Marchica
8. Performing Orgoni :ration Report No.
Fritz Laboratory No. 386.1
9. Performing Organization Name and Address 10. Work Unit No.
Fritz Engineering LaboratoryLehigh UniversityBethlehem, Pennsylvania 18015
11. Contract or Grant No.
PennDOT Research Project 72-313. Type of Report and Period Covered
~------------------------------......12. Sponsoring Agency Nome and Address
Pennsylvania DepartmentOf Transportation
P.O. Box 2926Harrisbur2. Pa. 17120
15. Supplementary Notes
Interim ReportU.S. Department of July 1 1972 to June 30 1974
Transportation ' ,Federal Highway AdministratibW, Sponsor,ing Agency Code
Washin2ton. D.C. 20590
o
Prepared in cooperation with the United States Department of Transportation,Federal Highway Administration. From the study, "High Cycle Fatigue ofWelded Brid2e Details".
16. Abstract
Strains were measured at several structural steel details of one of theLehigh Canal Bridges under normal traffic during a pilot study. Since thesebridges have several fatigue cracks in the tie plates connecting the floorbeams to the outrigger cantilever brackets, the primary focus was on the tieplates and the cause of fatigue cracking. Strain gages were mounted on fivetie plates, on a stringer, and on the longitudinal girders. An automaticcomputer-controlled data acquisition system was used to record the strain rangeoccurrences. In addition, an .analog trace recorder was used to determine thelive-load strain variations with time. Stress ranges in the stringer andgirders were comparable to measurements observed by others in girder bridges.However, the ranges of the horizontal in-plane bending stresses in the tieplates were found to be two to three times as high. The cause of the higherstress range was attributed to differential longitudinal displacements betweenthe deck-stringer system and the girders transmitted through the tie plates.The strain measurements on the tie plates and esti~ted truck traffic duringthe structure's life were used to estimate the cumulative damage in severaltie plates. Good correlation was obtained between the root-mean-square stressrange and constant cycle laboratory fatigue test results. Miners rule was alsofound to provide a good correlation.
17. Key Words 18. Distribution Statement
bridges (structures); crack growth;fabrication; fatigue; joints;structural engineering; structuralmembers; stress history; secondarystresses
19. Securi ty Classi f. (of thi 5 report) 20. Security Classif. (of thi s page) 21. No. of Pages 22. Price
,Unclassified
Form DOT F 1700.7 (8-69)
Unclassified 52
----.-...-.-..-- -- _._..__._-_...._._-_.._ .._----------------,-------------------1
COMMONWEALTH OF PENNSYLVANIA
Department of Transportation
Bureau of Materials, Testing and Research
Leo D. Sandvig - DirectorWade L. Gramling - Research Engineer
Kenneth L. Heilman - Research Coordinator
Project 72-3: High Cycle Fatigue of Welded Bridge Details
FATIGUE DAMAGE
IN THE LEHIGH CANAL BRIDGE
by
John w. FisherBen T. Yen
Nicholas V. Marchica
Prepared in cooperation with the Pennsylvania Department ofTransportation and the U. S. Department of Transpo'rtation,Federal Highway Administration.
The contents of this report reflect the views of the authorswho are responsible for the facts and the accuracy of the datapresented herein. The contents do not necessarily reflect theofficial views or policies of the Pennsylvania Department of'Transportation or the Federal Highway Administration. This report does not constitute a standard,specification,or regulation.
LEHIGH UNIVERSITY
Office of Research
Bethlehem, Pennsylvania
November 1974
Fritz Engineering Laboratory Report No. 386.1
Figure
1
2
3a
3b
4
5a
5b
6
7
8
9
10
11
12
13
14
15
16
17
18
19
LIST OF FIGURES
Lehigh Canal Bridge - Instrumentation Installed inLeft Side Span
Plan and Elevation of the Test Span
C~oss-section of Bridge
Tie Plate Detail at Floor Beam Bracket Connection
Schematic of Cracked Tie Plates in the Test Span
Crack in Tie Plate Originating at Tack Weld
Schematic of Instrumentation on Tie Plates and Girder
Typica1 Strain History Response in Tie Plates,Stringer and Girder
Comparative Response of Tie Plate C4s and South Girderfor Various Loads
Instantaneous Stress Distribution in Tie Plate C6N
Instantaneous Stress Distribution in Tie Plate C6S
Measured Stress Distributions in the Gaged Tie Platesduring Passage of Truck
Strain History with Truck in the Passing Lane
Comparison of Measured Strain History in Tie Plateswith Influence Lines for Girder Slope
Stress Histogram for Tie Plate C4s
Stress Histogram for Tie Plate C6N
Stress Histogram for North Girder
1972 -PennDOT Loadometer Survey Gross Vehicle WeightDistribution
1970 FHWA Nationwide Gross Vehicle Weight Distribution
Estimated ADTT at the Lehigh Canal Bridge Site
Comparison of Estimated Equivalent Stress Range withLaboratory Test Data
ii
Page
20
21
22
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
1. INTRODUCTION
1.1 Background and Objectives
Fatigue cracks have recently been detected on several steel
highway bridges in the United States. These cracks have occurred at
bridge details such as the en~s of coverplates, at web or flange attach
ments, and in tie plates connecting transverse floor beams and brackets.
Among bridges which have sustained some of these cracks are the Yellow
Mill ·Pond Bridge on the Connecticut Turnpike 1, the Lehigh River and
Lehigh Canal Bridges on U.S. Route 22 in Pennsylvania (reported herein),
and the Allegheny River Bridge on the PennsylvaniaiTurnpike2• All of
these bridges carried large volumes of truck traffic. The occurrence of
cracks in the tie plates of the Allegheny River Bridge and the Lehigh
River and Lehigh Canal Bridges provided a unique opportunity to observe
the stress history of details which were actually experiencing fatigue
crack- growth.
Recent laboratory studies on the fatigue strength of beams
with coverplates or welded attachments indicate that stress range (live
load and impact stress) controls the fatigue behavior of structural
details s ,4. No "run-out" or fatigue endurance limit was observed when
testing these beams up to 6,000,000 cycles of load application at 6 ksi
stress range.
For the purpose of furth~r exploring the fatigue behavior of
steel bridge details under relatively low stress ranges and high numbers
of cyclic loading, and to correlate the stress range history of bridge
-1-
102 Description of Bridges
The Lehigh Canal Bridges consist of twin bridges which carry
the eastbound and the westbound lanes of U~So Route 22 near Bethlehem,
Pennsylvaniae Each bridge is continuous for three spans with small
haunches at the interior piers as illustrated in Figse 1 and 20 Each
bridge has two riveted steel longitudinal girders with a floor beam
stringer system and a non-composite concrete decko An end span of the
eastbound bridge, Fig, 1, was chosen for testing because of its accessi
bilityo The plan and elevation of the 144 foot end span are shown in
Fig, 20
In most of the test span, the longitudinal girders are 8'-1/2"
deepo Nine floor beams (W27 x 94) and outrigger brackets, two exterior
stringers (W~4 x 34 and W16 x 40), and four interior stringers (two
\W16 x 40 and two 16 ~ 45) constitute the load carrying steel system~ A
typical cross section of the two-lane bridge is shown in Fig. 3ao The
details of tie plates which connect the brackets to the floor beams are
shown in Fig, 3bo The bridges were constructed :in 1951-53 and opened to
traffic in November 1953~
~2-
Inspections by Pennsylvania Department of Transportation per
sonnel had revealed several cracks in the tie plates in the spring of
1972, prior to in~service testingo The approximate location and length
of these cracked tie plates in the test span are summarized in Figo 4.
The cracks detected throughout the length of the Lehigh Canal Bridges and
the Lehigh River Bridges are summarized in Appendix AQ Most of the
cracke~ tie plates were at or near the piers and abutments 0 All cracks
were along the outside edge of the longitudinal girderso Several of the
plates had cracked across their entire width, some had only fine hairline
cracks. All observed cracks appeared to be through the thickness of the
tie plates as illustrated in Fig, 5a. The cracks did not pass through
the rivet holes. All cracks st'arted at the edge of the tie plates from
a tack weld toe. The tack welds were used to connect the plates to the
bracket prior to riveting.
-3-
2. STRAIN'DATA ACQUISITION
2.1 Strain Gages and Recording Syste~
In order to monitor stresses, electrical resistance strain
gages were mounted on the tie plates, on the main longitudinal girders,
and on a stringer beam. Most of the gages were mounted on the tie plateso
The tie plate gages were orientated parallel to the edges of the tie
plates and were generally above the cut-off point of the top flange of
the cantilever brackets, Fig. 5b.
The approximate location of the strain gages is indicated
schematic~11y in Fig. Sb. Gages were placed on tie plates, CIN, C4S,
C5S, C6S, and C6N. The gages on the girders were located on the top
and bottom flanges between floor beams 4,5,6 and 7. A stringer was
also instrumented between floor beams 5 and 6.
All gages were 1/4 in. long electrical resistance foil gageso
Weatherproof coatings were applied to protect the gages from moisture
and the environment. The gages were connected to temperature cODpensa
tion gages and plates, so as to minimize the effect of t~mperature changes 0
Strain variations due to traffic were recorded using two
separate systems: the FHWA automatic data acquisition system and an ana
log trace recorder. The FHWA system, which is installed in a van, con
sists of an amplifier, an analog-to-digi~al signal converter, and a
computer, The computer is connected to a teletype machine for input
outputSo In this study the strain histories at the gage points were
-4~
sought. As each vehicle traveled over the bridge, the magnitudes of
strain ranges were recorded. After a period of tine the number of occur
rences when the strain range was between predetermined values (levels)
wete printed. To exclude recording large numbers of very low strains due
to vibration and automobile traffic, a test level was defined for each
gage. The strain range had to exceed this level (between 15 to 50 micro
inches per inch) to be recorded. For accurate recording of strain ranges,
the zero-level of strains was checked periodically to prevent drifting.
Variation of strains due to traffic were also recorded by an
analog trace recorder. A typical recorded analog trace is shown in
Fig. 6 in which the live-load strain magnitudes of six gages were re
corded as a function of time. The analog recorder utilized the ampli~
fiers of the FHWA system, thus strain variations of several gages could
be monitored by the two systems simultaneously.
2.2 Traffic Identification and Recording Periods
In order to correlate traffic on the bridge with recorded
strains in the tie plates and girders, truck traffic counts were made$
This was done visually by observing the traffic flow and recording the
number and types of trucks for short periods of time. These observations
were also made while the analog traces were being taken so that strain
magnitudes could be correlated with. var'ious types of trucks.
Strain data were compiled during two intervals of time. Alto
gether 170 hours of data. were acquired for statistical evaluations and
stress analysis.
-5-
3. STRESSES IN VARIOUS COMPONENTS
From observing the analog traces of strain variations and the
traffic on the bridge, it was evident that eac4 truck crossing the bridge
caused one, stress range excursiC?n at all gage points on the bridge. This
was true of the tie plates and the girders as illustrated in Fig., 7. The
strain magnitudes in most of the tie plates were much higher than those
observed in the stringer and girders.
The recorded live load s tress distributions in typ.ical tie
plates are shown in Figs. 8 and 9 for a given tiEe. The stress distribu
tions indicate that all tie plates were subjected to bending in the hori
zontal plane. The gages at the centerline of the t~e plates had low
stress magnitudes, implying that there was little axial elongation or
vertical bending of the plates due to truck loading.
The stress distributions in the tie plates are superimposed on
the plan-view sketch of these plates in Fig. 10. From the stress data
for these plates, it was observed that the maximum live load horizontal
bending stresses were higher in the tie plates near the abutment and
towards the. pier., This distribution agrees with the. crack pattern
observed in the tie plates. Some cracks occurred at locations where the
live load stresses were primarily compressive in nature. This reflects
the dead load and residual tensile stresses which made the stress range
fully effective in crack growth3 ,4.
-6-
In order to examine, the r'elative stress magnitudes at' different
tie plates when a truck was crossing the bridge, one gage on each of the
tie plates was connected to the r~co~ding instruments simultaneously 0
Examples _of strain variation with time as recorded on the analog trace are
shown in Fig. 11 which show~ the response when a five-axle tractor semi
trailer was travelin.g in the p~ssing lane. As the vehicle crossed the
bridge, tie plate? in the slow lane (gages 5, 7 and 11) registered lower
strains.than tie,p~ates in the passing lane ,(gages land 10)0 Gage 20
was on the t~p flange of the slow lane girder. The small fluctuations of
strain ~or each gage were due to vibration of the bridge, under traffic.
The strain :'responses shown in Figs 0 6 and 11 confirm that
strains in the tie plates were dependent on the location of the tie
plates in the bridge span, and that maximum stresses were higher in tie
plates near the abutment and towards the piero
Live load stresses in the longitudinal girders and a stringer
of the bridge were low compared to those observed in the tie plates. The
magnitude of maximum live load stresses in the girders were in the order
of 8 and 9 ksi. These values are comparable to strains recorded by other
investigators in main longitudinal members 6,7,e,9.
Examples of the time variation of stresses due to a truck were
shown in Figs. 6, 7 and 11. The total stresses were the superposition of
the static response and the vibrational stresses. The static stress
variation for a point on a girder flange is analogous to the stress influ
ence line for that point. The stress-time tracing for gage 19 in Fig. 6
, -7-
is obviously similar to the stress influence line for a point in the left
span of a three-span continuous beam.
The strain variation traces shown in Fig. 7 indicate that, the·
pattern of stress excursion at a tie plate on the bridge was the same for
all types of trucks, only the magnitude of the strain range changed.
This suggested that the strain variation in the tie plates was directly
related to the strain' occurring in the longitudinal girders. Preliminary
analysis indicates that longitudinal displacement at the top flange of'
the girder would cause horizontal bending in the tie plates. Since the
longitudinal displacement of a point on the top flange of the girder is
the product of the slope of the deflection curve and the distance from
the neutral axis to that point, the strain influence line for the tie
plates would be analogous to the influence lines for slope in the longi~
tudinal girder at the tie plate. Figure 12 compares the strain variation
at gages. 5 and 10 on tie plates c4S and C6N with the influence lines fo'r
the slopes of the girder at the corresponding points. It is read~ly
apparent that these curves are compatible with each other. 'Detailed
evaluations of this behavior are under way.
-8-
4. STRESS RANGE UNDER RANDOM TRAFFIC
The data acquisition system was programmed to count the number
of strain range occurrences. More trucks traveled in the slow lane than
in the passing lane. Hence, the strain range occurrences for tie plates
under the slow lane were higher than those for the tie plates under the
passing lane.
The detailed strain range occurrence data are summarized in
Tables Bl and B2 of Appendix B. Table Bl shows the stress range levels
and frequency of occurrence of the stress excursions for the tie plates.
Similar data is given in Table B2 for the girders and stringers. The
total recording time and number of occurrences is also indicated.
The stress range occurrence data were plotted as histograms,
dep.icting the percentage of frequence of occurrence between the stress
range levels. Examples are shown in Figso 13 to 15 for gages on tie
plates C4S and C6N and on the north girder$ Although the recording peri
ods were not very long, some general observations could still be made$
The gage on tie plate c4S was subjected to stress reversal and the maxi
mum stress magnitudes were smaller than observations on other tie plates
(se-e Figs. 11 and 12). A large percentage of occurrences were observed
at lower stress range levels. Tie plate C6N was subjected to maximum
stre~s magnitudes that were higher and more frequent at higher stress
range levels. There was also a tendency for two peaks on the histograms.
The gages on the main longitudinal girders indicated much lower stress
-9-
range levels. From the analog traces and traffic monitoring, it was
observed that only larger trucks induced stresses above one ksi in the
girders. The single-peaked histograms reflect this response.
-10-
5. TRAFFIC RECORDS
During' the in-service te:sting of the Lehigh' C"anal Bridge", traf
fic counts were taken on the bridge on three consecutive weekdays. All
counts were carried out in the afternoon. The highest volume of trucks
consisted of five-axle tractor semi-trailers. Because the sample time
WqS -only limited_ ~n length the results were compared with existing
PennDOT traffic surveys.
Twenty-four hour traffic counts taken during 1972-73 at a site
near the Lehigh Canal Bridge were obtained from PennDOT. A comparison
showed that the twenty-four hour traffic surveys resulted in more two~
axle trucks and five-axle tractor semi-trailers than were observed at the
bridge site during the short sample periods. Overall, the "percentages of
various types of trucks from the twenty-four hour, reGords were consistent
with the observations a~ the bridge site. The PennDOT records were also
consistent with results from loadometer surveys throughout the state and
the nation.
"The results of a 1972 PennDOT loadome'ter survey for twenty
stations on main arteries in Pennsylvania are summarized in Fig. 16.
The frequencies of 'occurrences of trucks by weight are plotted in the
. figure. There are two peaks in the histogram, indicating large numbers
of loaded trucks at 60 to 75 kips, and large numbers of two-axle or
unloaded tractor semi-t,railers at 24 to 36 kips.
-11-
Figure 17 summarizes the results o~ a 1970 FHWA Nationwide
Loadometer survey. The relative distribution of truck frequencies from
this survey are comparable to the distribution shown ,in Fig. 16, There
are two peaks at approximately 25 kips and 70 kips, similar to the
twenty station survey from Pennsylvania.
Since the 'results of traffic counts at the bridge were consis
tent with the results of the twenty station main artery survey, which in
tum were comparable to results from other surveys, the resutts of' ·traf
fie counts at the bridge were considered representative of, the normal
traffic crossing the, Lehigh Canal Bridge.
The total number of trucks that have traveled over the bridge
during the nineteen year period (1953-1972) was estimated from the
PennDOT traffic count data. 'rhees·tirnated average daily- truck traffic,
counts (ADTT) from the PennDOT data are plotted in Fig. 18 on fl.- yearly,
basis. This ranged from 2038 trucks per day in 1953 to '3709 in 1972.
The average rate of increase was about 1.5% prior to 1960 and 4.5%' after
1960 when main arteries leading to the bridge were open to traffic. The
total volume from 1953 to 1972 was estimated to be 18.9 million trucks.
1 ..,
-12-
6. CORRELATION WITH FATIGUE TEST RESULTS
6.1 Fatigue Test Data
'Results from beam tests and girder tests in the laboratory
have indicated that stress range and the length of welded attachments
are controlling factors for fatigue strengtho If all other conditions
are the same, an increased attachment length results in a shorter, fatigue
li£e 4• Results of the laboratory studies on beams with short attachments
are plotted as a function of stress range and cycle life in Figo 190 The
dotted lines represent the 95% confidence limits for the test specimens.
The specimens used for determining the stress range-cycle life relation
ship were all subjected to constant stress range cyclesQ These test data
correspond to Category·n in the 1974 AASHTO Specifications10 and were
used to evaluate the tie plate performance because several pilot bending
tests of tie plates provided comparable test results 0
The riveted tie plates of the Lehigh Canal Bridge were sub~
jected to horizontal in-plane bending under random applied loads. Most
of the tie plates have tack welds one to two inche~ long connecting the
edges of the tie plates to the bracket top flange as was shown in Fig. 5ae
Load transfer between the flange of the outrigger bracket and the tie
plate was provided by both the rivets and the tack weldsQ Since the
tack welds were at the end of the joint which is the most highly stressed
region 11, the stress concentration at the end of the tack .weld appears to
be as severe as the condition at short attachments to beam flanges 4 as
-13~
was confirmed by the pilot studieso The initial discontinuity at the
tack weld toe was probably greater than discontinuities that commonly
exist at weld terminations and may be the reason for this low fatigue
strengtho When comparing laboratory fatigue test results with the mea-
sured stresses at the bridge, adjustments must be made to the ,measured
stresses because cracks occurred at the edges of the tie plates where
stresses were higher~ Furthermore, the measured stresses were not of
constant magnitude.
6.2 Root-Mean-Square Estimates
One procedure to account for the stress range sp~ctrum is. by
the root-mean-square (RMS) method 3 ,12,13. In this method the root-mean·-
square of the stress ranges in a spectrum is considered equivalent to a
constant cyclic stress range and is correlated with the number of stress
cycles corresponding to the spectrum stress ranges. The root-mean-square
stress range is defined as
2 1/2(L a. S )
1 r.~
where a. is the frequency of occurrence of stress range S1 r.
1
The stress range occurrences of the gages on the tie plates
were used to determine the RMS values of stress ranges. These values
were then adjusted by the distance from the gages to the plate edges
where the crack growth originated.
-14-
(1)
The SrRMS at the plate edge and the corresponding fatigue
cycles corresponding to the truck traffic are plotted in Fig. 19 and are
compared with the constant stress ,range laboratory data. The data from
the Lehigh Canal Bridge is seen to lie above or near the lower confidence
limit for Category D. Cracks were observed in tie plates CIN, C5S and
C6S. The cracked 'plates are seen to plot on or above the lower confi
dence limit in Fig. 19. Tie plates C4S and C6N had no visible cracks at
the time of measurement. The data points for these two plates are below
the lower confidence limit.
The comparison given in Fig. 19 indicate's that the RMS stress
range provides a reasonable correlation between the field data and the
laboratory test results. The lower confidence limit for Category D of
AASHTO provides a reasonable lower bound to the fatigue strength of the
tack welded tie plates of the Lehigh Canal Bridge. Tie plates that had
SrRMS and cycle lives above this limit could be expected to exhibit
visible fatigue cracking. Tie plates c4S and C6N can be anticipated to
develop fatigue cracks under additional truck traffic.
6.3 Correlation by Miner's Rule
The stresses at gages on tie plates CIN, C4S, C5S, C6N and C6S
were adjusted to give the stress range at the edge of the tie plates due
to truck traffic. The resulting distribution of stress range at the tie
plate edges was used with Miner's cumulative damage hypothesis 14 to deter
mine the total damage caused by the truck traffic.
-15-
Tie plates ClN, C5S and C6S had cracks; the cumulative damage
ratio based on Miner's Hypothesis was close to or higher than 1.0. No
visible cracks were detected at the tie plates C4S and C6N, which had
damage ratios less than unity.
By combining the relationship provided by constant cycle data 3
and Miner's rule 14 , an equivalent stress range S M- 15 was estimated as:r lner
S M.r lner
1/3) (2)
where a. is the frequency of occurrence of stress range S1 r
i
The S M. values of the tie plates are also plotted inr lner
Fig. 19. The results also demonstrate that Miner's Rule provides a rea-
sanable correlation between the observed tie plate behavior and the
laboratory results.
The results in Fig. 19 indicate that the Miner's Hypothesis can
be used to correlate field measurements and laboratory test results in
terms of stress range and cycle life. The lower confidence limit for
Category D of AASHTO Interim Specification 1974 provides a reasonable
estimate of the fatigue strength of the tie plates of the Lehigh Canal
Bridge.
-16-
7. SUMMARY AND CONCLUSIONS
From the pilot field studies and the a11alyses of results of the
Lehigh Canal Bridges, the following conclusions- can be reached:
1. The live load stresses in the main longitudinal girders were
similar to those observed in longitudinal members of other
bridges. Seldom did the live load stresses exceed 8 ksi.
2. The live load stresses in the tie plates connecting the outrig
ger brackets and floor beams were much higher than those observed
in the longitudinal members. The distribution of these stresses
in the plates indicate that horizontal in-plane bending was
occurring with little axial extension or twisting of the plates.
3. The development of live load stress in the tie plates was related
to displacements caused by longitudinal strain of the main girder
under bending. The relative longitudinal movement of the top
flange of the main girder with respect to the stringers, brackets
and concrete slab caused a horizontal bending of the tie plates.
4. The crack pattern in the tie plates agreed with the measured
strain ranges. In addition it appeared to be compatible wi-th
the predicted stresses due to displacement.
5. The measured stress history for short time periods was com
patible with the frequency distribution of trucks reported by
PennDOT on comparable roads.
-17-
6. The root-mean-square stress range, SrRMS' and an estimated traf
fic volume for the period of time from 1953 to 1972 provided
good correlation between crack occurrence on several tie plates
and the fatigue test results from laboratory studies.
7. Miner's Hypothesis also provided a reasonable estimate of the
damage that had occurred in the tie plates.
8. No stress measurements were obtained on completely cracked tie
plates. However, the crack pattern was compatible with the
expected stress distribution due to the in-plane displacements.
A more detailed analysis of the stress development mechanism in
the tie plates of outrigger bracket connections is being developed. This
should permit a better evaluation of the cause of the high stresses and
their possible occurrence in other structures. It is apparent that care
should be taken to minimize displacement induced moment and stresses such
as those occurring in the tie plates. Very short distances between the
first stringer on a bracket and a longitudinal girder results in high
displacement induced moments in th'e tie plates. These moments are reduced
when this distance is increasede
On the Lehigh Canal Bridge, more comprehensive studies are being
undertaken in order to develop more complete knowledge of the stress
range-life relationship as the bridge provides a unique opportunity to
relate laboratory studies to the random loading and corresponding bridge
responseo
-18-
8. ACKNOWLEDGMENTS
This investigation ,~s a part of a research project being con
ducted at Fritz Engineering Laboratory, Lehigh University, Bethlehem,
Pennsylvania. The project is sponsored by the Pennsylvania Department
of Transportation and the U. S. Department of Transportation, Federal
Highway Administration.
Thanks are due Mr. Hugh T. Sutherland for his assistance in
acquiring the data, Dr. J. Hartley Daniels for his help on the analysis
and the project staff for their participation. The assistance of the
Fritz Engineering Laboratory staff is acknowledged, and the~assistance
of PennDOT and FHWA personnel is sincerely appreciated.
-19-
1
Fig. 1 Lehigh Canal Bridge
(Instrumentation Installed in Left Side Span)
FloorBeam
Long.Girders
31-3 11
T~""'--~---'--"""'-~~----'-"""""""'*5@5'-SU" 211-01l cc=27' -6
11Girdersl --l
I8 Panels @ lSi = 1441-0" I! II~ _ 3-3
PLAN VIEW(Deck Not Shown)
50'-01'
---,~~~~~~~~~~~~l
1"81-0~2n
Abutment C
Pier IA
ELEVATION VIEW
Fig. 2 Plan and Elevation of the Test Span
-21-
I III3-94
W27x94
21~O"C/C Girders
Fig. 3a Cross-Section of Bridge
5'-9'1
Tie ~
FloorBeam
21
-3" -1-
LongitudinalGird·er
Flange ofBracket
-I.SECTION A-A
Fig. 3b Tie Plate Detail at Floor Beam BracketConnection to Girder
-22-
Girder EN<l
• • • • • • "ell..
Girder ES<l
I:~:: '·1·': :~:II
iA·· · f±-.:'· "-~· · .. · · .... IA 1· · · ·+.~
I I
C7 I (I' ···'1 I" I ·f··· '1• •• " ',,' ••• ill, , • • •• ',. • ••• ,
I
C6 · . .. . . " ... !... I ·,. · · .,, . . .. . ,.' .... ,
C5• • • •• • 0 • • •
• • !'. II • •••• • " " " * ••••
• • •• !' 0) ••••
1~::····I··r~ ~~II
. . .. .. ."..
C3
C4 I···· '1 '1 t· · .. ',,' " Bound
I
, Lanes
i
0 • • •• • •••• i. . .. " . ....
"""""""----=--4------"
C2
CI " . .. . . .... I~ ~ ~ ~ ··1· 'r ~ ~ ~ II
C ---I-- ~O. ·" .-E::~~Cl" '1~'''''' ~• • 0 • • • .. Q • • ••••
'------4~----J
ABUTMENT
Fig. 4 Schematic of Cracked Tie Plates in the Test Span
-23-
Fig. Sa Crack in Tie Plate Originating at Tack Weld
-24-
s
1-4
FloorBeam C CI
Gage Nos.
Abu t. C
5,6,29
-; 9,10,
27."/2I t 22/ I Girder
/ N
/C7 IA
Pier IA
Fig. Sb Schematic of Instrumentation on Tie Plates and Girder
-25-
I/--.../'---..-......-- '____ "', ~
!I
I~~
~/ II ;
f1OO1l1n.1tIt -....
l===i:==~==t===+~ttI=~~=t=~GlRDERESABUT. CI C2 C3 C4 C5 C6 C7 IA
Fig. 6 Typical Strain History Response in Tie Plates,
Stringer and Girder '
-26-
TIE f.o
2 AXLEGIR_D_ER_.~~~_~~~~
TIE~ ~"- 3 AXLE
GIR__....A.ft...-..._~~~""........",~~~L=.-__~ · _---:::::::::;..,,--~~.-~ _
4 AXLE
5 AXLE TRACTORSEMI- TRAILER
TIE--....,-..---""'-"r-----~
GIR.-c::3lo__.........-~:::.......L_- --=~ ~.",.e..- _
GIR
4 AXLE TRACTORSEMI- TRAILER
GIR__.........~~_~-.'---. ~ ---:~o&--- ~
TIE
Fig. 7 Comparative Response of Tie Plate c4s and South Girder for Various- Loads
-27-
1011\~r---------.........
Direction ofEast Bound
~
Traffic
o 10.8 ksi
F~g. 8 Instantaneous Stress Distribution in Tie Plate C6N
~28-
10"
I
Traffic
7/1---------Direction ofEast Bound..
C6S
1/2'
Fig. 9 Instantaneous Stress Distribution in Tie Plate C6S
-29-
GIRDER ENt
GIRDER ESt
-----....----1.
IA
C7
C6
C5
C4
C3
C2
CI
c
I I II
.....-.-.---+-______ ~:r t_-_~ . ~
I I ( II
I ~ II
( 13; II
East (-41
-+ I----t-----..I Bound--H
Lanes I
~~ I : II
I P II I J I
-+---- --+---+-- ---....- -E·=$£.=tABUTMENT
Fig. 10 Measured Stress Distributions in the Gaged Tie PlatesDuring Passage of Truck
-30-
tIOO~ In/In.·! kll
/1\
1=~=:t==i=t::t=~==:::::t::=J;:::::.t=:::::tGIRDEREN
~=+=:;::::I==+===~~;:t::::::t:=t=~GIRDER ESABUT. CI C2 C3 C4 C5 C6 C7 IA
Fig. 11 Strain History with Truck in the Passing Lane
-31-
Gage C4S-5
~~-~-~oo 6 sec.
3 e I ~:~~:t i~(Approximately 5.3 Sec. at 60 mph.)
Gage C6N-1OStrain
o 6 sec
Slope
,.... 4681
Fig. 12 Comparison of Measured Strain History in Tie PlatesWith Influence Lines for Girder Slope
-32-
18
[7
16
15
14
13
12
II
100/0
9
8
7
6
5
4
3
2
Gage C4S-559.67 % 1.5- 3.3 ksi40.33 % 3.3->9,6 ksi
III I
o 3.3 4.2 5.1 6.0 6.9 7.8 8.7 9.6 >9.6
ur (ksi)
Fig.!3 Stress Histogram for Tie Plate C4S
-33-
II
10
9
8
7
6
5
4
3
2
Gage C6N - 9
78.65 % .9-3.0 ksi
21.35 % 3.0- >13.5 ksi
f
•o 3.0 4.5 6.0 7.5 9.0 10.5 ,12.0 13.5 >13.5
o-r (ksi)
Fig. 14 Stress Histogram for Tie Plate C6N
-34-
18
17
16
15
14
13
12
II
10
0/0 9
8
7
6
5
4
3
2
I
GAGEGEN-1534.37°/0: 1.5-1.95 KSI65.63°/0: 1.95-> 2.85 KSI
1.95 2.10 2.25 2.40 2.55 2.70 2.85 >2.85
CFr (KSI)
Ffg~ 15 ,Stress Histogram for North Girder
-35..
10.0
9.0
8.0
7.0
~
>0 6.0Zl.LI:JoIJJ 5.0a::LL.
4.0
3.0
2.0
1.0
I-
~
~
~
~
""--
.......I-
~
~
r-- ~
I-
- -~
-
..
~..
~
..
- ~
rh--~ - -o 202428323640 45 50 55 60 65 70 75 80 85 90 95 100 105 110>110
GROSS VEHICLE WEIGHTI Kips
F,ig. 16 1972 PennDOT Loadorn.eter Survey Gross Vehicle Weight Distribution·
-36-
~
,....
-"""" -
,...~
10-I""-"
,....,
""""",....
,.. ............
~,..-1--",,- ....,.011
~ ... -""'""'"-
"'-
...~, .
3.0
25 35 45 55 65 75 85 95o 20 30 40 50 60 70 80 90 100
GROSS VEHICLE WEIGHT,Kips
1.0
4.0
~
>-oZlLI:::>02.0~a::LL
Fig. 17 1970 FHWA Nationwide Gross Vehicle Weight Distribution
-37-
•wco•
5
4
3
ADTT(in 1000)
2
(4.5%)~..,.!"'" •
(1.5%) ............ - .......
~ ........ --._-------~-A- •
1953 55 60YEAR
65 70 72
Fig. 18 Estimated ADTT at the Lehigh Canal Bridge Site
20.950/0 Confidence
~ Limits.......................
15t- ~ '-
.............'-...t:. '_'_.............
............
............
10J- L'- '-..........................Category 0 '-,-
f ~5S''''-
'-AASHTO
'-......... • -C6S" ",CIN~ '--...-C4S
Sr5
o -C6N
(k 5 i)
RMS Miner
I
48 Art..
w
Cracked
\.0I
6.No Crack 0
2 3 5 10
N (x 106 cycles)
20 30
Fig. 19 Comparison of Estimated Equivalent Stress RangeWith Laboratory Test Data
APPENDIX A
<lGIRDER ES
(I)lIJ
'"'"---............- ...... Z """"---............--"'"""ct..J
...---...........--...... C .------+---.......Z~
.-.._~_........ o~- .............-_..mt-
~--.I-=--CI) ,..,.,.....-........,..-.-...........,..~
«w
( rI
~lllj~Cl Cl
GIRDER WS GIRDER EN<l
GIRDER WN
Pier.~
vzf(J)
_. -Cl of Span - . ---- .-----
Fig. A.l Cracks in Tie Plates, Lehigh River Bridges (PennDOT)
-40-
I j I
enLLI
",,-_-+__IIIIIIQ Z .........- ..............--10I8III«..J
~---e--_... 0 '-----t-.....- ...z::)
...-_....... 0 .--_-+-__......CDt-........_--+0__'""'" en ...........-A:r.._...
«lJJ
enlJJ
,.----+--.... z ......- ........--..<t..J.....---+--_... a---.......--....z=t
,...----+--.... 0 .....- .......--...m....
~---I--_ .... tJ) ----.....--...\.LI~
Pier-&-
1.0Z
~(J)
___P_ie.r_
Fig. A.l Cracks in Tie Plates, Lehigh River Bridges (PennDOT) M continued-41-
bt
GIRDER ES
----.--
tGIRDER EN
U)
.........iIIIIIIlIlIiiP'~_~w~_ ...........~...............~.-J
~~=+=.:~~ ~-~~.............az
~~~_~:J~_...............~.--.oII)
~-"""""''''''''''''''''~~~-~'''''''''"''''''''''''''''en<
~-~~~W~~~~............
(Continued Next Page)...---t--.............
~I~l Iit t
GIRDER WN GIRDER WS
en............._ ........_~IJJ~_...pm_~
ZC
~_............._IIllil!IIIIlilI.....Jl!'JZl!i!l"lIl_..........I:IIlllIIIIIlII~CZ
~-~-~;:)~-.........-~om
~-~-........."""t--~-~~~entJJ
~-~-~~ ............- .........~~
Abl.e
PierfA
z~VII(J)WI-~
_·-t of Span-·_--
Fig. Aq2 Cracks in Tie Plates, Lehigh Canal Bridges (PennDOT)
-42-
If I
(J).......- ........--~ l&J ......- .........-II\IIII!IIIIlII..
Zc(
,...,..,..,..".......................--IIlIIIlll..J,.,....-........-"""'--1Iit=ilI
a~~~__~z~ ::) ..................................__-t
oCD
~~~=--liIIi!!li1/Il- ..........- ....................""""""'"(/)au
"""--~---~ .........- ..........~.......
Pier"3;\
~---.-_....
enz~CI)
IeI)<CLtJ
I I I_. Cl of Span _ . GIRDER WN
--Fig. A.2 Cracks in Tie Plates, Lehigh Canal Bridges (PennDOT) • continued
-43-
TABLE Bl STRESS RANGE OCCURRENCESTIE PLATES
Gage CIN-l ClN-3 CIN-4 c45-5 c45-6
Sr N. Sr Ni
Sr N. Sr No Sr NoLevel 1 1 1 1
(ksi) (ksi) (ksi) (k.si) (ksi)
0 159* 139 15 1
1 21.6 9.6 18.6 9s6 801
0 57 30 8 5
2 19.2 B07 16.5 80 7 702
0 49 42 22 34
3 16.8 7.8 14.4 7,08 6.3
2 53 99 62 164
4 1404 609 1203 609 504
24 117 233 185 364
5 12.0 600 10.2 600 405
34 350 339 338 470
6 9.6 5Q1 8.1 5.1 306
59 719 476 395 764
7 702 4.2 6.0 402 2 07
218 1035 591 765 944
8 408 3.3 309 3.3 108
1092 1117 2392 1453 910
9 204 2.4 le8 2 0 4 009
1608 1239 3133 1195 371
Min 1.5 0.75 1e5 0045
Total N 3037 4895 7474 4438 4027
RecordedHr:Min 27:21 57 :34 17:44 57:34 28:35
* Number of occurrences when stress range 'equaled or exceededthe stre,ss range for level 1
lD44~
TABLE Bl STRESS RANGE OCCURRENCESTIE PLATES (continued)
Gage C5S... 7 C5S~8 C6N-9 C6N-IO C6S~11 C6S*=12
S No S No S No S Ni
S No S NoLevel
r 1. r l. r 1 r' r 1. r ].
(ksi) (ksi) (Itsi) (ksi) (ksi) (les1)
10 1 1 15 1 1
1 9.6 801 13.5 1203 1305 1305
30 0 4 5 1 10
2 8.7 702 12QO 1101 1200 1200
93 0 26 6 0 8
3 708 ' 603 1005 909 1005 1005
148 0 58 6 1 22
4 6.9 504 900 807 900 900
163 1 31 5 67 13
5 6.0 405 705 705 705 705
108 16 40 35 241 16
6 5.1 306 600 603 600 600
91 29 46 118 175 24
7 402 207 405 .501 405 405
105 28 200 378 183 146
8 3.3 108 300 309 300 300
232 220 575 1423 529 870
9 2.4 009 105 207 105 105
306 519 921 2154 478 865
Min IllS 0045 009 105 009 009
Total N 1286 814 1902 4145 1656 2075. --
RecordedHr:Min .57:34 27:21 12:54 57:34 12 ~54 12~54
=45c=
: \ - ~
REFERENCES
1. Bowers, D. G.LOADING HISTORY SPAN NO. 10, YELLOW MILL POND BRIDGE 1-95,BRIDGEPORT, CONNECTICUT, Research Project HPR-175-332,Connecticut Department of Transportation, May 1972.
2. Marchica, N. V., Yen, B. T. and Fisher, J. W.STRESS HISTORY STUDY OF THE ALLEGHENY BRIDGE (PENNSYLVANIATURNPIKE), Fritz Engineering Laboratory Report No. 386.3,Lehigh University, May 1974. .
3. Fisher, J. W., Frank, K. H., Hirt, M. A. and McNamee, B. M.EFFECT OF WELDMENTS ON THE FATIGUE STRENGTH OF STEEL BEAMS,NCRHP Report No. 102, Highway Research Board, National Academyof Sciences-National Research Council, Washington, D.C., 1970.
4. Fisher, J. W., Albrecht, P. A., Yen, Be T., Klingerman, D. J. andMcNamee, B. M.
FATIGUE STRENGTH OF STEEL BEAMS WITH TRANSVERSE STIFFENERS ANDATTACHMENTS, NCHRP Report No. 147, Highway Research Borad,National Academy of Sciences - National Research Council,Washington, D. C., 1974.
5. Galambos, C. F. and Armstrong, W. L.LOADING HISTORY OF HIGHWAY BRIDGES, U. S. Department ofTransportation, Bureau of Public Roads, January 1969.
6. Cudney, G. R.THE EFFECTS OF LOADING ON BRIDGE LIFE, Highway Research RecordNo. 253, Highway Research Board, Division of Engineering,National Research Council, National Academy of -Sciences National Academy of Engineering, Washington, D.C., 1972.
7. McKeel, W. T., Jr., Maddox, C. E., Jr., Kinnier, H. L. andGalambos, C. F.
LOADING HISTORY STUDY OF TWO HIGHWAY BRIDGES IN VIRGINIA,Highway Research Record No. 382, Highway Research Board,Division of Engineering, National Research Council, NationalAcademy of Sciences - National Academy of Engineering,Washington, D.C., 1972.
-47-
8. Armstrong, W. L.DYNAMIC TESTING OF CURVED BRIDGE - HUYCK STREAM, Journal ofthe Structural Division, Proceedings of ASeE, Vola 98, 8T9,September 1972. .
9. Moses, F. and Carson, R. ,PROBABILITY THEOR~ FOR HIGHWAY BRIDGE FATIGUE,. STRESSES, FinalReport, Ohio Department of Transportation, SMSMD No. 50, CaseWestern Reserve University, Cleveland, Ohio, July 1973.
10. American Association of St~te H~ghway Transportation OfficialsInterim Specifications Bridges, 19740
11. Fisher, J. W. an4 Struik, J.GUIDE TO DESIGN, .CRITERIA FOR BOL'rED AND RIVETED JOINTS,Wiley Interscience, 19740,
12. Swanson, S. R.RAMDOM. LOAD FATIGUE TESTING: A STATE OF THE ART SURVEY, ASTM,Materials Research and Standards, Vola 8, Noo _4, -April 1968.
13. Barsam, J. M.FATIGUE CRACK GROWTH UNDER VARIABLE AMPLITUDE LOADING INASTM A514-B STEEL, ASTM STP536, 1973~
14. Miner, l1. A.CUMULATIVE pAMAGE ,IN FATIGUE, .Journa1 of ,the Applied Mechanics,Vol. 12, No.1, September 1945.
'15. Schilling, CoG., K1ippstein, K. H., Barsom, J. M., and Blake, Go T.FATIGUE OF WE~DED STEEL BRIDGE MEMBERS UNDER VARIABLE AMPLITUDELOADING, Research Resqlts D~gest. No. 60~ Highway ResearchBoard, 1974.
